Page 180 - Modern Control of DC-Based Power Systems

144 Modern Control of DC-Based Power Systems

Figure 5.21 ISPS I L15L25L3 I&I for ideal CPL and 30.9 - 53.4 MW.

5.4 DECENTRALIZED CONTROLS
The previous two sections presented nonlinear schemes, which are
based on a centralized architecture. In this section, an approach is intro-
duced that is based on a decentralized control architecture and in addition
is also linear.
To enable a decentralized control approach the ISPS has to be
decoupled. Logical decoupling points are every generation side interfac-
ing converter as depicted in Fig. 5.22.In [31] and [32] the concept of the
“virtual disturbance” for power system decoupling was introduced,
according to which, under proper modeling and the usage of an aug-
mented Kalman filter, it is possible to estimate simultaneously local states
and interactions with other subsystems.
It consists of two parts. The first part is a linear controller which is
based on the principle of the Linear Quadratic Regulator (LQR) [33],
which successfully stabilized a CPL in [34]. The second part is based on
an observer—Kalman filter [33]. A Kalman filter is similar to Leunberger
observer, except that the basic framework is stochastic rather than deter-
ministic. The concept was proposed in [35] and [36]. Both together form
the Linear Quadratic Gaussian (LQG) control.
This control approach has found successful application on the stabili-
zation of CPLs in [37,38]. The optimal control was chosen because it has
the following properties which the other small-signal approaches do not
have. On a small-signal basis the results obtained by linear optimal control