基于扩散机制的杂交粒子群优化算法*

为了解决标准粒子群优化算法容易陷入局部极小值的问题，模拟统计物理和热力学中的扩散现象，设计了一种扩散机制，根据扩散定律和扩散系数公式，给出了粒子的扩散能、种群的温度和粒子的扩散概率三个定义和扩散池的概念;并把这种策略和多父体杂交算子结合起来，提出了基于扩散机制的杂交粒子群优化算法。该算法在具有欺骗性的多模态函数优化和非线性模型参数估计等实际问题上取得了较理想的实验结果，证实了扩散机制和多父体杂交策略可以有效地改善粒子群优化算法的性能。

Crossover particle swarm optimization algorithm based on diffusion mechanism

In order to solve the defect that the standard particle swarm optimization algorithm is easy to fall into the local minimum, this paper designed a kind of diffusion mechanism by simulating diffusion phenomenon in the statistical physics and thermodynamics. According to the law of diffusion and the equation of diffusion coefficient, defined the diffusion energy of the particle, and the temperature of the swarm and the diffusion probability of the particle, and also introduced the diffusion pool. It proposed the hybrid particle swarm optimization algorithm (DCPSO), which combined with the diffusion strategy and multi-parent crossover operator. The experiment results on the deceptive multi-modal function optimization and the nonlinear model parameter estimation confirmed that the diffusion mechanism and multi-parent crossover strategy can effectively improve the performance of particle swarm optimization.