Study On Water Quality Assessment And Prediction Method Of Water Source

With the rapid development of our national economy, the increasingly prominent problems of water shortage, water waste and water contamination have threatened the security of drinking water and the sustainable development of economy. Therefore, the strengthening of water resources protection and the establishment of a sophisticated water quality early-warning system are significant to prevent water pollution events and ensure drinking water safety.Based on the research and application of water quality safety assessment and early-warning technology supported by the "Water Pollution Control and Treatment Project", a water quality assessment method based on combination weight and a prediction method of chaotic time series of water quality using support vector machine (SVM) are implemented in this thesis. The main content and innovative points are as follows:(1)The water quality safety assessment and early-warning system is improved, including water quality assessment method, water quality assessment index system, classification and weight calculation method of water quality indicators, and so on, which has been successfully applied in three demonstration areas.(2)A new water quality assessment method based on combination weight is presented. Firstly, the subjective and objective weights of the evaluation indicators are calculated by following methods:Analytic Hierarchy Process (AHP) method, dynamic entropy method and the weighting method based on the proportion of the pollutants over water quality standard level. These weight vectors are integrated based on Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method to gain the combination weights of evaluation indicators. Then some typical water quality indicators are taken as evidences and a new evidence combination rule is introduced to combine these evidences on the basis of the combination weights of the evaluation indicators. Finally the water quality class with the highest belief is regarded as the evaluation class. Compared with the traditional and single weight calculation method, the presented method is more reasonable and scientific with a consideration of the harm to human health, the content of pollutants in water and so on. Application of the presented method is carried out and the assessment class accords to water status, which shows that this method is objective and effective. Using this method, the water quality trend can also be analyzed by belief fluctuations of water quality classes.(3) Application of chaos theory to the prediction of time series of water quality is studied in this thesis. The chaotic characteristics of time series is studied by principal component analysis (PCA), largest Lyapunov exponent and saturated correlation dimension, and the results show that time series of water quality has chaotic property. Based on the study of chaotic characteristics of time series of water quality, a new prediction method of chaotic time series using SVM is proposed. In this method, the phase space of time series of water quality is reconstructed. Then a water quality prediction model based on SVM is established. In the meantime, the parameters of SVM are calculated by particle swarm optimization method. Finally this model is applied to the prediction of water quality. The prediction results show that this method is effective and applicable.