You are on page 1of 4

2012 UKSim-AMSS 6th European Modelling Symposium

Design and Tuning Approach of 3-DOF Emotion Intelligent PID(3-DOF-PID)


Controller

Dong Hwa Kim


Dept. of Electronic and Control Eng., Hanbat National University
16-1 San Duckmyong-Dong Yuseong-Gu
Daejon City, Korea
e-mail: koreahucare@gmail.com

Abstract—This paper suggests the 3-DOF PID Controller with species typical reactions to environmental changes (Damasio,
emotion function and tuning approach. The PID controller has 1994; Sloman, 2003). So, the emotion technology should
been widely used owing to its simplicity and robustness in study for coming technology. That is, emotion technology
various systems. However, the classical PID controller cannot will be used in various IT technology or design. Neural,
both function such as set-point following and disturbance. To neuro-fuzzy and evolutionary system and their derivatives
recover, the 2-DOF (Two-Degree Of Freedom) controller with followed later. Recently, biological information processing
a 2-DOF (Two-Degree Of Freedom) function controller which systems such as human beings have many interesting
has both set point-following function and disturbance-rejecting functions and are expected to provide various feasible ideas
function, at the same time has been suggested. On the other
to engineering fields [10]. Biological information such as
hand, emotion function is an important role in human being.
However, many researchers have been dealing its structure
immune system in living organisms can be mainly classified
and modeling. Herein, this paper suggests novel 3-DOF-EI-PID into the following four systems: brain nervous, genetic
controller with emotion function(3-DOF-EI-PID: 3-Degree Of system, endocrine system. Also, they are the important for
Freedom-Emotion Intelligent-PID). Also, this paper suggests artificial intelligence. This paper suggests the scheme and
tuning technique of the 3-DOF-EI-PID controller by immune tuning approach of 3-DOF Emotion Intelligent PID
algorithm. controller (3-DOF- EI-PID).

II. THE DESIGN OF 3-DOF EMOTION INTELLIGENT PID


Keywords-3-DOF-EI-PID control; Emotion control; Immune CONTROLLER DESIGN (3-DOF- EI-PID)
algorithm; PID control; 2-DOF PID control; Intelligent control; This type of the 3-DOF-EI-PID controller has a combined
Intelligent tuning. parameter for a 3-EI-DOF function as shown in Fig. 1 The
transfer functions between process value PV(s) and settling
I. INTRODUCTION value SV(s), and between process value PV(s) and
Emotion technology and PID controller is very useful one. disturbance DV(s) are given in the following equations,
However, emotion technology has some difficulty for respectively:
implementation in actual system. But PID control has been
using from long time ago. The proportional-integral- PV ( s) Gd ( s)
GPVDV ( s)   (1)
derivative (PID) controllers have been widely used owing to DV ( s)  1 
1  K p 1  
their simple structure. In industrial applications, more than Tis

90% of all control systems are PID type [1,2,3]. Also, owing  1  1   K p K d s 
to their popularity in the industrial world, several approaches K p 1    
PV ( s) Ti s
1  Ti s
1  Td s

for determining PID controller parameters have been GPVSV ( S )   SV (2)


SV ( s)  1 
developed for tuning [2,3,4,7,8,10]. These methods are 1  K p 1  
straight forward to apply since they provide simple tuning Ti s

formulae to determine the PID controller parameters. On the  K p K d s 


 
other hand, 2-DOF PID controller has been suggested for  1  T s 
GPVG 1 ( s)  d
G ( s)
(3)
recovering disturbance and set-point together [9-10] and its  1 
1

tuning approaches are also studied in many methods. On the 1  K p 1  



Ti s

other hand, it is very important to express human being mind


and many research results prove, which has an important role where, the filter transfer function is F (s)  1/ (1  Ti s) , the PI
to decision-making. [9] Therefore emotion is a popular topic controller transfer function is PI (s)  K p 1  1/ Ti s  , and the D
in AI research, but most of existing work focuses on the
controller transfer function is D(s)  K T s / (1  T s) . In
appraisal of emotions and mechanisms eliciting emotions p d d
[Plutchik, 1980; Lazarus, 1991; Frijda, 1986; Izard, 1993; equation (1), the numerator has a similar function to that of
Sloman, 2003]. Primary Emotions can rise by the external the conventional PID controller.
stimulus patterns. It can give an impact on enabling fast,

978-0-7695-4926-2/12 $26.00 © 2012 IEEE 70


74
DOI 10.1109/EMS.2012.93
Figure 2. Curve shape of transfer function G2(s)

Figure 1. Block diagram of the design principle of the 3-DOF PID


controller with a emotion function

That is, if the proportional gain K


p
goes to a greater value,
the efficiency of disturbance Gd is smaller. However,
Figure 3. Architecture of each function in emotion express.
equations (2) and (3), the process value PV(s) and the plant
G ( s) depend on the two degrees parameter , ,  . The
1
proportional gain could also be affected by the parameter ¢, Individual stimulations is expressed by exponentially,
, and  given for the two degrees function. Since the where L is the lifetime and is a constant that could be
disturbance can be reduced by gains K p , Ti , and  the selected to suit the application.

process value PV and the plant G1(s) are effectively  E s  Enew


s

L    old  if Eold (t )  Enew (t )
s s

controlled by the two degrees parameter , ,  . Then, a 3-  Rm  (5)


-  0(otherwise)
DOF-EI-PID controller can perform the two degrees of
function, completely. The result of this arrangement
distinguishes it from the conventional arrangement method. where is a positive constant and could be taken as unity.
Trigger level of emotion function depends on emotion
function. [Fig. 2]
III. EMOTION INTELLIGENT PID CONTROLLER (3-DOF-
Description For Emotion Engine EI-PID) TUNING BY IMMUNE NETWORK ALGORITHMS
Emotions in biological systems have been known to be
caused by imbalances in biochemical neurotransmitters and The coding of an antibody in an immune network is very
(hormonal) secretions. Our intended model for Emotion important because a well designed antibody coding can
Generation is based on the hormonal suppression/excitation increase the efficiency of the controller. As shown in Fig. 4,
dynamics in biological beings. It intends to emulate the there are three types antibodies in this paper: 1) antibody
similar dynamics. This paper has 7-segment as emotion type 1 is encoded to represent only gain in the 3-DOF
function as Fig. 3. In each emotion segment, when the PID controller; 2) antibody type 2 is encoded to represent I
emotional intensity of function crosses a threshold it gets gain; 3) antibody is encoded to represent gains. The value
excited and transmits messages that suppress or stimulate of the k locus of antibody type 1 shows gain allocated to
other function activity. The signals can be either stimulate or
route 1.
suppress function onto whose territory they are sent. This
That is, the value of the first locus of antibody type 1
behavior is embedded a priori in the system. Emotional
means that gain allocated to route 1 is obtained by 2 [11,
expression is stimulated and suppressed from other function
as the followings [17]: 12].

s
Eout (t  1)  Eold
s
(t  1)  Enew
s
(t ) (4a)
2 1 0.5  0.2 0.1
/ new (t )  exp( L) (4b)
s
Eold

75
71
2 1 0.5  0.2 0.12 Llim
n
: limit speed in PID gain
Tuning algorithm is those emotional parameters do well in
 2 1 0.5  0.2 0.1 system.
[step 1] Initialization and recognition of antigen: The
Figure 4. Allocation structure of , ,  gain in antibody. immune system recognizes the invasion of an antigen, which
corresponds to input data or disturbances in the optimization
On the other hand, the k locus of antibody 2 represents problem.
gain for tuning of the 2-DOF PID controller. Here, the [step 2] Product of antibody from memory cell: The immune
objective function can be written as follows. system produces the antibodies that were effective to kill the
antigen in the past. This is implemented by recalling a past
successful solution from memory cell.
 i   Ln  Lobject   f n
z
2
n (6) [step 3] Calculation for searching a optimal solution.
n1 [step 4] Differentiation of lymphocyte: The B - lymphocyte

 R I 
P cell, the antibody that matched the antigen, is dispersed to
Ln  i i ,n the memory cells in order to respond to the next invasion
i 1
quickly.
0 : Ln Llim it
[step 5] Stimulation and suppression of antibody: The
fn   n
expected value k of the stimulation of the antibody is given
1 : Otherwise
where, i : objective function by
: the number of processes for obtaining an optimal PID m
z !k
 (7)
gain k "
k
L : optimal level in process for selection of an optimal gain
n where " k is the concentration of the antibodies and is
object
L
n
: target optimal value in process for selection of an represented by
sum of antibodies with same affinity as m! k
optimal gain " k = "k  (8)
 : penalty constant sumof antibodies
f
n
: penalty function
[step 6] Stimulation of Antibody: To capture the unknown
P : the number of route for selection of an optimal gain antigen, new lymphocytes are produced in the bone marrow
R : gain level in route i in place of the antibody eliminated in step 5.
i
I
i, n
: subsidiary function

 ,QLWLDOYDOXH K p , Ki , Kd , , ,  , 

⁎Ⱂ
 
,0


K p , K i , K d ,  ,  
       B
69 1  T1 s
⁎Ⱂ 
   1  
1  T2 s  K p 1  
 * V 
 B Ti s

⁎Ⱂ 
 ⁎Ⱂ
K p Td s
 
   1  Td s
  
⁎Ⱂ 
⁎Ⱂ


Figure 5. Immune based tuning of the 3-DOF-EI-PID controller with emotion function

76
72
IV. SIMULATION AND DISCUSSION V. CONCLUSIONS
Table 1 the comparison of tuning results in 3-DOF-EI-PID The PID controller has been using in many areas. Also,
shows in Table 1. Table 1 expresses immune based tuning is approach for achieving an optimal PID gain is wide and
close to best solution. The mechanism of emotion function is various in many tools such as, Z-N, fuzzy, NN-fuzzy, GA etc.
biologically very complicated. That is, when the emotional The modified 2 - DOF PID controller is designed by
intensity of a B-function crosses a threshold it gets excited rearranging the 2 - degrees parameter to enable parameters of
and transmits messages that suppress or stimulate other controller to fit into the industrial areas. The emotion
segment activity. function is an important role in human being decision-
making. Recently, that area is emerging technology. Herein,
this paper suggests approach for application emotion in
industrial control and the other IT technology. For this
purpose, we suggest an immune algorithm based tuning
method for the 3-DOF-EI-PID controller. Parameters P, I
and D encoded in antibody are randomly allocated during
selection processes to obtain an optimal gain for 3-DOF-EI-
PID controller. The suggested controller can also be
experimented for effectively in the system.

ACKNOWLEDGMENT
This research was supported by Basic Science Research
Program through the National Research Foundation of
Korea(NRF) funded by the Ministry of Education, Science
Figure 6. Simulated result with emotion function. and Technology(2012R1A1A2042098).

TABLE I. TUNED GAIN BY IMMUNE NETWORK. REFERENCES

Tuning Methods Response [1] C. H. Lee and C. C. Teng, “A Novel Robust PID Controllers Design
Item Kp by Fuzzy Neural Network,” Asian Journal of Control, Vol. 4, No. 4,
Ti Td tr ts OV
pp. 433-438, 2002.
Z-N- [2] [J. X. Xu, Y. M. Pok, C. Liu, and C. C. Hang, “Tuning and Analysis
34.30 0.27 0.08 0.0774 1.2356 63.2
PID of a Fuzzy PI Controller Based on Gain and Phase Margins,” IEEE
PI 23.23 0.38 - 0.1299 2.338 77.2 Trans. on Systems, Man, and Cybernetics- Part A: Systems and
ISTE- Humans, Vol. 28, No. 5, pp. 685-691, 1998.
40.86 7.53 0.04 0.1441 4.7563 6.8
PID [3] Kraus, T.W., &Mayron, T.J, “Self-tuning PID controllers based on a
8.04 pattern recognition approach,” Control Engineering Practice, 106–111,
PI 11.90 - 0.1605 1.6713 18.6
4 1984.
21.6 [4] Hang, C. C., W. K. Ho and L. S. Cao, “A comparison of two design
IM 33.85 0.08 0.0263 2.4402 9.3
7 methods for PID controllers,” ISA Trans., 33, 147-151, 1994.
[5] Zhuang, M. and D. P. Atherton, “Automatic tuning of optimum PID
The messages on reaching other B-function territories controllers,” IEE Proc. Part D, 14, 216-224, 1993.
stimulate or inhibit them, creating a positive or negative [6] C. H. Lee and C. C. Teng, “Tuning PID Controller of Unstable
effect on the current emotion of the B-function. These Processes: A Fuzzy Neural Network Approach,” Fuzzy Sets and
messages form metaphors for biological neurotransmitters. Systems, Vol. 128, No.1, pp. 95-106, 2002.
They also have a life-time stamped on them by their creators [7] Dong Hwa Kim, “Tuning of a PID controller using a artificial
at the time when they emanate, which is proportional to the immune network model and fuzzy set” IFSA, July 28, Vancouver.
change in the associated emotion intensity. The messages [8] Dong Hwa Kim, “Intelligent Tuning of the 2-DOF PID Controller On
the DCS for Steam Temperature Control of Thermal Power Plant,”
can be either stimulate or suppress the B-function onto IEEE Industrial Application Society. I&CPS 2002, May 5 - 8, 2002,
whose territory they are sent. The decision as to which B- Savannah, GA, USA.
function to stimulate and which to suppress has been made [9] Valentin Lungu, “Artificial emotion simulation model,”
by intuitively classifying the affection of other emotions on a http://aimas.cs.pub.ro/people/valentin.lung.
basic emotion, as positive or negative. It is also proportional [10] Dong Hwa Kim. “Tuning of 2 - DOF PID controller by immune
to the input intensity to that B-agent among other factors algorithm," IEEE international conference on evolutionary
discussed later. computation, Hawaii, May 12 - 17, 2002.

77
73

You might also like