基于小波变换的火车车轮扁疤信号能量分析

火车车轮扁疤是火车行车事故的重大隐患之一，了解扁疤信号的能量分布情况对正确检测车轮扁疤有重要的意义．针对车轮扁疤信号具有持续时间短、突变快等特点，提出了一种运用小波变换的多分辨率分析和小波能量谱对扁疤信号进行分析的方法，通过比较扁疤信号在不同尺度下的能量密度，找出能量与各频段之间的对应关系，得出了扁疤信号的能量主要集中在2500Hz以下的频带范围内，在实际的车轮扁疤检测算法中只需分析频带范围为2500Hz以下的扁疤信号即可，通过实验验证了该能量分析方法的有效性．

The energy analysis of wheel-flat signal based on wavelet transform

Wheel-flat is one of the major hidden dangers, which can cause traffic accident of trains, so fully understanding the energy distribution of the wheel-flat signal is of great significance to detecting the wheel-fiat accurately. Based on the characters of short duration and abrupt change of wheel-fiat signal, the multiresolution analysis method of wavelet and wavelet energy spectrum are applied to analyze the wheel-flat signal. By comparing the energy densities of wheel-flat signals in different scales, the relationship between the value of energy and different frequency bands is acquired, and then a conclusion is drawn that the energies of wheel-fiat signals are mainly distributed in the frequency bands below 2 500 Hz, so just the wheel-flat signals below 2 500 Hz need to be analyzed in the actual wheel-fiat detection algorithm. The detection verifies the validity of this analysis.