A Rapid Diagnosis Technology Of Short Circuit Fault In DC Microgrid

DC microgrid has the advantages of low line loss,large transmission capacity,and high access efficiency of distributed power supply,which has become a hot research and application at home and abroad.In recent years,in response to the national requirement of "double carbon",DC microgrid is connected to intermittent distributed power sources such as photovoltaic power and wind power.The system structure of DC microgrid is gradually complicated and the fault diagnosis efficiency is low.Therefore,a rapid diagnosis technology for short circuit fault of DC microgrid is proposed,which can be divided into two fault diagnosis methods: with and without eigenvalue.The main research works in this article are:(1)Exploring the characteristics of DC microgrid short-circuit faults through theoretical analysis and fault simulation.Based on the topology and grounding mode of DC microgrid,the fault characteristics of common inter-pole short-circuit faults and single-pole grounding faults in DC microgrid are analyzed theoretically,different time periods experienced during fault transients are discussed in depth,a DC microgrid model is established in MATLAB/Simulink for fault simulation and field verification analysis,which provides theoretical support for the proposed fault diagnosis method.(2)For the fault diagnosis method with eigenvalues,a parameter identification-based short-circuit fault diagnosis method for DC microgrid is proposed.The classification criterion is constructed to identify the fault type using the transient current change rate at the first end of the line,and the converter fault location criterion is constructed to determine the location of the faulty converter.Multiple objective functions are constructed,a multi-objective optimization algorithm based on genetic algorithm is used to identify the fault line parameters,the line parameters are used for fault location,and the experiments show that the method is practical and effective.(3)For fault diagnosis method without eigenvalue,a short circuit fault diagnosis method for DC microgrid based on LSTM neural network is proposed by combining deep learning methods,extracting the current data of each positive line and negative line first end after the fault occurs under various fault conditions,constructing a feature database,building an LSTM neural network classifier,automatically learning and extracting fault state features,short-circuit fault classification and interval location,no longer need to extract fault features,greatly improving the fault diagnosis efficiency,and finally the practicability and effectiveness of the proposed method are verified by experiments.(4)Based on the above research,this paper develops the DC microgrid short-circuit fault diagnosis software by using the GUI in MATLAB software,which includes the login interface and short-circuits fault diagnosis interface,which is easy for users to operate directly and has high reliability.