摘要
从最大熵谱估计入手,用TH网络实时地实现了对船用惯导系统姿态测量信号的最大熵一维谱估计。研究表明,由于神经网络的高度运算能力,该方法有效地解决了信号频谱处理过程中处理性能与运算量之间的矛盾,为现代谱估计提供了新途径。
Based on the theory of maximum entropy spectral analysis ,a linear program neural network (TH Neural Network)is proposed in this paper ,which is used to analyse the attitude signal spectral for marine inertial navigation system.It is shown that owing to the high calculation ability of the ANN,this method is effective in solving the contradiction between the performance and the calculation volume in analysing signal spectral,and it supplies a new way for modern spectral analysis.
出处
《船舶工程》
CSCD
北大核心
1999年第3期44-45,56,共3页
Ship Engineering
基金
中船总国防科技应用
基础研究基金
关键词
人工神经网络
惯导系统
舰船
姿态信息谱
ANN,Maximum entropy spectral analysis,Inertial navigation system