Research On Fault Diagnosis For Non-health Information Distribution Network

Rapid and accurate fault diagnosis is a prerequisite to ensure the normal operation of distribution network.It can ensure the high reliability of power supply,reduce the economic loss caused by faults,and reduce the impact on the life of the residents and industrial production.Distribution network fault diagnosis has been the focus of research and made a lot of research results,but there are still many problems that need to be further studied.This paper focuses on two of these issues to study.First,how establish model and make fault diagnosis correctly under the non-health information.Second,in the future a large number of distributed power access distribution network,how establish unified adaptive fault diagnosis mode.The main work of this paper is as follows:First,a brief introduction to the domestic and foreign research on distribution network fault diagnosis methods is summarized.Then an information acquisition method based on the SACDA information platform is proposed,and an information fusion method based on the DS evidence theory is proposed.It can help to get better information according to the analysis the logical relationship between various fault information when distribution network fault happens.Second,analyzes the construction of distribution automation,a new improved matrix fault diagnosis based on sparse technique is proposed in the substation layer.It is applicable to the distribution network with DG.The matrixes are symmetric sparse Boolean matrix in the fault diagnosis of the distribution network.Their elements are only 0 or 1.So using the sparse matrix storage and computing technique the required storage space and calculation amount are greatly reduced,and the speed of fault diagnosis is improved,which make it more practical.Finally,in the master station layer improved fuzzy Petri network fault diagnosis is proposed,which can effectively utilize the time relationship between the information.This paper proposes a unified fault diagnosis model for faulty components.The model has nothing to do with the connection mode of the distribution network and does not need to be re-modeled when the network structure changes.The DG accessing makes element protection model changes,so it needs re-modeling.This paper simplifies the model for distribution network with DG.And the BP algorithm is used to train the parameters in the model,which reduces the influence of human subjective factors in the fuzzy theory.