Papers by Author: Wei Liang Cai

Abstract: In order to design a robot end-effector for picking the litchi, at first, the clamping and cutting characteristics of the litchi stem were measured. Based on the biological characteristics and cultivation growth pattern of litchi, the characteristic of litchi picking mechanism was analyzed, and the relationship between the moving mechanism and cutting mechanism was established. At last, a new type of end-effector for picking the litchi, which consisted of clamping mechanism, moving mechanism and cutting mechanism, was developed. The experimental results showed that the end-effector can satisfy the clamping, moving and cutting requirements when picking the litchi.

Abstract: To study the characteristic fruit manipulator in southern China, architecture of fruit manipulator virtual design system was build by the use of modular technology. Parameterize the feature properties of picking manipulator, so the reusable parametric model was build. Based on the database technology, the knowledge base was build, including examples base, constraint rules and research status. Moreover, visualization 3D simulation of the processing of picking was realized under the virtual environment. Finally, through the development of experimental prototype, the virtual design system and its key technologies can be verified.

Abstract: To improve the positioning accuracy of picking manipulator, research of stereo vision positioning for picking object in dynamic was studied. And system composition and positioning principle of stereo vision for vibratory object were introduced. Moreover, experimental platform, which simulated the vibration while picking, was designed for the stereo vision positioning experiment in static condition or vibratory condition. Therefore, influence of vibration condition on the depth information of vision positioning can be analyzed and the regression equation of depth error can be built. The results showed that when the object vibrating, the depth error increased. The vibratory frequency was the most important factor, and the depth error would increase with the frequency increased. The influence of vibratory direction and amplitude on depth error was also significant, but much less than frequency.