Active Power Dispatching Strategies For Power System With Large-Scale Wind Power Intergration

The development of wind power promote the optimization of China’s energy structure, energy saving and emission reduction of power system. Large scale wind power integrated to power system has increased the uncertainty of power system operation, the uncertainty and the poor predictability characteristics of wind power challenge the traditional active power optimal dispatch based on the load forecasting. This dissertation firstly analyzes the influence of wind power uncertainty on the system active power dispatching, and then in-depth analyzes the main bottleneck of active power dispatching in wind power integrating system, and furthermore we study the peak load regulation ability and the optimal active power dispatch model and method in different time scales etc.The uncertainty of wind power based on the maximum entropy principle is quantified and analyzed. The main bottleneck restricting wind power integration from the system active power dispatch is discovered, and system peak load regulation capacity assessment is introduced to the multi-time scale scheduling framework. In this dissertation, some northeast wind farm historical data is taken as the researching object, and the maximum entropy theory is taken to analyze the uncertainty of wind power, and by comparing the day output characteristic of wind power and the typical daily load curve, the influence mechanism of anti-peak characteristics of wind power on system peak load regulation is analyzed, and the power system is lack of down peak load regulation capacity and unit response capacity during the system valley load periods which is the main bottleneck for acceptance of wind power is pointed out, and from the perspective of day-ahead active power dispatch the influence of wind power uncertainty on dispatch cost is analyzed. By utilizing the characteristic of the shorter time scale is the higher prediction accuracy, a multi-grade and multi-time scale active power optimization framework is designed. This framework is for the peak regulation ability calculation and analysis, the day-ahead active dispatch, and coordination between day-ahead and real-time dispatch system based on the wind power integration system, and the technical route is given clearly.The static and dynamic models evaluating system peak load regulation capacity are researched, and the calculation and analysis method is proposed considering wind power fluctuation extreme scenario on peak load regulation. This dissertation analyzes the influence of wind power fluctuation extreme scenario on peak load regulation, the influences of peak load regulation capacity and the unit ramp rate on the system peak load regulation are both considered, and a peak load capacity computing model and method in valley period for wind power integration system is put forward. In this dissertation, a 0-1 integer programming model is proposed to solve the model, the object of which is the maximum of the down adjustment space during the valley load period in the study periods, and the minimum outage time and the minimum operating time of the normal units, and the climbing rate constraint are all considered. The extreme scenario influenced by the wind power fluctuation is also considered. By solving the proposed optimization model the limitation of the peak load regulation for some wind power fluctuation range in studied periods and relative unit commitment. The calculation results of the model can provide the necessary reference information for the wind power integration system.The system dispatch cost by the wind power uncertainty is analyzed to build the wind power uncertainty opportunity cost model which is used in the day-ahead optimal dispatch decision. In this dissertation, based on the existing wind power error analysis and wind power statistical analysis, the maximum entropy theory is used to evaluate the wind power probability density function, and an opportunity cost model reflecting the uncertainty of the wind power is established. The wind power cost model is introduced to the object function of the day-ahead unit commitment model by dispersing the wind power probability distribution, and then the 0-1 mixture integer programming model of day-ahead unit commitment model is set up in wind power integration system. Different to the traditional active power dispatch method in wind power integration system, the proposed model and method can take the wind power as an active decision variable, and the amount of the wind power depends on the opportunity cost and the related system operation technical constraints brought by the wind power uncertainty.To realize the multi-time scale coordination active power dispatch, in this dissertation by use of model predictive control theory method the ultra-short time dispatch model in a day is built which can satisfy the wind power total acceptation and wind power curtail minimization. With model predictive control method by each level detailing and multi-level coordination rule the intra-day ultra-short term dispatch structure is built. For the difference of the influence of wind power fluctuation on the economy and the safety two rolling optimal model which are wind power total acceptation and wind power curtail minimization. The intra-day active power optimal dispatch strategy in this dissertation considers coordination of the system operation economy and safety. By solving the multi-period optimal problem in the prospective period, the initial strategy is provided to the real-time dispatching and the real dispatching result and the new forecasting information are taken as the feeding information to execute rolling optimal dispatch. This strategy can link the day-ahead dispatching time scale and the real-time dispatching time scale to deal with the influence of predictive information fluctuation on system dispatching.