Fuzzy Cuckoo Search Algorithm And Its Application

Cuckoo search algorithm( CSA) proposed by Xin-she Yang and S. Deb in2009 is not only highly efficient to optimize, but also easy to search, less parameters to use,better path to choose, and powerful ability to solve multi-objective problem. The speed of convergence and quality of the solution are directly affected by the choice of step length and discovery of probability ap.This thesis mainly studied how to choose the step and to find the probability apsuch that the algorithm has more faster rate of convergence and accurate solution. Main contents are as follows:Firstly, we propose a new cuckoo search algorithm with adaptive step length(RDCSA).Since the step length of the original self-adaptive step cuckoo search algorithm is not update when the current position of the nest is the optimal position, the rate of convergence will descend. We simply modified the step length to solve it. Then the range of zoom factor of the cuckoo search algorithm, new algorithm is proposed to improve. The results of the simulation experiment show that RDCSA has better rate of convergence and accuracy.Secondly, the fuzzy control technique is applied to dynamically adjust the finding of probabilityap in cuckoo search algorithm. A fuzzy cuckoo search algorithm( FCSA) is therefore developed.The results of the simulation experiment show that FCSA compared with CSA not only improve the premature situation but also has better rate of Convergence and accuracy.Finally, the FCSA has better practical performance, when applied to Spring Design Optimisation problem.