Improving Vehicle Navigation by a Heading-enabled ACO Approach

Xiao, Yamei, Chaomin Luo, Mohan Krishnan, and Mark Paulik

A heading direction methodology is proposed in this project in conjunction with a colony optimization algorithm (ACO) during the motion planning in the vicinity of obstacles to plan safer trajectories for real-time navigation and map building of an unmanned ground vehicle (UGV). In real world applications, a UGV is required to plan a shortest and reasonable collision-free trajectory that, in this project, is capable of being implemented by a novel heading-enabled ant colony optimization model. A LIDAR-based local navigation algorithm is implemented to carry out obstacle avoidance missions.  As the robot plans its trajectory toward the target, unreasonable path will be inevitably planned. A heading-enabled navigation paradigm is developed for guidance of the UGV locally so as to plan more reasonable and safer trajectories. In addition, grid-based map representations are implemented for real-time UGV navigation. In this project, simulation results successfully demonstrate robustness and effectiveness of the proposed real-time heading-enabled ACO approach of a UGV.