The effect of flow rate, accelerometer location and temperature in acoustic chemometrics on liquid flow: Spectral changes and robustness of the prediction models

摘要: Abstract Prediction of chemical composition flowing liquids using passive acoustic measurements and multivariate regression (acoustic chemometrics) has been reported as a promising in-line measurement method. However, the results are also affected directly or indirectly by other factors than liquid, i.e. physical conditions flow equipment/pipe properties. The present study focuses on effects rate, accelerometer location temperature spectra prediction liquids. studied were two-component mixtures sucrose water, three-component ethanol, water. Multivariate models estimated both local global calibration full spectra, augmented frequency amplitude matrices derived from spectra. Flow rate had most pronounced effect recalibrated models. Temperature minor results. error for determination water increased with increasing rate. Changes in resulted considerable spectral variations, causing resultant model to perform poorly predicting new samples taken at conditions. Global performed well liquid all levels. models, however, needed higher number PLS led errors compared Using PLS/PPLS more parsimonious (4–6 factors) (10–12 factors).