Application of Machine Learning for Online Dynamic Security Assessment in Presence of System Variability and Additive Instrumentation Errors

Abstract

Large-scale blackouts that have occurred in the past few decades have necessitated the need to do extensive research in the field of grid security assessment. With the aid of synchrophasor technology, which uses phasor measurement unit (PMU) data, dynamic security assessment (DSA) can be performed online. However, existing applications of DSA are challenged by variability in system conditions and unaccounted for measurement errors. To overcome these challenges, this research develops a DSA scheme to provide security prediction in real-time for load profiles of different seasons in presence of realistic errors in the PMU measurements. The major contributions of this paper are: (1) develop a DSA scheme based on PMU data, (2) consider seasonal load profiles, (3) account for varying penetrations of renewable generation, and (4) compare the accuracy of different machine learning (ML) algorithms for DSA with and without erroneous measurements. The performance of this approach is tested on the IEEE-118 bus system. Comparative analysis of the accuracies of the ML algorithms under different operating scenarios highlights the importance of considering realistic errors and variability in system conditions while creating a DSA scheme.

Techniques of power system static security assessment and improvement: A literature survey

“…The SCADA system has some limitations such as relatively lower accuracy of measurement and low refresh rate [ 42 ] which makes it impractical for online security analysis [ 33 ]. To resolve these drawbacks, new monitoring and communication infrastructure called Wide Area Measurement System (WAMS) with Phasor Measurement Unit (PMU) becomes practical to monitor voltage and current phasors at higher resolution (sampling rate) and fidelity [ 32 , 43 ]. The PMU technology helps EMS operators detect abnormalities quickly in emerging power systems with intermittent RERs [ 31 ].…”

“…Methods such as continuous power flow solutions of P –V or Q-V curves, modal analysis, or computation of voltage stability margin for voltage stability [ 8 ], time-domain simulation for transient stability, and analysis of eigenvalues for small signal stability are used to carry out DSA [ 30 , 31 ]. To prepare appropriate preventive and corrective measures that can maintain system security, simulations covering several operating conditions (obtained from different combinations of load demand and generated power) and contingencies must be performed [ 32 ]. However, the aforementioned simulation techniques are not only computationally heavy but also require detailed information on the fault characteristics, network configuration and protection behavior.…”

Recent Developments in Machine Learning for Energy Systems Reliability Management

“…In [28], a decision tree model is compared to SVMs and neural networks for transient stability assessment. The same algorithms are compared in [58], but this time considering also the random forest algorithm. Reference [85] compares decision trees, SVMs, core vector machines and naive Bayes models while [91] compares several ensemble methods (XGBoost, Bagging, Random Forest, and AdaBoost) with Naive Bayes, k Nearest Neighbor (kNN) and decision trees, this time for voltage stability assessment.…”

Dynamic Security Assessment For Power System Using Attribute Selection Technique

“…DSA is an assessment of the power system's ability to withstand sudden disturbances and maintain a secure case in the event of a sudden failure in the electrical power system. Recently, power system has undergone drastic changes related to privatization and installation of new devices and technologies such as renewable generation sources [4,5]. The privatization process pushed electric power networks to work near their limits, especially with the demand for continuous loads [5,6].…”