近红外光谱在预测果品硬度和表面色泽中.doc

近红外光谱在预测果品硬度和表面色泽中 变量筛选方法的研究 专 业：2007测控-1 学 号：20070310110118 学生姓名：范徽 指导教师：郝勇 摘要 近年来随着人们生活水平的不断提高，水果的消费量也呈上升趋势，大批量的水果在进入销售环节时，首先需要对其成熟度进行评判。对于成熟度的评价指标，最常用的方法就是从果品的硬度和表面色泽进行考察。目前，对硬度的检测常采用M-T戳穿试验法方法，该方法属于有损检测，而对于水果表面色泽的测量常采用多指标表面色泽测量，十分繁琐。这些检测方法在大量样本逐个检测中无法满足实际生产。为此，选择一种方便可靠的水果硬度和表面色泽检测技术十分必要。 本文采用近红外光谱分析方法对梨的硬度和表面色泽进行定量分析，探讨建模变量选择方法对果品硬度和表面色泽指标预测模型的精度和稳定性的影响。为提高预测的精度，消除无信息变量对于模型稳健性的影响，在近红外光谱模型的建立中使用比对了蒙特卡罗无信息变量（MC—UVE）和基于小波变换的蒙特卡罗无信息变量（WT-MC-UVE）两种变量筛选方法。结果表明，WT-MC-UVE将建模的变量数从1451个降到210个，得到了和预测值相接近的结果。WT-MC-UVE提高了水果表面色泽的预测精度，预测均方根误差RMSEP在校准变量数从1451降到220个的情况下由1.06降至0.90，同样，相关系数也由0.975提高到0.981。为此，该种方法能从近红外光谱中筛选出更有效波长，在水果硬度和表面色泽的定量分析中使预测值更加稳健和准确。 关键词：近红外光谱；硬度；表面色泽；蒙特卡罗；无信息变量消除 The Discussion of Variable Selection Method in Detection of Firmness and Surface Color of Pear by Near Infrared Spectroscopy Abstract Now the consumption of fruit are on the ride with the continuous improvement of living standards. So when a large quantities of fruit are launched on the market. We need to inspect maturity of these fruit. The most common method which judge the maturity index is to measure the firmness and surface color of pear. At the present, M-T are always used in the test of firmness, which is destructive measurement. Multiple indicators which is applied to the measurement of fruit surface is very tedious. These method can’t meet the needs of actual production. So it’s necessary to choose a convenient and reliable method for the non-destructive and rapid measurement of firmness and surface color of pear. In this study, NIRS was applied to non-destructive and rapid measurement of firmness and surface color of pear. In order to improve the predictive precision, and eliminate the influence of uninformative variables for model robustness, monte carlo uninformative variables elimination (MC-UVE) and monte carlo uninformative variables elimination based on wavelet transform (WT-MC-UVE) methods were proposed for vari