The Research On Power Grid Fault Diagnosis Method Based On Information Fusion

Under the background of current power big data,the scale of power grid is expanding,the interconnection between regions is becoming increasingly close,the uncertainties affecting the stable operation of power system are increasing,and the risk of large-scale blackouts caused by power grid faults is rising,which put forward higher requirements for fault diagnosis of power grid in accuracy,real-time and other aspects.The traditional fault diagnosis methods of power grid are mostly based on the switching information such as protection and circuit breaker action.However,due to many uncertainties such as rejection,malfunction and information loss of protection and circuit breaker,it is difficult to obtain accurate results for fault diagnosis only based on switching information.When power grid fault occurs,the first change is the information of voltage and current,which has incomparable advantages in accuracy,real-time and completeness.In order to solve the accuracy problem of current fault diagnosis methods for power grid,this paper introduces electrical information into power grid fault diagnosis,and combines switching information to carry out comprehensive fault diagnosis method research.Firstly,by analyzing the switching information,the quantum revolving gate is applied to realize chromosome evolution,fast convergence and global optimal search,minimal decision table is reduced,diagnosis rules are extracted,and the fuzzy Petri net model is established to carry out efficient fault diagnosis of power grid.Then,by analyzing the electrical information,the current energy fault degree,current energy distortion degree,voltage energy fault degree and voltage energy distortion degree are extracted by multiresolution analysis wavelet transform.Finally,a fault diagnosis model based on hierarchical multi-source information fusion is proposed.The fault diagnosis model takes the characteristics of switching information and the electrical information as independent evidence bodies,applies the improved D-S evidence theory for information fusion,and uses the improved fuzzy C-means decision model for cluster analysis to obtain the fault component set and realize fault diagnosis of the power grid.The simulation example analysis based on PSCAD and MATLAB software shows that this method has the advantage of high accuracy in power grid fault diagnosis,and can provide certain theoretical basis and technical support for power grid fault diagnosis and recovery under the new situation.There are 33 figures,10 tables and 82 references in this paper.