欢迎访问《兵工学报》官方网站,今天是

兵工学报 ›› 2024, Vol. 45 ›› Issue (4): 1108-1116.doi: 10.12382/bgxb.2022.1078

• • 上一篇    下一篇

基于小波奇异特征约束的期望最大时延估计算法

朱小婷1,2, 张君3, 王璐2,4, 陈志菲2,*(), 鲍明2,3, 王翊1,5   

  1. 1 安徽大学, 安徽 合肥 230601
    2 中国科学院 声学研究所, 北京 100190
    3 西北工业大学 自动化学院, 陕西 西安 710129
    4 航天工程大学, 北京 101400
    5 华东光电集成器件研究所, 安徽 蚌埠 233000
  • 收稿日期:2022-11-21 上线日期:2024-04-30
  • 通讯作者:
  • 基金资助:
    国家自然科学基金项目(11774379); 安徽高校协同创新项目(GXXT-2022-080); 安徽高校协同创新项目(GXXT-2023-015)

Expectation Maximization Time Delay Estimation Algorithm Based on Wavelet Singular Feature Constraint

ZHU Xiaoting1,2, ZHANG Jun3, WANG Lu2,4, CHEN Zhifei2,*(), BAO Ming2,3, WANG Yi1,5   

  1. 1 Anhui University, Hefei 230601, Anhui, China
    2 Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, China
    3 School of Automation, Northwestern Polytechnical University, Xi'an 710129, Shaanxi, China
    4 Space Engineering University, Beijing 101400, China
    5 East China Institute of Optoelectronic Integrated Devices, Bengbu 233000, Anhui, China
  • Received:2022-11-21 Online:2024-04-30

摘要:

针对低信噪比条件下非平稳信号时延估计精度低的问题,提出基于小波奇异特征约束的期望最大时延估计算法。设计小波奇异性特征尺度广义互相关矩阵,构建多尺度小波奇异特征约束下的期望最大化模型。推导参数更新公式,利用期望最大化算法并行迭代,求取奇异性特征显著性最大条件下信号的自适应尺度以及该尺度下声源信号的最优时延估计值。仿真和实验结果表明,所提算法在低信噪比条件下,相较于传统广义互相关时延估计算法以及改进算法具有较高的时延估计精度,并且有效提高了误差约束范围内的有效估计成功率。

关键词: 低信噪比, 非平稳信号, 小波奇异性, 期望最大化, 时延估计

Abstract:

To improve the accuracy of time delay estimation for non-stationary signals under low signal-to-noise ratio (SNR), an expectation maximization time delay estimation algorithm based on wavelet singular feature constraint is presented. A generalized cross-correlation matrix of the wavelet singularity feature scale is designed, and an expectation maximization model under the constraints of multiscale wavelet singularity feature is constructed. The parameter update formula is deduced, and the expectation maximization algorithm is used to iterate in parallel to obtain the adaptive scale of the signal and the optimal time delay estimate of the sound source signal at the maximum singularity significance. The simulated and experimental results show that the proposed algorithm has higher accuracy of time delay estimation than the traditional generalized cross-correlation time delay estimation algorithm and the improved algorithm under low signal-to-noise ratio, and effectively improves the estimation success rate within the error constraints.

Key words: low signal-to-noise ratio, non-stationary signal, wavelet singularity, expectation maximization, time delay estimation

中图分类号: