Vibration control in flexible systems with multiple modes

Abstract

Industrial robots are widely used in the manufacturing and construction industries. The robotic industry generally aims at having a light system capable of achieving high precision and accuracy in the shortest possible duration. Vibrations are induced within a flexible system, if either the system involves inherently flexible parts or the system involves light-weight construction. The main objective of this dissertation, is to address vibration issues in systems with multiple vibration modes, due to presence of more than one flexible components. For computer controlled systems, one effective feed-forward technique is Input Shaping which makes use of the constructive cancellation principle. Input shaping is basically done by convolving a sequence of impulses with a desired base command, which in turn creates a self cancelling command signal. The input shaping techniques considered in this dissertation are Positive Zero-Vibration shapers, Specified Negative Amplitude Zero-Vibration Derivative-Derivative shapers and the S-curve command function. A rotary multiple-link flexible manipulator was used to analyse and validate the effect of different input shaping. A virtual model of the multiple-link flexible manipulator, along a model of the PM DC motor and an angular positional controller was implemented in a realistic simulation environment provided by MATLAB ® Simscape Multibody ™. Different input shaping techniques were implemented and their effect on the virtual model was analysed through three dimensional animations and vibration graphical representations. Furthermore, the input shaping techniques were digitally implemented on the DS1104 control board using MATLAB ® Simulink ® and ControlDesk D-SPACE software. One can conclude that the most effective and robust input shapers are those that consist of a higher number of impulses namely, the positive convolved shapers, the SNA-ZVDD shaper and the S-curve command as a base function to positive input shapers. The analysis carried out in this dissertation is based on vibration reduction, settling time reduction and robustness of the input shaping techniques.