Validation of Road-Preview-Based Predictive Gear Selection on Heavy-Duty Vehicle Transmission Control Unit

The eHorizon unit enables the possibility to get information of the road ahead in a defined prediction horizon. This data, like road gradient, curve radius, velocity limitation and road class, can be used by the onboard transmission control unit (TCU) via Controller Area Network (CAN) bus. This obtained predictive road information combined with the actual driving conditions can be used to optimize the shifting strategy by a model predictive control (MPC) algorithm which intends to reduce the fuel consumption. In order to solve the optimum problem inside the MPC with less memory, a pre-optimization based dynamic programming (PODP) approach is proposed. In this paper, the predictive gear selection (PGS) strategy will be compared to a conventional automatic gear shifting strategy in a simulation environment and validated on road by implementing it on a heavy-duty truck with a 16-speed automated manual transmission (AMT).