Prediction Of Water Quality Of The Shuangyu River In The Deep Bay Area Of Hong Kong Based On Data Mining

In recent years,with the rapid development of the national economy,water pollution has gradually become one of the important issues affecting human life and development.It is of great significance to predict water quality data which provide data support for the effective estimation of water quality,and an indirect way to protect the water resources environment.This paper takes the dissolved oxygen,ammonia nitrogen and total phosphorus data of the main polluted water quality indicators from 1986 to 2016,total of 340 months in the three stations of the Shuangyu River Basin in the Deep Bay area of Hong Kong as the research object,predicts the water quality of the Shuangyu River and establishes a prediction model.Considering that the indicators of water quality prediction in the process of inquiry contain many nonlinear correlation characteristics,a combined water quality prediction model based on complementary integrated empirical mode decomposition and gray wolf optimization algorithm is proposed to explore the multi-faceted exploration.The empirical analysis shows that the combined model proposed in this paper shows strong superiority in many aspects such as practicability and prediction accuracy.In the process of water quality prediction,the paper firstly uses the integrated empirical mode decomposition method to solve the influence of white noise generated by the original signal on the final prediction result,and then optimizes the parameters of SVR through GWO.The empirical analysis shows that the proposed combination model overcomes the difficulty of premature convergence compared with thetraditional model,and the gray wolf optimization algorithm is more efficient than other optimization algorithms in the parameter optimization process of support vector regression.