Switched system identification based on the constrained multi-objective optimization problem with application to the servo turntable

Multi-model switching controllers can handle the complexity of process plants and other engineered systems, based on the switched system model. In this paper, a switched Nonlinear Auto Regressive Moving Average eXogenous (NARMAX) model for a continuously mechatronic servo turntable system is proposed to make a better description than the single structure model. Then the identification issue is converted into a constrained multiobjective optimization problem (CMOP). Further, a kind of multi-objective particle swarm optimization algorithm is developed on the basis of Pareto optimality. The selection of global best location for each particle and the maintenance of the external archive are ensured in the same procedure in this algorithm. Finally, our theoretical findings are supported by simulation and experimental results.