MATLAB and Simulink Seminars

Transient Stability Assessment of Electric Power Systems using Predictive-SIME based on Machine Learning and Artificial Intelligence

Event Type Start Time End Time
Webex 27 Oct 2021 - 1:00 PM CDT 27 Oct 2021 - 2:00 PM CDT


The main effects and consequences of blackouts caused by power system instability show the importance of the transient stability phenomenon, which might be one of the main instability problems in power systems.  Electric power systems are permanently subjected to different types of disturbances that, eventually, trigger dynamic phenomena that could put in risk the electrical energy supply. The transient stability is part of these possible dynamic phenomena, and their fulfillment ensures the synchronous operation of the generators that are part of the system.

The main objective of this research is to protect the electrical power system against transient stability problems that may arise due to a large disturbance in the transmission lines. In the event of a disturbance that lead to the problem of transient stability, the identification and fast prediction of potentially dangerous conditions in the system is crucial, in order to have enough time to execute the necessary emergency control actions to avoid partial or total collapses in the power system. The status prediction and the transient stability margin explored in this research differs from the traditional transient stability assessment, which generally focuses on determining the Critical Clearing Time (CCT) in response to a contingency, also, this evaluation is carried out off-line to define maximum action times of the protections in the transmission lines of the system. 

Up to now, the analysis of multi-machine power system transient stability has been carried out mainly by offline simulations, which tend to implement computationally demanding and laborious methods. So it may not be applied as a tool for the analysis of transient stability in real time at control centers.  Unlike traditional methods, the proposed status prediction and transitional stability margin do not focus on a specific fault or disturbance, and mainly involves monitoring the SEP power electrical system in the period after the transmission line disturbance, using synchro phasor measurements.

The precision of synchro phasor measurements given by PMU devices, as well as the speed and reliability of telecommunications infrastructure are important and necessary preconditions for implementing prediction methodologies and improvement of transient stability in real time based on Wide Area Measurement Systems (WAMS).

This research presents an innovative methodology based on artificial intelligence and machine learning to predict the SEP transient stability margin in real time using synchro phasor measurements (PMU). This prediction allows evaluating, in real time, the status and transient stability margin of the SEP post-failure.

Based on probabilistic models of the input parameters, such as load variation and contingency occurrences, offline Monte Carlo simulations are performed to iteratively evaluate the transient stability margin responses of the power system using the methodology called Simple Machine Equivalent (SIME). Subsequently, the database obtained from the offline simulations is used to structure and train a classifier and an intelligent regressor based on artificial intelligence and machine learning using MATLAB to be used in the prediction of the transient stability margin from synchro phasor measurements. This variant of the SIME methodology that use synchro phasor measurements has been called Predictive SIME (P-SIME).


Several simulations are then performed on the New England benchmark power system. Results demonstrate the feasibility and effectiveness that could be achieved in predicting the transient stability margin for power system transient stability assessment.

About the Presenter

Diego Echeverría, PhD
Centro Nacional de Control de la Energía (CENACE), Ecuador

Diego Echeverría was born in 1982 in Puyo, Ecuador. He got his Electrical Engineer degree in 2006 from National Polytechnic School, Quito-Ecuador. He is currently completing his PhD studies in Electrical Engineering at the Electric Power Institute (IEE) of the National University of San Juan, Argentina. Between 2005 and 2008 he worked at Agency for Regulation and Control of Electricity. His research experience includes an internship at Institute of Electrical Power Systems, University Duisburg-Essen, Duisburg, Germany, as part of the DAAD scholarship in 2011. Nowadays, he works in Ecuador at the Independent System Operator CENACE as the Head of Research and Development Department. Since 2012, he is an active IEEE and Power & Energy Society (PES) member. He is at present an IEEE volunteer, serving as the Ecuador PES Chapter Chair and as the Chapter Vice-Chair in different periods. His special fields of interest comprise power systems control and stability in real time, synchrophasor measurement technology, wide area monitoring systems and development of Smart Grids.

Transient Stability Assessment of Electric Power Systems using Predictive-SIME based on Machine Learning and Artificial Intelligence

To continue, please disable browser ad blocking for and reload this page.

You are already signed in to your MathWorks Account. Please press the "Submit" button to complete the process.