الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 107 |  |  |
| | | from | 107 | | |
Abstract
In this thesis, the optimal power flow (OPF) problem is solved using the Firefly algorithm (FA) as an optimization tool. The proposed algorithm modifies the conventional FA so that it can handle the constraints easily. A parameter-less penalty function is used, so this algorithm can be applied to any optimization problem without adjustments of any parameters. Furthermore, the basic FA is integrated with opposition based learning (OBL) during the initialization of the positions of the fireflies and during their update process. Using OBL during initialization increases the possibility of reaching a feasible solution. Furthermore, integrating FA with OBL reduces the gap between worst solutions and best ones during the initialization and updating processes. The proposed algorithm is suitable for real time applications due to its relatively small computational time compared with other algorithms in literature. The proposed algorithm is tested on four power systems: 26-bus, 30-bus, 57-bus and 1354-bus and the results are compared with those of some existing algorithms. The proposed algorithm can deal with piecewise functions and multiple objectives easily and again gives competitive results. Furthermore, in this thesis, the firefly algorithm with a modified parameterless penalty function is used to determine the optimal settings of Flexible AC Transmission System (FACTS) devices including the Thyristor controlled phase shifting transformer TCPST, Thyristor controlled series capacitor (TCSC) and Static VAR Compensator (SVC). Use of FACTS in a power system reduces the active power losses and improves the voltage profile. The proposed algorithm solves an optimal power flow (OPF) problem for the system without FACTS devices as an initial step to determine the optimal settings for the generators’ voltages and active powers and to calculate the initial active power losses. The optimal locations for FACTS devices are heuristically chosen by ranking the lines and the buses of the power system according to the reduction in the power losses. Finally, the optimal settings of the FACTS devices are selected to minimize the active power losses and to enhance the voltage profile. The proposed algorithm is tested on the IEEE 30-bus and the IEEE 57-bus power systems and gives high quality solutions in a relatively small computational time. Finally, the proposed algorithm solves a multi-objective OPF problem with two and three different objective functions simultaneously. The two and three dimensional optimal pareto front graphs are obtained easily. Additionally, the algorithm is shown to be robust under load uncertainties using a Monte Carlo simulation based approach.