Research Of Water Quality Prediction Based On Improved Intelligent Algorithm

In recent years,with the leapfrog development of China's society and the global economy,the serious pollution of water resources has become increasingly serious.The management and effective protection of water quality and ecological environment in various regions and basins is of great strategic significance for sustainable development.Correct projections for river water quality is an important means of protection of water resources,based on the San Francisco bay basin five stations from 2001 to 2015,the main water quality parameters dissolved oxygen(DO)and Salinity(Salinity),Temperature(Temperature)data as the research object,research on the San Francisco bay watershed water quality forecast.In order to improve water quality prediction method,an improved intelligent algorithm for water quality prediction is proposed.Experimental results verify that the proposed algorithm has certain advantages in applicability and error control.In this article,the artificial intelligence algorithm and data analysis technology are used to study the surface water quality data of San Francisco bay,analyze the factors that have a great impact on the surface water quality of the study area,and then build an effective water quality prediction model.This article mainly completes the following aspects of work.1.In view of the complexity of the research subject,the research status of water quality prediction at home and abroad was firstly analyzed.Through the data analysis in the early stage,some problems related to the prediction of surface water quality were found,which laid a foundation for the research work of this paper.2.Research work is carried out on different empirical mode decomposition methods,mainly including EMD method,EEMD method and CEEMD method.In this paper,the principles and experimental analysis of these empirical mode decomposition methods are introduced.3.Based on the statistical analysis of surface water quality data of San Francisco bay in the United States,the water quality data of the study area were analyzed from three aspects: dissolved oxygen,temperature and salinity.The statistical data shows that the total data volume of the five sampling sites is 7330 pieces,with an average value of 1466 pieces.The mean salinity was 4.63 ‰.The mean temperature is 16.16 ?;The average mass concentration of DO was 8.85 mg/L.4.Based on the surface water quality data of San Francisco bay in the United States,this paper established three intelligent algorithms,LSTM,SVR and LSSVM,to predict surface water quality,and prioritized the LSTM algorithm for improvement.5.Aiming at the research topic,this paper establishes the emd-lstm model,eemd-lstm model and ceemd-lstm model to predict the water quality in the research area,and compares and analyzes the prediction results of each model.The results show that the improved intelligent algorithm,eemd-lstm model,performs well in the prediction of water quality in the research area.