Trajectory-Optimized Torque Vectoring Controller for a Two-Track Model FSAE Racecar

I used direct transcription trajectory optimization on a nonlinear state-space vehicle model in matlab to design a torque vectoring controller that increases cornering performance.

This was a project for the graduate class 6.832: Underactuated Robotics, which I took last spring (junior year). The class taught methods for the analysis and optimal control of nonlinear underactuated systems.

I learned a lot, especially about trajectory optimization, and was excited to apply knowledge from this class to my work on the FSAE team.

My technical paper for the project is shown below, as well as a brief video clip to summarize the project.

RMcCabe_6_832_TrajOpt_Racecar-post-submi

.

Download • 7.21MB