Research On Peak-shaving Power Generation Plan And Inner-plant Economic Operation For Cascade Giant Hydropower Plant

As an economical and clean renewable energy source,hydropower can not only bear the base load and reduce the carbon emission level in the power system,but also frequently adjust peak and frequency due to its quick adjustment ability.With the completion and commissioning of a number of giant hydropower stations in China,cascade giant reservoir groups in the river basin are gradually formed.The cooperative operation of the reservoir group has exerted huge economic and social benefits in flood control,power generation,shipping,and water supply.However,the huge hydropower station has a large installed capacity,a large number of units and transmission lines,and many models.There are complex hydraulic and electrical connections between cascade power stations,which poses huge challenges for the preparation,execution,and economic operation of the cascade giant hydropower station.This paper focuses on the core issues of short-term optimal dispatch of cascade giant hydropower stations,combined with the actual operational needs of the project,from power generation plan preparation,plan execution and real-time forecast of plan execution situation to research,in-depth study of cascade giant hydropower station joint peak shaving solution feedback mechanism for load distribution among power plants,sub-plant,and units,and has constructed a water level prediction method for power stations based on data mining methods,which has realized the short-term fine scheduling of water resources for cascaded giant hydropower stations.The power generation benefits of cascade power stations.Part of the results of this paper have been integrated into the "Lower Reaches of Jinsha River ～ Three Gorges Cascade Water Resources Decision Support System" and successfully deployed and applied in the Yichang Yangtze River Electric Power Escalation Center.The main innovations of the research work are as follows:(1)To solving the contradiction between the peak load demand of the power grid and the benefits of cascade power generation,considering the factors such as the peak-shaving operation mode of the cascade power station,the time lag between the power stations and the water level jacking,a peak-shaped corridor peak shaving restriction method was proposed.The maximum power generation model which takes the benefits of peak shaving into account has been established.The average daily output of the power station is determined by the daily available water volume determined by mid and long-term dispatch plan,and the initial peak shape is formed according to the peak shape parameters issued by the power grid and combined with the peak load ratio range for peak-shaped corridors.An improved step-by-step optimization successive approximation(POA-DPSA)hybrid algorithm is used to solve the model within the corridor,and cascade power plant power generation plans that meet the peak shaving requirements are quickly generated.Case studies show that the model can meet the complex constraints of cascade power plants while taking the peak shaving demand and power generation benefits into account,and provide decision support for cascade dispatch centers.(2)Thinking of the characteristics of giant power plant with multiple branches,multiple units,and multiple outgoing sections,the economic operation method of giant hydropower plant considering the constraints of multiple branches is proposed.The load is distributed to the branch plant,and then the optimal load distribution table of the hydropower station is combined with a practical load adjustment strategy to quickly distribute the load of the branch plant to the single machine.Case studies show that the model can meet the complex constraints of economic operation and timeliness of solution in the giant power plant.(3)In order to improve the prediction accuracy of the daily water level process of the power station under peak shaving and frequency regulation,and ensure the safe and stable operation of the power station,using the historical monitoring data of the power station for many years as a training sample,a downstream water level prediction based on support vector machine The model is used as the downstream water level prediction knowledge base to embed the daily water level process prediction model of the power station,and the prediction accuracy of the daily water level process of the power station under peak shaving conditions is significantly increased.