Engineering Portfolio: Chris Berthelet

          This project was for my Experiential Advanced Control Design class, in which a main topic was Model Predictive Control (MPC). The goal of our project was demonstrate the effectiveness of an MPC dynamic programming algorithm toward autonomous parking problems. The dynamic programming approach is well suited for parking problems because of its feedback characteristics and its computational efficiency in embedded applications.

         

          The algorithm is implemented in MATLAB and simulations are run using a simplified vehicle model. The algorithm computes the optimal path from an initial position and orientation to a desired terminal state, all while avoiding collisions with obstacles. 

         

          Results demonstrate robust, efficient, and optimal paths over a variety of parking scenarios while also illustrating key conclusions about the effects of certain parameters used to compute dynamic programming solutions (see attached report).