基于CEEMDAN与反射系数修正的电缆输入阻抗谱故障诊断方法OA

To address the challenges of inaccurate fault identification of fault types and low localisation precision in cable fault location,this paper proposed a novel diagnostic method integrating complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)and reflection coefficient correction.Firstly,a spectrum model of the input impedance at the cable's initial end was established to derive the propagation coefficient and reflection coefficient of the faulty cable.Secondly,An attenuation factor was incorporated to correct the oscillation amplitude of the reflection coefficient,an impedance integral transform model was constructed.The impedance spectrum was then decomposed into amplitude and phase spectra to identify the fault type.Subsequently,cable data was collected and cleaned.The impedance was decomposed into multiple intrinsic mode functions(IMFs)using CEEMDAN,with false components eliminated via fuzzy entropy(FE).Finally,the reconstructed data were input into the kernel function to obtain localization results.The effectiveness of the method was verified through simulation experiments.The results demonstrate that the average fault localization error of the proposed method is controlled at approximately 0.60%,with a minimum error as low as 0.28%.Compared with simulation results with the experimental results without data preprocessing,the proposed approach decreases localization errors by 2.48 percentage points and maintains strong robustness and anti-interference capability under severe noise conditions.The proposed method effectively improves the accuracy of fault type identification,fault location,robustness and anti-interference performance under severe noise conditions,providing reliable theoretical basis and practical technical solution for precise fault identification and high-accuracy localization in power cable fault diagnosis scenarios.