Study Of Fault Diagnosis In Distribution Substations Based On Hybrid Cause And Effect Nets

With the development of automation in distribution substation, it will need not anyone to keep watch. SCADA sends information to DMC——Distribution Management Centre. When fault occurs, operators will be send to judge and deal with it by DMC. So the absence and delay will form disadvantage factors. Intelligent fault diagnosis system can assistant operators to form fast recovery decision-making for faults, to ensure the security of system and to recover the current supply. So it has important meaning and use value to embed fault diagnosis module in monitor system and to send diagnosis outcome to management centre.Local area communication based in distribution substation the SCADA system can obtain different information forms, as the SOE information sequence in the fault process, switching information sequence of circuit breakers, fault recorder information, fault alarm information etc. According to obtainable information this paper presents the hybrid C-E (Cause and Effect) Nets based in use for fault diagnosis of distribution substations. It is applied that associative relation between information to make the redundant correcting technique and deep knowledge reasoning. It is effective that overcomes the C-E Nets based to rely on excessively shallow knowledge and that improves the fault-tolerance performance of fault diagnosis.Hybrid C-E Nets and BPNN are used to emulate threephase short-circuit. Via contrast, the advantage of C-E Nets in the fault-tolerance performance of fault diagnosis is incarnated prominently.