Kalman Filter Embedded MPC for Stochastic Systems

This paper considers a computationally efficient Model Predictive Control (MPC) framework to design control for stochastic systems. The probability distribution function of the disturbance is utilized in the design of control. The computational efficiency is contributed by three factors: monotonically weighted cost function, reduction in prediction horizon and the concept of event triggering. Kalman Filter is embedded in the MPC in order to achieve a more accurate value of states to be used in the optimization problem. Monte Carlo simulation is carried out on a benchmark system to verify the advantage of the proposed technique.