Research On Optimal Dispatch Of Hydropower Stations Considering The Uncertainty Of Incoming

With the large-scale planning,construction,and commissioning of water conservancy and hydropower projects in major river basins in my country,especially the construction and commissioning of the upper Yangtze cascade reservoir group with the Three Gorges Water Conservancy and Hydropower Project as the core,the research focus of my country’s hydropower industry has gradually changed from planning.The construction will shift to the safe and economic operation and management of hydropower stations.Among the many factors that affect the operation of hydropower stations,the hydrological runoff sequence with strong uncertainty and nonlinear characteristics has a particularly prominent impact on the safe and economic operation of hydropower stations.The hydrological runoff sequence contains uncertain features such as randomness and chaos,coupled with the impact of upstream reservoir storage and discharge regulation and the lack of information sharing mechanisms between hydropower stations,the uncertainty of runoff has further increased.The existing runoff forecasting methods are not enough to provide high-precision,the longforeseeable runoff forecast information restricts the full utilization of the power generation benefits of hydropower stations and is not conducive to the safe and economic operation of hydropower stations.Therefore,how to accurately describe the uncertain characteristics of the inflow of the reservoir,and quantify the impact of the uncertain inflow on the optimal dispatch of hydropower stations,and construct a hydropower generation optimal dispatch model that takes into account the benefits of power generation and the risks of undergeneration and over-generation,is an engineering problem faced by the research of optimal dispatching of hydropower stations.This thesis takes the Three Gorges Reservoir as the research object,focuses on the uncertain runoff impacts faced by the hydropower station during the safe operation of the hydropower station,studies the station runoff distribution law,and proposes a better applicable random runoff feature description method;on this basis,establishes the optimal dispatching model of hydropower generation under the influence of the uncertainty of incoming water considering the risk provides theoretical support and decision-making basis for the safe operation of hydropower stations.The main work content and innovative results of this paper are as follows:(1)In order to explore the uncertainty distribution of the runoff process and study the random simulation method of the runoff sequence,this paper introduces the Gaussian Mixture Model(GMM)to establish the joint distribution of the runoff sequence;and then builds the daily runoff stochastic simulation model based on the Gaussian mixture distribution,using the Expectation Maximization(EM)algorithm optimizes the model parameters;the stochastic simulation model is driven by the historical runoff data of Yichang Hydrological Station to simulate the long series of runoff data.The results show that,compared with the historical measured data,the correlation coefficient,mean value,standard deviation,skewness coefficient and other commonly used evaluation indicators of the simulated runoff series meet the accuracy requirements,indicating that the daily runoff stochastic simulation model used to describe the uncertainty runoff is rational and effective.(2)In order to describe the uncertainty distribution characteristics of the inflow runoff sequence of hydropower stations,the multi-stage runoff scenario set suitable for the stochastic power generation dispatch model is studied.This paper introduces the scenario analysis method to characterize the uncertainty characteristics of the runoff time sequence;In view of the low efficiency of the traditional scenario reduction method,which is not suitable for large-scale initial scenario set reduction,the K-Means clustering method based on improved mixed metrics is used to reduce the initial runoff scenario set to obtain a typical daily runoff scenario.The results show that the average deviation of the typical daily runoff scenario obtained by this method and the historical runoff data is within 2.0%,and the standard deviation is 5.4%,which retains the main characteristics of the historical runoff data,verifying that the method is used to characterize the inbound Validity and applicability of runoff uncertainty features(3)Aiming at the risk of power generation dispatch caused by the uncertainty of the incoming water of hydropower stations,the semi-variance risk measurement model in the field of economics is introduced to construct a semi-variance risk power generation optimal dispatch model.Taking the Three Gorges Power Station as a research example,a short-term power generation dispatching model with weekly dispatching period and daily dispatching period is established.The solution is based on stochastic dynamic programming.The results show that the model can achieve a balance between the power generation benefits and power generation risks of hydropower stations.On this basis,it analyzes the sensitivity of hydropower plants’ over-generating risks,under-generating risks and power-generating benefits to the risk aversion coefficient.By drawing power-generating benefits and powergenerating risk contour maps,it explores the optimal balance of power-generating benefits that can comprehensively measure different power-generation risks.The interval provides a theoretical basis and decision-making reference for the preparation of the actual power generation plan of the hydropower station.