Parameter Estimation of Chaotic Systems Using Fireworks Algorithm

摘要

Chaotic system is a nonlinear deterministic system, and parameter identification for the chaotic system is an important issue in nonlinear science, such as secure communication, etc. By setting up an appropriate objective function, the parameter identification can be converted into a multi-dimensional optimization problem which can be solved by evolutionary algorithms. Emerging as an evolutionary algorithm, Fireworks Algorithm (FWA) has shown its good computational performance and robustness. In order to expand the application of FWA, several types of FWA are applied to estimate the parameters for two typical chaotic systems in which three parameters are totally unknown, simulation results show most of FWAs can have better estimation precision and robustness, and FWA is a new effective parameter identification method for the chaotic systems.