A Prior Information Heuristic based Robot Exploration Method in Indoor Environment
The Rapidly-exploring Random Tree (RRT) based method has been widely used in robotic exploration, which achieves better performance than other exploration methods in most scenes. However, its core idea is a greedy strategy, that is, the robot chooses the frontier with the largest revenue value as the target point regardless of the explored environment structure. It is inevitable that before a certain area is fully explored, the robot will turn to other areas to explore, resulting in the backtracking phenomenon with a relatively lower exploration efficiency. In this paper, inspired by the perception law of bionic human, a new exploration strategy is proposed on the basis of the prior information heuristic. Firstly, a lightweight network model is proposed for the recognition of the heuristic objects. Secondly, the prediction region is formed based on the position of the
heuristic object, and the frontiers in this region are extracted by the method of image processing. Finally, a heuristic information gain model is designed to guide the robot to explore, which
allocates priority to the frontiers within the heuristic object area, so that the robot can make effective use of the prior knowledge of the room in the scene. Priority is given to the exploration
of one room completely and then to the next, which can greatly improve the efficiency of exploration. In the experimental studies, we compare our method with RRT based exploration method in different environments, and the experimental results prove the effectiveness of our method.
作者:
Lining Sun*,Guodong Chen*,Wenzheng Chi*,Yuan Yuan,Yong Lv†,Jie Liu
A Knowledge-Based Fast Motion Planning Method Through Online Environmental Feature Learning
The sampling-based partial motion planning algorithm has come into widespread application in dynamic mobile robot navigation due to its low calculation costs and excellent performance in avoiding obstacles. However, when confronted with complicated scenarios, the motion planning algorithms are easily caught in traps. In order to solve this problem, this paper proposes a knowledge-based fast motion planning algorithm based on Risk-RRT, which guides motion planning by constructing a topological feature tree and generating a heuristic path from the tree. Firstly, an online topological feature learning method is proposed to simultaneously extract the features during the motion of the robot by means of the dual-channel scale filter and the secondary distance fusion. The learning process is completed until the feature points can represent arbitrary obstacle-free grid points of the whole map. Secondly, the topological feature tree is constructed with environmental feature points and the heuristic motion planning can be carried out on the feature tree. For one map, once the construction of the feature tree finishes, it can be reused as a prior knowledge in the following heuristic motion planning process, which will further improve the efficiency of searching feasible paths. The experimental results demonstrate that our proposed method can remarkably reduce the time taken to find a heuristic path and enhance the success rate of navigation in trapped environments.
作者:
Lining Sun*,Guodong Chen*,Wenzheng Chi*,Jiankun Wang,Jie Liu,Yuan Yuan
