Study On Optimal Dispatch And Market Trading Strategy Of Cascade Hydropower Stations

With the completion of large-scale cascade hydropower stations in southwest China,their operational status will have a marked impact on the safety,stability and operational benefits of power systems.The new round electric sector reform of China provides development opportunities for hydropower,whose operation model is also changed meanwhile,which poses a new challenge for optimal dispatching of cascade hydropower stations,ensuring consumption scale under the new situation,and improving the utilization efficiency of hydropower resources.Therefore,carrying out research on optimal dispatching of cascade hydropower stations and participating in market transaction methods is a necessary means to realize hydropower system dispatch management and improve comprehensive benefits.It's the primary task for ensuring the safety,stability,economy and environmental protection of power systems,and it's also the only way to realized the "save energy and reduce emissions" and "energy transformation" strategy for China.Based on the background above,the optimal scheduling model,the method of cascade hydropower stations,the problems in electricity market environment and the market trading strategies are deeply studied to provide theoretical and technical support for the sustainable development and efficient utilization of hydropower resources.The main research contents are as follows:(1)Research on multi-time scale optimal scheduling model and method for cascade hydropower stations.With the increasing number of cascade hydropower stations,the more and more serious problem of "dimensionality disaster",and the actual needs of constantly enriched running goals,the multi-time scale optimization scheduling model and method research are carried out.In the mid-term and long-term optimization scheduling research,taking the analysis of operational characteristics and spatial distribution patterns of each hydropower station as the entry point,the cascade hydropower system is decomposed into multiple sub-systems based on the large system decomposition and coordination idea.Then scheduling and coordination optimization regional-by-region are carried out.Accordingly the "dimension disaster"problem is improved from the perspective of optimization space reduction.Based on the parallel feature of dynamic programming method,Matlab multi-core cluster parallel computing platform is built for parallel computing to improve algorithm execution efficiency.In the research of day-ahead and real-time optimization dispatching,the counter-regulating hydro station is used to stabilize the discharge flow fluctuation,which provide the actual demand for the upstream hydropower station to participate in the deep peak-regulating for power grid.The coordinated scheduling model for the upstream peak-shaving power station and the downstream counter-regulation power station is established,including the day-ahead dispatching layer and real-time scheduling layer:in the day-ahead dispatching,the upstream power plant peak-regulating output is prioritized,and the sliding average filtering method is applied to smooth the unsteady peak-regulating flow.The real-time scheduling focuses on tracking and correcting the deviation between the real-time operation and the current plan to ensure the smooth execution of the day-ahead plan.(2)Research on joint optimization model and method of medium-to-long term output plan and market transaction for cascade hydropower stations.In view of the characteristics that hydropower station output is limited by runoff variation and reservoir regulation capability,and its market transaction needs to be closely integrated with operational scheduling to ensure that the transaction power is matched with actual power generation to avoid the problem of "water abandoning"and "power lacking".Two-tier decision-making framework for med-term and long-term scheduling and electricity optimization allocation is established.In the med-term and long-term scheduling,optimizing annual power generation plan provides reference for electricity allocation.In the electricity optimizing allocation,considering the volatility of price,uncertainty of real power generation caused by runoff randomness,the return risk is measureed by using the value risk model.The sequence operation theory is applied to merge the multi-random variable probabilistic sequence into the transaction portfolio profit probability sequence,so that the profit risk value is solved intuitively.On this basis,the impact of transmission capacity constraints on market transactions is further considered.The analysis of the risk on spot transactions and cross-price zone contract transactions arc focused on.Multi-period conditional value risk is adopted when scheduling dispatched electricity is optimally allocated to participate in multiple transactions at various time periods.Multi-period conditional value risk model is used to measure portfolio return risk and seek a reasonable compromise between return and risk(3)Study on the joint optimization maintenance scheduling and med-to-long term dispatching of cascade hydropower stations in power market environment.Aiming at the problem of tightly combining and interacting with the trading methods,dispatching plans and maintenance plans of cascade hydropower stations in the electricity market environment,a two-layer optimization model for med-to-long term dispatching and maintenance planning is established.As the outer layer optimization,The med-term and long-term scheduling carry out the decision-making according to runoff and price forecasting,and genetic algorithm is applied to optimize the output of cascade hydropower stations in each period.The maintenance plan is optimized for the inner layer,and the influence of hydraulic and electric power coupling relationship of the upper and lower hydropower stations on the maintenance plan is considered.The intermediate results of med-term and long-term dispatch optimization are the boundary conditions.The minimum maintenance loss is the optimization goal,including repair loss and outage risk loss.The maintenance loss and med-term and long-term power generation income are taken into total income as a function of genetic algorithm fitness,so as to realize the combination of medium and long-term power generation plan and maintenance plan.(4)Study on the day-ahead market clearing model for the multi-operator cascade hydropower stations participating and the self-scheduled bidding strategy of downstream power stations.For the cascade hydropower upstream and downstream power stations under different operating entities,each power station independently participates in the spot market bidding.The downstream hydropower station with weakly regulation faces an imbalance between the bided electricity and the actual power generation,which makes the transaction results difficult to implement and affects market stability and utilization of hydropower resources.In this paper,the market clearing model for multi-operator cascade hydropower stations participating is established.As a price acceptor,the downstream power station is expressed as a linear function of the declared power of the upstream power station according to the hydraulic and electric power coupling relationship between the upstream and downstream power stations,and is embedded in the clearing optimization model to realize the joint clearing.Furthermore,the self-scheduled bidding strategy of downstream power station is further studied.On the one hand,the profit space of downstream power station is expanded,and on the other hand,the effectiveness of above-mentioned market clearing mechanism is verified.For the downstream power station,the winning curve lags behind the upstream power station,and it is easy to cause problems of missing load and peak price.The output adjusting strategy of "pre peak release reservoir and post peak impound" is proposed for thickening revenue.For the large-scale fluctuation of downstream power station storage flow caused by rainfall change,the two stages output optimize and regulate strategy of "storage and peak shaving" is put forward to discharge water flow smoothly and alleviate water abandon.The downstream power station will report the output adjustment curve to market operation organization by self-scheduled bidding to realize increasing output and profit.The research results above can provide theoretical support for the development of real power generation planning optimization software and market transaction aided decision-making system of hydropower enterprises in China,and have application prospects in large-scale cascade hydropower system optimization scheduling.