Research On Model Predictive Control Of Urban Drainage System

With the economic development, the industrialization and people's living standard of our country have remarkably improved. The increasing amount of discharged industrial wastewater and municipal sewage has highlighted the importance of the urban drainage system. However, there are large number of problems commonly exist in our country, such as the imperfect drainage facilities, lag of necessary construction on the pipe network, insufficient ability of sewage discharges, aging of elements of the pipe network, the uncertainties of pipe network underground, etc, which cause serious environment pollution, and obstruct the urban development.In relation to the problems such as, low degree of automation in pumping station and pipe network, especially low degree of intelligence, this paper shows a lot of research achievements on the optimization control strategy of pumping station and pipe network. On the basis of predictive control theory, predictive control is applied in urban drainage, and an overall energy-saving and optimization control strategy is presented for urban drainage system.First of all, we establish the hydraulics state-space model by choosing the flow as the state according to Saint - venant equations, Which is stemmed from the hydraulics of Drainage network, and then, we carried out further analysis of the flow of water, and change the state-space model into the uncertain discrete-time system mode for urban drainage system. Then analysis the stability of the water flow requirements of sewage treatment plants.Secondly, based on the hydraulic model and analysis the stability, predictive control is used in the urban drainage system. The design of control system for the pipe network is presented. Predictive control is imported in this method, the process of designing predictive controller and verifying using computer simulation is presented in detail. On the premise of making the utmost use of the system's water storage utilities, this control system could minimize overflows on certain area and energy wasting in pumping stations, and found the transformation from hand-act to automation for the real time control of combined sewer systems.Finally, discussing the problem of the energy-economized of pump. At first, the vector listtree database on the basis of urban GIS is built by analysising the flow of the pipe network, then the relation between the nodes of the drainage pipeline networks is illustrated through the hydraulics equation of Saint - Venant. At second, a functional relation between the flow and time is detailedly derived according to the theory of process analysed and least-squares regression, and then deduce the relation among the energy– consumed, flow and time. At last, the predicting model is built by the input data of pump power calculated and output data of pump power observed. Through predicting the flow of pump to the control the real operation, and combining with the feedback adjustment, the best energy– economized state of the pump station is reached.