Predicting the gear ratio changing time by a fuzzy linguistic approach - an application to truck driving

This research project presents a gearbox predictive fuzzy system. It was conducted as apart of cooperation between RENAULT and the information processing team at the University of Savoie. This predictor system, meant to propose to the driver the optimal gear ratio of a truck gearbox, is based on a fuzzy multicriteria evaluation on context variables (resisting force, road type, driver style...). The fuzzy symbolic method proposed enables to merge easily numerical measurements and linguistic evaluations. Preliminary tests carried out on a gearbox simulation bench have led to promising results.