Desgin And Implementstion Of The System For Short Term Load Forcasting Of Electric Power

The current information technology with computer and communication technology as the representative of the production activities in modern enterprise obtained the widespread application, become the basis for long-term development of enterprises. With the rapid development of our economy, the power supply of electricity user accuracy and reliability requirements continue to improve, especially climate change and social emergencies, will put forward higher requirements on power short-term load. Short term load forecasting is an important type of power load forecasting, power load mainly according to the procedures to quasi real-time calculation of the next period of time, the results for dispatching electric energy production and consumption. It is one of important significance to the regulation of power market activities.This paper first describes the power short-term load system research background, industry research trends, the system function needs, and other user needs of the non functional analysis. The design system consists of data management, load analysis, load forecasting, reporting management, meteorological analysis, data collection, system management and so on. The current most popular browser / server(client / server) architecture and Java2 Enterprise Edition(J2EE) technology, from the architecture to ensure the system flexible scalability and portability. The system realizes the data interface, collects the information of the power supply and the information of the electricity information, automatically imports the information such as weather and festival, and can communicate with the dispatch system. The overall logic architecture of the system is composed of the display layer, application layer, service layer, interface layer, platform layer and resource layer. The display layer uses CSS3, Javascript, HTML5 and Ajax technology, application layer and service layer adopts J2 EE service structure, interface layer and platform layer application Oracle database. The use of Java language design to achieve historical data preprocessing, the establishment of a comprehensive forecasting model of the time, and increase the weather, holidays and other factors, improve the short-term load forecasting accuracy.After the implementation of the system, the performance tests, concurrency, reliability and other indicators of the design, greatly improve the short-term power load forecasting management capabilities. The max number of uses is 2000. Because of adopting the standard J2 EE distributed system, the new system can be easily implemented and the peripheral system is easy to expand. It provides a flexible operating platform for detailed data analysis and high quality demand forecast for power system departments.The system integrated the new electric power market and its forecast theory of outstanding research achievements can provide more accurate and reasonable electricity market demand forecasting results, thus optimal unit combination, economic dispatch, optimal power flow, power market provides important technical support. Through the implementation of the development of the system, the power short-term load forecast platform aided management mode will break through the traditional management pattern based on tables, and flow of, data query and maintenance, data mapping and analysis, statistical computing and graphical display closely, five layer architecture for power system short-term load forecasting system based on.