【摘要】 为了提高模型预测精度,结合连续小波变换(CWT)的最优参数选择,优化小麦蛋白质光谱模型。对原始光谱进行CWT,利用主成分分析(PCA)选出5种小波db1、sym2、sym5、sym7、coif1;在不同尺度参数下利用偏最小二乘法(PLS)建模,确定尺度参数为15;在此基础上,利用CWT结合多元散射校正(MSC)及支持向量机(SVM)建模确定最优小波db1;在最优参数下用CWT结合无信息变量消除算法(UVE)和连续投影算法(SPA)及SVM建立预测模型,预测均方根误差为0.3930,优于CWT-UVEPLS-SVM的0.4558和CWT-SPA-SVM的0.4415,研究结果表明,CWT参数选择可有效优化近红外光谱模型。更多还原

【Abstract】 In order to improve the model prediction precision,the spectral model of wheat protein was optimized by optimal parameter selection of continuous wavelet transform(CWT).The original spectrum was processed by CWT,and five kinds of wavelet:db1,sym2,sym5 sym7,coif1 were selected by principal component analysis(PCA). The model in different scale parameters was built by partial least squares(PLS) method and the scale parameter was confirmed to be 15.Based on the results,the optimal wavelet db1 was determined by using CWT combined with multiple scatter correction(MSC)and support vector machine(SVM)model. The prediction model was built by using CWT combined with uninformative variables elimination(UVE),successive projections algorithm(SPA) and SVM under the optimal parameters. The Root-Mean-Square Error of Prediction was 0.3930, better than CWT-UVEPLS-SVM and CWT-SPA-SVM,0.4558 and 0.4415 respectively. Results show that the near-infrared spectroscopy models could be effectively optimized by CWT parameter selection.更多还原