城市燃气管网日负荷预测的灰色神经网络模型

将灰色预测理论和人工神经网络理论结合起来,利用灰色静态预测模型来弱化数据的随机性并建立规律的累加数据,再利用神经网络模型来解决数据的非线性,建立了既反映其时间序列的周期性变化趋势,又包括天气、气温等影响因素的燃气日负荷预测灰色神经网络模型。对哈尔滨市燃气管网系统的日燃气用量进行了预测,表明模型不仅有较高的收敛速度和精度,同时也具有较强的适应性和灵活性。

Grey-neural networks model for city gas network daily load forecast

Applying the theory of Grey forecast to artificial neural network, the randomness of the data is weakened and the disciplinary accumulated data is developed using static Grey forecast model, and the lineari- ty of the data is then solved with the model of neural network. The city gas network daily load forecast model established reflect the variation trend with periodicity of time serial, the weather and temperature influence fac- tors. The results obtained with typical examples show that the model built has better convergence and forecast preciseness, better applicability and flexibility than those of other models.