Research And Application Of Co-evolutionary Genetic Algorithm

Co-evolutionary genetic algorithm is a rising intelligence optimization algorithm in recent years. Based on the simultaneous evolution of one population or more, genetic algorithm is specialized to settle the complicate combinatorial optimization problem, which is proved to act superior performance in many application researches. Compared with traditional genetic algorithm, co-evolutionary genetic algorithm can avoid premature phenomenon, enhance capability of local search and promote operational efficiency. Considering the shortage of SOFM neural network optimized by genetic algorithm, we put forward cooperative co-evolutionary genetic algorithm to optimize SOFM neural network, and it can by applied to solve the problem of mine water bursting. The experiment indicates that the new algorithm shows better global convergence and higher operational efficiency, and succeeds in solving the death of neurons in SOFM. Apart from that, this paper designs the co-evolutionary genetic algorithm by dynamical niche technique which can work out the vehicle routing problem, the result shows that this algorithm can improve the solution efficiency.This paper does the main work as follows:1.The research history and research status of co-evolutionary genetic algorithm is briefly introduced, and the main value and content of this paper is summarized.2.The basic principle, process, parameter settings of the genetic algorithm is analyzed, together with advantages and disadvantages.3.It first introduces the basic idea Co-evolutionary genetic algorithm, then gives algorithm steps, and finally presents several typical Co-evolutionary genetic algorithm.4.Neural network is discussed, and designed by using cooperative Co-evolutionary genetic algorithm steps and algorithm process are given, the superiority of this algorithm can be verified as the distinguish of mine water bursting.5. Co-evolutionary genetic algorithm is also can be used in vehicle routing optimization problem, combined with dynamical niche technique, co-evolutionary model can be utilized to optimize vehicle path, the optimized results are shown in the experiment.