Parameter Reduction in Probabilistic Critical Time Evaluation Using Sensitivity Analysis and PCA

TL;DR

This paper introduces a combined sensitivity analysis and PCA approach to identify key parameters affecting probabilistic transient stability in power systems, enabling faster assessments by reducing parameter complexity.

Contribution

The paper presents a novel method that efficiently reduces the number of parameters in probabilistic stability analysis using sensitivity analysis and PCA.

Findings

Effective identification of influential parameters in IEEE 14 bus system

Significant reduction in computational time for probabilistic assessment

Method maintains accuracy while reducing complexity

Abstract

In this paper, we discuss a method to find the most influential power system parameters to the probabilistic transient stability assessment problem---finding the probability distribution of the critical clearing time. We perform the parameter selection by employing a sensitivity analysis combined with a principal component analysis. First, we determine the sensitivity of the machine angles with respect to all system parameters. Second, we employ the principal component analysis algorithm to identify the most influential parameters in the transient stability problem. By identifying such parameters, we can reduce the number of uncertain parameters to only the influential ones in the probabilistic assessment of transient stability, providing a significant speed-up in the probabilistic analysis of large power systems. The proposed algorithm was tested in the IEEE 14 bus systems and the results obtained show that our method can effectively find the most influential parameters.