摘要
基于α－β滤波的处理思想，用神经网络对目标数据进行滤波处理，不断调整神经网络的连接权值，经过多次训练，使目标滤波输出反映目标航迹的变化规律。仿真结果表明，神经网络方法用于船用雷达平滑跟踪是可行的，其跟踪误差是收敛的，而且通过合理选择网络参数，可使跟踪误差很小。
On the basis of α-βfilter,this paper is about the process of the target data in neural net-work method.The network adjusts the connected weight of neural network and makes theoutput of target filter reflect the changing pattern of target tracking by training.The simulat-ed result shows that it is feasible that neural network method is used in navigation radarsmooth tracking.The tracking error is convergent and very small by selecting network param-eter reasonably.
出处
《大连海事大学学报》
CAS
CSCD
1995年第4期25-28,共4页
Journal of Dalian Maritime University
