Fault detection and fault tolerant control in industrial robotics

Abstract

Fault detection and fault tolerance are more and more significant in recent industrial robotic manipulators, mainly for those who work in inaccessible and risky situation. There are numerous fields of applications of robot manipulators to execute the given task such as mechanical industries, electrical industries, electronics industries, medical applications and varies fields where high accuracy, and stable of operations are required. For this purpose industrial robot manipulators are required to have an exact knowledge of joint factors and kinematic parameters. The key components of the robots are the internal sensors and the motors. Faults in robotic manipulator can affect economic factors in industries and serious damages to the working atmosphere. So the autonomous robots need the capability to identify as well as tolerate failures, allow effective handling with interior failures and stay performing designated jobs without the requirement for instantaneous human intervention. Fault tolerance is beneficial for industrial robots in that it cuts down-time by tolerating failures, finds faulty equipment or subsystems to speed up the repair method and avoids the robot from injuring the products being manufactured. To support these fault tolerant abilities, methods of detecting and tolerating failures must be perfected in robot manipulator. The scope of the modern robot manipulator is to achieve preferred position and placement of the end effector or tool so as to complete pre-defined task. Attaining the above expressed objectives is easily earned through the learning of inverse kinematics. So many researchers have suggested fault detection and fault tolerance architectures for robotic manipulators using the model-based analytical, and redundancy methodology. The foremost problem in the design of fault detection system is to model the rigid link robotic manipulators with modelling uncertainties. In this research, a new approach for fault detection and fault tolerance for industrial robot manipulators is discussed for fault detection and tolerance PUMA 560 robot has been analysed, initially, for finding faults in robotic kinematics is essential. Kinematics is an analytical study of PUMA 560 robotic manipulator. It is a mathematical module to define the position of the end effector. A learning architecture, with kinematic analysis of planar robots using extended Jocobian as online approximates the off-nominal system performance, which is used for monitoring the robotic system for the fault detection. This produces the residual by relating the real output from robot. newline