Fuzzy PD control of an optically guided long reach robot

This thesis describes the investigation and development of a fuzzy controller for a manipulator with a single flexible link. The novelty of this research is due to the fact that the controller devised is suitable for flexible link manipulators with a round cross section. Previous research has concentrated on control of flexible slender structures that are relatively easier to model as the vibration effects of torsion can be ignored. Further novelty arises due to the fact that this is the
first instance of the application of fuzzy control in the optical Tip Feedback Sensor (TFS) based configuration.
A design methodology has been investigated to develop a fuzzy controller suitable for application in a safety critical environment such as the nuclear industry. This methodology provides justification for all the parameters of the fuzzy controller including membership fUllctions, inference and defuzzification techniques and the operators used in the algorithm. Using the novel modified phase plane method investigated in this thesis, it is shown that the derivation of complete, consistent and non-interactive rules can be achieved. This methodology was successfully applied
to the derivation of fuzzy rules even when the arm was subjected to different payloads. The design approach, that targeted real-time embedded control applicat.ions from the outset, results in a controller implementation that is suitable for cheaper CPU constrained and memory challenged
embedded processors.
The controller comprises of a fuzzy supervisor that is used to alter the derivative term of a linear classical Proportional + Derivative (PD) controller. The derivative term is updated in relation to the measured tip error and its derivative obtained through the TFS based configuration. It is shown that by adding 'intelligence' to the control loop in this way, the performance envelope of the classical controller can be enhanced. A 128% increase in payload, 73.5% faster settling time and a reduction of steady state of over 50% is achieved using fuzzy control over its classical counterpart.