Abstract
In the present work, a computational routine was developed in order to optimize the shifting strategies implemented in a Gear Shift Indicator to assist the driver of manual transmission vehicles according to a desired driving style. In order to achieve this goal, a simulation algorithm was developed to estimate the fuel consumption and acceleration performance of a vehicle equipped with a 5-speed manual gearbox propelled by a 1.4 L engine when submitted to the FTP-72 driving cycle. A multi-objective Genetic Algorithm was then employed to solve the optimization problem for both fuel consumption and performance. Among the set of optimized solutions, 3 shifting strategies were selected for experimental evaluation on a chassis dynamometer. The experiments showed that the strategy with focus on fuel economy was able to save 10.7% of fuel, being the best performance strategy responsible for enhancing the vehicle acceleration in 8.33%, when compared to the standard strategy provided by the vehicle manufacturer. Finally, the best-compromised solution presented 6.4% fuel-savings with an acceleration performance similar to the best performance strategy.
Acknowledgments
This work was conducted during scholarships supported by the National Council for Scientific and Technological Development (CNPq), Brazilian Federal Agency for Support and Evaluation of Graduate Education (CAPES), State of São Paulo Research Foundation (FAPESP), and the University of Campinas (UNICAMP).