论文部分内容阅读
利用近红外光谱分析理论原理,设计研究了煤粉发热量近红外检测系统。实验中选取100个煤粉样品建立煤粉发热量的近红外模型,然后利用验证集的50个样本对模型的精度和稳定性进行了验证分析。结果表明,采用主成分分析法建立的煤粉发热量预测模型最优,煤粉发热量的预测值与真实值的相关系数达到0.996 2,相对偏差<2%。该系统具有良好的预测精度和稳定性,能够满足对煤粉发热量的快速检测的需求,而且该系统具有体积小、结构简单、操作方便的特点,还具有很好的可移植性,维护方便。
Using near infrared spectroscopy theory, the design and study of the pulverized coal heat value near infrared detection system. In this experiment, 100 samples of coal were selected to establish the near-infrared model of the calorific value of pulverized coal, and then the accuracy and stability of the model were verified by 50 samples of the verification set. The results show that the predictive model of coal calorific value established by principal component analysis is the best, the correlation coefficient between predicted and real values of calorific value of coal reaches 0.996 2, and the relative deviation is less than 2%. The system has a good prediction accuracy and stability, to meet the rapid detection of coal calorie needs, and the system has the characteristics of small size, simple structure, easy operation, but also has good portability, easy maintenance .