IN the last years, distribution automation has gathered a
significant relevance in distribution systems planning and
operation. The network operator (NOp) looks for a suitable
configuration of the feeder topology as well as the protection
system, pursuing the reliability enhancement and a full energy
demand supply. Nevertheless, an efficient protection system
requires an adequate investment in such devices as reclosers,
fuses and sectionalizers. Thus, two conflictive objectives arise,
namely, NOp investment minimization and reliability maximization.
In this sense, the number and location of devices
in the system are critical variables to accomplish preceding
objectives. Here, we focus on ...view middle of the document...
As in the aggregating function approach, the output of
this optimization strategy depends on the decision maker point
of view, which provides the boundary limits of each objective.
In this way, several authors have developed protection
systems planning by applying the aforementioned methods.
Dehgani et al.  realize a compound index optimization such
that SAIFI, SAIDI and MAIFIe are minimized by using a
genetic algorithm (GA). In the case of distributed generation
(DG) enhanced feeders, Wang et al.  implement ant colony
system (ACS) to minimize an index composed by SAIFI and
SAIDI. In this line, Li et al.  apply a multi-population GA to
solve this index optimization problem. Pregelj et al.  include
MAIFIe in the composite index and make use of GA to solve
the decision making. Besides, Greatbanks et al.  optimize
the placement of reclosers as well as DG within the feeder.
These authors implement GA to reduce reliability indices.
However, MOOP transformation into SOOP may be considered
as an incomplete solution to multiobjective decision
problems. Hence, the Pareto approach arises as the main
alternative to obtain an efficient set of solutions that look
for compromises between considered objectives. As far as we
know, there has been no research done on Pareto multiobjective
planning of protection systems, with the exception of
the Pareto multiobjective optimization based on ACS, which
minimizes both total costs and reliability indices such as SAIFI
and SAIDI on a non-DG feeder .
This paper provides an approach that considers a Paretobased
multiobjective optimization to enhance SAIFI and
SAIDI at the same time that investment costs are minimized
on a DG enhanced feeder. The outcome of this optimization
is a set of efficient solutions, obtained with the application
of multiobjective evolutionary algorithms (MOEAs), which
provide the amount and position of NCRs in the feeder. At
first glance, this decision-making is described by nonlinear ob2
jective functions and combinatorial solutions. Thus, it leads to
a nonlinear combinatorial multiobjective problem (NCMOP),
which usually has a NP-Complete complexity level. Hence, it
is necessary to apply metaheuristic approaches.
The remainder of this paper is organized as follows. In
section II efficient planning of NCRs by applying a multiobjective
optimization approach is described. The optimization