Date of Award
Fall 2017
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Mechanical Engineering
First Advisor
Fleischmann, Jonathan A.
Second Advisor
Voglewede, Philip A.
Third Advisor
Bowman, Anthony
Abstract
A Model Predictive Control (MPC) LIDAR-based constant speed local obstacle avoidance algorithm has been implemented on rigid terrain and granular terrain in Chrono to examine the robustness of this control method. Provided LIDAR data as well as a target location, a vehicle can route itself around obstacles as it encounters them and arrive at an end goal via an optimal route. This research is one important step towards eventual implementation of autonomous vehicles capable of navigating on all terrains. Using Chrono, a multibody physics API, this controller has been tested on a complex multibody physics HMMWV model representing the plant in this study. A penalty-based DEM approach is used to model contacts on both rigid ground and granular terrain. Conclusions are drawn regarding the MPC algorithm performance based on its ability to navigate the Chrono HMMWV on rigid and granular terrain. A novel simulation framework has been developed to efficiently simulate granular terrain for this application. Two experiments were conducted to analyze the performance of the MPC LIDAR-based constant speed local obstacle avoidance controller. In the first, two separate controllers were developed, one using a 2-DOF analytical model to predict the HMMWV behavior, and the second using a higher fidelity 14-DOF vehicle model. In this first experiment, two controllers were compared as they controlled the HMMWV on two obstacle fields on rigid ground and granular terrain to understand the influence of model fidelity and terrain on controller performance. From these results, an improved lateral force model was developed for use in the 2-DOF vehicle model to better model the tire ground interaction using terramechanics relations. A second experiment was performed to compare two developed controllers. One used the 2-DOF vehicle model using the Pacejka Magic Formula to estimate tire forces while the second used a 2-DOF vehicle model with the newly developed force model to estimate lateral tire forces. As a result of this research, a smarter controller was developed that uses friction angle, cohesion, and interparticle friction coefficient to more accurately predict vehicle trajectories on granular terrain and allow a vehicle to navigate autonomously on granular terrain.