Research and development of a hybrid kinematic platform, with associated control methodology, for several low degree of freedom serial kinematic manipulators.

Abstract

Industry has placed focus on efficiency and flexibility, for this reason it is beneficial to propose a method in which serial manipulators can be reoriented into a parallel system when needed. This reorientation allows industry to implement less robotic manipulators, which reduces cost and space requirements; while still being able to receive the benefits that are associated with serial and parallel robotic manipulators. This research focuses on the development of a hybrid kinematic platform that consists of the combination of several low degree of freedom Serial Kinematic Manipulators (SKM) into a Parallel Kinematic Manipulator (PKM). Special attention was placed on the initial selection of the serial manipulators to be combined into a hybrid SKM PKM platform. The aim of the hybridization is to create the ability to have several serial manipulators that can operate within a collaborative workspace for tasks that require dexterity, such as assembly operations. As well as, become an individual parallel mechanism to perform tasks that require increased rigidity, such as machining. Several desktop models were produced via rapid prototyping methods. An electronic system that included stepper motors, limit switches, Arduino, RAMPS breakout board and motor drivers were installed onto each RobotArm. Several design optimizations were implemented to improve the potential of the final platform. Full serial kinematic analysis was performed on the individual serial RobotArms, including full 3D forward and inverse derivations via a geometric approach. A parallel inverse kinematic derivation was performed for the combined system, which includes a rotation matrix. The rotation matrix was required due to the added degrees of freedom that became available in the combined hybrid kinematic platform. The kinematics were used for the control system that were implemented via Arduino and a Python GUI. The kinematics also allowed for investigation into the workspace, and the singularities present in that workspace. Both the workspace and the singularity plots were performed via Monte Carlo randomization scripts performed in Octave/MATLAB and Python. Testing was then performed to identify how well the platform works as individual collaborative serial manipulators as well as a single combined hybrid platform. The tests focused on the accuracy and repeatability present within the different combination of the platform. It was found that the combined hybrid kinematic platform outperformed the individual serial kinematic manipulator in accuracy as well as displaying a lower level of deflection to load. The combined hybrid kinematic platform was able to achieve three further degrees of freedom compared to the individual serial kinematic manipulator. The results obtained validate the research question of whether it is possible to combine several low degree of freedom serial kinematic chains into a single hybrid platform and how such a platform would be controlled.