Application Research Of Artificial Neural Network Method On Water Quality Simulation And Assessment

Artificial Neural Network (ANN) is the forefront of complex non-linear science and artificial intelligence science. In recent years, with increasingly perfect of ANN theory and applied technology, ANN is gradually becoming the focal point of every subject. Based on general analysis of basic principle of ANN and optimized arithmetic, the paper suggested introduce Back-Propagation Network(BP network) into the field of water quality simulation and water quality evaluation. The paper did some explorative research in Artificial Neural Network modeling methods and its adaptability in water quality simulation and water quality evaluation in Three Gorges.The research content mainly contain, first ,detailed elaborates the relation between the biological neuron and artificial neuron and the basic principle and inferential process of BP neural network and then studied the diffusing and moving rules in the observation surfaces of Zhutuo and Huangqian of the river pollutant (CODmn,DO and BOD5). Using the improvement BP neural network——A Levenberg-Marquardt feed forward learning algorithm, the water quality model was built to reflect the relation of backward position's pollutant concentrations and upward position's influence factors. It was applied to open out the impersonal rules of upward and back ward's main water quality factors, analysis and validate the applicability of this method. The main results show that:The advantage was showed according to the stability and forecast capability of model..Discussed the lake and reservoir nutrition neural network by resilient bp algorithm,Changshou Lake, which is located in the downstream of Longxi River (a tributary of Yangtze River), is carefully evaluated。Based on quality monitor data of of Three Gorges Reservoir in the yerar of 2000, theauthor considerably and systemically evaluated the present water pollution situation of main stream and tributaries of Yangtz River using single-factor evaluation method.By creatively applying multivariate statistical method firstly used and the SPSS10.0 adopted, the author analyzes the main pollutants of the monitor sectors of Yangtze River to study the relations between the water quality indicators of the Three Gorges Reservoir,then the paper studies the advantages and disadvantages of Fuzzy System and Neutral Network, then combines them organically. The paper applies Fuzzy Neural Network to water quality evaluation of Three Gorges Reservoir, and the effect is better. It provides a new method to evaluate waterquality.Finally,By analyzing the advantages and disadvantages and the complementary quality of he grey forecast technology and the artificial neural network forecast technology we can summarize five integrated models of load forecasting based on grey theory and artificial neural network, compensated model.the synthetic series connection model is chosen for modeling. by calculates with the monitoring data of Ciqikou we can prove that the model will make the grey model prediction accurately.This research demonstrates that with its theoretical feasibility and great practical utility, ANN water quality assessment and simulation models has good prospects for further development and application.