Research On Intelligent Matching Method For Radar Antenna Components Based On Telecommunications Performance

As an important armament product,phased array radar antennas have not yet fully automated production due to their complex structure,high assembly difficulty,and strict assembly quality requirements.Their assembly work still relies on a lot of manual work,resulting in low assembly efficiency and qualified assembly.Problems such as low rate,poor quality stability,and widespread rework have seriously affected the development cycle.Therefore,it is of great significance to carry out research on intelligent assembly technology for the needs of intelligent assembly of radar antennas.This article takes radar telecommunications performance as the starting point,analyzes historical assembly data and combines production practices,integrating hierarchical mixture Gaussian model,LSTM model,and genetic algorithm to carry out the optimization design of radar antenna unit assembly process parameters,radar antenna component telecommunications performance prediction and the optimization of radar antenna assembly scheme to improve the assembly efficiency and quality of the radar antenna.Aiming at the problem that the telecommunications performance of the radar antenna unit does not meet the requirements after the assembly is completed,the assembly process parameter optimization design is based on the production data of the radar unit;for the problem that data noise will affect the analysis result,the feature tree algorithm is introduced to remove the data noise;aiming at the problem of redundant parameters occupying computing resources,the assembly process parameters are screened based on the random forest algorithm;in order to optimize the assembly process parameters of the antenna unit,the assembly data samples are clustered based on the Hierarchical Mixture Gaussian Model(HGMM)algorithm and the 3? guidelines are used to optimize the design Work.The antenna unit assembly process parameter optimization scheme will be verified based on the XGBoost performance prediction module.Aiming at the problem of not being able to optimize the assembly plan of antenna components directly for telecommunication performance,a radar antenna component performance prediction model is established based on the LSTM neural network.The component assembly process parameters are used as input,and the telecommunications performance of the component is output to achieve the quality of the selection scheme.At the same time,the performance judgment standard setting of the radar antenna unit is realized based on the performance prediction model.Aiming at the problem of decreased prediction accuracy caused by the high complexity of sample data and over-fitting of the model,on the basis of the gray correlation analysis model,combined with the actual assembly situation of the radar antenna components,the Shannon gray correlation model is established by fusing the information entropy theory to screen the key assembly process parameters of the components,and the possibility of model overfitting is reduced through parameter screening.Aiming at the problem of formulating the selection and matching plan of radar antenna components,an optimization model of the selection and matching plan of radar antenna components is constructed based on the genetic algorithm of tribal and generation elimination.The model uses the performance prediction results of the aforementioned LSTM model as the assembly constraint,and the assembly qualification rate is the final goal;the optimization of the matching scheme is realized based on the genetic algorithm of sub-tribe and generation elimination,and the effectiveness of the algorithm is finally proved through the matching experiment.Based on the research of the above theoretical methods,combined with the actual assembly production requirements of radar antenna components,design and develop a radar antenna component selection prototype system,provide intelligent decisionmaking for actual production,and provide powerful tool support for the intelligent assembly of radar antenna components.