基于差分进化改进的蝗虫优化算法

Improved Grasshopper Optimization Algorithm Based on Differential Evolution

蝗虫算法(Grasshopper optimization algorithm,GOA),是一种以蝗虫群体行为为理论基础的新群智能算法,其性能优越,但仍然存在求解精度不高,收敛速度慢,陷入局部最优的不足。为了提高蝗虫算法的求解精度和收敛速度,减低其搜索盲目性,提出了一种基于差分进化改进的蝗虫优化算法(DE-GOA)。在蝗虫算法的迭代后期,加入差分进化策略增加种群多样性,提高整个算法的全局寻优能力。最后将改进的差分进化蝗虫算法与基本蝗虫算法GOA,PSO算法,GA算法,FPA算法,FA算法,BA算法在11个标准测试函数进行寻优实验对比,实验结果表明DE-GOA算法相比其他算法具有明显的优势。

Grasshopper optimization algorithm is a new Swarm Intelligence Algorithms based on the theory of swarm behavior of Grasshopper.Its performance is superior,but there are still shortcomings such as low accuracy,slow convergence and fall-ing into local optimum.In order to improve the accuracy and convergence speed of Grasshopper algorithm and reduce its search blindness,an improved Grasshopper optimization algorithm based on differential evolution(DE-GOA)is pro-posed.In the later iteration stage of Grasshopper algorithm,differential evolution strategy is added to increase the diversi-ty of population and improve the global optimization ability of the whole algorithm.Finally,the improved DE-GOA algo-rithm is compared with the basic GOA algorithm,GA algorithm,PSO algorithm,FPA algorithm,FA algorithm and BA al-gorithm on 11 Standard Test functions.The experimental results show that DE-GOA algorithm has obvious advantages over other algorithms.