E3S Web Conf.
Volume 114, 2019International Conference of Young Scientists “Energy Systems Research 2019”
|Number of page(s)||5|
|Section||Reliability and Safety of Energy Systems|
|Published online||04 September 2019|
Prediction of the power shortage in the electric power system by means of regression analysis by machine learning methods
1 Melentiev Energy Systems Institute of the Siberian Branch of the Russian Academy of Sciences, Irkutsk, Russian Federation ;
2 Irkutsk National Research Technical University, Irkutsk, Russia
Modern electricity consumers place increasingly high demands on the level of reliability of power supply and, correspondingly, the reliability of electric power systems (EPS). These requirements should be directly addressed in the EPS development planning tasks. To this end, the evaluation of the level of EPS reliability is performed by employing software and computer systems that have various methods of reliability evaluation implemented therein. Among the variety of methods for identifying reliability indicators to evaluate resource adequacy the most appropriate one is the Monte Carlo method (the method of statistical tests): it enables to perform calculations within a reasonable time with the required accuracy, while the calculation of complex EPS-like systems by means of analytical methods proves impossible because of the large dimensionality of the problem. However, even when using the Monte Carlo method, the difficulties arise as well, namely the problem of the need to reproduce a large number of random states of the simulated EPS and the calculation of the operating mode of each of them. There are several ways to reduce the overall calculation time by either efficient random number generators and optimizers or alternative methods of the calculation of operating modes. The article deals with the issue of bringing up to date the method behind reliability calculation that makes use of the Monte Carlo method. We propose to use regression analysis methods when calculating operating modes of random states of the EPS. To this end, we adopt the support-vector machine and the random forest method. Experimental studies covered in the article attest to the efficiency of the new method, for the 24-node system IEEE RTS-96 the calculation speed has been increased by almost a factor of 4 while maintaining accuracy.
© The Authors, published by EDP Sciences, 2019
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.