Topology-adaptive noise-immune fault detection system using an associative memory network

摘要

This paper presents the fault detection system with bi-directional associative memory network (BAMN) for a looped system. BAMN has the function of noise filtering to process uncertain signals from the data acquisition. The proposed method makes use primary/back-up information of protection devices to detect the fault sections involving single fault, multiple faults, data communicate with errors, or fault with the failed operation of the relays and circuit breakers. With a looped system, computer simulation was conducted to show the effectiveness of the proposed system.