多普勒信号的Burg算法优化研究

Study on Burg Algorithm Optimization of Doppler Signal

随着高速列车的发展,测速精度高的多普勒测速方式被广泛应用。其将多普勒信号看作广义平稳随机信号,利用功率谱估计方法进行信号处理。针对现代谱估计中Burg算法基本原理分析其误差来源。分析了两种优化算法,一种是基于窗函数的优化算法,另一种针对改进的协方差法实现了基于预测误差功率最小意义的优化算法。通过仿真对比原算法与两种优化算法的频谱估计性能,分析算法复杂度,选择模型最优参数,并验证了在列车测速范围内不同频段的适用性。结果表明,优化算法在不增加运行时间的基础上可降低频谱偏移程度、改善频谱分辨率,可识别出列车测速范围内各个频段频率,并且谱估计频率误差小于1%。

With the development of high-speed trains, the Doppler speed measurement method with high speed accuracy is widely used. The Doppler signal is treated as a generalized stationary random signal, and the power spectrum estimation method is used for signal processing. The error sources are analyzed based on the basic principle of Burg algorithm in modern spectral estimation. Two optimization algorithms are analyzed, one is the optimization algorithm based on window functions, and the other is based on an improved covariance method which achieves the optimization algorithm based on the minimum meaning of prediction error power. By comparing the spectrum estimation performance of the original algorithm and the two optimization algorithms through simulation, the complexity of the algorithm is analyzed, the optimal parameters of the model are selected, and the applicability of different frequency bands within the speed range of the train is verified. The results show that the optimization algorithm can reduce the degree of spectrum offset and improve the spectrum resolution without increasing the running time. It can identify the frequency of each band within the speed range of the train, and the spectrum estimation frequency error is less than 1%.