A Classification Identification Method Based on Phasor Measurement for Distribution Line Parameter Identification Under Insufficient Measurements Conditions

Due to the limited quantity of phasor measurement units (PMUs) in power distribution systems, the measurement data cannot meet the observability requirements. Thus, traditional methods cannot identify the line parameters under these circumstances. According to the time-invariant characteristic of distribution line parameters in a short period, a classification identification method based on phasor measurement (CIMPM) is proposed for distribution line parameter identification (DLPI) under the condition of insufficient PMU measurements. We use the ability of extracting the main features of a large number of multitime measurements via a convolutional neural network (CNN). The proposed method obtains the estimated line parameters through classifying line parameters and extracting the features of the PMU measurements. Owing to lack of measurements, not all lines can be identified by this method, but some of them can satisfy the conditions of the proposed method. Furthermore, the application conditions for the proposed method with insufficient measurements are analyzed. Finally, the effectiveness of the proposed method is verified by simulation analyses of IEEE 33-bus and IEEE 69-bus systems.