What is MTSOS?

MTSOS is a set of C subroutines that find the control inputs necessary for a vehicle to traverse a specified path in minimum time subject to the dynamics of the vehicle and actuator limits. For developer convenience, MTSOS includes a MATLAB interface. The details of the problem that MTSOS solves can be found in Minimum-Time Speed Optimization Along a Fixed Path. To implement MTSOS, the user must write two subroutines: the first to describe the vehicle dynamics and the second to describe a barrier function of the actuator limits.

MTSOS was written by Thomas Lipp a graduate student in Stephen Boyd's lab at Stanford University. MTSOS uses the CSparse library written by Tim Davis of the University of Florida. MTSOS is made available under a GPL.

MTSOS is lightweight and designed to be run in real-time and, as such, is suitable for the inclusion in model predictive control like loops. Although we make no claims of this software being fully optimized for speed (we can make several suggestions for improvements if speed is essential), it will perform orders of magnitude faster than existing general purposes solvers (like SeDuMi or SDPT3).