Student Projects
Contact Information
University: UC Berkeley
Team Member(s): Andrew Kuo, Carson Schulze, Nick Okita, Toby Ricco
Faculty Advisors: Benson Tongue, Dennis Lieu, George Anwar
Email Address: tobyricco@berkeley.edu
Country: USA
Project Information
Title: INSTAR Traction Control and Electronic Differential, USA
Description:
Implementation of traction control and an electronic differentail on an electric go-kart. The differential was achieved by a torque-splitting control algorithm to modulate current to each of the independently driven rear wheels; traction control was realized by means of speed matching the rear wheels to the front wheels.
Summary:
UC Berkeley's ME135 class is a mechanical engineering project-based class that focuses on labview, microcontrollers, and embedded systems. Our ME135 team, Andrew Kuo, Carson Schulze, Nick Okita, and Toby Ricco, had the idea to collaborate with a research group, INSTAR, for our class project. INSTAR (INertial STorage And Recovery) focuses its research on optimizing vehicle efficiency through hybrid energy storage technology such as a flywheel and battery-pack combination. INSTAR currently uses an electric go-kart (constructed by members of INSTAR and the ME135 project group) as its research platform. The Kart features two brushless motors (about 15 horsepower each) powering the rear axle. The ME135 project group decided that splitting the rear axle and independently controlling the rear wheels, as well as designing a traction control system, would make for an ambitious class project. And it did.
Wheel speed sensors, steering wheel sensors, vehicle dynamics, signal noise, real-time signal processing, and issues with computational speed were some of the various challenges encountered. Most of these issues were resolved through a combinaton of filtering electronics, mechanical improvements, and code-based filtering and error detection. By the end of the Spring 2012 semester, our group had a vehicle that turned better with traction control 'ON' and was able to navigate a tarp covered in dish-soap and water more effectively with traction-control 'ON' than when these systems were turned off. These control systems are user-selectable by the driver. Traction control can be turned on or off, and so can the differential. The vehicle works with one ON, the other OFF, visa versa, both ON, or both OFF. Overall, the project was a fantastic learning experience, and there are few things more fun to validate than driving an out-of-control go-kart on a slippery tarp.
A big thank-you to NI for donating all of the relevant NI hardware as well as additional hardware to be used for research and flywheel controls in the upcoming year.
For a more-detailed project overview, refer to attached file: Projectoverview.pdf
Products Used:
• (1) NI cRIO-9076 compact microcontroller
• (1) PS-15 Power supply (781093-01)
• (1) NI 9207 analogue input module (with 778676-01 connector block and 778621-01 cable)
• (1) NI 9263 analogue output module
• (1) NI 9853 CAN bus module**
• (1) NI 9403 I/O module (with 778676-01 connector block and 778621-01 cable)
**CAN communications to be setup in the near future.
Kart Assembly Animation (credit to Andrew Cheng, INSTAR member)
http://www.youtube.com/watch?v=I7LOzm02Z4w
Nominate Your Professor:
UC Berkeley's professor, Benson Tongue, is a phenomenal dynamicist and helped brain storm ideas for how to implement the traction control and electronic differential. Professor Dennis Lieu is a master of motors and magnetics: he leads the research team on the flywheel/energy-recovery aspect of the vehicle. George Anwar teaches labview and controls at UC Berkeley. George's experience and guidance helped greatly in the process of implementing labview programming, sensors, and the relevant electronics.
