Machine Learning and Intelligent Communications. Third International Conference, MLICOM 2018, Hangzhou, China, July 6-8, 2018, Proceedings

Research Article

Study of Radar Target Range Profile Recognition Algorithm Based on Optimized Neural Network

Download
100 downloads
  • @INPROCEEDINGS{10.1007/978-3-030-00557-3_61,
        author={Xiaokang Guo and Tao Jian and Yunlong Dong and Xiaolong Chen},
        title={Study of Radar Target Range Profile Recognition Algorithm Based on Optimized Neural Network},
        proceedings={Machine Learning and Intelligent Communications. Third International Conference, MLICOM 2018, Hangzhou, China, July 6-8, 2018, Proceedings},
        proceedings_a={MLICOM},
        year={2018},
        month={10},
        keywords={1-D range profile recognition LVQ (Learning Vector Quantization) PSO (Particle Swarm Optimization)},
        doi={10.1007/978-3-030-00557-3_61}
    }
    
  • Xiaokang Guo
    Tao Jian
    Yunlong Dong
    Xiaolong Chen
    Year: 2018
    Study of Radar Target Range Profile Recognition Algorithm Based on Optimized Neural Network
    MLICOM
    Springer
    DOI: 10.1007/978-3-030-00557-3_61
Xiaokang Guo1,*, Tao Jian1,*, Yunlong Dong1,*, Xiaolong Chen1,*
  • 1: Naval Aviation University
*Contact email: gxk157@163.com, work_jt@163.com, china_dyl@sina.com, cxlcxl1209@163.com

Abstract

Neural network as an important aspect of artificial intelligence has received extensive research and long-term development. Radar target range profile recognition is a commonly used method in radar target recognition, in this paper, it is combined with neural network. The LVQ (Learning Vector Quantization) neural network has excellent classification and identification capabilities. This paper applies it to radar target one-dimensional range image recognition and achieves good results. This paper studies the problem of LVQ neural network sensitive to initial connection weights, and uses PSO (Particle Swarm Optimization) algorithm to optimize it of recognition classification. The experimental results show that the study of radar target range profile recognition algorithm based on optimized neural network can overcome the sensitivity of the LVQ neural network to the initial weight and improve its recognition ability.