Research For Power Systems Short-term Load Forecasting In Electricity Market

This thesis is divided into two major parts, the first part is electricity market, the second part carries for power system short-term load forecasting.In the 20 last years, the electric power industrial construction has started a meaning reform in the whole world includes China. They set out from the different purpose, proceed the reform with the different mode, but all incline to the same direction: in traditional field the monopoly has been broken and the competition hasbeen imported—electricity market. The electric power industry passes to decontrol,lead into the competition, and improve the economic performance of generation transmition and distribution the electricity. Along with our country socialism market economy system developments, the electricity market reformation of our country has been in the steady propulsion.Proceeding from modern economics and socialism market economy theories at first, the thesis researched the purpose and essence of the electricity market that our country is establishing, discussed and probed the operation mode of electricity market of China. On the other hand, the thesis analyses the problem that the market faced current and development directions in the future.Short-term load forecasting is an important mold piece of electricity market technique support system, and it has very important function in the reliable, security and economical operation of power system. The operation, dispatch and assignment of economic load of power system are all need accurate and fast load forecasting. In the operation of electricity market, the accurate load forecasting can help to arrange the startup and stopping of generation units economically and reasonably, maintain the safety and stabilization in power grid, arrange the plan of generation units to examine and repair reasonably. At the same time, generation enterprise can adjust the quote strategy reasonable according to the forecasting data, so as to cut the cost effectively. The power grid enterprise also can buy the electricity energy with the cheaper price in electricity market. Therefore, accurate and quick load forecasting can optimize the social resource.The principle, the origination and development, the general application ofpresent domestic and international power load forecasting has been introduced in this thesis. The research applies artificial neural network model to forecast regional power grid short-term load. In order to improve the prediction accuracy, according to the different regulation that carries, different forecasting models have been established respectively including workday model (Monday to Friday), off days model (Saturday, Sunday) and festival model. Models have many merits, such as little sample data required, simple and high accuracy; also they have stronger generalization ability.The thesis regards the power grid of Nanchang of Jiangxi province as the target of forecasting; apply the grouping artificial neural network to carry on the short-term load forecasting. The forecasting results show that the mode using for short-term load forecasting by artificial neural network is comparatively satisfied and can satisfy the request of actual production.