The Hybrid Electric Vehicle Team (HEVT) of Virginia Tech is an award winning multi-disciplinary team (electrical, software, and mechanical) that builds advanced vehicles to demonstrate leading-edge automotive technologies. The team participated in the 2011-2014 EcoCAR 2 competition in which the team re-engineered the power train of a donated GM vehicle to reduce well-to wheel (WTW) petroleum energy consumption, WTW greenhouse gas, and criteria emissions while maintaining vehicle performance, consumer acceptability and safety.
Over the course of three years, the team designed, built, and refined a 2013 Chevy Malibu to a hybrid electric vehicle. After balancing different powertrain designs to meet competition regulations and goals, a series-parallel plug-in hybrid electric vehicle with P2 and P4 motors was selected to be designed and built. The project was divided into three sub-teams:
As a member of controls team, it was vital to create a control system architecture and design process that could keep up with the accelerated build schedule. With all of the complex subsystems on a vehicle, a distributed control system is the most effective design. This allows the reuse of components and their accompanying electronic controller units (ECUs) in the vehicle without having to validate the controller base functionality (though system level validation is still necessary). This cut down on wiring and the reliability issues.
While a conventional vehicle has no main ECU, the team used a hybrid vehicle supervisory controller (HVSC) which was responsible for interpreting driver demand, responding to component level fault conditions reported by other ECUs, detecting system level fault conditions, determining mode of operation, and commanding powertrain components. The controls team needed to determine and analyze the correct data transfer and interactions between these controllers, the driver, and the power-train. As a result, a new controller area network (CAN) was designed with the HVSC to compliment the vehicle’s architecture. For example, the HVSC would need to determine the mode of operation the vehicle is in and thus which power-train components will power the vehicle.
Different colored arrows represent different types of interaction. Physical interaction (1) represents the amount the driver is pushing on the accelerator pedal. Analog interaction (1) passes this signal to the HVSC which determines the vehicle’s mode of operation. The accelerator pedal’s analog signal is either passed on the ECM in analog interaction (2) during charge sustaining mode or to the Traction Motor Inverter Module (TMIM) in analog interaction (3) during charge depleting mode. In charge sustaining mode, analog interaction (4) represents all the actuators such as ignition and injectors the ECM controls on the engine. Analog interaction (5) contains all the sensor signals that the ECM reads from the engine such as cam angle sensors, oxygen sensors, intake airflow sensors, ect. Analog interaction (4) contains the analog signals the ECM uses to control the 12v starter and to shut the engine off. CAN interaction (1) contains command signals from to ECM to control engine torque while CAN interaction (2) contains feedback signals such as engine speed and achieved engine torque. A similar scheme of data transfer is conducted during charge depleting mode as the HVSC used the rear traction motor and energy storage system to power the vehicle.
Once a hybrid electric controls architecture was in place, a rapid controls prototyping process was followed to build, test, and validate new functionalities that were to be added to the already existing vehicle controls system. This process started with developing a plant model to virtual simulate the physical system the controller needs to control.
Testing of the simulated plant models and controllers helped identify many issues prior to the physical integration. The team used a rapid controls prototyping (RCP) method as described above. The two simulation methods that were used by HEVT to validate the controls logic are Software-In-the-Loop (SIL) and Hardware-In-the-Loop (HIL). SIL testing focuses on the development of the plant model’s and control’s algorithms in the computer environment. The HIL testing is concerned with finalizes these interfaces where a real time plant model is loaded to a simulation target and the control code is loaded to the actual HVSC hardware. The last step in the development validation of the controls software is physical integration onto the Plug-in Hybrid Electric Vehicle.
Wiring of the new LE9 engine wiring harness was needed to accommodate new hybrid components. The stock engine-to-body harness and fuse block connectors remained. Signals from old stock components were recycled for new hybrid components. Continuity testing with a multi-meter was performed before physical vehicle testing with installed components. Functionality of engine bay wiring harness was validated with engine start and successful HVSC communication with engine bay ECUs. In addition, I took the lead in implementing the vehicle pedal emulator and configured to command the pedal signal to the HVSC. High voltage and rear traction motor wiring harness was validated with an energy storage system charge and rotation of rear wheels when commanded: validation of electric wheel spin
Year two’s goal was to have a 65% complete mule build of the vehicle. Both power trains ran the vehicle successfully and were ready for refinement during year 3.
