Thesis Examination Committee

Prof Kai CHEN, CSE/HKUST (Chairperson)
Prof Zexiang LI, ECE/HKUST (Thesis Supervisor)
Prof Yonghua CHEN, Department of Mechanical Engineering, The University of Hong Kong (External Examiner)
Prof Shaojie SHEN, ECE/HKUST
Prof Jungwon SEO, ECE/HKUST
Prof Kai TANG, MAE/HKUST
 

Abstract

Nowadays, the industrial robot is widely used to replace human operators, however, programming the robotic system to carry out the industrial manufacturing is still time-consuming and costly. The thesis studies robotic offline programming (OLP) to reduce downtime and improve efficiency. Firstly, the overall architecture of OLP is studied and divided into four levels, including the data layer, algorithm layer, interface layer and process layer. Then, the steps of offline programming are studied in detail and divided into three phases, including layout, programming and post-processing.
 
Next, tool path generation methods are discussed. The thesis proposes two toolpath generation methods based on the mesh surfaces. These two methods based on mesh intersection and mesh slicing respectively and a comparison related to efficiency and applicability between them are given.
 
Thirdly, motion optimization algorithms are designed for three different kinds of motion. For point to point motions, B-Spline is used to do the interpolation in the joint space concerning time, and then the energy consumption is taken as an objective function to do the optimization. For continues-path motion, it is also formulated as an optimization problem to get the time-optimal and energy-efficiency trajectories. For redundancy-path motion, energy consumption and singularity avoidance are considered when doing the optimization.
 
Finally, the thesis fulfils some applications based on OLP systems. The utilizing of OLP system in the robotic machining and robotic hybrid manufacturing application are analysis. And the real work cells are built to test these applications.