Research On Feeder Automation Mode Of Distribution Network With Distributed Generation

As far as the overall development level of power system is concerned,the development of feeder automation technology in distribution network is relatively slow,and problems such as long fault diagnosis time and poor adaptability of diagnosis methods can not be solved well.Based on the understanding of distribution network structure,communication mode,intelligent terminal and other related technologies,feeder automation is divided into local control mode,centralized intelligent mode and intelligent distribution mode according to the need of communication and different communication modes.In order to solve the problems existing in the feeder automation technology,the medium voltage distribution network is taken as the object,and puts forward improvement measures for the fault diagnosis methods of feeder automation under different modes.Aiming at the traditional matrix algorithm which is widely used in the centralized intelligent mode,an improved matrix algorithm based on the segment description matrix is proposed.The algorithm can reduce the dimension of matrix and the amount of calculation.The fault identification matrix simplifies the identification steps and methods,and the fault criterion is simple without normalization.In order to solve the problem of long fault processing time in the centralized intelligent mode,the reclosing protection at the outgoing circuit breaker in the local control mode will be combined with it to realize the collaborative feeder automation of instantaneous cut-off and rapid diagnosis.The improved matrix algorithm has poor fault tolerance and low accuracy in the case of abnormal information such as misinformation and missing transmission and multiple complex faults.In view of the disadvantages of the switch quantity diagnosis method in this respect,a multi-source information fusion diagnosis method based on the switch quantity and the electric quantity information is proposed.The wavelet theory is used as the fault feature analysis method of the electric quantity,and the improved D-S evidence theory is used to fuse the information of the fault degree evidence body to obtain the fusion fault degree of the fault elements.Finally,the fuzzy c-means clustering algorithm is used to improve the fusion result,the fusion diagnosis results are obtained by making diagnosis decision.On the basis of collaborative feeder automation,a fusion diagnosis layer is added to construct a hierarchical feeder automation with fast isolation,fast diagnosis and fusion diagnosis capabilities,so as to ensure the fault diagnosis tolerance and anti risk capability of distribution network feeder automation in complex and abnormal situations.In addition,aiming at the intelligent distribution mode of point-to-point communication mode,an improved locking mode based on logical value information judgment is proposed,which combines the fault information provided by the upper and lower level protection to study and judge the fault uniformly,so as to improve the accuracy and adaptability of the protection action.The improved way of feeder automation proposed in this thesis takes into account the situation of grid connection of distributed power supply,and has good adaptability to multi power supply or distribution network with distributed power supply.Combined with specific cases,PSCAD/EMTDC simulation environment and Matlab data processing platform are used to simulate and analyze different fault conditions.The results show that the improved method can effectively improve the fault diagnosis ability of distribution network feeder automation.