The robustness optimization of parameter estimation in chaotic control systems

Standard particle swarm optimization algorithm has problems of bad adaption and weak robustness in the parameter estimation model of chaotic control systems. In light of this situation, this paper puts forward a new estimation model based on improved particle swarm optimization algorithm. It firstly constrains the search space of the population with Tent and Logistic double mapping to regulate the initialized population size, optimizes the fitness value by evolutionary state identification strategy so as to avoid its premature convergence, optimizes the inertia weight by the nonlinear decrease strategy to reach better global and local optimal solution, and then optimizes the iteration of particle swarm optimization algorithm with the hybridization concept from genetic algorithm. Finally, this paper applies it into the parameter estimation of chaotic systems control. Simulation results show that the proposed parameter estimation model shows higher accuracy, anti-noise ability and robustness compared with the model based on standard particle swarm optimization algorithm. © 2015 Kavala Institute of Technology.