Abstract: Shearer working conditions have strong stochastic complex and variable factors, its whole life cycle course of stochastic load spectrum acquisition and discrete sequence processing, as well as the whole domain load spectrum synthesis and digitization is the key basic engineering technology common problems of high-end equipment reliability design analysis and its fatigue life prediction. Constructing the correlation matrix of shearer sensors and detection quantities, detection quantities and feature quantities, adopting the superposition processing of row and column coefficients of the comprehensive correlation matrix to give the comprehensive correlation and importance of sensors and feature quantities, and proposing the quantitative evaluation method of feature quantity correlation in the process of multi-sensor information detection. Considering drum parameters, working parameters of shearer, nature and distribution parameters of coal rock, proposed a longitudinal amplitude and transverse time-range cumulative normalization algorithm for the load spectrum, as well as a proportional synthesis method for the segmental load spectrum and whole domain load spectrum of the drum cutting coal rock. Giving fusion algorithms for cutting motor current, rocker arm vibration and heightened hydraulic cylinder pressure information, approximate estimation of scale factor algorithms, etc. Establish the instantaneous load model and the load spectrum random distribution superposition algorithm of the drum cutting coal rock, combine the experimental load spectrum of the single pick cutting coal rock with the theoretical model, and obtain the segmented theoretical load spectrum of the drum cutting coal rock. The numerical simulation and experimental load spectrum containing noise information are processed by the adaptive weight EMD-IECM load spectrum optimal noise reduction smoothing algorithm, which gives a graded cumulative time-range ratio statistical algorithm of the drum cutting coal rock segmented load spectrum, and puts forward the synthesis and digitization method of the whole domain load spectrum. The results show that the correlation between the cutting load and coal rock strength and the detection information of each sensor is more significant, and the detection information of vibration sensors, current sensors and pressure sensors can better reflect the cutting characteristic quantity and cutting state. Adaptive weighting EMD-IECM load spectrum optimal noise reduction and smoothing algorithm, which can automatically adjust the weights of the indicators according to the data itself, and obtain the reconstructed load spectrum with balanced similarity and smoothness. Through the application of examples, it is known that the amplitude graded cumulative time-range ratio of the whole domain load spectrum shows a normal distribution law, which is in line with the actual situation of the project, and verifies the validity and completeness of the synthesis and digitization method of the whole domain load spectrum of the drum cutting coal rock.