Real-time monitoring of alcalase hydrolysis of egg white protein using near infrared spectroscopy technique combined with efficient modeling algorithm

Figures & data

Figure 1. A: The raw NIR spectra; and B: the SNV preprocessed spectra of EWP enzymolysis samples.

Figure 2. Variation trend of DH values during the enzymatic hydrolysis of egg white protein.

Table 1. Reference measured values of DH of samples in the calibration and prediction sets.

Subsets	Units (%)	S.N.^a	Range	Mean	S.D.^b
Calibration set	mol/mol	51	0 ~ 0.2064	0.1453	0.0016
Prediction set	mol/mol	34	0.0552 ~ 0.2023	0.1460	0.0014

^aS.N.: sample number.

^bS.N.: standard deviation.

Figure 3. Spectral variables selected by A: UVE, B: CARS, and C: UVE-ACO.

Figure 4. Scatter plots between reference measurements and NIR predicted results in calibration and prediction sets for A: PLS, B: UVE-PLS, C: CARS-PLS, and D: UVE-ACO-PLS.

Table 2. Results of different regression models for predicting the DH values in EWP enzymolysis.

Methods	Variables	PLS factors	Calibration set		Prediction set
			RC	RMSEC	RP	RMSEP
PLS	1557	13	0.8610	0.0201	0.6652	0.0293
UVE-PLS	108	15	0.9927	0.0048	0.8582	0.0196
CARS-PLS	60	14	0.9995	0.0012	0.9078	0.0155
UVE-ACO-PLS	19	9	0.9828	0.0072	0.9528	0.0116

Figure 5. RE (%) values of three times NIR predicted values for DH.