A new hybrid optimization algorithm for recognition of hysteretic non-linear systems

In this article, a new two-stage hybrid optimization method based on the Particle Swarm Optimization and the Big Bang-Big Crunch algorithm (BB-BC) is introduced for identification of highly non-linear systems. In this hybrid algorithm, the term of the center of mass from the BB-BC algorithm is incorporated into the standard particle swarm optimizer to markedly improve its searching abilities. In order to investigate the effectiveness of the newly formed optimization algorithm in identification of non-linear and hysteretic systems, it is utilized to optimally find the Bouc-Wen model's parameters for a sample MR damper in which the damper's force is related to its piston's motion through a non-linear differential equation. The obtained results indicate that the proposed optimization method is highly robust and accurate and can be utilized successfully in such intricate non-linear identification problems.