一种改进的自适应Chirplet分解方法及其实现方法(英文)  

A New Modified Chirplet Decomposition Algorithm and its Implementation

作　　者：王殿伟[1] 李言俊[2] 张科[2] 郭俊锋[2] 

机构地区：[1]西安邮电学院,西安710061 [2]西北工业大学航天学院,西安710072

出　　处：《火力与指挥控制》2011年第6期13-18,共6页Fire Control & Command Control

基　　金：国家自然科学基金(60575013);航空科学基金(20070153005);航空支撑基金资助项目(07C53007)

摘　　要：参数自适应时频分布成为当前非平稳信号分析与处理的研究热点。针对自适应Ch irp let分解所用基函数的时频关系为线性,对含有非线性时变分量的信号逼近效果不好的问题,提出了一种改进的自适应Ch irp let分解方法,并给出其快速实现算法。首先在Ch irp let时频原子的基础上添加了弯曲算子,构造了新的时频原子,得到了一种改进的自适应Ch irp let分解算法;然后,利用匹配追踪算法做为快速分解实现方法,并给出了具体实现步骤。经仿真信号和蝙蝠回声定位信号验证表明,本算法在收敛性和准确性等方面具有优势。The parametric adaptive time-frequency distribution is an Important method in the analysis and processing of proeessing of nonstationary signals. Chirplet decomposition only has linear time-frequency relationship basis function so it can not approach signals with nonlinear time varying components accurately. Aiming at overcoming this limitation, a new modified Chirplet deeomposition algorithm based on traditional Chirplet decomposition was presented method. Firstly added bending operator to Gaussian Chirplet atom, the new modified Chirplet atom was eonstructed. Secondly used this new atom, the modified Chirplet decomposition algorithm was proposed. Thirdly the matching pursuit algorithm was used as fast implementation method to search the parameters of the optimal time-frequency atoms. Finally results of simulation signals and actually measured bats eeholocation signals had demonstrated the superiors in eonvergenee, aecuraey and resolution of the algorithm proposed in this paper.

关 键 词：时频分析 参数自适应时频分布 时频原子 改进Chirplet分解 匹配追踪 

分 类 号：TN911.72[电子电信—通信与信息系统]