AVEEK MANGAL

Optimal control of induction machines under

minimum energy in opencast mining machinery
Mining industry is one of the most important users of electric speeds, such as mining conveyors, crushers and draglines.
motors. The most commonly used in the contemporary Control for these motors is focused on reliable, safe starting.
mining industry is alternating current (AC) machines which In addition to controlling speed, you also have to consider
are used for converting electrical energy into mechanical torque requirements. Torque control focuses on providing
energy. This paper presents the solution of the nonlinear adequate torque for the load at all operating points, and on
optimal control problems of three-phase induction mining protecting the equipment and people from excess torque and
machinery (IMM). A fifth order nonlinear model has been currents. Motor torque and motor current are roughly
described in arbitrary rotating frame (d-q) of induction proportional (Paul and William, 2017).
machine which is used in this paper along with a quadratic
Electrical machinery for mines usually consists of AC
performance index (QPI). The problem has solved using the
power consumers based on a squirrel cage induction motor
Taylor expansion about an operating point method which
such as exhauster fans, deep well pumps for pumping water
converts the nonlinear optimal control problem into
out of mines, grinding machines for ore reduction, drilling
sequence of linear quadratic optimal control problems. The
machines, hoisting equipment, etc. (Linkov and Olizarenko,
control objective is to minimize the total energy, ensuring
2015). The main objective in conservation of electrical energy
torque and speed tracking control requirements. The
in mines includes replacement of conventional systems with
evaluation of a simulation for an electrical drive application
the present technology system which leads to reduction in
shows that operation at varying optimal supply voltages
energy consumption per annum for about 10-15% (Ganapathi
(Vdsopt, Vqsopt) preserves well speed, torque and flux tracking
and Manjunath, 2016).
performances, while increasing motor efficiency. The system
is stable with the properly selected settings of optimal The consumption of both thermal and electrical energy is
regulators, which are particularly applicable during the growing, no matter the political intentions, so sooner and later
operation of machinery used in mining exploitations. very large investments in transmissions lines and power
Keywords: Mining industry, induction motors, stations are required. For the end-user there is of course the
minimization of energy, Riccati equations, torque and speed simple incentive that gets the same work done with less
control. energy by using more efficient energy conversion processes
(M. Shepard, 1992).
1. Introduction Induction motors are also attractive in new electric

T
he basic need for efficient use of energy imposes generation shops, water pumps and electric transportation
requirements of world law on the mining companies. systems (Abrahamsen, 1997). These applications require
Mining industry is one of the recipients of electric different power levels ranging from a few to several hundred
motors. Mining machinery for coal exploitation and mineral horse power. In fact, electrical motors consume around 56%
processing, machinery for mining and equipment for mineral of the total consumed electrical energy, and of this, induction
raw materials, generally represent the industry in which motors account for 96% (M. Shepard, 1992). This shows that
development and manufacturing experience plays a around 53% of the total electrical energy is consumed by
considerable role (Pytel and Jaracz, 2012; Gumula and Pytel, induction motors. There are general purpose loss minimization
2014). Although DC motors are still widely used in draglines techniques approaches of which are different depending on
and shovels, for the purpose of this research paper, only the induction motor drives applications. By using the
three-phase AC machines will be addressed, particularly advances in power electronics and signal processing
induction machines. Some machines work well at fixed technologies, the control schemes of induction machines
have been improved from simple scalar or auto-tuning control
Mr. Aveek Mangal, Department of Mining Engineering, Indian Institute
of Technology, Indian School of Mines, Dhanbad, Jharkhand, 826004,
to field oriented and direct torque controls (Abourida, 2009);
India (Bazzi, 2009); (El Refaei, 2005); (Krause, 1986); (Siva Reddy,

JOURNAL OF MINES, METALS & FUELS 229

2000); (Vas, 1990). The field oriented control is successfully variables the projections of the stator currents and rotor flux
applied in real time control of industrial applications when (ids iqs dr qr) on a moving frame into – axis coil (d - q) model
dealing with high performance induction machines drives on both stator and rotor as described by Krause and Thomas
(Besanelle, 2001); (Holtz, 2006); (Inanc, 2007); (Novotny, (S. I. Seleme Jr, 1993); (A. Bentounsi, 2002). Fig.1 shows the
1996); (Singh, 2005); (Wang, 2007). Field oriented control axial view of an induction machine (L Kirtley, 2003).
implementation can achieve robust control performance in The nonlinear differential equations that describe the
induction machines. Estimating the magnitude and phase of dynamics of an ideal symmetrical induction motor in a rotating
rotor flux is the key of the control implementation
(Abrahamsen, 1997); (Camblong, 2006); (Jemli, 2000). Direct
methods in sensing the rotor flux by using suitable hardware
around the machine have proved to be inaccurate and
impractical at speed and torque variations (Geng, 2004).
Indirect methods of sensing the rotor flux employ a
mathematical model in induction machine by measuring state
variables of currents, voltages and rotor position. The
accuracy of this method depends mainly on accuracy of time
rotor constant variation (Grouni, 2008); (Marino, 2008);
(Wang, 2007). The machine parameters may change during
operation which introduces incracies in the estimated flux
(Besanelle, 2001). With an aim to improve performance of
systems control of induction machine, research has been
conducted to design advanced nonlinear controllers (Dhaoui,
2010); (Faiz, 2006); (Holtz, 2006). Most of these controls
operate at constant flux norms (Levi, 1995); (Krishan, 2001).
Fig.1 Axial view of an induction machine (L Kirtley. Jr. 2003)
In this situation, efficiency is maximized only when the system
operates at its nominal torque (Ren, 2008), (Ta, 2001). Away frame is as follows (S. I. Seleme Jr, 1994); (Park, 1991) shows
from this operating point the machine will dissipate a the assignation of these projections for components of the
considerable part of the injected electrical power as core five dimensional state vector.
losses and it may inefficiently store too much energy in its
coil inductances (Abrahamsen, 1997); (Thanga Raj, 2009). In ... (1)
most application, induction machines do not operate at the
nominal rate since the desired torque may change on-line or The inputs to the system are the projections of the motor
may depend on system states such as position or velocity. It supply voltage (Vs) on the (d–q) axes frame.
is then technically and economically interesting to investigate
other modes of flux operation seeking to optimize system ... (2)
performance (Holtz, 2006), (S. Grouni, 2010). Aware of these
facts, some previous research works have already used the The output is given by a nonlinear relation between the
reference flux as an additional parameter seeking to increase projections of the flux and currents on the (d–q) axes frame.
machine efficiency or to maximize the delivered torque in
minimum time (Besanelle, 2001); (Camblong, 2006); (Inanc, ... (3)
2007). The evolution of microprocessor technology and the
need for high-performance controllers for drives using For large machines, for which the minimum energy
induction motors have rise to intensive investigation on criterion is important, the mechanical dynamics of the motor
advanced control design for AC-drives. The induction motor (the mechanical speed wr) is slow enough compared to the w
is known for its versatility and simplicity. It is capable of electrical dynamics of the torque transient so that can be
operating at wide ranges of speeds as well as wide rangers of considered as invariant with respect to the state x.
torques. (S. I. Seleme, 1992). In this paper the author considers the problem of
In this paper, a new optimal control method is suggested analytically finding optimal voltages references (Fig.2) that
to obtain a minimum total energy of induction motors under minimize total energy.
optimal trajectories of the induction motor’s, voltages, torque
The model of one pole pair induction motor can be given
and speed.
by:
2. Dynamic and energy equations ... (4)
It is useful to describe the induction motor model by a set of
four non-linear differential equations and to choose as state ... (5)

230 JULY 2020

 control problem such as direct method, discretization method,
parameterization method and Taylor expansion about an
operating point method (C.J. Goh, 1988); (P.A. Frick, 1995); (J.
Vlassenbroeck, 1988); (H. Jaddu, 1999).
To solve this optimal control problem under nonlinear
state equation (6), in our paper we using the Taylor expansion
, about an operating point
, to convert the nonlinear
optimal control problem into sequence of linear quadratic
optimal problems which can be solved by solving sequence
of Riccati equations. The optimal trajectories of the induction
motor’s, voltages, torque and speed are presented in this
paper.
Fig.2. Simulation result of stator voltage input (axe-q) motor with
and without optimal controller Where,

3. Modelling hypothesis and model properties

Validity of the model AC-machines are normally, three phase
electrical devices which can be reduced to a two phase model
due to their symmetry. The classical two phase textbook form
of the induction motor model is: ... (7)

To convert our optimal control problem of induction

machine to sequence of linear quadratic optimal control
problems we first linearize the state space equation (6) by
Taylor expansion about an operating point (7) to get:

... (6)

Where, (ids iqs dr qr)T are the dq-axis equivalent stator
currents, dq-axis equivalent rotor fluxes and mechanical speed,
Vds Vqs Tl are the dq-axis equivalent stator voltages and
the torque load, Ls, Lr and Msr are the stator, rotor and mutual
... (8)
inductance, R s, R r are the stator and rotor resistances

respectively, is the rotor time constant,

is the leakage coefficient of the machine, F is the viscous 4. Optimal control design methods
friction coefficient of machine and load, J is the inertia Optimal control is branch of modern control theory that deals
constant of the machine, P is the number of pole pairs. with designing controls for dynamic systems by minimizing a
The validity of this model is based on the assumptions performance index that depends on the system variables. (Kh
already mentioned at the beginning of section five which Yousfi, 2005).
concern the attribution of constant values for the inductances In this paper, the author discusses the design of optimal
and resistances of the stator and rotor electric circuit. controller for nonlinear systems which quadratic performance
There are several methods to solve this non linear optimal index. The object of the optimal regulator design is to

JOURNAL OF MINES, METALS & FUELS 231

determine the optimal control law [Uopt (x,t)] which can Finally, equating (15) with (18), we obtain:
transfer the system from its initial state to the final state such
that a given performance index is minimized. The performance ... (19)
index is selected to give the best trade-off between
performance and cost of control. The performance index that ... (20)
widely used in optimal control design is known as the The above equation is referred to as the matrix Riccati
quadratic performance index and is based on minimum error equation. The boundary condition for (20) is p(t f)=0).
and minimum energy criteria. Therefore, (21) must be integrated backward in time. Since a
Equation (8) represents the linear fifth order model for the numerical solution is performed forward in time, a dummy time
induction machine which can be rewritten in a compact form: variable ( = t f – t ) is replaced for time (t). Once the solution
to (20) is obtained the solution of the state equation (13) in
... (9) conjunction with the optimum control equation (17):
The problem is to find the vector [k(t)] of the optimal ... (21)
control law:
ki = diag(K) ... (22)
... (10)
The design procedure is in stark contrast to classical
This minimizes the value of a quadratic performance index control design, where gain matrix (K) is selected directly:
(J) of the form:
... (11)
Subject to the dynamic plant equation (4) in (11),
Where: Q is a positive semi definite matrix and R is a real ... (23)
symmetric matrix, if all principal minors are nonnegative.
The choice of the elements of (Q) and (R) allows the This has the significant advantages of allowing the entire
relative weighting of individual state variables and individual control loop in a multi-loop system to be closed
control inputs. simultaneously, while guaranteeing closed loop stability. The
To obtain a formal solution, we can use the model of optimal control law is assuring the minimum energy (Figs.3
Lagrange multipliers. The constraint problem is solved by and 4), maximum torque (Fig.5) and optimal trajectory of state
augmenting (4) in (11) using the vector of Lagrange multipliers (IM speed), (Fig.6).
(). The problem reduces to the minimization of the following From the following velocity differential equation is
function. obtained:
... (12)
... (24)
The optimal values are found by equating the partial
derivatives to zero: To look at the reliability of the optimal controller on result
them of the active and reactive losses of the plant to consider:
... (13) P1 = wr.Te ... (25)
... (26)
... (14)

... (15) ... (27)

Assume that there exists a symmetric, time-varying The total active and reactive losses in the induction motor
positive definite matrix (p(t)) satisfying: are:
... (16) Pa = P1+P2+P3 ... (28)
Substituting (16) into (14) gives the optimal closed loop
... (29)
control law:
... (17) 5. Simulation results
Obtaining derivate of (16), we have: In this paper the author has confirmed the validity of the
proposed optimal control by Matlab software simulation for
... (18)
an induction machine drive system (Table 1). (Figs.7 to 9) are

232 JULY 2020

 Fig.3. Simulation result of active energy with and without optimal Fig.6 Simulation result of speed with and without optimal controller
controller

Fig.7 Simulation result of reactive power with and without optimal

Fig.4. Simulation result of reactive energy with and without optimal controller
controller
TABLE 1: INDUCTION MACHINE PARAMETERS
Parameters Nomenclature Item (Unit) Value
Vn Rated supply voltage Volt 220/380
fn Rated supply frequency Hz 50
P Number pair of poles / 2
j Inertia constant Kg.m 2 0.089
F Viscous friction constant Nm/rad/s 0.00187
Rs Stator resistance  0.435
Rr Rotor resistance  0.816
Ls=Lr Stator and rotor inductance H 0.754
Msr Mutual inductance H 0.261

shown the simulation models of three phase induction motor

with and out optimal controller and optimal controller
respectively.
The optimal trajectories of the induction motor’s voltages,
Fig.5 Simulation result of torque with and without optimal controller torque and speed are perfectly realized under optimal energy.

JOURNAL OF MINES, METALS & FUELS 233

 Fig.8. Schema block of three phase Induction Motor with and without optimal controller

beneficial in terms of money and also reduction in global

warning. This paper presented a review of the developments
in the field of efficiency optimization of three-phase induction
motor through optimal control and design technique. Optimal
control covered both the broad approaches namely, loss
model control and search control. Optimal design covers the
design modifications of materials and construction in order
to optimize efficiency of the motor, particularly for the mining
machinery such as mining conveyors, crushers and draglines
Fig.9 Schema block of optimal Controller etc.

6. Conclusion Acknowledgement

In this paper we obtained the optimal trajectory of state (IM The author of this paper is very much thankful to the all
speed) and control (IM torque) under optimal energy. These faculty members of Department of Mining Engineering, Indian
trajectories are obtained by minimizing the quadratic Institute of Technology (Indian School of Mines) Dhanbad
performance measure or total energy (the stored magnetic for their valuable advice, resourceful guidance and
energy and the coil losses) of the induction motor. Since the continuous inspiration throughout the preparation of this
induction machine was used to this nonlinear model, the paper. The views expressed in this paper is the author’s and
Taylor expansion about an operating point to convert the not necessarily of the organizations to which he serves.
nonlinear optimal control problem by sequence of linear
References
quadratic optimal problem. The proposed minimum energy
approach may be of great interest for application in areas 1. A. Bentounsi and K. Yousfi., Riyadh, (2002): Detection
where energy consumption is an import issue; i.e. high speed of Faults In Induction Motor Coupled To PWM
electric trains and the expected new generation of electrical Inverter, STCEX 2002, 26-30 October.
vehicles. Efficiency optimization is very much essential not 2. Abourida, S., Belanger, J., Shariah, U.E.A, (2009): Real
only to electrical systems, it require all the systems to get time platform for the control prototyping and

234 JULY 2020

 simulation of power electronics and motor drive, Pol. J. Environ. Stud. 23, 6, 2321-2325.
Proceeding of the Third International Conference on 15. Holtz, J., (2006): Sensorless Control of Induction
Modelling, Simulation and Applied Optimization Machines with or without Signal injection, Overview
(ICMSAO), 20-22 January. Paper, IEEE Trans. On Ind, Elect, Vol (53): No. 1.pp.7-
3. Abrahamsen, F., Blâbjer, F., J.K. Pederson, J.K. , 30.
Trondheim, Norway (1997): Analysis of Stability in low 16. H. Jaddu and E. Shimemura. (1999): Combination of
cost Energy Optimal Controlled PWM VSI fed optimal control trajectories using Chebyshev
induction motor drive, Proceedings of EPE’97, pp. polynomials: Parameterization and quadratic
3.717-3.723. programming, Optimal Control Appl. Methods,
4. Bazzi, A.M., Kerin, P.T., (2009): Asurvey of real time Vol(20): pp. 21-42.
power loss minimizers for induction motors, IEEE 17. Inanc, N., (2007): A robust sliding mode flux and speed
Trans. Ind. Elec. Vol (43), No. 1, pp. 226-231. observer for speed sensorless control of an indirect
5. Besanelle, A. S., R., (2001): Robust tuning of the speed field oriented induction motor drives, Science Direct,
loop in indirect field oriented control induction motors, E lectric Power System Research Vol(77): pp.1681-
Journal Automatica, Vol.03: pp.1811-1818. 1688.
6. Camblong, H., Martinez de Alegria, I., Rodriguez, M., 18. J. Vlassenbroeck and R. R. Doreen, A. Chebyshev. ,
Abad, G. , (2006): Experimental evaluation of wind (1988): Technique for solving nonlinear optimal control
turbines maximum power point tracking controllers, problem, IEEE Trans. Automatic. Control, Vol(33): pp.
Energy conversion and Management, Vol(47): pp.2846- 333-340.
2858. 19. Jemli, M., Boussak, M., Gossa, M., Kamoun, M.B.A.
7. C.J. Goh and K.L. Teo. (1988): Control parameterization: (2000): Fail-safe digital implantation of indirect field
unified approach to optimal control problem with oriented controlled induction motor drive, Simulation
general constraints, Automatica, Vol (24): pp. 3-18. Practice and Theory, ELSVIER, Vol(8): pp.233-252.
8. Dhaoui, M., Sbita, L. (2010): Anew method for loss 20. Kh Yousfi.(2005): Modeling of High – Order Systems
minimization in IFOC induction motor drives, with an Optimal Approximation, Umm Al-Qura
International Journal of Systems Control, IJSC, Vol University Journal of Science – Medicine, ISSN: 1319-
(1): Issue.2, pp.93-99. 8904 -Vol.17, No.1, 1425.
9. El Refaei, A., Mahmoud, S., Kennel, R., (2005): Torque 21. Krause, P. C., Wasynczuk, O., Sudhoff, S.D., (1986):
Ripple Minimization for Induction motor Drives with Analysis of electric Machinery, Series in Electrical
Direct Torque Control (DTC), Electric Power Engineering, MGGraw Hill, USA.
Components & Systems, Vol(33): No.8. 22. Krishan, R. (2001): Electric Motor Drives, Modeling,
10. Faiz, J., Sharifian, M.B.B. (2006): Optimal design of an Analysis and Control, Prentice Hall, USA (2001).
induction motor for an electric vehicle, Euro. Trans. 23. Levi, E. (1995): Impact of iron loss on behavior of
Elec. Power, Vol (16): pp.15-334. vector controlled induction machines, IEEE Trans. On
11. Ganapathi.D.Moger, Ch.S.N.Murthy and Manjunath. Ind. App, Vol(31): No.6, pp.1287-1296.
A, (2016): Electrical energy conservation in 24. L Kirtley. Jr., January 26 (2003): Analytic Design
underground mines. ICAER 2011 Evaluation of Induction Machines.
12. Geng, Y., Hua, G., Huangang, W. and Pengyi, G., Busan, 25. International Conference on Industrial Engineering
Korea (2004): A novel Control Strategy of induction 26. Linkov S.A.a, Olizarenko V.V.a, Radionov A.A.b,
motors for optimization of Both efficiency and Torque Sarapulov O.A.a, (2015): Energy-efficient power
response, Proceedings of the 30Th annual Conference supply system for mines, Elsevier conference.
of IEEE Ind. Elect. Society, pp.1405-1410. 27. Marino, R., Tomei, P., Verrelli, C.M., (2008): An
13. Grouni, S., Ibtiouen, R., Kidouche and M., Touhami O., adaptive tracking control from current measurements
(2008): Improvement approach on rotor time constant for induction motors with uncertain load torque and
adaptation with optimum flux in IFOC for induction rotor resistance, Journal Automatica, Vol(44): No. 44,
motors drives, International Journal of No.11, pp.2593-2599.
Mathematical, Physical an Engineering Science, 28. M. Shepard, S. Greenberg, G. Katz, A.T. de Almeida.
Vol(2): No.2, No.3, pp.121-125. (1992): Energy-Efficient Motor Systems, by American
14. Gumu³a S., Pytel K., Piaskowska-Silarska M. (2014): council for an Energy Efficient Economy, 1001
Polemical Remarks to the Claim that Carbon Dioxide Connecticut Avenue, N.W., suite 801, Washington
Strengthens the Greenhouse Effect in the Atmosphere. D.C.20036, ISBN O-918249-10-4.

JOURNAL OF MINES, METALS & FUELS 235

 29. Novotny, D.W., Lipo, T.A. (1996): Vector Control and Workshop on Motion Control for Intelligent
Dynamics of AC Drives, Oxford, Clarendon Press. Automation, Vol(1) : pp. 127-133.
30. Paul Blaiklock and William Horvath, (2017): How to 38. S. I. Seleme Jr. and E. Mendes and C. Canudas de Wit
select the best motors and motor controls for mining and A. Razek. (1993): Experimental Validation of the
applications. Journal of Equipment and technology, Minimum Energy Approach for Induction Motor
Automation and energy. Control, Proc. of the IEEE/SMC’93 Conference, Vol(5) :
31. Park, M.H. and Kim, K.S. (1991): Chattering reduction pp. 78-83.
in the position control of induction motor using the 39. S. I. Seleme Jr. and M. Petersson and C. Canudas de
sliding mode, IEEE Trans., Vol(6) : pp. 317-325. Wit. (1994): The Torque Tracking of Induction Motors
32. P.A. Frick and D.J. Stech. (1995): Solution of the Via Magnetic Energy Optimization, Proceedings of the
optimal control problems on parallel machine using IEEE Conference on Decision and Control.
epsilon method, Optimal Control Appl. Methods, 40. Siva Reddy, Y.V., Vijayakumar, M., Brahmananda Reddy,
Vol(16): pp. 1-17. T. (2000): Direct torque control of induction motor
33. Pytel K., Jaracz K., Gumula S., (2012): An impact of using sophisticated lookup tables based on neural
chosen construction parameter and operating networks, AIML Journal, vol.7, Issue 1, pp.9-15.
conditions on the quality of wind turbine energy
41. Ta, C.M., Hori, Y. (2001): Convergence improvement of
generation. ICCC 2012, High Tatras, Slovakia, 28-31
efficiency optimization control of induction motor
May, 592-595.
drives, IEE Trans. Ind. App, Vol(37): No.6,pp.1746-1753.
34. Ren, J.Q., Li, Y.H., Xu, W., Wang, K., (2008): Loss
42. Thanga Raj, C., Srivastava, S.P., Agarwal, P. (2009):
model establishment and efficiency optimization
Energy efficient control of three phase induction
control of single sided linear induction motor, IEEE
motor, a review, International Journal of Computer
Trans. Ind. Elec, Vol(43): No.1, pp.226-231.
and Electrical Engineering (IJCEE), Vol (1): No.1,
35. S. Grouni, R. Ibtioun, M. Kidouche, O. Touhami. pp.1793-1798.
(2010): Novel Loss Optimization in Induction
Machines with Optimum Rotor Flux Control, 43. Vas, P. (1990): Vector control of AC Machines, Oxford,
International Journal of Systems Control, Vol(1- Iss. Clarendon Press, USA.
4), pp 163-169. 44. Wang, B., Chiasson and Bodson, M. and Tolbert,
36. Singh, G.K., (2005): A research Survey of Induction L.M., (2007): An online rotor time constant estimator
Motor Operation with Non-sinusoidal Supply Wave for the induction machine, IEEE. Trans, Contr. Syst.
Forms, Electric Power System Research, Science Direct Techno. , Vol (15): pp. 339-347, No.2.
Elsevier, Vol(75): pp.200-213.
No part of the article in any format can be uploaded to any medium
37. S. I. Seleme and C. Canudas de Wit. , Perugia, Italy, other than that of Books and Journals Private Limited, without the
27-29 October (1992): Minimum Energy Operation executive permission. Such actions will be considered breach of faith,
Conditions of Induction Motors Under Torque control, for which appropriate actions will be taken.

JOURNAL OF MINES, METALS & FUELS

SPECIAL ISSUE ON
THICK SEAM MINING IN CHINA – AN UPDATE
Price per copy
Rs.500.00; £50.00 or $80.00
For copies please contact:
The Manager
Books & Journals Private Limited
Moon Plaza (2A, 2nd Floor), 62 Lenin Sarani, Taltala, Kolkata 700013
e-mail: [email protected]; [email protected]
www.jmmf.info • Mob: +919239384829

236 JULY 2020