Study On Methods Of Multi-objective Decision Making And Risk Analysis In Reservoir Operation

Hydropower is quality, clean, non-polluted and renewable energy. Strengthen the scientific management of hydropower is an important strategic choice to construct the resource-saving and environment-friendly society, also China’s "twelfth five-year" plan outline also places hydropower in very important position. In recent years, with the large-scale development and construction of hydropower stations and cascade reservoirs in our each large river basin, the risk management of reservoir operation is becoming particularly significant and it will also become the inevitable development trend of large river basin integrated management. The development and utilization of hydropower is a comprehensive, multidisciplinary, widely acknowledged and complex systems engineering, which involves a series of problems such as runoff forecasting, multi-objective decision making and risk analysis etc. Selected the (cascade) reservoir as the research object, research achievements of the paper is acquired by respectively making systematically researching on mid-term and long-term flood forecasting of reservoir, the methods of risk analysis of reservoir operation and multi-objective risk decision making etc, which are of great significance on completing the theory of risk management system for reservoir operation, improving the comprehensive utilization efficiency of water resources, and promoting the sustainable development of society, economy and ecological environment. The main achievements in the whole paper are as follows:(1) Research on risk analysis methods of reservoir operation. The randomness of inflow runoff is the important factors that affect reservoir operation, several stochastic simulation methods of inflow runoff is given in the paper by the example of monthly runoff series; Based on the summary of commonly used analysis risk methods of reservoir operation and uncertainty theory, the paper establishes several risk analysis modes of reservoir operation such as risk loss expectation mode, opportunity risk value model and minimization model of risk events measurement according to the practical characteristics of reservoir operation. Then combine with the practical feature of every model, the corresponding solving method is given.(2) Research on hierarchical structure combination method for reservoir mid-term and long-term runoff forecasting. Mid-term and long-term runoff forecasting is the important basis of reservoir operation for the mid-term and long-term planning. In view of the traditional reservoir runoff forecasting models which fail to fully consider forecast information of a single forecasting model and are lack of pretreatment of a single model and prediction results, taking prediction accuracy and prediction stability as the evaluation index, it established the hierarchical structure combination model for hydropower station reservoir runoff forecasting by means of sub step to determine the weight of each single model; The case study of Three gorges reservoir shows that the established hierarchical structure combination forecasting model can provide multi-model integrated forecast information for the incoming runoff of hydropower station reservoir.(3) Random fuzzy risk analysis on hydropower station forecasting generation operation. It analyzed the purposes, goals and principles of the hydropower station reservoir optimal operation, combined with forecast result of hierarchical structure combination forecasting model, and then established optimal forecasting generation operation model of hydropower station reservoir; Meanwhile, together with the influence that prediction errors possibly play on reservoir generation operation, it built the overall process framework of random fuzzy risk analysis on hydropower station reservoir forecasting generation operation; The case study of Three Gorges reservoir shows that random fuzzy risk analysis provides certain theoretical basis and reference value for reservoir operation management policy-makers to prevent and avoid risk.(4) Research on multi-objective risk decision model for cascade reservoirs joint operation. The joint operation of cascaded hydroelectric reservoirs which is affected by numerous factors involves many departments’requirements and regulation goal, so the decision making process is a multi-objective decision problem concerned on risk management. On the basis of analyzing the features of cascaded reservoirs and the requirements of multi-objective joint operation and the main risk factors in course of the operation, this thesis mainly discussed the influences of the uncertainty of reservoir inflow on flood control, electricity generation and shipping, and established risk decision model for cascade reservoirs multi-objective joint operation; Aiming at the multi-objective problem and complex constraints in the model solution, designed a hybrid intelligent algorithm coupling uncertainty simulation technology, genetic algorithm and compromise algorithm combined with stochastic simulation of reservoir inflow. The established model was able to compare and analyze the risk and benefits of cascaded reservoirs.(5) Research on integrated risk assessment of joint operation. Considering various regulation goals like flood control, electricity generation, shipping, water supply and ecological water, the multi-objective decision index system of cascaded reservoirs was established. On this basis, comprehensively analyzed various risk that existed in the multi-objective joint operation, set up integrated risk index system of multi-objective joint operation and multi-objective minimum risk model; Realizing ranking of numerous Pareto Optimal set by transforming risk type multi-objective problem contains random variable into scheme optimization problem based on minimum integrated risk by support vector machine and scheme evaluation nesting genetic algorithm.