近红外透射式谷物蛋白质含量在线监测系统

On-Line Near-Infrared Transmittance Type Inspecting System for Protein Content of Grain

采用近红外透射光谱分析技术 ,建立了用于谷物加工过程中的蛋白质含量在线监测系统。宽光束照射、透射光谱的二阶微分处理和基于人工神经元网络的标定方法是保证监测系统有效工作的关键。同时建立了用于小麦蛋白质含量现场在线监测的最佳 BP( Back Propagation)网络结构 ,其 SEC=0 .1 2 % ,SEP=0 .1 5%。该监测系统的研究不仅可以直接应用于谷物成份含量的在线监测 ,而且所涉及的方法也同样适合于其它光谱分析应用。

An on-line inspecting system for protein content of grain during grain processing process is set up by means of near-infrared transmittance spectral analysis technique. A wide beam illuminating, second-order differential processing for transmittance spectrum and a calibration method based on artificial neuron network are the key factors for ensuring the effective operating of the inspecting system. At the same time, an optimal Back Propagation (BP) network architecture for the field on-line inspecting the protein content of wheat is set up. Its SEC is 0.12% and SEP is 0.15%, respectively. The development of the system not only can directly on-line inspect the protein content of grains, but the related method also can be suitable for the other spectral analysis applications.