Research on Neural Network Model Based on Maximum Conditional Entropy and Its Application on Short Term Load Forecasting

摘要

With the development of electricity market and power system growth, short term load forecasting (STLF) is becoming more and more important In this paper, a neural network forecasting model based on maximum conditional entropy model is proposed for STLF. With power system growth and the increase in their complexity, many factors have become influential to load management. First, we introduced conditional entropy to select relevant ones from so many load influential factors. The second step is to establish the relationship between the load value and the selected influential factors. In this paper, we adopted Recursive Multi Layer Perceptron Neural Networks with back propagation momentum training algorithm to establish the forecasting model. Expetiments show that the presented model can obtained satisfacroty results.