Research on Radar Intelligent Recognition System of Space Targets Based on Multi-feature Fusion

Radar target recognition technology is an important research direction in the field of radar information processing. With the rapid development of aerospace industry, the requirements for space target recognition are constantly increasing. The fusion of multiple features of the radar information of...

Full description

Saved in:
Bibliographic Details
Published in2024 3rd International Conference on Robotics, Artificial Intelligence and Intelligent Control (RAIIC) pp. 250 - 253
Main Author Lin, Xinghan
Format Conference Proceeding
LanguageEnglish
Published IEEE 05.07.2024
Subjects
Accuracy
Analytical models
CNN
HRRP
LSTM
multi-feature fusion
Recurrent neural networks
Simulation
Spaceborne radar
Target recognition
Training
Online AccessGet full text

Cover

More Information
Summary:Radar target recognition technology is an important research direction in the field of radar information processing. With the rapid development of aerospace industry, the requirements for space target recognition are constantly increasing. The fusion of multiple features of the radar information of space targets, e.g., HRRP, time-frequency map, can extract useful information in various complex and uncertain situations, and improve recognition accuracy. This paper introduces the fusion of convolutional neural networks and recurrent neural networks. The CNN-LSTM network model is used for simulation, and the simulation results are analyzed. To verify the performance of the proposed method, training datasets and test datasets of targets with different SNRs (18dB, 8dB and - 2dB) are created. On the above dataset, the classification accuracy can reach 93.94%, which outperforms methods based on single feature.
DOI:10.1109/RAIIC61787.2024.10670861