Research On Prediction And Assessment Of Surface Water Quality Based On Artificial Intelligence

At present,water quality forecast and evaluation research has made some achievements,the representative single item prediction and evaluation model has been widely used in practice.However,the analysis of the water quality of samples from different angles will lead to loss some important information.In addition,the water environment system is a complex dynamic system,the relevant water quality parameters have been in a dynamic change.But for the moment the water quality forecast and evaluation can not accurately reflect the overall situation of water quality,so it is necessary and urgent to study water quality prediction and evaluation deeply.According to the complexity of the background,in order to adapt to the dynamic characteristics of water quality changes,improve the prediction accuracy and better combine with the application of artificial intelligence algorithm in intelligent water quality modeling,this paper proposes a dynamic correction of grey forecast model and a dynamic time-varying exponential smoothing forecast model as the single item prediction models for water quality.The two forecast models are combined to establish a combination forecasting model,which is based on the effective measure of single item prediction models.The combined model can take advantage of each individual model,combine them with an appropriate weight,and update the individual models to adapt to the dynamic changes of water quality.In order to verify the validity of the model,the forecast and evaluation in this paper are based on the water quality data of a river in Jilin Province,it is used to forecast the water quality of five water quality parameters,including dissolved oxygen,permanganate index,ammonia nitrogen,total phosphorus and total nitrogen.The result of experiment shows that the proposed model is better than the single item prediction model,the development trend of the sample water quality is in good agreement with the predicted results curve.It also has a good practical value in water quality prediction.Upon the work above,the support vector machine are used to establish the water quality evaluation model.It also introduces the method of constructing multi-classifications by support vector machine,and optimizes the relevant parameters of support vector machine by using particle swarm optimization algorithm.The experimental result shows that the multi-classification based on support vector machine has higher classification accuracy in water quality evaluation,and the evaluation result is accurate,reliable and conforms to the objective reality.