Model predictive control-review and case study

摘要

A multi-rate sampling formulation of the Model-based Predictive control (MPC) constrained optimal problem is introduced in this work. The penalty functions are used as the soft constraints for the state and output. This approach improves the constraint violation inside the sampling period as well as the problems related to specifying the weights for the sampling periods with varying length. This approach prevents from obtaining infeasible solution during the optimization iteration and reduces complexity and time consumption for optimization. The proposed algorithm with a continuous-time criterion was tested on PVC production control.