Rating is available when the video has been rented.
This feature is not available right now. Please try again later.
Published on Jul 2, 2013
Model Predictive Control (MPC) is an advanced control strategy that allows not only control of multivariable systems (such as the Quanser 3DOF Helicopter in this video), but also allows system constraints to be implemented.
In this video we have implemented our MPC algorithm on a Texas Instruments Delfino C28343 microcontroller running at a 50Hz sampling rate. The pitch angle (angle from horizontal of the two propellers) is constrained to be +-30 degrees, and the system is subjected to a range of large setpoint changes and unmeasured disturbances. The algorithm performs well, and the system remains stable before return to 'base'.
The controller is auto-coded from a specification in MATLAB. Google "jMPC Toolbox" for a free download of the MATLAB Toolbox. The controller is an implementation of linear MPC with online optimization, solving a quadratic program at each sampling instance. A Kalman filter is used to estimate unmeasured states (angular velocities), while the system state-space model is derived from a linearized version of the nonlinear equations of motion.