Abstract: In response to the pain points of low efficiency and insufficient real-time performance of traditional lightning arrester detection methods, this study proposes an live intelligent detection system that integrates harmonic vector inversion and dual-mode communication. By deploying high-precision sensors at the lowvoltage end of the lightning arrester to capture key parameters such as leakage current and fundamental components of resistive current in real time, harmonic vector inversion method has been developed. This method extracts the third harmonic component based on the nonlinear characteristics of the lightning arrester, inverts the phase of the power frequency fundamental voltage, and achieves high-precision measurement of resistive current without PT reference signal. In terms of data transmission architecture, Bluetooth 5.0 and LoRa collaborative mechanism are adopted: near-field communication achieves real-time interaction between the tester and mobile terminals through low latency Bluetooth, while long-distance synchronization relies on LoRa adaptive spread spectrum technology, combined with layered encryption protocol to effectively resist strong electromagnetic interference in substations and ensure data security. Engineering verification shows that the system significantly reduces operation and maintenance costs by 70%, improves fault response speed by 85%, and provides reliable technical support for the construction of high resilience power