The Short-term Power Load Forecasting Based On Improved Elman Neural Network

Power systerm load forecasting makes the electric power enterprises allot and arrangepower network operation mode better. Short-term load forecasting is mainly used forscheduling plan in advance, state estimation, electricity distribution and coordination,comprehensive economic unit quotation system, power plants quotation systerm, power plantoutput forecast and reasonable arrangement of the unit overhaul. Short-term load forecastinghas important meaning of the research.In recent years, neural network prediction algorithms have been widely welcomed, inthe electric power system short-term load forecasting of prediction method. The Elman neuralnetwork is a typical type of feedback neural network. This paper focuses on short-term powerload forecasting method based on the Elman neural network analysis. And the Elman neuralnetwork is improved, by putting an error feedback between the units of the neural networkoutput layer and input layer, and the error is the diference between the predicted values andthe actual value. First of all, this paper is mainly devoted to the analysis of the loadcharacteristic of network and preprocessing the data. Second, advantages and disadvantagesof the Elman neural network and the BP neural network are analyzed; Again, the improvedElman neural network for short-term power load forecasting is used; Finally, with the wavelettransform to preprocess the normalized data, the improved Elman neural network is used inpower systerm load forecasting. By using Nanjing power network data to verify, the improvedElman neural network prediction precision has high enhancement, it proves the effectivenessof the design plan.