Robotics Research

Applications: Control and Coordination

A fundamental task in robotic systems is the control and coordination of the various actuation devices required in the generation of motion. While the discipline of control engineering existed well before robotics, the requirements of robotic devices pushed the boundaries of established control methodologies and led to a large amount of research aimed at addressing the strict performance demands of these systems.

Robots require control systems which can achieve excellent performance and stability even when subjected to frequent stopping and starting and varying environmental and loading conditions. The systems must have excellent transient characteristics to cope with frequent changes in actuation, but must also be accurate enough to avoid introducing errors into other elements of the system. They must also be able to cope with noisy sensor data and highly non-linear dynamics.

In order to satisfy these demands conventional linear control techniques have been augmented with a myriad of alternatives, many inspired by biological systems. These include neural networks, genetic algorithms, fuzzy and heuristic systems. A significant amount of current research is concerned with introducing learning and adaptation into control systems, allowing them to operate more effectively in dynamic environments.

Control and Coordination Research in Australia

• Gamini Dissanayake from UTS has been working on enhancements to quay crane control including anti-sway and load micropositioning.
• Gu Fang at UWS is involved in research into robot control and modelling, in particular non-linear control and applications of neural networks.
• Quang Ha from UTS is conducting research into various control problems including tracked vehicle control and robust control and estimation methods.
• Dikai Liu, also from UTS is involved in research into optimal control of automatic transmissions.
• The ACFR is conducting research into control systems for tracked vehicle control, autonomous excavation and robust control of hydraulic systems and lifting gear.
• The Intelligent Systems Group at Deakin University is involved in a large range of control projects, including control of large-scale interconnected systems, robust control of uncertain systems, intelligent process control and heuristic based control.
• Gordon Wyeth from the University of Queensland has been supervising work on joint control in the GuRoo humanoid robot using biologically inspired approaches.

• 

  The GuRoo humanoid at the University of Queensland uses an evolutionary joint control algorithm.

• Koren Ward from the University of Wollongong is working on fuzzy control and genetic algorithms.