船舶电力系统故障重构通过改变系统供电路径来恢复重要负载的供电,对增强系统的存活性有着重要意义。在多智能体(agent)架构的基础上,构建船舶电力系统博弈重构模型并提出求解方法。该模型以船舶电力系统中不同区域的agent作为博弈参与者,将船舶电力系统的重构问题转化为各agent之间的合作一竞争问题,从而保证重要负载的恢复供电。论文进一步提出该模型的求解算法,该算法首先判断博弈模式的合作与非合作属性,之后通过agent或agent联盟之间的博弈过程,计算各agent所属的负载/发电机的投切及转移策略。仿真算例表明,该算法能够快速有效地求解船舶电力系统重构问题,最大程度保证系统的存活性。
The reconfiguration process for a shipboard power system (SPS) alters the topology of the system in order to achieve certain objectives, such as enhancing service continuity and minimizing power losses under cascading failures or battle damage while satisfying the system's operating constraints. Reconfiguration techniques can be classified into two categories: centralized and decentralized methods. The decentralized reconfiguration technique has outstanding advantages such as fast computing speed and the avoidance of single-point failure, and therefore is preferred for SPS instead of centralized techniques.