Due to the close relationship between power grid fault and impedance, there is a significant defect in the use of power line communication (PLC) equipment to monitor and locate power grid faults, which is the lack of real-time impedance information. Therefore, this study proposes a new fault monitoring method based on impedance estimation of power line communication equipment. The channel frequency response (CFR) obtained from the PLC receiver is used to estimate the high-frequency input impedance to monitor and locate the ground fault of the power grid in real time. Firstly, based on the multi-conductor transmission line theory, bottom-up channel modeling method and impedance estimation technology, the basic principles of fault monitoring and location methods are clarified. Then, modeling and analysis of cables in different situations are carried out, and the characteristics of the factors affecting the impedance in CFR are extracted by variational modal decomposition (VMD), and the estimation results of impedance are obtained by inversion analysis. Based on this, the impedance change and machine learning algorithm are used to track and identify the abnormal state of the power line to achieve high-sensitivity positioning of the cable fault. The simulation results show that the method has a good identification effect on high and low impedance cable fault, and has a better effect on low resistance fault monitoring. The fault location error is less than 2.13%, and has a good positioning accuracy.