Methods For Long-term Inter-basin Complementary Optimal Operations Of Southwest Large Scale Hydropower Systems

In China, the hydropower is developing rapidly, and the hydropower installed capacity has became the highest in the world. A largescale of cascaded hydropower systems crossing basins have presented in southwest regions with abundant water resource. For the scheduling by power grids, operation of the cascaded hydropower system is difficult, since the dimension limitation, the uncertainty of runoff and the coordination involving multiple basins, etc. Hence, it is necessary to study models foroptimization of cascaded hydropower systems, which play a significant role in power operating, resources utilization. By effective optimization models, the final benefits for basins can be enhanced. This article studied the operation rules and a long-term optimization model forlarge-scale cascaded hydropower system, and takes the Lancang basin, Hongshuihe basin and Wujiang basin, etc., as research background. The major(1) To solve the power generation scheduling problem for cascade stations in dry seasons, this paper proposed a chance constrained optimal hedging rules. Based on the original reservoir operation rules which made by the stochastic dynamic programming, this model developed two hedging rules. The objectives are to maximize the minimum power generated by each station, considering the constraints of destruction rate and reliability requirements rate. Then a multi-objective chance constrained reservoir operation model is constructed. In the proposed model, aggregated function method is used to transform the min-max objective functions into continues function, and the chances constraints are addressed by penalty function. This model is solved by a genetic algorithm, in which a fuzzy optimization method is used to evaluate the performance of individual station. The results for the cascade stations located in Lancang river show that the proposed model is a feasible way for making or revising operation rules, and can significantly reduce power shortages.(2) There are many large hydropower stations located in southwestern China, crossing different basins with multiple stakeholders. Coordinating the interests of all stakeholders while operating the large-scale cascaded hydropower systems is necessary. Therefore, this paper proposed a distributed and centralized energy availability-based operating rules and an optimization method for multiple cascaded hydropower systems. The distributed operating rules take the generation capacity of each cascaded system as state variables, and decompose generation scheduling of cascade system to stations by a generation allocation model. For the centralized rule, the state variable is generation capacity of multiple cascaded systems, and power production scheduling is allocated across several cascaded reservoir systems. The operating rules are optimized to maximize the minimum power of the whole system. To avoid unacceptable profit losses to subsystems, the minimum energy produced by each cascaded system is constrained. Through this interchange of objective and constraint, the multi-objective problem can be avoided, and the possible profit earnings or losses can be controlled to acceptable levels. The optimization models are solved using a GA. A case study for three cascaded systems (i.e. Hongshuihe cascade, Lancang cascade and Wujiang cascade) in south-west China shows the effect of the proposed method in directing the cooperation and compensation of cascaded systems, and has guiding significance to actual operation.(3) Proposed an algorithm for solving the problem of large-scale multiple cascaded systems in long-term operation. Based on the dynamic programming, the discrete differential dynamic programing, the progressive optimization algorithm and dynamic programming with successive approximation method, sub-problems within two adjacent periods are solved by the depth-first or breadth-first method in the proposed algorithm, and then the problem dimension and solving difficulty are effectively reduced. Take the China Southern Power Grid as research background, this paper developed a dynamic searching method for multiple cascaded stations in long-term operation, and proposed different solving methods for problems with different sizes and characteristics. The study results show that the proposed algorithms can solve long-term scheduling problems with different size and different objective functions for multiple cascaded systems, and the strong robustness and practicality of this algorithm are also demonstrated. This study provides a strong theoretical approach to realize inter-basin hydropower systems'compensation scheduling and exert the comprehensive benefits of hydropower systems.Finally, summarized the whole paper and discussed the further study of aforementioned issues.