Research And Implementation On Fault Location And Recovery Technology For Single Phase Grounding In Distribution Network

Fault location and recovery technology for single phase grounding in distribution network has not been solved for a long time. This paper presents a fault location method of Multi-information fusion based on Extreme Learning Machine. Terminal use transient energy method, wavelet method, The first half-wave method to Extract Feature vector, input to the Extreme Learning Machine Network which has the be-trained parameters and output search result. Using extreme learning machine to fuse fault zero sequence transient current can eliminate the inherent flaws of single positioning method effectively.After determining the fault troubleshooting section, we need to restore power of non-fault region, which is the network reconfiguration problem. This paper proposes an improved algorithm branch, preferentially selects a bus fault lines in the same branch circuit, adopt a "hot swap" operation in the case of meeting the conventional constraints in order to avoid blackouts load. If the "hot swap" cannot be achieved need to take a "cold switching" operation, select the fault lines in different segments of the branch, not only to meet the conventional constraints, but also the degree of compensation at this time.On the basis of these theoretical studies we developed transient section location and recovery system based on multi-information fusion, including 10 sets of "three remote" type terminal and a master system.We did RTDS simulation test, built a simulation model based on RSCAD software for demonstration line topology, simulated all segments of the occurrence of different types, plotted and analyzed the fault of each characteristic quantity waveforms. Test result show that the system can be achieved accurate positioning for all segments.