Pre-and post-disturbance transient stability assessment using intelligent systems via quick estimating of the critical clearing time

Abstract

Abstract The real time transient stability assessment is the important steps of the dynamic security evaluation of power systems. To do this, intelligent systems (ISs) such as neural networks as an effective method to accurately and quickly estimate the critical clearing time (CCT) has been widely used. But, choosing the proper inputs for these systems remains a major challenge for researchers, still. Variables related to energy functions such as minimum kinetic energy, the slope of the variation of the minimum kinetic energy and maximum potential energy contain useful information in estimating CCT. Accordingly, in this paper, these variables are used as the IS inputs. However, the time-domain simulation of the power system response to obtain these inputs is time consuming. To be able to use the energy function-based inputs for real time stability assessment, in addition to the main IS used to estimate CCT, another ISs are used. By those ISs, a very limited period of system response is simulated to obtain proper inputs of the main IS. The method is simulated on the 10-generator New England test system. Simulation results show, the CCT can be found by simulation just 0.05 s of the considered power grid.