Research On The Day-ahead Peak Shaving Optimal Operation For Cascade Hydropower Stations With Sensitive Hydraulic Connection

Peak shaving has always been a difficult problem for power system dispatching.Especially the large-scale operation of intermittent and random clean energy such as wind and sunshine has further increased the difficulty of peak shaving.Hydropower,as a high-quality adjustable power source,plays a vital role in ensuring the safety,stability and economic operation of power grid.Therefore,peak shaving of hydropower has become a research hotspot in the industry and academia.In recent years,with the improvement of refined requirements for reservoir water level,flow and output control in power grid,new challenges have emerged in the participation of hydropower in peak regulation of power grids,including the backwater jacking caused by sensitive hydraulic connection of cascade reservoirs,the uncertainty of peak regulation net load caused by intermittent energy access,and the weak practicability of model caused by multiple dispatching requirements.These dispatching problems and challenges are more complex in cascade reservoirs with sensitive hydraulic connections.For this reason,this paper takes Tianshengqiao cascade hydropower stations in Hongshui River Basin as the practical engineering background,and makes a deep research on the above problems existing in sensitive hydraulic connection cascade reservoirs.The main work and the theoretical and practical application results are as follows:(1)Aiming at the problem of backwater top support from downstream to upstream of cascade reservoir with sensitive hydraulic connection,a day-ahead peak regulation model of cascade hydropower stations considering the influence of backwater jacking is proposed in order to realize the fine control of dispatching.We build 96-point day-ahead peak shaving model,which takes the minimizing mean absolute anormaly of residual load as objective function,and linearized by SOS2 constraints method.And then solving it with MILP.An example of cascade dispatching of Tianshengqiao in Hongshui River Basin shows that better peak shaving effect can be achieved under the condition of satisfying the power constraints,and dispatching deviation caused by backwater jacking can be reduced.(2)Aiming at the uncertainty of the net load of hydropower peak-shaving caused by largescale intermittent clean energy grid-connected,and to avoid the peak-shaving deviation caused by the uncertainty of the net load of peak-shaving,a day-ahead regulation model considering the uncertainty of net load is proposed.The model uses Kmeans to generate wind power output scenarios to describe uncertainty,taking the minimizing expected mean absolute anormaly of residual load in multiple scenarios as objective function.An example of cascade dispatching of Tianshengqiao in Hongshui River Basin shows that the multi-scenario expected dispatching results are more robust than the single scenario,which reduces the risk of hydropower staggered peak shaving,and this advantage is more and more obvious with the increase of wind power permeability.(3)Aiming at the problem that the short-term optimal dispatching model is not universal and the search space is too discontinuous to solve,a practical dispatching method for cascade reservoirs with fuzzy identification in peak and valley period is proposed.Typical peak-shaving curves are used to arrange the peak-shaving output of hydropower stations.Typical peakshaving curves are created based on the division of peak-valley period of system load and peakto-valley output ratio.Finally,aiming at the problem of discarding water,a heuristic adjustment strategy of discarding water is put forward to realized the peaking shaving and minimizing discarded water.The idea of this discarding water adjustment strategy is to revise the output hydropower stations in the following order: from abandoned hydropower station to upstream reservoir in space and from abandoned water period forward in time.An example of cascade dispatching of Tianshengqiao in Hongshui River Basin shows that the proposed model can generate satisfactory solutions in a very short time without discarding water and satisfying all constraints,and has high application value.Finally,summarizing the whole paper and points out the research directions that need to be carried out in the next step.