STATIC STATE ESTIMATION AND LOAD FORECASTING VIA ADVANCED COMPUTATIONAL TECHNIQUES

Abstract

The system operator faced high level of uncertainties due to the evolving complexity of the electric power systems. Therefore, efficient solutions for power system prediction, monitoring and state estimation can be found in new methods. These methods improve the secure operations of the network. This thesis covers three different state estimations incorporating static, tracking, and dynamic state estimation in order to estimate all possible operating conditions of power systems. This research also introduces new meta heuristic methods such as stochastic and fractal search technique. The SFS technique is implemented in real time nonlinear power system applications under various scenarios. New methodology of multilayer neural networks exhibited in composite typologies are proposed in this thesis to improve the estimation performance. Optimized Neural Network by Stochastic Fractal Search technique is used and applied to both tracking and dynamic state estimation. The proposed methods are validated utilizing diverse benchmark optimization methods. The combination of conventional and synchronized measurement is also studied in this thesis. This is used to increase the reliability of electric power systems in realtime. Additionally, the research is extended to evaluate the benefits of multiarea state estimators and how it uses to reduce the computational time. Finally, all formulations proposed in this work were validated in different IEEE test systems.