Abstract: In view of problems that coal and rock identification systems use single sensor to monitor data and have low precision, reliability and stability, an identification method of coal and rock based on information fusion and neural network was proposed. A variety of necessary sensors are added to the existing shearer, which are used to collect current, pressure, vibration frequency, acceleration and other signals of the shearer under different situations. Wavelet packet is used for characteristics extraction, and BP neural network is used for data fusion, so as to achieve coal and rock identification. The test results of the real machine show that the identification error of the proposed method is within ±0.5, which verifies its validity.