H∞ robust control of a seat belt load-limiting device

Continuously adaptive restraint system is currently the ideal state of restraint systems for occupant protection. It operates in real time during the impact according to the sensed operating conditions, such as seat belt force, occupant motion, crash severity, etc. In this paper, a linear time invariant (LTI) model of a crash dummy/vehicle system is built based on the method of system identification. The order of the system is first reduced using singular value decomposition (SVD). Next, a robust H_∞ controller is designed for the reduced-order system using linear matrix inequalities (LMIs). Simulations have been performed for the closed-loop system with both designed controller and the MADYMO occupant and vehicle model to demonstrate the performance of the designed controller.