摘要
有关天气的输入变量不确定性会造成较大的预测偏差 .新方法比使用多层感知器的网络有更准确的预测均值 ,这个新方法包括传统的训练后的神经网络和用来计算预测负荷的均值及置信区间的一组方程 .
The uncertainness of variable by the weather can lead to significant forecastaing errors. The new method provides a more accurate mean forecast than a multiplayer perceptron networks alone. The new method consists of traditionally trained neural networks and a set of equations to calculate the mean value and confidence intervals of the forecasted load.
出处
《山东大学学报（工学版）》
CAS
2004年第5期55-58,共4页
Journal of Shandong University（Engineering Science）
关键词
载荷分析
神经网络
均值
电力网络分析
预测
loading analysis
neural networks
auerage value
electric network analysis
Predictions
