Automatic Transmission Shift Strategy Based on Greedy Algorithm Using Predicted Velocity

Abstract

Control strategies for the vehicle equipped with an automatic transmission greatly affects the fuel economy and drivability. In general, the gear shift of automatic transmission is controlled based on the two-dimensional lookup tables. The lookup tables are calibrated based on the experimental results at a steady state condition. However, this method has a limitation on improving the fuel efficiency in a dynamic driving environment like an urban condition. In order to improve the fuel efficiency, this study proposes an optimal gear shift strategy based on the greedy control method using the predicted velocity. Since future driving conditions can be estimated using predicted velocity, optimal gear shifting is searched using a greedy algorithm based on the predicted velocity. A PI-type driver model and powertrain model are designed to calculate the forecasting vehicle states after gear shifting with predicted velocity. The proposed strategy was validated through the simulation of the urban driving cycle using various time period predicted velocity. Results show fuel efficiency was improved by up to 1.6% while shiftbusyness is prevented compared with the shift pattern which focused on fuel economy. As a result, the proposed strategy is affordable for improving not only the fuel economy but also the drivability in the dynamic driving environment.