Fault Diagnosis Analysis of Distribution Network Based on Compound Statistics of Residual and Score

It is a key tool for reducing distribution network failure loss by identifying the fault cause promptly and correctly and eliminating the fault quickly. By examining whether the Q and T2 statistics exceed the control limit, the fault diagnostic method based on Principal Component Analysis (PCA) may determine whether a fault occurs and the source of the fault by using the contribution value of Q and $\mathbf{T}^{2}$. However, the outcome of this procedure is ambiguous, resulting in low diagnostic accuracy. To simplify diagnosis chores and enhance diagnosis accuracy, this research suggested a PCA fault diagnosis approach based on Compound Statistics of Residual and Score (CRS), which uses Q and T2 Statistics to produce CRS Statistics and CRS contribution value. Finally, an IEEE33 node distribution network model is created for simulation verification, and the results validate the PCA approach for defect identification based on CRS data.