论文部分内容阅读
应用人工神经网络进行目标识别是当前模式识别的重要方法之一。前向多层神经网络及其BP算法是发展较为成熟的一种。该文对BP算法加以改进 ,使得其性能有所提高 ,收敛速度加快。针对战场监视传感器系统中处于一级警戒的地震动传感器 ,对在良好土质地面实测的人员脚步、汽车、坦克的地震动信号进行分析 ,利用小波变换和小波包分解提取能量特征 ,采用两级级连网络进行目标识别 ,识别率在 94.5 %以上
Artificial neural network for target recognition is one of the most important methods for pattern recognition. Forward multi-layer neural network and its BP algorithm is the development of a more mature. In this paper, BP algorithm is improved so that its performance is improved and its convergence speed is accelerated. Aimed at the first level of ground motion sensor in the battlefield monitoring sensor system, the ground motion signals measured by the ground and the car were analyzed. The energy characteristics were extracted by wavelet transform and wavelet packet decomposition. Even the network for target recognition, recognition rate of 94.5%