Title:

Development of novel sensorless maximum power point tracking controllers for wind turbine generator systems

In recent years, wind energy has become one of the rapid growing renewableenergy
sources. According to the new power report from the European Wind Energy
Association (EWEA), it forecasts that by 2020 the European Union will achieve 20%
of power generation from renewableenergy sources, e.g. wind, solar and biofuels.
Wind energy is a clean and inexhaustible energy source. It is available in all locations,
especially remote ones with rich wind resources and plentiful land, which are suitable
for developing largescale wind farms.
Typically, there are two wellknown strategies for operating wind turbine generator
(WTG) systems, including a fixedspeed strategy and a variablespeed strategy. The
former strategy is suitable for largescale WTG systems, which are directly connected
to a grid via capacitor banks for adjusting the generated reactive power. Most of the
fixedspeed WTG systems employ pitch angle controllers for extracting maximum
wind turbine power from wind. The main disadvantages of the fixedspeed strategy
are: first, the mechanical torques are highly affected under rapid wind speeds, i.e.
wind gusts, which cause power surges on a grid and second, additional expensive
equipment, e.g. motors, actuators and drivers, are required to implement a pitch
angle controller. In literatures, the first problem was tackled by keeping the reference
pitch angle constant at rapid wind speed variations in order to decrease mechanical
stresses on a wind turbine tower. Whilst, the variablespeed strategy has been widely
employed for maximising the output power of WTG systems using maximum power
point tracking (MPPT) controllers, which can be applied via power electronic converters.
The power delivered by a WTG system is dependent on the swept area of
a wind turbine, wind speeds, power coefficients of a wind turbine and the current
v
drawn from a generator. The only controllable factor is the power coefficient, which
varies with operating tip speed ratios (TSR). For coming wind speeds, there is a
unique optimal TSR that keeps power coefficients at its maximum value. In order to
achieve the optimal TSR, it is required to control rotor speeds of a WTG system to
follow reference rotor speeds, which can be produced by a TSR controller based on
measurement or estimation of wind speeds.
In Chapter 2, a comparison study between a classic direct field oriented controller
(FOC) and an optimised direct FOC, has been presented. The proposed VTG system
comprises a verticalaxis wind turbine (VAWT), a permanent magnet synchronous
generator (PMSG), a threephase controlled rectifier and a standalone DC load. The
objectives of these controllers are for improving the efficiency and the dynamic performance
of a WTG system as well as minimising rotor speed overshoots under rapid
wind speed variations. The developed controllers are based on a wellknown FOC
method, through adjusting stator currents and consequently electromagnetic torque.
FOC transforms threephase stator currents into two currents in the rotational reference
frame, i.e. daxis and qaxis currents, using the Park transformation. These
daxis and qaxis currents act as DC currents. To apply FOC, reference rotor speeds
or reference electromagnetic torques are required to generate reference qaxis currents,
whilst reference daxis currents are usually set as zero for minimising loss. It is important
to note that the Park transformation needs the knowledge of rotor positions,
which can be measured by an encoder. In practice, an encoder cannot measure an
accurate initial position, which may lead to wrong calculations of daxis and qaxis
currents. It is worth noting that the parameters of a PI current controller are firstly
tuned using a classic zero and pole placement method and secondly optimised using
a particle swarm optimisation (PSO) algorithm. The PSO algorithm is adopted due
to the following advantages: such as easy to implement with simulations in realtime,
a high computational efficiency and stable convergence characteristics. An accurate
model for a PMSG is important for the design of a highperformance PMSG control
system, because the performance of such control systems is influenced by PMSG physical
parameter variations under real operation conditions. In this research, electrical
parameters of a PMSG are optimally identified, e.g. the stator resistance per phase,
the stator inductance per phase and the rotor permanent magnet flux linkage, using
also a PSO algorithm. It is important noting that the bounds of these parameters
are obtained using standard tests, e.g. an opencircuit test, a shortcircuit test and a
load test. The aim is to increase the accuracy of parameter identification, reduce the
search space of parameters and decrease the convergence time of a psa algorithm,
i.e. the computation time required to reach an optimal solution.
One of the difficulties for implementing the direct vector control strategy is the
requirement to fix an anemometer close to wind turbine blades in order to obtain
accurate wind speed measurements, otherwise inaccurate calculations of reference rotational
speeds are obtained causing a WTG system not to rotate at optimal speeds.
For cost and reliability consideration, a sensorless MPPT controller, which is based
on a novel TSR observer is developed. The purpose of the proposed TSR observer is
for estimating TSRs and consequently reference rotor speeds without the knowledge
of wind speeds. The proposed TSR observer is based on the wellknown perturbation
and observation (P&O) method. It is also known as the hillclimbing searching
method, which doesn't require any previous knowledge of wind turbine and generator
characteristics. In spite of these advantages, it has some problems, which considerably
decrease its dynamic performance. These problems include the steadystate oscillations
around a maximum power point, a slow tracking speed, a perturbation process
in a wrong direction and a high rotor speed overshoot under fast wind speed variations.
In this research, these problems are tackled by using adaptive perturbation step
sizes instead of fixed ones. For implementing the proposed MPPT controller, a costeffective
powerelectronics converter, which consists of a threephase diode rectifier
and a DCDC boost converter, is constructed for experiments. Furthermore, a complete
transfer function of the proposed system has been derived, which is employed to
design a speed observer for estimating rotor speeds and consequently, rotor positions
and for testing the stability of the developed rotor speed observers and controllers.
In this thesis, another robust sensor less MPPT controller has been proposed for
maximising the output power of a WTG system. A switchmode rectifier (SMR),
which includes a threephase diode rectifier and a DCDC boost converter without a
boost inductance with an input capacitor filter for harmonic mitigation, is employed
for implementing the proposed sensorless MPPT controller. The proposed sensorless
MPPT controller is based on two novel observers, i.e. an adaptive slidingmode observer
(SMO) and an adaptive P&O algorithm. The former is used for estimating
backEMFs and consequently rotor speeds without the knowledge of rotor positions
using an adaptive PMSG model in the stationary ex/3 reference frame, an adaptive
sliding gain and an adaptive cutofffrequency LPF. The purpose is to eliminate the
chattering effect (which occurs in conventional S1\,1Os ) and decrease estimation errors.
The adaptive P&O algorithm is developed to estimate reference rotor speeds
and optimal duty cycles based upon turbine coefficient errors and rotor speed errors,
respectively. It uses adaptive variables compared with some widely used P&O algorithms,
which use an adaptive perturbation step size but a fixed observation period.
The adaptive variables are: (i) a perturbation step size, which decreases steadystate
oscillations around optimal operating power points and (ii) an observation period,
which is another contribution of this work. It increases the tracking speed and ensures
that MPPT is always executed in the right direction with small rotor speed
overshoots under fast wind speed variations. It should be noted that the developed
sensorless MPPT controllers are experimentally validated using a WTG simulator.
The data acquisition and control stage of the power electronic converters are implemented
using a digital signal processing and control engineering (dSPACE) controller.
In this thesis, the analysis of experimental results has been undertaken to verify the
proposed observers and controllers. Finally, future research work is suggested.
