Fault Diagnosis Of Power Grid With Information Fusion Based On Fuzzy Petri Net And Improved Hilbert-Huang Transform

With the rapid development of China’s economy,the demand and dependence of the whole society on electric power is increasing day by day,which makes the power grid interconnected closely,the structure becoming more and more complex,and the uncertain factors of operation stability increase.At present,the rapid development of power grid automation makes the fault information quickly get to realize,when complex fault occurs,the dispatch center will obtain a lot of alarm and fault information in the short term,beyond the processing capacity of grid staff,it is difficult to judge the fault quickly and accurately,and the false positives and false negatives occur.Therefore,it is necessary to rely on real time and efficient power system fault diagnosis system to provide decision reference.It can provide auxiliary basis for dispatcher decision-making,enhance the reliability of power supply,reduce the time of electric energy recovery,improve the stability of power grid,and diagnose the fault line accurately and in real time.This article first analyses various fault information acquisition approaches for current fault diagnosis of power grid,The second chapter puts forward the traditional method of fault diagnosis and fault element identification using single switch information,will be subject to protection and circuit breaker tripping or malfunction and information transmission in the process of loss and other uncertain factors,leading to inaccurate diagnosis result,in order to solve the problem,this chapter uses the fuzzy Petri net the model of fault diagnosis and fault tolerance is improved.The fuzzy Petri net model to build a simple and reliable diagnosis speed and fault tolerance is improved,but the complexity of failure,will also appear misjudgment phenomenon,need to be further combined with electrical quantities analysis;the third chapter based on the single switch fault diagnosis the problem of feature considering the variation of electrical fault information after,and Hilbert-Huang transform in traditional electric characteristic extraction In the process of mixed mode fault diagnosis problem leads to overlapping,envelope error and the end effect of defects is not accurate,and puts forward the method of Hilbert Huang transform improved,and an example is used to verify the effectiveness of the algorithm,but the complex fault occurs,the fault probability characterization of electric single gas may appear identification the degree is not high,can quickly and accurately determine the fault components;the fourth chapter in view of the above problems will switch and electrical fusion fault diagnosis based on improved D-S evidence theory,combination switch fault diagnosis in actual physical meaning and the amount of electricity from gas,modify the original evidence and modify the two combination rules of D-S evidence theory to solve the evidence conflict distribution unreasonable;the fifth chapter based on the above algorithm,the multi source information fusion algorithm With the support of the corresponding software framework,the ring network experiment platform and algorithm in the laboratory are combined to build a set of fault diagnosis decision system that can provide assistant decision for dispatchers.