Abstract-- Logic based power management system for

automatic switching of lights’ banks is presented in this work. This system is developed based on a generated power from a wind turbine, a profile of the banks’ states provided by the user, and a sequence of priorities. The generated power is controlled to track the reference power profile by a pitch angle control strategy. In order to maintain a consistent brilliance of the lights and reduce the activity of the pitch actuator, a modified PI controller is presented to deal with fluctuation problems. Experimental results are provided to show the proposed power management system for a pitch controlled wind turbine.

Index Terms-- automatic switching, electronic load, pitch angle control, power management, wind turbine.

I. INTRODUCTION IND energy is an intermittent energy source that keeps changing under the effect of wind speed without taking

into consideration the load demand [1, 2]. Therefore, it is very important to develop a management system for the consumption of the generated energy from wind and the efficient use of the load if wind energy is the only source of energy. Such system can provide an adequate amount of energy based on the load demand, which can be requested manually by a user, and the availability of energy from the wind turbine. In case of high energy production, the load demand can be completely satisfied and partially satisfied based on the user request, which has the priority in making the decision by the power management system. In case of low energy production, the system can automatically disable some parts of the load by priority for an efficient use of energy.

The power management system changes the amount of the requested and provided power during its operation. Therefore, this power must be regulated to track a power profile requested by the user. Pitch angle controller can adjust the pitch setting at its optimum value to achieve power tracking [3-5]. However, the regulation of power through the pitch angle can cause high activity of the pitch actuator due to small

This work was supported in part by the Faculty of Graduate Studies at Saint Mary’s University under FGSR Grant.

A. Merabet, R. Keeble and R. Vigneshwaran are with Division of Engineering, Saint Mary’s University, Halifax, NS, B3H 3C3 Canada (e-mail: adel.merabet@smu.ca, raquel.keeble@smu.ca, vigneshwaran.rajasekaran@smu.ca).

R. Beguenane is with the Department of Electrical Engineering, Royal Military College, Kingston, ON, K7K 7B4, Canada (e-mail: rachid.beguenane@rmc.ca).

H. Ibrahim is with TechnoCentre éolien, QC, G4X 1G2, Canada (email: hibrahim@eolien.qc.ca)

fluctuations in the power during steady state operation, which can affect the flow of power to the load.

In this paper, a power management system, based on Boolean logic, is developed to power LEDs lights in six banks arranged to look like windows in buildings. This system is designed for an automatic switching of the banks depending on the request of the user, availability of power from wind and banks priorities. For power regulation purpose, PI pitch angle control is used and a modified one is proposed to reduce fluctuation in the power and maintain a consistent brilliance of the lights.

II. DESCRIPTION OF THE SYSTEM 1. Wind turbine workstation

The wind turbine experiment workstation has three major components: wind turbine, generator, and load.

The wind turbine, installed in a tunnel, has five (05) blades and is connected to a DC generator through a gearbox of ratio 1:1 to change the vertical rotation to a horizontal one as the generator is mounted vertically. The generator is directly connected to an electronic load of a lighting system (circuit board with LED) [6]. A variable speed blower, based DC motor, is used to emulate experimentally different wind speed profiles by controlling the speed of the machine. The wind turbine experiment system, with major components, is presented in Figure 1.

Fig.1 Wind turbine experiment workstation

Logic Based Power Management System for Pitch Controlled Wind Turbine

A. Merabet, R. Keeble, R. Vigneshwaran, R. Beguenane and H. Ibrahim

W

2012 IEEE Electrical Power and Energy Conference

978-1-4673-2080-1/12/$31.00 ©2012 IEEE 280

2. Electronic load The wind turbine powers the lights in a city sky (circuit

board with LED’s). The electronic load consists of 6 parallel equal banks of LEDs in series with a same value resistance. Each load bank consists of 20 LED’s arranged to look like windows in buildings. The load banks can be switched ON and OFF manually, using switches on the front panel, or they can be switched automatically from computer. Figure 2 shows load banks arranged as windows in buildings with some banks illuminated.

Fig. 2 Six load banks with LED windows

III. POWER MANAGEMENT SYSTEM FOR AUTOMATIC SWITCHING OF THE BANKS

1. Development of the Logic based Power Management System

The design of the logic for the priorities of the automatic switching of the banks is based on the following rules: 1. The number of banks turned ON will be determined by the

power generated from the wind turbine. 2. A bank is considered to be ON if it is illuminated without

flickering and with minor power fluctuation. 3. Based on the amount of generated power, the banks are

ranked by priority from left to right in the load (Bank 1, Bank 2, …, Bank 6). However, the priorities can be assigned to different banks by switching wires into the multiplexer.

4. The user can manually turn ON/OFF any bank (i.e. If Bank 3 is manually turned OFF, less power should be required to turn Bank 4 ON than if Bank 3 is also ON). In this case, the priority of automatic switching will be based on the captured power from wind and the required (regulated) power through pitch angle control. The Boolean expressions of the priorities of the automatic

switching of the banks are given as follows: P1 = AQ (Given that the manual switch of bank 1 is ON and the generated power equals or exceeds 0.1W, first priority bank is illuminated)

P2 = BR (A) + BQ (A')

P3 = CS (AB) + CR (AB' + A'B) + CQ (A'B')

P4 = DT (ABC) + DS (A'BC + AB'C + ABC') + DR (A'B'C + A'BC' + AB'C') + DQ (A'B'C')

P5 = EU (ABCD) + ET (A'BCD + AB'CD + ABC'D + ABCD') + ES (A'B'CD + A'BC'D+ A'BCD'+ AB'C'D + AB'CD' + ABC'D') + ER (A'B'C'D + A'B'CD' + A'BC'D' + AB'C'D') + EQ (A'B'C'D')

P6 = FV (ABCDE) + FU (A'BCDE + AB'CDE + ABC'DE + ABCD'E + ABCDE') + FT (A'B'CDE+ A'BC'DE + A'BCD'E + A'BCDE' + AB'C'DE + AB'CD'E + AB'CDE' + ABC'D'E + ABC'DE' + ABCD'E') + FS (A'B'C'DE + A'B'CD'E + A'B'CDE' + A'BC'D'E + A'BC'DE' + A'BCD'E' + AB'C'D'E + AB'CD'E' + AB'C'DE' + ABC'D'E') + FR (A'B'C'D'E + A'B'C'DE' + A'B'CD'E' + A'BC'D'E' + AB'C'D'E') + FQ (A'B'C'D'E')

where, Pn (n=1:6) is the priority of bank n. A, B, C, D, E, F: conditions of the manual switches for the banks (1…6) respectively. Q, R, S, T, U, V: conditions of the generated power (Pgen)

Q R S T U V Pgen ≥ 0.1 ≥ 0.2 ≥ 0.3 ≥ 0.4 ≥ 0.5 ≥ 0.6

Table1. Conditions of the generated power

Table of generated power conditions has been built based on experimentation and observation. Depending on the manual switching of the banks, it is found that for each bank a 0.1 W power is required to illuminate the bank’s LEDs and remain bright enough to be considered ON.

2. Error-Dependent Thresholds for Automatic Bank Switches

The measured power fluctuates around the reference power because of various factors such as noise measurement in sensors and the continuous activity of the pitch actuator. Therefore, an error-dependent threshold for automatic bank switches must be added to the logic of automatic switching. The appropriate threshold Pt is the lower end of acceptable power for the number of manually switched ON banks. The added logic subtracts the programmed allowable error from the original threshold as

Pt = Pref – ε (2) where, Pref is the reference power from the number of banks ON and ε is the error tolerance.

The acceptable power range P for the number of banks ON is given by

P = Pref ± ε (3)

IV. PITCH ANGLE FLUCTUATION CONTROLLER FOR POWER REGULATION

The reference power profile is generated from the banks’ switches manually turned ON by the user or the captured power from wind. Therefore, the power must be regulated to track the reference power. In this wind turbine, adjustment of the pitch angle of the blades is the only way to regulate the power required for each bank or a combination of banks and is done by the pitch servo-motor as shown in Figure 1.

In a basic control strategy, the controlled power is compared with its reference and the error signal is sent to a PI controller to produce the optimum pitch angle. Using such

281

controller, the pitch angle is continuously changing, which can damage the pitch actuator as it is in high activity.

The power management system is developed for step changes in the profile of the reference power based on banks’ switching states and the available power. Therefore, when the generated power is fluctuating around the reference one within a certain error, the pitch angle can be considered constant in order to reduce fluctuations in power and the activity of the pitch actuator.

A pitch angle fluctuation controller is developed based on the above idea, where the pitch angle is kept constant provided that the generated power is tracking the reference power within certain programmable allowable error (error tolerance ε). The controller works in two stages: an error assessment followed by a pitch angle generation. 1. The error assessment component works by determining if

the instantaneous error is within the allowable error band (± ε), as set by the user. If the error is allowable, the generated power can be estimated to be equal to the reference power. If the error exceeds the allowable error at any point while the system is running, the PI controller will regain control of the pitch angle to find the optimum. Simultaneously, a programmable pulse generator emits a time-based “true = 1” pulse. The pulse is essential because it acts as a reset to allow the opportunity for the pitch angle to be held constant.

2. The pitch angle generator interprets the signal from the error assessment and either emits the real-time signal from the PI controller or the previous “optimum” pitch angle. The optimum pitch angle was one that when held constant allowed the production of power that was within the allowable error of the reference power.

V. EXPERIMENTAL RESULTS Experimentation is carried out to validate the performance

of the proposed power management system and the pitch angle fluctuation controller. Algorithms and I/O data acquisition are implemented in Matlab/Simulink/QUARC. QUARC generates real-time code directly from Simulink model and runs the generated code in real-time on the Windows target - all on the same PC. The Data Acquisition Board seamlessly interfaces with Simulink using Hardware-in-the-loop (HIL) blocks provided in the QUARC Targets Library.

The choice of proportional gain Kp and integral gain Ki of the PI controller is determined by trial and error method, so that the controller is tuned to achieve good performance. The

wind speed pattern is artificially generated by the blower motor.

Initially, one bank must be turned ON in order to connect the generator to the load. Also, the banks can be manually turned ON or OFF before running the system. Then, the automatic switching will change the states of the banks depending on the generated power from the wind turbine and the bank’s priority.

The reference power is chosen by the user by manually switching ON some banks and, based on the available power from wind, the pitch angle controller adjust the pitch angle in order the generated power tracks the reference one as shown in Fig. 3. The banks are automatically switching ON/OF based on this power as shown in Fig. 4, where 1≡ ON and 0≡ OFF. The red dashed line presents the banks’ switching states following the reference power and black line represents the automatic switching based on the actual power. It can be observed that around time 120 s there is not enough power to turn Bank 5 ON as the generated power did not reach the reference power. Then, the power fluctuates in/out the error band, which makes Bank 5 oscillates ON/OFF during this period.

For the test of the pitch angle fluctuation, the value of the error tolerance is 0.03W and the pulse period is 4 seconds. Theoretically, reducing the values of these two parameters will improve the behavior of the controller. However, the values’ choice depends on the performance of the PI controller in transient operation at the change of the step. Figure 5 presents the power tracking under the effect of a standard PI controller. It can be seen that the power is fluctuating around the reference and the pitch angle is continuously changing. However, the power error is within the band error, which allows the automatic switching to follow the requested manual one by the user as shown in Fig. 6. In Fig. 7, under the effect of the proposed controller, less fluctuation in the power is observed during steady state operation compared to the result of the standard controller, which provides a consistent brilliance of the lights. However, some side effects occur in the automatic switching (Fig. 8) as some banks turn OFF during a short time period of about 1~2 seconds due to undesirable changes in the power out of the error band and wind speed variation.

Fig. 3 Reference power by the user and generated power

282

Fig. 4 Automatic switching of the banks for the reference power profile

Fig. 5 Generated power and pitch angle under standard PI controller

Fig. 6 Generated power and pitch angle under standard PI controller

283

Fig. 7 Generated power and pitch angle under PI fluctuation controller

Fig. 8 Generated power and pitch angle under PI fluctuation controller

VI. CONCLUSIONS Logic based power management system for automatic

switching of lights’ banks of windows in buildings has been proposed in this work. It has been developed based on a generated power from the wind turbine, the profile of the banks’ states provided by the user, and the sequence of priorities. The generated power has been controlled to track the reference power profile by a PI pitch angle control strategy. Then, a modified PI pitch angle control has been proposed to maintain a consistent brilliance of the lights and reduce the activity of the pitch actuator.

VII. APPENDIX

Wind turbine specifications: Blade Radius = 14cm, ρ=1.14 kg/m3, Number of blades =5, Pitch angle constraint [-1o 15o] DC generator specifications: R=2.47Ω, L = 500μH, Kb=7.05mV/rpm, Ki=0.015 A/mN , J=110 g/cm2.

VIII. REFERENCES [1] S. Keshav and C. Rosenberg, "Direct Adaptive Control of Electricity

Demand", CS Technical Report CS-2010-17, September 2010 [2] C-X Liu, Q-A Zeng and Y. Liu, "A Dynamic Load Control Scheme for

Smart Grid Systems", Eenregry Procedia, 12, pp. 200-205, 2011. [3] E. Muljadi, Pitch-Controlled Variable-speed Wind Turbine Generator,

IEEE Transactions on Industry Applications Vol.37, No.1, 2001. [4] A. Musyafa’, A. Harika, I. M. Y. Negara, I. Roband. Pitch Angle

Control of Variable Low Rated Speed Wind Turbine Using Fuzzy Logic Controller. International Journal of Engineering & Technology IJET-IJENS Vol: 10 No: 05, 2010.

[5] Y. El-Tous, "Pitch Angle Control of Variable Speed Wind Turbine", American Journal of Engineering and Applied Sciences 1(2), pp. 118-120, 2008.

[6] "Wind Turbine Experiment", Quick-Start Guide, Quanser inc. 2010.

284