Fault Diagnosis analysis of PCA distribution network based on $\mathrm{T}^{2}$ contribution value

Improving the distribution network's short circuit fault diagnosis capability is a significant way to reduce diagnosis time and avert large-scale power outages. By assessing whether the data exceeds the control limit, the classic fault detection method based on Principal Components Analysis (PCA) can determine whether a fault occurs. This strategy, however, does not identify the source of the failure. As a result, when the incidence of the defect has been determined by statistics, the $\mathrm{T}^{2}$ contribution graph approach is utilized to determine the cause of the power system's deviating from the normal functioning condition. Finally, a simulation model of the IEEE33 node distribution network was created, and the results show that the revised PCA method is useful in defect diagnosis.