A Fault Diagnosis Method of Tread Production Line Based on Convolutional Neural Network

摘要

As an important part of automobile, tire is the foundation of the development of automobile industry. The stable and optimized control of compound extrusion process of automobile tire tread is the premise for producing good quality tires. Convolutional neural network is a model of deep learning which was widely used in image recognition, speech processing and other fields and has a very good classification capacity. In this paper, a convolutional neural network model is proposed. The data of the rotating speed, pressure and main current of Φ 150 and Φ 120 extruder compounds tread extrusion lines are selected as the model input. The convolutional neural network is applied to the fault diagnosis of the tread compound extrusion line and compared with the traditional machine learning algorithm-BP neural network, decision tree and Logistic regression. The superiority of the algorithm is verified from the result data. A new method of fault detection and diagnosis for tire tread production line based on convolution neural network is proposed.