论文部分内容阅读
针对高压脉冲(high-voltage impulse,HVI)轨道电路的特点,在仿真模型基础上分别对HVI信号使用最小二乘法(least square,LS)、滑动平均(moving average,MA)滤波及端部平滑的滑动平均滤波进行处理.为了消除“端部效应”,在对比处理结果之后,提出了一种综合LS与MA原理的改进型滤波算法.分别向仿真模型中加入随机噪声、周期性脉冲干扰和高斯白噪声后,对4种滤波方法进行了对比.改进型滤波算法在未引入庞大计算量的前提下较其它3种滤波方法提高了信噪比(signal noise ratio,SNR).最后,通过实际的HVI轨道电路实验平台验证,改进型滤波算法可靠有效,可应用于实际工程系统中进行HVI信号的识别和检测.“,”According to the properties of high-voltage impulse (HVI) track circuit, we use the least square (LS) method, the moving average (MA) filtering, and the MA filtering with endpoint smoothing on the basis of simulation model to smoothen the HVI signal. To eliminate the“end-effect, ”we propose a modified filtering method combining the MA filtering and LS method based on the comparison of the above three methods.Then, these four filtering methods are processed with a simulation model under random noise, periodic impulse interference, and Gaussian noise conditions. Compared with the other three methods, the modified filtering method improves the signal noise ratio (SNR), without requiring a large amount of calculation. Finally, the experimental results show that the modified filtering method is reliable and effective, and can be applied to identify and detect HVI signals in an HVI track circuit system.