联网运行下微型电网电压自愈控制方法

传统联网运行下微型电网电压自愈控制方法使用电流差动保护措施，无法远程控制馈线开关，不能快速实现电压自愈。提出一种新的自愈控制方法，将MAS分布式控制系统引入联网运行下微型电网电压自愈控制中，通过双层MAS结构中下层各Agent将微型电网运行情况及故障信息反馈到上层微型电网Agent后，制定最终自愈控制方案，并下达命令进行优化自愈恢复；通过各Agent间协作快速定位电网故障和进行电压自愈控制，其中故障定位时，Agent基于改进差动电流判断故障发生依据和支路故障判断依据，实现故障定位分析；电压自愈控制时采用启发式搜索方法形成供电恢复路径，各个联络开关处Agent发送恢复命令至故障末端分段开关处Agent，并通知本地Agent开挂合闸，实现微型电网电压自愈。通过实验表明，该自愈控制方法故障定位精准，自愈恢复耗时仅为0.27 s，自愈控制性能佳。

Method of Voltage Self-healing Control for Micro-grid under Network Operation

Under the traditional network operation, the voltage self-healing control method for micro-grid uses a current differential protection measure, which cannot remotely control the feeder switch or quickly realize the voltage self-healing. In this paper, a novel self-healing control method is proposed, in which a multi-agent system(MAS) distributed control system is introduced into the micro-grid voltage self-healing control under network operation. After the micro-grid operation status and fault information are fed back to the upper micro-grid agent through the lower agents in the two-layer MAS structure, the final self-healing control scheme is formulated, and an order is issued to optimize self-healing. Through the cooperation between different agents, the grid faults are quickly located, and the voltage self-healing control is carried out. In the process of fault location, the agent realizes fault location analysis based on the improved fault occurrence judgment criterion of differential current and the branch fault judgment criterion. In the voltage self-healing control, the heuristic search method is used to form the power supply recovery path. Afterwards, the agent at each contact switch sends a recovery order to the agent at the fault end segment switch, and notifies the local agent to turn on the suspension gate, thus realizing the micro-grid voltage self-healing. Simulation results show that the proposed self-healing control method is accurate in fault location and the self-healing recovery time is only 0.27 s, delivering excellent self- healing control performance.