Dynamic security evaluation functions in the more care project

Abstract

This paper describes the security assessment functions of an advanced control system for secure operation of isolated networks with increased renewable penetration, developed within the MORECARE project. One of the key features of this project is related with the capability of assessing on-line dynamic security and providing preventive control measures that can assure a robust operation for the system regarding some disturbances. The paper describes with some detail the general approach followed to derive these evaluation functions, which are based in functional knowledge generated off-line through computational simulation. Techniques like Decision Trees, Artificial Neural Networks and Hybrid Regression Trees were successfully exploited and integrated with other mathematical technical do deal with the problem.