Study On Optimal Operation Of The Reservoir On Sediment-Laden River

Development of social economy and the relative shortage of water resources put forward more and more demands for utilization of the reservoir. However, the controlling the use of methods of the reservoir on Sediment-Laden River is more complex than the general river reservoir because of sedimentation problems. In addition to considering the fundamental role of reservoirs, such as:flood control, irrigation, water supply, power generation, even more important to solve the reservoir sedimentation and other objectives of the relationship to achieve multi-objective comprehensive utilization and maximize comprehensive benefits of reservoir. In this paper, some problems related to the optimal operation of Sanmenxia Reservoir on Yellow River, including the runoff and sediment characteristics in recent years, the rapid calculation and prediction of the reservoir sediment, optimization of the reservoir in non-flood season and flood season, were discussed and researched. The main work of this paper are summarized as follows:(1) The changed characteristics of Sanmenxia Reservoir runoff and sediment and the reasons for the change since the late'80s are analyzed systematically based on historical statistical data.(2) This paper calculate and predicte sediment of the Sanmenxia Reservoir by the new non-linear machine learning methods—Support vector machines(SVM). The results are accurate and enable managers to make judgments timely and accurate for the reservoir sediment erosion process on a variety of conditions of runoff and sediment and the water level of reservoir.(3) Considering the actual runoff and sediment characteristics of Sanmenxia Reservoir in non-flood season, the paper establish the model for optimization dispatch of the Sanmenxia Reservoir in non-flood season, which is taken the dynamic programming as a theoretical basis and the model is simulated by using historical data. The results are satisfactory and achieve the effect of optimization, also meet the needs of the actual reservoir operation.(4) Basing on comprehensive consideration for sedimentation and power generation, the paper establish the model for the multi-object optimization dispatch of the Sanmenxia Reservoir in flood season by Multi-objective Optimization theory and give the simulation results of various programs under different the weight by linear weight method which converse the multi-objective problem into single objective problem. From the results, the model simulate the actual running status of the reservoir commendably and increase the sediment volume as well as generated energy, also meet the needs of the actual reservoir operation. Meanwhile, the reservoir sedimentation calculation model based on SVM is embedded into the model of optimization dispatch of the reservoir for sediment deposition and it's a new attempt for the joint operation of water and sediment in the reservoir.