IEEE Electrification Magazine - December 2017 - 54

optimal flux-weakening control
scheme has been developed. The
proposed method is based on a
lookup table that provides the current set points according to the
speed and power voltage measurements. To implement this control
strategy, two steps are required.
The first presents a specific characterization scheme of PMSMs. The
second shows how to build lookup
tables for the flux-weakening control by using an optimization algorithm under constraints, based on
obtained identification results. An
example is given to illustrate each
step of this study. Please be advised that the control
scheme carried out in this article is under patent by
Piaton et al.

proposed to compensate for the
power demand of an electric-drive
tracked bulldozer; This method
does not consider the magnetic saturation, and the flux-weakening
control is not optimal.
A study of the flux-weakening
control method and the position
sensorless control for an interior
PMSM has been outlined by Matsumoto and Hasegawa (2012 and 2016).
It is based on an L d model, where
the maximum torque control frame
is estimated by the flux model of the
position observer. This approach is
based on a linear motor. A similar
flux-weakening algorithm is proposed by Halder et al. to
achieve high-speed operating points in terms of maximum torque per ampere control, without exceeding the
voltage and current limit. However, the magnet saturation
is not considered. In the same way, in Ekanayake et al., a
flux-weakening control with two current controllers with
maximum torque per voltage is discussed for segmented
interior PMSMs; this approach is based on a linear motor
and depends on the motor parameters.
A study of the current harmonic suppression problem
of the linear PMSM in the flux-weakening control is developed by Pan et al. It depends on the motor parameters,
and the magnetic saturation is not considered. In the
work done by Miao and Shen, a method is proposed to
utilize flux-weakening control to restrict the permanent
magnet synchronous generator output voltage at high
speed and, consequently, increase the maximum operation speed in applications, such as wind power generation
systems with variable operation speed (Inoue et al. 2009).
A hardware solution is discussed by Fang et al. (2016) for
flux-weakening control by using an auxiliary capacitor. Lei
et al., Song et al., and Yuan et al. (2010) propose a fluxweakening control method with a single-current regulator
and a voltage angle control, by using the d-axis control
only for motion and the q-axis control only for generation.
These methods have the advantage of a simple structure,
but the magnetic saturation is not considered. In many of
the works in the "For Further Reading" section, the q-axis
voltage is given by using lookup tables [Longya et al. 2008,
Matsumoto and Hasegawa (2012 and 2016), Miao and Shen
2012, Morimoto et al. 1994, Pan et al. 2016, Piaton et al.
2015, Song et al. 2007, Wei et al. 2014, and Yuan et al. (2010
and 2011)]. Consequently, a large number of experimental
scenarios are required. By using single-current flux-weakening control, the maximum torque per given single voltage remains under optimal value. By using both axis
voltages, we can reach the maximum and optimal torque
with the same level voltage.
In this article, the goal is to present an optimal fluxweakening control based on lookup tables. In the first step,

The flux-weakening
control is carried
out by adjusting the
direct current (id) to
cancel out some of
the magnetic field
produced by
permanent magnets.

Flux-Weakening Control Methods
Flux-weakening control is increasingly used in PMSMs,
control to expand the speed range. The flux-weakening
control is carried out by adjusting the direct current (id) to
cancel out some of the magnetic field produced by permanent magnets. Many flux-weakening control methods are
based on two current regulators. If the voltage saturations
of these regulators are not well managed, the current vector could be uncontrollable in transient operations. In
addition to this, the coupling between d- and q-axis currents makes tuning difficult.
The work by Hayakawa et al. and Matsumoto and
Hasegawa (2016) proposed the new feedback-type fluxweakening control using values of the voltage saturation
and the voltage margin; this algorithm can achieve the
antiwindup control automatically, and, therefore, this
method can be used even in the case of magnetic saturation. However, the flux-weakening margin is not optimal.
In similar work by Morimoto et al., a current regulator is
proposed to improve the current responses in the fluxweakening region, which includes the decoupling current controller and the voltage-command compensator
regardless of magnetic saturation.
In the work done by Jang-Mok and Seung-Ki, a fluxweakening level is adjusted inherently by the outer
voltage-regulation loop to avoid saturation of the current
loop. This method does not depend on motor parameters, but it does not guarantee an optimal torque under
magnetic saturation. The work by Fang et al. (2014a and
2014b), Wei et al., and Zhu and Wen carried out the fluxweakening by adjusting the q-axis voltage fully to make
the motor work at the optimal operating points. This
method is easy to achieve and does not depend on motor
parameters, but the magnetic saturation is not considered. In the work by Lin et al., an outer voltage-loop fluxweakening control strategy based on overmodulation is

I E E E E l e c t r i f i cati o n M a gaz ine / DECEMBER 2017

Table of Contents for the Digital Edition of IEEE Electrification Magazine - December 2017