【摘要】 为改善花授粉算法后期收敛缓慢和容易陷入局部最优的缺点,提出一种基于监测邻域改进花授粉算法的方法。引入监测邻域,对邻域中粒子,利用适应度指数定标来确定不同的选中概率,选择部分花粉粒子进行变异,提高粒子种群的多样性和算法的效率。仿真实验中,选择两个典型函数对改进算法进行检验,验证了该算法改进的合理有效,运用改进花授粉算法、基本花授粉、Z-N法及云自适应粒子群算法进行PID控制器参数整定比较,仿真结果表明,改进花授粉算法寻优能力优于其它算法。更多还原

【Abstract】 To solve the problems of slow convergence and easiness of falling into the local optimum in the flower pollination algorithm at the late stage of the algorithm,a method based on monitoring neighborhood to improve the flower pollination algorithm was proposed.The monitoring neighborhood was introduced.To determine the probability of various particles selected,exponential operation was used.Some of the pollen grains were selected for variation.The diversity of the particle population and the efficiency of the algorithm were improved.In the simulation experiment,two typical functions were chosen to test the improved algorithm.Results verify that the improved algorithm is reasonable and effective.The IFPA,FPA,Z-N,and CAPSO were used to set the parameters of PID controller.The results show that the IFPA is better than other algorithms.更多还原