一种改进的粒子群算法研究

Study on an Improved Particle Swarm Optimization Algorithm

为了克服粒子群算法易发生早熟收敛、后期迭代速度较慢、易陷入局部最优的缺点,提出了一种改进的粒子群算法。该算法采用非线性动态自适应的更新权重,进一步提高收敛速度;通过引入差分进化算法中的交叉算子,以提高算法的全局探索能力,利用差分进化算法的变异策略产生候选解,克服种群多样性的下降,以跳出局部最优。利用该算法对2个测试函数进行寻优,仿真结果表明,文章提出的算法是一种收敛速度快、收敛精度高的全局寻优算法。

In order to overcome the shortcomings in the particle swarm optimization （such as premature convergence, slower iteration and tendency to local optimum）, an improved patti cle swarm optimization algorithm is proposed. This algorithm adopts the nonlinear dynamic adaptive update weight to improve the convergence speed. The crossover operator in the dif ferential evolution algorithm is introduced to improve the global exploration ability of the al gorithm. The mutation strategy of differential evolution algorithm is used to generate candi date solutions to overcome the decline of population diversity and avoid the local optimum. The algorithm has been used to optimize the two test functions. The simulation result shows that the proposed algorithm is a global optimization algorithm with fast convergence speed and high convergence precision.