Adapted Parameter Differential Evolution Algorithm Based On The Water Network Optimization

Differential evolution algorithm for the practical application of the process is difficult to determine the best control parameters, control parameters directly related to the appropriateness of the performance of optimization algorithms, two types of control parameters proposed adaptive differential evolution algorithm; the same time, the water network optimization for complex problems were two kinds of modified differential evolution algorithm to achieve global optimization of water networks to obtain good results.Information mining groups were proposed based on the differential evolution algorithm with control parameter coevolution based on population information (CPCE-PI-DE) and differential evolution algorithm with coevolution of normal distribution control parameter (NDCPCE-DE). In both algorithms, each individual has a corresponding individual control parameters symbiotic; which CPCE-PI-DE by multiple regression analysis to extract the optimal control parameters symbiotic evolution of individual direction control parameters to achieve co-evolution; NDCPCE-DE control parameters by means of normal distribution and update the control parameters to achieve the co-evolution.Tested by standard test functions compared with other improved algorithms, indicating that the two types of algorithms have good global optimization.Two improved differential evolution algorithm will be applied to solve a three-water unit of the water network constrained optimization problem, the results show that:two types of algorithms are the global optimal solution can be obtained, indicating that the water system, intelligent optimization algorithm in global optimization feasibility.