基于介电特性与蛋黄指数回归模型的鸡蛋新鲜度无损检测

为了更准确、快速的实现鸡蛋新鲜度品质指标蛋黄指数的无损检测,基于介电特性建立了鸡蛋新鲜度无损检测模型,获取鸡蛋的蛋黄指数信息。试验以不同新鲜度鸡蛋为研究对象,采用平行极板法测量不同新鲜度鸡蛋在温度为20℃,相对湿度为72%~89%,频率为1~200 k Hz下的介电特性参数,分析鸡蛋介电特性的变化规律,并建立鸡蛋介电特性与蛋黄指数之间的数学模型。分析了鸡蛋介电特性随测量信号频率及新鲜度指标蛋黄指数的变化曲线,发现其介电参数随频率及蛋黄指数的增大而减小。构建了蛋黄指数与相对介电常数的拟合方程,鸡蛋样品的实际蛋黄指数与模型预测的蛋黄指数间的决定系数R2为0.9115,蛋黄指数的误差为±4.2%,试验结果表明:利用拟合方程预测检验鸡蛋蛋黄指数取得了较好的预测效果,为无损检测鸡蛋新鲜度提供了一种新的可行方法。

Non-destructive examination for freshness of eggs based on dielectric properties and yolk index regression model

Abstract: Egg is an important source of protein nutrition in China. Eggs which have high protein, low fat and a variety of nutritional elements, easy to digest and absorb, and so on are popular with people. Detection of freshness of egg effectively is not only to protect the interests of consumers, but also beneficial to the management of producers and operators. At present, fluorescence spectrum, infrared spectrum, image processing and dielectric property detection are the main methods of egg nondestructive testing. Dielectric property detection technology is a fast and nondestructive testing technology. The characteristics of dielectric properties have been shown to have special advantages in the field of agricultural product quality evaluation and grading. There are many achievements in non-destructive testing of agricultural products by using dielectric properties. In order to achieve a more accurate and rapid determination of egg yolk index, we studied the relationship between the dielectric properties and egg yolk index. Based on the dielectric properties, we established an egg fresh degree nondestructive detection model, and obtained the egg yolk index information. In this experiment, we used parallel plate method to measure the dielectric properties of different fresh eggs between 1 and 200 kHz. This experiment was based on different fresh degree eggs. We measured the dielectric properties of different fresh eggs between 1 and 200 kHz by using the parallel plate method. We used MATLAB to analyze the effects of frequency and freshness on the dielectric properties of eggs, and to draw the curve. The research results showed that the measurement frequency and the fresh degree had a significant effect on the dielectric properties. Between 1 and 200 kHz, the relative dielectric constant and dielectric loss factor of eggs decreased with the increase of the test frequency. When the frequency was less than 50 kHz, the decline was obvious. When the frequency was more than 50 kHz, the decline was not obvious. Between 1 and 200 kHz, the relative dielectric constant of the egg decreased with the increase of the egg fresh degree. When the egg yolk index was less than 0.2, the relative dielectric constant decreased significantly. When the egg yolk index was greater than 0.2, the relative dielectric constant change was not obvious. We used MATLAB to fit the equation of relative dielectric constant and egg yolk index. Comparison of the determination coefficient, we found that the best fitting effect of measurement data under frequency of 1.2 kHz. The determination coefficient was 0.995. In order to test the correctness of the model, we randomly selected 50 eggs to measure. Validation of the results was that the determination coefficient was 0.9115 and the yolk index error within was ±4.2%. The results showed that the egg yolk index obtained by the fitting equation can predict the egg yolk index, which provided a new feasible method for the measurement of fresh degree of eggs.