Jump to content, Georgia Institute of Technology, College of Engineering, School of Electrical and Computer Engineering (ECE), ECE Research, ECE Research Labs
College of Engineering
Search | Contact ECE | Feedback | BuzzPort
GT Home > COE Home > ECE Home > Research > Labs > ACES, Advanced Computational Electricity Systems
Georgia Tech
Advanced Computational Electricity Systems(ACES)



Ayusman Roy

State Estimator is an essential tool for real-time power system monitoring, and a prerequisite for real-time security analysis, control, and economic applications. The utilization of state estimators is becoming more important at all levels of transmission and distribution systems as an enabling technology for electricity control, automation, and smart grid applications. Because the success of control objectives depends heavily on the ability to accurately determining the state of controlled devices and subsystems, having flexible state estimation functions at all levels of the electricity system is of paramount importance to realizing future electricity systems.

The conventional state estimator currently used in the industry at both the transmission and distribution levels, has the fundamental limitation of assuming certainty about the statuses of switching devices. An undetected topology error can completely undermine the performance of even the most statistically and numerically robust state estimator, since the algorithm would be trying to solve for the wrong network topology. Generalize State Estimators (GSE) incorporate breaker status as variables to be estimated. In this project we develop a robust, unified GSE, to address three difficult and interrelated problems present in the conventional state estimators currently implemented in the industry:

a) Inability to detect topology errors. This can be addressed by using generalized state estimation techniques, which handle both topology and measurement errors, simultaneously. As part of this project we will demonstrate how pervasive topology errors can damage conventional state estimators and how GSE implementation resolves topology error issues.

b) Implementation complexity due to the two-model (node/breaker and bus/branch) architecture and implementation. This can be solved by utilizing a novel unified architecture, which allows transparent handling of any switching device topology. Since the core routine of the GSE consists in reducing and expanding portions of the network with suspect breaker statuses, handling switching topologies at will is critical for computational speed.

c) Poor handling of bad data and network parameter errors. This can be addressed by using robust estimation techniques.

The project goes beyond application prototyping to propose a unified-model estimation architecture that can be adopted by the industry and that is fundamental for the success of numerous emerging applications.


Last revised on Aug. 25, 2011