基于现场数据采集的配电网低压线损预测方法

Low Voltage Line Loss Prediction Method of Distribution Network Based on Field Data Acquisition

为加快配电网线损的预测时间,降低预测结果的误差,营造更加稳定、安全的预测环境,结合现场数据采集模式,提出一种配电网低压线损预测方法。设定多元负荷预测目标,布设双向长短期记忆网络(Bidirectional Long Short-Term Memory,BLSTM)交换预测节点,设计等效数据采集线损预测模型,采用等值电阻法实现线损预测。通过测试结果表明,所设计的现场数据采集低压线损预测组得出的预测时间相对较短,均匀控制在1 s以内,表明其在实际应用的过程中对于H配电网的低压线损预测速度更快、误差小、效果更佳,具有实际应用价值。

In order to speed up the prediction time of distribution network line loss,reduce the error of prediction results,and create a more stable and safe prediction environment,combined with the field data acquisition mode,a low-voltage line loss prediction method for distribution network is proposed.Set multiple load forecasting targets,deploy Bidirectional Long Short-Term Memory(BLSTM)exchange forecasting nodes,design equivalent data acquisition line loss forecasting model,and use equivalent resistance method to realize line loss forecasting.The test results show that the prediction time obtained by the field data collection low-voltage line loss prediction group designed in this paper is relatively short and uniformly controlled within 1 s,indicating that in the process of practical application,the low-voltage line loss prediction speed for H distribution network is faster,the error is small,the effect is better,and it has practical application value.