Improvement And Application Of The Bat Algorithm

Bat algorithm (Bat Algorithm, BA) is a new meta-heuristic search algorithm, which raised by Cambridge scholar Yang while he simulated echo location behavior in 2010. BA in principle is using the bat as individual solutions of the search space, by adjusting the loudness and pulse rate of bats, following the current optimal bat in solution space, so that the entire population generated from disorder to order during the search process. Bat algorithm has strong global search ability and convergence speed, but it has some problems as other global optimization algorithms,such as easy to falling into local optimal solution and the solution accuracy is not high.Firstly,This parper described current research and principle of bat algorithm and its process in detail, And analyzed the effect of the algorithm parameters such as ro and fmax,Then improving bat algorithm based on the analysis of it and Proposing two improved strategies. The main research work of this paper is:(1)This paper introduced solving optimization problem as beginning, then introduced the meta heuristic search algorithm principle and characteristics of it,and then introduced the bat algorithm in the domestic and foreign research status,and make a description of the bat algorithm process and the specific implementation in detailed.At last analyzed the function of bat algorithm in the relevant parameters of ro and fmax,making preparation for the improving bat algorithm in the following two chapters.(2)According to the defects of bat algorithm,This paper proposed chaotic bat algorithm, regulating bat velocity with the variable inertia weight factor, enhancing the algorithm’s local search ability though join the local search of chaotic mechanism. Accord ing to the experimental data,it compared the ability of solving the problem to BA,DEBA, and analyzed the accuracy variation of the algorithm in different dimensions.(3)Based on CBA,this paper fused the differential evolution theory into bat algorithm,and put forward a kind of differential evolution hybrid bat algorithm based on quadratic difference, The two different types of differential methods enhanced the global search and local search capability of the bat algorithm.The experiments show that the improved algorithm has a better ability to solve the problem. Then function convergence curve of the four improved algorithms are analyzed and compared.(4)In the paper, The bat algorithm and its improved algorithm are applied to function optimization, combinatorial optimization and constrained optimization. the advantages and disadvantages of each algorithm are compared By solving the nonlinear equations,0-1 knapsack problem. Finally, the design of the welding strip is used to illustrate the application of the bat algorithm in engineering practice.(5)Finally,this paper summarized the research work,and explained the next research direction:Trying to apply the bat algorithm and its improved algorithm into the field of uncertain optimization.