Research On The Flood Forecasting And Dispatching System Of Xinjiang Xiabandi Reservoir

The Taxkorgan river basin has the characteristics of complicated drainage, changeful topographical feature and obvious flood feature. The rainfall frequency of this river basin is lower, but the raininess is stronger, always companying with debris flow which can make big damages to people. Xinjiang Xiabandi hydro-pivot project is located in the middle reaches of Taxkorgan river, which serves as the backbone reservoir in cascaded development process.Hydrologic regime measurement and forecasting system is the important part of the National Flood Control and Command System (NFCCS). The nuclear research subject of this system is how to develop water resources decision-support system with minimum invest in nowadays water conservancy model of construction and management in China, which will show fine utility reliability, low cost high value, easy operation, little maintenance work and can realize forecast quickly and exactly.Based on previous research, this paper detailedly discusses the development procedure of hydrologic regime measurement and forecasting system associated with the current developing of communication, automatic technique and computer technique from aspects of base station arrangement,system integration, system design and communication mode; studies flood forecast system based on snowmelt runoff forecast model and BP Artificial Neural Network (ANN )model.The result indicates that the research content and scheme of measurement and forecasting system of Xiabandi Reservoir can meet certain standards, base station has proper arrangement, equipment type has advanced selecting, which basically satisfy automatic hydrologic regime measurement and forecasting and flood forecast demand of Xiabandi Reservoir.The flood forecast system mainly uses snowmelt runoff forecast model which can better reflect evolutionary mechanics of rainfall and runoff. By analyzing running result, we find that the snowmelt runoff forecast model can apply in flood forecast successfully. Meanwhile, for making further decision conveniently, the ANN model is firstly used in this paper to improve forecast precision through comparing with Xin' an jiang model, but the result still should be examined by practice.The flood dispatching software is developed with OPP development language Delphi 6.0 and the data base management system is developed with Access, which has the characteristics of powerful function, fine openness, friendly man-machine interface and cartoon effectiveness, flexible operation, convenient to fault analysis and stable running.Practically saying, the research and development of this system can relieve working strength of reservoir staff and may have the good economic benefits, socil benefits and environmental benefits.