在分析机车轮辋超声检测信号的基础上，构建了以数字滤波和包络检波为核心的数字处理技术。对比了有限冲激响应(FIR)和小波变换两种滤波算法的优缺点，选择采用小波变换法去除噪声，并基于三次样条插值法对去噪后的信号进行包络提取，设计了基于该组合算法的信号处理方案。经仿真验证，该方案具有辨识度高、抗干扰能力强和回波损失小的特点，能有效解决超声探伤信号的信噪比较低和回波辨识困难的问题。

With the analysis of locomotive wheels rims ultrasonic testing signals, the deeply integrated project of digital filter and digital demodulation is presented. By comparing advantages and disadvantages of finite impulse response and wavelet transform, the later one will be adopted to denoise and then the envelopes by the denoised signals will be extracted based on cubic spline interpolation. The signal processing method based on these combinatorial algorithms is designed. Simulation results illustrate that the method in this paper has the qualities of high distinguishing capability, good ability of anti-interference and low loss in echo signal. So it can effectively solve the problem of low signal-to-noise ratio and difficulty in echo recognition.