Abstract: Aiming at the problem that the noise components contained in rock acoustic emission (AE) signals are difficult to be filtered effectively,a denoising method based on variational mode decomposition (VMD) and sample entropy (SE) is proposed.Firstly,the acoustic emission signal of rock with noise is decomposed by VMD,and a series of intrinsic mode functions (IMF) with different center frequencies are obtained.Then,the sample entropy of each IMF component is calculated,and the IMF component is removed whose sample entropy is greater than the set threshold is regarded as the noise component,and the optimal component is retained.Finally,the optimal IMF component is reconstructed to obtain the denoising AE signal.The improved VMD denoising,EMD denoising and wavelet threshold denoising methods were used to analyze the denoising of rock acoustic emission signal with noise.The experimental results showed that the effect of EMD denoising was not ideal.Although the wavelet threshold denoising method could filter out the noise component in the signal,the signal waveform after denoising showed distortion.In contrast,the improved VMD denoising method proposed in this paper can better filter the noise components in the rock acoustic emission signal,which is helpful to extract the time-frequency characteristics of the signal and signal recognition.