基于神经网络的测色仪器系统误差分析和校正

为了提高工业生产过程中厂泛应用的光电积分测色仪的测试精度，分析、研究仪器误差成为仪器设计的重要内容之一。在
分析探测器的光谱三刺激非线性误差的基础上，进行了测色仪原理误差的研究，提出了基于人工神经网络的非线性滤波器的方法，并
建立了一个基于神经网络的非线性滤波器黑箱模型。实验表明:用此三刺激值计算的色品坐标、色差等参数均达到了设计要求。
关键词:色差计三刺激值神经网络

Abstract; In order to increase the measurement accuracy of widely used photoelectric integral colorimeten,anaiyzing and studying the error of the instrument has become one of the important tasks for designing the instrument. On the basis of analysis of spectral tristimulus non-linear error of the detector,the research on principle eror of colorimeters is achievcd and the method of non-linear filter bused on artificial neural network
is proposed, and a non-linear filter black box based on neural network is established. The experiment shows that the parameters calculated tristimulus,e.g. c hromaticityc umdinates,colord ifferenee,ete.，meetth e,m quhrmentso fd esign
Keywords Color-differencem eter Tristimulusv alues :Neuraln etwork