Design of perturbation observers and input shaping for sliding mode control of multi-axes-mechanisms
Date of Completion
January 1997
Keywords
Engineering, Mechanical
Degree
Ph.D.
Abstract
The problem of robust motion control of multi-axes-mechanisms (MAM) is studied within the framework of variable structure systems (VSS). A new motion control algorithm is developed for a representative class of such nonlinear systems.^ The main feature of the proposed algorithm is its ability to estimate and compensate for modelling uncertainties, utilizing partial state feedback. The influence of the uncertainties on the dynamics is measured using the concept of perturbation vector. A variable structure observer (VSO) is introduced in order to estimate the state and the perturbation vectors. The proposed observer is integrated into a variable structure controller (VSC). This new combination VSC/VSO forms the main contribution of this work. Its stability is assured by appropriately selecting the design parameters in the algorithm.^ A second contribution of this work is a simple input shaping technique in order to achieve an optimum compromise between tracking accuracy and unmodelled dynamics excitation. Given a desired motion for the MAM, it is shown that there is always a modified path for the state vector to follow so that this optimum compromise is achieved. Potential applications of this routine include motion control of flexible manipulators.^ Comparison with state-of-the-art techniques are presented through simulations and experiments. A two link SCARA type and a PUMA 560 manipulators are used for the experiments. A special criterion is developed in order to perform meaningful comparisons. It is based on the disturbance rejection feature of the closed loop dynamics. Utilizing this criterion, the superiority of the proposed algorithm is shown. ^
Recommended Citation
Moura, Jairo Terra, "Design of perturbation observers and input shaping for sliding mode control of multi-axes-mechanisms" (1997). Doctoral Dissertations. AAI9730892.
https://digitalcommons.lib.uconn.edu/dissertations/AAI9730892