拉曼光谱仪结合GABP神经网络测定石脑油馏程

Raman Spectrometer Predicting the Distillation Range of Naphtha with GABP Neural Network

介绍了拉曼光谱仪的特点及结构组成。在现有仪器研制的基础上,收集了一定数量的石脑油样品采用拉曼光谱技术和经RNCGA(实数编码遗传算法)优化的BP神经网络建立了快速、准确测定石脑油馏程的分析仿真模型,并对预测集样品的分馏点馏程逐个进行测算,之后又对样品预测值与样品真实值的相关性进行了分析。实验结果表明,该模型与传统馏程检测方法相比具有分析速度快,准确性高等特点,适用于对石脑油样品馏程进行预测。

Introduced the structural characteristics and structural composing of Raman spectrometer, on the basis of existing spectrometer, collected a certain number of experimental samples of naphtha, make use of Raman spectroscpy and a BP neural network model based on RNCGA （real number coding genetic algorithms）to build up a quick, precise analysis simulation model approach to determine the distillation range of naphtha,and measured cut point＇s distillation range of prediction samples and analysised the correlation of the predictive value of the model and the true value of sample.The simulation results of experimental data shows,that method is fast, and high accuracy with the comparison of the traditional distillation range detection method, suitable for predicting the distillation range of naphtha.