基于扩散机制的双种群粒子群优化算法*

为了避免标准粒子群优化算法（PSO）过早收敛的缺点，把热力学中的扩散现象引入到PSO算法的改进当中，提出了基于扩散机制的双种群粒子群优化算法(DPSO)。DPSO算法中定义了粒子的扩散能、种群的温度和粒子的扩散概率三个概念，两个群体中的粒子在进化过程中根据粒子的扩散概率被选入到各自种群的扩散池中，从而实现两个种群之间信息的交换和共享。通过解决典型的多峰、高维函数优化问题来证实DPSO算法的有效性，实验结果表明DPSO比标准PSO具有更高的性能。

Bi-swarm particle swarm optimization based on diffusion mechanism

In order to avoid premature convergence, this paper applied the diffusion phenomenon of thermodynamics to improve standard particle swarm optimization (PSO), and proposed a modified PSO algorithm, the bi-swarm particle swarm optimization based on diffusion mechanism (DPSO). In DPSO algorithm, defined the diffusion energy of the particle, the temperature of the swarm and the diffusion probability of the particle. During the evolution of DPSO, chose the particle of each swarm into the diffusion pool of each swarm. The diffusion pool of both swarms exchanged and shared information. Experimental results on multi-modal, high-dimensional numerical optimization problems show that DPSO outperforms standard PSO.