The Study And Application Of Improved Water Cycle Algorithm

The water cycle algorithm is an intelligent optimization algorithm abstracted from the natural water cycle process.On the one hand,the algorithm is based on the guided advance of the river,which can effectively prevent individuals from entering the non-optimal region.On the other hand,there is an evaporative mutation process,which is beneficial for the algorithm to jump out of the local optimal solution and find the global optimal solution more easily.However,the convergence speed and accuracy of the algorithm still have some defects.This paper makes some work to improve the water cycle algorithm.Aiming at the defects of the basic water cycle algorithm,this paper proposes three improvement ideas:Firstly,the improvement based on initial population,reverse learning is used to replace the traditional random initialization in the initial solution of the water cycle algorithm to improve the uniformity and diversity of the initial population and improve the quality of the initial solution.Secondly,the improvement based on iterative step size,in the iteration process of the water cycle algorithm,the iterative step size with exponential decline is used to replace the original fixed step size,so that the algorithm can rapidly narrow down the optimal neighborhood in the early stage of iteration,and then fine-tune in the small neighborhood in the late stage of optimization,making it easier to obtain better solutions.Finally,the improvement based on evaporation process,the combination of gaussian variation and chaotic variation is used to form the adaptive rainfall process,and the search mode combining global and local is realized,which overcomes the disadvantages of the original algorithm,such as low diversity and slow convergence speed.Benchmark functions were selected to compare and test the performance of the improved water cycle algorithm with other intelligent algorithms.The experiments verified that the improved water cycle algorithm has better stability,faster convergence speed and higher optimization accuracy.In general,the optimization effect of this algorithm has been improved,and it is obvious.Aiming at such problems as the k-means clustering algorithm is sensitive to initial value and the convergence speed of clustering center is slow,the improved water cycle algorithm and other intelligent algorithms are used to carry out comparative experiments on the clustering effect of the k-means algorithm,which further verifies that the improved water cycle algorithm has significantly improved the clustering effect of the k-means algorithm.