Continuous time model predictive control (CMPC) is a control algorithm similar to discrete time model
predictive control in that it uses a quadratic cost function to find the optimal control inputs within given constraints. Inputs which give the lowest cost function are the optimal control inputs. The benefit of CMPC over its discrete time counterpart is the use of a continuous time linear model to predict future system behaviour at each time sample. The control inputs are modelled with orthonormal basis functions.
This simplifies the calculation of the optimal control inputs in continuous time. The cost function uses an exponentially data weighting strategy to improve the stability and tracking performance of the controller.
Simulations with a non-linear F16 jet aircraft MIMO model show that CMPC can control the attitude of
the aircraft with good performance under certain conditions.
Traineeship report. - DC 2013.027