Research On Partition Fault Diagnosis Of Power Grids Based On Deep Learning And Grey Relational Degree

With the development of energy and power system,renewable energy grid and the access of electrical equipment,the scale of power grid is becoming larger and larger.The network topology of power grid is becoming more and more complex,which brings difficulties to fault diagnosis.In the case of power grid fault,the fault location and fault type identification are realized in time to ensure timely troubleshooting and rapid recovery of power supply and consumption.Most of the power grid fault diagnosis methods mainly use the fault switching value and fault electrical value as the input for fault diagnosis.Therefore,this paper studies the power grid fault diagnosis technology from the fault switching value and fault electrical value for different problems.The structure of large power grid is complex,large amount of data,and the timeliness of fault diagnosis is hard to guarantee.Starting from the fault switching value,a power grid partition fault diagnosis based on improved probabilistic neural network and gray relational analysis is proposed.Firstly,the large power grid divided into small areas for fault diagnosis through power grid partition,which reduces the difficulty of fault diagnosis.Then the PNN diagnosis module is established by the PNN optimized by GA-CPSO for diagnose the power grid fault.Finally,the faults in the overlapping area are reanalyzed by the GRA method,in order to realize the accurate fault diagnosis of the whole power grid.The feasibility and effectiveness of the method are analyzed by two cases.The diagnosis results show that the method can effectively identify the faults in the non overlapping area and the overlapping area,and has strong fault tolerance and high diagnosis accuracy.In this paper,a method of fault location and fault type identification of power grid based on variational mode decomposition and convolution neural network is proposed.Aiming at the problem that the fault feature extraction method will directly affect the fault accuracy in the process of fault diagnosis,a fault feature extraction method based on VMD and Hilbert-Huang transform(HHT)is designed.In this method,the VMD is used in the analysis the characteristic features from fault transient signals of the positive sequence current.After the intrinsic mode function(IMF)is obtained by VMD and the IMF component with more fault features is selected.The fault features of IMF is extracted through HHT.The extracted fault feature vector is used as the input of CNN to build fault diagnosis model.Finally,the fault diagnosis report is obtained by comparing and analyzing the output results of Soft Max layer.To evaluate the accuracy of the proposed method,the hybrid method is tested by small current grounding power system model of relay protectiondynamic simulation equipment.The experimental results show that this method can realize the fault location and fault type identification,and has good accuracy and generalization ability.Meanwhile,the method is less influenced by fault resistance and fault distance.