To address the challenges of high algorithmic complexity and implementation difficulty when identifying arcing in high-speed trains using current and voltage signals, as well as the high cost of arc recognition through high-speed camera imaging, a novel method based on port impedance is proposed for pantograph-catenary arc detection. First, the influence of airflow velocity under actual operating conditions on the arc voltage gradient is considered, and the voltage gradient is optimized to establish an arc voltage model that more accurately reflects real train operation conditions. Then, to facilitate more accurate simulation, the pantograph-catenary system is equivalently modeled as a two-port network consisting of system impedance and arc impedance under conditions of minimal load fluctuation, and a port impedance model incorporating airflow effects is developed. Subsequently, the feasibility of the optimized arc model is analyzed using the PSCAD/EMTDC simulation platform and a constructed experimental pantograph-catenary arc simulation setup. A support vector machine (SVM) arc recognition model optimized by the dream optimization algorithm is employed to identify the arc, thereby validating the feasibility of port impedance-based arc recognition. Finally, the port impedance-based arc recognition method is comprehensively compared with three traditional methods—current signal, voltage signal, and arc image recognition—across six evaluation criteria to verify its practicality and cost-effectiveness.