E.M. Farhadzadeh, A.Z. Muradaliev, S.A. Abdullayeva, A.A. Nazarov.
Quantitative assessment of the operational reliability of overhead power lines

The relevance of developing a methodology for assessing the operational reliability of overhead power lines of electrical power systems is due to the increase in the number and length of power lines, the service life of which exceeds the standard value. The lack of a methodology is due to unusual requirements for calculating reliability indicators. It is necessary to calculate a quantitative assessment of reliability based on statistical data only for operational data, i.e.,small enough time interval. All data preceding this interval cannot be used as characterizing an object in the "past". A calendar month is takenas the operational interval, in particular, because the economic operational efficiency is traditionally calculated on this interval. Let us recall that today, operational efficiency is understood as a combination of three properties: efficiency, reliability and safety. Such a transformation of the concept of efficiency is also determined by the consequences of the aging of EPS facilities. The recommended new methods for calculating average monthly estimates of reliability indicators, comparing these estimates, recognizing the expediency of classifying statistical data, allow making decisions that ensure reliable, safe and efficient operation of overhead power lines. Difficulties in developing a methodology for assessing operational reliability were overcome by taking into account the multidimensional nature of statistical data, using a fiducial approach when assessing the feasibility of information classification, conducting simulation of representative samples.

Keywords: Overhead power lines, Operational reliability, Aging of facilities, Quantitative assessments, Appropriateness of classification, Methodological support, Multidimensional data, Fiducial approach, Simulation
Copyright © 1997- Institute of Control Systems of Azerbaijan National Academy of Sciences. e-Mail: [email protected]