【摘要】 分析了上海市电力月峰值负荷时间序列的特点 ,选择趋势分量和周期分量作为神经网络的输入变量 ,给出 BP神经网络辨识预测此类系统的一般方法和具体步骤。从预测的结果可以看出 ,人工神经网络辨识和预测具有趋势性及周期性波动时间序列的有效性。更多还原

【Abstract】 The characteristics of time sequence of the monthly peak load va1ue in Shanghai has been analysed and then the trend component and periodic component are choiced for the input variable of the BPneural net and the common method and steps for power load identification and prediction using methematical model established by BP algorithm are described. The predicted results show that the predictionand identification of the artificial neural net has the effectiveness of trending and periodic surge of timesequence.更多还原