Báo Cáo Logic based error detection algorithm in a substation automation system

ABSTRACT

​This paper addresses a novel data error detection method in a substation. Unlike conventional
methods, the proposed algorithm uses logical reasoning taking general structure of a substation into
consideration.


A substation model is divided into several modules according to the voltage level and connection
status. This modular approach enables us to utilize as many data as possible, because the data in main
database do not contain all pieces of information measured at process buses. In addition, it reduces
search space compared with centralized approach.


For the analog data correction, the conservative reasoning is adopted, which means that a data value
is not evaluated to be “True” in case every evidence is proved. For digital data, a measure “Falseness”
is suggested, by which a data can be evaluated to be false or not. Every member of a data set is ranked
according to the Falseness which is calculated on every possible set of evidences. And the data with
highest Falseness become candidates of bad data.


After the basic module is investigated, the original basic module is expanded by including one of
adjacent modules. The new basic module is evaluated in the same manner as the original basic module.
Simulations on a standard substation of KEPCO demonstrate the effectiveness of the proposed
algorithm.