Nonlinear Model Predictive Control on a PLC - An Approach for the Use of MPC in Automation

Model predictive control (MPC) is a model based method which is well suited for controlling nonlinear systems with constraints. It relies on the solution of an underlying optimal control problem (OCP) and typically requires considerable computational effort. This paper discusses an efficient MPC approach for nonlinear input constrained systems with sampling times in the millisecond range that is suitable for implementation on a programmable logic controller (PLC). The efficiency of the MPC scheme is demonstrated by means of a laboratory crane setup.