Research on Radar Detection Range Conversion Based on Neural Networks
WANG Junyi
YANG Zichen
Abstract：A systematic study is conducted on the radar range conversion problem of five types of target models,Marcum and Swerling Ⅰ～Ⅳ,under the condition of constant false alarm probability Pf and variable detection probability Pd.Thes paper calcu-lates the distance conversion coefficients of different targets through radar equations,and then discretizes the Pd=f(SN R,N)formula for five types of targets to obtain a train dataset.Finally,MLP and MLP improved based on DeepONet are used to fit the relationship between the first five terms and conversion coefficients.The neural network model obtained after fitting has a normalized L2 loss val-ue below 0.004 when the pulse number N is between 0 and 150.By making certain improvements to the general MLP,its fitting per-formance can be improved,and the fitting effect of the neural network on this dataset meets the requirements for the accuracy of de-tection distance estimation in practical engineering.
Keywords：radar target modelradar detection rangeMLPDeepONet
Publication Date：2025-08-20
Online Publishing Date：2025-12-12(First online date of this platform, not the publication date of the document)
Pages：6( 89-93,162 )