基于改进粒子滤波的电力线载波通信语音增强

Speech Enhancement in Power Line Carrier Communication Based-on Improved Particle Filtering

针对噪声和混响环境下的电力线载波机语音通信系统聆听上的困难,研究了为实现多路电力线载波机复接器所采用的粒子滤波的语音增强算法,提出了一种基于粒子群优化的改进粒子滤波算法,它将语音增强问题转换为从带噪语音中对纯净语音的估计过程,引入粒子群优化的方法来产生建议分布,使降噪结果更接近纯净语音,从而得到更好的语音增强效果。

Since it＇s hard for the users of power line carrier communication system to listen in the noise and reverberation environment,a particle filtering speech enhancement algorithm to implement the multi-channel power line carrier multiplexer is researched and an improved particle filtering algorithm based on particle swarm optimization is designed. It converts the speech enhancement problem to the estimation procedure of pure speech from the speech with noise. The particle swarm optimization method is proposed to generate the proposal distribution and make the noise reduction result closer to the pure voice,thus it will obtain better speech enhancement effect.