Diagnosis, Analysis and Evaluation of the State of Tobacco Shreds and Smoke Blocks in Tobacco Machine Conveying Based on the Convent Model

The material flow interruption fault in tobacco machinery is a typical type of fault, mainly manifested as abnormal material conveying or abnormal airflow, which may lead to equipment shutdown, production losses, and even pose safety hazards. The rapid and accurate identification of flow interruption faults in tobacco machinery and the timely, effective prevention and mitigation of such faults are of great significance for improving tobacco production efficiency. This paper adopts artificial intelligence technology to study abnormal material conveyance in tobacco machinery flow interruption, focusing on the identification of tobacco shreds and tobacco clumps during conveyance. First, a dataset of abnormal material transportation is collected. Then, comparative experiments are conducted by training multiple models on the dataset, and the optimal model is selected. Experiments show that the final accuracy of the Convent model is as high as 91.5 %, higher than other classical convolutional neural network architectures. This research provides a feasible, high-performance solution for the diagnosis of tobacco machinery flow interruption faults.