Design of a Fuel Economy Oriented Vehicle Longitudinal Speed Controller with Optimal Gear Sequence

In this paper, a vehicle speed controller aiming at minimizing fuel consumption in car-following scenarios is developed. The optimal longitudinal control of torque and brake is calculated and applied under Model Predictive Control (MPC). For a given car-following distance modulation range, the controller searches for the optimal gear selection at each step in the prediction horizon to generate the minimum fuel control, and triggers gear shifts optimally by vehicle state. The search process with optimality in torque, brake, and gear all ensured is originally highly time consuming. However, in this study, a cooperative method is developed to achieve the optimal solution in real time. The numerically complicated problem is partitioned into a series of simplified subproblems by a deduction method. Fast quasi-optimal solutions are given to the subproblems using an optimal driving pattern derived by Pontryagin Minimum Principle (PMP). To solve for different speed scenarios, a Finite State Machine (FSM) is designed for problem partitioning and fast solving. As the result, the search space is efficiently narrowed down for each MPC step, and the computation speed is fast enough for real-time application. A simulation study shows that the developed controller can generate a fuel saving benefit of 14% for the tested scenarios.