Abstract: Based on the Particle Swarm Optimization (PSO) and the multi-modeling, a PSO-based Piece-Wise Affine (PWA) modeling method was proposed to deal with the complex nonlinear system. The proposed algorithm translated the PWA modeling problem to the Mixed Integer Quadratic Programming (MIQP) firstly, and then the PSO was employed to deal with it. The layering strategy was employed in the process of solving, which not only decreased the dimension of the optimization problem effectively, but also reduced the probability of local optimization. A simulation example indicates the effectiveness of the proposed modeling method.

Key words: nonlinear system, Piece-Wise Affine (PWA) multiple model, modeling, Particle Swarm Optimization (PSO)

摘要： 针对复杂非线性系统，将微粒群优化(PSO)算法与多模型建模相结合，设计了一种基于PSO算法的非线性系统分段仿射(PWA)多模型建模算法。该算法将PWA多模型建模问题转化为混合整数二次规划（MIQP）问题，并基于PSO算法对其进行优化求解。在求解的过程中，采用分层优化求解方法，有效降低优化问题的维数，减小了陷入局部最优的概率，并通过仿真验证了该算法的有效性。

关键词: 非线性系统, 分段仿射多模型, 建模, 微粒群优化算法