Due to the complexity of mine environment,the error of monitoring sensors and the interference of magnetic field,the measured blasting vibration signal inevitably contains a lot of high-frequency noise.In order to remove the noise components,adaptive CEEMD algorithm was obtained by introducing the correlation root mean square error.This algorithm was used to fine decompose the blasting vibration signals,and obtain an intrinsic mode functions(IMF) with frequencies from large to small.Furthermore,the random MPE test was carried out for each IMF.In order to achieve the purpose of noise reduction,the IMF components with MPE value which was greater than 0.6 were removed.The results show that the algorithm has good fidelity and denoising effect.It eliminates the high frequency noise effectively and has little influence on the real information.Comparative analysis shows that the adaptive CEEMD-MPE algorithm is superior to the EMD-MPE and the EEMD-MPE algorithm,which verifies the effectiveness of the algorithm.