Research On Energy Storage Fault Diagnosis Method Of Spring Operating Mechanism Of High Voltage Circuit Breaker

High voltage circuit breaker(HVCB)is an important task to quickly cut off and isolate fault lines in power system.In view of its importance in the power system,accurate and timely fault diagnosis can ensure the normal operation of the power system.Energy storage fault is a common mechanical fault in high-voltage circuit breaker,which will affect the movement of the spring operating mechanism of the circuit breaker,cause too long opening and closing time,even cause refusal to move,and threaten the safety of the power system.Therefore,it is necessary to study the energy storage fault diagnosis method of this kind of equipment.However,there are the following problems in the diagnosis of this kind of fault: on the one hand,due to the limited test conditions,it is difficult to obtain the data of energy storage fault;On the other hand,the energy storage fault of the circuit breaker is mainly the energy storage spring fault,and there are three common types: the closing spring is loose,the operating mechanism is jammed,and the fastening screw is loose.The three have similar effects on the movement of the spring operating mechanism,which is prone to misjudgment.In addition,if the severity of energy storage fault can not be evaluated timely and accurately,it may lead to the further development of such fault and cause great harm to the power system.Therefore,the multi-body dynamics simulation model of the spring operating mechanism of circuit breaker is established in this paper,which solves the problem of difficult acquisition of energy storage fault data.The time domain analysis of the travel signal of the moving contact and the spring force signal of the closing spring under the three energy storage faults of circuit breaker,i.e.the closing spring is loose,the operating mechanism is jammed,and the fastening screw is loose,are carried out respectively.The corresponding energy storage fault detection models are established based on the two signals by using the support vector machine classifier.Through the comparison of the fault detection effects of the two models,The energy storage fault detection method based on the closing spring force signal of circuit breaker is determined,and the accurate detection of three common energy storage faults of circuit breaker is realized.Based on the time-domain characteristics of the closing spring force signal of circuit breaker and the multivariate linear function relationship between the three kinds of energy storage faults,an energy storage fault severity evaluation model is established,and the accurate evaluation of the energy storage fault severity is realized.Based on Solid Works,the structure of multi-body dynamics simulation model of highvoltage circuit breaker spring operating mechanism is built in this paper.According to the motion process of circuit breaker spring operating mechanism,ADAMS software is used to reasonably set the important parameters in the simulation model,such as constraint force between parts,opening and closing spring stiffness coefficient,friction coefficient between rotating mechanisms and so on.Then,the travel peak value,peak time and motion time t in the travel signal of the moving contact of the circuit breaker spring operating mechanism are compared with the theoretical and measured results to verify the accuracy of the simulation model.It is verified that the deviation between the motion characteristics obtained from the simulation model and the measured results is less than 5%.The simulation model is used to simulate and obtain the energy storage fault signals such as the travel of the moving contact and the spring force of the closing spring under the working conditions of normal circuit breaker,relaxation of the closing spring,jamming of the operating mechanism and loosening of the fastening screw,and the wavelet transform is used to preprocess the two energy storage fault signals.By extracting the motion characteristics such as the travel peak value,peak time and motion time t in the travel signal of the moving contact of the circuit breaker and the spring force characteristics such as the initial value,minimum value,minimum value time and steady value of the closing spring,as well as the statistical characteristics of the two energy storage fault signals obtained from the statistical characteristic quantity and the inter class dispersion matrix,The time domain characteristics of the two signals are analyzed respectively.Based on the support vector machine classifier,the energy storage fault detection models under the two signals are established,and the parameters of the two models are selected through cross validation.Then the two energy storage fault detection models are tested according to the simulation data.It is verified that the energy storage fault detection method based on the closing spring force signal of circuit breaker has a higher accuracy than the energy storage fault detection method based on the moving contact travel signal.Based on the time-domain characteristics of the closing spring force signal of the circuit breaker and the multivariate linear function relationship between the severity of the three kinds of energy storage faults of the circuit breaker,the severity evaluation models of the three kinds of energy storage faults are established.According to the iterative operation of BP neural network,the parameters of the three severity assessment models are selected.Finally,the three severity assessment models are tested for many times with simulation data to verify the feasibility of the assessment method.