论文部分内容阅读
针对表面粗糙锻件的超声检测信号分别做了基于傅里叶变换的信号处理工作和基于小波变换的信号处理工作。其中基于傅里叶变换的信号处理主要包括频谱分析和FIR滤波,基于小波变换的信号处理主要包括时频分析和小波去噪。经FIR滤波后,表面粗糙锻件上平底孔缺陷信号的信噪比(SNR)与滤波前相比提高了3.9 dB,但是超声检测信号的均方根误差(RMSE)与滤波前相比减小了0.085;对信号做小波去噪处理后,平底孔缺陷信号的信噪比与去噪前相比提高了9.9 dB,而且信号的均方根误差与去噪前相比几乎没发生变化。通过实验对比发现,基于小波变换的信号处理技术可以对表面粗糙锻件的超声检测信号达到更好的去噪效果。
The signal processing work based on Fourier transform and the signal processing work based on wavelet transform are respectively carried out for the ultrasonic testing signals of the rough surface forging. Among them, the signal processing based on Fourier transform mainly includes spectrum analysis and FIR filtering, and the signal processing based on wavelet transform mainly includes time-frequency analysis and wavelet denoising. After FIR filtering, the signal-to-noise ratio (SNR) of the flat bottom hole defect signal on the surface roughness forging increased by 3.9 dB compared with that before filtering, but the root mean square error (RMSE) of the ultrasonic detection signal was reduced compared with that before filtering 0.085; After the signal is denoised by wavelet, the signal-to-noise ratio of flat-bottom hole defect signal is improved by 9.9 dB compared with that before denoising, and the root mean square error of signal is almost unchanged compared with that before denoising. The experimental comparison shows that the signal processing technology based on wavelet transform can achieve better denoising effect on the ultrasonic testing signals of the rough surface forgings.