摘要
为了有效地对植物电信号进行分类,提出了基于多种小波熵与信号熵的特征提取方法。小波熵由于结合了小波变换和信息熵理论的优势,能够快速、准确地提取植物电信号的特征;但是,由于植物电信号的非平稳性和多样性,依靠单一的小波熵可能出现分类困难和分类不准确等问题,需要结合多种小波熵和信号的熵信息进行特征提取和分类。为此,以4类干旱胁迫下的君子兰叶片信号为对象,对提取的特征向量利用KNN方法进行分类。试验结果表明,该方法能够对君子兰叶片的电信号进行有效识别,为植物电信号的识别提供了一种可行的新方法。
To solve the problem of diagnosis for plant electrical signals,a classification approach based on multi-wavelet entropy and signal entropy feature extraction is proposed.Wavelet entropy can pick up the signal characteristic quickly and exactly because it combines together the advantages of wavelet transform and Shannon entropy.But signal identification based only on single wavelet entropy may cause difficult or inaccurate results because of the non-stationary and diversified plant electrical signals.Therefore,several different wavelet entropies and signal entropies are extracted as eigenvectors.The experiment results show that this diagnosis method can recognise the electric signals of the laminae of Clivia effectively,so it is a feasible method for plant electrical signals diagnosis in quantification.
出处
《农机化研究》
北大核心
2013年第9期38-40,47,共4页
Journal of Agricultural Mechanization Research
基金
河南省高等学校青年骨干教师项目(2010GGJS-220)
关键词
植物电信号
多小波熵
信号熵
特征提取
plant electrical signals
multi-wavelet entropy
signal entropy
feature extraction
