Introduction - Control of Power Inverters in Renewable Energy and Smart Grid Integration - page 18

1.2.3 DC-AC Conversion
A DC-AC converter, also known as an inverter, generates an AC output from a DC source.
There are different types of inverters. According to the type of the DC supply, an inverter is
known as a current-source inverter (CSI) if the supply is a current source and a voltage-source
inverter (VSI) if the supply is a voltage source. Typically, an inverter is a VSI if there is a
large capacitor across the DC bus and is a CSI if there is a large inductor in series with the DC
supply. According to the type of the inverter output, an inverter is called current-controlled if
the output is controlled to be a current source and voltage-controlled if the output is controlled
to be a voltage source. Hence, there are current-controlled VSIs and voltage-controlled VSIs,
and there are also current-controlled CSIs and voltage-controlled CSIs. The details of CSIs can
be found from textbooks about power electronics, e.g. (Bose 2001), and will not be discussed
in the rest of this topic. The inverters dealt with in this topic are all VSIs. According to the type
of commutation, inverters can be line commutated (e.g. those built with thyristors) or forced
commutated (e.g. those built with IGBT and MOSFET). In the rest of this topic, only forced
commutated inverters are dealt with. The output voltage waveform of a voltage-controlled VSI
can be a square wave, a modified square/sine wave, multi-level or a pure sine wave. In the rest
of this topic, voltage-controlled VSIs are expected to have a purely sinusoidal voltage output
with minimal harmonic components.
The amplitude of the output of an inverter can be fixed or variable. Moreover, the frequency
can be fixed or variable as well, depending on the applications. These can be easily achieved
with pulse-width-modulation (PWM) techniques. There are many different PWM techniques
available (Asiminoaei et al . 2008; Cetin and Ermis 2009; Holmes et al . 2003; Holtz 1992,
1994; Lascu et al . 2007 2009; Wong et al . 2001). In this topic, the focus is not on PWM
techniques and the widely-used sinusoidal PWM is adopted in most cases. Note that the main
objective of PWM is to change a signal with possibly variable amplitude into a train of pulses
with variable widths to drive the switches. Hence, as long as the average of the pulses over one
switching period well approximates the original signal, then it should not considerably affect
the performance with a well-designed controller if the switching frequency is high enough,
according to the averaging theory (Khalil 2001). When the switching frequency is not high
enough, some particular PWM strategies should be adopted.
1.2.3.1 Sinusoidal PWM (SPWM)
Most inverters are required to provide a clean sinusoidal voltage supply with a fixed or
variable frequency, which is normally much lower than the switching frequency. In this case,
the desired clean sinusoidal output voltage, called the modulating signal, can be compared
with a triangular carrier wave at the switching frequency to generate a train of pulses, as
shown in the left column of Figure 1.17. The harmonic components of this signal are mainly
around the multiples of the switching frequency. If the pulses are amplified to drive a VSI,
then the output voltage of the inverter has the same shape. When the carrier frequency, i.e. the
switching frequency, is high enough, then the harmonic components can be easily filtered out
via a low-pass filter, which is often an LC or LCL filter. This type of modulation is called a
sinusoidal PWM (SPWM). The frequency of the reference signal determines the frequency of
the output voltage and its peak amplitude controls the modulation index and then in turn the