 Open Access
 Total Downloads : 848
 Authors : Manikiran. P, Ramesh. G
 Paper ID : IJERTV2IS120734
 Volume & Issue : Volume 02, Issue 12 (December 2013)
 Published (First Online): 19122013
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Intelligent PID controller Design for Extrusion Process
Manikiran. P1, Ramesh. G2
1P.G Student, EEE, Gudlavalleru Engineering College, Gudlavalleru, A.P, India.
2Assistant Professor, EEE, Gudlavalleru Engineering College, Gudlavalleru, A.P, India.
Abstract
An injection module machine is a basic requirement of every plastic product. So, it is very intended to maintain the desired temperatures. Temperature control of plastic extrusion system suffers problems related to longer settling time, and undesirable overshoot. Conventionally, PID controller offers major constraint in the selection of controller gains but unable to control the temperature process due to its nonlinear behavior. Logic based intelligent concepts are used to control the temperature process. FuzzyPID gives satisfactory results only when there is no disturbance; however it is unable to stabilize the temperature at disturbances. In this paper ANNPID controller is developed for temperature control in plastic extrusion system. The proposed ANNPID controller was simulated using MATLAB software. Relatively the methodology and efficiency of the proposed ANNPID methods are compared with that of traditional methods and results are provided. ANNPID controller which offers best performance compared to prior traditional methods.

Introduction
Use of polymer materials has greatly increased over last few decades due to their many attractive properties such as ease of forming into complex shapes, lightweight with high tensile/impact/tear strengths, high temperature resistance, high chemical resistance, high clarity, reprocessibility and low cost. This has resulted in new industrial applications for polymer materials while enabling products to be more cost effective, flexible, and efficient. The extrusion process is used for
the production of commodities in diverse industrial sectors such as packaging, household, automotive, aerospace, marine, construction, electrical and electronic, and medical applications. Despite of this success, it seems that effective thermal monitoring and control still remains a concern [2].
1.1 Polymer extrusion process
There are two basic types of polymer processing extruders [1] known as continuous and batch extruders (Rosato, 1998) [13]. Of these, single screw continuous extruders are the most commonly used in the plastics industry (Spalding and Hyun, 2003). The basic components of a single screw extruder are shown in Figure 1. The screw is the key component and has been divided into three main functional/geometrical zones (i.e. solids conveying, melting, and metering) based on their primary operations. The material fed into the machine through the hopper is conveyed along the screw while absorbing heat provided by the barrel heaters and through process mechanical work. Eventually, a molten flow of material is forced into the die which forms the material into the desired shape. More details of the mechanisms of polymer extrusion can be found in Rauwendaal (2001). Under poor thermal conditions, several processing problems can occur, e.g. thermal degradation, output surging, poor mechanical properties, dimensional instability, poor surface finish and poor optical clarity melt temperature homogeneity depends on the selection of processing conditions, machine geometry, and material properties. Moreover, it has been found that melt temperature non homogeneity increases with screw speed. Therefore, it is a challenging task to run extruders at higher screw speeds although the process energy efficiency then increases.
Figure.1 Basic components of Extruder plant
The temperature control in injection mould machine
[3] is a key part of the machine. So, this controlling process is achieved by designing the controller for the response of the transfer function in accordance with the desired set point applied. Conventionally, PID (Proportional + Integral + Derivative) controller is used. PID controller is simple in algorithm, good in stability, high in reliability, easy in design, and wide in adaptation, it is the most extensive basic controller used in the application of process control. It can obtain satisfactory control effects in variety of linear time invariant systems, particularly in systems whose parameters of controlled objects are fixed, nonlinear is not very serious. However, the PID control is crisp control, the selfturning of the P, I, D parameters is a
Where,
Static gain (K) = 0.92 Time constant (T) = 144sec
Lag delay time () = 30 sec

SYSTEM DESIGN WITH PID CONTROLLER
A simple strategy widely used in industrial control is PID controller. But the parameter selection for the KP, KI, KD gains is always a challenge. Many tuning algorithms were developed, but still it is a major pain to select particular algorithm for designing PID gain values for the particular system for a particular process control. Figure.2 shows the system architecture with PID controller. Equation (3) shows the mathematical description of a general PID controller [4].
Figure.2. System Design with PID controller
= + () + ()
quite difficult job, and sometimes the PID control makes overshoot, and also with regard to the temperature control system the characteristics of which
Where,
0
. (3)
are distributed parameter, nonlinear, large time delay and large inertia, the conventional PID controller is very difficult to obtain satisfactory control results. In
order to solve this problem, a control method which
U (t) = Control signal applied to plant KP = Proportional Gain
KI = = Integral Gain
uses the fuzzy logic technology in temperature control for injection mould machine is used. Fuzzy controller
KD =
KP Ã— TD
= Derivative Gain
can make full use of the successful operation experience of the operator which they get in real time nonlinear adjustment. Also it can give full play to the fine control effect of the PID controller, makes the whole system to achieve the good control effect. So, the paper also proposes the method of fuzzy logic, for tuning the PID controller gain parameters. The temperature process of an injection mould machine is a kind of common controlled object in temperature control system. It can be described qualitatively by the model shown in equation (1).
The selection of these KP, KI, and KD values
will cause for the variations in the observed response with respect to the desired response. In general, the dependency will be as per the Table.1.
Parameter
Rise time (Tr)
Overshoot (Mp)
Settling time (Ts)
Error (Ess)
KP
Decrease
Increase
Small change
Decrease
KI
Decrease
Increase
Increase
Decrease notably
KD
Minor Decrease
Decrease
Decrease
No effect
Parameter
Rise time (Tr)
Overshoot (Mp)
Settling time (Ts)
Error (Ess)
KP
Decrease
Increase
Small change
Decrease
KI
Decrease
Increase
Increase
Decrease notably
<>KD Minor Decrease
Decrease
Decrease
No effect
Table 1. Effect of increasing parameter values independently on the response
=
+1
(1)
And hence, for the given system the transfer function
[5] can be obtained as,Open Loop Transient Response Tuning method used for setting up the values of
KP = 7.4, KI = 0.1274, and KD = 107.411.
= 0.92
144+1
30 (2)

SYSTEM DESIGN WITH FUZZY LOGIC CONTROLLER
FLC is relatively easy to implement, as it usually needs no mathematical model of the control system. Fuzzy logic has rapidly become one of the most successful of today's technologies for developing sophisticated control systems. Fuzzy logic addresses such applications perfectly as it resembles human decision making with an ability to generate precise solutions from certain or approximate information
DeFuzzification is the process of producing a quantifiable result in fuzzy logic, given fuzzy sets and corresponding membership degrees. Figure 4, shows the system architecture with fuzzy logic controller. Fuzzy logic controller [6] takes two inputs namely, error and error change and produces control signal according to the Fuzzy Inference Structure (FIS) designed with assumed fuzzy rules. Each of the input and output quantity is described with its corresponding membership function. Table 2 indicates Ifthen Fuzzy Rules [11] for Developing FIS.
Figure.3 System Design with Fuzzy Logic controller
E 
NB 
NM 
NS 
ZO 
PS 
PM 
PB 

EC 
U 

NB 
PB 
PB 
PB 
PB 
PM 
PS 
ZO 

NM 
PB 
PB 
PM 
PM 
PS 
ZO 
NS 

NS 
PB 
PB 
PM 
PS 
ZO 
NM 
NM 

ZO 
PB 
PM 
PS 
ZO 
NS 
NM 
NB 

PS 
PM 
PM 
ZO 
NS 
NM 
NB 
NB 

PM 
PS 
ZO 
NS 
NM 
NM 
NB 
NB 

PB 
ZO 
NS 
NM 
NB 
NB 
NB 
NB 
E 
NB 
NM 
NS 
ZO 
PS 
PM 
PB 

EC 
U 

NB 
PB 
PB 
PB 
PB 
PM 
PS 
ZO 

NM 
PB 
PB 
PM 
PM 
PS 
ZO 
NS 

NS 
PB 
PB 
PM 
PS 
ZO 
NM 
NM 

ZO 
PB 
PM 
PS 
ZO 
NS 
NM 
NB 

PS 
PM 
PM 
ZO 
NS 
NM 
NB 
NB 

PM 
PS 
ZO 
NS 
NM 
NM 
NB 
NB 

PB 
ZO 
NS 
NM 
NB 
NB 
NB 
NB 
Table 2. Fuzzy rules for developing Fuzzy Inference Structure (FIS) for Fuzzy Logic Controller
Where, NB Negative Big; NM Negative Medium; NS Negative Small; PB Positive Big; PM Positive Medium, PS Positive Small, ZO Zero value.

SYSTEM DESIGN WITH FUZZYPID CONTROLLER
Fuzzy machines, which always tend to mimic the behavior of man, work the same way. However, the decision and the means of choosing that decision are replaced by fuzzy sets and the rules are replaced by fuzzy rules. Fuzzy rules also operate using a series of
ifthen statements. Table 3, shows the fuzzy rules for developing FIS. The fuzzy control rule is based on fuzzy decisionmaking, which satisfies some input conditions and has an output result. Figure 4, shows the design of the system with FuzzyPID controller [7] where the gain values of PID controller are tuned by Fuzzy controller.
Table 3. Fuzzy rules for developing Fuzzy Inference Structure (FIS) for Fuzzy – PID Controller
E
NB
NM
NS
ZO
PS
PM
PB
E C
KP KI KD
NB
PB NB PS
PB NB NS
PM NM NB
PM NM NB
PS NS NB
ZO ZO NM
ZO ZO PS
NM
PB NB PS
PB NB NS
PM NM NB
PS NS NM
PS NS NM
ZO ZO NS
NS ZO ZO
NS
PM NB ZO
PM NM NS
PM NS NM
PS NS NM
ZO ZO NS
NS PS NS
NS PS ZO
ZO
PM NM ZO
PM NM NS
PS NS NS
ZO ZO NS
NS PS NS
NM PM NS
NM PM ZO
PS
PS NM ZO
PS NS ZO
ZO ZO ZO
NS PS ZO
NS PS ZO
NM PM ZO
NM PB ZO
PM
PS ZO PB
ZO ZO NS
NS PS PS
NM PS PS
NM PM PS
NM PB PS
NB PB PB
PB
ZO ZO PB
ZO ZO NM
NM PS PM
NM PM PM
NM PM PS
NB PB PS
NB PB PB
Figure.4 System Design with Fuzzy – PID Controller
FuzzyPID performs well as per the requirement to the plant. But when problem rises means if any disturbance arises it is unable to handle. Disturbances are commonly took place in industries occurs either from internal source or external basis. In Injection Mould machine the disturbances caused to poor quality in product, disorder in shapes leads to great loss. In order to rectify these losses disturbance has to overcome. In this paper Artificial Neural Networks with
PID was proposed to overcome the losses occurred due to disturbances. Figure.5 shows the system designed with different disturbances. Table 4. Shows the Fuzzy rules for FuzzyPID with Disturbances.
Figure.6 Mathematical model of a neuron
n
f ( p) f w0 wi xi
(4)
i1
Figure.5 System design with FuzzyPID with Disturbances.
Table 4. Fuzzy rules for developing Fuzzy Inference Structure (FIS) for Fuzzy – PID with Disturbances
PID
Disturb
KP
KI
KD
NB
AA
BA
CA
NS
AB
BB
CB
ZO
AC
BC
CC
PS
AD
BD
CD
PB
AE
BE
CE

ARTIFICIAL NEURAL NETWORKS
Artificial Neural networks [9] are motivated by human way of learning and the network structure is motivated by human nervous system in the characteristics like learning from examples, fault tolerant, learns from experience, distributed in nature and etc. Neural networks has been successfully applies in the fields of load forecasting, pattern recognition, image processing, optimization and where the input output data is available. If spikes or transients are sufficiently involved in the trained data, then no doubt about that the neural networks will give accurate solutions.
NOTE: No mathematical calculations are required to train NN and to about results from trained Neural Network.
Figure.6 shows the mathematical model of the
There are many activation functions used for the neurons. Of them, commonly used are linear, tan sigmoidal and logsigmoidal activation functions. Multilayer feed forward neural network architecture which is used in this paper due to is flexible characteristics.
At this juncture training the Neural Networks plays a key role. There are several learning algorithms [10] such as Hebbian, Competitive, Gradient descent, and Back Propagation. Among the learning method Back Propagation is the best. In the proposed method tan sigmoidal used as activation function and back propagation used as learning method. In multilayer feed forward network appropriate input output and hidden layer neurons were used.
Figure.7 shows the design of system with Artificial Neural Networks PID [8] with different types of disturbances. Of them, used in this paper are repeated sequence interpolated, constant disturbance, chirp, whiteband noise, repeated sequence, step, repeated sequence stair. For all these applied disturbances proposed method was able to overcome the disturbance with appreciable results.
neuron.
x1, x2 ……..xn
are the scalar inputs and those
are multiplied with synaptic weights
w1, w2 ……..wn
respectively and they are summed with bias (W0). A bias is also like a synaptic weight with input signal one. The total sum is given as input to the activation function. Y is the output of the neuron. This is given by equation 1
Figure.7 System design with ANN_PID with disturbances.

SIMULATION RESULTS
The dynamic performance of the system can be analyzed based on the time response plot. Those performance parameters are called as time domain specifications, can evaluate the effectiveness of various control schemes proposed in this paper.
MATLAB is used to simulate the simulink models. Figure 8 gives result for PID controller, for Fuzzy controller then FuzzyPID controller. Now Figure. 9 shows the result for 10% disturbance to FuzzyPID controller. Drawbacks were rectified in figure 10 which is response of ANNPID controller.
Figure.8 System response with conventional PID, Fuzzy, and FuzzyPID controller.
Figure. 9 System response with FuzzyPID controller for 10% disturbance.
When a 10% constant disturbance was applied to FuzzyPID system gives the result of peak overshoot which is undesirable to the plant causes a great loss to industry. When the same 10% constant disturbance was applied to ANNPID system gives the result peak overshoot was eliminatedand and settling time also minimized shown in Figure. 10.
Figure.10 System response with ANNPID controller for 10% disturbance.
Figure.11 System response with ANNPID controller for Repeated Sequence Interpolated disturbance.
Different types of disturbances were applied to system with ANNPID controller[12] such as step, Repeating Sequence Stair, Chirp Disturbance, Repeating Sequence Interpolated Disturbance, Band Limited White Noise, Repeating Sequence, and the results are tabulated in Table 4, shown below.
Table 4. Comparision table of PID, FuzzyPID & ANNPID control system for different disturbances
S. No 
Type of Disturbance 
Dynamic Performance Specification 
For the Conventional PID Control System 
For the FUZZY PID Control System 
For the Proposed ANNPID Control System 
Improvement from PID to ANNPID in % 
1 
No Disturbance 
Delay Time (TD) in Sec 
11.623 
19.513 
20.401 
43.02 
Rise Time (TR) in Sec 
18.546 
33.768 
38.632 
51.96 

Settling Time (TS) in Sec 
295.210 
265.541 
56.720 
80.78 

Peak Overshoot (MP) in % 
38.07 
7.33 
0 
100 

% Steady state Error (ESS) 
0 
0 
0 
0 

Nature of Oscillations 
Oscillatory 
Smooth 
No 
100 

2 
A Step Disturbance of – 0.1R 
Delay Time (TD) in Sec 
12.502 
20.00 
21.120 
40.80 
Rise Time (TR) in Sec 
19.310 
35.356 
41.500 
53.46 

Settling Time (TS) in Sec 
293.320 
233.538 
120.421 
58.94 

Peak Overshoot (MP) in % 
34.20 
5.35 
0 
100 

% Steady state Error (ESS) 
0 
0 
0 
0 

Nature of Oscillations 
Oscillatory 
Smooth 
No 
100 

3 
A Step Disturbance of +0.1R 
Delay Time (TD) in Sec 
11.043 
19.052 
19.013 
41.91 
Rise Time (TR) in Sec 
18.512 
32.645 
37.510 
50.64 

Settling Time (TS) in Sec 
297.311 
288.368 
51.301 
82.74 

Peak Overshoot (MP) in % 
39.50 
9.32 
0 
100 

% Steady state Error (ESS) 
0 
0 
0 
0 

Nature of Oscillations 
Oscillatory 
Smooth 
No 
100 

4 
Repeating Sequence Stair Disturbance 
Delay Time (TD) in Sec 
11.323 
18..571 
18.303 
38.13 
Rise Time (TR) in ec 
17.534 
32.465 
31.346 
44.03 

Settling Time (TS) in Sec 
368.051 
361.843 
350.00 
4.90 

Peak Overshoot (MP) in % 
43.50 
15.50 
8 
81.60 

% Steady state Error (ESS) 
0 
0 
0 
0 

Nature of Oscillations 
Oscillatory 
Smooth 
No 
100 

5 
Repeating Sequence Interpolated Disturbance 
Delay Time (TD) in Sec 
11.531 
18.625 
19.453 
40.72 
Rise Time (TR) in Sec 
18.413 
31.651 
35.326 
76.18 

Settling Time (TS) in Sec 
362.00 
310.176 
210.00 
41.90 

Peak Overshoot (MP) in % 
39.50 
11.5 
3 
92.40 

% Steady state Error (ESS) 
0 
0 
0 
0 

Nature of Oscillations 
Oscillatory 
Smooth 
No 
100 

6 
Repeating Sequence Disturbance 
Delay Time (TD) in Sec 
11.864 
20.158 
19.651 
39.62 
Rise Time (TR) in Sec 
18.100 
36.974 
40.097 
54.85 

Settling Time (TS) in Sec 
297.105 
245.879 
148.750 
49.93 

Peak Overshoot (MP) in % 
38.30 
9.83 
1.80 
95.53 

% Steady state Error (ESS) 
0 
0 
0 
0 

Nature of Oscillations 
Oscillatory 
Smooth 
N0 
100 

7 
Chirp Disturbance 
Delay Time (TD) in Sec 
11.903 
21.927 
20.573 
42.14 
Rise Time (TR) in Sec 
18.756 
39.756 
38.852 
51.72 

Settling Time (TS) in Sec 
295.571 
280.694 
56.00 
81.05 

Peak Overshoot (MP) in % 
35.80 
13.62 
0 
100 

% Steady state Error (ESS) 
0 
0 
0 
0 

Nature of Oscillations 
Oscillatory 
Smooth 
No 
100 

8 
Band Limited White Noise Disturbance 
Delay Time (TD) in Sec 
12.090 
19.534 
20.122 
39.91 
Rise Time (TR) in Sec 
18.572 
34.476 
39.471 
52.94 

Settling Time (TS) in Sec 
352.651 
293.650 
61.534 
82.26 

Peak Overshoot (MP) in % 
35.50 
7.10 
1 
97.18 

% Steady state Error (ESS) 
0 
0 
0 
0 

Nature of Oscillations 
Oscillatory 
Smooth 
No 
100 
Figure.12 System response with ANNPID controller for Chirp Noise.

CONCLUSION
Various PID controllers including Fuzzy PID and ANNPID for the extrusion plant with several types of disturbances are tabulated above and results are shown. Results exhibited from proposed method provides less settling time and peak overshoot. Finally, ANN PID was succeded in overcoming the adverse effects of disturbances. Risetime and delay time were not improved but detereorated.
REFERENCES

AntÃ³nio Gaspar Lopes da Cunha Modelling and Optimisation of Single Screw Extrusion

Dr.I.Santi Prabha, K.Durga Rao, Fuzzy Logic Based Intelligent Controller Design for an Injection Mould Machine (IJEST) Vol. No 10, Issue No. 1, 98103.

Fons van de Ven Modelling of Industrial Processes for Polymer Melts_ Extrusion and Injection Moulding Eindhoven University of Technology, Eindhoven, Netherlands.

N.S. Marimuthu and P. Melba Mary, Design of selftuning fuzzy logic controller for the control of an unknown industrial process, IET Control Theory Appl., 2009,
Vol. 3, Iss. 4, pp. 428436

Hongfu Zhou, Simulation on Temperature Fuzzy Control in Injection Mould Machine by Simulink, Asian J. Control, Vol.5, pp. 176186, 2003.

Wolfgang Altmann, Practical Process Control for Engineers and Technicians, IDC Technologies, 2005.

Taur, J.S, C.W.Tao and C.C.Tsai, temperature control of a plastic extrusion barrel using PID Fuzzy controllers, Proc. Of 1995 IEEE conference on industrial automation and control: Emerging technologies Taipei, Taiwan, pp. 370375.

Liu Luoren, Luo Jinling Research of PID Control Algorithm Based on Neural Network Energy Procidia 13(2011) 6988 6993.

Yegna Narayana.B, Artificial Neural Networks, Pretence Hall of India pvt. Ltd, 2006.

S.Rajasekaran, G.A.Vijayalakshmi Pai, Neural Networks, Fuzzy Logic and Genetic Algorithms, PrenticeHall of India Pvt. Ltd., 2006.

JyhShing Roger Jang ANFIS: Adaptive NetworksBased Fuzzy Inference System IEEE transactions on systems, man, and cybernetics, vol. 23, no. 3, mayijune 1993.

F. Shahrakia, M.A. Fanaeib, A.R. Arjomandzadeha Adaptive System Control with PID Neural Networks Department of Chemical Engineering, University of Sistan and Baluchestan, Zahedan, Iran.

Chamil Abeykoon, Kang Li, Marion McAfee Peter J. Martin, George W. Irwin Extruder Melt Temperature Control With Fuzzy Logic 18th IFAC World Congress Milano (Italy) August 28 – September 2, 2011
