Generalized Predictive Control Based On Particle Swarm Optimization And Differential Evolution

The paper focuses on generalized predictive control based on hybrid differential particleswarm optimization algorithm. Based on the study of hybrid particle swarm optimization anddifferential evolution to calculate, the optimal control input increment can overcome thedeficiency of the iterative optimization. This algorithm gets an optimal objective function valuemore quickly and accurately. The main content of paper includes the following areas:Firstly, this paper introduces the basic principles of generalized predictive control and itsimproved implicit algorithm. Implicit generalized predictive control algorithm avoids solvingDiophantine equations, in order to reduce effectively the amount of computation and acceleratethe response time. The simulation examples prove the good control performance.Secondly, this paper investigates the generalized predictive control based on particleswarm optimization. As the traditional generalized predictive control is presented without theconstraint, so finding the method to solve the constrained industrial control systems is urgent.Particle swarm optimization is an algorithm which originates from swarm intelligence theory. Ithas of highly accuracy, easily operation, and fast convergence to solve the problem ofconstrained optimization. The paper describes PSO in generalized predictive control application.Besides simulations analyses prove the effectiveness and the good control performance.Finally, the paper adopts generalized predictive control based on hybrid PSO and DE tosolve the deficiency of the iterative optimization of GPC to make the output smoothly andaccurately tracking system output settings. Theory study and simulations analysis prove theeffectiveness and the good control performance. This method has high practical value of itswidely application range, fast response, highly stability.