Analysis of Centrifugal Clutches in Two-Speed Automatic Transmissions with Deep Learning Based Engagement Prediction
Bo-Yi Lin* and Kai Chun Lin*. “Analysis of Centrifugal Clutches in Two-Speed Automatic Transmissions with Deep Learning-Based Engagement Prediction”. arXiv preprint arXiv:2409.09755 (2024). paper
Numerical analysis of centrifugal clutch systems integrated with a two-speed automatic transmission is shown in this paper. Various clutch configurations and their effects on the dynamics of the considered transmission have been examined. Based on these configurations, torque transfer, upshifting, and downshifting behaviors under various conditions are discussed. This paper presents a Deep Neural Network (DNN) model for clutch engagements, whose parameters are spring preload and shoe mass. In this paper, a computationally efficient alternative to these complex simulations for the modeling is presented. Deep learning and numerical modeling further help in the critical insights required for improvement in the design parameters, performance, and efficiency of the clutch-transmission system.