025-52239040
EN

#### Industry dynamics

##### Error Analysis of Shelf Life Prediction of Energy Bar with Color as Deterioration Index2019-05-23

School of Food Science and Nutrition Engineering, China Agricultural University, Beijing 100083) studied various factors in accelerated experimental design and specific experimental methods for predicting the shelf life of energy bars, including the determination of repetition times, the number and time interval of fitting points.

And accelerate the temperature condition.

The results show that the repetition number of samples has little effect on the prediction accuracy. The detection points and intervals at each temperature have a great influence on the prediction accuracy.

In the design of acceleration experiment, the number and range of acceleration temperature have the greatest influence on the prediction error.

Through reasonable experimental design, the prediction accuracy can be controlled within 10%.

ASLT is a common method to predict the shelf life of food.

Generally speaking, the ASLT method is more common in industrial applications because it can shorten the time to market. The prediction method of the shelf life usually includes five steps: determining the main quality deterioration and the corresponding indicators in the storage process;

Predictive experiments; determination of sensory acceptable endpoints and corresponding chemical index limits; establishment of mathematical models (Arrhenius 2-3 or improved Arrhenius 4 equation) to describe the relationship between chemical index changes and temperature; calculation of shelf life and reality at room temperature

Verify the shelf life comparison.

In this field, Calligaris 5 establishes an ASLT dynamic model to predict the shelf life of baked food, and validates the model with bread as raw material. 6. Rahmouni0 has established a mathematical model to describe the oxidation stability of olive oil.

However, there is a fixed error in ASLT's shelf life, which is accumulated by each step error. Prediction process.

Determining the number of repetitions will affect the accuracy of measurement; the number of fitting points and the time interval of downward curve will affect the accuracy of rate constant; and the temperature point and temperature range of accelerated experiment will also affect the accuracy of shelf life prediction.

At present, some researchers have reported error analysis using accelerated experiments to predict shelf life. Zhang Ronghui 8 predicted the shelf-life of egg rolls by accelerated experiments, 20T; in constant temperature environment, in the absence of desiccants and desiccants, shelf-life prediction

The errors were 9.5% and 8.0% for 63 D and 75 d, respectively. Respectively. Magari 9 uses mathematical models to estimate factors affecting the accuracy of shelf sales, such as measurement repetition, batch size and experimental design.

However, in the actual prediction experiment, the influence of various factors on the prediction error is still uncertain.

This study will discuss the prediction accuracy of color difference model by parallel number, number of points, influence time interval of detection and acceleration temperature condition, in order to maximize the prediction accuracy and provide a basis for similar research.

Materials and Methods 1.1 Military Energy Bars for Materials and Instruments 1.1 Military Requirements of Beijing General Logistics Department

Prepared by the Research Institute of Tianjin Jinyuan Cake Industry Co., Ltd.

DC-P3 Automatic Colorimeter Beijing Starlight Chromometer Company; LHS-250HC-1 Constant Temperature and Humidity Box Shanghai Yiheng Science Instrument Co., Ltd.

1.2 Sensory evaluation was used to determine the shelf life at the end of the shelf. Weibull hazard analysis method determines sensory acceptable endpoints. Forty bags of energy bars were placed in a constant temperature and humidity chamber (RH = 60% c) at 37 C. Four samples were randomly sampled every 10 days and h for storage at 4T. After all the samples were taken out, the refrigerator,

Sensory evaluation was conducted together.

The sensory evaluation method, Dong Xinna experiment method, used the least square method to regress the data to the cumulative risk of 100% C. The shelf life prediction model of energy bars with color as deterioration index was established by ASLT method.

The XC of the Lnt-(LnK) index of the shelf life prediction model represents the life of food storage (d), and C represents the temperature of Celsius as a parameter.

Prediction error analysis of Arrhenius equation Sample parallelism/experimental detection point and detection time interval can predict the accuracy of Arrhenius equation. From Table 3, it can be seen that Arrhenius equation has a high fitting coefficient, and the room temperature rate is usually 3 parallel to the sample.

The repetition times (d) of fitting test time interval at different temperatures were determined. The velocity constant K Lenus 2 was constant K and the shelf life at room temperature was determined.

Prediction Error Prediction Value Prediction Value Prediction Value Prediction Value Table 4 Effect of Acceleration Temperature on Prediction Accuracy of Arrhenius Equation Conditions Temperature Number Temperature Gradient Temperature Range

Prediction error temperature condition (especially) Prediction value Prediction value Prediction value Prediction value Calculated value Ancient: Question / dish whole ancient: Question / dish whole ancient: Question / dish whole ancient: Question / dish whole ancient: Question / dish whole ancient: Question / dish whole ancient: Question / dish whole ancient: Question / dish whole ancient

The predicted value of the total is too small, the predicted value has too long shelf life, and the predicted error is too large.

When the number of experimental points and the time interval remain unchanged, the Arrhenius equation does not change much with the increase of parallel forces. When the prediction error is less than 0.99, the prediction error decreases slightly, indicating that the number of samples has little effect on the prediction accuracy of the model.

When the parallel number and time interval remain unchanged, the reaction rate constant k is 45,50T with the increase of test points, and the correlation coefficient of Arrhenius equation increases gradually, and the prediction error of shelf life decreases.

From 64.1% C to 24.2% c, 40% C is reduced, which indicates that the number of points has a greater impact on the prediction accuracy of the model. When the number of parallel experiments and the number of experimental points are fixed, the detection interval changes with the change of the number of parallel experiments and the number of experimental points.

The correlation coefficient of Arrhenius equation increases gradually, the predicted value of room temperature shelf life decreases gradually, and the predicted error decreases significantly.

From 55.9% C to 16.4% C, the decrease is about 40% c, which indicates that the time interval can predict the accuracy of the model.

The temperature conditions affecting the prediction accuracy of the Arrhenius equation are shown in Table 4. The Arrhenius equation has higher fitting coefficients than 0.97, with a maximum value of 0.9999 (except two temperatures). The acceleration temperature conditions of the Arrhenius equation are different.

The prediction accuracy has a significant impact, and the prediction error varies greatly. The maximum value was 257%, and the minimum value was 11.1%. Temperature number, temperature gradient and temperature range have significant influence on the prediction accuracy of Arrhenius equation, and have the number of accelerated temperatures.

With the increase of temperature gradient and temperature range, the prediction error decreases gradually, ranging from 214% to 12.4%. Using 37,50,60T to establish the prediction model at the overall temperature, accelerating the temperature and predicting the shelf life at room temperature.

For 479d, the prediction error is at least 12.4%, which is similar to the report's prediction error of shelf life (+10%).

3. Conclusion The prediction accuracy of Lnt (LnK) XC and Arrhhenius equation of accelerated experimental condition shelf life prediction model is Zhang Liping, Yu Xiaoqin and Tong Huarong. Application of Weibull model in shelf life prediction of salted duck.

Food Science and Technology, 2010, 35 (2): 111-113. Cao Ping Yu Yanbo and Li Peirong

Weibull Hazard Analysis is used to predict the shelf life of food.

Food Science, 2007, 8 (8): 487-491. Zhang Ronghui, Xiao Kaijun

The shelf-life of egg rolls was predicted by kinetic theory.

Food Research and Development, 2001, 2 (5): 51-53. Dong Xin Na

Research on the prediction technology of the shelf life of military energy bars.

Beijing: Zhong Jiaojiao, Xiangchuan Wan.

Flavor peanut powder on the oxidation rack according to the temperature change. Thank you Julan, Chen Long, Lei Xiaoling and so on.

Modem food science and technology for predicting shelf life of low-salt shrimp paste using volatile salt-based nitrogen dynamics model, 2013, 29 (1): 29-33. (Page 305 above) (12.36% to 12.89%)

These two fatty acids are unsaturated fatty acids, which play an important role in softening blood vessels, reducing blood lipids and preventing cardiovascular diseases such as atherosclerosis. ＃

## Nanjing Future Mechanical Equipment Co., LtdWelcome to leave messages online, you may have problems, or even suggestions on how to improve.  