Research Of Parallel Distributed Evolutionary Computation Based On Internet

As a research focus of the domain of computational Intelligence, evolutionary computation has been successfully applied to machine learning, process controlling, economy forecasting and engineering optimization etc. With the development of the problems, the searching space becomes more and more complicated. The optimization time and the optimization quality of evolutionary computation can't keep up with the actual demand.In order to solve the massive complicated optimization problems, the author analyzes the parallelization principle and the application environment of Parallel Evolutionary Computation, and presents Internet-based Parallel Evolutionary Computation (IPEC). On the basis of Genetic Algorithm, the author analyzes the implementation characters of Parallel Genetic Algorithm in different application environments in detail. The key implementations of Internet-based Parallel Genetic Algorithm (IPGA) are discussed, and the corresponding program is also given in C++. In order to improve the performance of the algorithm, three important improvements of IPGA are presented. At first, in order to prevent the premature convergence of Genetic Algorithm effectively, the author brings forward a novel dyadic floating-point supplementary mutation operator. Then, simulating the natural evolution, the author presents a novel topology, unoriented-connected topology, for Parallel Genetic Algorithm. In the end, an interval decomposed optimization method is brought forward for IPGA, which can improve the optimization performance of the algorithm. In the last section of the paper, IPGA is applied to optimization design of the network. Experimental results demonstrate that IPGA can not only save the optimization time evidently but also largely improve the optimization quality, which provides an effective solution to massive optimization problems.