小波包方法在车载FSK信号中的应用

FSK信号作为保障铁路安全运行的主要信号制式，在国内铁路上现在有两种，是法国引进的UT信号和国内自主开发的YP信号，小波变换是继傅里叶变换之后的重大突破，而小波包则是小波变换的进一步发展，克服了小波变换的一些不足，本文首先研究了车载FSK信号的特征，再利用小波包对车载FSK信号进行滤波处理，文中，给出了如何确定给定频率的信号在小波包分解树各个分解层中对应节点的算法，在滤波处理过程中，为了处理带内的噪声，也给出了采用阈值的方法来减少带内白噪声，阈值的选取充分应用到FSK信号的小波包分解的特点，最，我们给出了计算机产生的仿真FSK信号和现场采集的FSK信号的两种仿真，仿真结果表明，根据车载FSK信号的特性，小波包方法是处理车载FSK信号的有效方法。

Applications of wavelet packet theory on cab FSK signals

As the key signal system guaranteeing the safety for the railway, there are two types of FSK system used now by China railway: UT signal imported from France and the homemade YP signal. Wavelet transform is a big breakthrough after Fourier transforms, and the wavelet packet theory is the further development for wavelet transform, for it overcomes some shortcomings of wavelet transform. In this article, we study first the main characters of cab FSK signal, then the wavelet packets method is applied to the FSK signal filtering. We give an algorithm that determines the corresponding decomposition node for a signal with a pre-specified frequency in the decomposition tree. In order to reduce the in-band white noise in the denoising process, we give an algorithm based on threshold technology, and we also give a method to calculate the threshold that takes full advantage of the character of the wavelet packet decomposition of the FSK signal. In the last part of the article,a simulations on both computer simulation data and field data are given, and the simulation results show that the wavelet packet theory is an effective filtering method.