Method for Extracting Impact Signals in Falling Weight Deflectometer Calibration Based on Frequency Filtering and Gradient Detection

FWD is an important non-destructive testing instrument in the field of highways. It evaluates the pavement bearing capacity by continuously hammering the ground. However, due to noise interference, the current identification and extraction of the impact signals generated by the hammering are not accurate enough, which affects the calibration accuracy of the FWD results. To address this issue, this work proposes a novel method for impact point identification. The method integrates frequency domain filtering with gradient detection. Firstly, by analyzing the frequency domain characteristics of FWD impact signals using fast Fourier transform (FFT) and short-time Fourier transform (STFT), the primary response frequency band of the impact was identified. Subsequently, the impact signal segment was reconstructed using inverse fast Fourier transform (IFFT) to effectively suppress noise interference. Furthermore, gradient detection was employed to precisely determine the initiation moment of the impact. To validate the proposed method, a simulated acceleration signal incorporating interference noise was constructed. Comparative experiments were also conducted between traditional identification methods and the proposed method under high-noise conditions. The results demonstrate that the proposed method can accurately identify the impact point even under strong noise, thereby providing reliable data support for FWD measurements. This method exhibits strong environmental adaptability and can be extended to other engineering tests involving impact events and impact point identification.