Model-based gear-shift optimization for an automated manual transmission

Abstract

Gear shift quality is traditionally optimized in a vehicle on the road, through the evaluation of the driver’s perception. This classical human-feedback-based method faces different challenges, such as low reproducibility and high cost. In order to overcome these drawbacks, an innovative automatic virtual calibration framework is proposed in the Model-in-the-Loop environment. The devised framework is composed of a dynamic model of the transmission system, an adaptive gear shift controller, an objective assessment of the gear shift quality, and a multi-objective optimization algorithm. The Pareto-optimal front with multi-conflicting criteria, such as shift time and shift comfort, can be obtained within this framework. In particular, an improved hybrid multi-objective evolutionary algorithm is developed to enhance the optimization performance. The optimization of the shift trajectory for an automated manual transmission synchronization system is implemented as a case study. The obtained optimal shift trajectories are also validated on a transmission test bench.