河北大学学报(自然科学版) ›› 2021, Vol. 41 ›› Issue (6): 666-671.DOI: 10.3969/j.issn.1000-1565.2021.06.006

• • 上一篇    下一篇

基于GA-BP算法的太赫兹波鉴定苜蓿草品种

王芳1,张玉1,张春红1,夏红岩2   

  • 发布日期:2021-12-08
  • 通讯作者: 夏红岩(1966—)
  • 作者简介:王芳(1969—),女,辽宁盖县人,中国石油大学(北京)教授,主要从事光学与固体力学研究.
    E-mail:wangfang6402@163.com
  • 基金资助:
    内蒙古自治区科技计划项目(201602054)

Identification of alfalfa varieties by terahertz wave based on GA-BP method

WANG Fang1, ZHANG Yu1, ZHANG Chunhong1, XIA Hongyan2   

  1. 1.College of Science, China University of Petroleum, Beijing 102249, China; 2. Inner Mongolia Grassland Station, Huhehot 010020, China
  • Published:2021-12-08

摘要: 基于太赫兹时域光谱技术(THz-TDS)测得8类苜蓿草种的折射率光谱数据,利用遗传算法优化BP神经网络(GA-BP)模型对苜蓿草种的分类进行了研究.结果表明GA-BP模型对8个苜蓿品种的平均分类准确率达到94%,且对单个品种的分类准确率最高能达到94.6%.本研究为牧草品种的分类鉴别提供了一种新方法,对种质资源的鉴别也具有重要的参考价值.

关键词: THz-TDS, GA-BP, 苜蓿草种, 分类

Abstract: Based on the refractive index data of alfalfa species measured by the terahertz time domain spectroscopy(THz-TDS), the genetic algorithm optimized BP neural network(GA-BP)model was used to identificate alfalfa species. The results show that the average classification accuracy rate of 8 varieties is 94%, and the maxial classification accuracy rate of a single variety is 94.6%. A new method is provided for the classification and identification of alfalfa species, which provides important reference value for the identification of germplasm resources.

Key words: THz-TDS, GA-BP, alfalfa species, classification

中图分类号: