الفهرس 
Objective of the Thesis
CENTRALIZED AND DECENTRALIZED PROBLEM FORMULATIONOnly 14 pages are availabe for public view
 |  | from | 169 |  |  |
| | | from | 169 | | |
Abstract
Creative solution approaches in order to solve the multi-regional optimal power flow (OPF) problem for the smart grid are addressed. The two employed OPF approaches are: centralized (inter-bounded) and decentralized (holistic) methods. In multi-regional power networks, OPF variables will contain relevant entities from the neighboring connected areas. Therefore, these entities have to be shared with each other for achieving reliable, stable and effective performance for the entire grid. The multi-regional network is counted as one large grid in the decentralized method. While, the control variables of the inter-connected boundaries between the areas are only evaluated in the centralized approach.
In this thesis, the AC-OPF and DC-OPF formulations have been applied to solve the multi-regional OPF problem. This is utilized by considering three approaches, namely: 1- Adjusting the active power of generation unit; 2- managing the voltage magnitude and 3- Combining both, the changes in the real power of generation units and the voltage variations. The optimization methodologies applied to achieve the most optimum result including, the well-known Genetic Algorithm (GA), which is used as a tool to validate the solution reached through the other utilized new algorithms, the Whale Optimization Algorithm (WOA), and the Marine Predators Algorithm (MPA). The presented methodologies are investigated on the IEEE 48-bus inter-connected electric power system, that consists of two regions connected to each other via transmission lines. Initially, the parameters of WOA and MPA are fine-tuned to enhance the performance of them. Then, both algorithms; in addition to GA, are utilized to solve the DC-OPF and the AC-OPF for the both centralized and decentralized approaches.
In addition, the system is analyzed under stochastic conditions; the influence of variable renewable energy resources (Wind and/or Solar) is applied to the system including both approaches; centralized/decentralized. For each approach three conditions are considered; integration of wind generation alone just to cover 30% of the needed demand, integration of solar generation alone just to cover 30% of the needed demand, and integration of both wind and solar generation to meet 40% of the needed demand. In each case, the location of the renewable energy resources (RERs) will also be determined, in order to minimize the overall system cost. Moreover, the system is investigated under the effect of the load uncertainties (load profile variability). The system is studied for the whole week, considering both Solar and Wind RERs.
Furthermore, the system is evaluated based on different cases of contingencies that can be occur in the system. These contingencies happen due to the result of transmission line outages due to insulation failure or a lightning strike or generating unit failure due to a mechanical issue.
The results for each case study are depicted. Comparisons between the mentioned algorithms were executed for confirming and demonstrating the robustness, and revealed that the tuned WOA and tuned MPA surpassed the GA in terms of reaching better values for the DC-OPF and AC-OPF objective functions. As well as, the comparison showed and proven the small difference between the DC-OPF and AC-OPF from an economic standpoint, which indicate that DC-OPF can be utilized in order to tackle the economic problems. Additionally, the results by considering renewables and load variability are analyzed and showed the effectiveness of the inserted renewables in each case. Finally, the results under different cases of contingencies are presented and the showed the effect of the outage on the network in each case.
The results confirmed that while the decentralized approach provides a more optimal solution, it needs additional computational power since it analyzes the entire inter-connected system in the OPF model. On the alternative hand, the centralized OPF approach reaches a much quicker solution, however, it is less optimal compared to the decentralized approach. The centralized approach is appropriate for quick responses, and when the entities are controlled by an independent coordinator that only concerns about optimizing the inter-connected tie line flows.