Adaptive inverse control application on thermal system based on least squares support vector machine

An adaptive inverse control method based on support vector machine is studied in this paper. The initial inverse model of process is built based on least squares support vector machine, and the number of support vector is reduced by doing pruning algorithm. In this adaptive inverse control mechanism, the inverse model is updated through recursive least squares algorithm, and controller is adjusted online according to method of epsiv filter. Simulation results show that this identification method has excellent approximation ability, and system of adaptive inverse control has good robustness and performance.