Abstract: In the field of coal mine production，belt conveyor is a common transport equipment，its safety and stability directly affect the efficiency of coal mine production. Aiming at the problem that belt conveyor is prone to failure during transportation，an inspection robot for bulk material transportation system based on improved fuzzy cellular neural network (IFCNN) optimized by dune cat swarm algorithm (SC⁃ SO) is proposed. The inspection robot is equipped with a variety of sensors to collect the image，tempera⁃ ture，speed，sound and other information of the belt conveyor during operation. The signals collected by each sensor are converted into frequency domain and time-frequency domain information，and IFCNN is used to extract and fuse the frequency domain and time-frequency domain information respectively. Combined with the attention mechanism to select important features for fault diagnosis，so as to realize the intel⁃ ligent inspection of bulk material conveying system. The practical application shows that the IFCNN opti⁃ mized by SCSO can effectively extract the fault features of belt conveyor and improve the accuracy of fault detection，so as to improve the safety of coal mine transportation.

Key words: multi-sensor fusion, IFCNN, fault detection, SCSO