基于改进型BP神经网络的短期电力负荷预测

Short-term load forecasting based on improved BP neural network

科学、准确的短期电力负荷预测有利于提高电力系统运行的经济性和安全性,向用户提供高质量的电力。提出一种基于改进型BP神经网络的短期负荷预测方法,并充分考虑建模时复杂气候敏感因数的影响,对输入样本的选取、预测模型的建立进行了论述。算例表明所提出方法具有较高的预测精度,负荷预测结果的相对误差小于3.63%。

Scientific and actual short-term load forecasting is benefit to improve electricity power system economically and safely,and provides high quality power energy. The paper brings forth a new type of method of short-term load forecasting based on BP neural network, takes complicative climate sensitive factor into account when establishing model. It discusses the selection of input sample and the construction of forecasting model. The example shows that the method is of high precision and the relative error is less than 3.63%.