| The development of robot technology and the extension of application fields bring new challenges to the environmental adaptability of robot systems. With the outstanding obstacle-overcoming performance, small jumping robots are very suitable for narrow space and multi-obstacle environment. But limited by small size and battery capacity, it can not keep long time and long distance working. Besides, a single kind of jumping robots may not be able to complete the task in complex environments. Therefore, base on a legged jumping robot early developed in our laboratory, a carrying and detecting platform is designed, forming a marsupial robot system. The relative orientation and position detecting method between the mother robot and the child robot and the recycling strategies for the child robot are researched in this paper.The overall system is designed firstly, including a bigger carrying robot and a smaller jumping robot. For the jumping robot, two kinds of self-righting method and steering method are proposed, which enhances its locomotion performance effectively. A disc-like label with two different colored sides is mounted on the top of the jumping robot, which is used for orientation and position detecting by the carrying robot. The sensing and control board is designed with the function of wireless communication for the interaction with the operator or the carrying robot remotely. The hardware of carrying robot is composed of body structure module, RGB-D information feedback module, wireless communication module and the monitoring module.Then, based on the hardware design, the strategies for the jumping robot orientation and position detecting and recycling are designed. Based on the color recognition algorithm, the relative heading angle detecting methods of the carrying robot and the jumping robot are proposed. Based on the alignment principle of the color image and depth image, the distance between the two robots is obtained. In particular, an anti-interference detecting method based on the dynamic cooperation of the two robots is proposed, which improves the success rate for the jumping robot recycling.Finally, the experimental studies of the basic locomotion performances of the two robots and the related detecting method are conducted. The new jumping robot with the adding weight can jump about 75cm in height and traverse 30cm far. The self-righting and steering capabilities of the jumping robot are verified in the flat ground and outdoor uneven environment. At the set speed ratio, the horizontal and vertical deviations of the carrying robot are less than 1.5cm in the linear motion, while in the rotational motion test, the turning error is less than 5°. The maximum error of relative heading angle detecting of the carrying robot and the jumping robot are 2.9° and 7.4° respectively. The distance detecting error between the two robots is within 2cm. The experimental results for carrying and recycling the jumping robot show that the carrying robot can detect the jumping robot within 2m range and sense its relative orientation and position at a success rate of 92%, which verifies the effectiveness of the algorithm and the feasibility of the proposed system. |