A Tuning Scheme for Parameters of Generalized Predictive Controller Based on Mind Evolutionary Algorithm

This paper presents a scheme that the Mind Evolutionary Algorithm (MEA) tunes adaptively parameters of the generalized predictive controller. The value domain of parameters constitutes the solution space of MEA. The cost function of Generalized Predictive Control (GPC), the maximum value of the system output and its decay speed constitute the fitness function of MEA. During the control process, MEA adjusts constantly parameters so that the rapidity and robustness of GPC can be improved. Optimum experiments and simulation experiments show the scheme effectiveness and feasibility.