Aiming at the reliability challenges of mid-to-low-end UAV servos in strong electromagnetic pulse （EMP） environments， traditional static protection models employing simple technical superposition have limitations in adapting to dynamic strong electromagnetic environments. To overcome this limitation， this study innovatively proposes a “dynamic weight allocation—auxiliary fault-tolerant” collaborative protection architecture by introducing time-varying weight coefficient to achieve dynamic perception of EMP and adjust filtering and reconstruction strategies. First， considering that manufacturing defects and structural gaps in the material shielding layer can lead to attenuation of shielding effectiveness （SE）， the model of the silver/copper nanowire gradient coating was correspondingly revised based on experimental data inversion. Subsequently， an adaptive notch filtering algorithm based on dynamic filtering technology was employed to further suppress residual noise. Simulation and experimental results demonstrate that the proposed multimodal coupling method of material shielding， dynamic filtering， and dynamic reconstruction significantly enhances the electromagnetic interference protection of UAV servos， increasing the anti-interference performance from 31.5 dB with single shielding effectiveness to 93 dB.