基于BP网络的侦察雷达目标识别方法研究

Target Identification of Reconnaissance Radar Based on Neural Network

本文旨在探讨应用于侦察雷达的一种目标识别方法。选择目标回波信号的多普勒频率作为目标特征,采用目前被广泛应用的BP网络构建人工神经网络模型,对某型战场侦察雷达的目标样本进行训练,并进行测试。在识别过程中不断对训练样本及训练目标进行更新,提高了识别正确率和识别速度。

The aim is to find a method of target identification that will be used in reconnaissance radar. Doppler frequency is chosen as the target identity in this paper. And we construct the BP neural network, which is widely used now. We update the samples and the targets in the process of the identification, train the network with these samples and then test it. The identification rate is high and it can finish rapidly.