Investigation On Condition Monitoring And Fault Diagnosis For High-voltage Circuit Breakers Using Multiple Parameters

High-voltage circuit breakers(HVCBs)are key equipments in the power system,and they play a dual role of protection and control in the power network.The HVCBs can break the working current when they run in a normal state,while they are requested to break the short-circuit current and the overload current in the allotted time.Therefore,to guarantee the normal operation of the HVCBs is very important to improve the reliability of the power grid.In recent years,the emergence of artificial intelligence methods(AIM),making the state monitoring based on HVCBs' maintenance strategy developing rapidly.The use of the AIM improved the accuracy of condition monitoring and fault diagnosis of the HVCBs.In this research,we mainly study the mechanical properties of the SF6 HVCBs.By installing multiple sensors on the HVCBs to obtain data of various state parameters,we can extract the information that can reflect the mechanical properties of the HVCBs.Then information extracted can be used to evaluate the working status of the HVCBs and diagnose whether there exist a mechanical failure.Therefore,the basis of this study is to determine the monitoring parameters,select the type of sensors to measure these data signals and select how these sensors is installed.In this research,the probability of occurrence of multiple faults in the operation mechanism is classified statistically.Combined with the statistical data of the literature to determine the main monitoring parameters in this study.The monitoring parameters mainly including the trip/close coil current signal,the energy storage motor current signal and the vibration signal,and so on.Through analysing the characteristics of the amplitude and time distribution of the above monitoring parameters,the appropriate sensors and data acquisition modules are chosen.Followed by that,a test platform is build in our laboratory,so a variety of operating mechanism failure mode are simulated,from which we captured the multi-parameters data.According to the characteristics of the monitoring data,the contour method,multi-layer wavelet packet decomposition and reconstruction method,power spectrum analysis method and signal energy envelope statistics method are used to denoise the signal and extract the eigenvalues.Then the multi-dimensional mapping database between the characteristics of the multi-parameters and the status of the HVCBs is established,in order to achieve an accurate identification of the status of the HVCBs.The main fault components of the HVCBs include the control system,the operating system and the connecting rod system.The operating status of each component is related to the changing trend of the multi-parameters.Based on the state recognition above,a method of condition monitoring and fault diagnosis of the HVCBs based on multi-parameters and multi-dimensional mapping is developed by combining the particle swarm optimization(PSO)algorithm and the support vector machine(SVM)theory.At last,through integrating the state monitoring and fault diagnosing algorithm with the graphical user interface(GUI),a set of application software which has a good visual interface and operability is developed,which is the multi-parameters integrated system software for condition monitoring and fault diagnosis of the HVCBs.With the developed software,on the one hand,the energy storage motor current signal,the trip/close coil current signal and the vibration signal captured above can be analysed and processed,and the operation state of the HVCBs can be determined according to the signal processing results,on the other hand,by analysing the SOE signal exported from the EMS/SCADA system of the Oracle database,the inventory data can be analyzed and processed.At last,we can determine the history operating status of the HVCBs.In this research,part of the theoretical research based on multi-parameters of the on-line monitoring and fault diagnosis system of the HVCBs is finished,the design and development of the hardware and software of the acquisition unit based on the CompactRIO is finished.