| The future advancement of hybrid energy networks will require application specific integrated system based analysis and improvement of their energy network characteristics in order to tolerate and/or compensate the negative effects of associated perturbations. In the general network context, the hybrid energy network may provide very little global information on the individual profiles of the various types of connected components. In addition, natural supplies could generate dynamically variable power profiles, resulting in fluctuations and system imbalances on the network. Thus, the analysis of hybrid energy networks is problematic, especially when the networks are exposed to exogenous, convoluted perturbations that could impact the reliability, resiliency and restoration or the "3Rs" of the electric power system.;In this dissertation, graph theory based new analytical methods are presented to evaluate the resistance and resilience to perturbation characteristics of hybrid energy networks. This is achieved by computing the variability of power profiles of energy nodes and the cooperation of multiple specialized areas when such networks are exposed to a series of convoluted perturbations. Spatiotemporal metrics such as energy node core-ness, power transfer assortativity, energy clustering, energy modularity and clique of energy nodes have been developed and tested on various system models including the IEEE 57-bus system and real-world energy networks. It was identified that the proposed graph theory based metrics could reveal the resistance and resilience to perturbation characteristics in a selected IEEE 57-bus system exposed to convoluted perturbations, even in cases where conventional, load support computational methods such as the Newton-Raphson and Gauss-Seidel methods could not converge to a specific solution. Thus, the proposed analysis was applied to identify critical energy network components in a new perspective.;The results further indicate the requirement of an intelligent component that could perform the proposed analysis from a perturbation tolerance point of view. Thus, an integrated expert system has been designed that performed temporal forward-backward induction on the energy storage node profiles. The expert system computed the developed energy network parameters namely energy core-ness, power transfer assortativity, energy centrality, and made use of line loading factors to identify optimal operating margins of converter connected sources and storage units. The expert system was developed in an IBM ILOG studio whereas the hybrid energy network test cases were implemented using ETAP, Matlab-Simulink and RTDS. The results show that energy centrality and energy core-ness parameters have positive relations to resistance to perturbation at higher values of loading of lines, whereas power transfer assortativity is negatively related to resistance to perturbation, although the relation improves at higher values of loading of lines. The results also disclose that, using the integrated expert system to optimize the developed parameters, improvements in the resistance to perturbations factor of up to 47.9 % could be obtained. These findings and the designed application specific integrated expert system could be used by network operators for optimal planning and energy efficient implementation of hybrid energy networks worldwide. |
