改进的群搜索优化算法

摘要

标准的群搜索优化算法(GSO)是一种新的群智能优化算法,适用于解决高维函数的优化问题,而且简单高效,易于实现,但在其优化的后期容易陷入局部最优.为进一步提高其收敛速度和精度,对GSO算法进行了改进.保留其“发现者-加入者”模型,针对GSO算法发现者和游荡者搜索的无目的性,引进最大下降方向和杂交策略,发现者按角度搜索的同时也按最大下降方向进行搜索,游荡者通过基因突变策略的方式生成.通过23个基准测试函数对GSO算法和改进的GSO算法进行测试,结果表明改进的GSO算法在收敛速度和收敛精度上优于标准GSO算法.%Standard group search optimizer algorithm(GSO) is a new swarm intelligence algorithm,which has a superior performance on high-dimensional function optimization.It is simple and efficient,and easy to implement,but can't avoid entrapments by local minima.In order to enhance its convergence speed and precision,an improvement on GSO(IGSO) is presented.Inheriting the framework of "producer-scrounger" of GSO,in the light of the purposeless of producer and rangers,strategy of maximum descendent direction and hybrid of the group best member and the personal best member are presented.The producer searches according to the direction of standard GSO,and according to the maximum descendent direction at the same time.Tests through 23 benchmark functions on standard GSO and IGSO are carried out independently.The results show that IGSO has a preferable convergence rate and accuracy.