A trajectory learner for sonar based LEGO NXT differential drive robot
Electrical Engineering Congress（2014）
摘要
This paper presents a trajectory learning algorithm for sonar based LEGO Mindstorm NXT differential drive robot. The trajectory learning technique utilizes vehicle odometry and sensor scanning data. The aim of this technique is to find the obstacle free path for mobile robots. Consequently, a free configuration space from the higher dimension sensor data is extracted by employing highest eigen-vectors technique [2] in discrete time scans during robotic manipulation. Integration of these eigenvectors in discrete time results in a new trajectory. This new trajectory is free from dynamic and static obstacles. The trajectory generation algorithm was tested on a nonholonomic LEGO NXT differential drive robot equipped with sonar and position sensors. The trajectory learning proposed method has been tested in different scenarios which resulted into promising preliminary results and are shown in this paper.
更多查看译文
关键词
collision avoidance,distance measurement,drives,legged locomotion,sensors,sonar,trajectory control,discrete time scans,dynamic obstacles,eigen-vectors technique,mobile robots,nonholonomic lego nxt differential drive robot,obstacle free path,position sensors,robotic manipulation,sensor scanning data,sonar based lego mindstorm nxt differential drive robot,sonar sensors,static obstacles,trajectory generation algorithm,trajectory learner,trajectory learning algorithm,vehicle odometry,eigen spaces,non-holonomic,path planning andtrajectory learning,robot kinematics,trajectory
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要