基于LIBS和遗传算法的畜禽粪便中钙含量研究

发布时间：2018-12-16 07:29

【摘要】：为实现畜禽粪便高效、安全资源化利用,快速、准确定量表征分析畜禽粪便中钙(Ca)元素含量具有重要意义。以我国畜禽养殖比较集中地区的代表性畜禽粪便样本为对象,探讨了LIBS快速、定量分析其Ca元素含量模型构建和通过GA筛选变量优化模型的可行性。采集光谱的仪器参数设置如下:采集点数、激光能量、压制样压力、光斑大小和延迟时间分别为80个,15%,20T,400μm和1μs。建模结果表明:采用基于Ca元素特征波长所建三种线性模型精度较低;基于全谱的PLS所建模型中,R2v和RPD分别为0.85和2.13;选用化学计量学方法-遗传算法进行变量筛选可显著减少建模变量信息,提高建模效率,经GA优选所得12个变量所建PLS模型中R2v和RPD为0.90和3.04,具有较高精度和实际应用价值。研究表明,将激光诱导击穿光谱技术和遗传算法相结合可定量分析畜禽粪便中Ca元素含量。
[Abstract]:In order to realize efficient and safe utilization of livestock and poultry feces, rapid and accurate quantitative analysis of calcium (Ca) content in livestock and poultry feces is of great significance. Taking the representative samples of livestock and poultry feces from the concentrated areas of livestock and poultry breeding in China as the object, the feasibility of LIBS rapid, quantitative analysis of Ca element content model and optimization model of screening variables by GA was discussed. The instrument parameters are set as follows: collecting points, laser energy, pressing pressure, spot size and delay time are 80, 1520TX 400 渭 m and 1 渭 s, respectively. The results show that the accuracy of the three linear models based on the characteristic wavelength of Ca elements is low, the R2v and RPD are 0.85 and 2.13, respectively. The selection of chemometrics method-genetic algorithm for variable screening can significantly reduce the information of modeling variables and improve the efficiency of modeling. The R2v and RPD of PLS model established by GA are 0.90 and 3.04, respectively. It has high precision and practical application value. The results showed that the combination of laser induced breakdown spectroscopy and genetic algorithm could be used to quantitatively analyze the content of Ca in livestock and poultry feces.
【作者单位】：中国农业大学工学院;
【基金】：欧盟框架计划项目(690142) 国家重点研发计划项目(2016YFE0112800)资助
【分类号】：X713

【相似文献】