Font Size: a A A

Research On Storage Life Prediction Method Of Spaceflight Electromagnetic Relay Based On Data Driven

Posted on:2020-06-27Degree:MasterType:Thesis
Country:ChinaCandidate:J W WangFull Text:PDF
GTID:2392330590479125Subject:Control engineering
Abstract/Summary:PDF Full Text Request
Aerospace electromagnetic relay is one of the main basic electrical components for control,information transmission and switching in defense weapon equipment systems and aerospace systems.It is widely used in aviation,aerospace,defense and military engineering,and bears signal transmission and execution control.Functionality,which requires a very high level of reliability.In the field of military equipment,modern weapons and equipment often undergo a long-term storage process,that is,they have the characteristics of “long-term storage,one-time use”.Therefore,accurately predicting the storage life of a relay plays a key role in ensuring that military equipment maintains good tactical performance over its full life span.Based on the research status at home and abroad,this paper starts from the aspects of life test,test data analysis and theoretical modeling analysis,and establishes a complete storage life prediction program for aerospace electromagnetic relays.Firstly,the structure of aerospace electromagnetic relay is introduced.The three basic elements of storage failure environmental stress,failure mode and failure mechanism are studied.Aiming at the storage degradation characteristics of aerospace electromagnetic relays,it is determined that the temperature is the main environmental stress causing failure,and the constant stress accelerated degradation experiment is carried out under different stress levels.The experimental sample quantity and experimental test data are given,the experimental scheme is determined,and the software and hardware system of the test with the resistance test module and the time test module as the core is designed for the experimental scheme.After obtaining the experimental data through the storage accelerated degradation experiment,the visual analysis was carried out by using the scatter plot and the box plot diagram,and the conclusion that the contact resistance has an upward trend and the pull-in time has a downward trend is obtained.Then,the correlation between the contact resistance and the pull-in time is further analyzed using statistical analysis to determine the contact resistance and the pull-in time as sensitive parameters for predictive analysis.Because the original experimental data inevitably has noise,this paper uses the wavelet denoising method to preprocess the original experimental data.From the aspects of threshold function,wavelet base,soft and hard threshold,etc.,the wavelet denoising method suitable for contact resistance is determined.Considering the random influence of environment in the process of aerospace electromagnetic relay degradation,the applicability of stochastic process in this problem is verified by autocorrelation method.The stochastic process-based degradation model is established by using contact resistance as random degradation quantity.The maximum likelihood estimation method is used to estimate the parameters of the model,and the storage life model of the aerospace electromagnetic relay based on the failure threshold is constructed.The storage life of the aerospace electromagnetic relay at room temperature is accurately predicted.Combining with grey theory,all stochastic processes can be regarded as the view of the gray process,and based on the theory of gray and the theoretical characteristics of neural networks: high-quality prediction for the non-linear mapping problem of less data and poor information,thus constructing a gray neural network Storage life prediction for aerospace electromagnetic relays.However,the model still has the risks of local optimization and convergence.The fruit fly optimization algorithm is introduced to optimize the initial weight and threshold of the neural network,and the Drosophila gray neural network prediction model with faster optimization process is established.After the optimization of the fruit fly algorithm,the prediction model has higher prediction accuracy and shorter training time,which can effectively predict the contact resistance growth during the storage process of the aerospace electromagnetic relay.The prediction accuracy of the above life prediction model also illustrates the effectiveness of the experimental method for the storage life degradation of aerospace electromagnetic relays designed in this paper.
Keywords/Search Tags:Aerospace electromagnetic relay, Storage Life Degradation Experiments, Wavelet denoising, Stochastic process, FOA-GNN
PDF Full Text Request
Related items