Abstract: As one of the core components of light industrial machinery and equipment, the operating condition of slewing rings directly affects the safety and production efficiency of the equipment. Traditional fault diagnosis methods for slewing rings rely on manual feature extraction and model construction, which suffer from poor robustness and insufficient diagnostic accuracy when dealing with complex signals. To address this challenge, this study proposes a multi-modal signal fusion method based on Convolutional Neural Networks （CNNs）. The spectral features of vibration signals and acoustic emission signals are input into the CNN model in the form of dual-channel images, enabling efficient and accurate diagnosis of four types of slewing ring faults （normal condition, inner race fault, outer race fault, and rolling element fault）. The results show that the multimodal signal fusion diagnosis strategy can significantly improve diagnostic accuracy, achieving a classification accuracy of 93.92% on the validation set. The training process demonstrates good stability and generalization capability. The research provides effective technical support for fault diagnosis of mechanical equipment under complex working conditions.