Table of Contents

LIST OF FIGURES
ABSTRACT
Chapter-1 INTRODUCTION 6

Chapter-2 SOLAR CELLS

2.1 History of fuel cell 9

2.2 Structure 10

2.3 Operation 11

2.4 Output characteristics of solar cells 13

2.5. Applications 13

2.6 Solar Power advantages and disadvantages

2.6.1 Advantage 14

2.6.2 Disadvantage 14

Chapter-3 WIND ENERGY

3.1 Introduction 15

3.2 Need of wind mills in present generation 15

3.3 Concept of Distribution Generation 16

3.4 Micro grid 16

3.5 Models of power generation 17

1
Chapter-4 MODEL PREDICTIVE CONTROL

4.1 Introduction 20

4.2 Overview of model predictive control 21

Chapter-5 WIND-SOLAR SYSTEM DESCRIPTION 24

Chapter-6 CONTROL PROBLEM FORMULATION AND

CONTROLLERS DESIGN

6.1 Control problem formulation 29

6.2 Wind subsystem controller design 30

6.3 Solar subsystem controller design 31

6.4 Supervisory controller design 32

Chapter-7 MATLAB

7.1 Introduction to matlab 35

7.2 Components of matlab 35

7.3 Toolboxes in matlab 36

7.4 Simulink

7.4.1 Introduction 36

7.4.2 Concept of signal and logic flow 36

7.4.3 Connecting blocks 38

7.5. Sources and sinks 38

7.5.1 Continuous and Discrete systems 40
7.5.2 Non linear operators 40
7.6 Mathematical operations 41
7.7 Signals and data transfer 42

2
7.8 Making subsystems 43
7.9 Setting simulation parameters 44
7.10 Basic circuit designing and analyzing of results 45

Chapter-8 SIMULATION RESULTS

8.1. Constraints on the Maximum Increasing Rates of Pwref and Psref 46
8.2. Suppression of Battery Power Fluctuation 48
8.3. Varying Environmental Conditions 49
8.4. Consideration of High-Frequency Disturbance of Weather Condition 50

CONCLUSION
REFERENCES

3
 List of figures
Fig 1.1 Wind-Solar energy generation system 7
Fig 2.1 Structure of photovoltaic cell 10
Fig 2.2 Output characteristics of solar cells 13

Fig 3.1 Synchronous Generator circuit 18

Fig 4.1 Block diagram for Model Predictive Control 22
Fig 5.1 Supervisory control of a wind-solar hybrid energy system 24
Fig 7.1 Block diagram of Matlab components 35
Fig 7.2 Simulink library browser 35
Fig 7.3 Connecting blocks 36
Fig 7.4 Sources and sinks 36
Fig 7.5 Continous and Descrete systems 37
Fig 7.6 Simulink blocks 41
Fig 7.7 Simulink math blocks 42
Fig 7.8 Signals and systems 43
Fig 7.9 Simulation Parameters 44
Fig 7.10 Block diagram for analyzing of results 45
Fig.8.1. Waveforms showing Load Power Requirement,Power 48
Supplied From Wind and Solar Generation System, Power
Deficit Supplied by Battery
Fig.8.2.Waveforms showing gradual decrease in power supplied by 49

battery bank

Fig.8.3.Waveforms showing varying Environmental conditions and 50

Load current

4
 ABSTRACT
Energy technologies, like wind and solar based energy generation systems,
are receiving national and worldwide attention owing to the rising rate of
consumption of nuclear and fossil fuels. In particular, drivers for solar/wind
renewable energy systems are the environmental benefits, reduced investment risk,
fuel diversification, and energy autonomy, increased energy efficiency as well as
potential increase of power quality and reliability.
This work has been focused on the development of a supervisory model
predictive control method for the optimal management and operation of hybrid
standalone wind-solar energy generation systems.
The supervisory control system via model predictive control which computes
the power references for the wind and solar subsystems has been designed at each
sampling time while minimizing a suitable cost function. The power references are
sent to two local controllers which drive the two subsystems to the requested
power references. Simulations are carried out using MATLAB software.

5
 1. INTRODUCTION
Alternative energy technologies, like wind- and solar-based energy generation
systems, are receiving national and worldwide attention owing to the rising rate of
consumption of nuclear and fossil fuels. In particular, drivers for solar/wind
renewable energy systems are the environmental benefits (reduction of carbon
emissions due to the use of renewable energy sources and the efficient use of fossil
fuels), reduced investment risk, fuel diversification, and energy autonomy,
increased energy efficiency (less line losses) as well as potential increase of power
quality and reliability and in certain cases, potential grid expansion deferral due to
the possibility of generation close to demand. In a recent report of the California
Energy Commission, for example, the state’s target is to generate from renewable
sources the 33% of the energy needed by year 2020, with about 70% of that energy
being produced by wind and solar systems, many other states have similar goals.
However, achieving such major renewable energy production goals requires
addressing key fundamental challenges in the operation and reliability of
intermittent (variable output) renewable resources like solar- and wind-based
energy generation systems. Specifically, unexpected drops in energy production of
a solar or wind energy system may require quick start units to cover the shortfall
while unexpected increases require the ability to absorb the unscheduled
generation. One way to deal with the variable output of wind and solar energy
generation systems is through the use of integrated energy generation systems
using both wind and photovoltaic energy, which are also tightly integrated with
distributed energy storage systems (batteries) and controllable energy loads like,
for example, a water production system that operates at controllable time intervals
to meet specific demand.
There are few works that have focused on the control of standalone hybrid
wind-solar energy generation systems. A reduced-order nonlinear model was used
to design a controller to regulate the wind power generation to complement the
power generated by a photovoltaic subsystem and to satisfy a specific power
demand. Sliding mode control techniques were used to control the power generated
by a photovoltaic array in order to satisfy the total instantaneous power demand in
a highly uncertain operating environment. A supervisory control system was
developed to satisfy the load power demand and to maintain the state of charge of
the battery bank to prevent blackout. In a recent work, a cost-effective control
technique was proposed for maximum power point tracking from the photovoltaic
6
array and wind turbine under varying climatic conditions without measuring the
irradiance of the photovoltaic or the wind speed. However, no attention has been
given to the development of supervisory control systems for standalone hybrid
wind-solar energy generation systems that take into account optimal allocation of
generation assignment between the two subsystems.
The objective of the present work is to develop a supervisory predictive
control method for the optimal management and operation of hybrid wind-solar
energy systems. The supervisory control system via model predictive control
(MPC) which computes the power references for the wind and solar subsystems at
each sampling time is to be designed while minimizing a suitable cost function.
The power references are sent to two local controllers which drive the wind and
solar subsystems to the desired power reference values. MPC is a popular control
strategy

Fig.1.1. Wind-Solar energy generation system

because of its ability to account for state and input constraints as well as optimality
considerations explicitly in the evaluation of control actions. MPC uses a model of
the system to predict at each sampling time the future evolution of the system from
the current state along a given prediction horizon. Using these predictions, the
input/set-point trajectory that minimizes a given performance index over a finite-
time horizon is computed solving a suitable optimization problem subject to
constraints. In this work, we discuss how we can incorporate practical
considerations into the formulation of the MPC optimization problem by
determining an appropriate cost function and constraints. We present several
simulation case studies that demonstrate the applicability and effectiveness of the
proposed supervisory predictive control architecture.

7
 2.LITERATURE REVIEW
R. Spee and J. H. Enslin developed adaptive maximum power point tracking
(MPPT) strategy to implement an efficiency maximization loop in parallel with the
regular maximum tip speed ratio tracker, without the measurement of mechanical
quantities. The overall power output of the generation system is increased with a
minimal increase in controller cost.
P. Novak, T. Ekelund, Y. Jovik, and B. Schmidtbauer proposed that it is possible
to use variable speed on a fixed-pitch wind turbine without any major difficulties.
It is possible to control the output power and to operate the rotor at high as well as
at low tip-speed ratios, 
Yeong-Chau Kuo, Tsorng-Juu Liang proposed single-stage system has advanced
features such as small physical volume, low weight, and high efficiency. A novel
single-stage MPPT controller is used for rapid tracking of the PV array’s
maximum power point. The proposed algorithm reduces oscillation, resulting in
significantly improved tracking. The proposed PV energy conversion system
operates in two modes, providing solar generation when insolation is adequate and
active power line conditioning when insolation is inadequate.
D. Muñoz de la Peña and P. D. Christofides[2008] proposed a LMPC controller
for control of a broad class of nonlinear uncertain systems subject to data losses.
The main idea is that in order to provide guaranteed stability results in the presence
of data losses, the constraints that define the LMPC optimization problem have to
be modified to account for data losses. This fact also holds for other MPC
controllers. The proposed LMPC controller allows for an explicit characterization
of the stability region, guarantees practical stability in the absence of data losses,
and guarantees that the stability region is invariant if the maximum time without
measurements is shorter than a given constant that depends on the parameters of
the system and the controller used to design the lmpc controller

8
Robust sliding-mode control laws have been considered in this paper to fulfill the
different control objectives (i.e., power regulation or maximum energy conversion)
of the wind and solar subsystems. However, it is important to remark that the
validity of the supervisor decision framework is not merely restricted to those
particular control laws. On the contrary, the applicability of the decision
framework of the supervisor is general and admits the incorporation of any other
control law suitable to attain the aforementioned control objectives.This
development traces back to the work of F. Valenciaga and P. F. Puleston[2005]
M. Chinchilla, S. Arnaltes, and J. C. Burgos described the operation and control of
one of these variable-speed wind generators: the direct driven permanent magnet
synchronous generator (PMSG). This generator is connected to the power network
by means of a fully controlled frequency converter, which consists of a pulsewidth-
modulation (PWM) rectifier, an intermediate dc circuit, and a PWM inverter. The
generator is controlled to obtain maximum power from the incident wind with
maximum efficiency under different load conditions. Vector control of the grid-
side inverter allows power factor regulation of the windmill.

9
 3. SOLAR CELLS AND WIND ENERGY
Solar cells (as the name implies) are designed to convert (at least a portion of)
available light into electrical energy. They do this without the use of either
chemical reactions or moving parts.
3.1. HISTORY OF FUEL CELL:
The development of the solar cell stems from the work of the French
physicist Antoine-César Becquerel in 1839. Becquerel discovered
the photovoltaic effect while experimenting with a solid electrode in an electrolyte
solution; he observed that voltage developed when light fell upon the electrode.
About 50 years later, Charles Fritts constructed the first true solar cells using
junctions formed by coating the semiconductor selenium with an ultrathin, nearly
transparent layer of gold. Fritts's devices were very inefficient, transforming less
than 1 percent of the absorbed light into electrical energy.
By 1927 another metal semiconductor-junction solar cell, in this case made
of copper and the semiconductor copper oxide, had been demonstrated. By the
1930s both the selenium cell and the copper oxide cell were being employed in
light-sensitive devices, such as photometers, for use in photography. These early
solar cells, however, still had energy-conversion efficiencies of less than 1 percent.
This impasse was finally overcome with the development of the silicon solar cell
by Russell Ohl in 1941. In 1954, three other American researchers, G.L. Pearson,
Daryl Chapin, and Calvin Fuller, demonstrated a silicon solar cell capable of a 6-
percent energy-conversion efficiency when used in direct sunlight. By the late
1980s silicon cells, as well as those made of gallium arsenide, with efficiencies of
more than 20 percent had been fabricated. In 1989 a concentrator solar cell, a type
of device in which sunlight is concentrated onto the cell surface by means of
lenses, achieved an efficiency of 37 percent due to the increased intensity of the
collected energy.

10
3.2. STRUCTURE:
Modern solar cells are based on semiconductor physics -- they are basically
just P-N junction photodiodes with a very large light-sensitive area. The photo
voltaic effect, which causes the cell to convert light directly into electrical energy,
occurs in the three energy-conversion layers.

Fig.2.1. Structure of photovoltaic cell

The first of these three layers necessary for energy conversion in a solar cell
is the top junction layer (made of N-type semiconductor). The next layer in the
structure is the core of the device; this is the absorber layer (the P-N junction). The
last of the energy-conversion layers is the back junction layer (made of P-type
semiconductor).
As may be seen in the above diagram, there are two additional layers that
must be present in a solar cell. These are the electrical contact layers. There must
obviously be two such layers to allow electric current to flow out of and into the

11
cell. The electrical contact layer on the face of the cell where light enters is
generally present in some grid pattern and is composed of a good conductor such
as a metal.
The grid pattern does not cover the entire face of the cell since grid
materials, though good electrical conductors, are generally not transparent to light.
Hence, the grid pattern must be widely spaced to allow light to enter the solar cell
but not to the extent that the electrical contact layer will have difficulty collecting
the current produced by the cell. The back electrical contact layer has no such
diametrically opposed restrictions. It need simply function as an electrical contact
and thus covers the entire back surface of the cell structure. Because the back layer
must be a very good electrical conductor, it is always made of metal.
3.3. OPERATION:
Solar cells are characterized by a maximum Open Circuit Voltage (Voc) at
zero output current and a Short Circuit Current (Isc) at zero output voltage. Since
power can be computed via this equation:
P=I*V
Then with one term at zero these conditions (V = Voc / I = 0, V = 0 / I = Isc)
also represent zero power. As you might then expect, a combination of less than
maximum current and voltage can be found that maximizes the power produced
(called, not surprisingly, the "maximum power point").
Many BEAM designs (and, in particular, solar engines) attempt to stay at (or
near) this point. The tricky part is building a design that can find the maximum
power point regardless of lighting conditions.

12
3.4. OUTPUT CHARACTERISTICS OF SOLAR CELLS:
The output characteristics of solar cells are expressed in the form of an I-V
curve. An I-V curve test circuit and typical I - V curve produced by the circuit are
shown below.
 The I-V curve is produced by varying RL (load resistance) from zero to
infinity and measuring the current and voltage along the way. The point at
which the I-V curve and resistance (RL) intersect is the operating point of
the solar cell. The current and voltage at this point are Ip and Vp,
respectively. The largest operating point in the square area is the maximum
output of the solar cell.

Fig.2.2. Output characteristics of solar cells

3.5. APPLICATIONS:
• Calculators
• Indoor clocks
• Remote control units
• Indoor digital thermometers
13
3.6. SOLAR POWER-ADVANTAGES AND DISADVANTAGES:
There are many advantages worth considering when it comes to solar energy
and everything that it offers. There are many advantages that solar energy has over
oil energy. Not only does solar energy benefit your pocketbook, but it also benefits
the environment as well. However, there are two sides to everything, and there is a
list of solar power disadvantages to accompany the list of advantages.
3.6.1. ADVANTAGE:
Solar energy is a completely renewable resource. This means that even when
we cannot make use of the sun’s power because of nighttime or cloudy and stormy
days, we can always rely on the sun showing up the very next day as a constant
and consistent power source.
3.6.2. DISADVANTAGE:
The Solar Cells and Solar Panels that are needed to harness solar energy tend
to be very expensive when you first purchase them.
3.7 WIND ENERGY
3.7.1 INTRODUCTION:
Wind energy is nothing but the velocity by which the air flows. It is of kinetic
energy; now days we are using wind mills to generate electrical power.
A windmill is a machine which converts the energy of wind into rotational energy
by means of vanes called sails or blades. Originally windmills were developed
for milling grain for food production. In the course of history the windmill was
adapted to many other industrial uses. An important application was to pump
water. Windmills used for generating electricity are commonly known as wind
turbines.
3.7.2. NEED OF WIND MILLS IN PRESENT GENERATION:
The development of wind power in India began in the 1990s, and has significantly
increased in the last few years. Although a relative newcomer to the wind industry
compared with Denmark or the US, India has the fifth largest installed wind power
14
capacity in the world. In 2009-10 India's growth rate was highest among the other
top four countries.
The worldwide installed capacity of wind power reached 197 GW by the end of
2010. China (44,733 MW), US (40,180 MW), Germany (27,215 MW) and Spain
(20,676 MW) are ahead of India in fifth position. The short gestation periods for
installing wind turbines, and the increasing reliability and performance of wind
energy machines has made wind power a favored choice for capacity addition in
India.
Suzlon, as Indian-owned Company, emerged on the global scene in the past
decade, and by 2006 had captured almost 7.7 percent of market share in global
wind turbine sales. Suzlon is currently the leading manufacturer of wind turbines
for the Indian market, holding some 52 percent of market share in India. Suzlon’s
success has made India the developing country leader in advanced wind turbine
technology
3.8. CONCEPT OF DISTRIBUTION GENERATION:
Distributed generation, also called on-site generation, dispersed
generation, embedded generation, decentralized generation, decentralized
energy or distributed energy, generates electricity from many small energy sources.
Currently, industrial countries generate most of their electricity in large
centralized facilities, such as fossil fuel (coal, gas powered), nuclear, large solar
power plants or hydropower plants. These plants have excellent economies of
scale, but usually transmit electricity long distances and negatively affect the
environment.
3.9. MICROGRID:
A micro grid is a localized grouping of electricity generation, energy
storage, and loads that normally operate connected to a traditional centralized grid
(macro grid). This single point of common coupling with the macro grid can be

15
disconnected. The micro grid can then function autonomously. Generation and
loads in a micro grid are usually interconnected at low voltage. From the point of
view of the grid operator, a connected micro grid can be controlled as if it was one
entity. Micro grid generation resources can include fuel cells, wind, solar, or other
energy sources. The multiple dispersed generation sources and ability to isolate the
micro grid from a larger network would provide highly reliable electric power.
Byproduct heat from generation sources such as micro turbines could be used for
local process heating or space heating, allowing flexible tradeoff between the
needs for heat and electric power.
3.10. MODES OF POWER GENERATION:
DER systems may include the following devices/technologies:
 Combined heat power (CHP)
 Fuel cells
 Micro combined heat and power (MicroCHP)
 Micro turbines
 Photovoltaic Systems
 Reciprocating engines
 Small Wind power systems
 Stirling engines

16
 4. MODEL PREDICTIVE CONTROL
4.1. INTRODUCTION
In this chapter we consider model predictive control (MPC), an important
advanced control technique for difficult multivariable control problems. The basic
MPC concept can be summarized as follows. Suppose that we wish to control a
multiple-input, multiple-output process while satisfying inequality constraints on
the input and output variables. If a reasonably accurate dynamic model of the
process is available, model and current measurements can be used to predict future
values of the outputs. Then the appropriate changes in the input variables can be
calculated based on both predictions and measurements. In essence, the changes in
the individual input variables are coordinated after considering the input-output
relationships represented by the process model.
In MPC applications, the output variables are also referred to as controlled
variables or CVs, while the input variables are also called manipulated variables or
MVs. Measured disturbance variables are called DVs or feed forward variables.
These terms will be used interchangeably in this chapter. Model predictive control
offers several important advantages: (1) the process model captures the dynamic
and static interactions between input, output, and disturbance variables, (2)
constraints on inputs and outputs are considered in a systematic manner, (3) the
control calculations can be coordinated with the calculation of optimum set points,
and (4) accurate model predictions can provide early warnings of potential
problems. Clearly, the success of MPC (or any other model-based approach)
depends on the accuracy of the process model. Inaccurate predictions can make
matters worse, instead of better. First-generation MPC systems were developed
independently in the 1970s by two pioneering industrial research groups. Dynamic
Matrix Control (DMC), devised by Shell Oil (Cutler and Ramaker, 1980), and a
related approach developed by ADERSA (Richalet et al., 1978) have quite similar

17
capabilities. An adaptive MPC technique, Generalized Predictive Control (GPC),
developed by Clarke et al. (1987) has also received considerable attention. Model
predictive control has had a major impact on industrial practice. For example, an
MPC survey by Qin and Badgwell (2003) reported that there were over 4,500
applications worldwide by the end of 1999, primarily in oil refineries and
petrochemical plants. In these industries, MPC has become the method of choice
for difficult multivariable control problems that include inequality constraints. In
view of its remarkable success, MPC has been a popular subject for academic and
industrial research. Major extensions of the early MPC methodology have been
developed, and theoretical analysis has provided insight into the strengths and
weaknesses of MPC. Informative reviews of MPC theory and practice are available
in books (Camacho and Bordons, 2003; Maciejowski, 2002; Rossiter, 2003;
Richalet and O’Donovan, 2009); tutorials (Hokanson and Gerstle, 1992; Rawlings,
2000), and survey papers (Morari and Lee, 1999; Qin and Badgwell, 2003;
Canney, 2003; Kano and Ogawa, 2009).
4.2. OVERVIEW OF MODEL PREDICTIVE CONTROL
The overall objectives of an MPC controller have been summarized by Qin and
Badgwell (2003):
1. Prevent violations of input and output constraints.

2. Drive some output variables to their optimal set points, while maintaining other
outputs within specified ranges.

3. Prevent excessive movement of the input variables.

4. Control as many process variables as possible when a sensor or actuator is not

available.

18
 Fig.4.1. Block diagram for Model Predictive Control
A process model is used to predict the current values of the output variables.
The residuals, the differences between the actual and predicted outputs, serve as
the feedback signal to a Prediction block. The predictions are used in two types of
MPC calculations that are performed at each sampling instant: set-point
calculations and control calculations. Inequality constraints on the input and output
variables, such as upper and lower limits, can be included in either type of
calculation. Note that the MPC configuration is similar to both the internal model
control configuration and the Smith predictor configuration, because the model
acts in parallel with the process and the residual serves as a feedback signal.
However, the coordination of the control and set-point calculations is a
unique feature of MPC. Furthermore, MPC has had a much greater impact on
industrial practice than IMC or Smith predictor, because it is more suitable for
constrained MIMO control problems. The set points for the control calculations,
also called targets, are calculated from an economic optimization based on a
steady-state model of the process, traditionally, a linear model. Typical

19
optimization objectives include maximizing a profit function, minimizing a cost
function, or maximizing a production rate.
The optimum values of set points change frequently due to varying process
conditions, especially changes in the inequality constraints. The constraint changes
are due to variations in process conditions, equipment, and instrumentation, as well
as economic data such as prices and costs. In MPC the set points are typically
calculated each time the control calculations are performed.

20
 5. WIND-SOLAR SYSTEM MODELLING
In this hybrid system, there are three subsystems: wind subsystem, solar
subsystem, and a lead-acid battery bank which is used to overcome periods of
scarce generation. In the wind generation subsystem, there is a windmill, a multi
polar permanent-magnet synchronous generator (PMSG), a rectifier, and a dc/dc
converter to interface the generator with the dc bus. The converter is used to
control indirectly the operating point of the wind turbine (and consequently its
power generation) by commanding the voltage on the PMSG terminals. The
mathematical description of the wind subsystem written in a rotor reference frame
is as follows
𝑅𝑠 ω𝑒 ∅𝑚 𝜋𝑣𝑏 𝑖𝑞 𝑢𝑤
𝑖̂q=− 𝑖𝑞 − ω𝑒 𝑖𝑑 + −
𝐿 𝐿
3√3𝐿√𝑖𝑞 2 +𝑖𝑑 2
𝑅𝑠 𝜋𝑣𝑏 𝑖𝑑 𝑢𝑤
𝑖̂d=− 𝑖𝑑 − ω𝑒 𝑖𝑞 −
𝐿
3√3𝐿√𝑖𝑞 2 +𝑖𝑑 2
P 3P
ω
̂𝑒 = (𝑇𝑡 − ∅ 𝑖 )
2𝐽 22 𝑚 𝑞
Where 𝑖𝑞 and 𝑖𝑑 are the quadrature current and the direct current in the rotor
reference frame, respectively; Rs and L are the per phase resistance and inductance
of the stator windings, respectively; ω𝑒 is the electrical angular speed; Ϙm is the
flux linked by the stator windings ;Vb is the voltage on the battery bank terminals;
Uw is the control signal [duty cycle of the dc/dc converter], P is the PMSG number
of poles, J is the inertial of the rotating parts, and 𝑇𝑡 is the wind turbine torque. The
wind turbine torque can be written as
1
𝑇𝑡 = 𝐶𝑡 (λ)ρARv2
2

where ρ is the air density, A is the turbine-swept area, R is the turbine radius, v is
the wind speed, and 𝐶𝑡 (λ) is a nonlinear torque coefficient which depends on the
tip speed ratio .The power generated by the wind subsystem and injected into the
dc bus can be expressed as follows:

𝜋𝑣𝑏
𝑃𝑤 = √𝑖𝑞 2 + 𝑖𝑑 2 uw
2√3

21
The model of the wind subsystem can be rewritten in the following compact form:

̇ =fw (xw) + gw (xw) uw

𝑋𝑤
In the solar subsystem, there is a photo-voltaic (PV) panel array and a half-bridge
buck dc/dc converter. The solar subsystem is connected to the dc bus via the dc/dc
converter. In this subsystem, similar to the wind subsystem, the converter is used
to control the operating point of the PV panels. The mathematic description of the
solar subsystem is as follows:
𝑖𝑝𝑣 𝑖
𝑣𝑝𝑣̇ = − 𝑠 𝑢𝑝𝑣
𝐶 𝐶
𝑣𝑏 𝑣𝑝𝑣
𝑖𝑠̇ =− 𝐿𝑐 + 𝐿𝑐 𝑢𝑝𝑣
𝑞(𝑣𝑝𝑣 +𝑖𝑝𝑣 𝑅𝑠 )
𝑖𝑝𝑣 = 𝑛pIph - 𝑛pIrs (exp( )-1)
𝑛𝑠 𝐴𝑐 KT
where 𝑣𝑝𝑣 is the voltage level on the PV panel array terminals, i s is the current
injected into the dc bus, C and Lc are electrical parameters of the buck converter,
𝑢𝑝𝑣 is the control signal (duty cycle), 𝑖𝑝𝑣 is the current generated by the PV array,
𝑛𝑠 is the number of PV cells connected in series, np is the number of series strings
in parallel, K is the Boltzman constant, Ac is the cell deviation from the ideal p-n
junction characteristic, Iph is the photocurrent, and Irs is the reverse saturation
current. The power injected by the PV solar module into the dc bus can be
computed by Ps=𝑖𝑠 𝑣𝑏

Note that this power indirectly depends on the control signal u pv. The model of the
solar subsystem can be rewritten in the following compact form:
𝑥𝑠̇ =fs (xs) + gs (xs) upv
hs(xs)=0
where xs=[𝑣𝑝𝑣 is]T is the state vector of the solar subsystem and fs=[fs1 fs2]T and
gs=[gs1 gs2]T are nonlinear vector functions and hs(xs) is a nonlinear scalar function
whose explicit form is omitted for brevity.

The dc bus collects the energy generated by both wind and solar subsystems
and delivers it to the load and, if necessary, to the battery bank. The voltage of the
dc bus is determined by the battery bank which comprises of lead-acid batteries.
The load could be an ac or a dc load. In the case under consideration in the present
work, it is assumed to be an ac load; therefore, a voltage inverter is required. We

22
also assume that the future load of the system for certain length of time is known,
that is the total power demand is known.
Because all subsystems are linked to the dc bus, their concurrent effects can
be easily analyzed by considering their currents in the common dc side. In this
way, assuming an ideal voltage inverter, the load current can be referred to the dc
side as an output variable current. Therefore, the current across the battery bank
can be written as
𝜋
𝑖𝑏 = √𝑖𝑞 2 + 𝑖𝑑 2 𝑢𝑤 + 𝑖𝑠 − 𝑖𝐿
2 √3
where 𝑖𝐿 is assumed to be a known current. The lead-acid battery bank may be
modeled as a voltage source 𝐸𝑏 connected in series with a resistance 𝑅𝑠 and a
capacitance 𝐶𝑏 . Based on this simple model, the dc bus voltage expression can be
written as follows:
𝜋
𝑣𝑏 = 𝐸𝑏 + 𝑣𝑐 + ( √𝑖𝑞 2 + 𝑖𝑑 2 𝑢𝑤 + 𝑖𝑠 − 𝑖𝐿 ) 𝑅𝑏
2√3
where 𝑣𝑐 is the voltage in capacitor 𝐶𝑏 and its dynamics can be described as
follows:
1 𝜋
𝑣𝑐̇ = ( √𝑖𝑞 2 + 𝑖𝑑 2 𝑢𝑤 + 𝑖𝑠 − 𝑖𝐿 )
𝑣𝑏 2√3
The model of the battery bank can also be rewritten in the following compact form:
𝑣𝑐̇ = 𝑓c(xw,xs,vc)
𝑓 c(xw,xs,vc) is a nonlinear scalar function.
The dynamics of the hybrid generation system can be written in the following
compact form:
𝑥̇ = f(x) + g(x)u
h(x)=0
where x =[xwT xwT vc], u=[uw upv], f(x) and g(x) are suitable composition of fw ,fs
,gw, gs and fc, and h(x)=hs(xs). The explicit forms of f(x) and g(x) are omitted.
The maximum power that can be drawn from the wind and solar subsystems is
determined by the maximum power that can be generated by the two subsystems.
When the two subsystems are not sufficient to complement the generation to
satisfy the load requirements, the battery bank can discharge to provide extra
power to satisfy the load requirements.

23
 Fig.5.1. Supervisory control of a wind-solar hybrid energy system
However, when the power limit that can be provided by the battery bank is
surpassed, the load must be disconnected to recharge the battery bank and avoid
damages. In this work, we do not consider the power needed to charge the battery
bank explicitly. However, this power can be lumped into the total power demand.
In the reminder of this work, we refer to the total power demand as PT.

24
 6. CONTROL PROBLEM FORMULATION AND
CONTROLLERS DESIGN FOR SOLAR AND WIND SUBSYSTEM
6.1. Control Problem Formulation
We consider two control objectives of the wind-solar energy generation system.
The first and primary control objective is to compute the operating points of the
wind subsystem and of the solar subsystem together to generate enough energy to
satisfy the load demand. The second control objective is to optimize the operating
points to reduce the peak value of surge currents. With respect to the second
control objective, specifically, we consider that there are maximum allowable
increasing rates of the generated power of the two subsystems and that frequent
discharge and charge of the battery bank should be avoided to maximize battery
life. Note that the constraints on the maximum increasing rates impose indirect
bounds on the peak values of inrush or surge currents to the two subsystems.
The proposed control system is shown in which the supervisory control system
optimizes the power references and operating points of the wind and solar
subsystems, respectively. The two local controllers (wind subsystem controller and
the solar subsystem controller) manipulate and to track the power references,
respectively. In this work, we consider hybrid energy generation systems that
already operate in normal generating conditions, and do not address the issues
related to system startup or shut down. Moreover, we focus on the application of
the proposed supervisory control system and do not provide specific conditions
(and detailed theoretical derivation) under which the stability of the closed-loop
system is guaranteed. We also note that, in the case of a energy generation system
containing several solar and wind subsystems, the proposed supervisory control
approach can be extended to control the system in a conceptually straightforward
manner by letting the supervisory controller determine the power references of all
the subsystems or a distributed MPC approach (in which each MPC controls only
some of the subsystems) can be applied. However, the distributed MPC approach
is out of the scope of this work.

25
6.2. Wind Subsystem Controller Design
For the wind subsystem controller, the objective is to track the power reference
computed by the supervisory predictive controller. In order to proceed, we
introduce the maximum power that can be provided by a wind subsystem
2 2
Pw,max=Pw,max(x)=Koptωm3− 32 (𝑖𝑞 + 𝑖𝑑 ) rs
where Kopt=(Ct(λopt)ρAR3)/(2λopt2) and λopt is the tip speed ratio at which the
coefficient reaches its maximum, and is the torque coefficient of the wind turbine.
Specifically, the controller is designed as follows:
uw = uw1, if Pw,ref < Pw,max
= uw2 if Pw,ref ≥ Pw,max
where

uw1= −[6rs (iqfw1+ idfw2) −3∅sr (wefw1+iqfw3) + 2(γ|sw1(xw)| +ζmax

|∂sw1/|∂xw|) x sign(sw1 ( xw ) )]/( 6rs (iqgw1+ idgw2)- 3∅sr wegw1)
uw2= −fw1/ gw1 + 2Koptwefw3/(∅sr gw1)-iqfw3/( wegw1)+ 2(γ|sw2(xw)| +ζmax
|∂sw2/|∂xw|) x sign(sw2 ( xw ) )]/( 3∅sr wegw1)
with γ=1000 and ζmax=0.02 being design constants and

∂sw1 3 2
= √4rs (𝑖𝑞 + 𝑖𝑑 2 ) + ∅𝑠𝑟 2 (we 2 + 𝑖𝑞 2) − 4rs ∅𝑠𝑟 we 𝑖𝑞
∂xw 2

∂sw2 3 2 3
= √( ∅𝑠𝑟 we ) + (3K opt we − ∅𝑠𝑟 𝑖𝑞 )
2 2
∂xw 2 2

In the control design shown in (14), sw1=Pwref-Pw and sw1= Pwmax are the sliding
surfaces. When the power reference is less than the maximum power that can be
provided by the wind subsystem, the control law will operate the subsystem to
generate the desired power; when the power reference is greater than the maximum
power that can be provided by the wind subsystem, the control law will drive the
subsystem to operate at points in which the subsystem provides the maximum
power.

26
6.3. Solar Subsystem Controller Design
The objective of the solar subsystem controller is to force the subsystem to track
the power reference computed by the supervisory controller. The maximum power
operating point (MPOP) of the solar subsystem can be computed, in principle, by
the following expression :

∂Ppv ∂ipv
= v + ipv = 0
∂vpv ∂vpv pv
In the present work, the maximum solar power provided, Ppv,max , is computed
numerically through direct evaluation of the following expression in the region
close to zero
∂ipv ∆ipv
Ppv,max = Ppv,max (x)=- ∂v vpv 2 ≈∆v vpv 2
pv pv
Specifically, this controller is designed as follows:

If Ppv,max ≥ Ps,ref upv=1 if h1≥ 0

=0 if h1<0
Ppv,max < Ps,ref upv=0 if h2≥ 0
=1 if h2<0
∂ipv ipv
Where h1= Ps,ref –isvb and h2= +
∂vpv vpv

27
 APPENDIX
7. MATLAB
7.1. Introduction to Matlab:
Matlab is a high-performance language for technical computing. The name
matlab stands for matrix laboratory. It integrates computation, visualization, and
programming in an easy-to-use environment where problems and solutions are
expressed in familiar mathematical notation. Typical uses include Math and
computation Algorithm development Data acquisition Modeling, simulation, and
prototyping Data analysis, exploration, and visualization Scientific and engineering
graphics Application development, including graphical user interface building.
Matlab is an interactive system whose basic data element is an array that
does not require dimensioning. This allows you to solve many technical computing
problems, especially those with matrix and vector formulations, in a fraction of the
time it would take to write a program in a scalar no interactive language such as C
or FORTRAN.
7.2. Components of Matlab:
 Workspace
 Current Directory
 Command History
 Command Window

Fig.7.1. Block diagram of Matlab components

28
7.3. Toolboxes in Matlab:
 Simulink
 Fuzzy
 Genetic algorithm
 Neural network
 Wavelet
7.4. Simulink:
7.4.1 Introduction:
Simulink is a software add-on to matlab which is a mathematical tool
developed by The Math works,(http://www.mathworks.com) a company based in
Natick. Mat lab is powered by extensive numerical analysis capability. Simulink is
a tool used to visually program a dynamic system (those governed by Differential
equations) and look at results. Any logic circuit, or control system for a dynamic
system can be built by using standard building blocks available in Simulink
Libraries. Various toolboxes for different techniques, such as Fuzzy Logic, Neural
Networks, DSP, Statistics etc. are available with Simulink, which enhance the
processing power of the tool. The main advantage is the availability of templates /
building blocks, which avoid the necessity of typing code for small mathematical
processes.
7.4.2 Concept of signal and logic flow:
In Simulink, data/information from various blocks are sent to another block
by lines connecting the relevant blocks. Signals can be generated and fed into
blocks dynamic / static).Data can be fed into functions. Data can then be dumped
into sinks, which could be scopes, displays or could be saved to a file. Data can be
connected from one block to another, can be branched, multiplexed etc. In
simulation, data is processed and transferred only at discrete times, since all

29
computers are discrete systems. Thus, a simulation time step (otherwise called an
integration time step) is essential, and the selection of that step is determined by
the fastest dynamics in the simulated system.

Fig.7.2. Simulink library browser

7.6 Mathematical operations:
Mathematical operators such as products, sum, logical operations such as
and, or, etc. can be programmed along with the signal flow.Trigonometric
functions such as sin or tan inverse (at an) are also available. Relational operators
such as ‘equal to’, ‘greater than’ etc. can also be used in logic circuits.

30
 Fig 7.7.Simulink math blocks
7.7. Signals & data transfer:
In complicated block diagrams, there may arise the need to transfer data
from one portion to another portion of the block. They may be in different
subsystems. That signal could be dumped into a GOTO block, which is used to
send signals from one subsystem to another.
Multiplexing helps us remove clutter due to excessive connectors, and
makes matrix (column/row) visualization easier.

Fig.7.8. signals and systems

31
7.8. Making subsystems:
Drag a subsystem from the Simulink Library Browser and place it in the
parent block where you would like to hide the code. The type of subsystem
depends on the purpose of the block. For instance, the subsystem can be a triggered
block, which is enabled only when a trigger signal is received.
Open (double click) the subsystem and create input / output PORTS, which
transfer signals into and out of the subsystem. The input and output ports are
created by dragging them from the Sources and Sinks directories respectively.
When ports are created in the subsystem, they automatically create ports on the
external (parent) block. This allows for connecting the appropriate signals from the
parent block to the subsystem.
7.9. Setting simulation parameters:
Running a simulation in the computer always requires a numerical technique
to solve a differential equation. The system can be simulated as a continuous
system or a discrete system based on the blocks inside. The simulation start and
stop time can be specified.Simulation step size must be decided based on the
dynamics of the system. A thermal process may warrant a step size of a few
seconds, but a DC motor in the system may be quite fast and may require a step
size of a few milliseconds.

Fig.7.9. Simulation Parameters

32
 7. SIMULATION RESULTS AND DISCUSSIONS
7.1. Constraints on the Maximum Increasing Rates of Pwref and Psref
In this set of simulations, the control objective is to operate the hybrid wind-solar
energy generation system to satisfy the total power demand , subject to constraints
on the rate of change of Pwref and Psref. Because the constraints on the maximum
increasing rates of Pwref and Psref are considered as hard constraints in the
formulation of the MPC [i.e., constraints of (18d)–(18e)], in the cost function, we
only penalize the total power demand. The cost function designed for these control
objectives is shown as follows:
𝐿(𝑥, 𝑃𝑤,𝑟𝑒𝑓 , 𝑃𝑠,𝑟𝑒𝑓 ) = 𝛼(𝑃𝑇 − 𝑃𝑤,𝑟𝑒𝑓 − 𝑃𝑠,𝑟𝑒𝑓 )2 + 𝛽𝑃𝑠,𝑟𝑒𝑓 2
where 𝛼 = 1 𝑎𝑛𝑑 𝛽 = 0.01 are constant weighting factors. The first term,
𝛼(𝑃𝑇 − 𝑃𝑤,𝑟𝑒𝑓 − 𝑃𝑠,𝑟𝑒𝑓 )2 , in the cost function penalizes the difference between the
power generated by the wind-solar system and the total power demand, which
drives the wind and solar subsystems to satisfy the total demand to the maximum
extent. Because there are infinite combinations of Pwref and Psref that can
minimize the first term, in order to get a unique solution to the optimization
problem, we also put a small penalty on . This implies that the wind subsystem is
operated as the primary generation system and the solar subsystem is only
activated when the wind subsystem alone can not satisfy the power demand. In the
simulation, we assume that the environmental conditions remain constant with
wind speed 12 m/s, insolation 90 mW/cm and PV panel temperature 65 C. We see
in Fig7.1 that at 4 s there is a demand power increase from 2100 to 4000 W,and
that because of the constraints on the maximum increasing rates of Pwref and Psref
, the wind solar system cannot supply sufficient power and the shortage of power is
made up by the battery bank. Note that we assume that the future power demand
for a short time period is known to the MPC. Because of this, at 8 s, when the MPC
supervisory controller receives information about a power demand increase at 9 s,
and having information of the limits on the power generation of the two
subsystems, it coordinates the power generations of the wind and solar subsystems
to best satisfy the power demand by reducing the power generation of the wind
subsystem and activating the solar subsystem in advance at 8 s. This coordination
renders the two subsystems able to approach as much as possible to the total power
demand requirement at 9 s (even though they cannot fully meet this requirement
due to operation constraints of the wind and solar subsystems) by boosting their

33
power production at the maximum possible rate, i.e., about 1500 W boost in power
production from 8 s to 9 s. On the other hand, if there is no information of the
future power demand increase that is fed to the MPC, the wind-solar system would
not increase its production as fast to approach the total power demand requirement
because the solar subsystem would stay dormant up to 9 s (the power demand
requirement at 8 s can be fully satisfied by the wind subsystem only) and the
presence of a hard constraint on the rate of change of power generated by the solar
ubsystem would not allow to boost its production enough to meet the total power
demand requirement at 9 s (in this case, the total power demand requirement
cannot be achieved by operation of the wind subsystem only); as a result the boost
in total power production in this case would be only 1200 W.

Fig.7.1. Waveforms showing Load Power Requirement,Power Supplied From

Wind and Solar Generation System, Power Deficit Supplied by Battery

34
7.2. Suppression of Battery Power Fluctuation.
In this set of simulations, the cost function of (20) is modified to take into account
the fluctuation of the battery power in order to avoid frequent battery charge and
discharge. The cost function is modified as follows:
𝐿(𝑥, 𝑃𝑤,𝑟𝑒𝑓 , 𝑃𝑠,𝑟𝑒𝑓 ) = 𝛼(𝑃𝑇 − 𝑃𝑤,𝑟𝑒𝑓 − 𝑃𝑠,𝑟𝑒𝑓 )2 + 𝛽𝑃𝑠,𝑟𝑒𝑓 2 +𝑃𝑏2
where is the change of the power provided by the battery bank between two
consecutive steps and is a weighting factor. Note that this newly added
term requires that we store the trajectory of Pb. In this set of simulations, the
environmental conditions are set with wind speed 11 m/s, insolation 90 mW/cm2
and PV panel temperature 65 C. It is seen in Fig7.2 that there is a power demand
decrease at 3 s, and though the wind and solar subsystems are able to provide
enough power to satisfy the demand, the supervisory controller will not reduce the
power generated by the battery to 0 immediately at 3 s; instead, the supervisory
controller operates the system to make the power provided by the battery bank
decrease slower and reach its recharge state at 5 s fig shows the power trajectory of
the battery bank if no penalty on the change of the power provided by the battery
bank is applied.

Fig.7.2.Waveforms showing gradual decrease in power supplied by battery bank

35
7.3. Varying Environmental Conditions
In this subsection,simulations carry out under varying environmental condition.
Time evolution of wind speed, PV panel temperature and insolation are shown. fig
shows the trajectory of total power demand. It is seen in Fig7.4 that the wind/
solar/ battery powers coordinate their behavior to meet the load demand. Time
evolution of output power and maximum available power from the wind subsystem
and solar subsystem are plotted. When sufficient energy supply can be extracted
from the two subsystems such as during 0–60 s, 100–140 s, and 160–173 s, the
battery is being recharged. In other periods, load demand is relatively high and the
weather condition, which determines the maximum available generation capacity
of the two subsystems, cannot permit sufficient energy supply. Thus, the
supervisory controller drives wind/solar parts to their instant maximum capacity
and calls the battery bank for shortage compensation.
*

Fig.7.3.Waveforms showing varying Environmental conditions and Load current

36
Fig.7.4.Waveforms showing Standalone Power supply to Load during changing
environmental conditions

7.4. Consideration of High-Frequency Disturbance of Weather Condition

In the preceding scenario, it is assumed that the variation of weather-related

parameters, like wind speed and insolation, within each sampling time interval is
negligible. While this assumption is reasonable in most cases, additional attention
for robust system operation should be given under even harsher conditions where
high frequency disturbances that influence the values of wind speed and insolation
are present. This scenario is possible when the wind turbine encounters turbulent
flow or when insolation is affected by abrupt changes in atmospheric turbidity. To
study this case from a control point of view and evaluate the robustness of the
proposed control system in this case, disturbances in two parameters; specifically,
10% variation in the wind speed and 5% variation in the insolation. The profiles of
the wind speed and insolation are shown in Fig7.5 this conservative estimate of
insolation ensures that the predicted maximum power delivered by the solar
subsystem does not exceed what the weather permits. Both maximum power
generation capabilities of the two subsystems are perturbed as a result of the

37
weather disturbance, but both the wind subsystem and the solar subsystem operate
in a robust fashion and the total power demand is met as shown in Fig7.6

Fig.7.5.Waveforms showing high frequency Environmental conditions and Load

current

38
Fig.7.6.Waveforms showing Standalone Power supply to Load during high
frequency disturbances in environment

39
 CONCLUSION
In this work, It has been focused on the development of a supervisory
predictive control method for the optimal management and operation of hybrid
wind-solar energy generation systems. A supervisory control system via MPC
which computes the power references for the wind and solar subsystems is
designed at each sampling time while minimizing a suitable cost function. The
power references are sent to two local controllers which drive the two subsystems
to the power references. Simulation results demonstrated the effectiveness and
applicability of the proposed approach.

40
 REFERENCES

[1] California Energy Commission, “2008 energy policy report update,” 2008.

[2] R. Spee and J. H. Enslin, “Novel control strategies for variable-speed doubly
fed wind power generation systems,” Renewable Energy, vol. 6, pp. 907–915,
1995.

[3] P. Novak, T. Ekelund, Y. Jovik, and B. Schmidtbauer, “Modeling and control

of variable-speed wind-turbine drive system dynamics,” IEEE Control Syst. Mag.,
vol. 15, no. 4, pp. 28–37, Aug. 1995.

[4] T. Thiringer and J. Linders, “Control by variable rotor speed of fixed-pitch

wind turbine operating in speed range,” IEEE Trans. Energy Conv., vol. 8, no. 3,
pp. 520–526, Sep. 1993.

[5] M. G. Simoes, B. K. Bose, and R. J. Spiegel, “Fuzzy logic based intelli gent
control of a variable speed cage machine wind generation system,” IEEE Trans.
Power Electron., vol. 12, no. 1, pp. 87–95, Jan. 1997.

[6] K. Uhlen, B. A. Foss, and O. B. Gjosaeter, “Robust control and analysis of a

wind-diesel hybrid power plant,” IEEE Trans. Energy Conv., vol. 9, no. 4, pp.
701–708, Dec. 1994.

[7] F. Valenciaga, P. F. Puleston, R. J. Mantz, and P. E. Battaiotto, “An adaptive

feedback linearization strategy for variable speed wind energy conversion
systems,” Int. J. Energy Res., vol. 24, pp. 151–161, 2000.

[8] K. Tan and S. Islam, “Optimum control strategies in energy conver sion of
PMSG wind turbine system without mechanical sensors,” IEEE Trans. Energy
Conv., vol. 19, no. 2, pp. 392–399, Jun. 2004.

[9] F. Valenciaga, P. F. Puleston, and P. E. Battaiotto, “Variable struc ture system

control design method based on a differential geometric approach: Application to a
wind energy conversion subsystem,” IEE Proc.—Control Theory Appl., vol. 151,
pp. 6–12, 2004.

41
[10] M. Chinchilla, S. Arnaltes, and J. C. Burgos, “Control of perma nent-magnet
genrators applied to variable-speed wind energy systems connected to the grid,”
IEEE Trans. Energy Conv., vol. 21, no. 1, pp. 130–135, Mar. 2006.

[11] F. Valenciaga and P. F. Puleston, “High-order sliding control for a wind

energy conversion system based on a permanent magnet synchronous generator,”
IEEE Trans. Energy Conv., vol. 23, no. 3, pp. 860–867, Sep. 2008.

[12] D. Q. Dang, Y. Wang, and W. Cai, “Nonlinear model predictive control

(NMPC) of fixed pitch variable speed wind turbine,” in Proc. IEEE Int. Conf.
Sustainable Energy Technol., Singapore, 2008, pp. 29–33.

[13] T. A. Johansen and C. Storaa, “Energy-based control of a distributed solar

collector field,” Automatica, vol. 38, pp. 1191–1199, 2002.

[14] F. Coito, J. M. Lemos, R. N. Silva, and E. Mosca, “Adaptive control of a solar

energy plant: Exploiting accessible disturbances,” Int. J. Adapt. Control Signal
Process., vol. 11, pp. 327–342, 1997.

[15] E. F. Camacho and M. Berenguel, “Robust adaptive model predictive control

of a solar plant with bounded uncertainties,” Int. J. Adapt. Con trol Signal Process.,
vol. 11, pp. 311–325, 1997.

[16] N. Hamrouni, M. Jraidi, and A. Cherif, “New control strategy for 2-stage grid-
connected photovoltaic power system,” Renewable En ergy, vol. 33, pp. 2212–
2221, 2008.

[17] T. Yoshida, K. Ohniwa, and O. Miyashita, “Simple control of pho tovoltaic

generator systems with high-speed maximum power point tracking operation,”
EPE J., vol. 17, pp. 38–42, 2007.

[18] F. Valenciaga, P. F. Puleston, P. E. Battaiotto, and R. J. Mantz, “Pas

sivity/sliding mode control of a stand-alone hybrid generation system,” IEE
Proc.—Control Theory Appl., vol. 147, pp. 680–686, 2000.

[19] F. Valenciaga, P. F. Puleston, and P. E. Battaiotto, “Power control of a

photovoltaic array in a hybrid electric generation system using sliding mode
techniques,” IEE Proc.—Control Theory Appl., vol. 148, pp. 448–455, 2001.

42
[20] F. Valenciaga and P. F. Puleston, “Supervisor control for a stand-alone hybrid
generation system using wind and photovoltaic energy,” IEEE Trans. Energy
Conv., vol. 20, no. 2, pp. 398–405, Jun. 2005.

[21] N. A. Ahmed, M. Miyatake, and A. K. Al-Othman, “Hybrid solar pho

tovoltaic/wind turbine energy generation system with voltage-based maximum
power point tracking,” Electric Power Components Syst., vol. 37, pp. 43–60, 2009.

[22] C. E. García, D. M. Prett, and M. Morari, “Model predictive control: Theory

and practice—A survey,” Automatica, vol. 25, pp. 335–348, 1989.

[23] J. B. Rawlings, “Tutorial overview of model predictive control,” IEEE

Control Syst. Mag., vol. 20, no. 3, pp. 38–52, Jun. 2000.

[24] P. D. Christofides and N. H. El-Farra, Control of Nonlinear and Hybrid

Process Systems: Designs for Uncertainty, Constraints and Time-De lays.. New
York: Springer, 2005.

[25] P. Mhaskar, N. H. El-Farra, and P. D. Christofides, “Stabilization of nonlinear

systems with state and control constraints using Lyapunov based predictive
control,” Syst. Control Lett., vol. 55, pp. 650–659, 2006.

[26] P. Mhaskar, N. H. El-Farra, and P. D. Christofides, “Predictive control of

switched nonlinear systems with scheduled mode transitions,” IEEE Trans. Autom.
Control, vol. 50, no. 11, pp. 1670–1680, Nov. 2005.

[27] D. Muñoz de la Peña and P. D. Christofides, “Lyapunov-based model

predictive control of nonlinear systems subject to data losses,” IEEE Trans.
Autom. Control, vol. 53, no. 9, pp. 2076–2089, Oct. 2008.

[28] J. Peinke, E. Anahua, St. Barth, H. Goniter, A. P. Schaffarczyk, D. Kleinhans,

and R. Friedrich, “Turbulence a challenging issue for the wind energy conversion,”
presented at the Eur. Wind Energy Conf. Ex hibition, Brussels, Belgium, 2008.

[29] J. Hofierka and M. Suri, “The solar radiation model for open source GIS:
Implementation and applications,” in Proc. Open Source GIS—GRASS Users
Conf., Trento, Italy, 2002, pp. 1–19.

43