一种基于引导策略的自适应粒子群算法

为解决粒子群算法前期搜索“盲目”，后期搜索速度慢且易陷入局部极值的问题，对算法中粒子更新方式和惯性权重进行了改进，提出了一种基于引导策略的自适应粒子群算法。该算法在种群中引入4种粒子，即主体粒子、双中心粒子、协同粒子和混沌粒子对粒子位置更新进行引导，克服算法的随机性，从而提高搜索效率；为进一步克服粒子群优化算法进化后期易陷入早熟收敛的缺点，引入聚焦距离变化率的概念，通过聚焦距离变化率的大小动态调整惯性权重，以提高算法的收敛速度和精度，两者结合极大地提高了搜索到全局最优解的有效性。对4个标准测试函数进行仿真，实验结果表明IPSO算法在收敛速度、收敛精度以及成功率上都明显优于LDWPSO和WPSO算法。

An adaptive particle swarm optimization algorithm based on guiding strategy

In order to solve the problems of blind search in the early stage and slow search speed as well as easily trapped in local minima in the later period, an adaptive particle swarm algorithm based on guiding strategy (IPSO) is proposed by improving the particle updating way and inertia weight. Four kinds of particles were introduced in the population of the algorithm, which are the main particle, double center particle, cooperative particle and chaos particle. The algorithm decreases the randomness and improves the search efficiency through guiding particle position updating. Moreover, the focusing distance changing rate is introduced in the new algorithm. The inertia weight is adjusted dynamically by the size of the focusing distance changing rate to improve the convergence speed and accuracy of the algorithm. The effectiveness of the search for the global optimal solution is greatly improved by the combination of the both modes. The simulation experiments are conducted on the four benchmark functions. The results show that IPSO has obviously higher convergence rate, convergence accuracy and success rate than LDWPSO and WPSO.