摘要 为了减小低信噪比环境下时延估计误差,基于广义互相关法和谱减法,选择互功率谱相位加权法,提出了一种基于谱减法的互功率谱频域加权时延估计方法。通过仿真,分析了几种常见的加权函数随信噪比降低时的时延估计性能,以及改进的谱减法在不同信噪比环境下的去噪效果。对整体算法进行了联合仿真,结果表明:低信噪比环境下基于谱减法的互功率谱相位法,能有效锐化凸显互相关函数的主峰值,有效提高时延估计的精度。 In order to reduce the delay estimation error in low SNR environment,based on the generalized cross-correlation method and spectral subtraction,the phase weighting method of cross-power spectrum is selected,and a delay estimation method based on the frequency domain weighting method of cross-power spectrum combined with spectral subtraction is proposed.Through simulation,the time delay estimation performance of several common weighted functions with SNR reduction and the denoising effect of improved spectral subtraction under different SNR environments are analyzed.The results show that the spectral subtraction combined with the phase of the cross-power spectrum in low SNR environment can effectively sharpen and highlight the main peak value of the cross-correlation function and improve the accuracy of the delay estimation.
机构地区 西安工业大学光电工程学院
出处 《西安工业大学学报》 CAS 2021年第1期81-85,共5页 Journal of Xi’an Technological University
关键词 声源定位 时延估计 广义互相关 互功率谱 acoustic source localization time-delay estimation generalized cross correlation cross-power spectrum
