 And ISIS, if you could share your screen, that would be great. Could you show the agenda on this, ISIS? Please. Welcome, everybody. We are going to get started in just a few minutes. Going to let everybody come in. So we're going to get started in just a few minutes. All right, welcome, everybody, to this Opal RT boot camp. We are going to get started just about now. And it looks like since we are in a webinar format, we won't be able to capture all of your beautiful faces, but we wish we could. So throughout the webinar, we may be taking a few photos. Also, I wanted to let you know that this webinar will be recorded. Right, so on with the official welcome. My name is Brittany VanDorff. I am the public relations specialist for New Mexico EPSCORE, and I am helping host this meeting, along with ISIS Serna and Sarah Pashay. A few housekeeping things before we begin. You should know that this will be recorded, and you should have gotten a notification from Zoom. But in case you didn't, the webinar today will be two hours instead of one. Our instructors simply brought together so much astounding and awesome information and examples that we decided another hour was necessary. But if you can't stay for the second half, don't worry. It will be recorded, and we'll put this up on our YouTube channel and our website. I also wanted to let you know that you can ask questions at any time by typing them into the Q&A box, and my partner in crime, ISIS Serna, will shoot them to the relevant speaker, and we'll get those answered for you. You can also put them in the chat, but we prefer that they go into the Q&A box. All right, and next, we have, I just wanted to plug up a fantastic lineup that we have put together for 2021. In January, yeah, thank you, ISIS. In January, we're gonna have Carl Benedict and John Wheeler putting on a GitHub training. In February, today's presenters will possibly, February or later in the spring, be back to do a typhoon training. And then in March, April, and May, you will hear from new New Mexico Smart Grid Center faculty, Klaus Danielson, Hamed Nadimi. And Yu Ting Yang. And that will be shown on our website if you don't necessarily see that schedule up on the screen right now. Right, so enough about 2021. We've got a lot to cover today for today's webinar, so I would like to introduce Olga Lavrova from New Mexico State University, who will take it from here. Okay, thank you, Brittany, and welcome, everyone. It is great to see so many participants. I see that there are 28 participants attending and we are very glad to see you here. So we wanted to make sure that this OPAL-RT bootcamp is very productive to a wide variety of our EPSCOR researchers and our collaborators who are not necessarily in EPSCOR. I know that there are participants who are not part of EPSCOR who have signed up for this OPAL-RT training. So we wanted to make sure that we include three different examples of how OPAL-RT can fill in your research needs in terms of modeling some of the communication functions or some of the electrical hardware functions. So if we could possibly go back to our schedule in terms of showing the examples, the shows that we wanted to address communications and microgrid controls and integration of PV plus battery storage systems. And the structure of this class is such that we will go through an overview of what OPAL-RT hardware is and explain why it is needed. I hope that majority of you were able to complete your homework that we have assigned. I hope that homework was not extremely scary. So we hope that you were able to complete this homework and so that you already have some information about what OPAL-RT is and how we use it. So I hope that most of you are also familiar by now with the way to use Simulink with OPAL-RT. So I wanted to make sure you have some of the background so that we don't jump over anybody's head when we show our examples. So with this, I would like to introduce our research faculty who will be doing the examples in the tradition. First and foremost, Dr. Hamad Nademi, who is faculty at NMSU and Electrical and Computer Engineering Department. We are very grateful for his experience and being able to do those examples. Dr. Nataraj Pragala-Patti, who is also our host doc researcher at NMSU working on some of these projects. Also Dr. Sijo Augustin, who is also our host doc at NMSU working on similar projects. And we also have an undergraduate student assisting in this demonstration, Steven Lucero, who you might see on your screens shortly. And so with this, I'll hand this off to Dr. Nademi and I hope this will be all very productive. Thank you very much, Dr. Lavrova, for your kind introduction and also overview on this hopefully thoughtful webinar. So let me share my screen with you. So since my internet connection at home is not actually strong, so I turn off my video to boost internet connection. But for later on, I would turn on if it needed. Can you see my screen now? Yep. Yeah. Hello everyone and thank you very much to the all participants around the state, different locations for this seminar related to the hardware in the loop and also based on of course Opal RT Simulator, which is one of the very widely used real-time simulator and hardware platform to do the kind of real-time simulation and testing. Actually for many applications, maybe the examples for today is just cross-section between the computer science, machine learning techniques, and hardware development kits and applied to the all integrated into the power and energy systems, specifically microgrid. But these kind of platforms usually are really helpful and useful for doing the test and also for design validation for many applications from aerospace, power, also automation areas. So I know that there are participants from computer science and also from electrical engineering, some people from industry. So I try to make it in some way. It could be really useful, but hopefully my intention is that at the end of this actually time slot and also during the entire presentation for this webinar, you can't learn actually something because I'm working in power, my background is electrical engineering, but anyway, we need to learn from computer science other domains and vice versa. So this real-time simulator is very kind, so don't be worried about and this is not a monster hardware platform. But for students specifically, I made this slide to begin my presentation. This is very important for you to know that based on actually the recent development and also what is foreseen for the next future, next five, 10 years in many engineering and technology field is that this is something, this technology is really needed and for the students, they have to develop their skill sets. So I would say if you can be able to work model and work with this real-time hardware and simulator from the modeling to the end execution, then you gain a skill set which is in demand for the future. So this is not just being used, this technology is not being used obviously for educational institute or universities, but it has been a new norm actually for industries, utilities and many other agencies actually to use this kind of real-time simulation approach or techniques or technology during the design phases and processes. So this is obviously very good platform to be able to do the test and then later on as we move forward during my slide you can learn what would be the really key advantages by doing these real-time simulators and because the first step is to realize that why do we need this kind of real-time simulation and why we should not use the offline simulator for example. So the key part is that the real-time model should be developed to represent the fair and accurate method for the, for example, the system of interest for the issues we want to deal with or do some investigations, analysis actually. And then as I said that recently all of this OpalRT platform are being used for doing some analysis during the design phases for distributed energy resources for automation companies, utilities during the design phases and in the beginning of their model or solution validations. And as you see here at the left side of this picture, for example, the left side you will see the one just hardware bolt, right? This hardware could be the external device. The device could be controller for example and then the tall platform on the right side you will see that's the simulator, the hardware simulator. And then you will see what is inside that simulator is your model actually could be developed and run in a real-time manner on this real-time simulator why the external device under test or controller or could be motor or could be Raspberry Pi for computer science in video jets and machine learning algorithms executed on their computer science related hardware to talk to the simulator and then to be integrated into the like model of energy network, power network, right? And then as you can see there would be a lot of digital and analog input output could be replicated. The measurement devices, the actuators and so on. So as I said that this is the V-shape this diagram you will see in this city called V-design actually allows multiple of engineers from different domain to be involved in design modeling project to use that model to communicate the knowledge of the system under development the system we want to study in an efficient and organized manner. And as you will see the real-time simulation actually these days in industry sector covered the most part after online simulation is basically offline simulation when you just develop the key architecture the single line diagram of your system and then when it comes to the design prototyping integrating the different devices and from like a cyber physical modeling and system and then you can verify and validate your design solution modeling and then test it on the prototype at the very last stage. And the good thing is that you can also do the closed loop design changes in your study by changing some variables for examples controllers for example parameters and then you will see by changing those parameters you can adjust and then achieve the intended performance operation of your system or your algorithm for example. So the key point is that you can actually model the physical system or real system as close as possible resemblance to its physical counterpart. And you said that and as I just said you know these all power and energy systems or engineering domains are being interconnected to each other. So we are living in the multi-domain system. So we need some tools or some real-time platforms to be able to handle this kind of cyber physical modeling or analysis or model-based design solutions. So for all computer science and electrical engineering students or people working on we have heard a lot this smart grid concept or the world. So the smart grid itself means that we are integrating two infrastructures architectures. One is the electrical infrastructure the other one is actually the informing information systems communications infrastructures to be able to work together in a secure manner. And then you can see that the cyber security is the key for the whole platform. And then you will see for example from the distribution residential level commercial industrial to the transmission system a long way to the distribution plan. The distribution could be still the traditional power generation and also could be based on the renewable and distributed energy resources. Could be solar PV wind battery energy storage and also and so on. So the reason is that as you see to increase productivity reduce the CO2 emission efficiency will be increased empowering consumer and so on. For computer science the key point is that for to achieve such a smart grid system so we need a smart network usually based on the wireless communication system. For for example energy management to get the knowledge of all the loads transmission system and also the generation platforms whatever whether those the generation system is based on the renewables or traditional system. So we need to have a lot of information and all the subsystems should talk to each other. So then then you can this is just for power grid. The airplane or aircraft is another perfect example. The aircraft has mechanical aspect aerospace dynamics the electrical system electric motors right they are like a multi-domain the aircraft itself is like a multi-domain system and therefore we need the the real-time simulation and testing to be able to verify the both communication system protocols could be the the data analytics because we need to have a lot of collect a lot of information. Therefore to to develop such a system it's very costly actually to build the whole system we are dealing with. Then we model it using the existing or most advanced tools modeling tools modeling technologies or or methodologies and then we can be able to integrate all pieces in a real-time manner to get there to verify the the design the the concept or or to apply to the prototype and so on. So the purpose or the key contribution of real-time simulation is the the testing to accurately produce the simulation output as I said that in the same length of the time as it could be very close to the physical counterpart for example and then in this manner we can replace the the real hardware by real-time domain model. In this way we can just do some destructive tests we can actually identify the flaws and some design issues earlier during the design processes and then of course you can reduce the the development afford costs and then if you need to for example if you need to for example just integrate the controller you design for example the Raspberry Pi based on the machine learning algorithm into for example control the microgrid or power system as an example then that device could be device under test could be integrated as an external hardware or controller into the closed loop and then you can verify your design with the replicated or very close modeling for the system of interest could be run in real-time manner into the on the like opal artist simulator as we are focusing today. Let's dig into a little bit into the what does it mean when I or when you see a lot of discussion about the offline simulation and also the real-time simulation the offline simulation is that obviously when we deal the modeling the model could be even from very simple to very complex or complicated system could be the mathematical functions expressions equations in time domain for example and then those variable states are solved actually as a function of variable or a state at the end of the the time step but we should also so it seems that the simulation time step is that the time interval does all the state and the control variables all the system variables should be solved and find their solutions at the end of each time step when you do this offline simulation as you will see here for example if we start from time interval tn minus one the next step ahead is tn right and then the next one obviously is tn plus one focus on tn minus one and tn you see the the functions f tn is actually the function of all the variable states for the your system you are dealing with so the solution actually is identified like in the middle of time step or prior to the end of time instant tn so we say for the offline simulation this is faster than real time why in the second one at the bottom this is a slower than real time why because the solution for the the expressions of the system the function of all variables solution can be identified after tn right after the simulation time step so it seems that that your simulation is slower than the real time so the the conclusion is that for example the the accuracy and fidelity of the result would be under question for example if you are dealing with the fast dynamics in your system right then for example for measurement system you are missing some measurements some data for example in the the data analytics you have seen a lot have heard a lot related to the missing some information or bad data you know if you do the offline simulation definitely it could be a slower or faster and neither of them are quite precise representation of the dynamics of your system because the only way is to do the actually and also the computation time for computer science when you do the analysis the computation burden or time on the embedded hardware is another actually interesting point obviously if you do not do the real time testing or simulation it would be longer why the most bottom solution is exactly the synchronized that's what we call the the real time solution you see all the variables of the system the function would be identified and solve at the exactly end of each time step right and also the same for the next step ahead so in this manner the accuracy of the computational not only depends on the the precise dynamic representation of the system but also it depends on the lengths of the time used to produce that that that results if because for the like a PV system for renewable generation we are using a lot of these power electronics very high speed or fast switching electronic devices so we need simulation time a step even to be considered as one mic per second right not even millisecond and then that's the actually the strong contribution or powerfulness of the real time simulators so in this way we can accurately replicate as much as possible of course nothing is 100 percent perfect but at least based on based on the existing technology could simulate and get more inside and thoughtful related to the the physical counterpart of the if your system is airplane if your system is like a cyber security matters if it's cyber attacks if it's power and energy systems integration to the distribution systems and so on that's actually the the key part of the technology behind the real-time simulation requirements for example he means that hardware in the loop as I said that it could be for example external hardware can be integrated into the closed loop running into the the the virtual model runs into the opal arti platform for example hardware in a real-time manner just give you some example in the past for example we we had to for example use the power electronic systems is everywhere almost in every applications we have these power electronics systems could be run with like a one kilohertz pwm signal and then you will see the switching so the technology for power electronics demand for higher switching frequency to for example deal with some or reduce the power density or harmonic something like that and then we need to even achieve time a step below one microsecond then that's the technology of opal arti they have the cpu you can run the slow dynamics of the system on cpu in the range of like a millisecond or a hundred microsecond and then the fvga is even one or below one microsecond simulation time steps that's more powerful actually uh than the the at least the previous technologies we had so that's the uh used therefore the fvga the other a skill set for the or contribution of the opal arti simulation is that you don't need even to be very knowledgeable related to the vhdl programming or quite actually be expert on fvga analysis performance and so on if your model for example opal arti is compatible with matlab simulink when you develop your model on simulink and then when you run it and execute it it would be actually without having any writing of of hdl code it could be um compile or executed on powerful fvga uh both and then uh in in your very tiny uh like one microsecond simulation time step to to better actually simulate your uh system of interest so the the workflow is that you have a host computer and then you model your circuit with some existing tools like matlab simulink and then you you execute that on the real-time simulator opal arti there are multiple platforms but we just go through one of them today and then these opal arti actually features with the very advanced fvga processor boards and then of course you can also uh integrate external hardware whatever it could be controller or device like electric motors and so on or original battery energy storage or physical uh solar arrays into your system and then the second thing is that there are some communication protocols and then each system could communicate with each other for example and later i'll detail the some communication protocols and so on so again the key contribution of of real time simulator is that all the values should be sent to the external card or fvga uh should negotiate with the cpu actually and then the values from the cpu to the external card or fvga all should be done within one single time step should be like one microsecond for very fast or should be up to like hundred or fifty microsecond as well so that's the the key contribution and also in a very well organized manner so the the key point is that even for the this the technology suppliers for all real-time simulator is that the more values we exchange from the cpu for the external hardware and fvga fvga and then we need less time could be actually remain for for model computation that's the now still the old for example opal rt has made a tremendous amount of advancement but still they need to to do to reduce this this time to be able to actually provide some computational burden for the fvga to be able to actually reach to the accuracy level as close as possible so today actually we would have later on the second half of this training we use the the one one type of opal rt simulator which is called op 5600 the the most advanced one is op 5700 we have also in the lab you will see but basically the the key requirements are the same except as i said that the most recent generations are dealing with the very high sweet actually sampling frequency or sampling time this is this this platform you will see the physical picture can handle sampling time or frequency up to 100 megahertz but the the most recent 150 700 can handle up to 200 megahertz so that's the only thing but more or less in terms of the the i o monitoring the the analog or digital i o input output channels are the same and then the fbj is from vertex sixth generation and some some actually the cpu specifications as you will see and also you can connect the four pc i slides into this this real-time hardware and then when it comes to the as i said that the the host computer negotiate to the or communicate with the simulator using the the in a synchronized manner using the tcp ip link but the cpu and the fbj in the board inside this this platform also negotiate to each other in a synchronized manner using the the fiber optical links and then the this this opal rt simulator can support the ic 61850 communication protocols and the serial and also the canvas protocol we have very famous military industrialized protocol meal 1553 actually and also the aeronautical radio ink it called air ink standard that's kind of digital information transfer system standard technical standard is also support by this platform so we can can actually run multiple communication protocols for for the house pc your model to negotiate with the real-time simulator as well so that's another feature let's look a little bit into the what's inside the this this box and obviously you will see there is the the upper side and also the the lower side for this hardware platform if you cover up and then you will see the upper section is usually includes the the fbga all the fbj board for processing of the model and execution and then the digital analog i o modules in mesonians and also the lower section is basically the cpu ram similar to your personal computer hard drive pc i slots and then if we go to the very detailed overview or cross section of this platform you will see on the the upper section we have one fbj board to handle the i o board actually from the host pc to the simulator and then the second one is the most powerful actually xilinx vertex 6th generation the fbga board as you will see here to to and then the the technology of opal rt is also provide another powerful feature which is the they develop the e h s which is the electronic hardware solver you don't need to do any vhdl programming your model in the simulink can be converted and compiled to the to the opal to the fbga processing board and then afterwards can can negotiate to the external hardware and also back to your host species so this is the main fbga board as you will see some features like 32 digital i o lines when it comes to the analog i o lines is 16 channels but obviously you can also parallel to multiple fbga boards to be able to to to to if your system requires more i o connections and as i said that the sampling time for this specific 5600 is 100 mega 100 megahertz sampling time but the most recent can also handle 200 megahertz so the calculation inside the fbga can reach up to even from 250 nanosecond to 500 nanosecond and then you can see how how powerful and beautiful would be this this this platform for if you need the really accurate higher accuracy and high fidelity for your simulations for very critical and sensitive applications and the lower section as i said that includes the cpu so this is where the model can be executed without if your model doesn't need a very like one micro a second simulation time step still can be run in a real time on your cpu and then you will see some some ram or some power supply and so on placed at the bottom of this this platform and then the i o modules are organized in some mezzanine modules that that's the i o modules can communicate between the host pc your signal you sent from your laptop connected to the opal rt all the digital output analogs and then the another view from the the front side the back side of that hardware platform you will see the red and the green so the the red is actually connected to you will see this this pin architectures and then for for the for the analog side and then the green also from the front view of the platform can be reached or accessed at the front side of of the the platform for the i o boards the the analog in and analog out so these are the the analog signals channels and then we have also the digital i o as you will see the digital are actually in 32 channel 32 bit channels and again as you will see the backside or rear view are shown in the red color and then the front view of the hardware you can actually put the rj45 cable into this this platform and then connect to the for example the the port you can actually monitor the waveforms on the oscilloscope for example in the lab or you can send the signals to your connected external hardware and this is the front view as i said that the the monitoring and then you will see an each connector actually can drive four signals the the 16 i o boards and this is actually the the physical picture you will see the the front view of this opal rt hardware so this analog actually signals can actually be for the monitoring purposes as i said that you can see the waveform using this bnc cables on the oscilloscope in the lab and then the rear view the backside of of the platform are for the external hardware connections to the to the real-time simulator and then this is what do you send the signals and then the signals can be processed on the fvga boards and then the fvga board can convert using the analog to using the digital to analog converter to to monitor the the analog or real evolved application signals and then the you develop the model using the the rt lab rt lab is actually very close to the to the what we model in the the simulink and then when you install the rt lab you have rt lab i o facilities you have a lot of modules and then for example you can capture this this or drag and drop this this block for the opal rt communication i o blocks and then define what signals you want to send from your model to the external hardware so it's pretty easy and then when for example you click on one of these this i o block and then ask you to define the name for example this is some something you can go through the document and then you can design if it is for example the synchronization mode is can be arranged in two way for example if this is the master communication i o block or you can put it on a slave to to be actually identified by fbga and so on and then this is the very high level picture of input and output you have electrical signals usually the the analog signals are using the transducer converts to the electrical signals the analog digital and then this is the the entire real-time simulator platform and then we have an analog out and also digital out and then you can obtain the analog on the monitor or the digital could be sent to your external device and do some processing or or data analytics or controller monitoring things and then as i said that for example the front view you see you can extract the the analog signals by using this this rj-41 cables 45 cables and then connects these signals to the monitoring panel which is here and then each panel can actually cover the four bnc connectors as you will see here there are four bnc connectors you can capture that signal to the oscilloscope to monitor the the waveform you are designing and then you can also using the the computer do some changes and then see the the the actually online changes simultaneously on the oscilloscope and then for for data acquisition purposes as well you can use the the oscilloscope so this is my last slide but again similar to the first slide and i think all the especially electrical engineering and computer science students should work with these real-time simulators this is the real picture based on the what i did last year and the year before in new york and the new york estate energy research and development authority as part of new york government developed the new york advanced grid innovation lab for energy they called agile lab this agile lab is full of actually three different or three existing real-time simulation technologies opal rt i'm just put the opal rt rtds and so on and then they can actually do some monitoring control or performance operation of the system even without need of any people to be in that site so this is the something just to to actually draw your attention this is the technology or i would say a skill sets everyone should should hold and carry similar to the python programming for example because this platform are quite user-friendly there are built-in features you can do the python programming also using some of the other real-time simulation technologies available in the market currently and then for your professional career these are the skill sets in demand for the future so thank you very much i think this is my my last slide questions and just to remind everybody please put the questions in the q&a box or in the chat and um sara pashay will shoot them over to the presenters hello everyone currently there are no questions to be asked or to be answered there's many questions to be asked yeah for for for computer science or people working on the communication or information systems maybe as a additional information for example when you connect the i don't know external hardware could be controller and then sometimes you need the fast rate with very small latency right because i know that the latency for the people working in this area is is one of the the challenges or interests so this opal rt actually of course it depends on the number of optical fiber this platform is used and the optical fiber optic links and also the complexity of the system then the latency i think this platform can handle is somehow from one to five microseconds based on the usually it's more four to five microsecond because the one microsecond is not maybe so realistic for the medium to very advanced system in terms of the complexity of your model but still is in the range of a couple of microseconds which could be really helpful for people doing the the cyber security for example analysis if you inject some signals to emulate that for example the attack and then you can see how your controller or solution system can respond to do is kind of the applications or or investing or assessments actually that's the one thing i just wanted to to mention okay so thank you very much dr. Nademi i think we are ready to move to the example where we will show how to integrate and how to communicate with raspberry pi with opal rt so if you're ready please go ahead could you please give me access control um you should you should have it yes hello guys i will be showing you guys how to interface with the raspberry pi and rt lab with using opal so uh with the opal if you communicate with several uh computers and with this example we're going to be using a raspberry pi a raspberry pi in many cases could be used as a gateway to talk to the internet talk to other computers so it's a great great platform to get sensor readings let's say from an arduino setup to the network and to a opal that is simulating or is reading in data so um how to set up the uh raspberry pi is to get a micro sd card either pre-downloaded with the debian os or download the os with the image writer that raspberry pi provides you then um you download python onto the uh raspberry pi then you jump down to um opening up the raspberry pi and from there you enable the uh ssh by going into the raspberry pi configure windows then from there uh you can go into moba exterm basically it is a terminal to communicate with other computers or uh run computers headless by using my piece so with this uh picture here you get kind of like an overview of getting the ip address from the raspberry pi and naming it specifically so it can pop up later on then from there um you open up a uh terminal or session and basically type in your log information if you have it changed from the default to your own um so we step down to the checking of the python install and if you type in python 3 you will get this output then from uh from there you get to go into the uh um the download from uh overall team knowledge base then from there um you have two python scripts a main and a server python script and you drag and drop into the moto x term and this will basically run the python scripts remotely from your host computer to your slave computer and from there you have to change out some of the information uh to the host and target ip addresses and then get the ports correct in there so this right here this picture shows what you need to have done on the host side then um this is where we got hung up because of the firewalls that mcu has up we could not go into this window here on the opal side and uh basically from here on out uh we could not get any type of configuration any type of acknowledgement or anything of that sort because of the uh of the firewall but if it did have you know successful communication it will have all these numbers all these commands and matrices coming up and you can change the numbers and values to what you need on the rt lab side and from there basically you'll have acknowledgement sent out along with your commands if things were successful and that's and uh that's uh that's about it so this is um this is all there again i'm just going to jump in and say that this uh example um was demonstrating how if you are a computer scientist or if you're a network um professional you could use opal rt to integrate a simulation of any of your network code or any of your other codes that you might be writing in raspberry pi with opal rt so as dr nademi explained at the beginning there are multiple applications of opal rt where you could demonstrate a real hardware application or you could just integrate with a software code that may be running on various platforms and so raspberry pi could be one of those platforms and since if you know how to connect to raspberry pi then basically the world is your oyster then you can connect to any other networked piece of equipment so i will stop here and i understand we wanted to take a break and also possibly take a picture with everybody who is attending the webinar so i will hand this back to bradney and see if there is a picture or if there are questions um so because of the setup of this webinar it doesn't look like we can actually allow them to turn on their cameras unfortunately but we can definitely answer questions so if anyone has questions there will be happy to shoot those over to any of our presenters sorry about the the lack of photo if you don't have any questions i think i think it's it's fair game to maybe take a quick break and reconvene at one as shown on the schedule if you just need like a quick break but do come back and if you can't join us for the second half uh no worries we'll be recording this and it will be up on the new mexico website in january at some point and as the schedule shows there are more real-time integration examples so please stay on we'll see you in five all right everybody it is exactly one o'clock um i'm going to uh we can get started on our next example by uh dr nadimi i did want to remind everybody that if you have a question please type it into the question q and a box or the chat and um we will shoot it over the over to the center to have them answer uh we want this to be useful for you so please do ask question questions if you have them um dr nadimi are you you ready to go yes of course outstanding i'll stop sharing my screen and you can take over yes let me share my screen hi dr nadimi um you just let me know i will uh i will exclude your files yes yes hello everyone yeah as you will see the third session would be to show you the the simple demo and later it would be a little larger demonstration based on the the this macro grid model but this the the system i'm talking about now is just a pv system and then integration with the battery energy storage as one of the the actually lithium ion battery as one of the technologies related to the energy storage devices and then how we can develop the model and then later on dr nadimi my colleague will execute the same model i'll show in the slides and then show you the the model inside the blocks and then execute simultaneously on and afterwards on the opal artist simulator i just described at the beginning of this this webinar so yeah to see what actually this is the different view or look into the okay so that platform is quite powerful that the hardware is is actually a strong to execute the the very fast dynamics or very tight requirements in terms of the operational requirement to be very close to the the real world application but there are some challenges as i said that this is something some skill sets you have to gain in demand and it is quite a normal or new norm for designing processes even in industry so but before that at the when it comes to the modeling and the tool is still there is a huge actually huge demand for developing a still the very advanced models and also what kind of modeling approach or technique we need to at least adopt or or adjust the the existing models to be able to run it on in real-time manner because the at least in the power industry or in the control design solution it was not a so long time that they use the real-time simulations right or this this technology deployed in the past for for verification or validations on the prototype we used to build for example in power electronics we used to build the prototype or real test bed and then do the whatever we want to to design or evaluate or assess so but we know that the renewable energy resources the energy storage technologies in general the energy domain the telecommunication the IT the also power electronics are like fast-moving technologies so it seems and for all the systems are now considered to be integrated and then these impose a lot of challenges first we need to develop the model to be able to actually real-time model to be able to run on the the available real-time simulators we have on the market and then this integration of several systems into the for example the entire power grid or power networks is a actually impose a lot of challenges for integration from integration standpoints and then for example the battery energy storage of course the battery lithium ion batteries or other types of batteries like a lead acid or nickel cadium and so on they have been in around for years and then we have quite good models but even the the platforms like a or simulation or tools like mass works or MATLAB simulink they have some models but they cannot actually do in real-time to to achieve that level of actually accuracy or fidelity we need to do the really outperforming designs and then as I said that also we should include some communication aspects because all of these distributed energy resources the battery itself has its some power electronics converters controller the solar inverter also had features its own controller wind diesel generator whatever we have the the traditional grids the the transmission systems and the data we need to exchange from the bottom to the up and then onward needs to consider some some communication system to full field so those kind of some aspects actually this is still open challenges but some aspects of those time instances or latency should be included in the the simulation the the simulation process or a strategy we are going to to develop at the scratch or in the beginning of our design and then when we for example look into the the battery technologies or the model we have for example as I said that the ongoing trends actually call for for accurate model and then the fast modeling which is quite quite significant for the like lithium ion batteries which is currently one of the widely used or type of batteries we usually use for microgrid or new energy power systems and then how we do the control monitoring for charging discharging of the the batteries could be also this battery actually deployed in electric vehicle because batteries inside obviously so that's the the the challenges so we we need an accurate model the real-time model also for batteries for PV systems for example for PV modules for you know that the PV cell is just a very small piece with the very low capability in terms of power and then can be connected in series or parallel connections to create the PV modules with the intended power level capacity for example 10 kilowatt hour 20 kilowatt hour and so on so the how we can actually how how do we model that that systems with the very accurate PV cell model or PV module model and then integrate with the the battery for example this is just the two exemplary application but this is basically the same thing and also concerning to actually do some appropriate control and also the communication protocols into the the modeling design because we should not deteriorate for example the the communication requirements you see all are interconnected and then when it comes to the or testing into the real-time capability then we need the real-time model as I said that the the model for for example for lithium ion battery MATLAB actually provided is not very powerful when it comes to the real-time simulation in terms of for example the accuracy in terms of the computational time burden if you need higher accuracy in your system and on the some exemplary embedded platform then you need to be careful if that computational burden for that embedded platform can be able to handle the time consuming or computation time of the the battery model for example or the PV model and so and also the hardware resource usage so I'm going to show you the the PV system modeling and then the benchmark with the existing model mass work had like in the last two recent version of MATLAB simulating in 2018-19 B revisions and then compare each other in real-time simulation execution on on the opal arti platform we just discussed and then compare them in terms of the computational burden in terms of the hardware resource consumption to to be able to run very simple circuit to to also fulfill the time step for example for this system is is 20 microsecond for example is not that very fast but still and then do some comparative analysis at the end for example so the mass work model for example for battery is is really powerful but for offline simulation and then therefore we use that one for benchmarking against the the model we develop actually is just very simple based on the simple mass equations and expressions and then we integrate the PV panel with the PV module let's say with its converter and then integrated with or pair with the battery energy storage and then we run it on the opal arti the other thing is that for example the challenge for industry is that if they use the opal arti but they shouldn't be limited themselves or the technology should not be limited to the only the opal arti technology we have for example the typhoon we have currently the the DS space real-time simulator and so on these are also being used in industry as well so the real-time model we should develop should be able to actually meet what is called a cross-platform real-time simulation the model should not be only compatible with opal what happened to the other two or three real-time simulators right this is a very significant limitation if we make the model for example to be compatible only on on opal arti simulators right and also the same thing these real-time simulators maybe in next two, three years they were limited for example had some limitations you cannot built in like python-based programming into your design but now since last two, three or four years they just started to actually add those kind of built-in features to their their technology or to their hardware simulators to be able to do this kind of real-time analysis that's the the purpose of this very actually simple simulation and later on a demo to get some feeling for example what you can do if you are doing the the computer science if you are also working on the power or energy area as an electrical engineering you can achieve at the end and the other thing is that for example in the past mass work for example provides the again accurate model detail but it would be like a holistic or general model for four types of the batteries not only lithium ion nickel cadium or lead acid and so on but usually okay, Nademi if I may jump in here we are down to 18 minutes left in you in your demonstrations I think it is better to jump to the file yeah I'm actually moving forward so maybe I need like 10 minutes and then we can leave some minutes for question and also five minutes for Dr. Nataraj to show the demo yeah, I think maybe we can show the demo because I hope there will be questions in the actual demo okay, so let me just explain a little bit and then like two minutes and then we move to the demo because I have also so I don't want to dig into the how we model the like the simple PV cell and so on but usually for the PV systems we have a state of the charge we need to monitor and control the state of the charge based on the operating point of the battery and then the to keep the output voltage of the battery actually within the acceptable tolerance or limits and then you can develop the for example the those kind of model and the control system for the battery to get the monitor the state of the charge and then you can also integrate with some MATLAB function or script into that one this could be also done in conjunction with the the Python programming into the MATLAB and then execute on the real-time simulator then obviously there would be some algorithms for that to keep the the voltage and the current of the like a PV panels corresponding to the manufacturer specifications and data sheets and then when you develop the real-time model usually if you have two sets of blocks or modules that's the simulated model as you will see on the left and then the blue one on the right side is actually the user console and then if we look into the model then for example you can develop the model as you will see here the solar panel is shown in the light blue and then we have the rest can be connected to the the battery model from MATLAB Simulink and then you need this op communication as I showed in the beginning of this webinar that's the I.O. module you can just drag and from the RTLAB software and then add what is the input and then the output can communicate with the hardware side and then this is the detail of the the PV cell model for like a set of the charge monitoring of the the battery as well and the PV and then again we can also look into the we can build ourselves whatever we need to monitor for the demo and then what we need to capture and then to actually acquire the data measurements on the like you can save it on your laptop or can be saved as a CSV or a spreadsheet format this is the inside of the for example the console or the user model when you run it you can actually extract the signals you define and then showed on the oscilloscope or as I said that in the beginning you can also capture the analog signals and then monitor and they can be visible on oscilloscopes and then if you need to set some set point as well you can provide those set points for your system and then we just as I said that benchmark the analysis and then compare with the mass fork or MATLAB simulink model for the PV cell and also the against the propose and also the existing one which was proposed in the some literatures and then for the battery also you can see the state of the charge for example when the battery is discharged what would be the benefit if we use the real time model against the mass fork model which is more friendly with the offline simulation and so on so at the end as I said that compare the mass fork and then the proposed PV model you will see the memory usage the other thing is that the computational time burden and then you will see the existing models we have on the available tools also are a little bit time consuming when it comes to implementation on embedded platform so based on what I just explained to you the models I think now Dr. Nathraj can show the the real-time execution of this model on the Opal RT simulator yes let me stop sharing can you see my screen yes yep yes here this is the folder Dr. Rohitana I just mentioned that this is the actually the first page view of the RT lab so when you build the model on the simulink then you should incorporate it using the RT lab and then you need this RT lab to be able to communicate with the Opal RT platform yes yes this is the model file in that model it contains PV module and battery I will show you that file and yeah this is the real-time simulation file here in this in main block it contains the real-time system here here we can see left side we have PV panel including algebra non-linear equations between the PV voltage and PV current and then this one we can format as like a current source and then we have the boost converter and then we have battery model okay now yeah this is the system now I will execute this file on Opal RT okay for any for any software it is the common processor first we need to compile the file and then load and then execute here I'm just compiling the file you can negotiate with or talk or communicate with real hardware Opal RT hardware yes okay now it is this file is compiling to check whether are there any errors in the file or not and also it will generate these some files to execute on the core of the Opal RT so here we can see for compilation it took around one minute 29 seconds okay now we can see there are no errors in the file and we can load the generator file on the Opal RT load yes it is loaded and we can execute the file okay now we can see the results here in console block and now now the system is running on the Opal RT that is in real time okay now here we can see the PV side voltage and current waveforms and these are the waveforms corresponding battery side okay this is the state of charge second one battery current third one is the battery voltage yes let's just tip the model just I would like to also add for example as you see you can use the same ascope you have in the simulink but the thing is that since it's running in the real time and then if if not not touch please show the main circuit the general overview the circuit and then you will see for this one we have actually defined the 20 microsecond simulation time step since that's the scope since it's running in the real time and because of lack of capacity then you will see some maybe it's some strange waveform on the oscilloscope but the better view is for the monitoring of the waveform is that when you actually captured it on the oscilloscope and then you can see very high resolution you can go down to the very high resolution and details of the waveform to see for example the the tiny sections of the the waveform or you can also capture the result or data measurements for example for computer science we can for example apply the fault for example on the output of the pv panel or pv system and then the data can be acquired and also extracted and then for the computer science usually the times to the very advanced machine learning they can process tremendous amount or huge amount of data sets and measurements then you can develop for example such algorithms to actually predict the failure and then before failure happens then we can for example disconnect the affected system for example the solar system right to avoid damaging the rest of the the physical damage or the rest of the system or avoid the physical deterioration of the pv panels for example so these these kind of activities can be done and you can actually exchange the data you can store lots of measurements based on the different operating scenarios for the people who are working on the developing the machine learning or data analytics algorithms and then this you see the the the block opcom is exactly the the opal rt actually can communicate and then capture the the analog data and then can be shown or send the the signals the digital signals to the to the analog ones and then the analog IO can can be used for monitoring or or capturing the the measurements so can I move to my presentation is there any question we just had one come in it's from Chris Scott if you write a MATLAB script to control your simulink model is it possible to include the script in the rt lab compilation yes actually this is the the very recent changes opal rt has made I think since the last year I believe it would be possible because they have realized that the python script or programming is one of the key and then again they have made that changes and currently yes it's possible for you to incorporate your python script and programming whether that's the control or monitoring purposes you can also write python for capturing the measurement data as well wonderful thank you there are no further questions at this time yes go ahead please start with the presentation yes hi all my name is Natras and now I will show the real-time simulation of AC and DC micro grids okay now I will show the first system description and then I will explain some control strategies for inverter and DC to DC converter for the DC and AC micro grid application and then I will show the two real-time simulation examples those are developed by me only and first simulation file is the related to AC micro grid and second file is related to the DC micro grid okay now first first real-time simulation system is somewhat like this here generally micro grid as you know it is a collection of distributed energy resources and loads and that can operate it's like it's a part of main grid and here if you see the this is the IEEE 13 bus system and let's say here I have incorporated like solar PV on 611 652 and 611 on it's like a single phase solar PV source and 652 walls single phase solar PV source and at 680 it is like a 100 kva inverter and at 675 100 kva 3 phase inverter now here if you see the let's say if you incorporate this circuit breaker between 632 and 671 let's say if there is any fault on the upstream side that is on 650 bus the circuit breaker we can open and remaining lower part we can operate like a system like that is in micro grid that means that is in island mode operation let's say if you close the circuit breaker now whatever energy is there in the micro grid we can put it into the main grid that is nothing but grid connected mode okay now for implementing this system in real time on OPALRT what we need like first we need like line models here we can see there are like line models nothing but here there is a overhead lines and underground lines between the here buses here I have used three phase pi section model for line models and here and also we need some load models those are nothing but constant impedance type loads constant pq loads and constant current source type loads and coming to the renewable energy source based interfacing inverter that is nothing but here we can call like a dzu you the sugar generation unit too okay now here in this inverter control we we have to implement like single phase version like control strategies and three phase version control strategies I will show you the control strategies later and here you can see here I am not going into deep why because there is a time constant presentation here that's why I will explain overview of the control strategies here we can see let's say single phase solar pv here this is the dzu unit that is a single phase here dzu side either solar pv or battery source or wind energy anything here here for this let's say if you want to integrate this inverter with the utility main grid okay we we need some control strategies like power active power and reactive power control strategies and also we need some phase lock loop technique so PLL PLL is a technique we will get the information of the utility grid that is the ac voltage template that is the magnitude and frequency and the angle okay now here if you see the control strategy here let's say this is the active power reference equation that is the off of vd times id let's say vd is the condition why because let's say if there is a utility grid grid side voltage if you assume the constant vd is constant let's say if we want to feed the some power into the grid let's say 10 kilowatt reference based on that we will get the id reference commands and and here there are some internal current control loop for that internal current control we need i reference that i reference will get from the based on the inverse or transformation technique okay now here here i have i have used the like stationary reference frame based control strategy to implement this control strategy here okay this is the control strategy regarding single phase version let's say if it is like a three phase version let's say let's say if i'm using like a three phase inverter here also we need to control the active power and reactive power and also we need utility grid voltage information that is by using the PLL technique here here i have assumed like utility grid side voltage is unbalanced so if voltage is unbalanced we cannot use conventional techniques to get the information now for utility grid AC voltage so we need some advanced technique PLL type here i have used double synchronous reference frame based PLL technique to get the information of positive sequence voltage template okay now here here this is the active power reference command and this is the reactive power reference command okay these are the equations for unbalanced voltage conditions here vd positive id positive is nothing but a positive sequence component in the synchronous rotating reference frame okay now here if you see here by using the inverter we there is a one one restriction like we need to pump the power it's like a balanced power that means positive sequence power only we have to inject into the grid so further further here if you assume the negative sequence component to zero we will get the id positive reference iq positive reference here id positive reference iq positive reference and also id negative and iq negative reference so once we get the reference commands id reference and iq reference so this is the like a positive sequence current control this is like an inner current control loop and we will get the some modulation mix and then we can generate the pulse it get pulses by using the sinusoidal pulse with modulation technique and this is the way we can feed the we can feed the power into the grid under the unbalanced voltage condition okay this is the control strategy here i have used double synchronous reference frame based current control strategy okay and then next yeah before going to the show before going to show the real-time simulations of ac microgrid and dc microgrid i would like to show the some research of the tools like where we are working on in the lab here here if you see the here here opal rt5700 this is the opal rt real-time simulator and here this is the host pc okay now now let's say here our main objective is we need to implement this iterably certain bus system on the opal rt okay let's say by using the same power system model blocks and rt lab model blocks we can implement this entire power system model on the opal rt and then we can run in the real-time manner okay that is a it's like a software in loop actually that is a thing let's say let's say if this system want to coordinate with the real physical hardware systems like let's say some part of the system is on the opal rt and so like a like a dc inverter is on the external side that is the real time it's like a real physical hardware system okay here right side we can see here we have in the lab n phase micro inverter that is like a single phase version actually but here we are working on one research object is like a hybrid solar pv and energy system and this research work working like dr sizo and myself we are working on this research object too and this is the thing hybrid solar pv and battery by using the commercial micro inverter and then and then we have implemented grid forming inverter physically in the lab here this is the like a three phase inverter and okay this work we have done like hardware and myself we have done this one grid forming inverter implementation and we have in the lab here lab world test bench that means here this test bench having like a synchronous generator and inverter and loads and then we can we can implement residential residential distribution system that is like a secondary distribution system we can implement in the lab here and also we have the relays to do some analysis on the fault production production analysis okay why i why i'm focusing here research object to some let's say all these physical hardware systems we can integrate through the power amplifier to the system on the polarity so so in this way we can we can run the some part of the real time some part of the system in real time on the polarity and some part of the system on the external side okay this method is nothing but this method we are calling like a power hardware and the loop power hardware and the loop and and one more we have done some work on the rapid control prototyping so rapid control prototyping means here this let's say this is the inverter grid forming inverter this inverter is externally outside but getting the control strategy of this inverter is executing on the polarity okay in this way this method we can call this like a rapid control prototyping okay and one more i'm working on research objectives like cyber communication here here here i have implemented some xb modules outside that are interfacing with the opal rt through analog ios and digital ios and here what i've been here let's say let's say system is on the real time but control control port is on the external side so this method we are calling like a control hardware in the loop here here i'm concentrating like cyber cyber cyber attacks so and cyber models cyber attack models and also here i'm just implementing some resilient control strategies that should be work any type of communication type topology okay this communication topology is outside thus that communication topology is outside but we can use raspberry pi or otherwise xb communication device so this communication system we can implement outside that communication system we can integrate with the real time simulator the system and in that way in that way we can we can do the some resilient control strategies and the cyber security some research work we can do and also here this is physically this is the we're implementing like a communication layer and also we can use some open ed software or ns3 simulator we can we can form it like a course simulation it's like a course simulation that means here we can use one software like we can implement communication topology by using some communication type software and that software we can integrate with the open ed and in that way we can we can analyze some latency issues in the communication topology would there any effect on the control strategy of the of the inverters these inverters okay now okay now i will show you the real time simulation real time simulation on the in software loop method only okay here i will show the first ac microgrid system okay this microgrid is connected in like grid type grid mode why because due to the time constraint i'm just showing the let's say if the microgrid is connected in the grid connected mode how it is operating i will show that thing and then next real time simulation i will show the dc microgrid okay dc microgrid with the with the droop control technique and and and secondary control strategies okay now now i will ship to the real time software okay now here first i will show you the ac microgrid that is the idruby13 pi bsi here yes let's say if we if you want to run the file on the opal rt okay we need to follow some steps so for executing the file on the opal rt we have to we have to create blocks on the top hierarchy level like this is the system that is a like a main master and second one is the leg console let's say if you have if you want to let's say let's see in this manner in this manner that means this main system is executing on the one core okay one core of the opal rt here we have eight cores okay similar type of these systems we can execute on the eight cores okay here core one let's say if you want to execute second if you want to execute some files on the core two okay and next sub block we had to create slave mode yes this underscore any file name okay now here i used only one core that means that name of the block is main sm underscore any file name okay now here we here you can see the idruby13 bus system and with some distributed energy resources here we can see idruby13 bus system having like we need some live models okay these are these things I have implemented by using the three phase uh five section model here and also we need some like ppu type loads here okay all these like these load blocks are available in the some sim power system library and these things live models are developed on my wall actually by using the three phase five section concept and here first we will see the single phase inverter okay and here you can see the single phase uh this is the like a single phase h bridge how bridge h bridge and followed by the lc filter to adornate the high frequency components and then followed by the one transformer to integrate the low voltage distribution filter that is the 4.16 kb and here for control strategy for this inverter here okay let's say if you want to operate the inverter in the grid connected mode we need the information of the grid voltage so for that here I have used the SOGI base PLL technique SOGI means second order generalized integrator method okay okay this is the model for that PLL technique here we will get the information like frequency and maximum voltage that means peak voltage okay once we have that information we can implement the control strategy for the single phase inverter okay here here I'm just giving some reference command from the console okay once once we have the active power reference command we can generate the id reference command okay this is the id reference command and then inner current control loop here and then after that this modulation brakes we can here comparing with the high frequency carrier waveform this is the like a PWM technique sinusoidal pulse with model technique to generate the pulses to the inverter switches okay now this is the single phase this is the single phase inverter is operating in the like a grid connected mode only okay and now next I will show you the three phase inverter control strategy okay and our unbalanced world is conditioned okay why because here if you see i313 bus system having the like loads pq type constant impedance constant current source but all these are like a asymmetric load configuration so and also asymmetric line impedance due to that here at the bus voltage you will see the unbalanced voltages so let's say if you want to integrate the three phase inverter with the with the distribution feeder so we need some some resilient control strategy to operate the inverter in the active power and reactive power reference control mode okay now here you can see here three phase inverter three phase inverter and then followed by the filter to alternate the frequency components and then here here DC side voltage is 700 volts so to generate the 480 volts at the AC side here but here 4.16 volts 4.16 kilo volts on the distribution feeder so we need one transformer the transformer is in the delta two star ground configuration okay and here we will see the inverter control so here inverter control here first we will see the PLL okay this is the three phase PLL but this is not conventional control strategy this is the like a decoupled synchronous reference frame base PLL so first we need to decompose the actual voltage to the positive sequence and negative sequence once we get the positive sequence components we can use the conventional PLL technique so here we are getting the positive sequence voltage and then we can decompose into the DQ and once once we once if you regret the quadrature voltage to zero we will get the positive sequence voltage frequency frequency and magnitude here okay once we're done with the three phase PLL and then we we have to implement the inverter control here we can see inverter control is here we need to regulate the active power and reactive power and for that this is the like a control strategy current control strategy and here first we need to sense the inverter output currents and then we need to decompose that currents into the positive sequence and negative sequence and this is the positive sequence positive sequence current current control strategy here lower side is the negative sequence current control strategy okay here here objective is we are attenuating this sequence components to zero we are injecting only positive sequence components into the distribution system okay and okay that is the inverter control okay now now we have okay now we are done with the simulation file with the like actual distribution system model that is the attribute that inverse system and also single phase uh uh distribute generation units and also three phase distribution generation units okay now we will see the um how to implement this file on the real time okay for that first we need to um first we need to compile the file and then load and then execute same procedure okay here i'm just uh compiling the file and also it will generate the files to be loaded on to the polarity okay now it is compiling the file are there any errors in the file or not it is checking but because it is the large file that's why it is taking more time to compile this here we can see um it took around two minutes 49 seconds for the compilation okay once it is done uh next load this file okay loading done and next execute this file on the polarity now it is executing okay now this file is running on the polarity here we can see here and uh and one more thing here uh i have given the simulation step size is 20 microsecond okay here we can see the uh bus voltages as we can see here bus voltage is here and and here we can see the per unit fashion in the bus voltage is here here we can see per unit voltage let's say here at 632 bus there is a phase here voltage is around 1.04 per unit phase b is around 1.018 something 1.017 on the case here okay and then we can see the yes back to power interact to power here you can see here now here uh just see here i have given the 100 kilowatt that is a reference command i have given here and here you can see uh at 680 bus there is a three phase inverter okay the three phase dg unit here we can see uh it is around 100 kilowatt here and here react to power reference command i have given the zero so corresponding here approximate it is zero okay similarly here we can observe at 675 bus there is also one more three phase dg unit okay for this also i have given the 100 kilowatt okay and here we can change and we will see whether that is regulating properly or not okay now here here i'm changing the value at 680 bus that is 250 kilowatt okay okay here we can observe here at 680 there is a drop in the active power 100 kilowatt to the 50 kilowatt okay there is a reference but but and also you can observe here at 632 that is like a source side here at the renewable side it is decreasing the power means at the source side it will increase here that minor change we can observe here add to power this is the like a add to power in phase a phase b phase c actually here this is the individual phases here power and and also i'm just i'm just changing the reference value at 675 let's say suddenly drop to some 20 kilowatt suddenly there is a cloudy situation on the solar side let's see here we can observe at 675 bus there is a certain change 100 kilowatt to the 20 kilowatt that means that means here three phase inverter control strategies and single phase inverter control strategies are working with respect to the with respect to the scenarios on the renewable side and also load side also okay and now this is the regarding that microgrid simulation and this this this thing we can call it as like a active distribution system why because there are some active components on the system we can call it as like a active distribution system okay i'm just stopping the simulation here okay now i will show i will show you the dc microgrid um and we're actually coming up on time we've got about a minute left yes i will show i think it's at least i will expand that file okay now uh this is the dc microgrid and let's say one minute i will okay here we can see this is the this is the dc microgrid system this uh this is the system i have implemented on the opal rt by using the cm power system block models and rt lab models here you can see here dc units dc dc converters based okay all these are connected in parallel and corresponding control strategies like primary control is nothing but it's like a group controlled voltage source uh by using the primary control we can operate this converter in the group control voltage source okay by using the secondary control um we can uh correct the uh voltage deviations across the dc bus okay this is the secondary control and here i have used this is the primary control here we can see inner current and outer voltage loop technique and uh with the group control and this is the primary control this is nothing but a group controlled voltage source and secondary control is nothing but here i have used average voltage and average current control technique and also here i have used some consensus based algorithm technique for regulating the for regulating the dc bus voltage and also improve the power sharing across all the you know all the converters okay this is the secondary control now i'll i'll just move to the simulation file here we can see here i have implemented like four converters here here all four converters inputs having battery source and the main dc dc network that is with dc dc converters i have implemented by using the efpgs same that is uh ehs solver here we can see that file yeah this is the file okay four converters connected in parallel through the cable impedances okay okay this is the um efpgs file okay and i'm sorry to interrupt um but just so we respect our attendees time because um we're two minutes over i'm going to say that people can definitely head out if you need to um sorry for for going over time and um dr well net netra i can't say her name i'm so sorry but if do you think you could maybe wrap up in the next three minutes for this yes it's definitely awesome okay now yeah now this is the system here and now here we'll see the control for each dc unit here we'll see here in our current control loop here outer voltage control loop here and also here this is the um this is the group technique and also here this is the primary control technique okay this is the primary control uh to operate the converter in the group control voltage source okay and and this control this secondary control is for the to improve the dc bus voltage and power sharing all the inverters across all the inverters here we can see the this is the average control technique here i have used uh different average technique that is the conventional averaging and also i have used dynamic consensus algorithm this is the current and also i have um i have modified something like dynamic consensus average technique here for improving the voltage and power sharing okay this is the current side and here voltage side also similarly i have implemented same conventional averaging and consensus based algorithm and also modified dynamic consensus algorithm here okay okay this is the cyber point of view actually and here yeah this is the main system now i will exclude this file we will see the some results okay now i'm just to compiling the file it may not take more time i think oh and it's so close um but i am going to have to call it unfortunately um because we are at 205 after thank you so much for your time i wanted to thank all of the presenters um and i also want to remind everybody that we're going to have another webinar in january um and it will be looking at github um so go ahead and uh log off for now we want to wish everybody a wonderful winter break and again thank all of the presentation the presenters for their work on this today thank you so much