摘要
静电目标特性是静电探测和目标识别的基础。构建了一个基于USB数据采集卡的静电探测器数据采集系统,给出了低通滤波器算法的基本原理和设计实现过程。为满足高速弹目交会条件下对信号处理速度的要求,采用线性神经网络和非线性Elman神经网络设计了目标特性预测器,仿真结果表明:非线性Elman神经网络预测器能够有效给出目标特性预测结果,提高了静电探测系统实时性。
A data collecting and processing system is proposed in this paper for electrostatic target detection. A low-pass filter is designed to overcome the influence of noise The neural networks are applied to the prediction filter for the system. Simulation results show that the Elman neural network can approximate the expected signal more correctly after weight training, thus the prediction filter providing an effective way to ensure the realtime work of the electrostatic target detection system.
出处
《探测与控制学报》
CSCD
北大核心
2007年第2期20-22,26,共4页
Journal of Detection & Control
关键词
静电探测器
滤波器
神经网络预测
ELMAN
electrostatic sensor
signal processing filter
neural network prediction
Elman
