Research On Cuckoo Search Algorithm And Its Application

The Cuckoo Search algorithm is a new heuristic intelligent optimization method designed by the scholars of the Cambridge university, UK, Xin- She Yang and Suash Deb in 2009. The algorithm is simple in searching process, involving less parameters, the search path is good and easy to understand, so the algorithm has been successfully applied into solving multi-objective problem, engineering optimization problem, the neural network training and so on. All of these showed very good performance. However, there is also some disadvantages to the algorithm itself. For example, slow speed of convergence, easily fall in local optima. It needs to improve the performance of CS from the basic theory, algorithm design and so on. And put it into application.The paper makes some researches of CS to solve above problems, and the obtained results can be summarized as follows:(1)In view of the disadvantage that the Cuckoo Search algorithm convergence precision is low and convergence speed is slow, get the Cauchy distribution into the cuckoo search movement formula of searching bird's nest. This paper proposes a Adaptive Cauchy-mutation Cuckoo Search Algorithm. To a certain extent, the results of experiments show that the algorithm improve those deficiency. At the same time,in terms of algorithm, the improved algorithm was applied into optimizing the the parameters of the air-suction roller precision metering device of notoginseng.(2)In view of the Cuckoo Search algorithm easily fall in local optima, sometimes affect the defects of global search results. In this paper, by introducing a Hybird-mutation operator to improve the performance of CS. The improved algorithm introduces Guass-mutation and Differential Evolution into the cuckoos search their nests progress. Which makes the algorithm easy to jump out of local optima. And the improved algorithm was applied to solve nonlinear equations and an engineering example.