GPS Position Hold

UAV holding position and altitude using GPS and Barometric sensor

Category:

Science & Technology

Tags:

• GPS
• RC
• Autopilot
• UAV
• Micro
• Helicopter
• Quadrocopter
• Quadrotor
• VTOL
• Radio
• Control
• Flying
• XUFO
• Mikrokopter
• UAVP
• NG
• Draganflyer

[VIPER4054]

I'm astounded! well done! - I have a few questions... whats the max altitude? - what is the most economical altitude that will give the best loiter time? - can inputs be controlled via a laptop? - what is the lift capacity ( i'm thinking of 2 small cams - 1 for visual flight, the other for vertical positioning ) if all lighting is removed... increased flight time would be what? ( considering weight reduction / power consumption ) - would you build me 1? at what cost? - and yea, I'm serious.

[infernoed]

@VIPER4054 Thanks. Flown to 120m (UK legal limit), can go much higher - limit is configurable in software. Loiter time is mainly down to g/W efficiency - typical 11g/W with this set-up. 43minutes has been achieved with a new version. Climbing to altitude doesn't consume too much power in comparison.

Flight with laptop possible - drag & drop waypoints.

More details of later prototype on RCGroups under "airbotix"

Commercialisation of controller likely

[monkeyking114]

hey! nice! what electronic did you use?

[infernoed]

@monkeyking114 Thanks, own design hardware and software - includes 3x gyros, accelerometer, barometer, magnetometer and gps

[petleh82]

Very impressive, the small drift, is it caused by the small gps drift, or is it just normal flight movment. Have you tried/what do you think about using a optical mouse sensor looking at the ground, to hold position, perhaps indoors to ?

[infernoed]

Thanks - I will do a new video sometime, algorithms have now been improved, the feedback was a little soft in this video. Raw static GPS data logged over 20mins gave 4x4m box, so some drift is due to GPS and some control loop firmness.

You are right, I am also looking at lensing a mouse sensor for indoors capability.

[VIPER4054]

@infernoed seriously, I think all I would need would be a decent FPV system and an OSD and I'm right to go! - see, py problem is my eye sight, and because of this problem, I cant fly the usual RC plane but from a cockpit perspective...no problems! I was wondering if YOU were to build me one like yours and include an OSD as well as a good FPV setup & cam... and yes my friend, I AM serious.

[Marquall]

Absolutesly amazing, I hopw to build one sooooon!

[Catz0125]

@IHeartRoboticsDotCom Uhh yeah. That little "String" You see is for the GPS Sherlock.

[infernoed]

Thanks for the encouragement - no strings attached ;-)

Uses accelerometer integration augmented with single SCP1000 running at 9Hz - Kalman filter to control offsets in estimated acceleration, velocity and height. Control loop works mainly on velocity feedback. Vertical acceleration is resolved independent of tilt attitude. Vibrations can upset things if accelerometer measurements get clipped. SCP1000 needs good layout and power supply to reduce noise. No GPS data used for altitude in this test