Design And Implementation Of Fault Diagnosis For Communication Network Device In Smart Substation

With the rapid development of the smart grid,the smart substation has become an significant part in the construction of the smart grid.While promoting the intelligent construction of substations,the functions and structures of substations are becoming increasingly complex,which puts forward higher requirements for the fault diagnosis system of smart substations.At present,the fault diagnosis research of substation is mostly focused on primary equipment.According to the characteristics of networked secondary equipment in the smart substation,this thesis proposes a method for communication network fault diagnosis based on Substation Configuration Description(SCD)file,and diagnoses the fault on the determined message transmission port,which improves the accuracy of fault diagnosis.In order to meet the needs of fault diagnosis and abstract the communication network configuration information into a mathematical model,this thesis establish a communication network topology model of smart substation,including the physical network topology matrix model and logical network topology matrix model.In order to quickly obtain the information of the substation communication network in SCD file,a parallel parsing method of SCD file is proposed.The method adopts a master-slave resolution mechanism,and the master threads extract the structure of the SCD file,and the slave threads parse the name and attribute of the label.Based on the establishment of the topological model of the substation communication network,a step-by-step method is introduced to narrow the fault range.In the first step,the rough set theory is used to diagnose the fault range from the whole station to the determined virtual local area network.Aiming at the problem of a large amount of knowledge reduction in rough set theory,the genetic algorithm is used to accelerate its convergence speed.In the second step,the mapping relationship between the physical network topology model and logical network topology model is constructed,and the process of fault diagnosis based on matrix network topology analysis is given.The fault range is diagnosed at the determined port.The virtual local area network,sparse matrix and symmetric matrix technology are used to reduce the computational complexity of the network topology matrix method.This project takes the Smart Substation Laboratory of the State Grid as an example to verify the proposed method.The parallel parsing algorithm and DOM parsing algorithm of SCD are validated and compared.Then,taking the fault of smart substation as an example,the project tests the proposed fault area diagnosis based on rough set theory and the fault port diagnosis based on matrix network topology analysis method for function and performance.The test results show that the fault diagnosis scheme for the communication network of smart substation speeds up the diagnosis speed and improves the accuracy of diagnosis.