SGER: Robust Gain Scheduled Control Design in Power Systems

Iowa State University, Ames IA

Investigators

Abstract

The electric power system in North America has undergone unprecedented changes with the advent of a competitive market place and deregulation. These changes in the system have resulted in higher levels of loading on the system, and have further increased its stress. At the same time, the transmission grid has seen very little expansion due to the prevailing economic conditions and the lack of incentives in the market. As a result, available transmission and generation facilities are highly utilized with large amount of power interchanges among companies and geographical regions. It is envisioned that this trend will continue to grow and result in more stringent requirements to maintain reliability and adequate system dynamic performance. To a large extent, critical controls like excitation systems, power system stabilizers (PSS), static VAR compensators (SVC), and a new breed of control devices driven by modern power electronics and referred to as flexible ac transmission systems or FACTS play a key role in maintaining adequate system dynamic performance. Proper design of these controls resulting in robust system-wide stability and performance is essential. With the increased emphasis on reliability and system dynamic performance there is a greater need to analyze and design controls in an integrated fashion, taking into effect the interaction between the various kinds of controls. We propose to apply gain-scheduling techniques to enable the use of a family of robust controllers. More specifically, we propose to explore the use of linear parameter varying (LPV) approach to gain scheduling. This is a relatively recent approach that has not been seriously explored for large power systems. On the other hand, LPV gain scheduling methods are gaining acceptance in aerospace applications following their successful use in flight control systems. In these applications, one observes wide changes in the operating conditions (e.g. altitude, airspeed) similar to those seen in power systems. Specifically, we will address the following topics: Development and application of linear parameter varying (LPV) gain scheduling methods of control design to large power systems accounting for large changes in operating conditions. A novel decentralized control design for large-scale power systems. Broader Impact The project addresses a critical need for the reliability of the national electric grid dealing with the efficient design of controllers. The project also propose a novel decentralized approach to controller design for power systems that will significantly reduce the computational burden for large power systems but still take into account the uncertainty associated with changing operating conditions. This will provide initial testing of the approach and provide insight into the development of more elaborate techniques for designing controls. The analytical basis for the approach will also be developed. The project will support 1 PhD student for a year. This will contribute to the nation's scientific work force and produce an engineer with advanced capabilities to tackle the important issues related to the reliability of national electric grid.