Study on Controlling Mechanical Manipulator for Timering Tunnel Support

点击翻译按钮获取中文摘要

The present investigation deals with a hybrid controlling method for a mechanical manipulator which works for timbering of tunnel support. The manipulator was equipped with two cameras and a force and moment sensor consisting of four elements to implement visual and tactile functions.The vision systems allow the manipulator to find a support member placed on the foor from an image taken by a camera, and also to determine a destination, on which the work piece should be hang, from an image detected by another camera. A new scanning method of binary image data was developed to determine fast geometrical features of a object such as its centroid and inertia axis, resulting in a considerable reduction in time for image processing. The force and moment sensor complements the vision systems when the manipulator approaches, grasps and releases the work piece, by sensing contact with the object or surroundings.A laboratory experiment was carried out to validate the hybrid controlling method which combines a position control scheme with a force control scheme. The experimental result confirmed that the mechanical mamipulator with visual and tactile functions could manipulate skillfully timbering of tunnel support.The present investigation also includes planning a path along which the work piece grasped by the manipulator is free from collision with the wall of tunnel and the manipulator itself. The investigation concludes the manipulator can avoid any collisions whenever it carries the work piece along a spiral trajectory starting from the grip point and terminating at the release point.