Page 62 - Mathematical Models and Algorithms for Power System Optimization

52 Chapter 3

installed capacity is 1000MW, and the minimum is 500MW, which means the two largest
generating units cannot be maintained simultaneously because it will exceed the maximum
allowable maintenance capacity.

3.1.2 Basic Requirements for Annual Generator Maintenance Scheduling

GMS is a medium- and long-term planning problem. GMS is used to arrange the generator
maintenance sequence and schedule. The units should be considered as being in an outage state
from the start to completion of maintenance. Therefore, the basic requirements of GMS are, on
the one hand, not only to determine the capacity of the maintainable units in the maintenance
period, that is, deducting the capacity (disabled capacity) of the units unable to reach their full
capacity from the generating capacity of the whole system, but also to meet the load demand
and minimum reserve capacity during the maintenance period. On the other hand, GMS also
takes into overall consideration the maintenance durations of all units, makes feasible
combinations of unit maintenance under the previously mentioned conditions, and further
confirms each unit and its capacity are available to be maintained in the maintenance period.

3.1.3 Overview of This Chapter

This chapter studies the optimization method of the annual GMS based on the dynamic fuzzy
logic method and takes the GMS problem in a province as an example. This chapter uses the
expertise of professionals to form the ES knowledge base, simulates the inference method of the
power grid dispatcher to form GMS solution rules and the basis for logic judgment, and replaces
the conventional rigid constraints with flexible ones based on FDP. It provides a rapid,
effective, and convenient tool for the comparison of several GMS solutions as well as a scenario
in which the maintenance schedules of the generating plant need to be updated as soon as
possible after system failure or when load conditions have a great change.

This chapter presents the basic idea of optimizing modeling, specifically in Section 3.2 as
reference. Section 3.3 presents the optimization mathematical model of GMS. Section 3.4
introduces the fuzzification of a GMS mathematical model. Section 3.5 introduces the ES for
solving GMS. Section 3.6 introduces the optimization calculation algorithm and process of
GMS dynamic programming. Section 3.7 presents detailed results of actual calculation
examples and analysis.

3.2 Basic Ideas of Developing an GMS Model

GMS is a medium-term planning problem, whose scheduling interval is quite long (about
1 year). When considering all necessary operating conditions and constraints, the GMS
problem is very complicated.