论文部分内容阅读
从仿生学和心理学角度出发,提出一种深度扩展记忆的仿人粒子群算法,以解决标准粒子群及其主流改进算法易陷入局部最优等问题.算法对粒子认知进行群体共享,并采用深度扩展记忆积累粒子认知,通过仿人遗忘函数配置不同时期认知对当前决策的影响权重.仿真分析表明,所提出算法对遗忘函数和遗忘因子高度敏感,算法寻优多维多极值函数时,在收敛精度、成功率和优化成本等方面较标准粒子群及其改进算法有显著提升.
From the perspective of bionics and psychology, a novel particle swarm optimization algorithm with extended memory is proposed to solve the problem that standard particle swarm optimization (PSO) and its mainstream improved algorithms are easy to fall into the local optimum, etc. The algorithm shares the particle cognition and adopts Depth expansion of memory accumulation of particle cognition, through human-like forgetting function to configure the cognitive weight of different periods of the current decision-making. Simulation shows that the proposed algorithm is highly sensitive to the forgetting function and forgetting factor, Compared with the standard particle swarm optimization and its improved algorithm, the convergence precision, success rate and optimization cost are significantly improved.