基于人工神经网络的城市煤气短期负荷预测

Applying BP Artificial Neural Network to Forecast Urban Gas Short-term Load

应用人工神经网络误差反向传播模型对城市煤气管网的短期负荷进行了预测。根据城市煤气短期负荷变化的特性建立了既反映煤气负荷连续性、周期性及其变化趋势 ,又包含天气、气温、节假日等因素影响的短期负荷预测模型。

According to characteristics of short-term load changes for urban gas,this paper sets up a new model for urban gas load based on the BP artificial neural network.This model reflected both time sequence periodical trend and nonlinear affection factor such as weather, temperature and holiday.Through improving algorithm of BP neural network,we forecast day and hour gas load of Harbin gas network system. The result of forecasting system showed that it is feasible and efficient and could apply in practice.