【作者】 傅荟璇；

【导师】 刘胜；

【作者基本信息】 哈尔滨工程大学 ， 控制理论与控制工程， 2007， 硕士

【摘要】 船舶电力系统网络重构是船舶电力系统故障的重要组成部分，是提高船舶生命力的重要途径之一。本论文对网络重构过程中涉及的网络拓扑表达及分析、故障拓扑跟踪及处理、网络重构求解方法等问题进行了深入研究。针对大型船舶电力系统的特点，采用扩展关联矩阵法进行网络拓扑结构在线分析，其中的扩展关联矩阵是将支路及节点按照供电关系进行编码并依次扩展而成，直观自然地反映了电网结构。同时运用了带有网络拓扑结构识别的实用潮流计算方法—改进的节点电势法，仿真试验结果证明了方法的准确性。利用扩展关联矩阵结构描述方法进行了故障时的拓扑跟踪及处理，将故障时的网络拓扑结构表达与此时的潮流计算相结合，给出了简便的故障后系统潮流计算方法。研究了一种新的故障跟踪和提取方法，运用数据库实时反映故障情况，进行故障定位提取，提高了故障跟踪速度，并用实例验证了其合理性。分析了遗传算法求解故障恢复问题的优越性，并针对标准遗传算法及其在故障恢复应用中存在的不足，充分利用系统对象的特点及进化过程中的信息，提出了一种改进的遗传算法。优化问题的测试结果表明，改进的算法加快了搜索速度，有效地避免了不成熟收敛。根据故障后的网络拓扑跟踪结果，采用遗传算法及其改进方法求解了网络重构问题，将遗传操作中的个体映射成不同的网络拓扑结构，并生成扩展关联矩阵，运用故障后的潮流计算方法进行了可恢复判断，通过仿真和试验，验证了恢复方法及其与系统结合操作的可行性及准确性，保证了最重要负载的恢复供电及各级负载的最大限度恢复。更多还原

【Abstract】 Reconfiguration of the electrical network in a shipboard power system is an important activity, which is required to be done either to restore service to a section or to meet operational requirements of the ship, is one of the most important methods to restore power supply and improve survivability.In this paper, the problems involved in SPS reconfiguration, such as network topology expression and network analysis, fault tracing and handling, reconfiguration algorithm, are all studied in detail.Based on analyzing the characteristics of the large SPS, adopt the expanded associated matrix, carries on the on-line direct and natural method analysis. The expanded associated matrix is presented to identify the network topological structure, which expands the associated matrix by using nodes and branches in term of their power supply connection. At the same time, the improved node voltage method, a practical system calculation method with the ability of topology identification is used, and the accuracy of the method is verified by tests and simulations.Using the expanded associated matrix representing the network topological structure to trace the fault network and handle. By combining the topology expression of fault network with the power flow calculation, a simple calculation method of the fault system is obtained, Studied a new fault trace and processing method, utilized the database real-time reflection fault situation, carried on the fault localization to extract, accelerate the fault tracing and its validity is tested by examples.The superiority of the genetic algorithm(GA) to solve the service restoration problem is analyzed. To deal with the defects of simple GA and its application in restoration, a improved genetic algorithm is presented, which makes full use of object characteristic and the information of the system and evolutionary process. The results of simulation tests on optimization problems show that the improved GA increase the convergence speed and avoid the premature convergence.Based on the network topology tracing result after the faults, the improved GA are applied to solving the reconfiguration problem. The individuals of GA are mapped into different network topological structures and the expanded associated matrixes are created, and then, the restoration judgement is made by using the fault system calculation method. Simulation and testing results demonstrate the feasibility and accuracy of all the restoration methods which are combined with the operation on the system. Thus, the system can guarantee the service restoration of the most important loads and restore each grade load maximumly.更多还原