Large scale modeling, model reduction and control design for a real-time mechatronic system

Qiong Li, Purdue University

Abstract

Mechatronics is the synergistic integration of the techniques from mechanical engineering, electrical engineering and information technology, which influences each other mutually. As a multidisciplinary domain, mechatronics is more than mechanical or electronics, and the mechatronic systems are always composed of a number of subsystems with various controllers. From this point of view, a lot of such systems can be defined as large scale system. The key element of such systems is integration. Modeling of mechatronic system is a very important step in developing control design of such products, so as to simulate and analyze their dynamic responses for control design, making sure they would meet the desired requirements. The models of large scale systems are always resulted in complex form and high in dimension, making the computation for modeling, simulation and control design become very complicated, or even beyond the solutions provided by conventional engineering methods. Therefore, a simplified model obtained by using model order reduction technique, which can preserve the dominant physical parameters and reveal the performance limiting factor, is preferred. In this dissertation, the research have chosen the two-wheeled self-balancing scooter as the subject of the study in research on large scale mechatronic system, and efforts have been put on developing a completed mathematical modeling method based on a unified framework from varitional method for both mechanical subsystem and electrical subsystem in the scooter. In order to decrease the computation efforts in simulation and control design, Routh model reduction technique was chosen from various model reduction techniques so as to obtain a low dimensional model. Matlab simulation is used to predict the system response based on the simplified model and related control design. Furthermore, the final design parameters were applied in the physical system of two-wheeled self-balancing scooter to test the real performance so as to finish the design evaluation. Conclusion was made based on these results and further research directions can be predicted

Degree

Ph.D.

Advisors

zhang, Purdue University.

Subject Area

Electrical engineering|Mechanical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS