Ultra-short Term Dispatching Strategies For The System With Large-scale Wind Power Intergration

Under large-scale wind power grid-connected conditions,the two-stage scheduling model coordinated by the current scheduling and AGC gradually shifts to daily scheduling,ultra-short-term scheduling,and coordinated multi-time-scale scheduling of AGC.At this stage,wind power forecasting accuracy is difficult to meet engineering needs.Using traditional scheduling strategies and models will increase the pressure of AGC threat system security.Ultra-short-term scheduling in wind power grid-connected systems is an important part of reducing planned deviations.In this paper,based on the analysis of wind power uncertainty and its impact on active dispatch,an ultra-short-term scheduling strategy for large-scale wind power integration is designed,thereby reducing the pressure on the AGC link and solving large-scale wind power to some extent.The problem of grid connection is difficult.Firstly,from the perspective of analyzing the uncertainty of wind power,the forecasting error of wind power is analyzed and described,revealing the time scale characteristics of wind power forecast error;At the same time,the impact of wind power uncertainty on active dispatch is analyzed,and the mechanism of unfavorable impact of wind power uncertainty on dispatching AGC is clarified;Comprehensive consideration of time scale,fully invoking crew resources and introducing AGC related constraint indicators,built a framework for ultra-short-term scheduling of large-scale wind power integration.Secondly,combined with wind power uncertainty analysis,a probabilistic active power optimization method based on model predictive control is proposed.The method introduces the scene probability optimization theory in the rolling optimization module of model predictive control;utilizes model predictive control methods and scene probability optimization to handle the advantages of uncertain factors,and reduces wind power uncertainty in the scheduling process in which wind power accuracy is not high.The effect of scheduling results;through a wind power grid-connected system,it is verified that the scenario probability optimization method based on model predictive control has strong robustness and traceability.Finally,a multi-time scale intra-day ultra-short-term scheduling strategy combined with intra-day rolling optimization and real-time optimization is proposed.The scheduling strategy is divided into two levels.The first level of the strategy is the four-hour rolling optimization link.The strategy introduces a rapid-start-and-depress group pressurization operation method,aiming at fully invoking crew resources,adopting a scenario probability optimization method based on model predictive control to solve the problem,aiming to reduce the impact of wind power uncertainty on the scheduling results;The second stage of the strategy is the 15 min real-time optimization link.The strategy introduces the AGC margin criterion and feasibility adjustment in order to maintain the safety of the system AGC link and avoid frequent start and stop of the unit.To test the validity of ultra-short-term scheduling strategy to adapt to large-scale wind power system by 10 units.