Study On Economic Dispatch Of Power System For Accommodating Large Scale Wind Power

Wind power is renewable energy, the development of wind power is conducive to improving the energy structure. Because wind power has the characteristics of non-control, stochastic volatility, and anti-peak regulation, the traditional economic dispatch method of power system will be not applicable when the large-scale wind power integration on power system.Based on research of economic dispatch/scheduling of power system with wind power connected at home and abroad, this paper points out that advanced dispatch technology is the foundation and guarantee for accommodating the wind power fully. Aiming at the problem of large scale wind power connected, this paper improves the existing economic dispatch/scheduling method from the following three aspects: Firstly, the multi-timeframe dispatch/scheduling model is established, and the influence of prediction error of wind power on economic dispatch/scheduling reduced step by step. Secondly, by using the advanced dispatch/scheduling technology to describe wind power output in each time frame, it realizes maximum the integrated benefits of the dispatch/scheduling. Thirdly, considering various controllable loads, this article set up a source-load coordination dispatch model which can improve the wind power utilization. Specific tasks of this paper are as follows:First, the robust optimization theory and the basic principle of dual interior point method, genetic algorithm and particle swarm algorithm were proposed, which laid the foundation for the following chapters.Secondly, a multi-timeframe robust dispatch/scheduling was proposed for accommodating the large-scale wind power. On the one hand, by means of the multi-timeframe dispatch/scheduling method(day-ahead scheduling, intraday dispatch, and real-time dispatch), the influence of prediction error of wind power on economic dispatch/scheduling reduced step by step; on the other hand, the robust scheduling method is used to satisfy the security of power system when pursuing the economy of power system dispatch/scheduling. According to the case of a provincial power system, the robust dispatch/scheduling approach presented in this paper can be looked as a more accurate way to consider the spinning reserve, and it also can guarantee security of transmission.Again, the relationship between economy and robustness of robust dispatch/scheduling is a conflict. In order to solve this conflict, we converted the robustness into risk cost by calculation of the wind power curtailment cost and the load shedding cost. Then took the composite cost(the sum of generation cost and risk cost) as the objective function, and the uncertainty set optimization model was set up, which could be solved by a two-level optimization algorithm. The inner optimization is the process for solving the robust dispatch/scheduling model using internal point method. Outer optimization is the process for searching for the optimal uncertain set using hybrid particle swarm algorithm. The proposed method solved the selection problem of uncertain set in robust dispatch/scheduling, and an example analysis of 10 machine 39-bus system is carried out.Finally, based on conventional unit commitment model, a source-load coordination dispatch model was set up for improving the wind power utilization. It included the interruptible load and the transferable load. The result shows: the source-load coordination dispatch can make the peak valley difference decreased by 389 MW, and it also can make the maximum acceptance ratio of wind power up from 11% to 21%.