基于Bi-LSTM的家庭用电量预测

电网智能化升级改造将传统电网与先进的信息、智能技术相融合,实现电力行业的根本性变革。智能电表是智能电网系统中收集用户用电信息的代表性边缘设备,当前智能电表收集的用电量数据存在维度低、波动性强等特征,造成对未来用电情况难以预测的问题;同时对于未来边缘设备端用电量的预测,其他相关特征信息的不可得,此时研究基于单变量特征的用电量预测至关重要。为此,提出一种基于双向长短期循环记忆循环神经网络(Bi-directional Long Short-Term Memory,Bi-LSTM)的单变量家庭用电量预测模型,Bi-LSTM模型能够充分利用上下文的信息实现更准确的预测效果。通过西班牙某市真实的智能电表数据对提出的模型进行了验证,实验结果表明,该模型的预测性能相比传统LSTM、SVM方法有进一步的提高。

Household Electricity Consumption Forecast Based on Bi-LSTM

Smart meters are representative edge devices for collecting user electricity information in smart grid systems.The power consumption data collected by current smart meters has the characteristics of low dimension and strong volatility,which makes it difficult to predict the future power consumption.At the same time,for the prediction of the power consumption of the edge devices in the future,other relevant feature information is not available.At this time,it is important to study the power consumption prediction based on the univariate characteristics.To this end,this paper proposes an univariate household electricity consumption prediction model based on two-way long short-term memory(Bi-directional Long Short-Term Memory,Bi-LSTM).The Bi-LSTM model can make full use of context information to achieve more accurate prediction results.