Optimal gearshift strategy based on greedy algorithm using predicted velocity in automatic transmission vehicle

Several powertrain control technologies have been developed in order to cope with a greenhouse gas regulation of vehicles. Especially, control strategies for an optimal gear ratio have been researched since the gear ratio of the transmission greatly affects the fuel economy and drivability of the vehicle. In the case of the automatic transmission, a gearshift is conducted by the transmission control unit (TCU). Therefore, the optimal gearshift strategy for TCU becomes important. In general, gearshift of an automatic transmission is controlled based on two-dimensional lookup tables which are called shift pattern. In order to improve fuel economy and drivability, a lot of shift patterns should be calibrated and shift patterns are switched according to the driving environment. Therefore, an enormous time and effort are required to prepare calibrated shift patterns or determine proper shift pattern depending on the driving environment. In order to overcome the weakness, the gearshift strategies based on a calculation using drive model have been researched. The gearshift strategies based on calculation compute a driving force and cost function which consists of fuel economy and drivability firstly. Then gearshift is conducted based on the calculated values. Furthermore, techniques about applying driving environment information to the gearshift have been researched for higher fuel economy and drivability. In order to apply driving environment information for optimal gearshift, some issues should be considered. First, traffic information such as traffic volume is changed with time. Therefore, it is necessary to formulate a shift control strategy every time according to changing the driving environment. Second, an optimal control strategy should be applied with little computation in order to establish a shift control strategy in real time. This thesis proposed an optimal gearshift strategy based on a greedy algorithm using predicted velocity to solve these issues. The proposed gearshift strategy was designed based on the greedy algorithm which has little computation. A simple powertrain model was used to estimate the vehicle states after shifting the gear, and optimal gearshift was conducted based on the greedy algorithm using the estimated states. A predicted velocity which reflects the changing driving environment was applied to gearshift strategy based on the greedy algorithm in order to perform the optimal gearshift. The proposed gearshift strategy was validated through the FTP 72 driving cycle using commercial simulation software. Validation results showed that the proposed strategy improved the fuel efficiency up to 2.2% than the shift pattern which is focused on fuel economy, while eliminating the frequent gearshift and decreasing the number of shifts by 14.6%.Maste