宽带毫米波雷达目标时延神经网络识别新方法

基于一维距离像和神经网络研究宽带毫米波雷达目标识别问题，研究了用于雷达距离像序列识别的时延神经网络模型及其学习算法，并提出了基于距离像序列的宽带雷达目标时延神经网络识别方法，还利用三种飞机缩比模型的暗室测量数据，研究了时延神经网络分类器中时延单元数目对分类精度的影响以及分类器的分类性能。实验结果表明：本方法具有良好的应用前景。

NEW TARGET RECOGNITION METHOD OF WIDE-BAND MILLIMETER WAVE RADAR BY USING TIME DELAY NEURAL NETWORK

Target recognition by using neural network based on high resolution range profile (HRRP) with wide band millimeter wave (MMW) radar was researched. Time delay neural network (TDNN) model and its learning algorithm to range profile sequence input were studied first, the TDNN target recognition method based on range profile sequence with wide band MMW radar was proposed. The effect of time delay unit number on classification precision and the performance of TDNN classifier using three typical aircraft dark room data measured with scale model were studied. The results show that the proposed method has good application prospects.