The Water Quality Predicition Based on the Gray Model and Curve Fitting ⋆

Water quality prediction can be applied to guard against all kinds of emergency events and provide decision support to the relevant departments. Many water quality prediction methods have supplied, such as time series analysis method, fuzzy algorithm, artificial neural network, wavelet analysis. Time series analysis method is suitable for the changing apparent data sequence, the predicted data are very random; fuzzy algorithm are used in forecast of the water quality, establishing the correspondences between predicting factors and predicting objects is more difficult; artificial neural network method is not a good method for the network topology; study and application of wavelet analysis is not very perfect in water quality prediction.Combining the water quality characteristics of the region, this paper uses the combination forecast method of gray model and curve fitting. Before prediction, it pretreats and classifies the corresponding data, using gray model predicts the periodic parameters, using the curve fitting method predicts the cyclical parameters. Then combining forecast result gets the final trend. Experimental data analysis shows that the fusion method has the higher forecast accuracy than the single method, if data are enough, this method may also be made better effect.