“MATLAB and Simulink allowed the students to construct mathematically complex simulations quickly and easily.”
Dr. Mark Duvall, University of California, Davis
The University of California at Davis (UC Davis) was among 15 North American universities to participate in the 2001 FutureTruck Competition. This competition gives students an opportunity to apply skills learned in school to a real-world engineering project: Participants must reengineer a sport utility vehicle so as to reduce emissions and fuel consumption without sacrificing vehicle performance or safety. As a FutureTruck sponsor, The MathWorks provided its software for students to use throughout the competition.
UC Davis won the 2001 FutureTruck championship and the MathWorks Vehicle Modeling & Dynamics award. Their "plug-in" hybrid electric version of a conventional 2000 Chevrolet Suburban was judged to have the best combination of low tailpipe emissions, high fuel economy, low global warming gases, and good powertrain performance.
The team used MATLAB, Simulink, and a Simulink based vehicle modeling program called PSAT to develop the powertrain strategy for their vehicle. "I feel that our vehicle won because we had the best success in implementing our hybrid powertrain and control strategy, and MATLAB and Simulink played a very important role," says UC Davis Future Truck advisor Dr. Mark Duvall. "Using Simulink, thestudents could develop systems models that were both accurate and easy to modify."
Each team needed to develop a hybrid electric vehicle powertrain control strategy to determine how the vehicle’s engine and transmission should react to driver inputs. The strategy had to balance three objectives: reduce fuel cycle greenhouse gas emissions by 66%, reduce the fuel consumption to half that of the standard Suburban, and meet California’s Super Ultra Low Emissions Vehicle standards. Students were also evaluated on the completeness of their representation of the vehicle model.
The UC Davis team developed a comprehensive vehicle model based on PSAT (Partnership for a New Generation of Vehicles Systems Analysis Toolkit), a Simulink based modeling program developed by Argonne National Laboratory. They simulated their Suburban’s four-wheel-drive powertrain to determine a control strategy, settling on a combination of charge-depletion and charge-sustaining strategies. In charge-sustaining vehicles, the onboard battery is recharged by the combustion engine while the vehicle is being driven. During charge-depletion, electric energy from the battery powers the vehicle. The vehicle is recharged when the driver returns home that night.
Since the PSAT powertrain control strategy is intended for charge-sustaining vehicles only, the team had to rewrite many of the model’s control algorithms. Because PSAT is based on Simulink, this was easy to do: “One of the strengths of Simulink is the open source nature of the program,” Duvall says. “The students can go into the Simulink code for PSAT and make improvements or modifications that better suit their objectives.”
The UC Davis team embedded portions of the vehicle control software directly in the PSAT simulation. They imported the Suburban’s C-language microcontroller code into PSAT using a Simulink S-function. This allowed them to accurately simulate several modes of operation. They improved the control algorithms and simulated the changes in PSAT. Once the simulation results met their targets, they transferred the final algorithms directly to the vehicle controller for immediate use in the vehicle.
“The students could have developed models using C, but they would have expended more effort developing the software,” Duvall says. “With MATLAB and Simulink, they could dedicate more time to developing the vehicle models and optimizing the designs.” The students acquired engineering experience that cannot be taught in a classroom. They worked with software tools used by engineers in industry and learned firsthand how to apply these tools to problems in the real world. As a result of their work on FutureTruck, the team has been awarded several patents for hybrid vehicle control strategies.
To confront students with real-world engineering experiences by challenging them to modify an SUV to achieve lower emissions and fuel consumption without sacrificing vehicle performance
Use MATLAB and Simulink to model strategies for the powertrain control system for a hybrid electric vehicle