Analysis Of Short-Term Load Forecasting Based On Artificial Neural Network

Load forecasting is one of the important tasks of the power enterprise, and the level of load forecasting has become one of the significant symbol which shows whether the management of a power enterprise has achieved modernization. With the economic and social development, high accuracy of short-term load forecasting wil be more and more important.This paper first analyzes the characteristics of the load, and the principles of load analysis should be followed. Various types of data of January to August of the last year and the year before last which affect load characteristics are collected for statistics and analysis to describe the load characteristics of large cities.The second part describes the basic theory of artificial neural networks(ANN), which focus on the BP neural network structure, algorithm and improved algorithm, and genetic algorithm is proposed to optimize the BP network. The simulation results show that the optimized BP network prediction results are more accurate. In the third part, the establishment of the concept mapping database related factors is proposed, and optimization strategy is proposed for mapping database. Example shows that the trained mapping values of quantitative factors are more reasonable , and the predicted effect and stability has been further improved. Then it studys short-term load forecasting( STLF) based on BP neural network. Comparing the simulation of the input sample which takes into account relevant factors and the input sample which doesn't take into account relevant factors, the simulation results suggest the establishment of a database significantly help to improve the accuracy of the short-term load forecasting. Last, a neural network madel based on difference degree of weather is proposed. Numerical example shows that load forecasting is more accurate when the model is used for severe weather. The last part introduces a short-term load forecasting system development ideas based on VBA and PI database. After conducting a detailed investigation in a district in the East China, the load characteristics of the region and limitations of the original loadforecasting system used in the region is detailed analyzed, a short-term load forecasting system framework is developed. The application shows that the system meets the requirements of load forecasting.