An algorithm for inferring master agent rules in an agent based robust Zone 3 relay architecture

In order to prevent cascading events caused by hidden failures in zone 3 relays, agent based relay architectures have been suggested in the recent past. In such architectures each zone 3 relay contains agents that require communication with other agents at various relevant relays in order to distinguish a real zone 3 event from a temporary overload. In the presence of hidden failures such temporary conditions may cause unnecessary trip in a Zone 3 relay aggravating the situation towards cascading failures. In a recently proposed scheme, a local master agent is consulted by all Zone 3 agents before a tripping decision is made. The Master agent maintains a rule base which is updated based on the local topology of the network and real time monitoring of the status of other relays and circuit breakers. In this paper we provide an algorithm that uses local grid topology and relay/breaker status information and creates these rules at the master agent. When queried, the rules allow the master agent to appropriately advise the Zone 3 relays on the local conditions which enables them to make robust decisions.