KR20180055356A - System and Method for Detecting Milk Adulteration Using Stalagmometer - Google Patents

Abstract

The present invention provides a system and a method for detecting adulteration contained in milk with an automated method based on a phenomenon in which surface tension of the milk is changed in accordance with contents of adulteration such as urea, detergent, or the like in the milk. Moreover, the present invention measures surface tension by an amount of milk collected for a predetermined time with a stalagmometer, and calculates a type and contents of adulteration in a milk sample from a value measured by a controller.

Description

Technical Field [0001] The present invention relates to a system and a method for detecting milk impurities using an aqueous solution,

The present invention relates to a system and a method for detecting milk impurities using an aqueous system. More specifically, the present invention is based on the phenomenon that the surface tension of milk changes according to the content of impurities (for example, urea, detergent, etc.) in the milk, but the amount of milk collected for a certain period of time To a system and method for detecting impurities in milk in an automated manner by measuring the surface tension and calculating the type and content of impurities in the milk sample from the measured values by the controller.

Consumer interest in antimicrobials and other antimicrobials in agricultural and livestock products has increased and the demand for safe livestock products has been strongly increased. Therefore, it is required to strengthen standards management and sanitation management system to produce safe livestock products.

In recent years, demand for livestock foods, especially milk, has been increasing as dietary habits change, and production volume is steadily increasing accordingly. In terms of livestock management, the type of feed, the quality, and the method of paying are changing drastically as the grouping, and furthermore, the change of the breeding form to the facility livestock are made.

Milk is an essential source of nutrients that are essential to the growth of infants and children and is essential for the maintenance of adult health.

In particular, milk is a complete food that is easy to ingest and absorb into the body. Especially, it is the only natural food for infants and children. It is also a high-value source of high quality protein, fat, carbohydrate, vitamins and minerals, is inexpensive, and is also needed to maintain adult cellular tissue. As such, milk is a nutrient-rich natural food, which is an important nutrient source for maintaining the health of the human body. Since milk is widely consumed from infants to adults, quality control is an important factor.

However, in the livestock sector, a variety of adulterants are often added for the purpose of cost reduction or prolonging the shelf life. Such impurities in the milk not only reduce the milk quality (nutritional value, flavor, etc.) Can lead to danger. In this connection, "impurities" are generally understood to mean any ingredient which reduces the safety or value of the food.

In particular, chemicals such as detergents, urea, baking soda, and the like are ingredients added to milk to increase the shelf life and, in the case of urea, to increase the SNF value. However, the hazards of these additive ingredients are still controversial.

Typically, milk is judged to contain impurities or deteriorate in quality if either of the following is true: (i) the addition of ingredients that reduce the value of milk or cause adverse health effects; (ii) ) Low or low valued components are wholly or partly incorporated; and (iii) high valued or essential components are wholly or partly removed.

In this regard, an example of a method for detecting impurities in milk is as follows: in the case of urea, p-dimethylamino benzaldehyde reagent is added at 16% (w / v) to 10% hydrochloric acid (w / v) Dissolved) is added to the mixture, and the mixture is well mixed and detected by color. When the solution changes to yellow, it is determined that the milk sample contains urea. The change in hue with respect to the concentration of urea is exemplified in Table 1 below.

In the case of detergent, 1 to 2 drops of Bromocresol purple solution (1% (w / v) in water) may be added to the milk, followed by mixing well. At this time, when purple appears, it indicates that detergent is contained in the milk.

As described above, in order to add impurities in the milk in the laboratory, it is necessary to use a reagent according to the kind of impurities. Therefore, a detection test must be performed manually for each individual impurity, thereby increasing detection work and time. Furthermore, the presence or absence of impurities can be confirmed by such a detection test method, but quantitative analysis is difficult.

In addition, U.S. Patent Publication No. 2014/0065712 discloses a method for rapidly detecting impurities (e.g., urea) in milk using a microfluidic system in order to alleviate the above-mentioned problem of the manual method, The method is also limited to specific impurity detection applications.

Thus, there is a need for a method of detecting impurities in milk in a simpler, more accurate and automated manner.

Therefore, in the embodiment of the present invention, it is possible to solve the problems caused by the conventional detection method of impurities in the milk, and to provide an improved method for detecting the impurities in the milk Detection system and method.

According to one embodiment of the present invention,

A method for detecting the type and content of impurities in a milk sample in an automated manner,

a) dropping the milk sample through the enemy water system into a container located beneath it;

b) measuring the weight of the milk sample collected into the vessel for a predetermined time using a pressure sensor; And

c) transmitting a signal of the pressure sensor according to the weight value of the measured milk sample to the controller; And

d) using the controller to determine a component of the impurities in the milk sample and its content;

/ RTI >

The controller provides a method of deriving the type of impurities in the milk sample and its content by comparing the data on the surface tension obtained from the delivered signal with the reference data fitted by the controller.

According to an exemplary embodiment, the controller may be an arduino microcontroller.

According to an exemplary embodiment, the controller may be equipped with an ATmega328P chip.

According to an illustrative embodiment, step d) is performed using an artificial neural network based on the Naive Bayes Classifier to determine the presence of an already known impurity-containing milk sample and an already known impurity for at least 1000 samples And a step of learning the relationship between the two.

According to another embodiment of the present invention,

A system for detecting the type and content of impurities in a milk sample in an automated manner,

An aqueous system configured to drop the introduced milk sample downward;

A container disposed in the underside of said aqueous system for collecting said distilled milk sample;

A pressure sensor attached or connected to the vessel to measure the weight of the milk sample and transmit a signal about the weight of the measured milk sample;

A controller for deriving the type of impurities in the milk sample and its content by comparing data on the surface tension of the milk sample obtained based on the signal transmitted from the pressure sensor with fitted reference data; And

A display unit for displaying the kind and content of the derived impurities;

/ RTI >

The system and method provided in accordance with embodiments of the present invention are based on the phenomenon that the surface tension of milk changes according to the content of impurities (for example, urea, detergent, etc.) in the milk, The surface tension is measured by the amount of collected milk and the type of the impurities in the milk sample and its content are compared with the reference data fitted through the learning process by the artificial neural network with the measured value by the controller It is possible to solve the complexity of the detection method and the difficulty of quantitative evaluation caused by the prior art at once by calculating it in an automated manner in real time. Therefore, it is expected to be widely applied in the future.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a diagram showing a typical configuration of an in-water system usable in an exemplary embodiment of the present invention;
2 is a schematic diagram illustrating a system for detecting the type and content of impurities in a milk sample in an automated manner in accordance with one embodiment of the present invention;
3 is a view schematically showing a series of processes of learning through an artificial neural network based on preliminarily measured experimental data and fitting learned data based on a microcontroller in one embodiment of the present invention;
FIG. 4 is a graph showing the relationship between the concentration of impurities and the surface tension, obtained by preliminary measurement according to an embodiment of the present invention, And a series of processes for detecting the type and the content of the sample; And
5 is a graph showing the surface tension of milk measured according to the concentration for each of the three impurities (urea, detergent and baking soda) in the examples.

The present invention can be all accomplished by the following description. The following description should be understood to describe preferred embodiments of the present invention, but the present invention is not necessarily limited thereto. It is to be understood that the accompanying drawings are included to provide a further understanding of the invention and are not to be construed as limiting the present invention. The details of the individual components may be properly understood by reference to the following detailed description of the related description.

In this specification, when an element is referred to as "comprising ", it means that it may further include other elements unless otherwise stated.

Also, in this specification, the terms "located on the upper side" or "located on the lower side" can be understood to express a relative positional relationship in a state of being in contact with not only a state in contact with a specific object.

Where reference is made herein to any component or element that is "connected" or "communicating" with another component or element, unless otherwise stated, the component or element is directly connected or in communication with the other component The present invention can be understood to include not only the case but also the case where it is connected or communicated under the interposition of another component or member.

As described above, the present invention measures the surface tension of a milk sample measured using an in-water system, and calculates the surface tension of the milk sample by using fitting data derived based on data learned by the artificial neural network, The present invention provides a system and method for obtaining the kinds and contents of impurities in a milk sample in an automated manner in real time.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagram showing a general configuration of an aqueous system usable in an exemplary embodiment of the present invention. FIG.

The aqueous system includes a glass tube in which a bulb-shaped chamber 1 is integrally formed, and receives a sample of milk through an inlet 2 and discharges a droplet of a milk sample through an outlet 3 in a downward direction As shown in FIG. At this time, the formation of the droplet of the liquid sample depends on the viscosity of the liquid sample as well as the radius of the outlet 3 of the aqueous system.

The relationship between the weight of the droplet discharged from the red water system and the surface tension of the sample can be expressed by the following equation (1).

[Equation 1]

mg = 2? r

Where g is the weight of the sample, g is the gravitational acceleration, rho is the surface tension of the sample, and r is the radius of the emitter outlet.

At this time, since the weight of the droplet of the liquid sample depends on the surface tension, the number of droplets discharged from the aqueous system per unit time changes according to the surface tension of the sample. The result is a change in the total weight of the liquid sample collected over a specific time interval. Thus, the surface tension of the sample can be measured by measuring the weight of the sample droplet falling through the enemy water system.

In this connection, the surface tension of the milk changes depending on the content of the impurities contained in the milk. Therefore, theoretically, the surface tension can be obtained by collecting the milk sample falling from the red water system for a predetermined time interval and measuring the total weight thereof, and the content of the impurity in the milk sample can be obtained based on the calculated surface tension. Furthermore, the weight or surface tension of the collected sample differs depending on the type of impurities contained therein.

2 is a schematic diagram illustrating a system for detecting the type and content of impurities in a milk sample in an automated manner in accordance with one embodiment of the present invention.

In the illustrated embodiment, when the milk sample is injected into the inlet 2 of the aforementioned aqueous system, the milk sample transferred into and held in the chamber 1 drops downward through the discharge port 3 while forming a droplet. The droplets thus separated can be transferred to the container 7 located at the lower side without loss by the funnel 4.

At this time, as shown in FIG. 1, an overflow siphon 5 may be provided as a kind of passive valve in order to prevent the falling droplet from overflowing the funnel 4. A valve, specifically, a solenoid valve 6 is disposed between the funnel 4 and the container 7 to control the flow of droplets that are opened and closed by a signal of the controller 9 to be described later and introduced into the container 7 have. As such, the overflow siphon 5 and solenoid valve 6 serve to regulate the flow of droplets collected into the vessel over a predetermined time interval.

On the other hand, the container is provided with a sensor, specifically a pressure sensor 8, to generate a signal corresponding to the weight of the milk sample collected in the container 7. At this time, the pressure sensor 8 measures the pressure generated by the total droplet collected in the container 7, and the pressure sensor 8 measures the pressure of the droplet collected in the container 7 according to the kind and / The numbers become different.

Thus, the signal generated by the pressure sensor 8 corresponding to the weight of the milk sample collected from the red water system for a predetermined time interval is transmitted to the controller 9. At this time, the controller 9 generates data about the kind of the impurities in the milk sample and its content by comparing the data about the surface tension obtained from the transmitted signal with the reference data fitted in the controller 9 do. The data thus generated is transmitted from the controller 9 to the display unit 10 and is recognized by the measurement person.

3 schematically shows a series of processes of learning through an artificial neural network based on preliminarily measured experimental data and fitting learned data based on a microcontroller in one embodiment of the present invention.

As shown in the figure, in order to generate the reference data fitted in the controller 9, experiments are conducted in advance to obtain experimental data on changes in surface tension according to the kind of the impurities and their contents, It can go through a learning process and accumulation process.

Specifically, a preliminary experiment can be performed on, for example, at least 1000 samples, specifically at least 1500 samples, more specifically at least 2000 samples. To obtain more accurate results, preliminary experimental results are learned in artificial neural networks. In particular, Naive Bayes Classifier, which is widely used as a classification algorithm among machine learning algorithms, can be used during the learning process. The algorithm that is the basis of the Naive base classifier has recently attracted attention as one of statistical based machine learning algorithms.

In one embodiment, an artificial neural network based on the Naive Bayes classifier is used to learn the relationship between already known impurity-containing milk samples and known impurities for at least 1000 samples. As a result, a learned classifier can be formed for various concentrations of milk impurities. The network thus learned is used to detect the type and content of impurities in the unknown milk sample. Throughout this learning process, we will produce a learned classifier of various milk impurities and their concentrations.

In an exemplary embodiment, an Arduino microcontroller is provided as a controller 9 for generating reference data by performing a fitting process based on a learning operation on the preliminary experimental results described above, and more specifically, An ADNO microcontroller equipped with an ATmega328P chip can be used.

In this regard, the ADINO microcontroller with the ATmega328P chip is typically able to receive information from outside through the input pins of the sensor and can be programmed using the Aldino programming language.

In order to be able to carry out the above-described series of operations, according to an exemplary embodiment, the microcontroller comprises, for example, 14 digital input / output pins, 6 analog inputs, 16 MHz ceramic resonator, USB connector, A header, and a reset button.

FIG. 4 is a graph showing the correlation between the individual impurity concentration and the surface tension by preliminary measurement according to one embodiment of the present invention, and data on the weight (or surface tension) of the milk sample measured after learning it by the artificial neural network And detecting the kind and content of the impurities based on the detection results.

According to the figure, an experiment is first performed to measure the surface tension of the milk sample versus the impurity concentration for various impurities (chemical components) (step 1).

Then, the data obtained from the experiment are learned through the artificial neural network of the controller (step 2).

In the case of the value obtained through the learning network, a step of verifying (verifying) the validity of the value using a sample having an already known impurity concentration is performed (step 3).

Once validated by step 3, the controller (specifically, the ATmega328P microcontroller) is uploaded to the learned artificial neural network (step 4). At this time, USB cable can be used.

The milk sample to be analyzed is then introduced into the aqueous system, which has an arbitrary impurity concentration (step 5).

The weight of the milk sample collected in droplet form from the enemy water system produces a signal for the surface tension or weight by the pressure sensor, which is input to the controller (step 6).

In this way, the data input to the controller is compared with the data (or reference data) previously learned by the artificial neural network (step 7).

Finally, the data on the surface tension of the measured milk sample and the fitted reference data are compared with each other to output on the display unit the result of the impurity and its content (step 8).

Through the process described above, the controller can perform qualitative and quantitative analysis of impurities in the milk sample in an automated manner in real time based on the fitted data (learned irregularities). The system according to embodiments of the present invention can provide reliable analysis results in the detection of milk impurities by using an automated algorithm.

The present invention can be more clearly understood by the following examples, and the following examples are merely illustrative of the present invention and are not intended to limit the scope of the invention.

Example

The present inventors evaluated the feasibility of the system and method for detecting impurities in milk according to this embodiment by determining the correlation between the change of the impurities in the milk and the content thereof and the surface tension of the milk.

(Urea, detergent and baking soda) were mixed in a milk sample and the resultant was dropped for a predetermined period of time using an adiabatic system to measure the surface tension. To this end, milk samples were artificially prepared by adding impurities such as urea, detergent (or any other surfactant), baking soda, and the like to the milk.

Specifically, various concentrations of urea, detergent and baking soda were each incorporated into pure milk, with a range of 0 to 250 g / l at 50 g / l at a time. That is, five solutions (milk samples) containing urea impurities were prepared by adding 5 different amounts of urea to each 1 liter of milk (50 g, 100 g, 150 g, 200 g and 250 g, respectively). Similarly, a solution (milk sample) containing impurities was prepared in the same manner for each of detergent and baking soda.

Thereafter, the milk samples were loaded in the aqueous system and droplets of each of the milk samples were collected for one minute. The weight of the collected droplets was then measured according to the concentration of the individual sample. The sample weight thus measured was converted into a value related to the surface tension by the above-mentioned Equation (1) and plotted. The results are shown in Figs. 5A to 5C.

The figure shows the surface tension of the milk versus urea concentration (Fig. 5A), the surface tension of the detergent to the milk (Fig. 5B) and the surface tension of the milk to the baking soda (Fig.

Considering the above experimental results, the surface tension of the milk sample (or the weight of the collected milk sample) collected over a specific time interval from the red water system is similar (linearly) according to the concentration change of each of the three impurities Change can be seen. This suggests that the detection system and method provided in the present invention is useful for detecting not only the presence of impurities in milk but also the concentration of impurities.

It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present invention as defined by the appended claims.

Claims (6)

A method for detecting the type and content of impurities in a milk sample in an automated manner,
a) dropping the milk sample through the enemy water system into a container located beneath it;
b) measuring the weight of the milk sample collected into the vessel for a predetermined time using a pressure sensor; And
c) transmitting a signal of the pressure sensor according to the weight value of the measured milk sample to the controller; And
d) using the controller to determine a component of the impurities in the milk sample and its content;
/ RTI >
Wherein the controller derives the type of impurities in the milk sample and its content by comparing the data on the surface tension obtained from the delivered signal with the reference data fitted by the controller. 2. The method of claim 1, wherein the controller is an arduino microcontroller. 3. The method of claim 2, wherein the controller comprises an ATmega328P chip. 2. The method of claim 1, wherein step (d) is performed using an artificial neural network based on a Naive Bayes classifier to determine a difference between an already known impurity-containing milk sample and an already known impurity for at least 1000 samples And learning the relationship. A system for detecting the type and content of impurities in a milk sample in an automated manner,
An aqueous system configured to drop the introduced milk sample downward;
A container disposed in the underside of said aqueous system for collecting said distilled milk sample;
A pressure sensor attached or connected to the vessel to measure the weight of the milk sample and transmit a signal about the weight of the measured milk sample;
A controller for deriving the type of impurities in the milk sample and its content by comparing data on the surface tension of the milk sample obtained based on the signal transmitted from the pressure sensor with fitted reference data; And
A display unit for displaying the kind and content of the derived impurities;
/ RTI > 6. The method according to claim 5, further comprising an artificial neural network based on a Naive Bayes Classifier to study the relationship between already known impurity-containing milk samples and already known impurities system.

Images