Improved Invasive Weed Optimization Based on Adaptive Niche Algorithm
Xuemin Tian · Journal of Shanghai Dianji University · 2012
This paper aims to improve population diversity of the standard invasive weed optimization(IWO) to ensure better global convergence of the algorithm in dealing with the high dimension multimodal problems.By combining the Niche algorithm,the IWO algorithm is improved,named Niche invasive weed optimization(NIWO).This algorithm is enlightened by the idea of birds of a feather flock together.Individuals in the weed populations are first adaptively classified according to the Euclidean distance,and other operations are then completed.As a result,diversity of population is enhanced to improve the algorithm's capability of global optimization and convergence precision.The searching capability of the algorithm is verified based on four standard test functions.Experimental results show that,regardless of the low dimensional or high dimension multimodal function,The NIWO algorithm's search accuracy and stability are significantly better than the standard IWO algorithm.