Position Control Method for a Planar Acrobot
Based on Fuzzy Control
CAO Jun-Qing
1
, LAI Xu-Zhi
2
,WUMin
2
1. School of Information Science and Engineering, Central South University, Changsha 410083, P. R. China
2. School of Automation, China University of Geosciences, Wuhan 430074, P. R. China
E-mail: laixz@cug.edu.cn
Abstract: A position control method for a planar Acrobot based on fuzzy control is presented. The system control objective is
to control the endpoint position from any initial position to any reachable target position. Firstly, the state constraint relationship
between the active joint and the passive joint can be obtained by employing the complete integral characteristics of the planar
Acrobot. Next, another constraint relationship between the endpoint position and two angles of the joints can be obtained
according to the physical structure of the planar Acorobt. Then, two target angles of the joints can be searched by using the
particle swarm optimization algorithm based on the above two kinds of constraint relationship. Finally, a fuzzy controller is
designed for the planar Acrobot, and it can achieve the control objective quickly. When the active link is controlled to the target
angle by the fuzzy controller, the passive link is also controlled to its target angle because of the state constraint relationship.
That is, the system control objective is achieved. Simulation results demonstrate the effectiveness and rapidity of the proposed
method.
Key Words:
Planar Underactuated Mechanical System, Planar Acrobot, Position Control, Particle Swarm Optimization Algo-
rithm, Fuzzy Control
1 Introduction
The underactuated mechanical system involves more chal-
lenges than the full actuated mechanical system, because its
control inputs are less than its degrees of freedom
[1,2,3]
.At
present, the vertical underactuated mechanical system (grav-
ity) is widely researched
[4,5]
, and its study has achieved some
results. However, to the planar underactuated mechanical
system (no gravity), its control problem is more complicat-
ed than that of the vertical underactuated mechanical system
because of its linear approximation model at the equilibrium
uncontrollable
[6]
.
The control objective of the planar two-link underactuated
system is usually to move its endpoint position from an ini-
tial position to a target position. And there are two kinds of
planar two-link underactuated system: the planar Pendubot,
which has a passive second joint, and the planar Acrobot,
which has a passive first joint. The concept of small-time
local controllability (STLC) is proposed to solve the posi-
tion control problem of the planar Pendubot
[7]
. Based on the
concept of [7], [8] and [9] have studied tracking control of
the planar Pendubot. But the planar Acrobot is not STLC.
So this system is more difficult to be stabilized at the equi-
librium position. Although there are few reports analyzing
the motion of this system, [10] has studied the integral char-
acteristics of the planar Acrobot and demonstrates the planar
Acrobot is complete integrable. Afterwards, Cao et al.
[11]
present a position control strategy for the planar Acrobot
based the integral characteristics. However, it mainly studies
the control problem with the special system parameters, and
it spends a long time to reach to the control objective.
This paper presents a position control method based on
fuzzy control to quickly achieve the control objective for the
planar Acrobot. Firstly, its dynamic equation is obtained by
using the Lagrangian equation. Then, the state constraint re-
This work is supported by National Natural Science Foundation (NNS-
F) of China under Grant 61374106.
lationship (SCR) is obtained by employing the integral char-
acteristics. Next, another constraint relationship between the
endpoint position and two angles of the joints is obtained
based on the physical structure of the planar Acrobot. The
two target angles of the joints are searched by using the parti-
cle swarm optimization (PSO) algorithm based on the above
two kinds of constraint relationship. Finally, the fuzzy con-
troller is designed to move the active link to the target angle,
and the passive link is also controlled to its target angle be-
cause of the SCR. Hence, the planar Acrobot is controlled to
the target position. Simulation results demonstrate that the
proposed control method not only accomplishes the control
objective, but also spends less time.
2 Characteristics Analysis for Planar Acrobot
In this section, the dynamic equation of the planar Acrobot
is given. Then, the state constraint relationship (SCR) is an-
alyzed according to the integral characteristics of the planar
Acrobot.
2.1 System Model
The physical structure of a planar two-link system with a
passive first joint (planar Acrobot) is showed as Fig. 1.
In Fig. 1, m
i
is the mass of the ith link; L
i
is the length
of the ith link; l
i
is the length from the center of mass of the
ith link to the ith joint; J
i
is the moment of inertia of the
i
th link;
q
i
is the angle of the
i
th joint;
(
x, y
)
is the endpoint
position.(i =1, 2)
According to the Lagrangian equation, the dynamic equa-
tion of the planar Acrobot is written as
M (q)¨q + H (q, ˙q)=τ (1)
where q =[q
1
,q
2
]
T
is the angle vector of the planar Ac-
robot, ˙q is the angular velocity vector and ¨q is the angular
acceleration vector; M (q) is the symmetric positive definite
matrix; H (q, ˙q) is the combination of the Coriolis and cen-
trifugal forces; τ is the vector of driving torque. They are
Proceedings of the 34th Chinese Control Conference
Jul
28-30, 2015, Han
zhou, China
923