Earticle

초록

영어

A method of feature extraction in non-stationary signals is proposed to improve the correct classification rates of plant electrical signals. The samples are composed of plant electrical signals datum which includes the period of four circumstances. For wavelets have the trait of arbitrary distinction and decomposition, eigenvector which reflect different state of plant electrical signals are extracted from different frequency segments with the technology of wavelet decomposition. Then we take them input neural network as samples to establish the model of BP neural network. Experimental results demonstrate that the classification accuracy of the proposed feature extraction method for experiment plant electrical signals is high.

목차

Abstract
 1. Introduction
 2. Basic Principles
  2.1. Wavelet Decomposition Principle
  2.2. Electric Signal of the Plant
  2.3. BP Neural Network Establishment and Training
 3. Energy Characterization Extraction Algorithm
 4. Experimental Results and Analysis
 5. Conclusion
 Acknowledgments
 References

키워드

저자

• Zhang Xiaohui [ College of Electrical Engineering, Henan University of Technology, Zhengzhou 450001, China ] Correspondence author
• Zang Haihe [ College of Electrical Engineering, Henan University of Technology, Zhengzhou 450001, China ]
• Su baoping [ College of Electrical Engineering, Henan University of Technology, Zhengzhou 450001, China ]
• Zhang zhixia [ School of Information Engineer, Henan Institute of Science and Technology, Xinxiang 453003, China ]

참고문헌

발행기관

간행물

표지보기 관심저널 등록

• 간행물명

  International Journal of Control and Automation

• 간기

  월간

• pISSN

  2005-4297

• 수록기간

  2008~2016

• 십진분류

  KDC 505 DDC 605