Computer Engineering and Applications ›› 2024, Vol. 60 ›› Issue (19): 80-96.DOI: 10.3778/j.issn.1002-8331.2401-0078

• Theory, Research and Development • Previous Articles     Next Articles

Improved Particle Swarm Optimization Algorithm Combining Differential Evolution and Sine Chaos

MA Lejie, ZOU Dexuan, LI Can, SHAO Yingying, YANG Zhilong   

  1. School of Electrical Engineering and Automation, Jiangsu Normal University, Xuzhou, Jiangsu 221116, China
  • Online:2024-10-01 Published:2024-09-30

融合差分进化和Sine混沌的改进粒子群算法

马乐杰,邹德旋,李灿,邵莹莹,杨志龙   

  1. 江苏师范大学  电气工程及自动化学院,江苏  徐州  221116

Abstract: Combining differential evolution with Sine chaos, an improved particle swarm optimization algorithm is proposed. It uses Sine chaotic mapping to optimize the initial population and improve convergence speed. This algorithm introduces a speed update formula for asynchronous learning factors and random inertia weights, enabling the algorithm to better balance global search and local optimization. It draws on the crossover operation in differential evolution algorithm, adopts a random search strategy with elimination mechanism to improve the algorithm’s global search ability and convergence speed. To verify the performance of improved particle swarm optimization algorithm, compared with PSO related algorithms such as yield-based particle swarm optimization (YPSO), self-adaptive particle swarm optimization (SPSO), as well as the latest algorithms such as spider wasp optimization (SWO) and energy valley algorithm (EVA) in 2023, the effectiveness of the improved particle swarm optimization algorithm (IPSO) that integrates differential evolution and Sine chaos is verified. It solves 12 commonly used benchmark functions in different dimensions, conducts experiments on 12 test functions, and compares them with other algorithms. The experimental results show that the improved PSO algorithm has fast convergence speed and high convergence accuracy.

Key words: particle swarm optimization algorithm, Sine mapping, differential evolution algorithm, cross operation, random search strategy

摘要: 将差分进化与Sine混沌相结合,提出一种改进的粒子群算法。利用Sine混沌映射对初始种群进行优化,提高了收敛速度;该算法通过引入非同步变化的学习因子的速度更新公式,引入随机惯性权重,使算法能够更好地兼顾全局搜索与局部优化;借鉴差分进化算法中的交叉操作,采用淘汰机制随机搜索策略, 提高算法的全局搜索能力,提高算法收敛速度。为了验证融合差分进化和Sine混沌的改进粒子群算法(improved particle swarm optimization algorithm, IPSO)的性能,与基于压缩学习因子的粒子群算法(yield-based particle swarm optimization, YPSO)、自适应加权粒子群算法(self-adaptive particle swarm optimization, SPSO)等PSO相关算法以及蜘蛛蜂优化算法(spider wasp optimization, SWO)、能量谷算法(energy valley algorithm, EVA)等2023年最新算法相比较,验证融合差分进化和Sine混沌的改进粒子群算法(IPSO)的有效性。在不同维度下解决12个常用基准函数,对12个测试函数进行实验,并与其他的几种算法进行比较,实验结果表明,改进后的PSO算法收敛速度快,收敛精度高。

关键词: 粒子群优化算法, Sine映射, 差分进化算法, 交叉操作, 随机搜索策略