Abstract: To address heterogeneous calibration and risk misjudgment caused by mixed weighting of long-term aging and short-term environmental catalysis in partial discharge monitoring of transmission lines, a comprehensive evaluation strategy based on multi-source dimensionless processing and stress separation decoupling is proposed. An evaluation system covering equipment status, environmental severity and physical aging is constructed, and a dual-track calibration method is proposed to distinguish absolute electrical measurements from relative electromagnetic features. FAHP and entropy weight method are combined to calculate integrated weights, with a lower-limit protection mechanism introduced for slow variables. Furthermore, annual-scale thermal degradation and hourly-scale meteorological disturbance are separately modeled to construct a multi-stress correction model. Typical condition verification shows that, compared with fixed-threshold and uncorrected baseline models, the proposed model can identify latent risks under high humidity and high altitude, elevating relevant risk levels from “normal” to “attention” or higher. The hierarchical warning mechanism clarifies the safety boundary of auxiliary decision-making and provides quantitative support for predictive maintenance.