Design And Application Of Simulation Control System For Water Treatment And Research On Water Quality Forecast

Tap water is the basis of national economic development and people's lives. It is one of the city's most important infrastructure. With the development of industrialization and urbanization in China,problems like water resources scarcity and demand increase, water dispersion,raw water diversification, water quality complex are becoming more and more serious. Traditional water treatment is meanwhile facing enormous challenges. It is of great significance to apply advanced computer, machine learning and data mining techniques to the production of tap water and realize the informationization of water production.The study Discuss the use of advanced computer science, data mining and artificial intelligence in tap water production processes from several aspects such as system design, industrial visualization and so on.In the study, we use a water plant in Beijing as an engineering object to analyze the water treatment process and combined with production needs. An innovative simulation control system for water treatment with B/S structure is designed and constructed. It separately realizes the dynamic demonstration of process flow, water control, equipment maintenance, operation analysis and other functions, which effectively improve the overall quality of water services. Platform design with high mobility, multi-platform availability, no need to install the client,scalability, and other advantages. The system has been on-line through the performance test, and was put into actual production use.Based on the data of water quality analysis of the platform, this paper uses the optimized neural network model to predict the chemical oxygen demand (COD) of raw water in the water plant. In the optimization model, the PSO-BP (Particle Swarm Optimization - Back Propagation) algorithm is used to improve the global search ability of the traditional BP neural network. Then,a dynamic inertia weight which is nonlinear with the iteration number is introduced to the PSO-BP algorithm and effectively improves the convergence speed of the optimization algorithm, and enhances the generalization ability. In this paper, the optimized neural network model is applied to the prediction of the main pollutant COD in a water plant in Beijing,and compared with the traditional BP neural network. Simulation results show that BP neural network based on dynamic inertia weight can effectively predict the raw water quality and reduce the forecast error. Therefore, it can better forecast the water quality of the future water plant.The main contribution and innovation of this paper are as follows:filling the gaps in the water plant without the simulation control system.The platform provides functions such as production process simulation,plant pipeline production data forecasting, production data analysis integration and visualization, plant structure fault repair simulation and load estimation, state correlation display, water monitoring and warning and so on. The BP neural network water quality prediction model based on particle swarm optimization (PSO) with dynamic inertia weight can predict the change trend of raw water quality of water plant more accurately, which provides theoretical and technical support for the establishment of water quality forecasting system...