Abstract
The gearshift schedule of an automatic transmission (usually implemented in the form of shift map calibrated by engineers) has a great effect on fuel economy, performance and drivability of the motor vehicle. Calibration of shift maps traditionally is performed by engineers heuristically based on their experiences, intuitions and existing reference shift maps. This traditional shift map calibration process not only is time consuming, but also does not guarantee the utility of the resulting shift map versus the inherent potential of the power train configuration. Furthermore, if the power train hardware are unfamiliar to calibration engineers (for instance, engine with unfamiliar characteristics, or transmission with new gear ratios); the calibration becomes more difficult with little or no confidence regarding its optimality. To change this situation, a math-model-based shift map generation tool has been developed. The tool uses Dynamic Programming technique to achieve an optimal solution of shift map generation. The problem is formulated as a performance and drivability constrained fuel economy maximization problem. The tool greatly accelerates shift map calibration process, and provides a measurable objective achievement.
Keywords:Automatic, Transmission, Shift, Schedule , Optimization, Dynamicprogramming.