作指导瓦:(1. 电子科技大学 自动化工程学院,四川 成都 610054; 2. 电子科技大学 光电信息学院,四川 成都 610054) 摘要:非线性系统参数优化的多种群并行遗传算法,以群体遗传算法搜索整个解空间实现参数优化。通过群体中个体间的信息交换,淘汰劣质基因,并用优秀个体反映解空间信息,使系统稳健收敛。该方法首先选取编码方案、交叉和变异概率、适应度函数、策略及遗传算子以及优秀个体的迁移率,并通过饱和非线性环节、速率限制环节和三阶环节验证了非线性系统的参数优化。 英文题名:Multi-Population Parallel Genetic Algorithm for Parametric Optimization in No-Linear System Abstract: Multi-population parallel genetic algorithm for parametric optimization of no-linear system is that while solving space is searched with the population genetic algorithm to realize the parametric optimization of non-linear system. The bad gene was eliminated through information exchange between individual in population, and information of solving space was reflected with fine individual to make the constringency of system is steady. Firstly, coding scheme, across and variation probability, function for degree of adaptability, strategy, genetic operator and mobility of individual were selected, the parametric optimization of no-linear system was validated with multi-population parallel genetic algorithm through saturated no-linear element, speed limitation element and third-order element. |