Parameter Identification And Life Prediction Of Aluminum Electrolytic Capacitor In Converter

With the rapid development of science and technology,power electronic equipment plays an important role in various fields of society,and the requirements for its safety and reliability are also gradually improved.Current converter is the most commonly used power electronic equipment,and has been widely used in various field.The bus aluminum electrolytic capacitor is one of the components of the converter which is easy to fail.In this thesis,the bus aluminum electrolytic capacitor of the converter is taken as the research object.On the basis of analyzing the basic characteristics and aging mechanism of aluminum electrolytic capacitor,researches on the parameter identification and life prediction of aluminum electrolytic capacitor are carried out.The specific research contents are as follows:(1)Firstly,the basic characteristics,aging mechanism and influencing factors of aluminum electrolytic capacitor are introduced.The impact of capacitor ripple current and ambient temperature on the aging of aluminum electrolytic capacitor is emphasized.The characteristic parameters of aluminum electrolytic capacitor that need to be monitored are determined,which lays a foundation for the parameter monitoring and life prediction of aluminum electrolytic capacitor.(2)Secondly,according to the equivalent model of aluminum electrolytic capacitor,an experimental platform for parameter identification of bus capacitance is built,including hardware system and software system.The ripple voltage,ripple current and temperature signals of the capacitor are extracted by the sensor installed on the converter,and then transmitted to the upper computer by the NI USB-6002 data acquisition card.The aluminum electrolytic capacitor data monitoring system based on Labview is designed.The capacitance parameter identification system is constructed based on five-fold cross multilayer perceptron.The surface temperature of the capacitance,ambient temperature,ripple current and ripple voltage of the capacitance are used as the input of neural network,the capacitance value and the equivalent series resistance value are the output of neural network.The parameter identification of the converter bus capacitance is realized,and experiments are performed to verify the feasibility of the method.(3)Finally,the factors affecting the service life of aluminum electrolytic capacitor and failure criteria are analyzed,and on this basis,use the Least Squares Support Vector Machine to predict the change trend of the value of capacitance and equivalent series resistance,so as to realize short-term prediction of the rest service time of capacitance.Experiments are performed to verify the feasibility and effectiveness of the method.Based experiments and correlation analysis,the following conclusions are drawn: in this thesis,by adopting parameter identification and life prediction,online monitoring for the status of the busbar capacitance of the converter is achieved,and life prediction can be realized by the recognized characteristic parameters of capacitance,so as to improve the reliability of converter.The method has good engineering application value.