International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 6545(Print), INTERNATIONAL JOURNAL OF ELECTRICAL ENGINEERING &

; ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), IAEME TECHNOLOGY (IJEET)

ISSN 0976 6545(Print) ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), pp. 47-55 IAEME: www.iaeme.com/ijeet.asp Journal Impact Factor (2013): 5.5028 (Calculated by GISI) www.jifactor.com

IJEET
IAEME

PERFORMANCE ANALYSIS AND CONTROL DESIGN OF TWO DIMENSION FUZZY PID CONTROLLER
1*

Manikandan P,

2*

Geetha M,

3*

Jubi K,

4**

Hariprasath P,

5*

Jovitha Jerome

*Department of Instrumentation and Control Systems Engineering **Department of Electronics and Communication Engineering PSG College of Technology, Coimbatore, India

ABSTRACT Tuning the parameters of PID controller is very important in system performance. Ziegler and Nichols tuning method is simple and cannot guarantee to be effective always. In order to overcome the parameter uncertainties, enhance the fast tracking performance of a process system, a brand-new two-dimension fuzzy PID controller, fuzzy PI+ fuzzy ID, is proposed in this paper. The self-tuning fuzzy PI+ fuzzy ID controller is fast; computing on-line easily and can reduce stability error. To demonstrate the effectiveness of the fuzzy PI+ fuzzy ID controller has been experimented to high order systems. The simulation and implementation were executed and its results show that the proposed control scheme not only enhances the fast tracking performance, but also increases the robustness of the system. From the simulation it is clear that there is substantial improvement in the Two Dimension Fuzzy PID (2D Fuzzy PID) controller in terms of peak overshoot, settling time, peak time, rise time, Integral Square Error (ISE) and Integral Absolute Error (IAE). Keywords- PID Controller; Fuzzy Logic, 2D Fuzzy PID; CSTR; ISE; IAE. 1. INTRODUCTION Process control is of vital importance in the operation of chemical, pharmaceutical, power plant, paper and bleach processes. Chemical process present many challenging control problems due to their nonlinear dynamic behavior and time varying parameters, constraints on manipulated variable, interaction between manipulated and controlled variables, unmeasured and frequent disturbances, dead time on input and measurements. Because of the inherent nonlinearity, most of the chemical process industries are in need of traditional control techniques. Fuzzy logic control (FLC) is one of the most successful applications of fuzzy set theory, introduced by L.A Zadeh in 1973 and applied (Mamdani 1974) in an attempt to control the system that are structurally difficult
47

International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 6545(Print), ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), IAEME

to model. In the last three decades, FLC has evolved as an alternative or complementary to the conventional control strategies in various engineering areas. Fuzzy control theory usually provides non-linear linear controllers that are capable of performing different complex non-linear linear control action, even for uncertain nonlinear systems. Unlike conventional control, designing a FLC does not require precise knowledge of the system model such as the poles and zeroes of the system transfer functions. Nonlinear tanks are a used in many process industries as storage of cryogenic liquids, fuels and other liquids also as surge tanks and find wide application in gas plants. Control of a level in a tank is important, because the change in shape gives rise to the nonlinearity. Fuzzy logic has gained great attention in the area of process control due to its ability to incorporate human intuition in the design process. Three Three- dimensional fuzzy controller [2] which has a good control effect and performance, needs great capacity in computing, and is not suitable for on-line on line transaction for the design sophisticates. Self Self-organized fuzzy controller with neural network demands large quantity of training time. The parameters affecting its control effect primarily include gain factor, scaling scaling factor and proportion factor, and in a two dimensional fuzzy controller these factors are usually fixed. When the model or parameters of the object are changed, two dimensional fuzzy controllers [3] cant self-tune tune and the control performance becomes bad. So it needs to seek a kind of controller which has self-tuning tuning function. For above reasons the paper puts forward a kind of controller of which factors can be self-adjusted adjusted according to the status and adopt a load observer to observe load torque. In order to evaluate the performance of the fuzzy PI+PD controller the paper adopts the ITAE index [4], rules of the fuzzy controller analysis born. The de-fuzzy de fuzzy method is based on centroid together with phase surface [5]. There are several types of control systems systems that use FLC as an essential system component. The majority of applications during the past two decades belong to the class of fuzzy PID controllers. Several forms of decomposed proportional-integralproportional - derivative fuzzy logic controllers (fuzzy P + fuzzy zzy I + fuzzy D form, fuzzy PD + fuzzy I form, fuzzy PI + conventional D form, fuzzy P + conventional ID form and fuzzy PI + fuzzy PD form) have been tested and compared [1]. To obtain simple structures, the activities of the proportional, integral and derivative ivative parts of the fuzzy PID controller are defined with simple rules in proportional rule base, integral rule base and derivative rule base. The analysis, design and simulation of the proposed self-tuning self tuning fuzzy PI+ fuzzy ID controller are described. The self-tuning tuning fuzzy PI+ fuzzy ID controller is fast, computing on-line on line easily and can reduce stability error. Good control performance, both in the command-tracking command tracking and parameters robust of the level process, is achieved. This paper is organized in the following following manner. In Section II, the experimental setup is described. In Section III& IV, self-tuning self tuning fuzzy PI+ fuzzy ID control scheme is presented. To verify the methodology, computer simulations and the discussions are provided in Section V. A brief conclu conclusion is outlined in Section V& VI. 2. . STRUCTURE OF CONTROLLER 2.1 Coventional PID Controller The general expression of traditional PID controller can be stated as: (1) The structure of PID controller is shown in Fig.2, where the error value e is the deviation of input and feedback values. The controller regulates the output according to the PID parameters.

48

International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 6545(Print), ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), IAEME

Figure.1 Block diagram of PID controller Conventional linear PID controller signals in digital form can be represented by three controller actions with respect to the error response. The incremental controller output signal can be found by taking

1
The error state variables are defined as,

(2) (3)

1 1

(4) (5) (6)

Ts- sampling time, r(n) - desired response, y(n) - actual plant response, n - sampling instance (at time t=0, n=0, e(0)=0), KP, KD and KI are Proportional, Derivative and Integral constants related to a linear PID controller. e(n)-error, e(n)-error change, e(n)-rate of error change. According to the Eq. (2), we can make up of the next structure. 1 1 2 1

(7) (8) (9)

2.2 Self-Tuning Fuzzy PD+Fuzzy ID Controller Based on the Eq. (5), a new 2D Fuzzy PID controller, which are composed of fuzzy PD controller FC1 and fuzzy ID controller FC2, namely incremental fuzzy PD+ fuzzy ID, is designed. r(kt) the reference signal, y(kt) the process output, input variables, including error e(kt), change of error ec(kt), rate of change of error ed(kt), u(kt) the output variables. GI (gain for error) is the input scalar for e(kt), GP (gain for error change) the input scalar for ec(kt), GD (gain for rate of error change) the input scalar for ed(kt) and GU (gain for controller output) the output scalar of the FLC.

49

International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 6545(Print), ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), IAEME

ec

FLC 1

GP GD
FUZZ Y PD

U1(kt )

e r (k)

PLANT

y (k)

GI GD
edc
FUZZ Y ID
U2(kt )

FLC 2

Figure.2 Block diagram of two dimension Fuzzy PID Controller 3. DEVELOPMENT OF TWO DIMENSION FUZZY PID CONTROLLER 3.1 Input Stage and Fuzzification The incremental fuzzy PD+ fuzzy ID controller employs three inputs: error e(kt), change of error ec(kt) and rate of change of error ed(kt).The inputs times the input scalars GP, GI, GD, they can express below. The fuzzy set error e(kt), the scaled change of error ec(kt) and the scaled rate of change of error ed(kt) has seven members PB(Positive Big),PM(Positive Medium), PS(Positive Small), Z(Zero), NS(Negative Small), NM(Negative Medium), NB(Negative Big). The scaled error e(kt), the scaled change of error ec(kt) and the scaled rate of change of error ed(kt) are within the range [-L,L] of the fuzzification algorithm shown in Fig.2.

1/ 1/

(10)

(11)

(12)

3.2 Ouput Stage and Fuzzification The incremental fuzzy PD+ fuzzy ID controller has two outputs, called the incremental control outputs and are denoted by u1 (kt), u2 (kt). The fuzzy set output1 and output2 has seven members PB(Positive Big),PM(Positive Medium), PS(Positive Small), Z(Zero), NS(Negative Small), NM(Negative Medium), NB(Negative Big) for the fuzzification of the incremental outputs of fuzzy control FC1 and FC2. The out of the incremental fuzzy PD+ fuzzy ID controller:

1 1 2 2

(13)

50

International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 6545(Print), ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), IAEME

3.3 Fuzzy Rule Base The control structure of the 49 control rules are established for the fuzzy PD controller (FC1) shown in the table 1. Using the aforementioned membership functions, the following control rules are established for the fuzzy PD controller (FC1). TABLE I. OUTPUTu1 (kt) FUZZY CONTROL RULE TABLE
u1(kt) NB NM NS Z PS PM PB PB Z PS PM PB PB PB PB PM Z Z PS PM PM PB PB PS NM NS Z PS PS PM PB GI*e(kt) Z NM NS Z Z Z NS NM NS PB PM PS PS Z NS NM NM PB PB PM PM PS Z Z NB PB PB PB PB PM PS Z

GP * ec(kt)

The control structure of the 49 control rules are established for the fuzzy ID controller (FC2) shown in the table 2.Using the aforementioned membership functions, the following control rules are established for the fuzzy ID controller (FC2).
u2(kt)

TABLE II. OUTPUT u2 (kt) FUZZY CONTROL RULE TABLE

NB NM NS PS PM PB NB PB PM PS NM NB NB NM PM PS Z NB NS Z NS PB PM PS PS PM PB G I* e(kt) Z PB PM PS PS PM PB PS PB PM PS PS PM PB PM Z NS NM Z PS PM PB NB NB NM PS PM PB

GD* ed(kt)

4. RESULTS AND DISCUSSIONS Effectiveness of the proposed scheme is verified through simulation experiments on second and third-order processes [9] with dead-time (L). In addition to response characteristics, performance of the proposed Two dimension Fuzzy PID (2D FUZZY PID) is compared with those of ZN-PID and Conventional Fuzzy logic Controller (FLC) , with respect to a number of indices, such as percentage overshoot (%OS), integral-absolute-error (IAE), and time integral absolute error (ITAE). The detailed performance analysis for various types of processes is discussed below. 5.1 Linear processes In our simulation experiments, we consider the following second-order linear process used in [6, 9]:
GP ( s) = Ke Ls . (1 + Ts ) 2

(14)

Where, K is process gain, L is dead time, and T is time constant. For all the simulation examples, we have taken K = 1, T = 1.

51

International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 6545(Print), ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), IAEME

Fig.2 (a) shows the response characteristics of the process in (2) with L= 0.2s under ZN-PID, 2D FUZZY PID and TL-PID. Various performance indices for these PID controllers are recorded in Table I. Fig. 2(a) exhibits considerably improved performance of 2D FUZZY PID over ZN-PID and FLC, specifically during load disturbance. Table III clearly indicates the superiority of 2D FUZZY PID over ZN-PID, FLC.
1.5
1.8

2D FUZZY ZN-PID FLC

1.4 1.6 2D FUZZY ZN-PID FLC

Responses

1.2

Responses

1 0.8

0.5

0.6 0.4 0.2

10

15

Time

20

25

30

35

10

15

Time

20

25

30

35

(a)

(b)

Figure 3. Response of the second-order linear process,a) L=0.2, b) L=0.3 To further study the robustness of the proposed controller, a 50% higher value of dead-time, i.e., L = 0:3s is considered with the same controller setting as that of L = 0:2s. Corresponding responses of different PID controllers are shown in Fig. 2(b), which also exhibits the same level of improvement in 2D FUZZY PID compared to ZN-PID and FLC. TABLE I: PERFORMANCE INDICES OF SECOND ORDER LINEAR PROCESS
L % Overshoot 0.2 IAE ITAE % Overshoot IAE ITAE 2D FUZZY PID 28.2 70.49 262.4 29.28 69.69 288.1 ZN-PID 48.84 71.3 264.7 71.0 76.39 341.5 FLC * 87.56 607.2 * 89.41 717.6

0.3

* No overshoot 5.2 Second-order marginally stable process The transfer function of the second-order marginally stable process is given by
GP ( s ) = Ke Ls s (1 + Ts )

(15)

Figs. 3(a) and 3(b) depict the responses of this integrating process with L = 0:2 s and 0.3 s, respectively and detailed performance comparison is illustrated in Table IV. Even for a 50%
52

International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 6545(Print), ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), IAEME

perturbation in dead-time i.e., for L = 0.3 s, 2D FUZZY PID still provides much better performance compared to ZN-PID and FLC.
1.8 2D FUZZY 1.6 1.4 1.2 ZN-PID FLC

2 1.8 1.6 1.4 1.2 2D FUZZY ZN-PID FLC

Responses

1 0.8

Responses 1
0.8

0.6

0.6
0.4 0.2 0

0.4 0.2
0 5 10 15 20 25 30 35 40 45 50

Time
(a)

10

15

20

Time

25

30

35

40

45

50

(b)

Figure 4. Response of the second-order marginally stable process,a) L= 0.3, b) L=0.4 TABLE I I: PERFORMANCE INDICES OF SECOND ORDER MARGINALLY STABLE PROCESS
L
0.3 % Overshoot IAE ITAE % Overshoot IAE ITAE

0.4

2D FUZZY PID 49.8 73.3 382.6 57.8 62.5 394

ZN-PID 68.55 75.99 417.3 83.22 68.97 482.7

FLC * 92.97 944.2 * 84.24 1026

* No overshoot

5.3 Third-order linear process A second-order plus dead-time (SOPDT) model is usually considered to be a fair approximation for most of the industrial processes. However, to study the effectiveness of the proposed scheme, third-order linear process models are also tested. The transfer function of such a linear system is:
GP ( s) = Ke Ls (1 + Ts ) 3

(16)

Figs. 4(a) and 4(b) depict the responses of the third-order linear process in (8,4) with L = 0:2s and 0.3s, respectively and various performance indices of 2D FUZZY PID, ZN-PID and FLC are provided in Table V. Results reveal that there is a considerable improvement in %OS compared to ZN-PID.

53

International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 6545(Print), ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), IAEME
1.5 2D FUZZY ZN-PID FLC

1.4

1.2

1
1

Responses

0.8

0.6
0.5

0.4

2D FUZZY Z-N

0.2
0

FLC

10

20

30

40

50

60

70

80

Time
(a)

10

20

30

Time

40

50

60

70

80

(b)

Figure 5. Response of third-order linear process with a) L = 0.4 b) L=0.5 TABLE I I I: PERFORMANCE INDICES OF THIRD - ORDER LINEAR PROCESS
L

0.4

0.5

% Overshoot IAE ITAE % Overshoot IAE ITAE

2D FUZZY PID 19.3 84.0 515.9 25.58 65.98 511.7

ZN-PID 33.18 84.43 512.6 40.15 67.27 517.7

FLC

* 115.7 1575 * 99.1 1640

* No overshoot From the above results for various processes it is evident that in each case the proposed 2D FUZZY PID shows consistently improved overall performance compared to ZN-PID and FLC under both set-point change and load disturbance. 6. CONCLUSION The adaptive self-tuning fuzzy PI+ fuzzy ID controller is fast, computing on-line easily and can reduce stability error. The results show that the proposed control scheme not only enhances the fast tracking performance, but also increases the robustness of the process plant. Thus Fuzzy PID can be applied not only to precise process control, as its results shows it can be applied to obtain control with precision in various other applications like Medical robotics, Industrial Testing and measurements. 7. REFERENCES [1]. M. Golob, Decomposed fuzzy proportional-integral-derivative controllers, Applied Soft Computing, vol. 1, Issue 3, pp. 201-214, Dec. 2001. [2]. Mudi, R. K., Pal, N. R.,A Robust Self-Tuning Scheme for PI and PD Type Fuzzy Controllers, IEEE Trans. on Fuzzy Systems,Vol.7, pp.2--16 ,1999.
54

International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 6545(Print), ISSN 0976 6553(Online) Volume 4, Issue 5, September October (2013), IAEME

[3]. Palm, R., Scaling of Fuzzy Controller Using the Cross-Correlation, IEEE Trans. on Fuzzy Syst,Vol.3, pp.116--123 ,1995. [4]. Mudi R. K., Pal, N. R., A Self-Tuning Fuzzy PI Controller, Fuzzy Sets and Systems,Vol.115, pp.327--338 ,2000. [5]. Mudi, R.K., Pal, and N. R., A Self-Tuning Fuzzy PD Controller, IETE Journal of Research,(Special Issue on Fuzzy Systems), pp.177189, 1998. [6]. Pal, A.K., Mudi, and R.K.,Self-Tuning Fuzzy PI controller and its application to HVAC system, IJCC (US),vol. 6,2008. [7]. Palm, R.,Sliding Mode Fuzzy Control, Proc. 1st IEEE Int. Conf. on Fuzzy Systems,pp. 519-526 ,1992. [8]. K.H. Ang. G. Chong., and Y. Li, PID control system analysis, design, and technology, IEEE Transaction on control system technology, vol.13, No.4, pp.559-576, 2005 [9]. K.J. Astrom and T. Hagglund, PID controller: Theory, Design and Tuning, USA, Instrument Society of America, 1995 [10]. C. Dey and R.K. Mudi, An improved auto-tuning scheme for PID controllers, ISA Transactions, vol. 48, no. 4, pp. 396-409, 2009 [11]. VenkataRamesh.Edara, B.Amarendra Reddy, Srikanth Monangi and M.Vimala, Analytical Structures for Fuzzy Pid Controllers and Applications, International Journal of Electrical Engineering & Technology (IJEET), Volume 1, Issue 1, 2010, pp. 1 - 17, ISSN Print: 0976-6545, ISSN Online: 0976-6553. [12]. B.Rajani and Dr.P.Sangameswara Raju, Comparision of Pi, Fuzzy & Neuro-Fuzzy Controller Based Multi Converter Unified Power Quality Conditioner, International Journal of Electrical Engineering & Technology (IJEET), Volume 4, Issue 2, 2013, pp. 136 - 154, ISSN Print: 0976-6545, ISSN Online: 0976-6553. [13]. L.Raguraman and A.Gnanasaravanan, Performance Optimization and Comparison of Variable Parameter using Genetic Algorithm Based Pid Controller, International Journal of Electrical Engineering & Technology (IJEET), Volume 4, Issue 4, 2013, pp. 42 - 47, ISSN Print: 0976-6545, ISSN Online: 0976-6553.

55