| With the higher requirements on the relay performance from many industries such as aerospace and ship,the performence of high voltage,miniaturization,large capacity and long life is the development direction.Relay can be sealed in the high purity protective gas,which can effectively improve the contact resistance,arc durations,electrical life and other electrical properties to enhance its capacity of arc quenching and reliability.Therefore,the study on arc characteristics in gas is particularly important.This dissertation focuses on the research of influence of gas on arc durations by the method of experiment,theoretical analysis and data fitting.Firstly,the experiment is carried out in Air,N2,et al and the conditions of 14/28/42V.Then,a quantitative mathematical model,which aims to indicate dependences of break arc durations on gas parameters such as molecular mass,viscosity,et al,is established.By the method of Multiple Linear Regression,two types of model that including eight independent parameters model and gas parameters being composite variables of the voltage and current model are got.In order to optimize the model,a neural network data fitting method is used to establish its corresponding model.By conducting the analysis on the new neural work model,each of the gas parameters value is found to follow a composite function of both the voltage and the gas parameters itself.So a new model is established,which is consistent with the experimental results.Finally,the applicability of the optimal model is explored under three different experiments.The results show that this model can predict the ability of arc quenching of gas which is better,but the applicability is relative to the experimental conditions.Generaly,the applicability is not ideal under the codition of high voltage?above 100V?,but good under the condition of low voltage?below 100V?.As a conclusion,this model has a certain guiding significance in the preferred choice of atmosphere. |
PDF Full Text Request