A Fault Location Method For Small Resisitance Grounded Distribution Network Based On Information Fusion

With the continuous expansion of the distribution network and the gradual increase in the number of cables,the traditional neutral point ungrounded system has a larger and larger capacitive current to the ground,and the arc cannot be extinguished by itself,and the overvoltage generated is extremely high,which may easily cause equipment damage.The low resistance grounding system can effectively solve this problem,so the system is widely used.After a fault occurs,it is necessary to accurately locate the fault point of the small resistance grounded distribution network,solve the fault,and restore the power supply.Because it is difficult to obtain accurate results for fault location based on single fault information,this paper proposes a small resistance grounded distribution network fault location method based on the fusion of electrical and switching values.First of all,in view of the high cost and redundant data of the fault location collection device based on electrical quantities,a method of compressing sensing to process electrical quantities is designed.Use the synchronous phasor measurement device(PMU)to collect the electrical quantity information before and after the fault.The positive sequence voltage sag is obtained by calculation,combined with the node impedance matrix,reconstructed using compressed sensing algorithm to divide the fault area.The positive sequence voltage sag vector of the adjacent first-order bus is solved,and the reconstruction is performed again to obtain the fault current,which is transformed to obtain the electrical fault degree.Then,the switch value is collected by the Supervisory Control And Data Acquisition(SCADA).Determine the nodes as components,relay protection and circuit breakers,and build the Bayesian network topology according to the principle of relay protection.According to a large number of historical operating data and equipment data,the nodes are assigned,and the probability of failure of each component’s switching value is obtained by probabilistic reasoning,and it is converted into a switching value failure degree.Finally,the combination rules of evidence theory are used to fuse these two fault degrees to obtain the fault location.Based on the traditional DS evidence theory information fusion,this paper uses a multi-sensor data fusion algorithm based on the enhanced belief divergence measure for information fusion.Two examples are set up in the IEEE33-node system to verify the effectiveness of the proposed method,and the two fusion methods are compared and analyzed to improve the accuracy and accuracy of fault location based on information fusion.