Gearshift Calibration for Automatic Transmission Using a Model-Based Optimization Algorithm

Software development for automotive application requires several iterations in order to tune parameters and strategy logic to operate accordantly with optimal performance. Thus, in this paper we present an optimizer method and process used to tune calibration parameters related to gearshift patterns for immediate usage during gear box tuning for an automatic transmission application. The algorithm is developed based on the forward-looking vehicle simulator, "Autonomie" and the optimization algorithm called "POUNDERS". The calibration process is simulated for a conventional internal combustion engine based vehicle to allow the efficient optimization of the gearbox shift patterns and thus the reduction of transmission calibration effort in the development process. In the process the calibration process is used by the calibration engineer to speed up the in-vehicle development and plausibility check results quickly.