基于RLLMP与倒谱预滤波的二次相关时延估计算法

Quatic correlation delay estimation algorithm based on RLLMP and inverted spectral pre-filtering

本文针对单声源的远场检测定位问题,旨在提高低信噪比情况下时延的估计精度。传统的广义互相关算法在噪声和混响状态下性能不佳,为了提高时延估计精度,在对广义互相关算法进行分析以后,引入增强型递归最小平均功率算法(RLLMP)和倒谱法进行预滤波,提出基于RLLMP和倒谱预滤波的二次相关算法来提高时延估计精度。以传感器接收到的语音信号叠加高斯白噪声作为单点声源,固定混响时间,通过MATLAB仿真实验测试在不同信噪比情况下的时延估计精度。结果表明:信噪比在0 dB-10 dB范围内,本文提出的改进算法均方根误差与其他两种算法接近,而信噪比在-10 dB-0 dB的低信噪比阶段,本文提出算法的均方根误差比GCC算法低25.6%左右,比二次相关算法低10.2%左右。

In this paper,the far field single point sound source positioning aims to improve the accuracy of the state delay estimation with low signal to noise ratio(SNR).The generalized cross-correlation algorithm has poor performance under noise and reverberation.In order to improve the accuracy of time delay estimation,after analyzing the generalized cross-correlation algorithm,RLLMP and cepstrum are introduced for pre-filtering,and a quadratic correlation algorithm based on RLLMP and cepstrum pre-filtering is proposed to improve the accuracy of time delay estimation.The speech signal received by the sensor is superimposed with Gaussian white noise as a single point sound source,the reverberation time is fixed,and the delay estimation accuracy under different signal-to-noise ratio is tested by MATLAB simulation experiment.The results show that the root-mean-square error of the improved algorithm is close to that of the other two algorithms when the SNR is in the range of 0 dB-10 dB,and the root-mean-square error of the proposed algorithm is about 25.6%lower than that of GCC algorithm and about 10.2%lower than that of the quadratic correlation algorithm when the SNR is in the low SNR stage of-10 dB-0 dB.