Abstract: In order to solve the problem of the particle swarm algorithm in slow search speed and easy to fall into local optimum. This thesis analyzed the iterative formula of the algorithm based on PID control theory, revealed that the speed update mechanism of the algorithm essentially adopts a proportional integral (PI) approach, and improved the iterative formula of particle swarm algorithm based on the theory of the PID control mechanism. In order to verify the effectiveness of the proposed strategy, It achieved the function of the algorithm by using MATLAB programming, and made a detailed comparison with the standard particle swarm algorithm by the benchmark test function. The results showed that the convergence rate of the improved particle swarm optimization algorithm is improved obviously, and the algorithm can avoid falling into local optimum.

Key words: Particle swarm, local optimum, proportional integral, PID control

摘要： 针对粒子群算法搜索速度不足和易于陷入局部最优的问题，基于PID控制理论的方法从其本质出发分析了该算法的迭代公式，揭示了该算法的速度更新机制实质上采用的是一种比例积分(PI)的方式，基于该理论采用PID的控制机理对粒子群算法的本质特性进行改进。为了验证所提策略的有效性，借助MATLAB编程实现了算法的功能并利用benchmark测试函数与标准粒子群算法进行了详细的实验对比。结果表明，改进后的粒子群算法收敛速度得到了明显的提高并且可以有效避免陷入局部最优。

关键词: 粒子群, 局部最优, 比例积分, PID控制