Pole Environment Detection And Autonomous Grasping With A Biped Climbing Robot

A biped climbing robot is a new kind of climbing robot which can climb in space pole environment. It consists of a multi degree of freedom manipulator for its main body, and the two grippers installed on its two ends. By controlling the two grippers to grasp poles alternately, the robot can climb and transit between two poles. This kind of robot has excellent performances in transition and over-obstacle capacity, with respect to other climbing robot. It can take place of human beings in the high space tasks.In order to achieve autonomous grasping with its target pole and the autonomous climbing in pole environment, the biped climbing robot is equipped with a motion sensing senor called “Prime Sense Carmine 1.09”. The sensor can acquire a pole environment information with a depth image. The algorithms of pole extraction and detection are proposed by processing the depth image of the pole environment. The algorithms provide a basis for autonomous grasping and local pole environment construction. The pole environment detection in this paper is used for autonomous climbing with a biped climbing robot. This paper mainly includes the following parts:First of all, this paper presents a mathematical model and its parameters of graspable pole, and based on its model, the pole extraction and detection algorithms are founded. The depth image of the pole environment provided by Carmine 1.09 is processed, then the data of the target pole information is delivered to robot controlling central computer.Secondly, an efficient solution of autonomous grasping with a biped climbing robot is proposed in this paper. Here, the two different configuration constraints of Climbot in its grasping round pole and square pole are addressed in this section. Combined with the target pole information extracted and detected by using Carmine 1.09, it finally put out the principle of Climbot autonomous grasping and its implement steps.Thirdly, a method for local pole environment constructing is proposed in this section. Considering the complexity of a local pole environment, it simplify the mathematical model of graspable pole, and also the algorithms of pole extraction and detection are improved with regard to the multiple poles condition. Finally it gives a mathematical model of local pole environment and its construction method.Finally, the feasibility and effectiveness of the theories above are verified by experiments, and also their shortcomings and improvement methods are given. Experiments are divided into three parts: the pole extraction and detection experiment, the Climbot autonomous grasping experiment and the local pole environment construction experiment. Research of this paper can also be extended to requirements of pole environment testing in other types of climbing robots.