Research And Application Of Shuffled Frog Leaping Algorithm

Shuffled frog leaping algorithm, a post heuristic optimization algorithm simulated nature frog group foraging process, is a kind of new intelligent optimization algorithm. Although it has a strong global search ability, it still has the defects of premature convergence and local optimum when solving complex optimization problems.As a typical representative of stochastic optimization algorithms, genetic algorithm has the advantages of parallel, global search ability, strong robustness and so on. Therefore, on the analysis of the advantages and disadvantages of shuffled frog leaping algorithm, and drawing lessons from the idea of genetic algorithm, improved shuffled frog leaping algorithms are studied in this thesis. The main contents are as follows:1. Drawing on the idea of genetic algorithm, the natural selection and crossover operations are used in the global search and the mutation operation is used in the local search, the shuffled frog leaping algorithm with genetic operations is put forward. Constituting a test environment with typical test functions for optimizing experiments, the results are much better than the basic shuffled frog leaping algorithm, which indicates that the introduction of genetic operations makes the performance is greatly improved. The proposed algorithm is used to estimate the parameters of the proton exchange membrane fuel cell model, and the experimental results verify the validity.2. Using the DNA encoding method, the chromosomes in genetic algorithm are coded, the crossover operator of shift reconstruction is designed, and the mutation operator is used. Using the idea of the species diversity, the multiple group cycle strategy is proposed in order to expand the search scope of the solution space, and the genetic algorithm with multiple group cycle strategy is presented. To generate the initial solutions for shuffled frog leaping algorithm, the mixed shuffled frog leaping algorithm based on the genetic structure with multiple group cycle strategy is proposed. Through the typical unconstrained test functions, and compared with the results of other algorithms, the effectiveness of the proposed algorithm is verified. The proposed algorithm is used to estimate the parameters of heavy oil thermal cracking process model, the results are compared with the other algorithms in the literatures, which shows the superiority of the proposed algorithm.