Cultural Algorithm And Its Application In Optimal Scheduling Of Applied Research

Cultural Algorithm (CA) provides an explicit mechanism for acquisition, storage and integration of individual and group's problem solving experience and behavior. The main idea of CA is individuals from the population component interact with knowledge in the belief space during the search process,individual experience can modify knowledge stored in the belief space and that knowledge can likewise be used to influence individuals in the population.This can improve the rate of search and convergence. Therefore CA performs better on some problems than the traditional evolutionary algorithms, especially in solving constrained optimization problems, the global optimization ability and computational efficiency is better than traditional evolutionary algorithms. So far, CA has been successfully applied to many areas,such as machine learning, automatic control, semantic networks, production scheduling. For each specific problem, the difficulties of CA are the design of belief space and the implementation of influent function.This paper proposes the cultural algorithm based on evolutionary programming and hybrid algorithm based on cultural algorithm and modified differential evolution after well study CA. Using improved CA to solve constrained optimization problems and Flow Shop problem. The main research work is as follows:(1) The paper elaborates the sources, the basic principles, mechanisms,features and application of CA in detail.(2) Cultural algorithm based on evolutionary programming (CAEP) is proposed in this paper. This algorithm embeds evolutionary programming into the CA framework, which utilizes knowledge extracted during evolution to guide the evolution of population space. In this way,it speeds up the evolutionary process.(3) A new hybrid algorithm based on cultural algorithm and modified differential evolution (CAMDE) is put forward, which embeds modified differential evolution into the CA framework. The algorithm implements the acquisition of knowledge and guides the evolution of population space based on modified differential evolution. In the way, it assures the diversity of population space and the speed of convergence.(4) Simulation tests are performed based on benchmark functions and the results indicate that the two algorithms proposed perform better than traditional evolutionary algorithms.(5) CAEP and CAMDE were used to solve the production scheduling problem of butene alkylation, and CAMDE was used to solve Flow Shop problem. The results suggest that the two algorithms can improve the rate of search, and confirm the feasibility of using CA to solve production scheduling problem.