1. Trang chủ
  2. » Thể loại khác

SM PID controller using fuzzy tuning approach for manipulator

5 250 0

Đang tải... (xem toàn văn)

THÔNG TIN TÀI LIỆU

Thông tin cơ bản

Định dạng
Số trang 5
Dung lượng 107,52 KB

Các công cụ chuyển đổi và chỉnh sửa cho tài liệu này

Nội dung

One simple way to decrease the sensitivity of sliding mode controller to the parametric uncertainties and external disturbances is using of high control gain which decrease also the reac

Trang 1

1 INTRODUCTION

A robot manipulator is a nonlinear system with high

coupling term whose dynamics consists of uncertainty and

encountered with payload changes, friction and disturbance

[1] On the other hand, sliding mode control (SMC) as a

nonlinear technique with the capabilities of robustness

against the model uncertainties and ability of the disturbance

rejection has been considered in many researches [2-4]

Although the robustness of the SMC is one of its main

characteristics, this is achieved only in the sliding phase and

the system is sensitive to the structured uncertainties and

external disturbances in the reaching phase to the sliding

surface Therefore, different approaches for improving the

performance of the SMC has been proposed which one of

them is intelligent control method such as fuzzy control

system [5-7] Because of the relations between SMC and

fuzzy control, [8], the combination of these two approaches

has been considered as a research topic in last years [9-13]

such that the advantages of both approaches can be used

One simple way to decrease the sensitivity of sliding mode

controller to the parametric uncertainties and external

disturbances is using of high control gain which decrease

also the reaching time and tracking error However, high

control gain increases the oscillations in the control signal

that may lead to the excitation of high frequency unmodeled

dynamics which is an undesired phenomenon To overcome

this drawback, the fuzzy logic can be used for tuning of this

gain In this regard, in [14], a nonlinear sliding surface and

fuzzy logic have been used in the design of a fuzzy terminal

SMC for a robot manipulator Also, a fuzzy variable sliding

surface based method has been proposed in [15] in order to

improve the tracking performance In [16], in addition to

using variable sliding surface, the idea of fuzzy gain tuning

and boundary layer has been presented to achieve more improvements

In this paper, in addition to using the integral term in the sliding surface, [17, 18], at first, the SMC witch is including PID part is designed and its stability guarantee is proved in a lemma Then, in order to improve the controller performance, a fuzzy system is used to tune the gain of reaching phase and also PID part gain Thus, a chattering free SMC is achieved in which the tracking error and reaching time to sliding surface has been reduced without need to variable sliding surface

The reminder of the paper is organized as follows In the section 2, the mathematical model of the robot manipulator

is given The SMC including the PID loop to which is denoted as SMC-PID is presented in section 3 The design

of fuzzy SMC-PID is described in section 4 In section 5, the simulation results are provided and finally, summary and some conclusions are presented

2 THE SYSTEM MATHEMATICAL MODEL The dynamical equation of an n-link robot manipulator in the standard form is as follows [1]:

τ

+ + +C q q q G q d q

q

where M(q)∈Rn×n is a symmetry and bounded positive definite matrix which is called inertial matrix Moreover,

n

R q q

q , , ∈ are the position, velocity, and angular acceleration of the robot joint, respectively The matrix

n n

R q q

C( , )∈ × is the matrix of Coriolis and centrifugal forces such that the matrix H(q)−2C(q,q)is asymmetry, i.e., for a nonzero n×1 vector x we will have:

Sliding Mode PID-Controller Design for Robot Manipulators by Using

Fuzzy Tuning Approach

Mohammad Ataei1, S Ehsan Shafiei2

1 Electronic Department- Engineering Faculty, University of Isfahan, Isfahan, Iran

E-mail: mataei1971@yahoo.com

2 Electrical and Robotic Engineering Faculty, Shahrood University of Technology, Shahrood, Iran

E-mail: sehshf@yahoo.com

Abstract: In this paper, a chattering free sliding mode control (SMC) for a robot manipulator including PID part with a

fuzzy tunable gain is designed The main idea is that the robustness property of SMC and good response characteristics

of PID are combined with fuzzy tuning gain approach to achieve more acceptable performance For this purpose, in the

first stage, a PID sliding surface is considered such that the robot dynamic equations can be rewritten in terms of sliding

surface and its derivative and the related control law of the SMC design will contain a PID part The stability guarantee

of this sliding mode PID-controller is proved by a lemma using direct Lyapunov method Then, in the second stage, in

order to decrease the reaching time to the sliding surface and deleting the oscillations of the response, a fuzzy tuning

system is used for adjusting both controller gains including sliding controller gain parameter and PID coefficient

Finally, the proposed methodology is applied to a two-link robot manipulator including model uncertainty and external

disturbances as a case study The simulation results show the improvements of the results in the case of using the

proposed method in comparison with the conventional SMC

Key Words: Sliding Mode Control, Robot manipulator, PID control, Fuzzy control, Lyapunov theory

Trang 2

0 )]

, (

2

)

(

[H q − C q q x=

gravity vector, τd∈Rn is the bounded disturbance vector

such that τd ≤TD andτ∈Rn is the control input vector

In the following,H(q), C(q,q) and G(q) are shown by

H, C, and G, respectively

3 SLIDING MODE CONTROL WITH PID

The objective of tracking control is design a control law for

obtaining the suitable input torque τ such the position

vector q can track the desired trajectoryqd In this regard,

the tracking error vector is defined as follows:

q q

e= d − (2)

In order to apply the SMC, the sliding surface is considered

as the relation (3) which contains the integral part in

addition to the derivative term:

+ +

=e e tedt s

0 2

1 λ

λ (3) whereλi is diagonal positive definite matrix Therefore,

0

=

s is a stable sliding surface and e→0as t→∞ The

robot dynamic equations can be rewritten based on the

sliding surface (in term of filtered error) as follows:

τ

+ +

= Cs f d s

Where

G edt e

q C e e q

M

f = ( d + + )+ ( d + + t )+

0 2 1 2

(5) Now, the control input can be considered as follows:

) sgn(

f + v +

=

τ (6)

Where

G edt e

q C e e q

M

0 2 1 2

+

(7)

+ +

K

0

λ

tracking loop, and Kv,K are diagonal positive definite

matrices and are defined such that the stability conditions

are guaranteed The sgn(s) is also the sign function

We have also:

F G edt e

q C e e q

M

~

0 2 1 2

λ

(8) where f~= f − fˆ , M~ =M −Mˆ , C~=C−Cˆ ,and

G

G

G~= − ˆ F can also be selected as the following

relation:

G edt e

q C e e q

M

(

~ ) (

~

0 2 1 2

+

(9)

In order to reach the system states ( ee, )to the sliding

surface s=0in a limited time and remain there, the control

law should be designed such that the following sliding condition is satisfied [2]:

2

1

s s Ms

s dt

η

< for s≠0 (10) This subject is proved in the following lemma

Lemma- In the SMC design of a system with dynamic equation (1) and sliding surface (3), if the control input τ is selected as (6), by considering F as (9) and

) , , , (K11 K22 Knn diag

components:

K

i D v

Then, the sliding condition (10) is satisfied by equation (4) Proof: Consider the following Lyapunov function candidate:

Ms s

2

1

= (12) Since M is positive definite, for s≠0we have V >0and

by differentiating of the relation (12) and regarding the symmetric property of M, it can be written:

s M s s M s

V = T + T 2

1

(13)

By substituting (4) in (13) and considering that

0 ) 2 (M − C s=

) (

) (

2 1

τ τ

τ τ

− +

=

− + +

=

d T

d T

T T

f s

f s Cs s s M s V

(14)

By replacing the relation (6) in (14), Vcan be rewritten as:

=

− +

=

− +

=

n i

i ii v

d T

v d

T

s K s

K f

s

s K s K f f

s V

1 )

~ (

)) sgn(

ˆ (

τ

τ

(15)

Since the following inequality (16) is valid and regarding the relation (11), we have:

s K f

T s K

F+ v + D ≥ ~+τd − v (16)

i i v d

K ≥ [~+τ − ] +η (17) Finally, it can be concluded that:

=

≤ n i i

i s V

1

η (18)

This indicates that V is a Lyapunov function and the sliding condition (10) has been satisfied

The use of sign function in the control law leads to high oscillations in control torque which is undesired phenomenon and is called chattering To overcome this drawback, there are some solutions which one of them is using the following saturation function instead of sign function in the discontinuous part of the control law:

Trang 3

<

<

=

ϕ

ϕ ϕ ϕ

ϕ

ϕ

s

s s

s s

sat

1

1

(19)

By this, there is a boundary layer ϕ around the sliding

surface such that when the state trajectory reach to this layer

will be remaining there

4 THE DESIGN OF FUZZY SMC-PID

As it was mentioned before, by using a high gain in SMC

(K), the sensitivity of the controller to the model

uncertainties and external disturbances can be reduced

Moreover, a high gain in PID part of the control system

)

(Kv can reduce the reaching time to sliding surface and

tracking error However, increasing the gain causes the

increment of the oscillations in the input torque around the

sliding surface Therefore, if this gain can be tuned based on

the distance of the states to the sliding surface, a more

acceptable performance can be achieved In other words, the

value of gain should be selected high when the state

trajectory is far from the sliding surface and when the

distance is decreasing, its value should be decreased This

idea can be accomplished by using fuzzy logic in

combination with SMC to tune the gain adaptively

For this purpose, two-input one-output fuzzy system is

designed whose inputs are s and swhich are the distances

of the state trajectories to the sliding surface and its

derivative, respectively The membership functions of these

two inputs are shown in figure (1) The output of the fuzzy

system is denoted by Kfuzzand has been shown in figure

(2) For applying these gains to the control input, the

normalization factors N and Nvas the following relations

are used:

fuzz K N

K= ⋅ (20)

fuzz v

v N K

K = ⋅ (21) These factors can be selected by trial and error such that the

stability condition (17) is satisfied

0

0.2

0.4

0.6

0.8

1

input variable "s"

(a) Fig 1: The membership functions for a) input s

0 0.2 0.4 0.6 0.8 1

inpu variable sd

(b) Continue Fig 1: The membership functions for b) input s

0 0.2 0.4 0.6 0.8 1 0

0.2 0.4 0.6 0.8 1

output variable K

Fig 2: The membership functions of the output Kfuzz

Tab 1: The fuzzy rule base for tuning Kfuzz

PB

PS

Z

NS

NB

s s

B

S

M

B

B

N

B

M

S

M

B

Z

B

B

M

S

B

P

The maximum values of K and Kv are limited according to the system actuators power, and the minimum value of K should not be less than the provided amount in relation (17) The fuzzy base rule has been given in table (1) in which the following abbreviations have been used: NB: Negative Big; NS: Negative Small; Z: Zero; PS: Positive Small; PB: Positive Big; M: Medium For example, when s is negative small (NS) and sis positive (P), then Kfuzzis small (S)

5 THE CASE STUDY AND SIMULATION RESULTS

In order to show the effectiveness of the proposed control law, it is applied to two-links robot with the following parameters:

+

+ +

+

=

β γ

β

γ β γ

β α

2

2 2

cos

cos cos

2 )

(

q

q q

q

+

=

0 sin

sin ) (

sin )

, (

2 1

2 2 1 2

2 q q

q q q q

q q

q C

γ

γ γ

(23)

Trang 4

+ +

=

) cos(

cos(

cos )

(

2 1 1

) 2 1 1 1 1

q q

q q q

q

G

γδ

γδ αδ

(24) where α =(m1+m2)a12 , β=m2a22 , γ =m2a1a2 ,

1

a

g

=

δ , and m1, m2, a1=.7, a2 =.5are the masses

and lengths of the first and second links, respectively The

masses are assumed to be in the end of the arms and the

gravity acceleration is consideredg=9.8 Moreover, the

masses are considered with 10% uncertainty as follow:

2 ,

4 ,

2 2

2 2

1 1

1 1

0

0

∆ +

=

∆ +

=

m m

m m

m m

m m

(25)

0

0

m , and Mˆ , Cˆ , and Gˆ are estimated The desired state trajectory is:

=

t

t

qd

π

π

cos 2

cos 1

(26) and the disturbance torque is considered as follows:

=

t

t

d

π

π τ

2 sin 5 0

2 sin 5 0

(27)

which leads to =

5 0

5 0

D

The design parameters are determined as follow:

=

15 0

0 15

1

40 0

0 40

2

λ (28) The values of the ϕ and η are selected as

167

0

=

1 0 1 0

=

η Moreover, the factors N

and Nvare selected as follow:

=

5 0

0 50

10 0

0 5

v

N (29)

In order to show the improvement due to the proposed

method of this paper (Fuzzy SMC-PID), the simulation

results of applying this method are compared with the

related results of the conventional SMC The tracking error

and control law in the case of conventional SMC have been

shown in figures (3) and (4), respectively The

corresponding graphs for the case of applying fuzzy

SMC-PID are also provided in figures (5), and (6)

-0.05

0

0.05

0.1

0.15

time(sec)

-0.5

0

0.5

1

1.5

2

time(sec)

Fig 3: The tracking errors in the case of using conventional SMC

-50 0 50 100 150

time(sec)

-50 0 50 100

time(sec)

Fig 4: The control inputs in the case of using conventional SMC

-0.05 0 0.05 0.1 0.15

time(sec)

-0.5 0 0.5 1 1.5 2

time(sec)

Fig 5: The tracking errors in the case of using Fuzzy SMC-PID

-100 0 100 200

time(sec)

-100 -50 0 50 100

time(sec)

Fig 6: The control inputs in the case of using Fuzzy SMC-PID

As it is seen in these figures, the proposed fuzzy SMC-PID has faster response and less tracking error in comparison with conventional SMC In order to show more clearly the difference between the tracking errors in two cases, the enlarged graphs have been provided in figures (7) and (8)

Trang 5

0 2 4 6 8 10

-0.01

-0.005

0

0.005

0.01

time(sec)

-5

0

5x 10

-3

time(sec)

Fig 7: The enlargement of the tracking errors in the case of using

conventional SMC

-5

0

-4

time(sec)

-1

-0.5

0

0.5

-3

time(sec)

Fig 8: The enlargement of the tracking errors in the case of using

Fuzzy SMC-PID

6 CONCLUSION

In this paper, design of a sliding mode control with a PID

loop for robot manipulator was presented in which the gain

of both SMC and PID was tuned on-line by using fuzzy

approach Then the stability guarantee of the system was

proved by direct Lyapunov method The proposed

methodology in fact tries to use the advantages of the SMC,

PID and Fuzzy controllers simultaneously, i e., the

robustness against the model uncertainty and external

disturbances, quick response, and on-line automatic gain

tuning, respectively Finally, the simulation results of

applying the proposed methodology to a two-link robot

were provided and compared with corresponding results of

the conventional SMC which show the improvements of

results in the case of using the proposed method

REFERENCES

[1] M W Spong, and M Vidiasagar, Robot Dynamics and

Control, Wiley, New York, 1989

[2] J J Slotine, and W Li, Applied Nonlinear Control,

Englewood Cliffs, NJ: Prentice Hall, 1991

[3] W Perruquetti, and J P Barbot, Sliding Mode Control in

Engineering Marcel Dekker, Inc New York, 2002

[4] J Y Hung, W Gao, and J C Huang, Variable Structure Control: A Survey, IEEE Trans Ind Elec., Vol 40, No 1, 2-22, 1993

[5] L X Wang, A Course in Fuzzy Systems and Control, Prentice Hall, NJ, 1997

[6] C C Lee, Fuzzy Logic in Control Systems: Fuzzy Logic Controller-Part I, IEEE Trans Sys Man and Cyb Vol 20,

No 2, 404-418, 1990

[7] C C Lee, Fuzzy Logic in Control Systems: Fuzzy Logic Controller-Part II, IEEE Trans Sys Man and Cyb Vol 20,

No 2, 419-435, 1990

[8] R Palm, D Driankov, and H Hellendoorn, Model Based Fuzzy Control: Fuzzy Gain Schedulers and Sliding mode Fuzzy Controllers Springer-Verlag Berlin Heidelberg, 1997 [9] J C Lo, and Ya H Kuo, Decoupled Fuzzy Sliding Mode Control, IEEE Trans on Fuzzy systems, Vol 6, No 3, 426-435, 1998

[10] L K Wang, H F Leung, and K S Tam, A Fuzzy Sliding Controllers for Nonlinear Systems, IEEE Trans Ind Elec., Vol 48, No 1, 32-37, 2001

[11] Q P Ha, Q H Nguyen, D C Rye, and H F Durrant-Whyte, Fuzzy Sliding Mode Controllers with Applications, IEEE Trans Ind Elec., Vol 48, No 1, 38-46, 2001

[12] Z Lin, Q Zhu, and Y Yan, Variable Structure Control Based

on Fuzzy Law for Underactuated Robot Manipulators, IEEE Int Conf on Mechatronics and Automation, 1908-1913,

2006

[13] C M Lin, and Y J Mon, Decoupling Control by Hierarchical Fuzzy Sliding Mode Controller, IEEE Trans Cont Sys Technology, Vol 13, No 4, 593-598, 2005

[14] Y C Huang, and T H S Li, Fuzzy Terminal Sliding Mode Controller for Robotic Manipulators, IEEE Int Conf on Mechatronics, 858-863, 2005

[15] Q P Ha, D C Rye, and H F Durrant-Whyte, Fuzzy Moving Sliding Mode Control with Application to Robotic Manipulators, Automatica, Vol 35, 607-616, 1999

[16] H Javaheri, and G R Vossoughi, Sliding Mode Control with Online Fuzzy Tuning: Application to a Robot Manipulator, IEEE Int Conf on Mechatronics & Automation, 1357-1362,

2005

[17] T C Kuo, Y J Huang, A Sliding mode PID-Controller Design for Robot Mnipulator, IEEE Conf On Computational Intelligence in Robotics and Automation, 625-629, 2005 [18] I Eker, Sliding mode control with PID sliding surface and experimental application to an electromechanical plant, ISA Trans Vol 45, No 1, 109-118, 2006

Ngày đăng: 24/11/2016, 00:04