Design of a 7-Level Single-Stage/Phase PUC Grid-

Connected PV Inverter with FS-MPC Control

Rita Khawaja1,2, Fadia Sebaaly1, Hadi Y. Kanaan1
1
Saint-Joseph University of Beirut, Faculty of Engineering – ESIB, Mar Roukoz, Mkalles, Lebanon
2
Lebanese University, Faculty of Engineering – Branch III, Hadath Campus, Hadath, Lebanon

Abstract— Multilevel, transformer-less inverters are Based on the number of power processing stages, PV
becoming more and more a subject of interest in grid connected inverters are classified in two different categories: Multi-
photovoltaic system due to their low cost and high efficiency Stage Inverters (MSIs) and Single-Stage Inverters (SSIs)[9].
features. However, most of the designed grid tied systems show A MSI employs more than one power processing stage that
more than one stage conversion process to inject the amount of achieve the task of boosting the DC output from PV array
power generated to the grid. Mainly, a DC-DC conversion stage
is always requested between the Photovoltaic (PV) panel and the
and/or galvanic isolation. In general, most of the PV inverters
inverter. In this paper, a single-stage grid connected design employ two-stage power conversion process. During the
based on seven level packed U-cell inverter controlled by a initial stage, PV array output is increased to a higher level by
finite-set model predictive algorithm with a maximum power a DC-DC boost converter while tracking maximum solar
point tracking capability is presented as a competitive single- power. In the second stage, this DC signal is converted into
stage/phase grid-connected topology. The overall design is less AC one of high quality. Even though these two stage inverters
complex compared to existing topologies where a maximum have been in existence for a while and work well, however
power point operation of the PV system is well achieved without due to the high number of part count, they show a drawback
the use of a boost converter. A predictive algorithm takes in of lower efficiency along with higher costs, bigger size and
action of the inverter control. Simulation results in
low reliability. Therefore, SSIs which boost the PV output,
MATLAB/SIMULINK are presented to validate the proposed
design in both constant and variable irradiation conditions. employ MPPT and invert the boosted DC signal into AC
using only a single power electronic circuit give the
Keywords—PV system, Grid connected, Packed U-Cell Inverter, advantage of reduced and robust circuitry along with
Model Predictive Control, Single Phase/Stage Inverter, MPPT, reliability and efficiency [10].
Perturb and Observe. The control circuits are a key element in the operation
and performance of grid tied systems. Control techniques can
I. INTRODUCTION be classified into two main groups: linear as Proportional-
Integral (PI) regulators and nonlinear techniques. Due to the
Solar systems can be generalized as grid-direct systems. improvement of microprocessor-computing power ,
The major component in all grid-connected energy nonlinear current control techniques as : hysteresis control,
conversion systems is the DC-AC inverter which takes in deadbeat control, sliding mode [3, 11] and Finite Set Model
charge of flow energy control. The inverter requirements Predictive Control (FS-MPC)[12, 13] have been paid more
include operation over a wide range of voltages and currents, attention and started recently to emerge fast in grid tied
regulated output voltage and frequency while providing AC systems. Among them, FS-MPC has taken a major
power and good power quality that includes low Total importance and development during the last decades since
Harmonic Distortion (THD) and Unity Power Factor (UPF) it’s an easy, simple and flexible technique, it presents an easy
operation. Besides, to get the highest possible efficiency for inclusion of system nonlinearities and uncertainties [14, 15].
all solar irradiance levels [1] in all weather conditions, a This technique showed a good dynamic performance, strong
Maximum Power Point Tracking (MPPT) algorithm is current tracking ability, and good adaptability to different
adopted. variables when adopted to control grid tied multilevel
Nowadays, transformer-less Multilevel Inverters (MLIs) inverters [16].
gained so much popularity in PV systems [2-4]. MLIs In [17], FS-MPC was adopted to control the 7L-PUC
promise many advantages over conventional inverters as the inverter operating as a PV grid tied one, however two-stage
improvement in the output waveform since they produce a design was proposed where a boost (DC-DC) converter was
staircase, nearly sinusoidal output voltage. Hence, lower introduced. In this paper, a new design of single-stage PV
THD with a reduction in the switching losses are achieved 7L-PUC grid tied inverter is adopted. The proposed design is
and less filters are needed. In the end, the system is cheaper, based on 7L-PUC inverter controlled by a FS-MPC technique
much compact and lighter[5]. To choose the best inverter, a while an MPPT technique takes in action of maintaining the
comparison based on the number of semiconductor devices, PV panel operating at its maximum power. The overall
DC bus capacitors, THD of the output voltage, reduction of design of single-stage system shows less complexity while
leakage current and system efficiency is adopted [6, 7]. For no-need for boost converter. Simulation results are presented
seven-level inverters, a comparison between the most to prove the good performance of the proposed design in both
adopted topologies such as 7-level Flying Capacitor (FC), 7- constant and variable irradiation conditions.
level Neutral Point Clamped (NPC), 7-level Cascaded H- This paper is detailed as follow: Section II elaborates the
Bridge (CHB) and 7-level Pack U-Cell (7L-PUC) reveals that 7-level PUC inverter and MPC control. Section III presents
the 7L-PUC is the most promising one for single-phase the MPPT algorithm. Section IV resumes the overall design,
applications since it shows less components [8]. Section V introduces simulation results. Finally, section V
concludes the work done.

978-1-7281-5754-2/20/$31.00 ©2020 IEEE 751

II. GRID TIED PV 7L-PUC INVERTER: STRUCTURE AND value of the cost function. MPC is based on the system model
CONTROL equations and measures. It consists of controlling three
variables: DC side voltage , grid line current and DC
Grid-tied, on-grid, utility-interactive, grid intertie and capacitor voltage .
grid back feeding are all terms used to describe the solar
system that is connected to the utility power grid . It allows
saving more money with solar panels through better
efficiency rates, lower equipment and installation costs. Grid-
tied solar systems are therefore generally cheaper and simpler
to install. These latter rely on solar inverters that regulate the
voltage and current received from the panels and connect
these panels to the grid. As a derivation of CHB in which two
switches have been removed and two cells are connected
together directly, PUC converter was introduced in early
2008 [18]. PUC offers high power quality using a small
number of passive and active components in comparison with
other topologies of 7-level inverters[8] .
Two power switches with one capacitor in between to
configure the U-shape cell construct PUC inverter. Repeating
this cell increases the number of output voltage level
exponentially. The major advantage of this topology is that
only one DC bus is needed which simplifies the control of
maximum power point tracking and eliminates the need for Fig.1: Pack U-Cell converter
sourcing load balance [18]. It is made of six active switches
The system under study is 7L-PUC inverter connected to
and two DC links as it is depicted in Fig.1. Each pair of
the utility as depicted in Fig.1. The grid is represented as an
switches and , and , and is working AC voltage source and its connection with the inverter is
complementarily. It is clear that eight existing switching
modelled by a resistor and an inductor representing the line
states can produce different paths for current to flow through.
filter. The grid current and the capacitor voltage are measured
Table I illustrates the details of the switching states. Taking for each sampling time k. Based on these values and using
into account that the output voltage levels numbers depend Euler forward approximation v (k + 1) and i (k + 1) are
on DC sources voltages, using unequal DC sources result in
predicted using the following equations (Where is the
having different level numbers in output voltage waveform.
sampling time):
Assuming = 3 = 3E, seven voltage levels would be
generated as 0, ±E, ±2E, ±3E. The output voltage is closer to
+1 = 1− ⁄ + ⁄ − (1)
a sine wave, which will decrease the error compared to the
grid sine voltage. +1 = − − ⁄ (2)
TABLE I
SWITCHING STATES OF PUC INVERTER
The cost function at the same instant k is calculated using
States S1 S2 S3 Output Voltage
1 1 0 0 V1
the values of, and at the instant k + 1.Therefore, g
2 1 0 1 V1-V2 equation will be as follows:
3 1 1 0 V2 g(k) = +1 − ∗ ²+ +1 − ∗ ²
4 1 1 1 0 + +1 − ∗ ² (3)
5 0 0 0 0
6 0 0 1 0 k1, k2 and k3 represent the weighting factors; they are
7 0 1 0 V2-V1 coupling effects in case of using several variables in the cost
8 0 1 1 -V1 function. Indeed, there are no analytical or numerical
methods as such to adjust the weighting factors [19].
In this study for an inverter application, the DC link with The flowchart of the MPC for PUC inverter depicted in
higher voltage will be connected to a PV panel where the Fig.2. includes two loops. The outer loop, consisting of
DC link with lower voltage will be connected to a capacitor, measuring the grid current and the capacitor voltage, is
which its voltage will be regulated. Besides, the output executed every sampling time, while the inner loop is
current should be controlled to have low harmonic content executed for each possible state. The inner loop consists of
with unity power factor operation. Several control techniques calculating the predicted values of the grid current and the
have been applied on PUC topology such as hysteresis capacitor voltage, and then calculating the cost function g(x)
current control, sliding mode control and model predictive to store the optimal values. After executing the possible
control. FS-MPC is one of the most promising control states, the minimum value of g(x) is chosen and the
technique since it is very simple and intuitive [4]. It has a fast corresponding switching states are applied to the
dynamic response and accurate tracking ability. MPC semiconductor devices. Note that the variables with stars
measures the variable x ( ) then it calculates the future value denote the reference values.
x ( + 1) using the predictive control each possible switching Moreover, the reference current is generated based on
state. After, the cost function is calculated using a reference the grid-voltage angle generated by a Phase-Locked Loop
value x* and the calculated predicted value x ( + 1). At the (PLL) block.
end, the optimal switching state is chosen for the minimal

752
 Startup

Measure
i1(k),vl(k)andV 2(k)

x=0

x=x+1

Vad=(S1-S2 )V1+(S2-S3)V 2
il(k+1)=(1-RT s/L)i1 (k)+(Ts/L)(Vad(k)-vl(k))
V2(k+1)=V2(k)-(Ts(S2-S3)/C2)il(k)

g(i)=k1((il(k+1)-il*)2+
Wait for next
k2((V2(k+1) -V2* )2
sampling instant
+k3(V 1(k+1)-V1*)2

No
X=N Fig.3: P&O flow chart [20]
Yes
Minimum {g} and apply
IV. OVERALL CONTROL DESCRIPTION
optimum switching state
The overall system model detailing FS-MPC applied on
one-stage 7L-PUC inverter is depicted in Fig.4. The
Fig.2: Flowchart of FS-MPC algorithm 7L-PUC inverter[19] photovoltaic block was formed by twenty-nine strings
connected in parallel each one composed of five modules in
III. MPPT FOR SINGLE STAGE PUC INVERTER WITH MPC series. The module chosen had an open circuit voltage of
36.06V. Besides, PUC inverter model consists of six
Various MPPT algorithms have been developed to MOSFETs, a capacitor and one DC input where the PV panel
improve the PV array efficiency, achieve faster and accurate is connected. Three switches are controlled and their
tracking performance and to reduce the oscillations around complementary switching state are connected to the gates of
MPP. These algorithms are compared based on the chosen the other three MOSFETs. The circuit outputs are: the
convergence speed, the number of sensors used, the inverter output voltage, grid voltage, grid current and
complexity and the implementation cost. A comparison for capacitor voltage. These outputs are measured and entered
the different types of MPPT controllers can be found in [20, as inputs in the control loop to predict their future values,
21]. MPPT methods can be classified in two main categories: compare them to the reference values and calculate the
on-line and off-line techniques. On-line methods consist of minimal cost function that will precise the optimal switching
regulating the PV array voltage at the MPP voltage and are state that should be applied to the MOSFETs. Moreover, to
the most used methods such as Perturbation and Observation connect the PV to the inverter a capacitor was added in
(P&O) and the Incremental Conductance (IncCond) parallel to insure the power decoupling. Note that the control
algorithms, while Off-line methods are model-based system and the P&O algorithm are developed as
techniques such as Constant Voltage and Short-Circuit MATLAB/SIMULINK function. P&O function takes the
Current algorithms[22]. instantaneous voltage and current at the entrance of the
Perturb and Observe (P&O) algorithm is considered the inverter as inputs, calculates the power and compare them to
most popular MPPT technique. It shows less complexity in the previous value then respectively increment or decrement
comparison with other methods. The flowchart of this the voltage by a deltaV to get the corresponding output
technique is depicted in Fig.3. Its operation depends on the voltage. At this stage, the output voltage of the MPPT is taken
observation of the voltage, current and the power in different as a reference value. The voltage of the PV was compared to
two points. Referring to the maximum power point by its it. The maximum amplitude of the injected power was
power as #$%% ,its current as &$%% and its voltage by $%% ,if calculated considering that the efficiency of the inverter is
the operation zone is at the right of this MPP, it can be seen 100%. Accordingly, the power at the DC side is calculated by
that incrementing the voltage increases the power while multiplying the voltage of the PV by the root mean square of
decreases the power when moving at the right side of the the current and then it is divided by the root mean square
MPP. In the other side, decrementing the voltage at the left value of the voltage at the output of inverter multiplied by the
side of the MPP increases the power and decreases the power cosine of the phase angle.
when operating at the right side of it. Hence, if there is an
increase in power, the subsequent perturbation should be kept V. SIMULATION RESULTS
in the same direction to reach the MPP and if there is a
decrease in the power, the perturbation should go reverse In order to verify the robustness and the performance of the
direction [23]. proposed design, the overall PV single-phase/stage grid tied
7L-PUC inverter with FS-MPC control was designed and
simulated on MATLAB/SIMULINK.

753
Fig.4. Schematic model of MPPT for PUC inverter with MPC
algorithm is integrated is higher than the case of operation
The simulated system parameters adopted are illustrated in without MPPT by around 300W as it shown in the simulation
Table II. It should be noted that the sampling time is set to results of Fig.6 and Fig.7 respectively.
= 20)*.
TABLE II B. Operation with Variable Irradiation
SIMULATED SYSTEM PARAMETERS When the irradiation varies between 300 and 800 kW/m², the
PV Panel Output 180 V power extracted from the PV panel controlled by a MPPT
Grid frequency 60 Hz algorithm is higher than the power without it as it is shown in
DC Capacitor 1000 μF Fig.8 and Fig.9 respectively.
Line Inductor 2.5mH The previous discussed results validate the good
Parasitic Resistor r 0.1 Ω performance of the proposed design of a single-stage grid PV
AC Grid Voltage 120VRMS
inverter operated by a MPPT technique
Sampling Time 20 µs
VI. CONCLUSION

Fig. 5 to 9 show the simulation results obtained. In the In this paper, a new design of 7-Level Single-
steady state operation mode illustrated in Fig.5. it’s well Stage/Phase grid-connected PV Inverter based on PUC
shown that seven-level output voltage is perfectly generated topology with FS-MPC control is introduced. The innovation
at PUC inverter output. The grid voltage and current are well of the new design is that the P&O MPPT algorithm controls
synchronized and a unity power factor is achieved (the phase directly the power given to the 7-Level PUC inverter without
angle is set to zero). The injected grid current follows the the integration of a DC-to-DC boost converter. The main
reference current, which its maximum amplitude was set to advantage of the proposed system is less complexity since no
10A. Moreover, a low content of harmonic distortion need for additional conversion stage. The voltage extracted
approximately equal to 2.7% is achieved confirming that the from the MPPT algorithm is adopted as a reference value to
injected current complies well with the grid connectivity the control system. The system was modeled and simulated
requirements and standards (less than 5%). In addition, DC on MATLAB/SIMULINK. Simulation result were
capacitor voltage has reached one third of the DC voltage introduced to prove the proposed system and validate its good
across the PV panel and it is well verified that low ripples are performance in various conditions as steady-state, operation
obtained. The previous discussed results confirm the good with and without MPPT algorithm in both constant and
operation of the FS-MPC controller in steady-state mode. variable irradiations cases.
The following section presents the simulation results
obtained when a MPPT algorithm that controls the PV panel ACKNOWLEDGEMENTS
to operate at its maximum power point is integrated in the
generation of the current reference amplitude. Both cases of The authors gratefully thank the Agence Universitaire de
constant and variable irradiance operation are studied and la Francophonie (AUF), the Lebanese National Council for
compared to the case without integration of a MPPT Scientific Research (CNRS-L), and the Research Council of
algorithm Saint-Joseph University for their financial support.
A. Operation with Constant Irradiation

For a constant irradiation of 500kW/m², it is well verified

that the power extracted from the system when a MPPT

754
Fig.5: Steady state operation

Fig.6: PV and grid powers without MPPT implementation at constant irradiation

Fig.7: PV and grid powers with MPPT implementation at constant irradiation

Fig.8: PV and grid powers without MPPT implementation under variable irradiation

755
 Fig.9: PV and grid powers with MPPT implementation under variable irradiation

REFERENCES [17] O. A.-R. a. H. Funato, "Gird-Connected Boost Inverter for low power
PV applications with Model Predictive Control," The Journal of
Engineering 2017.
[1] J. V. Nimrod Vázquez, "Photovoltaic System Conversion," Power
[18] Hani Vahedi, Hadi Kanaan and Kamal Al-Haddad, "PUC Converter
Electronics Handbook (Fourth Edition),, 2018.
Review: Topology, Control and Applications " presented at the
[2] F. Sebaaly, H. Vahedi, H. Y. Kanaan, and K. Al-Haddad, "Novel
IECON 2015 - 41st Annual Conference of the IEEE Industrial
current controller based on MPC with fixed switching frequency
Electronics Society, 2015.
operation for a grid-tied inverter," IEEE Transactions on Industrial
[19] Julie Metrie. Hani Vahedi, Hadi Y. Kanaan and Kamal Al-Haddad,
Electronics, vol. 65, pp. 6198-6205, 2018.
"Real-Time Implementation of Model Predictive Control on Seven-
[3] F. Sebaaly, H. Vahedi, H. Y. Kanaan, N. Moubayed, and K. Al-
Level Packed U-Cell Inverter," IEEE Trans. on Industrial Electronics,
Haddad, "Design and implementation of space vector modulation-
vol. 63, 2016.
based sliding mode control for grid-connected 3L-NPC inverter," IEEE
[20] S. E. M. Kathiresan .R, S. Mohanvel and Dr. T.K. Santhosh4,
Transactions on Industrial Electronics, vol. 63, pp. 7854-7863, 2016.
"Parameter Comparison of Online and Offline MPPT Algorithm
[4] F. Sebaaly, H. Vahedi, H. Y. Kanaan, and K. Al-Haddad,
Review," International Research Journal of Engineering and
"Experimental Design of Fixed Switching Frequency Model Predictive
Technology (IRJET), vol. VOL. 5, 2018.
Control for Sensorless Five-Level Packed U-Cell Inverter," IEEE
[21] A. B. a. M. S. Arjav Harjai, "STUDY OF MAXIMUM POWER
Transactions on Industrial Electronics, vol. 66, pp. 3427-3434, 2019.
POINT TRACKING (MPPT) TECHNIQUES IN A SOLAR
[5] N. M. S. Nurul Aisyah Yusof, Hazlie Mokhlis, Jeyraj Selvaraj, "A
PHOTOVOLTAIC ARRAY," Bachelor of Technology, Department of
Comparative Study of 5-level and 7-level Multilevel Inverter
Electrical Engineering, National Institute of Technology, Rourkela-
Connected to the Grid " presented at the 2012 IEEE International
769008, Orissa.
Conference on Power and Energy (PECon), Kota Kinabalu Sabah,
[22] J. Metri, H. Vahedi, H. Kanaan and K. Al-Haddad, “Solar Energy
Malaysia, 2012.
Processor based on Packed U-Cells 7-Level Inverter for Grid
[6] V. C. Sanoop P, "Seven Level Inverter Topologies: A Comparative
Applications”, in Proc. 42nd Annual Conference of the IEEE Industrial
Study," International Journal of Innovative Research in Electrical,
Electronics Society (IECON’16), Florence, Italy, October 23-27, 2016,
Electronics, Instrumentation and Control Engineering, vol. Vol. 3,
pp. 6528-6533.
2016.
[23] A. F. M. Hadeed Ahmed Sher, Abdullah Noman, Khaled E.
[7] M. I. a. S. Mekhilef, "H6-type transformerless single-phase inverter for
Addoweesh and Marcello Chiaberge, "A New Sensorless Hybrid
grid-tied photovoltaic system," IET Power Electron,The Institution of
MPPT Algorithm Based on Fractional Short-Circuit Current
Engineering and Technology vol. VOL. 8, 2015.
Measurement and P&O MPPT," IEEE TRANSACTIONS ON
[8] B. N. a. H. V. Ali Asghar Shojaei, "Standalone Operation of Modified
SUSTAINABLE ENERGY, vol. VOL. 6,, 2015.
Seven-Level Packed U-Cell (MPUC) Single-Phase Inverter,"
[24] H. Vahedi, K. Al-Haddad and H. Y. Kanaan, “A New Voltage
Electronics vol. Vol. 8, 2019.
Balancing Controller Applied on 7-Level PUC Inverter”, in Proc. 40th
[9] D. A. S. K. K. Kirubasankar, "Inverter Power Stage Connected with
Annual Conference of the IEEE Industrial Electronics Society
PV-Grid," Scientific Research Publishing, vol. VOL, 7, pp. 4113-4123,
(IECON’14), Dallas, TX, USA, October 28 - November 1, 2014, pp.
2016
5082-5087.
[10] S. K. S. Ankita, Sukruedee Sukchaib, Franco Fernando Yanine,
[25] M. Sleiman, K. Al-Haddad, H. F. Blanchette, L.-A. Grégoire and H.
"Review and comparative study of single-stage inverters for a PV
Kanaan, “A New 7L-PUC Multilevel Cells Modular Converter for AC-
system," Renewable and Sustainable Energy Reviews, 2018.
AC and AC-DC Applications”, in Proc. IEEE International
[11] F. Sebaaly, H. Vahedi, H. Y. Kanaan, N. Moubayed, and K. Al-
Conference on Industrial Technology (ICIT’15), Sevilla, Spain, March
Haddad, "Sliding mode fixed frequency current controller design for
17-19, 2015, pp. 2514-2519.
grid-connected NPC inverter," IEEE Journal of Emerging and Selected
[26] F. Sebaaly, H. Vahedi, H. Y. Kanaan and K. Al-Haddad, “Fixed
Topics in Power Electronics, vol. 4, pp. 1397-1405, 2016.
Switching Frequency Model Predictive Based Controller for Sensor-
[12] M. K. K. a. S. K. B. Sourabh Rathore, "Simulation of Cascaded H-
less Five-level Packed U-Cell (PUC5) Single Phase Inverter”, in Proc.
Bridge Multilevel Inverter Using PD,POD,APOD TECHNIQUES,"
19th IEEE International Conference on Industrial Technology
Electrical & Computer Engineering: An International Journal (ECIJ),
(ICIT’18), Lyon, France, February 20-22, 2018, pp. 1915-1919.
vol. Vol. 4, 2015.
[27] J. Metri, H. Vahedi, H. Kanaan and K. Al-Haddad, “Model Predictive
[13] R. Nazir, "Advanced Repetitive Control of GridConverters for Power
Control for the Packed U-Cells 7-Level Grid Connected Inverter”, in
Quality ImprovementUnder Variable Frequency Conditions," Doctor
Proc. IEEE International Conference on Industrial Technology
of Philosophy, Electrical and Computer Engineering, University of
(ICIT’16), Taipei, Taiwan, March 14-17, 2016, pp. 1214-1219.
Canterbury,Christchurch, New Zealand, 2015.
[28] M. Babaie, F. Sebaaly, M. Sharifzadeh, H. Y. Kanaan and K. Al-
[14] C. B. a. C. Montero, "Basic Principles of MPC for Power Converters:
Haddad, “Design an Artificial Neural Network Control Based on
Bridging the Gap Between Theory and Practice," IEEE Industrial
Levenberg-Marquart Algorithm for Grid-Connected Packed U-Cell
Electronics Magazine, 2015.
Inverter”, in Proc. 20th IEEE International Conference on Industrial
[15] F. Sebaaly, H. Vahedi, H. Y. Kanaan, N. Moubayed, and K. Al-
Technology (ICIT’19), Melbourne, Australia, February 13-15, 2019.
Haddad, "Model predictive controller with fixed switching frequency
[29] F. Sebaaly, H. Vahedi, H. Y. Kanaan and K. Al-Haddad, “Design of
for a 3L-NPC inverter," in Industrial Electronics Society, IECON
Pulse Width Modulator based Sliding-Mode Control (SMC-PWM) for
2016-42nd Annual Conference of the IEEE, 2016, pp. 6536-6511.
Sensor-less Single-Phase Packed U-Cell Inverter”, in Proc. 45th Annual
[16] C. L. Jingang Han, Tengfei Yang, and Jun Han, "Simplified Finite Set
Conference of the IEEE Industrial Electronics Society (IECON’19),
Model Predictive Control Strategy of Grid-Connected Cascade H-
Lisbon, Portugal, October 14-17, 2019.
Bridge Converter," Journal of Control Science and Engineering-
Hindawi Publishing Corporation, 2016.

756