Analysis And Application Of Urban Wastewater Discharge Forecast Model Based On Government Purchase Of Service

With the deepening of China’s urbanization process,environmental problems are becoming more and more prominent,especially the phenomenon of high energy consumption and low efficiency in urban sewage treatment,Some sewage treatment plant scale construction moderately ahead,but the annual growth of sewage volume is small,Sewage treatment capacity is not up to the design scale and the sewage treatment rate is low;The actual sewage treatment capacity of some sewage treatment plants built and operated in PPP mode can not reach the guaranteed sewage treatment capacity purchased by the government,which leads to the government’s sewage treatment cost higher than the actual income,resulting in useless expenditure up to tens of millions of yuan.Based on the this phenomenon,This paper takes Qin huangdao sewage plant in PPP project treatment as the research object,uses historical data to establish a according the prediction results predict the long-term sewage volume of sewage treatment plant,provides reference for timely increase and expansion of sewage treatment plant.cost expenditure.According to the design scale of each sewage plant and the actual sewage treatment situation,the water allocation scheme is provided to ensure that the sewage plant participating in PPP project is in the saturation operation state,to achieve the guaranteed water quantity purchased by the government,and to provide a reference for the government to maximize the benefit of fiscal expenditure in sewage PPP project procurement service.The research of the paper contains the following aspects:First,review the literature and collect the factors affecting the amount of sewage.Data were obtained from the website of Qinhuangdao Statistics Bureau,and principal component analysis was used to determine the main influencing factors.Set up social-economic-urban indicators as the input layer of the prediction model.In addition,,construct the forecast model of sewage quantity.Three types of prediction models are adopted in this paper,establish ARIMA model based on time series,GA-BP model based on artificial intelligence and LSTM model based on deep learning,The model with the highest prediction accuracy is determined by comprehensive comparison。Finally,Carry out sewage plant planning and construction and optimize PPP project purchase plan,Based on the established mathematical model,the long-term sewage volume of five sewage treatment plants in Qinhuangdao city was predicted.According to the forecast results,the water allocation scheme between seasonal urban sewage plants is designed from the three aspects of planning and construction,existing infrastructure and the operation of sewage PPP projects.The results show that.compared with ARIMA model and GA-BP model,LSTM model has higher accuracy,learning ability and robustness in water quantity prediction models.The fitting value is 0.93,the root mean square error is 1.6%,the mean absolute error is 2.2%,It can process prediction events with relatively long intervals in time series.LSTM model can be used to predict the amount of sewage and form a basic model of urban sewage optimization and adjustment.According to the calculation,the sewage northward and sewage west transfer schemes proposed can save 36.52% of the sewage treatment cost of the government’s purchase of the guaranteed water amount in the sewage PPP project every year.