Event-Triggered Feedforward Predictive Control for Dimension Quality Optimization in BIW Assembly Process

Abstract

The assembly process of car body in white (BIW) in body shop is a very important link in the whole vehicle manufacturing process. Hundreds of stamping parts must be located and joined on the correct location of the platform, and the related accuracy influence the performance and quality of the cars. To guarantee the BIW manufacturing accuracy, the dimension control is a critical task in each automobile production line. In Comparison to traditional offline dimension control, this article proposed an event-triggered feedforward predictive control method to realize real-time inline dimension control during BIW assembly process. Experiments based on the actual production data from BMW Brilliance Automotive Ltd. demonstrate this control strategy can significantly improve the final products key point pass rate and decrease the fixture adjustment frequency.