一种带自适应飞行时间因子的粒子群算法

Particle Swarm Optimization with Adaptive Flying Time Factor

作为一种新型智能算法,粒子群算法具有概念简单、易于实现等特点,但也存在容易陷入局部最优的缺点。为了尽可能找到问题的最优解,提高粒子群算法的收敛速度,提出一种带自适应飞行时间因子的粒子群算法,在算法中引入种群多样性和种群进化度两个参数,并根据这两个参数对算法性能的影响,让飞行时间因子随着这两个参数自适应改变。通过对4个基准函数的测试表明,改进后的粒子群算法较其他几种粒子群算法在收敛速度和收敛精度上都有一定提高。

As a new intelligence algorithm,particle swarm optimization with advantages of simple conception and easy implement also suffers the high risk of trapping in a local optimum. To find the optimal solution,a particle swarm optimization algorithm with adaptive flying time factor was presented. Two parameters of species diversity and popu[a tion evolution degree were introduced,and the flying time factor changed with these two parameters adaptively. The test of the four benchmark functions shows that the modified particle swarm optimization algorithm is better in convergence speed and convergence accuracy compared with other kinds of particle swarm optimization algorithms.