Autonomous state estimation for the smart grid - Laboratory implementation

The centralized power grid is now evolving to smart grid characterized by distributed generation and automation technologies. Additionally, the development of inexpensive communication and control technologies enables the interconnection and coordination of distributed energy resources to the main grid. This paper focuses on the basic tool in any automation and control system, and on the validation of data and models in real time, i.e. the state estimator. We propose a robotic approach to the state estimator. The proposed scheme requires that data acquisition systems transmit not only measured data but the model of the device they control as well as the connectivity of the device to the system. An intelligent electronic device (i.e. a PC) uses the data model and connectivity to perform state estimation, detection of bad data and bad data rejection for the purpose of extracting a reliable real time model. The concept of the SuperCalibrator is utilized to make the process fully distributed. The proposed robotic approach enables true automation and plug-and-play capability. The proposed robotic approach to state estimation and the control of the system is being implemented in the laboratory. Specifically a laboratory comprising a small scale power system has been developed at Georgia Tech. The scaled model has been instrumented with many numerical relays including GE Hard Fiber system. The operation of the system is controlled by a computer, function generator, and a set of amplifiers that provide the necessary power to operate the system. The function generator is designed to create various non-ideal conditions for the purpose of simulating the impact of distributed generation in the operation of the system. Data communication follows IEC61850, a standard for communication networks and systems in substations.