The Design And Implementation Of The Rapid Detection System For The Number Of Frying Times Of Edible Frying Based On NIRS

Edible oil is an indispensable cooking material in daily life and is widely used in the food industry at home and abroad.However,after repeated high-temperature frying of edible oil,harmful chemicals such as esters,ketones,and aldehydes will be produced,which will not only affect the quality of fried foods,but also endanger human life and health.The content of harmful substances in edible oil will gradually increase as the frying times continue to increase,and because of this,the quality of edible oil will gradually deteriorate.At present,in many researches at home and abroad,the quality of edible oil is tested by chemical analysis methods,and physical and chemical indicators such as acid value,peroxide value,and carbonyl value are used to evaluate the edible oil quality.However,traditional chemical analysis methods require strict experimental conditions and professional technical personnel,so the detection efficiency is low and it is difficult to achieve rapid real-time detection.The near infrared reflectance spectroscopy(NIRS)detection technology has the unique advantages of non-destructive,efficient,environmental protection and green.Therefore,this paper proposed to use the NIRS technology and chemometric algorithms to detect the quality of edible oil under standard experimental procedures.That was,the quality evaluation method of edible oil was transformed into a quantitative detection for the frying times of edible oil under a standard frying experimental operation procedure.In this research,the DS2500 near-infrared spectral analyzer produced by the Danish Foss company was used to collect samples' spectra.And 150 raw spectra for each of soybean oil,peanut oil,and rapeseed oil were obtained.A manual method was used to divide the spectral data of each edible oil into a training set and a test set.A variety of preprocessing methods were used to optimize the spectral data and the best preprocessing method was determined according to the evaluation index.The correlation coefficient method was used to extract the characteristic wavelength,and the characteristic wavelength with the greatest common correlation among the three edible oils was selected at 1170 nm.1168 nm and 1172 nm wavelength pairs with a 2nm interval around 1170 nm were selected.Differential,normalize and average the original spectral absorbance of the two wavelengths to obtain an averaged difference normalized curve.The curve was used to determine the spectral absorbance threshold corresponding to each frying time of each edible oil,and then the difference threshold model of each edible oil was obtained.It was also necessary to select a fixed standard value and standardized the sample tested,and the standardized value was input into the differential threshold model of each edible oil to obtain the predicted frying times of the sample tested.The results show that the accuracy of the method for detecting the frying times for three edible oils was all 100% within the tolerance range(± 1).Therefore,the differential threshold model established in this study could effectively and accurately detect the frying times of three edible oils.Then designed a windows-based human-computer interaction interface that matched the differential threshold model of each edible oil to achieve visualization and systematization for detection the frying times of three edible oils.Simultaneously,this detection system also provided theoretical basis and technical reference for edible oil quality detection in the food industry.