中文摘要：

针对粒子群算法容易陷入局部极值、进化后期收敛精度低的缺点，提出了一种基于扰动的精英反向学习粒子群算法。算法采用在粒子迭代的过程中，以一定的概率对当前的最优个体进行动态一般反向学习生成其反向解，引导粒子向最优解空间靠近；用一种非线性递减的方式改变惯性权重，以提高算法的收敛速度和收敛精度;采用扰动的方式增强算法的局部探索能力，帮助粒子跳出局部最优解。在14个标准函数上进行仿真测试，结果表明改进算法具有更高的收敛速度和收敛精度，能有效地避免陷入局部最优，适合求解函数优化的问题。

英文摘要：

In order to overcome the shortcomings of particle swarm optimization( PSO) algorithm, such as easily falling into the local optima and low precision at later evolution process, this paper developed a modified PSO algorithm based on disturbances and elite opposition-based learning. In every iteration, the current best individual executed dynamic generalized oppositionbased learning to generate their opposite solutions with a certain probability, which guided the particle to approximate the optimum space. Meanwhile, the algorithm used a non-linear decrease method to change inertia weight, it could improve convergence speed and accuracy of the algorithm. And it used a disturbance approach to enhance the ability of local exploration and helped the particle escape from local optima. The simulation experiments are conducted on fourteen benchmark functions, the results show that the improved algorithm has higher convergence rate and accuracy, it also can effectively avoid being trapped in local optimal solution and is suitable to solve the function optimization problem.