Variable Structure Neural Network Used In Forecasting And Evaluation Of Water Quality Parameters

The populace,natural resource,environment and ecology are problems confronted by the world. Water is one of the unreplaceable natural resouces,which closely connected with the other three .In the past, the exploitation,utilization,administration,protection and management are crude to a great extend,with the growth of economy,there are lots of inharmony problems exposed under this development mode,such as the soil erosion and water pollution. The shortage of water and polluted water are seriously problems to the growth of economy of our country.Therefore,to intensify the water quality warning,forecasting,preventing and harness of water pollution,improving the comprehensive usage of water were already become one of the problems we have to pay great attention to. Intensify the management and protection of water environment is the most important conditions to realizing the sustainable development of our economy.On the base of researching water quality parameters,water quality evaluation methords and water quality parameters forecasting mode,this paper dewells on the features of BP algorithm and It's shortness,a new way are given to improve the accuracy of forecasting;because of It's hard to decide the structure of NN,we introduce the method of variable structure neural network to practice,comparsions are also given of these two methods's result. Cureent water quality evaluating methods are given in the paper and Variable BP NN evaluation method is introduced to the evaluation of water quality.ANN is the hottest arena of artifical intelligence,Proceess NN are introduced in the last chapter hoping It can be used in reality.