Hindawi Publishing Corporation

International Journal of Photoenergy

Volume 2014, Article ID 938526, 13 pages
http://dx.doi.org/10.1155/2014/938526

Research Article
A Newton-Based Extremum Seeking MPPT Method for
Photovoltaic Systems with Stochastic Perturbations

Heng Li,1,2 Jun Peng,1,2 Weirong Liu,1,2 Zhiwu Huang,1,2 and Kuo-Chi Lin3
1
School of Information Science and Engineering, Central South University, Changsha 410075, China
2
Hunan Engineering Laboratory for Advanced Control and Intelligent Automation, Changsha 410075, China
3
Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, FL 32816, USA

Correspondence should be addressed to Jun Peng; [email protected]

Received 2 May 2014; Revised 6 July 2014; Accepted 6 July 2014; Published 24 July 2014

Academic Editor: Dimitrios Karamanis

Copyright © 2014 Heng Li et al. This is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Microcontroller based maximum power point tracking (MPPT) has been the most popular MPPT approach in photovoltaic systems
due to its high flexibility and efficiency in different photovoltaic systems. It is well known that PV systems typically operate under
a range of uncertain environmental parameters and disturbances, which implies that MPPT controllers generally suffer from some
unknown stochastic perturbations. To address this issue, a novel Newton-based stochastic extremum seeking MPPT method is
proposed. Treating stochastic perturbations as excitation signals, the proposed MPPT controller has a good tolerance of stochastic
perturbations in nature. Different from conventional gradient-based extremum seeking MPPT algorithm, the convergence rate
of the proposed controller can be totally user-assignable rather than determined by unknown power map. The stability and
convergence of the proposed controller are rigorously proved. We further discuss the effects of partial shading and PV module
ageing on the proposed controller. Numerical simulations and experiments are conducted to show the effectiveness of the proposed
MPPT algorithm.

1. Introduction control system to regulate operating current or voltage to the

maximum output power point.
Recent years have seen a growing interest in the research of With the rapid development of embedded technology,
solar energy, which is mainly due to the advantages solar microcontroller based MPPT control system has been the
energy has over traditional fossil energies, including safety, dominated approach in current photovoltaic systems [5]. This
sustainable source of energy, less environment pollution, and approach is favoured because researchers and engineers pre-
numerous market potentials. As the main available solar tech- fer to program elegant and advanced MPPT algorithms rather
nology, photovoltaic array has been widely used in satellites than design complicated and expensive MPPT control cir-
and spacecrafts, solar vehicles, and domestic power supply; cuits.
see [1, 2] and the references therein. There are a number of MPPT control algorithms that
The process of guiding a photovoltaic array to its maxi- have been proposed in the literature, among which the two
mum power point is called maximum power point tracking. primary algorithms are incremental conductance (IncCond)
Due to the inherent nature of photovoltaic cells, the power- algorithm [5] and perturb and observe (PO) algorithm [6]. In
voltage curves of PV cells depend nonlinearly on temperature PO algorithms, a step perturbation is added into the control
and irradiation intensity; see [3–5]. This fact, however, means signal and used to monitor the direction of changes in power.
that the operating current or voltage that maximizes the out- PO has been a commercial algorithm because of its ease of
put power will change with environmental conditions. To implementation. The main drawback of PO algorithm is that
maintain the maximum output power regardless of environ- it will oscillate at the maximum power point. Also, it has
mental parameters, one common way is to design MPPT been shown that PO fails to track the rapidly changing
2 International Journal of Photoenergy

irradiance [7]. The IncCond algorithm tracks the maxi- Ipv

mum power point by comparing the instantaneous and
incremental conductances of the PV array. Thus it can track Iph Rs
the rapidly changing irradiance. However, the harmonic Id
components of array voltage and current need to be measured PV Upv Load
Rp module
and used to adjust the reference voltage, which implies that
errors at the maximum power point occur due to the low
precision sensors used [8].
A promising new MPPT algorithm is the method of extre-
mum seeking (ES) control; see [9–12]. As a model-free, real-
time optimization method, ES is well suited for applications Figure 1: The equivalent electrical circuit of a photovoltaic module.
with unknown or partly known dynamics and external
perturbations, such as photovoltaic systems [11]. ES uses per-
turbation signals (either from external disturbances or from 2. Photovoltaic Modelling and MPPT
converter ripples) as probing signals to estimate the gradi-
ent of the power map and then update the control signal In this section, we will introduce some preliminary knowl-
according to the estimation [12]. It has been shown that ES edge about PV array and MPPT. A photovoltaic array typ-
has the advantages of both rigorously provable convergence ically consists of several photovoltaic modules connected in
and simplicity of hardware implementation [13]. However, series or parallel to obtain the desired output voltage and cur-
existing extremum seeking MPPT controllers still suffer from rent. Thus we first introduce the electrical characteristics of
the following two limitations. PV modules. More detailed background information can be
found in [16].
(i) In existing extremum seeking MPPT controllers, per-
A photovoltaic module can be seen as a black box that
turbation signals are generally assumed to be peri-
produces a current 𝐼pv at a voltage 𝑈pv . The inside of that black
odic. On the one hand, the assumption is rather ideal
box can be described by an electric circuit with 4 components,
since external disturbances are typically unknown
as shown in Figure 1.
and stochastic; on the other hand, the requirement of
orthogonality makes periodic ES difficult to extend to (i) Current source: this is the source of the photo current,
multivariable cases. which can be denoted by
(ii) In existing extremum seeking MPPT controllers, the
convergence speed is typically defined by the gradient 𝐼ph = 𝑆 ⋅ 𝐻 ⋅ 𝜉, (1)
of power map of PV systems, which implies that
control systems will be highly influenced by unknown where 𝑆 is the module area, 𝐻 is the intensity of
and changing environmental conditions. To construct incoming light, and 𝜉 is the response factor.
a practical MPPT control system, the convergence
(ii) Diode: this nonlinear element reflects the dependence
of MPPT controller should be user-assignable rather
on the band gap and losses to recombination. It is
than being dependent on environmental conditions.
characterized by the reverse current 𝐼𝑑 .
Thus the aim of this paper is to provide an improved
(iii) Shunt resistor 𝑅𝑝 : it represents losses incurred by
extremum seeking MPPT control algorithm to solve these
conductors.
problems. Recent progress in stochastic averaging theory
has made it possible to consider more general stochastic (iv) Serial resistor 𝑅𝑠 : it also represents losses incurred by
perturbations when designing extremum seeking controllers; nonideal conductors.
see [14, 15]. In [14], Liu and Krstic provide a systematic design
of extremum seeking when stochastic perturbations exist The relationship between current 𝐼pv and voltage 𝑈pv of a
and prove the stability of the stochastic extremum seeking photovoltaic module is then expressed by
via a developed stochastic averaging theory. The theoretical
𝑈pv + 𝐼pv 𝑅𝑠 𝑈pv + 𝐼pv 𝑅𝑠
part of this paper is based on the findings of [14, 15] and 𝐼pv = 𝐼ph − 𝐼𝑑 [exp ( ) − 1] − , (2)
we further propose a Newton-based stochastic extremum 𝑈𝑇 𝑅𝑝
seeking MPPT controller considering physical features of
photovoltaic systems. We conduct extensive simulations and where 𝑈𝑇 = 𝑞𝑘𝑇/𝑒 with ideality factor 𝑞, temperature 𝑇,
experiments to show the effectiveness of the proposed control Boltzmann constant 𝑘 = 1.38𝑒−23 , and the elementary charge
algorithm. 𝑒 = 1.602𝑒−19 .
The remainder of the paper is organized as follows. In Then the output power 𝑃pv of a photovoltaic module is
Section 2, preliminary knowledge about photovoltaic mod- denoted by
elling and maximum power point tracking is introduced. In
Section 3, a new Newton-based stochastic extremum seeking 𝑃pv = 𝑈pv 𝐼pv = 𝑓 (𝑈pv ) , (3)
MPPT controller is designed. The tracking performance
of the proposed controller is evaluated in Section 4. We where function 𝑓(⋅) is determined by manufacturing and
conclude the paper in Section 5. environmental parameters, such as module area 𝑆, irradiance
International Journal of Photoenergy 3

4 3. MPPT Controller Design

3.5
In this section, we propose a new Newton-based stochastic
3
extremum seeking MPPT controller to handle the above
Current (A)

2.5 two difficulties. The proposed controller not only inherits

2 the advantages of classical extremum seeking in model-free
1.5 optimization but also shows its distinctive features in dealing
1 with stochastic perturbations and assigning convergence
0.5 rates. In what follows, we will introduce the control struc-
0
ture, implementation, and stability analysis of the proposed
0 5 10 15 20 25 controller. Some further discussions will be provided in
Voltage (V) Section 3.5.
H = 1000 W/m2 , T = 25∘ C H = 600 W/m2 , T = 25∘ C
H = 800 W/m2 , T = 25 ∘C Maximum power point 3.1. Controller Structure. In order to illustrate the motivation
of designing Newton-based extremum seeking MPPT con-
Figure 2: I-V characteristics of KC65T PV module at various irra- troller, we first introduce the application of classical gradient-
diance levels. based extremum seeking in MPPT of photovoltaic systems.
Figure 4 shows the control structure of a gradient-based
extremum seeking MPPT controller.
As shown in Figure 4, 𝜏 is the duty cycle of PWM waves,
𝐻, response factor 𝜉, and temperature 𝑇, and thus is generally ̂pv is the estimation of
𝜂 is the stochastic perturbation, and 𝑈
unknown or partially known a priori.
desired voltage that optimizes the power map. The model of
Figure 2 shows the I-V characteristics of the KC65T DC-DC driver and converter can be simplified as follows:
photovoltaic module at various irradiance levels; see the
datasheet of KC65T [18]. With light varying its intensity ̂pv ) ,
𝜏 = 𝑔 (𝑈pv , 𝑈 (4)
throughout the day, the maximum power point moves to
different voltages and currents. Thus a MPPT control system 𝑈pv = ℓ (𝜏) . (5)
is typically adopted between the photovoltaic module and the
load to adjust the output voltage or current and maintain the The maximum power point (MPP) of the photovoltaic
maximum power output. system is defined as
The schematic of a photovoltaic MPPT control system ∗
𝑃pv ∗
= 𝑓 (𝑈pv ) = 𝑓 (ℓ (𝜏∗ )) ≜ max 𝑃pv , (6)
with stochastic perturbations is shown in Figure 3. The pho-
∗
tovoltaic panel converts solar energy to electrical energy via where 𝑈pv , 𝜏∗ are the desired voltage and duty cycle at the
the photovoltaic effect. A MPPT control system is designed MPP.
between the photovoltaic panel and load to optimize the con- We then denote the voltage estimation error 𝑈̃pv by
version efficiency. The adopted MPPT control system consists
of three components: MPPT controller, DC-DC driver, and ̃pv = 𝑈
𝑈 ̂pv − 𝑈∗ . (7)
pv
DC-DC converter. The proposed MPPT system can be either
voltage control or current control depending on which one For the purpose of illustration, assume that the power
is used as the control variable of the MPPT controller. In map 𝑓(⋅) is of the quadratic form; then the averaged system
this paper, we design the MPPT control system with voltage is obtained by
control, but the proposed method is also suited to current
̃̇ pv = 𝑘H𝑈
𝑈 ̃pv , (8)
control cases.
As the DC-DC driver and converter are relatively mature where H is the second derivative (Hessian) of the power map.
in electrical industry, the main work in MPPT control Equation (8) shows that the convergence rate is governed by
system is the design of MPPT control algorithm. Such a the unknown Hessian H, which is highly influenced by envi-
research interest has continued for decades, since even small ronmental conditions, such as irradiance and temperature.
improvements in performance can lead to great economic A practical MPPT control system often requires that the
benefits. Though considerable progress has been made in this convergence of the controller should be designer-assignable.
field, the following two problems have not been effectively We next show that this goal can be realized with the Newton-
solved: based extremum seeking.
If the power map is known, the following Newton opti-
(i) how to optimize the output power with unknown mization algorithm can be used to find the maximum power
power map when stochastic external disturbances point:
exist; −1
𝑑𝑃pv 𝑑2 𝑓(𝑈pv ) 𝑑𝑓 (𝑈pv )
(ii) how to assign the convergence speed by designers = −( 2
) . (9)
rather than unknown power map. 𝑑𝑡 𝑑𝑈pv 𝑑𝑈pv
4 International Journal of Photoenergy

Sun irradiation
DC-DC converter
Ipv
SW L

C D Load
Upv

Upv Ipv
∗ Upv
Photovoltaic panel Upv +−
MPPT
controller +− PI DC-DC driver

Stochastic
perturbations

Figure 3: The schematic of a photovoltaic MPPT control system with stochastic perturbations.

Upv where 𝜂 is a general stochastic signal and 𝑆(𝜂) = 𝑎 sin(𝜂),

Ppv = f(Upv ) 𝑀(𝜂), and 𝑁(𝜂) are the output of signal generator Ψ with the
original signal 𝜂.
DC-DC
converter
𝜏
Remark 1. The signal generator outputs 𝑀(⋅), 𝑁(⋅) play an
DC-DC ̂
driver Upv+ k important role in the Newton-based stochastic extremum
×
s seeking, since they are used to estimate the first-order and
second-order derivative of power map 𝑓(⋅), respectively.
𝜂(t) Theoretically, they can be any bounded and odd continuous
Figure 4: The structure of gradient-based extremum seeking MPPT
functions. However, considering the practical stability of the
controller. photovoltaic system, we choose 𝑀(⋅), 𝑁(⋅) by following the
design principle in [19] as

1
𝑀 (𝜂) = sin (𝜂) ,
If 𝑓(⋅) is unknown, then an estimator is needed to approx- 𝑎𝑊0 (𝑞)
imate 𝑑𝑓(𝑥)/𝑑𝑥 and 𝑑2 𝑓(𝑥)/𝑑𝑥2 . Then the purpose of the (11)
controller design is to combine the Newton optimization 1 2
𝑁 (𝜂) = (sin (𝜂) − 𝑊0 (𝑞)) ,
algorithm (9) with estimators of the first and second deriva- 𝑎2 𝑊02 (√2𝑞)
tives of power map to achieve the maximum power point
tracking. where variables 𝑊0 (𝑞) and 𝑊1 (𝑞) are defined as follows:
Let 𝐺̂ be the estimation of the first-order derivative, H
̂
the estimation of the second-order derivative, and Γ the
estimation of inverse of second derivative. Then we pro- 𝑊02 (√2𝑞) = 2 (𝑊1 (𝑞) − 𝑊02 (𝑞)) ,
pose a Newton-based stochastic extremum seeking MPPT 1 2
controller in Figure 5. As shown in Figure 5, the first-order 𝑊0 (𝑞) = (1 − 𝑒−𝑞 ) , (12)
̂ = 𝑃pv 𝑀(𝜂) and the second- 2
derivative of 𝑓 is estimated by 𝐺
order derivative of 𝑓 is estimated by H ̂ = 𝑃pv 𝑁(𝜂). It is 3 1 −𝑞2 1 −4𝑞2
𝑊1 (𝑞) = − 𝑒 + 𝑒 .
̂ directly when H ̂ is 8 2 8
typically difficult to derive Γ = 1/H
̇
close to zero. Instead, a dynamic estimator Ξ = −ℎΞ + ℎH ̂
is employed to approximate Γ. 3.2. Controller Implementation. In this subsection, we pro-
We rewrite the proposed control algorithm in state space vide a detailed flow chart of the implementation of the
as proposed Newton-based extremum seeking MPPT method,
as shown in Figure 6. The basic idea of the controller imple-
mentation is that the controller measures the output voltage
𝑈pv = 𝑔 (𝜏) , and current and then computes the output power. Then the
first derivative and second derivative of the power map are
̂pv ) ,
𝜏 = ℓ (𝑈pv , 𝑈 estimated, respectively, based on the product of output power
(10) and stochastic signals. If the termination criterion is satisfied,
̂̇ pv = −𝑘Γ𝑀 (𝜂) 𝑃pv ,
𝑈 go to the next iteration. If not, regulate the duty cycle and
output voltage to update the output power and go to the next
Γ̇ = ℎΓ − ℎΓ2 𝑁 (𝜂) 𝑃pv , iteration.
International Journal of Photoenergy 5

Upv
Ppv = f(Upv )

̂
H
DC-DC ×
converter 𝜏 DC-DC ̂ pv
U
driver k ̂
G
+ ̂
−ΓG ×
s
M(𝜂(t))
S(𝜂(t)) Signal N(𝜂(t))
generator Ψ

𝜂(t)

Figure 5: The structure of Newton-based extremum seeking MPPT controller.

Initialization

𝜏
Measure
Upv (k) and Ipv (k)
𝜏
Power

Calculate power
Ppv (k) = Upv (k) · Ipv (k) A B C

Estimate the first derivative of power map

̂(k) = Ppv (k) · M(𝜂(k))
G Upv Upv
Estimate the second derivative of power map
̂ (k) = Ppv (k) · N(𝜂(k))
H
Voltage

Yes Figure 7: Three different regions of an illustrative P-V curve.

̂
G(k) ̂
= 0, H(k) < 0?

No
Compute the estimation voltage
̂ pv (k + 1) = −kΓ(k)M(𝜂(k))Ppv (k)
U in understanding the proposed MPPT algorithm. A more
2 rigorous stability analysis is provided in Section 3.4.
Γ(k + 1) = hΓ(k) − hΓ (k)N(𝜂(k))Ppv (k)
In the I-V characteristics of a PV module, there are three
Update the duty cycle of DC-DC driver
different regions, namely, current source region, MPP region,
𝜏(k) = 𝓁(Upv (k), U ̂ pv (k)) and voltage source region [20]. These three regions corre-
spond to the regions A, B, and C in the P-V curve, as shown
Update the output voltage of DC-DC converter
Upv (k) = g(𝜏(k)) in Figure 7. We will introduce the actions of the MPPT con-
troller in these three regions and explain how the MPPT is
achieved.
The update of control signal 𝜏 in (10) can be further
k← k+1
simplified as

𝜏 (𝑘 + 1) = 𝜏 (𝑘) + Δ𝜏 (𝑘) , (13)

Figure 6: Flow chart of the Newton-based extremum seeking MPPT
method.
where Δ𝜏(𝑘) is the update of the control signal 𝜏 in one itera-
tion.
3.3. A Heuristic Analysis. In this subsection, we provide a In what follows, we will introduce the update of control
heuristic analysis of the Newton-based extremum seeking input 𝜏 and output voltage 𝑈pv in the following three different
MPPT controller. This illustrative analysis will be helpful regions.
6 International Journal of Photoenergy

Region A. In this region, the first derivative of the power map Proof. We assume that the stochastic perturbation 𝜂(𝑡) con-
𝑑𝑃pv /𝑑𝑈pv > 0 and the second derivative 𝑑2 𝑃pv /𝑑𝑈pv 2
< 0. sidered in this paper has invariant distribution 𝜇(𝑑𝑥) =
2 2
Then the Newton optimization algorithm (9) turns to be (1/√𝜋𝑞)𝑒−𝑥 /𝑞 𝑑𝑥. Denote the estimation error 𝜏̃ = 𝜏̂ − 𝜏∗ ,
Γ̃ = Γ − H−1 . Then the error system can be denoted by
−1
𝑑𝑃pv 𝑑2 𝑓(𝑈pv ) 𝑑𝑓 (𝑈pv )
= −( 2
) > 0, (14) 𝜏̃̇ = −𝑘 (Γ̃ + H−1 ) 𝑀 (𝜂) 𝑓 (𝑔 (𝜏∗ + 𝜏̃ + 𝑎 sin (𝜂))) ,
𝑑𝑡 𝑑𝑈pv 𝑑𝑈pv
Γ̃̇ = ℎ (Γ̃ + H−1 ) (17)
which means that the update of the control signal Δ𝜏 > 0.
2
Then the duty cycle 𝜏 in (10) is increasing. The larger 𝜏 will − ℎ(Γ̃ + H−1 ) 𝑁 (𝜂) 𝑓 (𝑔 (𝜏∗ + 𝜏̃ + 𝑎 sin (𝜂))) .
drive the DC-DC converter to produce larger voltage 𝑈pv
and then 𝑈pv will move in the direction of converging to the
∗ For simplicity, a quadratic power map is considered:
desired voltage 𝑈pv .

Region B. At the MPP, the first derivative 𝑑𝑃pv /𝑑𝑈pv = 0 and 𝑓󸀠󸀠 (ℓ (𝜏∗ )) 2 H 2
𝑓 (ℓ (𝜏)) = 𝑓∗ + (𝜏 − 𝜏∗ ) = 𝑓∗ + (𝜏 − 𝜏∗ ) .
the second derivative 𝑑2 𝑃pv /𝑑𝑈pv
2
< 0. Then the Newton opti- 2 2
(18)
mization algorithm (9) is
Then the error system can be simplified as
2 −1
𝑑𝑃pv 𝑑 𝑓(𝑈pv ) 𝑑𝑓 (𝑈pv )
= −( ) = 0, (15) H 2
𝑑𝑡 2
𝑑𝑈pv 𝑑𝑈pv 𝜏̃̇ = −𝑘 (Γ̃ + H−1 ) 𝑀 (𝜂) (𝑓∗ + (̃
𝜏 + 𝑎 sin (𝜂)) ) ,
2

which means that the change of control signal Δ𝜏 = 0. Then Γ̃̇ = ℎ (Γ̃ + H−1 ) (19)
the control input 𝜏 and output voltage 𝑈pv keep stable at the
2 H 2
desired 𝜏∗ and 𝑈pv
∗
, which implies that the MPPT is realized. − ℎ(Γ̃ + H−1 ) 𝑁 (𝜂) (𝑓∗ + (̃
𝜏 + 𝑎 sin (𝜂)) ) .
2
Region C. In this region, the first derivative of the power map
To guarantee the stability of closed-loop system, the
𝑑𝑃pv /𝑑𝑈pv < 0 and the second derivative 𝑑2 𝑃pv /𝑑𝑈pv 2
< 0.
generated stochastic signals 𝑀(𝜂) and 𝑁(𝜂) are chosen as
Then the Newton optimization algorithm (9) turns to be
1
−1 𝑀 (𝜂) = sin (𝜂) ,
𝑑𝑃pv 𝑑2 𝑓(𝑈pv ) 𝑑𝑓 (𝑈pv ) 𝑎𝑊0 (𝑞)
= −( 2
) < 0, (16) (20)
𝑑𝑡 𝑑𝑈pv 𝑑𝑈pv 1
𝑁 (𝜂) = (sin2 (𝜂) − 𝑊0 (𝑞)) ,
𝑎2 𝑊02 (√2𝑞)
which means that the change of control signal Δ𝜏 < 0. Then
the duty cycle 𝜏 in (10) is decreasing. The smaller 𝜏 will where 𝑊02 (√2𝑞) = 2(𝑊1 (𝑞) − 𝑊02 (𝑞)); variables 𝑊0 (𝑞) and
drive the DC-DC converter to produce smaller voltage 𝑈pv 𝑊1 (𝑞) are defined as follows:
and then 𝑈pv will move in the direction of converging to the
∗
desired voltage 𝑈pv . 1 2
𝑊0 (𝑞) = (1 − 𝑒−𝑞 ) ,
2
(21)
3.4. Stability Analysis. Stability is a basic requirement of a 3 1 2 1 2

practical control system. In this subsection, we analyze the 𝑊1 (𝑞) = − 𝑒−𝑞 + 𝑒−4𝑞 .
8 2 8
stability of the proposed MPPT algorithm using stochastic
averaging theory. Then the main theoretical contribution of Then we obtain the average system
this paper is concluded as follows.
𝜏̃̇ = −𝑘 (̃
ave
Theorem 2. Consider the photovoltaic MPPT control system 𝜏ave − Γ̃ave H̃
𝜏ave ) ,
shown in Figure 3 with DC-DC driver (4), DC-DC converter (22)
Γ̃̇ = −ℎ (Γ̃ave − ℎ(Γ̃ave ) H) .
ave 2
(5), and unknown power map (3). Assume that the output
∗
power 𝑃𝑝V reaches its maximum 𝑃𝑝V at the desired control input
∗ ∗
𝜏 and desired voltage 𝑈𝑝V . Then under the control law (10) for Thus, according to the stochastic averaging theory [14],
the update of 𝜏, the stability of the closed-loop system is guaran- system (22) has a locally exponentially stable equilibrium at
teed. Moreover, the output power 𝑃𝑝V exponentially converges 𝜏ave , Γ̃ave ) = (0, 0). When 𝜏̃ → 0, we have 𝜏 → 𝜏∗ and then
(̃
∗ ∗
to the maximum power 𝑃𝑝V , which implies that the maximum 𝑃pv → 𝑃pv . Thus the maximum power point tracking is
power point tracking is achieved. realized. This completes the proof.
International Journal of Photoenergy 7

Remark 3. From the equilibrium equation (22), we can find

that the convergence of closed-loop system is totally deter-
mined by parameters 𝑘 and ℎ, which implies that, in the

Power
design of control algorithm (10), the convergence rate can
be arbitrarily assigned by the designer with an appropriate
choice of 𝑘 and ℎ. Generally speaking, relatively large 𝑘 and ℎ
will lead to a good convergence rate, but too large 𝑘 and ℎ will 0
also result in oscillations during the convergence. Hence, 𝑘 Voltage
and ℎ should be chosen by fully considering both the dynamic
and the steady responses. Global MPP
Local MPPs

3.5. Further Discussions. Besides the stability property dis- Figure 8: An illustrative P-V curve for PV systems under partial
cussed above, there are some more issues that should be shading.
considered in the implementation of MPPT controllers, such
as tracking efficiency, effects of partial shading, and PV
5
module ageing effects. In what follows, we will discuss these
topics that may be helpful in the application of the proposed 4
MPPT method.

Current (A)
3
(1) Tracking Efficiency. Tracking efficiency has been the most
2
important consideration of the MPPT method in many
commercial applications. The tracking efficiency of a MPPT 1
algorithm can be calculated by the following equation [6]:
𝑡 0 20 40 60 80 100 120 140 160 180
∫0 𝑃pv (𝑡) 𝑑𝑡
𝜂𝑇 = 𝑡 , (23) Voltage (V)
∫0 𝑃max (𝑡) 𝑑𝑡 I-V characteristics at BOL
I-V characteristics 4 years later
where 𝑃pv (𝑡) is the measured power produced by the PV
array under the control of MPPT algorithm and 𝑃max (𝑡) is the Figure 9: Degradation of 20 a-Si modules after 4 years of operation
theoretical maximum power that the array can produce. in [17].
The discrete definition of the tracking efficiency can be
denoted by [21]
seeking control designs have been available; see [23–25] for
1 𝑛 𝑃pv (𝑘) more information.
𝜂𝑇 = ∑ , (24)
𝑛 𝑘=0 𝑃max (𝑘)
(3) Effects of PV Module Ageing. Reliability and lifetime of
where 𝑛 is the number of samples. PV modules are key factors to PV system performance and
Tracking efficiency evaluates the overall performance are mainly dominated by the PV module ageing effect [26].
of a MPPT algorithm. In the next section, we will verify In order to separate the ageing effect from the irradiance
the tracking efficiency of the proposed MPPT method with and temperature, it is necessary to evaluate the Beginning
numerical results. of Life (BOL) performance. Then the deviations between the
experimental data and BOL performance explain the ageing
(2) Effects of Partial Shading. The partial shading effect has of the PV system [17].
received much attention recently in the implementation of Figure 9 shows the degradation of 20 a-Si modules after 4
MPPT controller, which is partly due to the rapid develop- years of operation in the literature [17]. We can find that the
ment of building integrated photovoltaics (BIPV) [22]. As the MPP declines slowly during the 4-year time. Mathematically,
BIPV is commonly installed on the rooftop, it typically suffers the maximum power point tracking for an ageing PV system
from partial shading due to the space limitation, clouds, and is essentially an optimization process for a slowly varying
so forth. unknown cost function [13]. The proposed Newton-based
When a PV system is subjected to partial shading, the P- extremum seeking MPPT controller optimizes the power
V curve often exhibits a global extremum and several local map with real-time measurements and calls for no knowledge
extremums, as shown in Figure 8. Hence, in order to track of model information. Thus it can deal with the PV module
the global MPP of power map, the adopted MPPT algorithm ageing effect well.
should have a global convergence. However, from (22) we
find that the proposed controller is locally stable at the MPP, 4. Tracking Performance Evaluation
which implies that it may get trapped at local extremums
and may not be a good choice for PV systems under partial In this section, we provide several simulation and experiment
shading. Fortunately, several global or semiglobal extremum results to show the effectiveness of the proposed MPPT
8 International Journal of Photoenergy

Table 1: Data sheet of KC65T at 800 W/m2 , 25∘ C.

Maximum power (𝑃max ) 47.7 W
Multimeters Maximum power voltage (𝑈mpp ) 15.5 V
Maximum power current (𝐼mpp ) 3.08 A
Auxiliary power supply (APS) Upper computer
DC-DC
Light source converter
S7-200 PLC Table 2: Data sheet of KC65T at 1000 W/m2 , 25∘ C.
Light source
control box Maximum power (𝑃max ) 65.3 W
PV panel
Maximum power voltage (𝑈mpp ) 17.4 V
Maximum power current (𝐼mpp ) 3.75 A

Table 3: Parameter setting of the proposed MPPT controller.

Parameter Value
Figure 10: The experiment setup of the light source-driven photo- 𝑎 0.1
voltaic MPPT system.
𝑘 1
ℎ 0.08
𝑞 40
algorithm. In order to evaluate the tracking performance
of the proposed MPPT algorithm under arbitrary desired
irradiance, we adopt the light source-driven photovoltaic module is 47.7 W. Then the simulated tracking efficiency of
MPPT system, as shown in Figure 10. the proposed MPPT algorithm is about 96.6%.
In Figure 10, there are two incandescents working as the Figure 12 shows the comparison between the proposed
irradiation source. The KC65T-PV panel converts the solar Newton-based extremum seeking MPPT method and clas-
energy to electrical energy via the photovoltaic effect. The sical extremum seeking MPPT method. We compare the
Siemens S7-200 PLC works as the MPPT controller, driving tracking performance of the two MPPT methods in terms
the DC-DC converter and regulating the output voltage of of tracking efficiency and convergence time. The numerical
the converter. The upper computer is used to collect and comparison of the two MPPT methods is shown in Table 4.
store experimental data and multimeters are used to display In addition to the simulations, experiments on the light
the measurements. The light source control box is used to source-driven photovoltaic MPPT system are conducted for
regulate the irradiation intensity of light sources. the purpose of verification. The irradiance is set to be
In what follows, we will provide simulation examples and 800 W/m2 and the temperature is about 23∼27∘ C. The PV
experiment results to illustrate the tracking performance of panel comprises four KC65T PV modules, with two modules
the proposed MPPT algorithm under uniform irradiance and connected in parallel and two in series.
nonuniform irradiance cases, respectively. The data sheets The experiment results of the proposed MPPT algorithm
of KC65T photovoltaic module under different irradiance are shown in Figures 13 and 14. From Figure 13(a), we can
conditions are shown in Tables 1 and 2; see [18] for detailed see that the output power converges to the MPP within 6 s.
information. The average power at steady state is about 178 W, whereas the
maximum power is 190.8 W; then the tracking efficiency is
about 93.2%.
4.1. Tracking under Uniform Irradiance. In this subsection,
we consider the tracking performance evaluation of the
proposed MPPT algorithm under the uniform irradiance. A 4.2. Tracking under Nonuniform Irradiance. In this subsec-
simulation and an experiment are provided, respectively, to tion, we will consider the tracking performance of the pro-
illustrate the implementation of the algorithm. The simula- posed MPPT algorithm under nonuniform irradiance with
tion is conducted in MATLAB/Simulink and the parameter an abrupt change from 800 W/m2 to 1000 W/m2 at 10 s. For
setting of the proposed control algorithm is shown in Table 3. the purpose of comparison, we also provide the experiment
For simplicity, we adopt only one KC65T PV module in the results of the classical gradient-based extremum seeking
simulation. MPPT method under the same experimental conditions.
Figure 11 shows the simulation results of the proposed Figure 15 shows the experiment results of the proposed
MPPT algorithm under uniform irradiance of 800 W/m2 at Newton-based extremum seeking MPPT method under the
25∘ C. The convergence of the maximum power point is shown nonuniform irradiance. It is shown in Figure 15(a) that the
in Figure 11(a). The output power oscillates in the direction new MPP is located within 2 s. The new steady power is
of MPP which is detected by the product of the power and about 243 W, while the maximum power is 261.2 W; then the
perturbations. We can see that the output power converges tracking efficiency is about 93%.
to the MPP within 5 s. The output power in the steady state Figure 16 shows the experiment results of the classical
is about 46.1 W, whereas the maximum power of the PV gradient-based extremum seeking MPPT method under the
International Journal of Photoenergy 9

50 18
16
40 14
40 12 18
30
30 10
P (W)

U (V)
17
20 8
20
6 16
10
10 4
0 15
0 0.2 0.4 0.6 0.8 1 2 0 0.2 0.4 0.6 0.8 1
0 0
0 5 10 15 20 0 5 10 15 20
t (s) t (s)
(a) Power (b) Voltage
3.5 1

3 1
0.9
2.5 3 0.9

2 0.8 0.8
2
I (A)

𝜏 0.7
1.5 0.7
1
0.6
1 0 0.2 0.4 0.6 0.8 1
0 0.6
0.5 0 0.2 0.4 0.6 0.8 1

0 0.5
0 5 10 15 20 0 5 10 15 20
t (s) t (s)
(c) Current (d) Duty cycle

Figure 11: Simulation results of the proposed MPPT method under uniform irradiance of 800 W/m2 at 25∘ C.

Table 4: The simulated comparison of the two methods. nonuniform irradiance. It is shown in Figure 16(a) that the
MPPT method Efficiency Convergence new MPP is located within 4 s. The new steady power is
Newton-based ES MPPT 96.6% <4 s about 231 W, while the maximum power is 261.2 W; then the
Gradient-based ES MPPT 92.1% >6 s tracking efficiency is about 88.4%.
In order to illustrate the effectiveness of the proposed
MPPT method more intuitively, we provide the numerical
50 comparison between the proposed algorithm and classical
extremum seeking MPPT algorithm under different irradi-
40
ance. Figures 17 and 18 show the comparison of the two meth-
ods under irradiance 800 W/m2 and irradiance 1000 W/m2 ,
30
respectively. We can see that the proposed method shows its
superiority in both tracking efficiency and convergence rate.
P (W)

20
5. Concluding Remark
10
In this paper, we propose a new Newton-based extremum
seeking MPPT algorithm for photovoltaic systems. The pro-
0
0 5 10 15 20 posed algorithm benefits the following advantages: it calls
t (s) for no knowledge of the power map; it has a good tolerance
of stochastic perturbations and its convergence rate can be
Newton-based ES MPPT method
Gradient-based ES MPPT method
totally designer-assignable. The stability and convergence of
the proposed MPPT controller are rigorously proved with
Figure 12: The comparison between the proposed MPPT method stochastic averaging theory. Some topics related to applica-
and classical extremum seeking MPPT method. tions are discussed, such as partial shading effects and PV
10 International Journal of Photoenergy

200 7
180
6
160
140 5

Current (A)
Power (W)

120 4
100
80 3

60 2
40
1
20
0 0

0
0.94
1.88
2.82
3.76
4.7
5.64
6.58
7.52
8.46
9.4
10.34
11.28
12.22
13.16
14.1
15.04
15.98
16.92
17.86
18.8
19.74
0
0.98
1.96
2.94
3.92
4.9
5.88
6.86
7.84
8.82
9.8
10.78
11.76
12.74
13.72
14.7
15.68
16.66
17.64
18.62
19.6
Time (s) Time (s)
(a) Power (b) Current
Figure 13: Experiment results of output power and current collected in the upper computer.

30 V 4V
3V
2s
20 ms

(a) Voltage (b) PWM signal

Figure 14: Experiment results of output voltage and PWM signal measured in the oscilloscope.

300 9
8
250
7
200 6
Current (A)
Power (W)

5
150
4
100 3
2
50
1
0 0
0
0.98
1.96
2.94
3.92
4.9
5.88
6.86
7.84
8.82
9.8
10.78
11.76
12.74
13.72
14.7
15.68
16.66
17.64
18.62
19.6

0
0.94
1.88
2.82
3.76
4.7
5.64
6.58
7.52
8.46
9.4
10.34
11.28
12.22
13.16
14.1
15.04
15.98
16.92
17.86
18.8
19.74

Time (s) Time (s)

(a) Power (b) Current

4V
33.2 V
3V 30 V
2s 20 ms

(c) Voltage (d) PWM signal

Figure 15: Tracking performance of the Newton-based extremum seeking MPPT algorithm under nonuniform irradiance.
International Journal of Photoenergy 11

300 9
8
250
7
200 6

Current (A)
Power (W)

5
150
4
100 3
2
50
1
0 0

0
0.94
1.88
2.82
3.76
4.7
5.64
6.58
7.52
8.46
9.4
10.34
11.28
12.22
13.16
14.1
15.04
15.98
16.92
17.86
18.8
19.74
0
0.98
1.96
2.94
3.92
4.9
5.88
6.86
7.84
8.82
9.8
10.78
11.76
12.74
13.72
14.7
15.68
16.66
17.64
18.62
19.6
Time (s) Time (s)
(a) Power (b) Current

4V
32.4 V
29.5 V
3V
2s 20 ms

(c) Voltage (d) PWM signal

Figure 16: Tracking performance of the gradient-based extremum seeking MPPT algorithm under nonuniform irradiance.

94 12

93 10

92
8
91
(%)

(s)

6
90
4
89

88 2

87 0
Newton-based ES MPPT Gradient-based ES MPPT Newton-based ES MPPT Gradient-based ES MPPT

Tracking efficiency Convergence time

(a) Tracking efficiency (b) Convergence time

Figure 17: The numerical comparison between the proposed MPPT algorithm and classical extremum seeking MPPT algorithm under
irradiance 800 W/m2 .

module ageing effects. Simulation and experiment results are extremum seeking, which requires that the power map of
provided to show the effectiveness of the proposed method. photovoltaic systems is rather slower than extremum seeking.
It is also worth mentioning that both the Newton-based When irradiance changes quickly, the tracking performance
ES and gradient-based ES belong to the so-called static of ES MPPT algorithms will be worse. In order to propose
12 International Journal of Photoenergy

94 4

93 3.5

92 3

91 2.5
(%)

90

(s)
2

89 1.5

88 1

87 0.5

86 0
Newton-based ES MPPT Gradient-based ES MPPT Newton-based ES MPPT Gradient-based ES MPPT

Tracking efficiency Convergence time

(a) Tracking efficiency (b) Convergence time

Figure 18: The numerical comparison between the proposed MPPT algorithm and classical extremum seeking MPPT algorithm under
irradiance 1000 W/m2 .

a commercial ES MPPT algorithm, we will further investi- [6] D. P. Hohm and M. E. Ropp, “Comparative study of maximum
gate the dynamic extremum seeking and its applications in power point tracking algorithms,” Progress in Photovoltaics:
maximum power point tracking in future work. Research and Applications, vol. 11, no. 1, pp. 47–62, 2003.
[7] T. Kawamura, K. Harada, Y. Ishihara et al., “Analysis of MPPT
characteristics in photovoltaic power system,” Solar Energy
Conflict of Interests Materials and Solar Cells, vol. 47, no. 1–4, pp. 155–165, 1997.
[8] C. Hua, J. Lin, and C. Shen, “Implementation of a DSP-
The authors declare that there is no conflict of interests controlled photovoltaic system with peak power tracking,” IEEE
regarding the publication of this paper. Transactions on Industrial Electronics, vol. 45, no. 1, pp. 99–107,
1998.
[9] R. Leyva, C. Alonso, I. Queinnec, A. Cid-Pastor, D. Lagrange,
Acknowledgments and L. Martı́nez-Salamero, “MPPT of photovoltaic systems
The authors would like to thank the anonymous reviewers using extremum—seeking control,” IEEE Transactions on
Aerospace and Electronic Systems, vol. 42, no. 1, pp. 249–258,
for their valuable comments. This work is supported by the
2006.
National Natural Science Foundation of China (Grant nos.
[10] S. L. Brunton, C. W. Rowley, S. R. Kulkarni, and C. Clarkson,
61071096 and 61379111), the Fundamental Research Funds
“Maximum power point tracking for photovoltaic optimization
for the Central Universities of Central South University, and
using ripple-based extremum seeking control,” IEEE Transac-
Hunan Provincial Innovation Foundation for Postgraduate. tions on Power Electronics, vol. 25, no. 10, pp. 2531–2540, 2010.
[11] H. Yau and C. Wu, “Comparison of extremum-seeking control
References techniques for maximum power point tracking in photovoltaic
systems,” Energies, vol. 4, no. 12, pp. 2180–2195, 2011.
[1] E. F. Camacho, M. Berenguel, F. R. Rubio, and D. Martinez, [12] A. Ghaffari, M. Krstić, and S. Seshagiri, “Extremum seeking for
Control of Solar Energy Systems, Springer, 2012. wind and solar energy applications,” ASME Dynamic Systems &
Control Magazine, pp. 13–21, 2014.
[2] E. F. Camacho, M. Berenguel, and F. R. Rubio, Advanced Control
of Solar Plants, Springer, 1997. [13] D. Dochain, M. Perrier, and M. Guay, “Extremum seeking
control and its application to process and reaction systems: a
[3] T. Esram and P. L. Chapman, “Comparison of photovoltaic array survey,” Mathematics and Computers in Simulation, vol. 82, no.
maximum power point tracking techniques,” IEEE Transactions 3, pp. 369–380, 2011.
on Energy Conversion, vol. 22, no. 2, pp. 439–449, 2007. [14] S. Liu and M. Krstic, “Stochastic averaging in continuous time
[4] V. Salas, E. Olı́as, A. Barrado, and A. Lázaro, “Review of the and its applications to extremum seeking,” IEEE Transactions on
maximum power point tracking algorithms for stand-alone Automatic Control, vol. 55, no. 10, pp. 2235–2250, 2010.
photovoltaic systems,” Solar Energy Materials and Solar Cells, [15] S. Liu and M. Krstic, “Newton-based stochastic extremum
vol. 90, no. 11, pp. 1555–1578, 2006. seeking,” Automatica, vol. 50, no. 3, pp. 952–961, 2014.
[5] K. H. Hussein, I. Muta, T. Hoshino, and M. Osakada, “Max- [16] http://www.greenrhinoenergy.com/solar/.
imum photovoltaic power tracking: an algorithm for rapidly [17] A. Abete, F. Scapino, F. Spertino, and R. Tommasini, “Ageing
changing atmospheric conditions,” IEE Proceedings: Generation, effect on the performance of a-Si photovoltaic modules in
Transmission and Distribution, vol. 142, no. 1, pp. 59–64, 1995. a grid connected system: experimental data and simulation
International Journal of Photoenergy 13

results,” in Proceedings of the Conference Record of the 28th IEEE

Photovoltaic Specialists Conference, pp. 1587–1590, Anchorage,
Alaska, USA, 2000.
[18] http://www.kyocerasolar.com/assets/001/5170.pdf.
[19] S. Liu and M. Krstic, Stochastic Averaging and Stochastic
Extremum Seeking, Springer, London, UK, 2012.
[20] P. Bhatnagar and R. K. Nema, “Maximum power point tracking
control techniques: state-of-the-art in photovoltaic applica-
tions,” Renewable and Sustainable Energy Reviews, vol. 23, pp.
224–241, 2013.
[21] A. Reza Reisi, M. Hassan Moradi, and S. Jamasb, “Classification
and comparison of maximum power point tracking techniques
for photovoltaic system: a review,” Renewable and Sustainable
Energy Reviews, vol. 19, pp. 433–443, 2013.
[22] K. Ishaque and Z. Salam, “A review of maximum power point
tracking techniques of PV system for uniform insolation and
partial shading condition,” Renewable & Sustainable Energy
Reviews, vol. 19, pp. 475–488, 2013.
[23] Y. Tan, D. Nešić, I. M. Y. Mareels, and A. Astolfi, “On global
extremum seeking in the presence of local extrema,” Automat-
ica, vol. 45, no. 1, pp. 245–251, 2009.
[24] F. Esmaeilzadeh Azar, M. Perrier, and B. Srinivasan, “A global
optimization method based on multi-unit extremum-seeking
for scalar nonlinear systems,” Computers and Chemical Engi-
neering, vol. 35, no. 3, pp. 456–463, 2011.
[25] W. H. Moase and C. Manzie, “Semi-global stability analysis
of observer-based extremum-seeking for Hammerstein plants,”
IEEE Transactions on Automatic Control, vol. 57, no. 7, pp. 1685–
1695, 2012.
[26] R. Khatri, S. Agarwal, I. Saha, S. K. Singh, and B. Kumar,
“Study on long term reliability of photo-voltaic modules and
analysis of power degradation using accelerated aging tests and
electroluminescence technique,” Energy Procedia, vol. 8, pp.
396–401, 2011.
 Photoenergy
International Journal of
International Journal of Organic Chemistry International Journal of Advances in
Medicinal Chemistry
Hindawi Publishing Corporation
International
Hindawi Publishing Corporation Hindawi Publishing Corporation
Analytical Chemistry
Hindawi Publishing Corporation
Physical Chemistry
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

International Journal of

Carbohydrate Journal of
Chemistry
Hindawi Publishing Corporation
Quantum Chemistry
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Submit your manuscripts at

http://www.hindawi.com

Journal of
The Scientific Analytical Methods
World Journal
Hindawi Publishing Corporation
in Chemistry
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of International Journal of International Journal of Journal of Bioinorganic Chemistry

Spectroscopy
Hindawi Publishing Corporation
Inorganic Chemistry
Hindawi Publishing Corporation
Electrochemistry
Hindawi Publishing Corporation
Applied Chemistry
Hindawi Publishing Corporation
and Applications
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of Chromatography Journal of Journal of International Journal of

Theoretical Chemistry
Hindawi Publishing Corporation
Research International
Hindawi Publishing Corporation
Catalysts
Hindawi Publishing Corporation
Chemistry
Hindawi Publishing Corporation
Spectroscopy
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014