基于相思树NIR数据理论模型的软件设计  

Software Design Based on NIR Data and Theory Model of Acacia

作　　者：王聪[1] 张文杰[1] 刘胜[1] 

机构地区：[1]北京林业大学理学院,北京100083

出　　处：《现代农业科技》2012年第6期199-200,203,共3页Modern Agricultural Science and Technology

基　　金：国家科技支撑计划子课题"速生纸浆材特性数据库及浆纸性能预测模型的研究与开发"(2006BAD32B03-5)

摘　　要：根据朗伯—比尔定律所建立的近红外光谱数据相思树综纤维素含量的预测模型公式,采用Java语言开发了利用近红外光谱数据来预测木材化学成分含量的web版软件V1.0。首先对近红外光谱数据进行预处理和分组,然后使用拟合方法建立了许多非线性子模型,列出每个吸光值及其所对应的子模型当中的系数参数,采用Java5.0语言将以上数据和模型公式编译成软件。该软件使用方便,操作简单,预测结果准确度高(如综纤维素测量值与预测值之间的相关系数为0.935 2),为木材纸浆性能分析提供了可靠的依据和可操作的方法,并有望用于其他树种的化学成分预测。According to the Lambert-beer＇s law,the mathematical model of Acacia was built by NIR data,and we developed software by using Java language. This software which called V1.0 could forecast component content of wood material by using NIR data. The NIR data were pretreated and divided into groups ,and many nonlinear sub mathematical models were built using the fitting method. Each OD value had its corresponding parameter in sub models. According to these parameters and the formula,we used Java 5.0 to compile a software named V1.0. The V1.0 is easy to use,and the predicted results are high accuracy. （E.g. the coefficient of correlation between the experiment value and the predicted value is 0.935 2）.The software provides a reliable basis and workable methods for the wood pulp performance analysis,and is expected to use the chemical composition forecasts of other trees.

关 键 词：相思树 近红外光谱 数学模型 JAVA语言 

分 类 号：S792.99[农业科学—林木遗传育种] TP319[农业科学—林学]