摘要
本文提出了一种新的带有马尔可夫跳跃的PSO算法(MJPSO),在MJPSO中,通过评估每一代的进化因子,粒子的惯性权重参数和加速度系数均可以根据一个齐次马尔可夫链自适应地调整。设计的新型变异策略,可以帮助粒子根据适当的概率,提高逃离局部优化陷阱的可能性。该策略不仅降低了计算成本,还改进了全局搜索,提高了粒子的全局搜索效率和搜索能力。一系列广泛使用的优化基准实验的结果表明,所开发的MJPSO算法优于现有的六种流行的PSO算法的变体。
In this paper, a new Particle Swarm Optimization algorithm with Markov Jumping (MJPSO) is pro-posed, in which both the inertia weight parameter and the acceleration coefficient of a particle can be adaptively adjusted according to a chi-square Markov chain by evaluating the evolution factor of each generation. A new variational strategy is designed to help the particles to improve the possibil-ity of escaping the local optimization trap according to the appropriate probability. The strategy not only reduces the computational cost, but also improves the global search and increases the global search efficiency and search capability of the particles. The results of a series of widely used opti-mization benchmark experiments show that the developed MJPSO algorithm outperforms six ex-isting variants of the popular PSO algorithm.
出处
《理论数学》
2023年第3期732-749,共18页
Pure Mathematics
