Etection Algorithm Research Of Railway Signal Frequency Based On All Phase FFT

Measuring the frequency of railway signal in real-time and with high precision is very important to the train running safety and drawing train working curve graph. The detections of baseband frequency and side frequency are the main measure object in the detection of railway signal. It has become the main research contents in train control system. China railway signal has two kinds of modes. For the compatibility, it needs the algorithm to be able to detect the baseband frequency and the side frequency. For the detection in real-time and with high precision, the algorithm needs to be easy to realize with low complexity. But it is difficult to detect the frequency of railway signal with high precision, especially in the detection of side frequency.There are the following methods for the detection of frequency of railway signal. Detection method based on under sampling technique can improve the frequency resolution of the FFT transform. But the drawback is that the sampled waveform is distorted. This will affect the accuracy of measurement. The other method is based on digital orthogonal I&Q binary channels and resample algorithm. But it is not compatible with the two main China Railway Systems. The current detection method based on FFT has the following disadvantages. This algorithm is very complex because the side frequency detection does not use the results of baseband frequency detection. And it uses different spectrum correction methods. The side frequency uses the phase invariance to judge the boundary. Its anti-noise ability is very low and is prone to error of judgment. The detection of baseband frequency and side frequency are completely separate in all of the methods mentioned above.All these methods have the problems of low compatibility and the detection of baseband frequency and side frequency is running separately. To solve these two problems, this paper puts forward a method that can combine the detection of baseband frequency with side frequency. In order to improve detection precision, this paper uses all phase FFT to reduce spectrum leakage. In order to simplify the complexity of the algorithm, the same frequency correction method is applied to both baseband frequency and side frequency detection. This reduces the complexity of the algorithm and is conductive to real-time detection. The extraction algorithm is added in order to ensure the signal frequency detection of baseband frequency and side frequency at the same sampling frequency. This method can meet the two modes of railway signals and it has the higher practical application value.The difficulty of the side frequency detection is to judge the boundary point. A new algorithm to judge the boundary is put forward to solve this problem. By using baseband frequency extracting sequence, the side frequency boundary algorithm of railway signal is designed according to the baseband frequency signal amplitude jump. Then the stable side frequency sampling sequence is calculated according to the results of baseband frequency detection. Compared with the phase invariance algorithm, the improved algorithm of boundary determination reduces one FFT transformation. The complexity of the algorithm is significantly reduced. And anti-noise performance is only affected by baseband frequency amplitude. The performance of the boundary identification is excellent.Finally, the railway signal is simulated on Matlab and Quartus software. Then the feasibility of the algorithm is verified. And the reason of influencing the frequency detection accuracy is analyzed. The experimentation proves that the algorithm can detect the baseband and side frequency accurately with high precision when SNR is8dB.