New developing reconfiguration method for naval shipboard power distribution system after fault situation

The most serious scenario among the possible scenarios that a warship may encounter in a combat environment is damage or injury. Damage or injury can occur due to any conflict, impact, malfunction, or fire. It is vital for a warship that is incapacitated due to injury or any malfunction to regain control of its situation, return to its mission as soon as possible, or be able to withdraw to a safe area where assistance can be obtained. Ships have complex power systems that encompass propulsion/drive systems, platform systems, communication systems, navigation systems, combat/weapon systems, and many other critical systems that carry electrical loads. While the missions performed by warships in normal or combat situations vary, the loads that need to be present in the ship's power system and their importance levels also vary depending on the types of missions. This situation necessitates the ship's power system to perform multifaceted task-based functions. In the event of possible damage, the ship's power system first detects the damage, isolates the fault, identifies the fault location, operates the reconfiguration process of the power system considering the priorities of loads and the damage situation, performs the necessary switching for the transition of the power system to the new state, and checks the final status of the system. Preserving the ship's operational capability, reducing the power outage duration, and the personnel's abilities are as important as the method used in the success of a reconfiguration process that is effective and safety-oriented. In this thesis, the focus of the study is on the reconfiguration (restoration) process after fault isolation and fault location processes have not been addressed. In the reconfiguration method used, intelligent algorithms are incorporated into the method task-based with operator knowledge. With the intelligent algorithm-based reconfiguration method developed in the thesis, the system status is analyzed, and the best switching result is obtained. Reconfiguration is performed by controlling the switches in the distribution system obtained as a result of the method. During the reconfiguration process in the ship's power distribution system, manual or automatic bus transfer devices fed from dual sources (normal and alternative) become particularly important for vital and semi-vital loads, while load shedders are used to collectively shed non-vital loads from the system. In classical approaches, it should be noted that when the energy of the normal line is cut off, the bus transfer devices transfer the load to the alternative line without any reasoning. In this case, with the classical approach, encountering the pessimistic scenario where the fault spreads, even the power-generating generator systems go offline due to overload after load transfer, and ultimately encountering the scenario where the distribution system goes offline. The proposed intelligent reconfiguration method aims to apply the most accurate solution to the power system according to load priorities, taking into account the overall evaluation of the system and especially the objective functions such as power generation-consumption balance. The developed reconfiguration method has been analyzed under different scenarios using three different optimization techniques: 0-1 integer programming, genetic algorithm, and pattern search. The proposed method in the thesis has provided consistent results with all three optimization techniques. Particularly, the 0-1 integer programming method, which seems sufficient for objective functions using sets of linear equations, has yielded faster results compared to other optimization techniques with the algorithm proposed in the thesis. However, genetic algorithm and pattern search optimization techniques have emerged as the strongest due to their ability to adapt to complex problems and work with equations of complex structure, owing to their inherent randomness. Before using evolutionary algorithms such as genetic algorithm and pattern search, the calculation/adjustment of their parameters affects the success of the solution, whereas the integer programming method can be applied more easily in the proposed reconfiguration process compared to other methods. In this thesis study, a method aimed at the reconfiguration of a sample warship's power system after damage or malfunction using intelligent methods has been developed, and the results obtained have been examined. The study focused on the ship's power distribution system with a radial structure. The dynamically changing importance levels and priorities of loads based on the ship's mission type were utilized, and operator experience was incorporated into the method. Simulation-based analyses were conducted on the ship's power distribution system. The proposed intelligent reconfiguration method has been tested in challenging scenarios created for the ship's power system, and the results obtained in the thesis study have shown that it prevents a complete power outage on the ship and successfully maintains the system operational. The proposed reconfiguration method utilized genetic algorithm, pattern search, and 0-1 integer programming methods, and the results of the three different optimization techniques were comparatively presented in the thesis study. The proposed method yielded consistent results with all three optimization techniques. The calculation/adjustment of parameters before using evolutionary algorithms such as genetic algorithm and pattern search affects the success of the solution. Due to their inherent randomness, the adaptation of these optimization methods to complex problems and their ability to work with complex equations are their strongest aspects. However, it has been observed in this thesis study that solution times for sets of linear equations are longer compared to the 0-1 integer programming method. Dynamic load priority lists, operator knowledge, and intuitive rules defined with penalty scores on a load basis have been included in the method content. In the thesis study, the results obtained from the developed method and three different optimization methods have been compared, and it has been observed that intelligent reconfiguration methods are highly successful in a radial distribution system compared to classical reconfiguration methods Considering the future potential of renewable energy systems, a simulation scenario including a ship model with solar energy panels has been included in the thesis content to examine the response of the developed method under configuration changes in the ship's power system. The simulation results have also evaluated the results of the proposed reconfiguration method presented within the scope of the thesis along with conventional damage control systems without intelligent algorithms, considering their effects on the ship's power system.