Relay Protection Based On Multi-Agent System

This paper is an in-depth study of the Agent and Multi-Agent System, the Agent for alltypes of structure, organization of Multi-Agent Systems, Agent collaboration model, solvingmechanisms and communication methods are in-depth analysis, based on existing technology,Design of a Multi-Agent-based relay system. System is divided into three layers, there arenetwork reconfiguration layer, middle layer and the bottom of the execution levelcoordination. Network reconfiguration layer is a network reconfiguration Agent, on the relayfrom the system level, Multi-Agent Systems for Management, responsible for the overalloperation of the system, adjust. Detect the middle of the coordination layer contains Agent,Agent protection and search Agent, real-time computing systems detect Agent identificationparameters; Protection Agent includes a variety of protection schemes, to detect Agentcalculated based on fault component, select the protection action; Search Agent is responsiblefor all of Agent communications and search, Agent communication and collaboration is thefoundation. The bottom layer contains the implementation and enforcement of dataacquisition Agent Agent, Agent collecting data acquisition system, real-time protectionparameters for the system decision-making; Executive Agent management protection systemin each of the protected action devices, mainly on the lines of breakers, receive protectionAgent to issue protective action, failure to complete resection.In the system design, based on this in-depth study of the Multi-Agent Systems ofcommunication and collaboration process. To KQML Knowledge Query Language as acommunication infrastructure, KQML Agent translation of various beliefs and ideas, asKOML for each Agent communication language is an internal transparent, we use the XMLlanguage KQML statement on the package, the use of mature computer communicationprotocol used transmission in the network. Supplemented by the existing power grid means ofcommunication, and preliminary estimates to achieve quality of communication between theAgent. The results show that the system can completely meet the requirements of relayprotection. At the same time, this article on Multi-Agent implementations have been studied,the choice of Agent-Oriented Development Method MADE, select the protection system insearch Agent, and Agent Java language development platform MADE, multi-protectionsystem in the search for Agent of Agent the software.Multi-Agent in the completion of the preliminary design of protection systems, thisarticle in order to protect the type of adaptive current protection, current protection of theadaptive mechanisms of protection, the scope of protection, compared with the traditionalcurrent protection, in order to verify the specific instance of adaptive protection Multi-Agentsuperiority, and with PSCAD platform, reached the fault clearing waveform chart. Simulationresults show that adaptive protection than traditional Multi-Agent-current protection to, the reaction rate, eliminate tripping, malfunction, action to improve protection system responsetime, etc., has significantly increased.This study was designed based on Multi-Agent Protection System, to achieve amulti-Agent technology and the combination of relay protection, the use of advanceddistributed artificial intelligence technology to solve some of the problems encountered in theexisting relay, through simulation, we It can be seen on Multi-Agent Protection System isfeasible, but has its advantages.