基于改进神经网络的农村电力系统短期负荷预测

Rural Short-term Load Forecasting of Power System Based on Improved Neural Network

为进一步提高农村电力系统短期负荷预测模型的性能,实现准确与快速预测农村电力系统短期负荷的目的,采用基于优化理论的Levenberg-Marquardt算法来改进传统的BP算法,并构造电力系统负荷预测模型。结果表明,基于L-M算法的神经网络预测模型具有较高的预测精度,在农村电力系统短期负荷预测方面具有较高的使用价值。

In order to improve capacity of rural short-term load forecasting mode/of power system and make short term load forecasting more accurate and fast, Levenberg-Marquardt algorithm based on optimization theory was adopted to improve the traditional BP algorithm, and the power system load forecasting model was constructed. The results showed that the neural network forecasting model based on L-M algorithm has higher prediction accuracy and a high value in the rural power system short-term load forecasting.