不同预处理方法对柴油十六烷值预测模型的影响

Effect of different pretreatment methods on prediction model of diesel cetane value

十六烷值是衡量柴油在压燃式发动机中发火性能的指标。采用近红外光谱方法对柴油的十六烷值进行快速检测,针对多种光谱预处理方法的组合进行了建模效果的对比筛选,研究结果表明,采用SNV标准正态变换和3点SG平滑能够整体提升柴油十六烷值模型的预测能力。根据外部验证结果,该模型结果满意,具备应用价值。

Cetane number is an index to measure the ignition performance of diesel oil in pressurized combustion engine.Using near infrared spectroscopy to detect the cetane number of diesel oil quickly.The results show that the SNV standard normal transformation and 3-point SG smoothing can improve the prediction ability of diesel cetane number model.According to the external verification results,the model results are satisfactory and has application value.