The Study On Real-time Optimization Of Water Diversion Project Scheme In Lowland River Based On The Varation Of Dissolved Oxugen

Hydraulic regulation and control for water diversion project is of great significance to protect and improve the ecological environment of the urban river network by accelerate the water circulation in the river network and improve the self-purification capacity of water with the water conservancy facilities.The waterways of urban river network in lowland area are dense and the velocity of water in the river network is quite insufficient which contributed to the serious deposition of sediment.On the premise of meeting the requirement of improving water quality,the cost of water diversion project could be saved by the real-time optimal regulation of the sluice and pump system under different water quality conditions.For this purpose,this study paid attention on the variation of the water quality with the hydraulic condition varying and the optimization of the real-time online technology.The main research contents and research conclusions are listed as follows:(1)The study area was located in the lowland area of Jiaxing.It was a river network covering an area of about six square kilometers which was surrounded by the water conservancy facilities.This study concluded the variation of the water quality after the water diversion and the response and influencing factors of water quality to the hydraulic condition varying.With the hydraulic condition improving,dissolved oxygen(DO)in the water as the primary indicator of water quality increased rapidly.The improvement of DO concentration in the water is related to temperature by statistical analysis of monitoring data during years.The DO concentration declined more slowly in winter than summer.And there was an exponential relation between the long of maintain time in water quality and initial saturation of DO.Therefore,it was of great significance to optimize the real-time regulation of sluice and pump system based on the water quality response mechanism.(2)This study found a computational model for DO which was suitable for heterotrophic rivers in lowland area based on the DO balance equation and empirical formula for the coefficient of aeration.Then the parameters in the DO balance equation including the coefficient of aeration,photosynthesis rate and integrated respiration could be calculated by the extreme value method.(3)To solve the contradiction between the real-time scheme generation of the water diversion project and the long time taking by the water quality model of the river network,this study created a real-time optimization technique based on the offline database for the lowland rivers.This database consisted of 1000 sets of data which included the data of water quality(DO,NH3-N,NO3-N and TP)and corresponding hydraulic data(q,t)computing by the accurate model.Then a BP neural network could be trained by those data.A trained BP neural network could be used as a computational model to get the water quality data with different conditions.The optimal scheme of the water diversion project could be real-time by using BP model.It took only less than 20s to calculate a general condition by using the BP model.On the contrary,it took 34min by using the business software.With the increasing complexity of the model,the efficiency of the model increased significantly and the accuracy maintained.This model did not need the connection between self-programmed optimization algorithm program and commercial software and saved a lot of computation time.It could replace the previous method by manual trial algorithm to determine the scheme of the water diversion project because that it could realize the real-time optimization of the scheme.(4)This study established a real-time optimization platform for hydraulic control in the study area which covering an area of about six square kilometers.It was a conclusion that the water quality of rivers in this area had been greatly improved.The average concentration of DO improved 53.12%?172.09%,the NH3-N improvement rate was 37.57%?69.83%,and the COD(Mn)improved 75.64%?82.07%.The real-time optimization for hydraulic control in this study could be used with the online monitoring system in rivers to realize the real-time optimization and the effect feedback of the hydraulic control scheme.It has an excellent application prospect with the demand of water pollution control and the development of smart water management system in big cities.