The EcoCAR challenge

  • 06-Jul-2009 03:29 EDT
EcoCAR student.jpg
Hugues Marceau, a student from the University of Ontario Institute of Technology, uses model-based design technology to test a virtual version of the team's vehicle before the actual design is assembled. EcoCAR, established by the U.S. Department of Energy, has more than 25 government and industry sponsors providing hardware, software, training, and support to the college teams.

They created green vehicles in the virtual world, and in the coming months 17 collegiate teams get physical in the second phase of a three-year, government-industry sponsored challenge.

Student engineers are re-engineering 2009 Saturn Vue compact SUVs that were part of a General Motors internal test fleet to achieve improved fuel efficiency and reduced emissions without diminishing vehicle performance. EcoCAR's first phase, which concluded in June, was heavily focused on model-based design, a key enabler for embedded systems development.

The University of Victoria, one of three Canadian universities participating in the challenge, "used advanced optimization techniques for model validation, an important aspect of model-based design. It also presented a comprehensive plan on model reuse as it progressed from model-in-the-loop to software-in-the-loop to hardware-in-the-loop simulation," said Tom Egel, Principal Application Engineer at The MathWorks, which provided its Matlab and Simulink software to EcoCAR teams.

Computer tools are helping teams make informed design decisions relative to fuel efficiency and emissions. "Many teams are using unique engine, motor, battery, and/or fuel cell combinations requiring unique control algorithms. An algorithm that works in a simulation environment may fail in the actual implementation because of I/O latency, processor limitations, and even faulty wiring," explained Paul Mandeltort, Product Manager for Automotive and Embedded Networks at National Instruments.

"By using LabVIEW and National Instruments hardware early in the algorithm engineering process, EcoCAR teams are quickly eliminating problems and developing more robust algorithms in real-world scenarios many months before the vehicle is ready to test. This results in shorter development time, faster integration, and ultimately a more advanced vehicle," said Mandeltort, an electrical engineer.

Ohio State University's team, working to redesign the Vue as an extended-range electric vehicle, was the first-place overall year-one EcoCAR winner. OSU also won dSPACE's Embedded Success award for showcasing "the best usage of dSPACE HIL simulation tools in their controls development. The team excelled in all aspects of HIL application, including vehicle modeling, failure simulation, test plan development, and automated test execution," said Santhosh Jogi, Director of Engineering at dSPACE Inc.

In the second phase of the EcoCAR challenge, teams will work to refine vehicle design architecture en route to having a functional prototype before entering the third and final competition phase. "With the aggressive development schedule, teams like OSU will need to utilize their HIL system throughout the process just as industry professionals do. During the next two phases of the EcoCAR competition, control software will go through several iterations, so validating with a full regression test on the HIL system is the recommended practice. It is important to do that for all iterations in order to make sure an unintended effect was not introduced and to validate any new functionality that was added," said Jogi.

OSU's team will modify its Vue by removing the engine, transmission, and the corresponding power electronics. The OSU EcoCAR vehicle will be fitted with a front electric machine (FEM), a rear electric machine (REM), a 21 kWh lithium-ion battery pack from A123Systems, and a 1.8-L, four-cylinder Honda compressed natural gas engine—reconfigured to run on E85 ethanol—for charging the 363-V battery pack in range-extending mode. The FEM will act as a primary generator in range-extended mode and as an electric drive during all-electric operation, while the REM—the same drive unit used in GM's EV1 all-electric car—will function as the primary drive in all-electric and range-extended modes, according to Eric Schacht, OSU's EcoCAR team leader.

When OSU and the 16 other college teams from the U.S. and Canada begin the physical vehicle conversion process in the coming weeks, computer modeling and simulation will facilitate the design and integration of various advanced hybrid vehicle technologies. "OSU's controls team will work to make the vehicle and component models much more dynamic. The controls team also will be adding many more detailed simulations and parameters to the controls system, such as gear selection. And our engine team will develop an initial model and test it on the HIL system to be certain it works properly," said Schacht, an electrical engineering OSU student.

HTML for Linking to Page
Page URL
Rate It
4.75 Avg. Rating

Read More Articles On

Related Items

Training / Education
Technical Paper / Journal Article
Training / Education