The ever-growing requirement for reliability and quality of power supply suggests to enable self-healing features of smart distribution network using intelligent communication and control. In this article, a concept of fully decentralized multi-agent system (FDMAS) automation is proposed to build a unified restoration service framework for distribution network with distribution generators (DGs), where an FDMAS interaction mechanism is designed for establishing a reduced model which can significantly reduce the computational dimensions of service restoration. Furthermore, an FDMAS-based strategy is proposed for service restoration by combining network reconfiguration with intentional islanding; especially a network reconfiguration algorithm based on network flow model is presented, which, along with parameter justification, can mitigate the variations of loads and intermittence of DGs. The simulation studies are carried out on the 84-bus and 22-bus distribution system, respectively, using MATLAB and java agent development framework (JADE) simulation system and dynamic model test platform. The test results show that the proposed strategy can maximize restoration of out-of-service loads with minimum switching times and has an excellent performance on service restoration time.

中文翻译：

---

分布式配电网的完全分散多代理服务恢复

对电源可靠性和质量的不断增长的需求表明，必须通过智能通信和控制来实现智能配电网的自我修复功能。在本文中，提出了一种完全分散的多代理系统（FDMAS）自动化的概念，以建立一个具有分配生成器（DG）的配电网络的统一恢复服务框架，其中设计了一种FDMAS交互机制来建立简化模型，该模型可以大大减少了服务还原的计算范围。此外，提出了一种基于FDMAS的策略，通过将网络重新配置与有意孤岛相结合来进行服务恢复。特别是提出了一种基于网络流模型的网络重配置算法，并结合参数说明，可以减轻负荷的变化和危险品的间歇性。使用MATLAB和Java代理开发框架（JADE）仿真系统和动态模型测试平台分别在84总线和22总线配电系统上进行了仿真研究。测试结果表明，所提出的策略能够以最少的切换时间最大程度地恢复服务中断负载，并在服务恢复时间上具有出色的性能。