Occupant kinematics in abrupt vehicle maneuvers are highly variable, yet previous active human body models provided only deterministic predictions for a limited range of body sizes. This study bridges the gap by developing and validating an efficient tool capable of stochastic predictions, thereby capturing behavioral variability across diverse occupant characteristics during pre-crash maneuvers.

A computationally efficient version of the midsize male GHBMC simplified model (GHBMCsi-pre) was first developed by rigidizing non-deformable body components in vehicle maneuvers while preserving key geometric and joint configurations. Closed-loop proportional-integral-derivative (PID) controllers were implemented at key joints to simulate active muscle responses. Twelve parametric models were then generated by morphing GHBMCsi-pre to represent diverse occupant characteristics (age, stature, and BMI). The models were validated against subject test data under abrupt braking and turn-and-brake maneuvers from a previous study.

Results showed that age and BMI significantly affect head excursions, with older and higher BMI occupants exhibiting smaller excursions, likely due to behavioral adaptations. The parametric models accurately captured occupant variability, covering the full range of corridors for subject-tested head excursions without requiring stiffness adjustments for stature. The developed GHBMCsi-pre model also reduced computational time by 80% compared to the original GHBMC model, making it feasible for long-duration pre-crash simulations.

This study presents a robust and scalable tool for simulating diverse occupant responses during pre-crash scenarios with stochastic predictions, supporting the design and development of adaptive safety systems. Further work is needed to better understand age and BMI effects on pre-crash occupant kinematics.