基于双高斯函数的一种高效鸟群优化算法  

An efficient bird swarm optimization algorithm based on double Gaussian function

作　　者：彭君君 刘勇进 PENG Junjun;LIU Yongjin(College of Computer,Shenyang Aerospace University,Shenyang 110136,China;College of Science,Shenyang Aerospace University,Shenyang 110136,China)

机构地区：[1]沈阳航空航天大学计算机学院,辽宁沈阳110136 [2]沈阳航空航天大学理学院,辽宁沈阳110136

出　　处：《现代电子技术》2018年第23期106-112,共7页Modern Electronics Technique

基　　金：国家自然科学基金(11371255)~~

摘　　要：针对采用鸟群算法求解实际问题中的复杂函数时存在易陷入局部最优、学习能力差、缺乏收敛性理论分析等问题,提出基于双高斯函数的一种高效鸟群优化算法。该算法增加了鸟群的挑食行为,巧妙地避免初始寻优值易陷入局部最优点或鞍点的问题。同时,通过构建智能学习行为提高算法的自适应学习能力;然后构建双高斯函数更新法提高种群的多样性以增强算法全局搜索能力;最后,对于高效鸟群优化算法,给出时间复杂度分析。对多种标准测试函数进行仿真实验,实验结果表明,对于复杂函数优化,高效鸟群优化算法在达到收敛时其迭代次数相对基本鸟群算法减少50%左右,寻优成功率提高10%左右。Since the bird swarm optimization algorithm used to solve the complex functions in practical problem is easy to fall into local optimum,has poor learning ability,and lacks of theoretical convergence analysis,an efficient bird swarm optimization algorithm based on double Gaussian function is proposed.By adding the picky behavior of bird swarm,the problem that the initial optimization value is easy to fall into local optimal point or saddle point is avoided in the algorithm.The intelligent learning behavior is constructed to improve the self-adaptive learning ability.A double Gaussian function update method is constructed to improve the diversity of the population,and enhance the global searching ability of the algorithm.The time complexity analysis of the efficient bird swarm optimization algorithm is given.The simulation experiments are carried out for several standard test functions.The experimental results show that,in comparison with the basic bird swarm algorithm,the iteration time of the efficient bird swarm optimization algorithm is reduced by about 50%,and the success rate of optimization is increased by about 10%.

关 键 词：高效鸟群优化算法 双高斯函数 局部最优点 时间复杂度分析 全局搜索能力 迭代次数 

分 类 号：TN02-34[电子电信—物理电子学] TP301.6[自动化与计算机技术—计算机系统结构]