As shown in Fig. 3.1 and Fig. 3.2, both the photovoltaic voltage and the photovoltaic current at the MPP can represent the MPP. For a particular operating condition, the control of MPPT normally regulates either the voltage or the current to a certain value that represents the local MPP.
However, the mapping between MPP and these variables is time variant, as it is a function of changing irradiation and temperature. Ideally, however, this relationship is constant or changes slowly within a range. The photovoltaic voltage is a preferable control variable when it comes to the maximum power point tracking. The advantages of the voltage-based tracking are described in the following paragraph.
Photovoltaic Voltage, VPV(V) Photovoltaic Current, IPV(A)
8
6
4
2
40 20
0
1 kW/m2
0.75 kW/m2 0.5 kW/m2
Different Irradiation Levles
Large Variation
Small Variation
Fig. 3.1. Photovoltaic current vs. photovoltaic voltage under different irradiation levels.
22
8
6
4
2
40 20
0
50°C
85°C 15°C
Different Ambient Temperatures
Photovoltaic Voltage, VPV(V) Photovoltaic Current, IPV(A)
Fig. 3.2. Photovoltaic current vs. photovoltaic voltage under different ambient temperatures.
Changing radiation causes the photovoltaic current to vary dramatically, as illustrated in Fig. 3.1. The fast dynamic of insolation is usually caused by the cover of mixed rapidly moving clouds. If the photovoltaic current is used as the set point, the MPPT requires a fast dynamic to follow a wide operating range from 0A to the short-circuit current because the current is heavily dependent on weather conditions. In contrast, the changing insolation only slightly affects the voltage of MPP.
Fig. 3.2 shows that the cell temperature is the dominant factor varying the voltage of MPP when the temperature changes. However, cell temperature has a slow dynamic and is always within a certain range.
Unlike the current of MPP, the photovoltaic voltage of MPP is usually bounded to 70%~82% of the open-circuit voltage. This gives the tracking range a lower bound and an upper limit. When regulation of the photovoltaic voltage is implemented, the MPPT can quickly decide the initial point according to the percentage of the open-circuit voltage.
The photovoltaic current value at MPP is close to around 86% of the short-circuit current. Because the photovoltaic current dramatically varies
23
with insolation, the transient response of MPPT can occasionally cause the photovoltaic current to reach its saturation point, which is the short-circuit current. This shall be prevented because its nonlinearity causes a sudden voltage drop and results in power losses. However, for the regulation of photovoltaic voltage, the voltage saturations can be easily avoided because a controller knows the operating range is bounded around 70%~82% of the open-circuit voltage. Furthermore, a good-quality measurement of voltage signal is cheaper and easier than that of current measurement. Considering all the benefits of the voltage-based tracking aforementioned, as a result, the voltage-based tracking algorithm is preferred than the current-based tracking.
The voltage-based tracking technique, aside from the advantages of robustness and low solar irradiation dependence, holds also the advantage of fast tracking speed and low implementation complexity.
According to the relationship between the maximum power point voltage (VMPP) and the open-circuit voltage (VOC), the maximum power point tracker can take the advantage of this attribute to tack the MPP with high tracking speed. As illustrated in Fig. 3.3, the operating point can quickly jump to the point VMPP(approx) near the MPP if the open-circuit voltage is known. The tracker easily calculates the voltage of MPP from the information of open-circuit voltage. This can be achieved simply by dividing or multiplying the open-circuit voltage.
24 0
MPP
VMPP(approx)
PV Voltage, VPV(V) PV Power, PPV(W)
MPP’
Irradiation Reduction
Voc Voc’
70%~82%
Fig. 3.3. The concept of voltage-based MPPT technique.
When the weather condition changes, say the irradiation reduces, the tracker has to disconnect the solar array and the power converter to create an open circuit on the system. This way, the new open-circuit voltage (VOC’), as indicated in Fig. 3.3, is detected. After getting the information of the new open-circuit voltage, the tracker can further calculate the new maximum power point (MPP’) and therefore regulate the operating point to be around this point. All in all, the voltage-based tracking algorithm benefits from its low design complexity, only a divider or a multiplier is needed to calculate the maximum power point. The fast tracking speed is also guaranteed in this technique, since there is no complicated calculation and iteration.
The voltage-based tracking algorithm, on the other hand, has some undesirable drawbacks as well. For example, the low power efficiency and the poor tracking accuracy.
The working principle of the voltage-based tracking technique is based on both the open-circuit voltage and the pre-determined percentage between maximum power point voltage and the open-circuit voltage. In order to constantly track the maximum power point, the open-circuit voltage of the
25
solar array needs to be updated periodically. Therefore, one has to disconnect the connection between the solar array and the power converter so that the real time open-circuit voltage can be recorded. In this way, the power delivery path between the solar array and the output stage is interrupted. It causes power loses as a result. Namely, the power delivery is discontinuous on the PV system and low power efficiency is presented, which is obviously undesirable for a PV system.
The second obstacle of the voltage-based tracking technique toward its perfection is the low tracking accuracy. As mentioned before in Chapter 2, the voltage-based tracking technique depends on the information of open-circuit voltage and the pre-determined or pre-measured coefficient kV. This coefficient, however, depends greatly on the manufacture material, the sunlight intensity, the environmental temperature and other factors. These factors are either time-varying or temperature-varying. As a result, it makes the determination and the measurement of the coefficient extremely difficult. Even though one can determine a desirable coefficient for a system, the environmental parameters, temperature, insolation, etc., vary along with the time, this makes the pre-determined coefficient inaccurate.
Therefore, a high tracking accuracy cannot be guaranteed when the voltage-based tracking technique is used.
3.2 The Pros and Cons of Perturbation and Observation Tracking Algorithm
The perturbation and observation tracking technique takes the advantage of calculating the slope of the power-voltage characteristic curve
26
to determine the maximum power point. Once the slope condition is determined, the perturbation direction can be consequently decided. It forces the system to operate toward the maximum power point. Based on this tracking idea, the tracking accuracy or the tracking effectiveness is guaranteed even under a varying environmental condition. The varying environmental condition changes the power-voltage characteristic curve and undoubtedly alters the maximum power point. To find out the new maximum power point, the perturbation and observation technique recalculates the slope of the new characteristic curve and determines the next perturbation direction. In this way, the changing atmospheric factors do not affect the tracking functionality. Instead, a high tracking accuracy can be further ensured with smaller perturbation steps.
The downsides of the perturbation and observation tracking technique, however, are the oscillation problem, low tracking speed, and the demanding design complexity.
Since the working principle of the perturbation and observation algorithm is to change the operating point step by step, the system can never operate exactly at the maximum power point. Instead, the operating point jumps back and forth around the actual maximum power point. This, therefore, causes an oscillation problem when the system reaches near the maximum power point. The power delivery loss and fluctuation, as a result, present on the PV system.
Before concluding the next perturbation direction, the controller has to deliberately calculate the slope of the characteristic curve at that time instant. This slows down the entire tracking process, compared to the voltage-based tracking algorithm. However, a faster tracking speed using
27
perturbation and observation method can still be achieved if the variable perturbation step method is used. As shown in Fig. 3.4, the size of perturbation step is not fixed. Variable step size is used to accelerate the tracking process. The farer the operating point away from the maximum power point, the larger the perturbation step size is. When the operating point is near the maximum power point, a smaller step size is used to guarantee an accurate tracking. Nevertheless, this may require a more complicated MPPT controller in the PV system.
Output Voltage, VPV(V)
Output Power, PPV(W) Pmax
Variable Step Sizes
Fig. 3.4. The Perturbation and Observation Tracking Algorithm with Variable Perturbation Step Sizes.
The design complexity of the perturbation and observation tracking technique is higher than the voltage-based method. The MPPT controller needs to calculate the slope at any time instant, which means it requires a lot of arithmetic circuits and the circuits to determine the next perturbation direction. On the other side, the voltage-based tracking algorithm requires only a few circuits to find out the maximum power point from the open-circuit voltage.
28
To summarize, TABLE 3.1 gives the comparisons between the voltage-based tracking algorithm and the perturbation and observation tracking algorithm in terms of the tracking effectiveness, the tracking speed and the implementation complexity.
TABLE 3.1 Comparisons between Voltage-Base Tracking and Perturbation and Observation Tracking
Voltage-Based Tracking Perturbation and Observation Tracking
Tracking Accuracy Worse Better
Tracking Speed Faster Slower
Implementation Complexity Easier More complicated
3.3 The Proposed Tracking Algorithm
Fig. 3.5 demonstrates the concept of the proposed MPPT algorithm.
Before the solar system turns on, the operating point of the PV system is located on the open-circuit voltage point, VOC. Conventional MPPT algorithm, such as slope detection algorithm [31] [e.g.
perturbation/observation (P&O) algorithm or hill-climbing (HC) algorithm]
calculates the slope of characteristic power-voltage curve to determine the slope condition and then to track the maximum power point (MPP).
Nevertheless, as mentioned before, a PV system adopting this algorithm requires a lengthy amount of time to track the operating point from points VOC to MPP during the system power-on period, as depicted in Fig. 3.5.
29
0
MPPConventional Tracking Algorithm
(1) OCT
0.7×Voc
Output Voltage, V
PV(V) O u tp u t P o w e r, P
PV(W )
(2) SDT
(3) SDT
MPP’
Irradiation reduces
Voc
Fig. 3.5. The concept of the proposed MPPT algorithm when irradiation reduces.
Other tracking algorithms such as the constant voltage algorithm use a fixed ratio of maximum power voltage to open-circuit voltage VOC to approximate the MPP. Theoretically, 0.7 fractions of open-circuit voltage VOC is close to the MPP [32]. Therefore, periodically disconnecting the solar array and power stage to measure VOC and multiplying it to 0.7 can rapidly detect the current MPP, as described before. The fraction factor (0.7) varies when different solar cell materials are used. Moreover, it is considerably susceptible to environmental conditions such as ambient temperature and solar irradiation level. In this sense, the MPP cannot be guaranteed when varying environmental conditions are taken into consideration. Moreover, consistent disconnection between the solar array and power stage causes power delivery interruption during the sampling period, thereby resulting in the low power efficiency of the PV system.
To increase both tracking speed and accuracy while maintaining high power efficiency, the open-circuit tracking (OCT) algorithm and the slope detection tracking (SDT) algorithm are both adopted to track the MPP in this study. Disconnection between solar array and the power stage occurs
30
only one time; that is, in the beginning of the system power-on period. This way, unnecessary power loss can be avoided while maintaining the system power efficiency.
Fig. 3.6 illustrates the timing diagram of the proposed tracking algorithm. Eslope is a digital signal used to indicate the slope condition.
Logic-high Eslope means the slope condition, dPPV/dVPV, of the solar panel is positive. On the contrary, logic-low Eslope means dPPV/dVPV is negative. The signal OCT Enable indicates whether the OCT tracking algorithm is enabled or not. The signal VG is the gate signal of the power NMOS in the boost converter. The signal IL1 shows the inductor current of the boost converter, which can also indicate the current of the solar panel.
VG
(1) Open Circuit Tracking Period (2) Slope Detection Tracking Period
Eslope
(3) Slope Detection Tracking Period
Fig. 3.6. The timing diagram of the proposed MPP tracking algorithm.
Basically, the tracking procedure can be divided into the following sequences.
31
The first step is: Before the solar system turns on, VOC is detected by the controller to set the PV panel voltage, VPV, close to 0.7×VOC for improving the tracking speed compared to the disadvantage of slow tracking speed in the conventional P&O and HC algorithms. Besides, the switching duty cycle of the boost converter is set to its maximum value in order to accelerate the tracking speed during the open-circuit voltage detection period.
The second step is: the SDT technique takes over the tracking procedure to continually and accurately track the MPP to make sure that the power stage receives the most energy from the solar panel.
The third step is: If the environmental condition changes, say, the irradiation level reduces, the slope condition changes from positive to negative, according to the solar cell characteristic shown in Fig. 3.1. Then, Eslope transits from high to low. Aforementioned in Section 2.1, the voltage is inversely proportional to the current. As a result, the SDT technique will increase the switching duty cycle. The current of solar array, therefore, increases to pull down the operating voltage and ensures the system operating move to a new maximum power point, MPP’, as illustrated in Fig.
3.5.
On the other hand, if the environmental temperature reduces, as shown in Fig. 3.7, the new maximum power point, MPP’, is higher than the old one, MPP. The PV system can still locate the new maximum power point through the proposed algorithm and then work on this new operating point.
32 0
MPP
(1) OCT
0.7×Voc
Output Voltage, VPV(V) Output Power, PPV(W)
(2) SDT
(3) SDT MPP’
Voc
Temperature reduces
Fig. 3.7. The concept of the proposed MPPT algorithm when temperature reduces.
The flowchart in Fig. 3.8 summarizes the overall tracking topology of the proposed MPP tracking algorithm. After the solar system turns on, the OCT technique is enabled until the operating voltage is set to be around 0.7×VOC. After that, the SDT technique consistently monitors the slope condition and makes sure the system operate at MPP regardless of any condition change in the environment. When the signal Eslope is set to be logic high, this may mean the irradiation level increases or the temperature reduces, the controller will decrease the switching duty cycle to the boost converter in order to pull down the inductor current. Meanwhile, the PV current is reduced as well and the PV voltage is increased conversely, which forces the system to operate toward the MPP. On the other hand, if the signal Eslope is detected to be logic low, the inductor current and the PV current is raised by the increased duty ratio. Therefore, the PV voltage is pulled down to track the MPP.
33
Fig. 3.8. The flowchart of the proposed MPP tracking algorithm.
The proposed AMPPT technique includes both the advantages of the OCT technique and the SDT technique. That is, the OCT technique can rapidly but roughly locate the MPP with fast tracking speed. In the meanwhile, the SDT technique improves the tracking accuracy which cannot be guaranteed by the OCT technique.
3.4 Shading Effect and Global MPPT Algorithm
For a large scale photovoltaic system, unavoidable shadow from the nearby trees, cloud and buildings frequently cause the energy degradation of the solar array. The so called “partial shading effect” [33] poses a great threat to highly efficient utilization of the PV system. As depicted in Fig.
34
maximum power point. As a result, the designed MPP tracker may easily misjudge the optimal operating point and therefore the system cannot provide its maximum energy. The proposed MPP tracking algorithm in this paper can be further improved by the algorithm provided in Fig. 3.10 to ensure the robustness of the PV system.
Output Power, PPV(W)
Output Voltage, VPV(V) 0
Uniform Irradiation
Non-Uniform Irradiation
Fig. 3.9. P-V Characteristic Curve due to non-uniform irradiation level.
Fig. 3.10. Flow chart of the global maximum power point tracking.
LMPP Tracking
LMPP Reached?
Update Pmpp_last & Vmpp_last
Timer/Incidence
Global Maximum Power Point Tracking No
No
35
The system primarily determines whether it has reached its local maximum power point (LMPP). The measured power (Pmpp_last) and voltage (Vmpp_last) are then stored for later comparison. The control loop enters the global maximum power point tracking (GMPPT) stage when triggered by the timer. The default timer is set to 1 s. That is, the global maximum power point (GMPP) tracker is enabled every second to check whether the current operating point belongs to the GMPP. If the current operating point is already the maximum point on the power-voltage plane, the system will continuously operate at this point. Otherwise, the GMPP tracker locates the GMPP and forces the system to operate on the located point. During the GMPPT stage, the system is perturbed by a voltage difference ΔVperturb, which is approximated to 60%–70% of Voc [34], as sketched in Fig. 3.11.
Thus, an MPPT technique guarantees full robustness of the connected PV system.
Output Power, PPV(W)
Output Voltage, VPV(V) 0
LMPP
GMPP ΔVperturb
MPP Tracking
Fig. 3.11. P-V characteristic curve sketching the tracking of the global maximum power point due to non-uniform irradiation level.
36
Chapter 4
The Proposed Circuit of the MPPT Controller
4.1. Maximum Power Point Tracker Interface Topologies
This chapter provides a comparative study with a goal of choosing the suitable converter topology for the application of maximum power point tracking in a PV system. Both the DC-DC buck converter and the DC-DC boost converter are usually used in the PV power system because of their simplicity and efficiency. The following paragraphs analyze both of these converter topologies and draw a conclusion of choosing the appropriate one for the maximum power point tracker in a PV system.
4.1.1. The Basic Working Principles of Buck Converter
Fig. 4.1 shows a schematic of a conventional DC-DC buck converter connected to a solar array. It comprises of a solar array as the power source, input and output capacitor, C1 and C2, respectively. A switch S and a diode D are also presented in the figure. Rout is the summarized loading of the next stage, which is the DC-AC inverter in the proposed PV application.
Under the condition of continuous current mode, the operation of buck converter can be divided into two phases. In phase 1, as shown in Fig. 4.2, the switch S is turned on, which means it creates a short-circuit path
37
between the input node Vpv and the output node Vbus through the switch S and the inductor L. The energy from the solar array is then charging the inductor L and providing the energy for the output node Vbus. The timing diagram is depicted in Fig. 4.4. The symbol d represents the duty cycle to the switch. Ts is the switching period of the converter and VC2 is the voltage across the output capacitor C2, which equals to the output voltage Vbus. During this phase, the inductor current is increasing with the time and the
between the input node Vpv and the output node Vbus through the switch S and the inductor L. The energy from the solar array is then charging the inductor L and providing the energy for the output node Vbus. The timing diagram is depicted in Fig. 4.4. The symbol d represents the duty cycle to the switch. Ts is the switching period of the converter and VC2 is the voltage across the output capacitor C2, which equals to the output voltage Vbus. During this phase, the inductor current is increasing with the time and the