Multimodel Control of Nonlinear Systems Using A New Weighting Method Based on Gap Metric

Multi-linear model approach is employed to control a strongly nonlinear process where a global controller is constructed based on a weighted combination of local controller outputs. In this paper, a new weighting method is presented for multimodel control of nonlinear systems, in which the gap metric is used to design weighting functions for local controllers combination. Compared to conventional weighting approaches, the suggested new weighting approach requires just one tuning parameter. It is therefore simple and easy to adjust, and improves closed-loop performance.