Application Research Of Genetic Neural Network In Surface Water Evaluation

Due to the defects of traditional water assessment method, this article constructs a model based on data mining technology——genetic neural network model. This paper analyzes the BP neural network and describes the basic principles and characteristics of GA, Discusses the theoretical foundation of optimizing the weights and thresholds of BP neural network based on GA, and lay a foundation for achieving scientifically water assessment method. The reasonable conversion of fitness function and BP network error of genetic algorithm implements the combination between genetic algorithm and BP network. This model is used to actual water evaluation, which proved the optimized network has been improved in convergence rate and convergence property. The result is compared to actual water level, fuzzy comprehensive method's results and index method's results, which indicate the genetic neural network model has scientificity and feasibility on water assessment.