基于最小二乘向量机结合双向时序长短期记忆的台区用电特征提取

Power Consumption Feature Extraction Based on Least Square Vector Machine and Bidirectional Time Series Long Short-Term Memory

为了实现泛在电力物联网精细化管理,实现电网安全经济的调控,需要精确把握台区用户用电特征。对此,基于最小二乘向量机结合双向时序长短期记忆网络方法,提出了台区用户用电特征提取方法。首先,给出了最小二乘支持向量机回归模型的计算方法;在此基础上,提出了双向时序长短期记忆的最小二乘支持向量机回归模型,并以此作为边缘计算模型,将台区天气环境信息、人文活动信息、经济社会状态作为用电特征的输入数据,提取台区用电特征;最后,以实际电网台区为例,对所提方法进行验证。实验表明了该方法能够有效提取用电特征。

In order to realize the refined management of ubiquitous power Internet of things and realize the regulation of power grid security and economy, it is necessary to accurately grasp the power consumption characteristics of users in the substation area. Therefore, based on least square vector machine and bidirectional time series long-term and short-term memory network method, a power consumption feature extraction method is proposed. Firstly, the calculation method of least squares support vector machine regression model is given. On this basis, the least squares support vector machine regression model with bidirectional time series long-term and short-term memory is proposed, which is used as the edge calculation model. The weather environment information, human activities information, economic and social status of the station area are taken as the input data of power consumption characteristics, and the power consumption characteristics of the station area are extracted. Finally, taking the actual power grid as an example, the proposed method is verified, which shows that the method can effectively extract the power consumption characteristics.