Improving Resilience in Smart Grids Using Advanced Predictive Analytics and Computational Intelligence Optimization

In recent decades, the integration of intermittent and un-schedulable renewable energy brings more uncertainties into the real-time operation of the power systems that can adversely impact the power grid resilience. In many reported research works, advanced predictive analytics and computational intelligence optimization for the real-time operations and controls of the power system, based on online monitoring, parameter optimization, and real-time decision making, have been proposed to improve power system resilience. In this thesis, novel predictive and large-scale optimization methodologies are proposed to respond quickly and proactively to uncertain changes in the power systems to be able to make intelligent long- and short-term planning and decision-making strategies to improve the power system resilience. The proposed optimization methodologies involve multi-period, multi-objective, large-scale optimizations that can utilize parallel programming taking into consideration the increasing penetration levels of distributed energy resources (DERs), such as the small-scale renewable energy resources (RES), the energy storage system in the form of a battery energy storage system (BESS), and the increasing demands for a fast response to fluctuations on the real-time price and load demand. This thesis presents a set of predictive analytics and real-time computational intelligence algorithms and strategies to solve the large-scale optimization problems in modem power systems to improve the power system resilience considering the uncertainties of the sources, loads, and electricity prices