Tuning interconnection relays using a multi-agent system to detect weak infeed conditions

Abstract In this paper, a multi-agent system (MAS) is used to enhance the performance of protection under dynamic operating conditions that result in a weak source contribution. A traditional relay setting that is established for the worst-case scenario is not suitable for systems with a wide operating range, mainly due to the presence of non-regulated sources. A MAS is proposed that operates on-line to modify relay settings in response to topological changes due to the output of lines, loads or generation, and identifies low contributions from sources. The gang architecture for the relay coordination is simulated in Matlab/Simulink and controlled using Jade. Matlab enables modeling that integrates the diverse dynamics of the elements of the electrical network, and Jade allows for a multi-agent environment that facilitates the different actions that are activated in the network. The MAS architecture improves the relay performance under fault conditions by detecting weak infeed conditions caused by both non-regulated sources and low-level conventional sources. The on-line coordination between interconnection relays and line relays uses the MAS architecture to send transfer trips, and uses non-conventional time curves to improve the operation times. The results show that the MAS improves the sensitivity and operation times of the interconnection relays.