Page 71 - Decision Making Applications in Modern Power Systems

42 Decision Making Applications in Modern Power Systems

TABLE 2.1 Classified uncertain parameters in power system [7].
Technical parameters Economical parameters

Operational parameters Topological parameters
Load demand Line outage Economic growth
Generation output Generator outage Price levels
Governmental regulation
Unemployment rate
Fuel price

Ref. [7], technical uncertain parameters are related to a topological network
such as failure rate of transmission lines or generators. The other technical
uncertain parameters that affect the operating decisions are generation and
demand values in system. The economical uncertain parameters contain
energy price, economic growth, environmental policies, etc., which face the
decision-making process with multiple challenges [8]. Table 2.1 categorizes
all the possible uncertain parameters in the power system that should be han-
dled for suitable operation of the system.
However, in scheduling and operation of power system, the main objec-
tive is cost minimization or profit maximization [9]; therefore the main
uncertain parameters faced by us include generated power of renewable ener-
gies, load demand of consumption, and energy price. So, we focus on the
modeling of these parameters. In the following section a comprehensive
review on uncertainty handling in power system will be provided. Then, we
will choose one of them and provide the example to describe the uncertainty
modeling in power system.

2.2 Uncertainty management in power system: a review
There are multiple methods for handling uncertainty in power systems. The
main feature that causes the discrepancy between multiple methods is in line
with the different techniques used to describe the uncertain input data and
parameters [10] and the degree of uncertainty. In the following section, some
of the main uncertainty handling approaches will be described.

2.2.1 Probabilistic method
One of the simple and earliest methods, which assumes all the input uncer-
tain parameters as random variables, is a probabilistic approach [11]. In this
method, each variable has a known probability density function (PDF), and