This article presents further developments of our earlier vector polar histogram method (VPH+) for unmanned ground vehicle(UGV) real-time obstacle avoidance and navigation in complicated unknown environments. Two other strategies, wall-following and move-to-goal are combined into the VPH+ method to overcome the local minimum problem. Meanwhile, a simple yet effective coordination mechanism that determines when to activate the corresponding strategy is carefully designed. The state memory and position prediction strategies are employed in the coordination mechanism to ensure that the corresponding strategy is evoked only at the appropriate moment. Contrastive simulation tests and real UGV vehicle tests confirmed that the proposed method is robust, stable, and efficient in complicated environments.
展开▼
