A Robust Line Parameter Identification Method Based on LSTM and Modified SCADA Data

Abstract

Transmission line parameters are the basis of power system calculation, and their accuracy directly affects the safe and stable operation of the system. At the same time, with the improvement of computing power and the rapid growth of the amount of power grid operation data, deep learning has developed rapidly and applied in power systems. However, there is little research on line parameter identification combined with deep learning. Therefore, from the perspective of line model and deep learning, combined with median estimation and modified Supervisory Control And Data Acquisition (SCADA) data based on line model, this paper proposes a robust method for parameter identification of transmission line based on Long Short-Term Memory (LSTM) and modified SCADA data. Speciffically, the overall process of the line parameter identification method based on LSTM is given first. Then the settings of the proposed method are introduced in detail. Combined with the line model, a modified input data based on the SCADA data is constructed, and a multi-case training set considering different operating conditions and different line parameters is established. According to a large number of tests, the optimized parameter configuration of the LSTM is given. Finally, the median estimation is used to give the parameter identification results. Case studies in simulated and measured data verify the effectiveness and robustness of the proposed method.