A simulator for autonomous and semiautonomous controllers performing obstacle avoidance in the presence of delay

Model Predictive Control (MPC);Receding Horizon Control;Simulator;Dijkstra;Potential Field Control;Human Control;Cost-to-go

Zearing, Joseph T. ; Stipanović ; Dušan M.

关键词: Model Predictive Control (MPC);    Receding Horizon Control;    Simulator;    Dijkstra;    Potential Field Control;    Human Control;    Cost-to-go;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/16483/Zearing_Joseph.pdf?sequence=2&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

This thesis discusses a MATLAB based simulator designed to aid in the study of various controllers performing obstacle avoidance tasks in the presence of delay.The simulator is divided into functional blocks, each of which is described in detail.Several controllers were developed and their performance analyzed using the simulator.These controllers include several Model Predictive Control based controllers as well as an artificial Potential Field controller.Additionally, humans acted as controllers and their performance was quantified.The Model Predictive Control based controllers developed here include a simple application in which a two dimensional distance is employed as the cost-to-go, and a more advanced application that uses Dijkstra’s algorithm to find a more accurate cost-to-go estimate.