基于聚乙烯本体温度与其凝聚态结构具有--对应关系,通过对拉曼光谱进行偏最小二乘法(PLS)分析,建立了聚乙烯本体温度的PLS同归预测模型.聚乙烯本体温度的模型预测值与真实值的相关系数、平均相对误差和预测均方根误差分别为0.999,1.01%,1.32.完全能满足实际工业生产对聚乙烯本体温度控制的要求.对1 350～1 530 cm-1 内的拉曼光谱进行PLS分析发现,第一项PLS成分的载荷分布能反映聚乙烯本体温度与聚乙烯凝聚态结构间的关系,即与无定形相含量成正相关,而与结晶相含量成负相关.研究结果表明,利用拉曼光谱可实现聚乙烯本体温度的快速检测.%Based on relationship of polyethylene aggregation structure with its bulk temperature, a regression model was established to predict polyethylene bulk temperature by analyzing the Raman spectrum between 1 350 -1 530 cm-1 with the partial least square (PLS) method. The correlation coefficient, average relative error and root mean square error of the prediction are 0. 999,1.01％ and 1.32,respectively, which can fully satisfy the requirements to control bulk temperature during industrial production. According to the PLS analysis, the loading distribution of factor 1 reflects the relationship between the bulk temperature and polyethylene aggregation structure, that is, the bulk temperature positively correlates to the content of the amorphous phase,but negatively correlates to the content of the crystalline phase. The results indicate that determination of polyethylene bulk temperature by Raman spectrum is feasible.
展开▼
