Multibody Dynamics Modeling and Control of Wheelchair Balancing System

Doung, Sokmengkeang (2021) Multibody Dynamics Modeling and Control of Wheelchair Balancing System. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 02111950087002-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (4MB) | Request a copy

Abstract

An electric wheelchair has the potential to increase the mobility of people with disabilities. The four-wheeled wheelchair is the most popular design because it seems to stabilize in every condition, but the problem is spending too much space. Hence, we have a strategy for transforming a four-wheeled wheelchair into a two-wheeled wheelchair. Two-wheeled mobility is the most challenging idea for control of position control for the forward and backward direction.

The purpose of this study is to obtain the dynamic model and control of the self-balancing wheelchair. Based on the wheeled inverted pendulum principle, a state-feedback control is designed to stabilize the wheelchair in the upright position. Simscape Multibody is used to model and simulate the system dynamics. First, the two-wheeled balancing wheelchair system's mathematical model is presented in a Newtonian approach, then linearise to obtain the linear model used to design the controller. The obtained controller is then applied to the nonlinear model. Three types of controller namely LQR, LQR-PID, and LQG are studied. For this study LQR considered all-state variable can be measure while LQG controller uses only two-state variables to be measures is position and tilt angle. An obstacle on the ground and different slope angles are used to evaluate the control system's performance. We also study the robustness of the control by varied the weight of the person.

The simulation result shows that the LQR controller gives the velocity to the wheelchair faster. The drawback of LQR control is that it required that all state variables can be measured. The performance of the LQG controller showed not good, it can't control perfectly while passing through the slop and hill because the gain that we calculated is still not fit with our system. LQR-PID controller is robust than the LQR controller with changing the person's weight. Among three controllers LQR-PID controller exhibits a better performance.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Multi-body modeling, Wheeled inverted pendulum, LQR, LQR-PID, LQG
Subjects:	T Technology > TJ Mechanical engineering and machinery > TJ217 Adaptive control systems
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Mechanical Engineering > 21101-(S2) Master Thesis
Depositing User:	Sokmengkeang Doung
Date Deposited:	31 Jul 2021 09:11
Last Modified:	31 Jul 2021 09:11
URI:	http://repository.its.ac.id/id/eprint/84643

Actions (login required)

View Item