天气数据模糊化处理对神经网络短期负荷预测模型的影响研究

介绍了一种简洁实用的基于模糊集的神经网络电力系统短期负荷预报方法，计及了天气和日期特征量．应用模糊集理论将天气和日期特征量模糊化后作为BP神经网络的一部分输入进行训练，同时考虑实际历史负荷数据构造了短期负荷预测模型，预测未来24h负荷．通过典型算例与普通BP方法预测结果相比，表明该方法是有效的并具有较高的预测精度．

Research on Impact of Fuzzy Processing of Weather Data on Neural Network Model of Short-term Load Forecasting

A simple and practical short-term load forecasting approach is presented by using artificial neural network based on fuzzy set and considering weather variables and date variables. Fuzzy processing to some sensitive factors such as weather variables and date variables was finished using fuzzy set technology. Processed data were used to train BP neural network as a part of network inputs. Taking into account the historical load data, the forecasting model is constructed to forecast hourly loads for the next 24 hours. The comparison between typical example and usual forecasting result by BP model indicates that this method is effective and of high accuracy.