A Multi-Model Predictive Control Strategy Based on Data-Centric

A multi-model predictive control strategy based on data-centric used to deal with control problem of strong nonlinear systems is proposed. Firstly a multi-model modeling method based on local models is given. The modeling idea is to find some data matching with the current working point from vast historical system input-output datasets, then to develop a local model using local polynomial fitting algorithm. With the change of working points, multiple local models are built. Then a multi-model predictive control strategy is developed by combining the obtained multiple models with predictive control, thus solves the control problem of a sort of unknown-structure nonlinear system which based on vast history data only. The simulation results of the designed control system show that the control strategy can obtain satisfactory performance.