基于Rao-Blackwellised粒子滤波的岩石声发射信号去噪

针对声发射数据采集过程易受随机噪声干扰，提出了Rao—Blackwellised粒子滤波的信号降噪方法。在地震波模型的基础上，建立了声发射信号的状态空间模型。由于两者在某些方面有相通性，从而把降噪问题转为在状态空间模型下的滤波问题。首先，通过对构造信号的仿真实验，验证Rao—Black—wellised粒子滤波能够用于信号滤波；其次，利用Rao—Blackwellised粒子滤波，对石灰岩声发射信号进行降噪处理，并取得较好的滤波效果，从而为处理声发射数据提供一条新途径。

Acoustic emission signals denoised with Rao-Blackwellised particle filter in rock

As the acoustic emission data measurement is easily subjected to random noise contamination, Rao-Blackwellised particle filter based noise removal method is put forward for signals in dissertation. Because of the similarity between seismic signals and acoustic emission signals in some respects, the state space model of acoustic emission sig- nals according to seismic wavelet model is established, thus denoising problems are turned into filtering problems under the state space model by the simulation of the synthetic signal.The results show that the RBPF can be applied to signal filtering. Acoustic emission signals denoised with RBPF in limestone can be as a good effectiveness of filtering, which can provide a new way for dealing with acoustic emission data.