Research On Fault Diagnosis Of High Voltage Circuit Breaker Operating Mechanism Based On Current Signal Of Opening And Closing Coil

The safe and stable operation of the power system is related to the national energy security and the lifeline of the national economy.As an extremely important primary electrical equipment,once the high-voltage circuit breaker fails,the safety and stability of the whole power system will be threatened.In recent years,the number of high-voltage circuit breakers above 220 k V put into use in China’s power system has increased year by year,and the frequency of forced outage accidents caused by high-voltage circuit breakers has also increased year by year.Therefore,the research on fault diagnosis of high voltage circuit breaker is of great significance.In this paper,the high-voltage circuit breaker fault diagnosis research is carried out based on the current signal of the opening and closing coil,and the main work is as follows:Based on the structure and closing process of high-voltage circuit breaker and the breaking,the variation process and reason of the opening and closing coil current of high voltage circuit breaker are analyzed emphatically.With this theoretical support,the opening and closing coil of high-voltage circuit breakers is simulated and modeled.On the basis of the simulation model,fault experiments such as voltage abnormality,resistance abnormality,coil inter-turn short circuit,and iron core sticking are conducted to provide data support for the fault diagnosis model in subsequent sections.In order to solve the problem of high misdiagnosis rate caused by the deviation of time domain feature point selection in the traditional circuit breaker diagnosis method based on the opening and closing coil current,this paper proposes two improved diagnostic methods: The first method combines frequency-domain and time-domain characteristics of the opening and closing coil current signals to reduce diagnostic errors caused by the deviation of characteristic point selection;the second method uses the Gramian angular field to transform the temporal opening and closing coil current signals into image data,which are then input into a two-dimensional convolutional neural network for fault diagnosis,thus avoiding diagnostic errors caused by manual feature point selection from the source.Comparative analysis of the two proposed fault diagnosis models is carried out.After parameter and structure optimization,the convolutional neural network model is slightly more accurate than the XGBoost model,and the convolutional neural network model has a higher recall rate on fault samples,which is more practical in practical production.Considering that in practical application scenarios,the impact of misdiagnosing faults as normal is far more serious than that of misclassifying normal as fault,this paper proposes a subhealth state assessment method for circuit breakers based on gray correlation analysis.The method effectively reduces the probability that the fault diagnosis model misclassifies fault-like samples as normal samples and improves the reliability of the fault diagnosis model.