Software Design Of Mathematical Model Of Acacia By Near Infrared Spectroscopy Data

The chemical composition of the paper wood is in connection with its pulp properties. The analyze of the chemical composition of the paper wood with traditional chemical analysis is complex operation, time-consuming, big error and is destructive analysis. It will call for a lot of money and manpower. Near Infrared Reflectance Spectroscopy is a dynamic non-destruction measurement. It's fast, efficient with low costs and good reproducibility, and it can also be convenient to use. This paper offers a new modeling method through analyzes the traditional chemometrics and its relative merits, focuses on the application of NIR in wood science. It also developed special software with this new mathematical model. According to the Lambert-beer's law, the mathematical model of acacia is builded by near infrared spectroscopy data, and we developed software by using java language. This software which we called V-1.0 can 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. Every OD value has its corresponding parameter in sub models. According to these parameters and the formula, we compiled a software using java5.0. 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 0.9352.) 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.