Algorithm Study On Water Quality Monitoring And Evaluation In Hulan River Basin

Due to the rapid population growth and the rapid socio-economic development,pressures for water environmental protection have continued to increase.This project is mainly based on the pollution of the Hulan River water quality,and the establishment of water quality monitoring system is to achieve monitoring of important water quality factors,and the establishment of water quality assessment model is to analyze the status quo of the water body,and further objective data on the specific water quality factors and the scope of changes in the objective forecasting,and predicting future water quality conditions.The paper firstly builds the Hulan River water quality monitoring system based on the Internet of Things technology.The system mainly includes data acquisition module,sensor module,and automatic water distribution module,which can realize the collection and monitoring of water quality parameters such as chemical oxygen demand,p H,and ammonia nitrogen.Then,a water quality evaluation and prediction model is established by using the LS-SVM classification and regression algorithm that is very advantageous in solving high-dimensional nonlinearity,small sample size,and other issues.For the establishment of a water quality assessment model,a multi-class least squares support classification algorithm(LS-SVC)is used to learn and train the water quality data of the national surface water monitoring center,and an adaptive mutation particle swarm optimization algorithm and a cross validation principle are used to perform parameter optimization.The optimal multi-class LS-SVC model is used to evaluate the unknown water quality of the Hulan River.In order to establish a water quality prediction model,partial principal least squares is used to extract the main components of the water quality data and reduce their multiple correlations.Then,the least squares support regression(LS-SVR)is used for regression training to predict future water quality factors.Combine the fuzzy information granulation method with LS-SVR to establish a water quality time series model and realize the prediction of the trend changes of water quality data in the next three days.Finally,use the predicted value of the important water quality factors and their interval changes to achieve the prediction of future water quality levels and their interval changes,and then predict the future water quality conditions.Through field testing at the monitoring site,we can see that the accuracy and stability of the water quality monitoring system established in this project can meet the environmental requirements of the Hulan River monitoring station.According to the experimental simulation results,the accuracy of the water quality assessment model established in this paper can achieve an accuracy of 94% for the classification of unknown samples.The water quality prediction model established in this paper can realize the prediction of the specific values of important water quality factors and the changes in the next three days,and bases on the results to realize the prediction of future water quality grades.