Solving an Inverse Control Problem Using Predictive Methods

Model Predictive Control (MPC) is the class of control methods that optimize a specified performance index in order to minimize weighted future output deviations from a set point trajectory. MPC operates on a receding horizon, calculating a set of inputs at each time step. The controller then implements the first input and the process begins again. The performance index can also include a weighted and/or constrained control sequence which can be of different length than the output horizon.