Page 293 - Mathematical Models and Algorithms for Power System Optimization

Local Decoupling Control Method for Transient Stability of a Power System 285

In theory, given the previous premise, it is possible to use the computer to perform stability
control calculations on the power system, but in reality, it is far from being so simple. On the
issue of how to make decisions and implement control, we will encounter more than just
theory; it is also a practical difficulty.

First, to analyze a specific system, it is often necessary to carry out a great number of
calculations, simply in the hope of getting some answers in the more typical cases in emergency
operation situations. For the moment, if a computer is used for centralized online control, even
if we make the calculations in advance and just save the calculated control schemes in the
computer for retrieval, it is often the case that its scale and speed cannot keep pace with the
changing operation requirements.

Second, the information required for collection from the control center for direct control will be
extensive across the system, including not only direct information but also skip-level
information. Collecting skip-level information is not only of low reliability but also costly.
Moreover, the information gathered in this way is often used only for local operations.

Therefore, there will be two practical barriers when pursuing online centralized control. One
is the difficulty of real-time data collection, and the investment in data collection systems is
huge; the other is that the speed and reliability of computer systems are not enough to
accommodate the demand for real-time closed-loop control of a power system.

This shows that centralized stability control is not appropriate and difficult to achieve for such a
large interconnected power system in current reality conditions.

In addition, according to the large system theory in modern control theory, from the perspective
of reliability, it can be proven that, for complex large systems, the centralized control structure
scheme is not the optimal one, whereas decentralized control is suitable in accordance with the
interconnection of the controlled objects.

In so-called “decentralized control,” that is, the respective stability control of each part of the
system, usually only limited information is obtained because it is generally difficult to obtain all
information of the system from a certain part of the system (even if it is available, the reliability
is very low). What kind of impact can a group of local emergency measures based on this
limited information have on systemwide stability?
From a certain point in the system, it is difficult to give a clear answer in advance, because the
power system is a complex interconnected system with strong interactions. When the system
has a stability problem in an emergency, local stability control is performed based on limited
information obtained only in each part of the system. Some are feasible in the local view, but
they are not necessarily feasible from a global perspective and may even lead to a systemwide
collapse. That means, for complex power systems, there is no simple causal relationship
between local stability requirements and global stability requirements.