Single-phase Grounding Fault Line Selection Method For Small Current Grounding System Based On Multi-Criteria Importance

Small current grounding system has been widely used in medium and low voltage distribution network system in China.However,when a single-phase grounding fault occurs,due to the complexity of operational environment and the unclarities of fault characteristics of the system leading to inaccuracy of,there is a limitation in application of single line selection methods leading to the lower accuracy of selection line results;therefore,how to select the fault line is still a challenge in relay protection field.Because different single line selection methods are suitable for different operational conditions,to improve the accuracy of selection line results,employing information technology to fuse different single methods has been widely concerned.The main idea is to establish fault measure based on the selection line results of the single methods.Furthermore,according to the fault measure,the sample data is constructed to describe the degree of line fault symptoms.Finally,the artificial intelligence methods are applied to select the fault line by extracting features from fault samples or reasoning about knowledge based on the sample.Line selection method based on fusion technology mainly focus on how to improve the ability of the artificial intelligence methods to reason about knowledge and extract features,but little attension is paid to sample data.However,the accuracy of line selection results using the artificial intelligence methods depends on the sample data;thereofore it is necessary to construct the sample data that can comprehensively reveal the fault features.In view of the importance of the data sample,this paper constructs the fault simple considering the importance of the different single methods so that the sample can completely reveal the fault characteristics under different opertrional conditions.Moreover,the fault line selection of low current single-phase grounding are studied using multiple criteria fused by support vector machine.The following work has been completed:1.Based on Simulink/Matlab platform,a distribution network system with eight feeders is established in the paper and the system is employed to simulate the several single line selection methods based on steady and transient states,and then the advantages and disadvantages of each single method are analyzed and their applicability is pointed out.2.An improved fault measurement method based on interrelation is studied to quantify the symptoms of feeder faults.Furthermore,to analyze the relationship between data and methods of fault line selection in multi-dimension,this paper chooses three algorithms: decision tree,grey correlation analysis and rough set,to quantify the importance of single fault line selection method from three perspectives of mapping relationship,change situation and uncertainty,to establish a sample data set considering the importance of different single method.3.According to the characteristics of the sample data set,support vector machine is used for comprehensive line selection.Because it is difficult to select the parameters of the support vector machine,a cross-validation parameter optimization algorithm to optimize the parameters of the support vector machine based on cross-entropy is proposed.The simulation results show that compared with the genetic algorithm and particle swarm optimization algorithm,the algorithm has strong optimization ability,fast convergence speed,and can give full play to the training and prediction ability of support vector machine model.4.A fault line selection diagnosis model based on cross-entropy parameter optimization and support vector machine is established,and the eight-feeder distribution network system is taken as the diagnosis object.The simulation results show that the proposed fault line selection diagnosis mode has stronger ability of fault identification and prediction.Moreover,considering the importance of single fault line selection method can effectively improve the accuracy of fault line selection,the influence of different combination of fault line selection methods on the accuracy of fault line selection is explored,which provides a reliable basis for using information fusion technology to select a single fault line selection method.