REDUCTION OF TRANSVERSE VIBRATION OF A SERPENTINE BELT SYSTEM: MODEL PREDICTIVE CONTROL OF A BELT TENSIONER

Abstract

The serpentine belt systems are widely used in today’s automotive industry. The preference of serpentine belts in the market over the traditional belts due to the advantages such as consuming less space in the engine compartment, reduced slips etc. makes the vibration control of this type of belt crucial. In order to control the vibration of the belt, number of tensioner types are used. In this study, a tensioner which is controlled by model predictive algorithm is introduced to a belt system and system performance is compared to the belt system which is not controlled for various case studies. The belt and pulley system is designed with multibody dynamic simulation program in this study. The internal model of MPC, which is used to control the belt drive model, is established with the help of artificial neural networks. Finally, the study is completed by making comparisons on the results of the multi body dynamic model with MPC and the model controlled by PID.