基于Tent-Chebyshev切换的粒子群优化算法

Particle Swarm Optimization Algorithm Based on Tent-Chebyshev Switching Mapping Strategy

针对种群位置初始化过程中由伪随机数引起的局部最优与早熟收敛问题,提出一种基于Tent-Chebyshev切换映射的粒子群优化算法。在使用Tent混沌映射函数计算粒子初始位置时,基于Chebyshev混沌映射函数对种群中陷入不动点的粒子引入随机扰动,使其重新进入混沌状态。通过多次迭代的数据分布结果和人体行为识别问题的应用实例,均验证了所提出算法在确保种群初始解能够随机遍历取值空间的同时,也有效降低了局部最优解对初始粒子的束缚,为后续更新过程提供多样性丰富、粒子特征显著的优质初始粒子群。

Aiming at the problems of local optimum and premature convergence caused by pseudo-random sequence during the process of particle initialization,this paper proposed an improved particle swarm optimization algorithm based on Tent-Chebyshev mapping switching strategy.In order to enable particles trapped in a fixed point to re-enter the chaotic state,a random disturbance calculated by the Chebyshev chaotic mapping function was introduced when using the Tent chaotic mapping function to calculate the initial position of particles.According to iterative distribution results and the application examples of human action recognition problems,the proposed algorithm was proved to be able to randomly traverse the value space of the population initial solution,but also effectively reduced the constraint of the local optimal solution on the initial particles,and provided a high-quality initial particle swarm with rich diversity and prominent particle characteristics for the subsequent update process.