MRAS trực tiếp với các điều kiện an toàn ứng dụng điều khiển robot di động hai bánh

Dinh Thi Gia et al Journal of SCIENCE and TECHNOLOGY 127(13): 21 - 28

21

DIRECT MRAS WITH SAFE CONSTRAINTS APPLIED

FOR TWO-WHEELED MOBILE ROBOT

Dinh Thi Gia, Tuan Manh Tran, Son Que Tran*

University of Technology – TNU

ABSTRACT

Most two-wheeled mobile robots (TWMR) are controlled and moved by two DC motors. The

heading angular velocity depends on the changing velocity of two wheels mounted on the two DC

motors respectively. During moving, if the heading angular velocity and the linear velocity are too

high, it can lead to flip and slide phenomena. In addition, under the effects of noises (internal and

external), TWMR may be unstable. To solve these problems, we use Euler-Lagrange method to

model for TWMR, build safe conditions against flip, then apply the Model Reference Adaptive

System (MRAS) to construct an adaptive controller for TWMR to ensure the required motion,

stability and safety. Simulation results and analysis point out the effectiveness of the designed

controller.

Keywords: Direct MRAS, Two wheeled mobile robot.

INTRODUCTION *

Two-wheeled mobile robot is shown in Fig. 1

including two wheels, a chassis and a

pendulum. In fact, TWMR - a nonlinear,

unstable and underactuated system - is built

based on the principle of the inverted

pendulum dynamics. To model TWMR, two

widely used methods are Newton and Euler￾Lagrange [1]. With this configuration, It has

been considered as anuseful prototype for

representing nonlinear systems when testing

control algorithms.

To design control for TWMR, the moments

which put into two wheels to control

movement and stability are computed.

When designing controller, the following

parameters are interested: the title angle is

stable at the reference and there is no

overturn while TWMR moving. Although

the system is unstable, difficult to control,

the TWMR is usually used because of the

ability to move in tight space, various terrain

and sharp corners [2].

After linearization, ignoring nonlinear,

coupling attributes, the linear algorithms as

PID, MRAS, etc are applied because they are

* Tel: 0988039336; Email: [email protected]

quite simple, quick converge, and have small

area stability. On the other hand, the

nonlinear algorithms are complexity, huge

computation, and long response time, but they

have the larger area of stability. However,

under the effect of disturbance, most

conventional controllers can not warrant the

robust performance of system. Normally,

adaptive controller would be the best choice

for this case. It can be easily seen that when

TWMR is drived by human, the TWMR is

affected by unknown forces or disturbance.

This domination is one first daresay in safe

control. The second task must be concerned

that the suitable controller must guarantee that

there are no overturn happening with human

and TWMR. It is quite scarce to find a

controller solving with human safety accept

for author in [2]. In this publication, the

author used a reduced-order disturbance

observer to estimate the disturbance acting on

the TWMR. This estimates disturbance and

compensates in the controller to reduce the

error signal.

In this paper, the model of TWMR is

expanded to three dimensions by using

unequal torque acting on each wheel of

TWMR. It clearly seen that TWMR will