Knowledge domain states Mapping Concept for intelligent power flow control

Abstract

The power system can be viewed as a dynamical changing system with many components being modified depending upon the utility requirements. Under such operational scenario, the basic objectives of system’s control have been to ensure stable behavior even during changing dynamics. The present work is an attempt to propose a power control concept as outcome of research work to stabilize system even under varied dynamical changes. The work presented in thesis reports an intelligent multi-area power control with dynamic knowledge domain inference concept. It is known that ongoing operational shift leads to unacceptable states variation, which may result in power oscillations, and if the controllers are not suitably designed, the system may be interactive and oscillations can aggravate. To address these issues, Knowledge Domain States Mapping Concept for intelligent power flow control has been presented in this research work. The work includes a new concept of updating control parameters, which is linked with operational shift, initially in offline mode in building respective knowledge domain that fits into the framework of changing situations, to ensure states regulation. The proposed concept also provides flexibility to update the knowledge domain over and above offline data with newer data set combining the nearest data clusters to derive an averaged data (controller parameter) within predefined boundary to change the controller functioning. The knowledge retrieval, as operational shift proceeds, has been mapped utilizing dynamical inference concept. The control so derived, effectively ensures best damping well within time for large network reliability and security. The structure of the controller so obtained is termed as the intelligent power flow controller. Three well-known optimization techniques (Particle Swarm Optimization (PSO), Gravitational Search Algorithm (GSA) and Firefly Algorithm (FA)) have been used to develop knowledge domain and find the optimal parameters of the controllers associated with operational shift. This approach demonstrates the capability of controller performance based on knowledge domain driven tuning of controllers employing dynamic knowledge domain inference concept. It is to emphasize that this approach is general enough to accommodate any system complexity on modular basis thus extendable to large power system as such. Power System Stabilizer (PSS) is used as local supplementary excitation control to enhance the damping of the electromechanical mode of oscillations in power system by individual generator regulated participation. The FACTS controller, due to fast power flow control, not only enhances the steady-state stability but also improves transient stability along with damping of low-frequency oscillations and thus ensures better network availability. New controllers (Integrated Multi-Stage LQR-POD controller and Modified-Multi Stage LQR (M-MSLQR) Controller) have been reported in this work to improve the system stability under various operating conditions.