Short-Term Load Forecasting In Power System Based On Fuzzy Neural Network

With the continuous development of China’s society and economy, construction and operation of the power system has made remarkable achievements, constantly moving towards modernization and intelligence,also increasing level of grid management. Among the numerous measurable indicators of management level about grid operation, the accuracy of the power system load forecasting is undoubtedly more important, and among these, the daily work of grid scheduling departments and short-term load forecasting accuracy are more closely linked with greater significance. There are many factors that affect the accuracy of short-term load forecasting,such as the trends of the region’s history load. It’s also affected by a number of non-load factors such as temperature、weather conditions、 operation mode and so on. The primary work done in short-term load forecasting in this paper is as follows:This paper describes the application meaning and analysis methods of power system load forecasting first, then the paper explains the general theory and meaning of artificial neural network (ANN) and fuzzy inference system(FIS).The third chapter makes a more in-depth and meticulous research on the basic theory of artificial neural network and establishes a neural network load forecasting model. The forecast model takes the history load data and temperature characteristics of the similar days into account,at the same time,it makes a series of pretreatment to the input load values,such as missing data repair, data smoothing and normalization disposal. Finally, the paper describes a short-term load forecasting system based on Adaptive Network based Fuzzy Inference System(ANFIS),and trains and simulates the system with actual load data to forecast short-term power load. From the view of forecast effect, the adaptive network fuzzy inference system model proposed is better than the traditional BP network on performance, not only accelerating the speed of network training and learning, but also complying better with the requirements of practical application.The load forecasting model based on ANFIS not only paves a new way for power system load forecast, but also provides a new idea for its applications. Therefore, the research has the theory significance and the practical value.