All Rights Reserved © 2016 IJARCET
2515MODIFIED SENSORLESS CONTROL OF
WIND TURBINE WITH PERMANENT
MAGNET SYNCHRONOUS GENERATOR
SYSTEM FOR POWER GENERATION
Alade Tunde A.I, Nwosu A. W,
Olisakwe C.O
Abstract -
This work presents modified frequency orientation control FOC of variable-speed direct-driven permanent magnet synchronous generator (PMSG) in wind energy conversion system (WECS). The objective is to optimize the power capture from the wind utilizing frequency orientation control FOC scheme without wind speed sensors. The model is developed with time framed meteorology forecast. the Mean wind speed forecast are substituted as starting wind speed (sensed value). The Simulation results are presented for various changes in wind speed at 480m above sea level using matlab powsys. The ability of the system to operate at the optimum coefficient of performance without the need for a speed sensor is subsequently analyzed, Mmaku in Awgu local government council of Enugu Nigeria is use as our test case.Index terms: Frequency Orientation Control FOC, Mmaku, Metrological Forecast, Permanent Magnet Synchronous Generator (PMSG)
I. INTRODUCTION
Energy is an essential input for economic growth; the lack of electricity restricts the development of all aspects of a normal modern life style. Most developing nations are experiencing hampered economic and industrial growth primarily on account of inadequate and inefficient electricity generation and supply systems. The African Development Bank (AfDB) estimated Nigeria electricity supply par capita from the national grid to be less than
0.8kwh and below 50% access in 2009 from
151.3kWh/capita and 67% access in 2000 (Afican development bank, 2008). This is against the estimated par capita of 1416.87KwH and 95% access to public power supply based on country with similarity in terms of education and life style (Olayande, 2010)
On the contrary renewable energy sources have the advantages that it is widely distributed, abundant, clean, and becoming increasingly economical. Among various types of renewable energy sources, wind energy is one of the fastest growing renewable energy sources (T. Burton,
2011). The wind industry set a new record of 44%
increase in the annual market about 51.473 GW (GWEC,
2014). The global cumulative wind power capacity has been explosively increased from about 6.1 GW in 1996 to 369.5 GW at the end of 2014, and the growth rate is expected to continue in the coming years.
Wind turbine is fast becoming viable alternative to gas and nuclear power plants, majorly on its environmental friendliness and limited infrastructural requirement. Despite of the numerous successes of wind energy (renewable energy) has played in meeting the increasing demand for clean electricity in the developed nation (Bakwa, 2002) (GWEC, 2014), its significance is yet to be fully appreciated in most developing nation such as Nigeria due to installation, operational and maintenance complexities considered tasking for most developing nations (GWEC, 2014)
Reduction of wind turbines overhead installation
All Rights Reserved © 2016 IJARCET
2516overhead mounted sensor removal without compromise to the overall system reliability is key.
II. WEC OVERVIEW
DD PMSGs have the capability to generate power in a wide speed range. Figure 2 is a basic block diagram of a WEC system. The variation in voltage are smoothen by rectification and inversion. However effective conversion requires coordination and constant feedback of the wind speed between MPPT and the generator side converter for manipulation of frequency and voltage to keep various control mechanism optimum. The expression of the mechanical power captured by the wind turbine is:
1
Where Cp (λ, β) is the wind turbine power coefficient
which is a function of λ and β, ρ is the air density, r is the radius of wind turbine blade,Vis the wind speed, β is the blade pitch angle, and λ is the tip speed ratio:
λ = 2
Substituting 2 into 1 expression for maximum power Pmax
is;
3
The only controllable term is the power coefficient. As shown in figure 1 (the mechanical power versus rotational speed output characteristic graph of wind turbine for different wind speeds), the dotted line shows the
maximum power points for different wind turbine rotational speed ω and different wind speedV.
Each P–w curve is characterized by a unique turbine speed corresponding to the maximum power point for that wind velocity (V. Agarwal, 2010). At maximum power points, the gradient P–ω curves CITATION RDa03 \l 1033 (R. Datta and V. T. Ranganathan, 2003). Different power coefficient curves correspond to different blade angles.
For each case, there is an optimal tip speed ratio,,which contributes to a peak power coefficient value which, in turn, leads to a maximum power capture.
III. . MAXIMUM POWER POINT TRACKING Wind turbine generators needed to be controlled to operate in three different modes; park when the wind speed is below the cut-in speed or extremecut-out speed. The third state is when the wind speed is within cut in speed and below cut out limit of the wind generator. In order to maximize the power capture maximum power point tracking (MPPT) controls schemes are implemented to adjust aerodynamics of the blade. The optimal tip speed ratio based maximum power point tracking (MPPT) control are achieve using several control models. However, Field Oriented Control FOC and Direct Torque Controls DTC are two of the most applied methods to PMSG. FOC is preferred due to its added advantages of Independent torque and flux control (Quang, 2008) and its as easily as DC machines. (Blaschke, 1972)
IV. Field Oriented Control FOC
In the FOC approach, the dq-axes are rotating at the rotor electrical angular speed with the dq-axis aligned with the rotor flux direction. Figure 3, is simple cross sectional diagram of SPN diagram., the flux producing current component and the torque producing current component, are along the d-axis and q-axis, respectively figure 4 refers
The torque expression of the PMSM is show below:
[image:2.595.40.292.135.311.2]= 4
[image:2.595.311.547.213.365.2]FIGURE 2: POWER COEFFICIENT
All Rights Reserved © 2016 IJARCET
2517For a surface mounted PM machine (SPM) which is applied in the case study, P = Number of Poles, the d-axis and q-axis inductances are equal (Ld = Lq). Thus, the torque expression can be simplified and rewritten as follows:
= 5
In order to achieve the maximum torque per ampere, the d-axis current is set at zero ( ). In expression(4),is the flux linkage due to the permanent magnets which is a constant. Thus, there will be a linear relationship between the electromagnetic torque and the q-axis current,, such that the electromagnetic torque can be easily controlled by regulating the q-axis current.
From figure 4 the voltage functions of the PMSM in the dq-axes reference frame were expressed as follows:
Vdc = Rsids+ Ld- ωeLqiqs 6
Vqs = Rsiqs + Ld+ ωeLqids+ ωeλr
Transforming the voltage functions into the αβ-axes reference frame, the state space functions of the PMSM can be rewritten as follows:
= - iαs- + - 8
= - iβs- + - 9
Where, iαs andiβs, are the stator currents in the α β-axes
reference frame, and are the stator voltages in the -axes reference frame. Here, and, are the αβ –axes back EMFs which can be expressed as follows (H. Li, Nov.-Dec. 2005.):
= - λrωe 10
= λrωe 11
Where, ωe, is the electrical angular speed of the rotor,
while, λr, is the peak/maximum phase flux linkage due to
the rotor-mounted PMs. As can be seen in equations 10
and 11, the back EMF contains the information of the rotor speed, ωe, as well as the rotor angle, which means that by
measuring the back EMF, we will be able to obtain the information on the rotor position and speed. Knowledge of the rotor flux vector position (angle) at any instant time it is a necessary requirement for field oriented control
(Quang, 2008), (Trzynadlowski, 2000).
Accuracy of recognition of the rotor flux angle is crucial in control, because the calculation of stator current components isd, isq depends directly on this angle
orthogonal (R. Datta and V. T. Ranganathan, 2003). Angular position of the rotor flux vector estimation θrit is
based on the relative difference between synchronous and rotor speed
ωs= ω1– ω
as follow:
θr= = + = + θ 12
Where θ is electrical angular of rotor displacement and ω1
is electrical synchronous speed. The required value of slip speed, can be calculated from dynamic model of the motor under rotor field orientation conditions
ω
s.The estimated [image:3.595.55.261.172.350.2]ω introduces error in iteration and consequently delays
FIGURE 3 THE CROSS-SECTIONAL VIEW OF A THREE-PHASE, 2 -POLE PMSM
V
si
ds=
d-
q-V
ds=
Vqs=λsr
s
[image:3.595.331.508.184.357.2]v i
Figure 5 Phasor diagram of the FOC
[image:3.595.323.527.605.745.2] [image:3.595.60.345.746.817.2]All Rights Reserved © 2016 IJARCET
2518convergence leading to observed chattering in the PMSG. As shown in fig 6
V. MODIFIED SENSORLESS CONTROL A. WIND POWER LAW
The power law wind exponent α is the parameter needed in extrapolating the wind speed above the surface from the available or known height
= 13
Where U is the wind speed at H, U0wind speed at H0
This have been found to vary with: season they are higher in wet season and lower in the dry season, heights and a number of other parameters. Power law wind exponent α increases with increasing height, (Akinlade, 2012). Established that wind speed increases with increasing height but tends towards a constant value at heights above 500m, these values were found to vary along four major
time frame (00hrs, 04hrs, 13hrs and 20hrs) as shown figure 6. its values are higher at nights than at day-time. However, it has its minimum value between 0400 and 0500hrs.
Relatively, its values are higher during the wet months than the dry months. Figure 9 projected variations across the months of the and its average value for the dry months being 0.33 while its average value for the wet months is 0.45. Applying this time based wind speed forecast as the starting wind speed i.e. time slots (00, 04, 13 & 20 hrs) into FOC as starting values. It should facilitate fast convergence of MPPT and possibly reduce or remove chattering effect.
B. SIMULINK MODEL
Basic Algorithm of modification of sensorless control is shown in figure 10 and matlab model developed as shown in fig 5 the MPPT utilizing the fed metrological wind speed forecast Wm1to calculate the PMSG optimum power
generatedPcalfrom 3. This is compared with the measured
generated power Pm. if Pcal > Pm them Wm1is increased
by a predetermine value Δw, if Pm>Pcal them the wind speed is reduce by Δw. The new value wmand feed into
MPPT otherwise the wind speed value is retained.
This iteration continues until the next time slot when new Wm2 projection for the slot. These iterations will be done
for 00hrs or wm0, 04hrs or wm1, 13hrs or wm2and 20hrs or
wm3
[image:4.595.62.281.126.354.2]VI. . SIMULATION AND EXPERIMENTAL SITE Matlab simulation model of this method was developed using Vestas Wind Systems A/S, 2014 turbine. The electrical system is formed by the permanent magnet machine, a passive rectifier, a dc-to-dc converter and a voltage source inverter, or he generator side converters as shown in Fig. 11
[image:4.595.316.531.567.716.2]FIGURE 9MEAN DAILY WIND SPEED PROFILE (WHOLEDATA) (Akinlade, 2012)
FIGURE 10 modified FOC control algorithm Figure 7 Wind speed input for the wind power
generation system PMSG
Figure 8 Actual electrical angular speed of the PMSG
0 0. 1 1. 2 2. 3 3.
4 6 8 1 1 1 1
[image:4.595.60.294.636.783.2]All Rights Reserved © 2016 IJARCET
2519A. Test site
Enugu is situated at 6.44° North latitude, 7.49° East longitude and 248 meters elevation above the sea level. (Most complex maps for all cities in the world, 2015). The State occupies much of the highlands of Awgu, Udi and Nsukka. Mmaku hills in Awgu local government areas are 300 to 1,000 metres (3,300 ft) above sea level are flanked by the rolling lowlands of border towns of Uturu and Okigwe (Egboka, 1985). This offers ideal site for installation of wind turbine due to reduction of in-wind turbulence and higher wind speed at the elevated plains. Therefore, Enugu is chosen as test study site in this work. Plate 1 the area view of Mmaku hills of Agwu LG in Enugu State Nigeria. and 25 years mean wind speed of Mmaku hills at Awgu local government areas in Enugu State Nigeria. The mean wind speed value is within 4.5 to 8m/s. using these mean value to extrapolate the wind speed at 130m height above sea level (Akinlade, 2012) see table 1 and 2.
VII.
Results Analysis
[image:5.595.50.278.170.437.2]The model for the maximum power tracking
algorithm and the dc-to-dc controller are shown in fig.
25. The embedded MATLAB function contains the
mathematical model that represents the maximum
power tracker and developed PMSG control model.
The triggered is set at a frequency of 6 Hz; therefore,
running algorithm 6 times per second. Every time the
model is run the power is calculated from projected
wind speed and the controller reacts to the reference
voltage commanded by the maximum power tracker.
FIGURE 11 SIMULINK CONTROL MODELPlate 1.0 Area view of Mmaku hills Awgu
All Rights Reserved © 2016 IJARCET
2520The generated E.m.f is compared with calculated and
the difference is applied to MPPT in subsequent
iteration. Input wind speed are based on mean of
metrological wind speed of several years for the
sampled area.
VIII. . Conclusions
This work has presented and analyzed several
important control algorithms for the wind turbine
PMSG systems. the simulation results, of the
modified FOC MPPT method shows capability of
controlling the wind turbine PMSG to generate the
maximum power at meteorological wind speeds.
Furthermore, a fast convergence was observed each
time there is a speed change, the generator reacted to
the wind speed change very fast. Thus, for the wind
turbine PMSG systems that require high dynamic
performance and high power capture efficiency, this
control algorithm is proven to be efficient.
2 Future Work
In this work, a simulation study was carried out to
validate the proposed control algorithms for a low
improved speed 2 MW wind turbine PMSG system
and 580m above sea level using metrological data of
Enugu (real site scenario). A real site implementation
of same for the selected area will be required for
further study
References
1. Acharya S. (2013, June 2013 4). Renewable EnergyRevolution. Retrieved may 5th, 2016, from
www.sunilacharya.wordpress.com.
2. Afican development bank. (2008). annual report. afdb.
3. Akinlade, G. a. (2012). INVESTIGATION OF POWER LAW WIND EXPONENT WITHIN THE LOWER BOUNDARY LAYER AT ILE-IFE, NIGERIA. Ife Journal of
Science , vol. 14(no. 2), 325-335.
4. Bertling, J. R. (2005). Survey of failures in wind power systems with focus on Swedish wind power plants during 1997. .22 (no. 1 ).
5. Blaschke, F. (1972). The principle of field orientation as applied to the new TRANSVECTOR closed loop control system for rotating field machines,. Siemens
Rev,, 217–220.
6. Egboka, B. C. (1985). Water resources problems in the Enugu area of Anambra State, Nigeria. WREPU
Department of Geological Anambra State University of Technology, pp. 95, 97 .
7. GWEC. (2014). GLOBAL WIND ENERGY COUNCIL
ANNUAL REPORT. Global statistics. (GLOBAL WIND
ENERGY COUNCIL) . Retrieved oct 10, 2015, from http://www.gwec.net:
wpcontent/uploads/2012/06/Global-Annual-Installed-Wind-Capacity-1997-2014.jpg
8. H. Li, K. S. ( Nov.-Dec. 2005.). Neural-network-based sensorless maximum wind energy capture with compensated power coefficient. IEEE Transactions on
Industry Applications, vol. 41, no. 6, pp. 1548-1556,.
9. Hau, E. (2006). Wind Turbines: Fundamentals, Technologies,Application, Economics, 2nd edition.
Berlin, Germany: Springe.
10. Most complex maps for all cities in the world. (2015, nov 13). Enugu road map, Enugu terrain map, Enugu
satellite view. Retrieved from
www.maps-streetview.com:
[image:6.595.42.289.94.371.2]All Rights Reserved © 2016 IJARCET
2521 11. Olayande, J. a. ( 2010). Nigeria electric power sector,workshop on power generation, International center for theoretical physics. abuja: nerc.
12. Oti M.I. (1995). Manufacture and Installation of Multi-bladed windmill. NJSE, 110 -117(13).
13. R. Datta and V. T. Ranganathan. (2003). A method of tracking the peak power points for a variable speed
wind energy conversion system. IEEE Trans. Energy
Conversion, pp. 163–168.
14. V. Agarwal, R. K. (2010). A novel scheme for rapid tracking of maximum power point in wind energy generation systems. IEEE Trans. Energy Conversion, vol.
25(no. 1), pp. 228–236.
15. Vestas Wind Systems A/S. (2014). Hedeager 44 . 8200 Aarhus N . Denmark.
Alade Tunde A. I MNSE, Cictp, MIEEE. A charted Electrical Engineer, holds B.Tech hons from Ladoke
Akintola University of technology Ogbomosho (LAUTECH), He started his professional career as a service engineer in Physical Planning Unit of Nnamdi Azikiwe university Awka Nigeria, presently a principal technologist in Electrical Engineering Department.
A post graduate student in COOU Anambra state with research interest in renewable energies, system and machine optimization.
Dr Nwosu Arinze W MNSE, MIEE holds a Ph.D. in electrical machines, a senior lecturer at COOU
Anambra state. He is currently the managing director of NASENI Nnewi.
His work interest in: Electrical Machines and Power System Optimization
Olisakwe Christian
o
. MNSE. A registered engineer by COREN, Presently, he is a Chief Labtechnologist in Electrical Engineering Department of Nnamdi Azikiwe University Awka.
