Security Constrained Economic Dispatch Of Power System With The New Energy

It is a significant approach to deal with the crisis of fossil energy and curb the emission of air pollution by exploiting and utilizing renewable energy, which rapidly develops in recent years and occupies a growing proportion in the total generation of power grid. As to the property of intermittency and fluctuation of wind or light, the safety and reliability will be challenged. It is well known that the tolerance ability of fluctuant source can be improved by implementing economic dispatch for the safe, reliable and economic operation of power grid. Thus, it is of vital importance to research the security constrained economic dispatch ?SCED? problem considering the fluctuant power source, which has a realistic meaning to the power system at present.This paper mainly researches the wind power as the representative of renewable energy.This paper makes a research on SCED combining the technology of the prediction of the wind power, aiming at the SCED in the transmission network with wind power. the main contributions are shown as follows:Firstly, according to the new demand of dispatching problem-energy saving and emission reduction, we introduce the constrain of carbon emission to the traditional SCED model and simplify the nonlinear and complex constraints as well, The sensitivity factor method is applied to solve the dynamic changes of net flow during the dispatching process. Simulation has been implemented to compare the deviation between GSDF ?Generation Shifted Distribution Factor?and quasi-steady-state GSDF while solving the line power flow, and to choose a more precise method as the solving method to handle the power flow in dispatching process.Secondly, in order to accurately grasp the power fluctuations in the system as much as possible, so as to lay the foundation for the subsequent system scheduling research, then this paper make research on the wind power prediction and load forecasting. At first, a forecasting system that combines the short-term prediction and the ultra-short term prediction is developed in this study, through the summary and analysis of domestic and overseas wind power prediction results. Then, this paper analyzes the related factors affecting wind power. A BP neural network method is used based on the characteristic of wind power prediction, choosing a wind farm of certain area in California America as the testing objective. Considering that the power load also has certain fluctuations, this paper also makes a forecasting on the load in system using the BP neural network in order to obtain accurate power fluctuations in system.Thirdly, based on the wind power sequence and power load sequence got though the prediction, we research the SCED problem in electrical power system. Because the total operation time considered is short in the ultra-short term scheduling, it is impossible to consider the start-stop of large unit fully and reasonably, so this article adopts the method of the double-layer scheduling,that is to say,at first this paper develops the start-stop plan for the units though day-ahead scheduling, according to the condition of the power fluctuation in system, and makes the ultra-short term precise scheduling according to the result of the day-ahead scheduling.This paper introduces the method of the day-ahead scheduling and the ultra-short term scheduling firstly, then according to the different demand of computational accuracy and speed of the two methods,the PSO and mixed programming algorithm is proposed to solve this two issues,and the feasibility of the double-layer scheduling method and the solution tools is tested on IEEE_RTS₉6 test system.At last,aiming at the actual scheduling,this paper makes a research on the rolling scheduling and SCED problem with the implementation of double layer rolling scheduling in the modified single regional IEEE 24 Bus test system, which has been proved effectiveness and feasibility with the outcome of simulation.