Multi-cl model for multi-equipment anomaly identification and prediction on a gear manufacturing production line

The increasing demand in the NEV industry necessitates efficient and reliable gear production. Modern multi-equipment lines face challenges with delayed anomaly detection and untimely warnings, hindering proactive maintenance and impacting production. This study proposes a Multi-CL model framework for proactive anomaly prediction in multi-equipment collaborative manufacturing. The network integrates 1D Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and Transformer encoder advantages, enhanced by a mixed-precision algorithm, to improve prediction accuracy and efficiency. Comparative experiments show the model proposed in this paper achieves superior performance, significantly outperforming baseline models. This framework offers crucial technical support for intelligent anomaly management and predictive maintenance, thereby enhancing production line stability and efficiency.