Abstract: This paper addresses the limitations of existing fault diagnosis methods for mine hoists by employing an intelligent fault diagnosis method based on multi-sensor information fusion. This method uses DS evidence theory for decision-level fusion to identify faults and utilizes statistical evidence to obtain basic probability assignments. An example demonstrates that this method is superior to single-sensor information fault diagnosis methods in mine hoist fault diagnosis. Keywords: information fusion; fault diagnosis; statistical evidence 0 Introduction Equipment failure refers to "equipment malfunction," meaning the equipment cannot achieve its expected working state and cannot meet its required performance and functions. Equipment fault diagnosis technology is a comprehensive discipline that has developed over the past 40 years to meet the actual needs of engineering. It is an inevitable result of the interdisciplinary development of various disciplines, and the resulting economic benefits are significant. Multi-sensor information fusion involves processing information from multiple sensors at multiple levels and in multiple aspects to derive new and meaningful information. This new information cannot be obtained by any single sensor; it represents the behavior of the detected target object. Through multi-sensor information fusion, the spatiotemporal coverage can be expanded, confidence can be increased, and the reliability of the detection system can be improved [sup][1][/sup]. This paper studies the identification of faults in mine hoists using a decision-level fusion method based on the DS evidence theory, which obtains basic probability assignments based on basic probability assignments. 1 Multi-sensor data fusion algorithm When applying the DS evidence theory to multi-sensor data fusion, the relevant values ​​obtained from the sensors are the evidence in this theory. They can constitute the confidence function assignment of the target pattern to be identified, representing the credibility of each target pattern hypothesis. Each sensor forms an evidence group. Multi-sensor data fusion combines several evidence groups into a new, comprehensive evidence group using joint rules. The confidence function assignments of each sensor are then combined using these rules to form a fused confidence function assignment, thus providing comprehensive and accurate evidence for target pattern decision-making. 1.1 Basic Concepts of DS Evidence Theory Evidence theory was initially proposed by Dempster in 1967. He used multivalued mappings to derive upper and lower bounds on probabilities. Later, Shafer generalized this into evidence reasoning in 1976, hence it is also known as the Dempster-Shafer theory. Let U represent a universe of discourse for all possible values ​​of X, and all elements within U are mutually exclusive. Then U is called the identification frame of X. U can be finite or infinite; in expert system applications, it is finite. According to the fusion decision method, the fusion decision result is O1, indicating that the reducer is malfunctioning. The calculation results show that after fusing phases A and B, the uncertain basic probability assignment is greater than the set threshold, and the fault type cannot be identified. After fusing the result with sensor C, the uncertain basic probability assignment of the fault decreases, and the fault type can be identified. It is evident that using a single sensor for fault diagnosis of mine hoists has relatively low reliability and sometimes fails to accurately identify the fault type. However, using multiple sensors can improve the fault identification rate. 4. Conclusion The analysis results show that using only single-sensor information for fault diagnosis of mine hoists results in many false diagnoses and false negatives. Fault diagnosis based on multi-sensor information fusion can fully utilize information from multiple sensors, reducing diagnostic uncertainty and identifying faults that a single sensor cannot detect. Furthermore, the reliability of the diagnosis increases with the number of fused sensors. References [1] He You, Wang Guohong, et al. Multi-sensor information fusion and application [M]. Beijing: Electronic Industry Press, 2007, 40-45 [2] Huang Ying, Tao Yungang, et al. Application of DS evidence theory in multi-sensor data fusion [J]. Journal of Nanjing University of Aeronautics and Astronautics, 1999, 31(2): 172-177 [3] Wang Jiangping. Fault diagnosis based on multi-sensor fusion information [J]. Mechanical Science and Technology, 2000, 19(6) [4] Lu Yan. Fault monitoring and diagnosis of mine hoists. Metal Mines, 1998.3