| Hydropower is a clean,low-carbon,flexible operation,and multiple comprehensive utilization of renewable energy.Its development and utilization play a very important role in my country’s energy development.The scientific overall planning of hydropower energy is a key measure to promote structural reforms on the energy supply side and build a safe and efficient energy system.Therefore,this paper takes the cascade power stations in the lower reaches of the Yalong River as the research object,and conducts research on optimal dispatching at various time scales.At the same time,it considers the nested coupling relationship of different time scale models to study the control mutual feed mechanism of reservoir dispatching under different time scales.The main research work and results are as follows:(1)In order to improve the power supply reliability of the power system and coordinate the power generation and capacity benefits of cascade hydropower stations,a medium and long-term power generation optimization dispatch model with the largest cascade power generation and the largest power-capacity comprehensive benefit is established.The results show that the two-year average power output of the cascade power station in the multi-objective model is 80.618 billion kWh,which is 0.19%less than the single-object model(80.773 billion kWh).But its minimum output during the period is 5.6099 million kW,which is increased by 9.66%compared to the single-object model(5.1155 million kW).Studies have shown that the multi-objective model can increase the minimum output during the dry season on the basis of ensuring the cascade power generation,thereby increasing the safety and reliability of the hydropower system operation.(2)In order to give full play to the peak-shaving capacity of hydropower units,around the problem of preparing short-term peak-shaving plans for cascade power stations in the lower reaches of the Yalong River,the "using water to determine electricity" mode and the synchronous peak-shaving strategy of cascade reservoirs are adopted to construct the model with the largest amount of power,the smallest residual load fluctuation based on the load trend and the largest power-peaking comprehensive benefit.The results show that the electricity benefit model only pursues the benefit of power generation,and fails to play the role of valley filling and peak shaving,and cannot give full play to the peak shaving advantages of hydropower stations.Compared with the power efficiency model,the standard deviation of the peak shaving model has decreased by 63.00%,72.15%and 64.43%on each typical day,and the peak-valley difference has decreased by 61.56%,70.47%and 62.15%.At the same time,its cascade power generation also decreased by 1.04%,1.06%and 15.07%.Although the peak shaving model can better track the trend of the grid load curve,its power generation benefit is reduced,which is not conducive to the full use of water resources.The standard deviation of the grid residual load of the multi-objective model on each typical day is respectively 1760MW,962MW and 1764MW,the peak-to-valley difference of the residual load is 5456MW,2955MW and 4078MW,respectively.The power generation of the cascade power stations is 123739MWh,109157MWh and 133746MWh.The research shows that the power efficiency and peak shaving efficiency of this model are all between the two single-objective models.The model guarantees the power generation of the cascade power station to a certain extent,and at the same time can better track the grid load trend,smooth the grid residual load fluctuation process,and be more conducive to the grid operation of new energy such as wind and solar in the future.(3)In order to solve the problems such as the disconnection between the scheduling models of various time scales and the inability to carry out effective information interaction,the nested coupling relationship of multiple time scales is analyzed,and the gradual constraint mechanism and the gradual feedback mechanism based on PID control principles are proposed.And then the multi-time-scale nested generation optimization scheduling model is established based on the control-feedback mechanism.According to the scheduling plan of the upper-level model,the boundary constraints of the lower-level model are generated,and then the deviation information of the real-time scheduling and the scheduling plan is transmitted to the upper-level model through the feedback mechanism,thereby forming a control mutual feed system for scheduling at various time scales.The simulation results of the TOPMODEL are used as input data,and the nested model is applied to the cascade reservoirs on the lower reaches of the Yalong River.The research results show that the operation process of the nested model in the short-time scale scheduling is closer to the actual scheduling process.In the non-feedback model,the sum of the absolute deviations between the guide water level of Jinping I and Ertan Hydropower Station and the actual water level during the remaining period of the representative month is 17.99m and 5.70m.The sum of the absolute deviations of each power station in the nested model considering the feedback mechanism is 3.56m and 2.71m,which are 80.21%and 52.46%decrease to the non-feedback model.In the long-term scheduling,the total power generation of the cascade power stations during the remaining period of the nested model is 115.448 billion kWh,which has an increase of 155 million kWh over the non-feedback model(115.293 billion kWh),an increase of 1.34‰,and the result of nested scheduling is more in line with the actual optimal decision.In summary,the nested model can feed back the scheduling information in real time while grasping the rationality of the global scheduling,so as to achieve dynamic control of the remaining period benefits,so that the revised scheduling decision is more in line with the actual operation process,and can play a more practical guiding role in the formulation of reservoir dispatching plans. |
PDF Full Text Request