The rapid advancement of aerospace technology has imposed stringent requirements on displacement resolution accuracy for flight control actuator axes. To address signal quality degradation caused by noise interference in actuator-axis displacement sensors, this study proposes a hybrid denoising method integrating improved Wavelet Threshold Denoising (WTD) with Savitzky-Golay (SG) smoothing filters. The methodology employs a three-stage decomposition strategy: (1) Multi-scale decomposition of raw displacement signals through wavelet transform to extract high-frequency detail coefficients and low-frequency approximation coefficients; (2) Implementation of SG filtering for local smoothing of high-frequency coefficients, coupled with a piecewise adaptive threshold function designed to mitigate oscillation artifacts and amplitude deviations inherent in conventional hard/soft thresholding approaches; (3) Signal reconstruction via inverse wavelet transform using optimized coefficients to balance noise suppression and feature preservation. Experimental results demonstrate that the WTD-SG method achieves a remarkable SNR enhancement from 1.7626 to 14.0646 dB, along with a 98.5% reduction in RMSE (0.5931 → 0.0205). Compared with state-of-art denoising methods, the proposed methodology achieves superior noise suppression while preserving essential displacement characteristics in actuator shaft signals.