摘要
针对多智能体遗传算法收敛速度慢,求解精度有待提高的问题,提出一种新的反馈多智能体遗传算法。该算法融合了均匀设计思想,丰富了初始种群的多样性并予以验证;添加反馈算子,提升了算法的收敛速度,大大降低了函数评价次数。同时,对邻域竞争,变异和自学习算子大幅改进,结合算术交叉,以及二进制竞争的方式保留精英个体。高维函数优化实验表明,改进后的算法在很大程度上能避免陷入局部极值窘境,具有很好的全局寻优能力和更高的求解精度。
Aiming at the problem that the multi-agent genetic algorithm has slow convergence speed and low accuracy, an improved multi-agent genetic algorithm is proposed. The algorithm combines the idea of uniform design, increase the diversity of the initial population of the algorithm, and adds feedback operator to accelerate the convergence speed of the algorithm and greatly improve the neighborhood competition, mutation and self-learning operators. Using arithmetic crossover, and the way parents and children compete to retain the best individual, and the algorithm largely avoids the algorithm falling into local excellent or The dilemma of non-global optimal values. The high-dimensional function optimization experiments show that the algorithm has good global search ability and fast convergence speed, and avoids the algorithm being trapped in local excellent.
作者
杨超
石连栓
施承尧
武琴
YANG Chao;SHI Lian-shuan;SHI Cheng-yao;WU Qin(Institute of Information Technology Engineering,Tianjin University of Technology and Education,Tianjin 300222,China)
出处
《软件》
2020年第7期81-90,共10页
Software
基金
天津市科普重点项目(16KPXMSF00210)。
关键词
反馈算子
多智能体
均匀设计
高维函数优化
遗传算法
Feedback operator
Multi-agent
Uniform design
High-dimensional function optimization
Genetic algorithm
