弹道目标进动周期特征提取研究

A Study on Precession-period Extraction Method of Ballistic Targets

建立了弹道目标进动回波模型,利用该模型对循环自相关法、循环平均幅度差法和方差分析法等回波序列幅度周期特征提取方法进行了比较,分析了各种方法的优缺点,提出了基于特征值熵的方差分析法。该方法通过对回波序列进行特征值分解计算特征值波形熵,然后利用方差分析法估计幅度周期。基于特征值熵的方差分析法有效地解决了上述传统方法易受噪声干扰的缺点,既保持方差分析法周期提取的精确性,又利用特征值波形熵准确判断周期性,减少了方差分析法的误判率,在低信噪比条件下具有较为稳健识别性。通过仿真实例比较了各种信噪比条件下该方法性能,并与方差分析法进行了比较,结果表明该方法具有良好的抗噪声性能。

A ballistic targets echo model is established in this paper.The traditional methods such as CAUTOC,CAMDF and variance analysis method are compared according to the echo model,and their characteristics are analyzed.Then an eigenvalue-entropy based variance analysis method is presented.This algorithm obtains the eigenvalue entropy using the eigenvalue decomposition of echo series,and the amplitude period is estimated by the variance analysis method.The presented approach can effectively maintain the precision o f the typical variance analysis method as well as decrease the error rate.This method is robust even in the low SNR condition.Finally,simulations are operated to evaluate this approach with variant SNRs.The amalysis shows that this approach has a strong anti-noise ability.