*3.2. MAS Applied in Energy Domains*

The energy sector is becoming more complex and consists of multiple hybrid systems, which includes various interactions and amounts of knowledge. MAS is being studied in many areas of power engineering including diagnostics, condition monitoring, power system restoration, market simulation, network control and automation, and hierarchical decision making, as smart grid (SG) and microgrids (MG) [18,20]. The development of simulation platforms based on MAS is increasing as a good option to simulate real systems in which stakeholders have different and often conflicting objectives [21].

#### 3.2.1. MAS for Grid Control

According to [18], research on using MAS in power engineering mainly focuses on distributed control architectures and simulation. MAS is a decentralized scheme that utilizes distributed controllers for energy management and optimization, and it is an alternative approach for smart system optimizers (SSOs) implementation within a typically integrated energy system (IESs) [22]. MAS is an obvious and promising choice for the smart grid control system because MASs can overcome the threat of SPOFs (single-point-of-failure) due to their distributed characteristic [23]. Meanwhile, Considering the agent properties, the variety of components used in power transformer and the huge amounts of data involved, MAS provides the best possible choice for the purpose of monitoring, automating, controlling and diagnosing the power transformer components [24]. MAS has proven to be suitable for addressing the demands of SGs both theoretically and practically [25].

Most of the research work in this area have focused on hierarchical control, optimization, and power restoration using MAS. For instance, [21] proposes a MAS-based optimal energy management solution for the optimization problem of the interactive operation of generation units and DR [26]. Similarly, introduces a decentralized agent-based approach for optimal residential demand planning [27]. A MAS is used in [28] to restore power in case of failure, and [29] introduces a flexible and versatile MAS for fault isolation and power restoration. Meanwhile, [30] presents a MAS automated management and analysis of SCADA and Digital Fault Recorder Data. Furthermore, a multi-agent system is used to control the voltage of the power system with co-ordination in [31].

Other distributed MAS-based solutions to grid control are also presented microgrids, islanded microgrids, and multiple microgrids [8]. The applications of MAS in a microgrid is similar to the smart grid control, e.g., Microgrid control, optimal energy exchange, and multi-level management, but also link to buildings or demand-side management. For instance, [32] presents a MAS for Microgrid control and a classical distributed algorithm. [33] proposes a MAS microgrid system for optimal energy exchange between the production units of the Microgrid and local loads. based on MAS, [34] proposes an Intelligent Distributed Autonomous Power System (IDAPS) to increase the reliability of the critical loads. [35] proposes a multi-level management and control scheme for microgrid systems taking into account the interaction among agents at different levels. [36] presents a consumption scheduling framework in small residential areas.
