Professional Documents
Culture Documents
Control of Flow Rate in Pipeline Using PID Controller: Conference Paper
Control of Flow Rate in Pipeline Using PID Controller: Conference Paper
net/publication/334004148
CITATIONS READS
13 797
4 authors:
Some of the authors of this publication are also working on these related projects:
A Novel Linear PID Controller for an Upper Limb Exoskeleton View project
All content following this page was uploaded by Cristobal Vargas on 01 July 2019.
Abstract— In this paper a PID controller is utilized in order to 16]. Fadlun et al. [17] implemented the concept of [18] to a
control the flow rate of the heavy-oil in pipelines by controlling the finite-difference methodology where a staggered grid is used. In
vibration in motor-pump. A torsional actuator is placed on the [19] a digital pulse feedback flow control system utilizing
motor-pump in order to control the vibration on motor and microcontroller as well as feedback sensing element is
consequently controlling the flow rates in pipelines. The necessary developed. Surprisingly, even though the flow control methods
conditions for asymptotic stability of the proposed controller is are widely spread, investigating the stability of the control
validated by implementing the Lyapunov stability theorem. The system is very rare. In [20-21] the P, PI and PID controls are
theoretical concepts are validated utilizing numerical simulations proposed for flow over a cylinder with a Reynolds number blow
and analysis, which proves the effectiveness of the PID controller
the 200. The aim of control in these studies is the attenuation or
in the control of flow rates in pipelines.
annihilation of vortex shedding behind a bluff body. The only
Keywords—component, formatting, style, styling, insert investigation on the implementation of PI and PID controls to
the flow over a bluff body is carried out in [21]. In the recent
years artificial neural networks (ANNs) have become popular
I. INTRODUCTION and many interesting ANN applications have been reported in
Classic PID approaches as well as controllers are updated engineering field[22-30]. In [31] the authors have develop flow
and expanded during the years, from the primary controllers on modeling and optimization methods using neural networks. In
the basis of the relays as well as synchronous electric motors or [32] a neural network controller approach is proposed in order
pneumatic or hydraulic systems to current microprocessors. to control the pump flow rate. Surprisingly, even though the
Currently, many techniques for the tuning as well as design of flow control methods are widely spread, investigating the
PI and PID controllers are proposed [1]. The method proposed stability of the control system is very rare.
in [2] is the most widely utilized PID parameter tuning
This paper deals with the modeling and control of flow rate
methodology in chemical industry and is considered as a
in heavy-oil pipelines. For this aim, the PID control algorithm is
conventional technique. Basilio and Matos [3] suggested a new
utilized to control the flow mechanism in pipelines. A torsional
method with less complexity in order to tune the parameters of
actuator is placed on the motor-pump in order to control the
PI controllers of the plant with monotonic step response. The
vibration on motor. The stability of the PID controller is verified
methodology of internal mode principle is utilized in [4] and [5]
using Lyapunov stability analysis. The stability analysis of the
in order to extract the gains of PID and PI controllers.
controller results in a theorem, which validate that the system
Exhaustive investigation [6-8] revealed that the outcomes of P
states are bounded. The theoretical concepts are validated using
control are very sensitive to the sensing location as well as the
numerical simulations and analysis, which proves the
quantity of phase shift. By suitable selections of these variables,
effectiveness of the PID controller in the control of flow rates in
the P control can be completely efficient in annihilating the
pipelines. Further, it is the first attempt to place a torsional
vortex shedding or minimizing its strength. Furthermore, it is
actuator on the motor-pump in order to control the vibration on
demonstrated that the increment in the proportional gain can
motor and hence control the flow rates in pipelines.
results in the decrement of the velocity fluctuations in the wake
and the strength of vortex shedding. Nevertheless, a large gain This paper is structured as follows. Firstly, in Section II the
causes instability in the system [9-10]. pump model system is established. The PID control method is
described in Section III. In this section, sufficient conditions for
In order to implement the control law, the primary step is to
the controller under the Lyapunov stability theorem are
determine a desired output response of a particular system to an
designed. A numerical example is presented in Section IV to
arbitrary input over a time interval, that can be carried out by
illustrate the results. Finally, the conclusions are provided in
system identification [11]. Generally, it is feasible to generate a
Section V.
model on the basis of a complete physical illustration of the
system. Nevertheless, this model contains complexity, also has
high calculation costs [12]. The secondary step is to define the II. MATERIALS AND METHODS FOR MODELLING OF THE
parameters of the PID controller. There exist various literatures SYSTEM
associated with the methodologies for tuning of PID controllers Flow control loop system is basically a feedback control
applied in various controller structures [13].
system. The structure of the pump model system is shown in
Flow control is a major rapidly evolving field of fluid Fig.1. which is an open loop system. If there is unwanted
mechanics. There have been various concepts of flow control in vibration in the motor, the stability of the flow rate will hamper.
drag reduction, lift enhancement, mixing enhancement, etc. [14- Therefore, it is important to change the open loop system to a
closed loop system by implementing a controller so as to control where ܮis taken to be the distance between two sections. Now
the stability of flow rate by controlling the vibration in the motor. let,
ߙ ൌ ିଵ ǡ݅ ൌ ͳǡ ǥ ǡ ݊ (5)
where ߙ is termed as the coefficient of pressure changes in
sections
߲ݑ ͳ (6)
ൌ െ ሺͳ െ ߙሻ ܨ ǡ݅ ൌ ͳǡ ǥ ǡ ݊
߲ݐ ߩܮ
The loss of the friction under the conditions of laminar flow
conforms with the Hagen–Poiseuille equation [35-36]. For a
circular pipe having a fluid of density () and Kinematic viscosity
߭, the hydraulic slope ܨ can be described as,
Ͷ ݑଶ Ͷ߭ ݑ (7)
Fig.1. Scheme of open loop model ܨ ൌ ܨ ൌ ܨ
ܴ݁ ʹ݃ܦ ʹ݃ ܦଶ
where ݃ is the gravity, ܦis the diameter of the pipes and ܨ is
A. Modelling of the pipeline the shape force vector in pipes. By substitution of (7) in (6), the
The proposed model consists of induction motor, which causes following equation can be extracted,
a rotation in pump and consequently can lead to flow of heavy- ߲ݑ ͳ Ͷ߭ ݑ
oil in pipelines as shown in (Fig.1). This flow model can be ൌ െ ሺͳ െ ߙሻ ܨ ǡ
߲ݐ ߩܮ ʹ݃ ܦଶ (8)
illustrated in the form of partial differential equation (PDE) [33]. ݅ ൌ ͳǡ ǥ ǡ ݊
The linear global theory associated with flow stability is rooted
The pressure ሺݔǡ ݐሻ in the pipeline can be described as,
in eigendecompositions of the linearized flow operators. The ܨ (9)
direct as well as adjoint eigendecompositions associated with ൌ
these kinds of operators generate information related to the ܣ
where F is the force inside the pipeline, also A is the cross
stability of the operator, the acceptance of initial conditions as section in the pipe.
well as external forcing, also the sensitivity to spatially localized ௗమ ௫ ௗ௫
disturbances. By taking into consideration ܨൌ ݉ܽ ൌ ݉ మ ǡ ݑൌ ,
ௗ ௧ ௗ௧
From the viewpoint of incompressible, constant-density, also by substitution of (9) in (8) the following equation is
constant-viscosity flows associated with Newtonian fluids, the extracted,
nonlinear Navier–Stokes equation is defined for a ߲ ߲ݔ ሺͳ െ ߙሻ ߲ ଶ ݔ Ͷ߭ ߲ݔ (10)
nondimensional velocity field ሺݔǡ ݐሻǣ Թ ൈ Թ ՜ Թ , pressure ൌെ ଶ
ଶ
ܨ
߲ݐ߲ ݐ ߩݐ߲ ܮܣ ʹ݃ݐ߲ ܦ
field ሺݔǡ ݐሻǣ Թ ൈ Թ ՜ Թ and Reynolds number ܴ݁ Ͳ as Therefore,
below equation, ሺͳ െ ߙሻ ߩ ߲ ܮܣଶ ݔ Ͷ߭ ߲ݔ (11)
߲ݑ (1) െ ଶ
ଶ
ܨ ൌ Ͳ
ൌെ െ ݑǤ ݑ ܨ ߩܮܣ ߲ݐ ʹ݃ݐ߲ ܦ
߲ݐ ߩ If an external force, ݂ generates by pump, (11) can be
Where,
rewritten as follows,
ߩ is the density in య Ȟݔሷ Ȱݔሶ ݂ ൌ ݂ (12)
u is the flow velocity in , where אԹଶ , Ȟ אԹଶൈଶ , Ȱ אԹଶൈଶ , ൌ ሾ݂ଵ ݂ଶ ሿ் א
௦
is the divergence, Թଶൈଵ ,݂ ൌ ሾ݂ଵ ݂ଶ ሿ் אԹଶൈଵ .
Since in this work two pipelines are used, (12) can be rewritten
is the pressure in మ,
Ǥ௦ as,
ݐis time in ݏ,
ܨ is termed as the summation of external force and body ߛଵ ݔሷଵ ߮ଵ ݔሶ ݂ଵ ൌ ݂ଵ
(13)
forces ߛଶ ݔሷଵ ߮ଶ ݔሶ ݂ଶ ൌ ݂ଶ
By implementing the mass balance into the Equation (1) the Since the pump is supplying pressure to the pipes for the
following is concluded [34], maintaining the flow rates, so the pipes will have the same
Ǥ ݑൌ Ͳ (2) external force, ݂ଵ ൌ ݂ଶ .
Equation (1) can be rewritten as,
߲ݑ ͳ ߲ (3) B. Modelling of the actuator
ൌെ ܨ
߲ݐ ߩ ߲ݔ In order to reduce the vibrations of the motor caused by the
డ external forces (݂ ) a torsional actuator is placed on the motor,
Let be the change of pressure in two different points,
డ௫
moreover for achieving a numerical stability of computation, it is see Figure. 2.
essential to partition pipeline into various segment (generally
identical), hence the flow in pipeline can be stated as,
߲ݑ ͳ (4)
ൌ െ ሺ െ ିଵ ሻ ܨ ǡ݅ ൌ ͳǡ ǥ ǡ ݊
߲ݐ ߩܮ
3URFHHGLQJVRIWKH,(((WK,QWHUQDWLRQDO&RQIHUHQFHRQ1HWZRUNLQJ6HQVLQJDQG&RQWURO0D\%DQII$OEHUWD&DQDGD
3URFHHGLQJVRIWKH,(((WK,QWHUQDWLRQDO&RQIHUHQFHRQ1HWZRUNLQJ6HQVLQJDQG&RQWURO0D\%DQII$OEHUWD&DQDGD
௧ ௧ ௧
න ݂݀ ݔൌ න ܨ ݀ ݔ න ݀௨ ݀ݔ (31)
௧ ௧
The lower bound of is stated as െܨ also
ܨ ݀ݔ ݀௨ ݀ݔ
as -
௨ . Therefore, the lower bound of ȳ is defined as,
ܦ
ȳ ൌ െܨ െ ܦ ௨ (32)
The stability analysis of PID control approach is given by the
below mentioned theorem.
Theorem. By taking into consideration the structural system
Fig. 4. PID controller of equation (12) controlled by the PID control approach of
equation (22), the closed-loop system of equations (28) is taken
The PID control is expressed as, to be asymptotically stable at the equilibriums ൣߴ െ
௧ ்
Ȳሺݐሻ ൌ െߢ ݁ሺݐሻെߢ න ݁ሺݐሻ݀߬ െ ߢௗ ݁ሶ ሺݐሻ (22) ݂ሺͲሻǡ ܺǡ ܺሶ ൧ ൌ Ͳ, if the following gains are satisfied,
ଵΤଶ
ͳ ͳ
where ݇ , ݇ , as well as ݇ௗ are positive definite and ݇ is the ߣ ሺߢௗ ሻ ቆ ߣ ሺȞሻߣ ൫ߢ ൯ቇ ͳ
integration gain. For the flow control,ܺ ௗ is desired reference and Ͷ ͵
݇
also ܺ ௗ ൌ ܺሶ ௗ ൌ Ͳ. Hence, equation (22) is rewritten as below, ൨ െ ߣ ሺȰሻ
௧ ߣெ௫ ሺȞሻ
Ȳሺݐሻ ൌ െߢ ܺെߢ න ܺ݀߬ െ ߢௗ ܺሶ (23) ߣெ௫ ሺߢ ሻ (33)
ଵΤଶ
ͳ ͳ ߣ ൫ߢ ൯
For analyzing the PID controller, equation (22) can be stated ቆ ߣ ሺȞሻߣ ൫ߢ ൯ቇ ቈ
as below, ͵ ߣெ௫ ሺȞሻ
͵
Ȳሺݐሻ ൌ െߢ ܺ െ ߢௗ ܺሶ െ ߴ ߣ ൫ߢ ൯ ሾȳ ȩሿ
௧
(24) ʹ
ߴ ൌ ߢ න ܺ݀߬ ǡߴሺͲሻ ൌ Ͳ
IV. NUMERICAL RESULTS
The closed-loop system of eq. (12) along with the PID control For the numerical analysis purpose and for the validation of
(eq. (23)) is demonstrated as below, the novel control strategy, the various parameters associated with
Ȟݔሷ Ȱݔሶ ܨ ൌ െߢ ܺ െ ߢௗ ܺሶ െ ߴ the flow control are described in Table1:
(25)
ߴሶ ൌ ߢ ܺ
In matrix form, the closed-loop system is defined as, TABLE 1. PARAMETERS ASSOCIATED WITH THE FLOW CONTROL
ߢ ܺ
݀ ߴ ߩ 1240 ݇݃Τ݉ଷ
ܺ ൩ ൌ ܺሶ (26) ߭ ͳǤͲͶ ൈ ͳͲିଷ ݉ଶ Τݏ
݀ ݐሶ
ܺ െȞ ିଵ ሺȰܺሶ ܨ ݇ ܺ ݇ௗ ܺሶ ߴሻ ݃ 9.81 ݉Τ ݏଶ
Here the stability of the PID control demonstrated by eq. (23) 100 ݉
is analyzed. The equilibrium of eq. (26) is presented ߙ 0.95 -
ܦ 0.05 ݉
byሾߴ ܺ ܺሶ ሿ ൌ ሾߴመ Ͳ Ͳሿ. As at equilibrium point ܺ ൌ Ͳ as
ܣ ͳǤͻ ൈ ͳͲିଷ ݉ଶ
well asܺሶ ൌ Ͳ , the equilibrium is ሾ݂ሺͲሻǡ Ͳǡ Ͳሿ. For moving ܹ 30 ݇݃
the equilibrium to the origin, the following is defined, ݎ௧ 0.3 ݉
ߴመ ൌ ߴ െ ݂ሺͲሻ (27) ݉௧ 1.5 ݇݃
Therefor, the final closed-loop equation is defined as,
Ȟܺሷ Ȱܺሶ ܨ ൌ െߢ ܺ െ ߢௗ ܺሶ െ ߴ ݂ሺͲሻ The implemented software in this paper is Matlab/ Simulink.
(28) Simulations are presented to show that the motor vibration can
ߴ ൌ ߢ ݔ
For analyzing the stability of equations (28), the following be attenuated to a signicant level by using the torsional actuator
properties are required, with the developed controllers thus validating the effectiveness
Property 1. The positive definite matrix Ȟ should satisfy in of the proposed control approach using PID controllers. A
the below condition simulation period of 20s is considered for evaluation. For the
Ͳ ൏ ߣ ሺȞሻ ԡȞԡ ߣெ௫ ሺȞሻ ߛҧ (29) simulation purposes, the weight of the torsional actuator is
such that ߣ ሺȞሻas well as ߣெ௫ ሺȞሻ are considered as the considered to be 5% of the motor and pump weight in
minimum and maximum eigenvalues of the matrix Ȟ , combination.
The Theorem proposed in this paper generates sufficient
respectively also ߛҧ Ͳ is taken to be the upper bound.
conditions for the minimum amounts of the proportional as well
Property 2. ݂ is taken to be Lipschitz over ݔ and ݕ if
as the derivative gains. This Theorem validates that both
ሻ െ ݂ሺݕሻԡ ȳԡݔ
ԡ݂ሺݔ െ ݕԡ (30)
proportional and derivative gain must be positive as negative
As ܨ is first-order continuous functions also satisfies in
gains can make the systems unstable. The PID gains are selected
Lipschitz condition, Property 2 is hereby established. within the stable range by the stability analysis in order to ensure
The lower bound of ܨ can be calculated as below, the efficiency.
3URFHHGLQJVRIWKH,(((WK,QWHUQDWLRQDO&RQIHUHQFHRQ1HWZRUNLQJ6HQVLQJDQG&RQWURO0D\%DQII$OEHUWD&DQDGD
3URFHHGLQJVRIWKH,(((WK,QWHUQDWLRQDO&RQIHUHQFHRQ1HWZRUNLQJ6HQVLQJDQG&RQWURO0D\%DQII$OEHUWD&DQDGD
[8] E. Berger, "Suppression of vortex shedding and turbulence Conference on Systems, Man, and Cybernetics, Budapest,
behind oscillating cylinders," Physics of Fluids, vol. 10, p. Hungary, 2016.
S191–S193, 1967. [26] S. Razvarz., R. Jafari., O.Ch. Granmo., A. Gegov, "Solution of
[9] K. Roussopoulos, "Feedback control of vortex shedding at low dual fuzzy equations using a new iterative method,," In
Reynolds numbers," Journal of Fluid Mechanics, vol. 248, pp. Proceedings of the 10th Asian Conference on Intelligent
267-296, 1993 . Information and Database Systems, Lecture Notes in Artificial
[10] S. Hiejima, T. Kumao, T. Taniguchi, "Feedback control of Intelligence (subseries of LNCS), pp. 245-255, 2018.
vortex shedding around a bluff body by velocity excitation.," [27] S. Razvarz., R. Jafari., A. Gegov, "A New Computational
Int. J. Comput. Fluid Dynam., vol. 19, pp. 87-92, 2005. Method for Solving Fully Fuzzy Nonlinear Systems,," In:
[11] S. W. Sung , J. Lee, I-B. Lee , Process identication and PID Computational Collective Intelligence. ICCCI 2018. Lecture
control., IEEE Press, 2009. Notes in Computer Science, Springer, Cham, , vol. 11055, pp.
503-512, 2018.
[12] A. Kusters , "MIMO system identica-tion of a slab reheating
furnace," in Proceedings of the Third IEEE Conference on [28] R. jafari., w. Yu., "Fuzzy Modeling for Uncertainty Nonlinear
Control Applications. : IEEE Conference Publications, USA, Systems with Fuzzy Equations," Mathematical problems in
1994. Engineering., vol. 2017, 2017.
[13] A. O. Dwyer, Handbook of PI and PID controller tuning rules, [29] R. Jafari.,S. Razvarz., A. Gegov, S. Paul., , "Fuzzy modeling
London:: Imperial College Press, 2009. for uncertain nonlinear systems using fuzzy equations and Z-
numbers," Advances in Computational Intelligence Systems:
[14] H. Choi, W. P. Jeon, J. Kim, "Control of flow over a bluff Contributions Presented at the 18th UK Workshop on
body," in Annual Review of Fluid Mechanics, 2008. Computational Intelligence, September 5-7, 2018,
[15] S. S. Collis, R. D. Joslin, A. Seifert, V. Theofilis, "Issues in Nottingham, UK . Advances in Intelligent Systems and
active flow control: theory,Issues in active flow control: Computing, Springer,, no. 2018, pp. 66-107, 840.
theory, control, simulation, and experiment," Progress in [30] S.Razvarz, R. Jafari, A. Gegov, W. Yu, S. Paul, , Neural
Aerospace Sciences, p. 237–289, 2004. network approach to solving fully fuzzy nonlinear systems,
[16] M. Gad-el-Hak, A. Pollard, J. P.Bonnet, Flow control: Fuzzy modeling and control Methods Application and
fundamentals and practices, Berlin, Germany: Springer, 1998. Research, NewYork: Nova science publisher, Inc,, 2018.
[17] E. A. Fadlun, R. Verzicco, P. Orlandi, and J. Mohd-Yusof,, " [31] S. Muller., M. Milano., P. Koumoutsakos, "Application of
Combined immersed-boundary finite-difference methods for machine learning algorithms to flow modeling and
three-dimensional complex flow simulations," J. Comput. optimization," Center for Turbulence Research Annual
Phys., vol. 161, pp. 35-60, 2000. Research Briefs , vol. 1, pp. 169-178, 1999.
[18] J. Mohd-Yusof, "Combined Immersed-Boundary/B-Spline [32] A. A. Aly, "An Artificial Neural Network Flow Control of
Methods for Simulations of Flow in Complex Geometries,," Variable Displacement Piston Pump with Pressure
Annual Research Briefs (Center for Turbulence Research, Compensation," INTERNATIONAL JOURNAL OF
NASA Ames and Stanford University, pp. 317-327, 1997. CONTROL, AUTOMATION AND SYSTEMS, vol. 4, no. 1, pp.
[19] A. Nandy , S. Mondal, P. Chakraborty, G.C. Nandi, 1-7, 2015.
"Development of a Robust Microcontroller Based Intelligent [33] A.Herrán-González ., J.M.De La Cruz ., B.De Andrés-Toro.,
Prosthetic Limb," International Conference on Contemporary J.L.Risco-Martín, "Modeling and simulation of a gas
Computing, pp. 452-462, 2012. distribution pipeline network," Applied Mathematical
[20] D. S. Park, D. M. Ladd, E. W. Hendricks, "Feedback control Modelling, vol. 33, no. 3, pp. 1584-1600, 2009.
of von Kármán vortex shedding behind a circular cylinder at [34] R. D. Whitaker, "An historical note on the conservation of
low Reynolds numbers," Phys. Fluids, vol. 6, no. 1, p. 2390– mass," Journal of Chemical Education, vol. 52, no. 10, pp.
2405, 1994. 658-665, 1975.
[21] E. Berger, "Suppression of vortex shedding and turbulence [35] L. F. Moody, "Friction factors for pipe flow," Transactions of
behind oscillating cylinders.," Phys. Fluids, vol. 10, pp. 191- the ASME, vol. 66, no. 8, p. 671–684, 1944.
193, 1967. [36] G.O. Brown, "The History of the Darcy-Weisbach Equation
[22] R. Jafari, W. Yu , "Uncertainty Nonlinear Systems Control for Pipe Flow Resistance," in Environmental and Water
with Fuzzy Equations," IEEE International Conference on Resources History. American Society of Civil Engineers, 2003.
Systems, Man, and Cybernetics,, pp. 2885-2890, 2015. [37] C. Rolda´n , F. J. Campa, O. Altuzarra , et al., "Automatic
[23] R. jafari., w. Yu., "Artificial neural network approach for identification of the inertia and friction of an electromechanical
solving strongly degenerate parabolic and burgers-fisher actuator," Mechanisms and Machine Science, vol. 17, p. 409–
equations,," 12th International Conference on Electrical 416., 2014.
Engineering, Computing Science and Automatic Control,, [38] L. wozniak, "A graphocal approach to hydrogenetor governer
2015. tuning," transaction on energy conversion , vol. 5, no. 3, pp.
[24] R. jafari., w. Yu., "ncertain nonlinear system control with 417-421, 1990.
fuzzy differential equations and Z-numbers," in 18th IEEE [39] K. De Cock,B. De Moor ,W. Minten , et al., A tutorial on PID
International Conference on Industrial Technology, canada, control. Report ESAT-SISTA/TR., Leuven, Belgium:
2017. Katholieke Universiteit Leuven, 1997.
[25] R. jafari., w. Yu., X. li, "Solving Fuzzy Differential Equation
with Bernstein Neural Networks," in IEEE International
.