Abstract: FFT technique is used to study the characteristics of load spectrum of the cutter pick on cutting coal and rock. Then,two kinds of wavelets,Db4 and Sym5 are applied to decompose and reconstruct the load spectrum and extract features of high-frequency and low-frequency components of the spectrum. Finally,fractal theory is used to calculate the box dimension in different layers of load spectrum after wavelet transform. Experimental results show that there is a steady-state value on the amplitude of load spectrum and it fluctuates mainly from 0 to 5 Hz. In addition,the fitting-curve of the load spectrum in different reconstruction layers based on Db4 and Sym5 are obtained. To be specific,the difference of maximum and average value in the fourth layer between the fitting curve and original load spectrum are 0.17% and 0.12% respectively. Furthermore,the low-frequency reconstructed load spectrum satisfied the distribution of original load spectrum,so it can be used as an efficient method to calculate the average value of the load. The decomposition and reconstruction precision can be measured based on the relationship between the low-frequency box dimension in previous layer and the low and high-frequency box dimension in the next layer.