Fuzzy Logic Based Control Scheme for Power Optimization of ... - IJECT

22 downloads 1567 Views 395KB Size Report
potential candidates for remote and off-grid power generation in standalone or hybrid applications [1]. This paper proposes a control strategy to optimize power ...
IJECT Vol. 2, Issue 2, June 2011

ISSN : 2230-7109(Online) | ISSN : 2230-9543(Print)

Fuzzy Logic Based Control Scheme for Power Optimization of a Small Wind Turbine System with DC-DC Converter 1 1,2

Shakil Ahamed Khan, 2Md. Ismail Hossain, 3Mohammad Jakir Hossain

Dept. of EEE, Rajshahi University of Engineering & Technology, Bangladesh Dept. of EEE, Dhaka University of Engineering & Technology, Bangladesh

3

Abstract This paper proposes a control strategy to optimize power output performance of wind energy conversion system. In order to obtain optimum power output from a wind turbine generator system, it is necessary to drive the wind turbine at an optimal rotor speed for a particular wind speed. Fuzzy logic based control algorithm is implemented with the embedded microcontroller which will track the maximum power point (MPP) by generating appropriate generator load references. The designed controller then forces the system to operate towards the dictated load reference. This paper discusses the low cost implementation of fuzzy logic based MPP algorithm in a 8-bit microcontroller using the tools and techniques to generate optimized real time code in C which will demonstrate how maximum power point tracker might provide elegant and efficient solution for increasing the efficiency of wind energy conversion systems, based on experimental results rather than on mathematical models. The proposed concept is verified and results are presented.

mechanical sensors, polynomial to approximate the wind turbine power coefficient, statistical model to predict the wind speed, lookup table based mapping, searching methods. These methods increases the cost, reduce the reliability, complex and time consuming calculation, significant memory space, long searching time respectively to locate the maximum power point [3]. This paper proposes calculation of appropriate generator load references to obtain maximum power by measuring frequency of generator using frequency sensor. The designed controller then forces the system to operate towards the dictated load reference. The proposed system includes a PMSG, a PWM controlled DC-DC converter and sensor circuits as shown in Fig. 3. The designed controller regulates the converter output load according to MPP load reference by varying duty cycle of the PWM signal applied to Boost DC-DC converter. Fuzzy logic is implemented on MPPT algorithm for swift maximum power point tracking to avoid energy loss [4].

Keywords Wind energy systems, MPPT control, fuzzy logic, microcontroller, DC/DC boost converter

II. Aerodynamic characteristics of the rotor A wind turbine is characterized by its power-speed characteristics. The power extracted from a wind turbine is a function of the wind power available, the power curve of the machine and the ability of the machine to react to wind variations. The power extracted from the wind can be expressed as given by (1) [5].

I. Introduction As energy sources are depleting due to the energy needs of the world, alternative energy sources are being sought after. The search for clean and low cost energy alternatives has yielded wind energy as an excellent candidate against conventional fossil fuel based power generation. Large wind turbine based grid connected wind farms have proven to be a commercial success in Europe and North America. Small wind energy conversion systems (WECS) are potential candidates for remote and off-grid power generation in standalone or hybrid applications [1]. This paper proposes a control strategy to optimize power output performance of a permanent-magnet synchronous-generator (PMSG) under variable wind speed, fixed-pitch wind turbines without the aerodynamic controls. In order to obtain optimum power output from a wind turbine generator system, it is necessary to drive the wind turbine at an optimal rotor speed for a particular wind speed. Wind rotor performance is fixed for fixed-pitch variable-speed wind turbine. So, the restoring torque needs to be adjusted to maintain optimum rotor speed at a particular wind speed to obtain maximum aerodynamic power output. Since, the restoring torque is proportional to generator load current (TG=f(IG)), So, the function of maximum power point tracker is to provide the required load on the generator [2].

(1) Where, ρ is the air density, A is the swept area of the turbine and Vw is the wind velocity. The Cp parameter is called the power coefficient. The power coefficient is the ratio of the mechanical power at the turbine shaft to the power available in the wind, given as a function of tip speed- ratio λ as given in (2). λ

ωm R Vw =

(2) Where R= radius of blades, ωm= angular velocity. Power coefficient CP versus tip speed ratio relationship of the wind turbine is shown in Fig. 1. The maximum power coefficient Cpmax corresponds to the optimal tip speed ratio λopt. Hence for a given wind speed the available rotor power can be plotted verses the alternator frequency, this is done for variable wind speed, as shown in Fig. 2. In order to extract peak power, the rotor must be held at its optimal tip speed ratio. Clearly the turbine speed should be changed with wind speed so that optimum tip speed ratio is maintained.

To implement maximum wind power extraction, most controller designs of the variable speed wind turbine generators employ

18  International Journal of Electronics & Communication Technology

w w w. i j e c t. o r g

IJECT Vol. 2, Issue 2, June 2011

ISSN : 2230-7109(Online) | ISSN : 2230-9543(Print)

system, FLC looks very promising for this application. The inputs to a MPPT fuzzy logic controller are usually an error E and a change of error ∆E as given in (4) and (5) respectively.

C oefficient of performa nce vs T ip s peed ra tio

0. 5

Coefficient of performance-Cp

Coefficient of performance (Cp)

0. 45 0. 4 0. 35 0. 3

Rectifier

0. 2

Gear Box ωm ω g CP PM T T m g

0. 15 0. 1

Wind

0. 05 0

0

5

10

PW

15

Rotor blade

(3) III. Maximum power point tracking The generator torque is controlled with the optimum torque curve as shown in Fig. 4 according to the generator frequency by varying the effective electric load on the generator. We assume that the load power is determined by the curve marked ‘sp, as given in Fig. 2 and that the wind speed is 8.4m/s. The turbine will be operating at point ‘s’. If the wind speed increases to 10.4m/s, the turbine torque exceeds the load torque and the turbine accelerates toward point ‘p’. But this point is not maximum power point for this wind speed. If the load on the generator is forced to increase, the load power will increase, causing the turbine to slow down. The new operating point would then be point ‘m’. To reach maximum power point for this wind speed, the accelerating/de-accelerating torque is calculated from the difference between the turbine mechanical torque and the torque given by the optimum curve as shown in Fig. 4. Hence the effective load is calculated and operated according to this load. Finally, the generator will reach the point “m” where the accelerating torque is zero. So (1-.6=0.4pu) power will be more extracted using MPPT. A similar situation occurs if the wind velocity decreases. 2 6 m/s 8. 4 m/s 10. 8 m/s 12 m/s 13. 2 m/s

1. 6

Rotor power(pu)'

1. 4

Optimum load line

1. 2

m

1

b

0. 8 0. 6

a

0. 4

s d

p

c C

0. 2 0

50

100

150

G enera tor frequency(H z)

Frequency Sensor

IV. Fuzzy logic implementation Fuzzy logic controllers have the advantages of working with imprecise inputs, not needed an accurate mathematical model, handling nonlinearity [6]. For controlling such a complicated w w w. i j e c t. o r g

C

M

DC/AC

Gate Isolation Circuit

Voltage Sensor Fuzzy Logic Control

PWM Signal

1. DC Loads 2. Battery Storage

Microcontroller

Fig. 3: Control structure of a PMSG-based stand-alone variable speed wind turbine The inputs to a MPPT fuzzy logic controller are usually an error E and a change of error ∆E as given in (4) and (5) respectively.

E (n) = I ref − I d

(4)

∆E (n) = E (n) − E (n − 1)

(5)

0. 9 0. 8 0. 7 0. 6 0. 5

Tg

0. 4 0. 3 0. 2 0. 1 0

f 0

50

100

150

G enera tor frequency(H z)

Fig. 4: Generator torque reference versus frequency Where Iref is calculated by measuring the rectifier output voltage Vd, (i.e. Iref =Tg×ω/ Vd). Reference torque Tg is calculated by using optimum power curve as shown in Fig. 4. E and ∆E are calculated and converted to the linguistic variables during fuzzification. Linguistic variables are non-precise variables that often convey a surprising amount of information. Fig. 5 shows the relations between measured error and the linguistic term, such as positive small, positive medium and positive big. At some point the error is positive small and at some point the error is positive big the space between positive big and positive small indicates an error that is, to some degree, a bit of both. The horizontal axis in the following graph shows the measured or crisp value of error. The vertical axis describes the degree to which a linguistic variable fits with the crisp measured data. μ

Fig. 2: Wind turbine maximum available rotor power (optimal load curve) and rotor power curve for changing wind speed.

VL

DM

PMSG

ADC Module

Optimum Torque Optimum torqueReference reference(pu)(pu)

Since rotor angular velocity is proportional to the alternator frequency, at optimal tip speed ratio, the maximum available rotor power also varies as the cube of the alternator frequency for the same rotor-generator combination. The maximum power from a wind turbine can be written as (3).

1. 8

L VS

PG CF

Fig. 1: Wind rotor characteristics

IL

IS

Current Sensor

T ip s pe e d ratio

0

Boost Converter

0. 25

NB

NS

ZE

PS

PB

1

-1.0

-0.8 -0.6 -0.4 -0.2

0

0. 2

0. 4

0.6

0.8

1.0

E(n)

Fig. 5: The relationship between linguistics variable and error To add the linguistics variable positive big to a computer

  International Journal of Electronics & Communication Technology  19

IJECT Vol. 2, Issue 2, June 2011

ISSN : 2230-7109(Online) | ISSN : 2230-9543(Print)

program running in an embedded microcontroller, translation the graphical representation into meaningful code is needed. The following C code fragment gives one example of how to do this. The function error- Positive Big ( ) returns a degree of membership, scaled between 0 and 1, indicating the degree to which a given error can be positive big. This type of simple calculation is the first tool required for calculations of fuzzy logic operations. char Error_ Positive Big (int CRISP) { if (CRISP < 0.4) return(0); else { if (CRISP >= 0.4) return(CRISP * 2.5); else { return(1); }} }

NS NS NS ZE ZE

NS NS ZE PS PS

ZE ZE PS PS PS

8

80

60

6

4

Searching Method

0

10

20

30

40

50

60

70

80

90

Wind Speed

Power (W) (W) Power Power (W)

10

2

100

Time (s ) Time (s)

Fig. 8: Simulated performance of the wind turbine

4

n =1

Fuzzy

Wind Speed

100

0

ZE PS PS PB PB

∑ Y [i] × F [i] ∑ Y [i]

Fig. 7: System performance wind speed step change

20

4

n =1

Generator Speed

40

The final step in the fuzzy logic controller is to combine the fuzzy output into a crisp systems output. The result of the defuzzification has to be a numeric value which determines the change of duty cycle of the PWM signal used to drive the MOSFET. There are various methods to calculate the crisp output of the system. Centre of Gravity (COG) method is used in our application due to better results it gives. The COG for our application is expressed mathematically as given in (6).

∆D =

Fuzzy control output surface using Matlab/Simulink simulation is shown in Fig. 6 which represents the change of duty cycle for corresponding error and change of error.

Speed Reference

TABLE 1 : FUZZY RULE BASE TABLE E NB NS ZE PS PB ∆E NB NB NS NS ZE

(7)

Fig. 6: Fuzzy control output surface

Rule evaluation is done by using an algorithm where loops compare the antecedent value depending on the rule being evaluated in a repeated fashion until all rules are evaluated. The fuzzy logic controller output is typically a change in duty ratio ∆D of the power converter. The linguistic variables assigned to ∆D for the different combinations of E and ∆E as shown in Table 1 which is based on a boost converter.

NB NS ZE PS PB

Dnew = Dold + ∆D

(6)

Where Y[i] is the ith members of the output vector and F[i] are the multiplying coefficients of the output membership function as shown in Table 1, and ∆D is the change of duty cycle, and this number represents a signed number which is added or subtracted from the present duty cycle to generate the next system response for reaching the MPP as given by (7).

20  International Journal of Electronics & Communication Technology

V. Result The designed controller is capable of rapidly locking into the maximum power point under variable wind speed. The model of the system along with FLC based MPPT algorithm as shown in Fig. 9 is established using Matlab/Simulink and microcontroller based hardware implementation. The system response to step changes in wind speeds is simulated to examine the response of the MPPT controller. As shown in Fig. 7, the step change in wind speed the system has a good dynamic response. Hardware implementation of fuzzy based gate signal control for changing input is shown in Fig. 10. With the given wind speed data, energy output over 100s is simulated. Results show that, system with Fuzzy Logic controller performs better than searching method. Simulated performance is shown in Fig. 8. VI. Conclusion The use of fuzzy logic in gate signals control in MPPT is tackled, analyzed, and implemented in this paper where Matlab/ Simulink simulation and experimental results are described. The results show how well this controller eliminates the complexity and maximizes the power extracted from the wind turbine and delivered to the load.

w w w. i j e c t. o r g

IJECT Vol. 2, Issue 2, June 2011

ISSN : 2230-7109(Online) | ISSN : 2230-9543(Print) Input Membership Functions

Id

Calculate E, ΔE

Fuzzification Calculate

Iref=T*ω/V

Rule List

Rule Evalution

Output Membership Functions Defuzzi -fication

Calculate

T=Kopt *ω2

Data Base

Update D

Plant Process

Get ω(n), V(n), I d

Fig. 9: Fuzzy MPPT firmware algorithm

(a)

(b) Fig. 10 Photos of MPPT controlled gate signal, (a) Output gate signal of MPPT controller (b) Tuned fuzzy MPPT controlled gate signal for changed input. References [1] M. J. Khan, M. T. Iqbal, “Dynamic Modeling, Simulation and Control of a Small Wind Energy Conversion System,” Proceedings of the International Conference on Mechanical Engineering, Dhaka, Bangladesh, 26- 28 December 2003 [2] E. Muljadi, K. Pierce, P. , “Soft-control control for variablespeed stall-regulated wind turbines”, Wind Engineering, Vol.85, pp277-291, Migliore, 2000 [3] Wei Qiao, Wei Zhou, José M. Aller, Ronald G. Harley, “Wind Speed Estimation Based Sensorless Output Maximization Control for a Wind Turbine Driving a DFIG,” IEEE Trans. Power Electron, vol.23, no.3, pp. 1156-1169, May 2008 [4] Simoes, M.G. Bose, B.K. Spiegel, R.J., “Design and performance evaluation of a fuzzy-logic-based variablespeed wind generation system” IEEE Trans. On Industry Applications, Vol. 33, no. 4, pp. 956 – 965, July-Aug. 1997. [5] D Le. Gourieres, “Wind Power Plants Theory and Design”, Oxford Pergamon press, 1982. [6] Bor-Sen Chen, Chung-Shi Tseng, Huey-Jian Uang, “Robustness Design of Nonlinear Dynamic Systems via Fuzzy Linear Control,” IEEE Fuzzy Systems, vol. 7, no. 5, pp. 571-585, Oct. 1999.

w w w. i j e c t. o r g

Shakil Ahamed Khan received his B.Sc Engineering Degree in Electrical and Electronic Engineering from Rajshahi University of Engineering and Technology, Bangladesh in 2009. He is now working as Lecturer in EETE Department in Dhaka International University, Bangladesh; His research intertest include Smart Grid, Power Electronics, Embedded Systems, Industrial Automation and Controls. Ismail Hossain received his B.Sc Engineering Degree in Electrical and Electronic Engineering from Rajshahi University of Engineering and Technology, Bangladesh in 2009. He is now working as Lecturer in EEE Dept. in International Islamic University Chittagong, Bangladesh; His research intertest include Smart Grid, Industrial Power Electronics, Embedded system, Power system operation and protection and efficient utilization of renewable energy source. Mohammad Jakir Hossain received his B.Sc Engineering Degree in Electrical and Electronic Engineering from Dhaka University of Engineering and Technology, Bangladesh. He is now working as Assistant Professor in EEE Dept. in Dhaka University of Engineering and Technology (DUET) Gazipur, Bangladesh.

  International Journal of Electronics & Communication Technology  21