Statistical inference of multivariable modal stability margins of time-delay perturbed power systems

Abstract This paper proposes a modal statistical inference algorithm to define multivariable operational stability limits of time-delay perturbed dynamic systems by employing remote sensor signals. Our proposal overcomes the drawbacks of linear independence and inter-area geometry of time-delays to analyze multiple-input, multiple-output dynamic systems. The manuscript contributes with: (a) the study of stability margins under distributed uncertainty and time-delays for large power systems, (b) the determination of stability conditions of inter-area oscillations through a new probabilistic modeling approach under the influence of intermissions, and (c) the usage of the proposed methodology to derive controlled stability limits and assess the modal resilience of perturbed power systems. Studies on the multi-scale signals sensitivity and multivariable polynomial intersection from empirical perspectives in modal stability analysis are also explored. Results on an IEEE 16-generator 68-bus system are presented to illustrate the effectiveness of the proposed algorithm. The estimation of multivariable operational stability limits and time-delays of inter-area oscillation modes are verified with the vector fitting procedure and first-order Pade approximation.