多策略融合的改进狮群算法及其工程优化

Multi-strategy Fusion Improved Lion Swarm Optimization Algorithm and Its Application of Project Optimization

狮群算法是近年提出的一种智能优化算法,已经应用于多个领域,然而该算法仍存在搜索效率不足、易落入局部最优等问题.因此,基于狮群算法,提出了多策略融合的改进狮群算法.首先,使用Tent混沌种群的初始化方法,增强种群分布的均匀性的历遍性,提高算法初始解的质量和搜索效率;其次,采用柯西变异机制,在狮群最优位置采用柯西扰动操作,提升算法逃离局部极值的能力;再次,改进母狮位置更新方式和步长公式,提高算法后期的收敛精度;最后,融合精英反向学习,提高解的质量.选取国际通用的13个基准函数和部分CEC2014函数进行实验仿真,结果表明所提算法寻优性能和搜索精度有明显提升;另外通过对两个工程实例进行优化,结果表明改进算法在工程应用中具有优势.

The lion swarm optimization(LSO)is an intelligent optimization algorithm proposed in recent years and has been applied in many fields.However,the algorithm still has problems such as insufficient search efficiency and easy to fall into local optimum.Therefore,based on the lion swarm algorithm,an improved multi-strategy fusion lion swarm optimization(MFLSO)is proposed.Firstly,the initialization method of Tent chaotic population is used to enhance the ergodicity of the uniformity of population distribution and improve the quality and search efficiency of the initial solution of the algorithm.Secondly,the Cauchy mutation mechanism is used to improve the ability of the algorithm to escape from the local extremum by using Cauchy disturbance operation in the optimal position of the lion group.Again,improve the lion position update method and step formula to improve the convergence accuracy of the algorithm late;finally,the integration of elite reverse learning to improve the quality of the solution.Thirteen international benchmark functions and some CEC2014 functions are selected for experimental simulation.The results show that the optimization performance and search accuracy of the proposed algorithm are significantly improved.In addition,by optimizing two engineering examples,the results show that the improved algorithm has advantages in engineering applications.