Research on rapid determination of calcium content in stone powder using near infrared spectroscopy
Stone powder is a kind of inorganic substance that serves as an importance source of calcium in livestock and poul-try rations.In this paper,a quantitative model for determining calcium content in stone powder is proposed based on near infra-red spectroscopy(NIRS)technology in the hope to find out a way of determining calcium content in stone power with high pre-cision and reliability,which may offer reference for NIRS-based analysis of other inorganic feed samples.In the study,alto-gether 248 stone powder modeling samples were collected to undergo NIRS and wet chemical tests.A quantitative model of near infrared was established using modeling software to determine the calcium content in stone powder.Furthermore,21 blind samples were randomly collected to verify the model.The results suggest:(1)the determination coefficient of the proposed model R2 is 85.28,the root mean square error of cross-validation(RMSECV)is 0.308,and the residual prediction deviation(RPD)is 2.61,which collectively prove the competent predicting performance of the proposed model;(2)as shown by the blind sample verification results,the model could effectively identify abnormal alarm samples,because for the 20 normal sam-ples,the relative errors between wet chemical values and predicted values are within 0.01％-0.56％,the root mean square er-ror of prediction(RMSEP)is 0.229,the average bias is 0.077,and the residual prediction deviation(RPD)is 3.98.All the da-ta have testified the competent verification effect of the proposed model.
near infraredstone powdercalciumquantitative model
万方数据
维普
