| Abstract |
Recently Delta robots parallel system are gaining more attention from researchers to investigate studies of its kinematics and dynamics. In order to build an autonomous system has the ability to perform the desired tasks without any interference from humans, it is necessary to provide the control of the system by the kinematic and dynamics models as well as besides providing it by the control task to be performed. A very common situation when dealing with robots especially the serial structured robots is the autonomy of operation, researchers paid a lot of attention to this type of robots to make the robots to be fully automated and to have the ability to take decisions in the different situations without any interaction from humans. Few researchers tried to evolve the delta robots in such these systems.
This work tries to focus on the parallel systems to make benefits from their great advantages of accuracy and stiffness…etc. This work aimed to add some intellectuality to this type of robots, to illustrate the potential of delta robots, it has the ability to perform 200 operating cycles per minutes which could be used for inspection of products in an industrial application such as PCB boards’ test which consist of millions of electronic circuits to be tested at a short time.
Adding the feature of object detection and tracking to this type of parallel robots enlarge the scope of use of them. First a kinematics analysis of the end effector will be studied, and this work will try to find out a suitable kinematic procedure to be applied in an autonomous system using delta robots. Among the methods of modeling a system, Multibody Systems (MBS) modeling is the only method which enables a control algorithm to be aware about the kinematics and dynamics of all the bodies which the system consists of. This work aims to integrate the modeling methods using the MBS modeling in conjunction with the object detection methods to implement an autonomous parallel robot system.
To handle and pick up an arbitrary object which moves with unstructured motions in a camera video surveillance, it needs to be observed by means of some tracker system, then an accurate control signal based on the information obtained by means of camera sensor, would be able to steer the end effector of a parallel robot to the exact location where is the object exist.
This research focus on handling an object by means of delta robot system by applying a suitable object detection algorithm which should not require complex computations in order to make advantage of the massive productivity of this type of robots, as the on line control of this system should have a good response to the sudden abrupt motions made by the object to be tracked. This situation of moving objects may not be in high demand nowadays, however this research aims to make a step forward to widen the scope of use from this type of robots.
|
