CN112630293B - Method and system for identifying freshness of pork - Google Patents
Method and system for identifying freshness of pork Download PDFInfo
- Publication number
- CN112630293B CN112630293B CN202110255589.XA CN202110255589A CN112630293B CN 112630293 B CN112630293 B CN 112630293B CN 202110255589 A CN202110255589 A CN 202110255589A CN 112630293 B CN112630293 B CN 112630293B
- Authority
- CN
- China
- Prior art keywords
- pork
- freshness
- mass spectrum
- sample
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/62—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating the ionisation of gases, e.g. aerosols; by investigating electric discharges, e.g. emission of cathode
- G01N27/64—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating the ionisation of gases, e.g. aerosols; by investigating electric discharges, e.g. emission of cathode using wave or particle radiation to ionise a gas, e.g. in an ionisation chamber
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Toxicology (AREA)
- Evolutionary Computation (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Electrochemistry (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Other Investigation Or Analysis Of Materials By Electrical Means (AREA)
Abstract
The invention discloses a pork freshness identification method and a pork freshness identification system, wherein the method comprises the steps of obtaining pork samples with different freshness grades, detecting all the pork samples by using a rapid mass spectrometry device combining dynamic headspace water bath heating sampling, vacuum ultraviolet photoionization and ion trap mass spectrometry, and dividing the obtained mass spectrometry data into a training set and a verification set as a data set; building a pork freshness discrimination model based on a decision tree according to the training set, and verifying the model by using a verification set; when the model is verified to be effective, determining K pieces of biological marker information influencing the freshness of the pork; and acquiring mass spectrum data of the pork sample to be detected, inputting the mass spectrum data into a pork freshness discrimination model for discrimination, and outputting the freshness category. According to the method, pretreatment of the sample is not needed, mass spectrum data of the pork sample is directly obtained, and then a pork freshness distinguishing model is constructed, so that not only can the pork freshness be quickly distinguished, but also the information of the biological marker influencing the pork freshness can be intuitively obtained.
Description
Technical Field
The invention relates to the technical field of meat freshness detection, in particular to a pork freshness identification method and a pork freshness identification system.
Background
During the transportation and storage of pork, the freshness of the pork is continuously reduced, even decayed and deteriorated, and toxic and harmful chemical products are generated due to environmental factors such as storage conditions and storage time, and self factors, so that the freshness is an important measurement index for evaluating the quality of the pork.
The national standard detection method has the problems of complex pretreatment, complex process, long time consumption and the like, and is not suitable for large-batch rapid identification and detection because detection personnel with certain professional qualities need to be equipped. At present, the widely used high performance liquid chromatography, gas chromatography-mass spectrometry, near infrared spectroscopy and the like all need to carry out complex pretreatment on pork, extract a liquid to be detected and then carry out differential analysis, so that the time consumption is long, or the resolution ratio is low, and the rapid differential analysis of the pork quality is not facilitated. By computer technology simulating human senses, such as electronic nose, the accuracy of the technology is low due to the selectivity and limitation of the sensor and sensitivity to the influence of environmental humidity, temperature and other interference, only an identification result can be obtained, and biomarker information influencing pork freshness cannot be obtained.
Disclosure of Invention
The invention provides a pork freshness identification method and a pork freshness identification system, which are used for solving one or more technical problems in the prior art and at least providing a beneficial selection or creation condition.
In a first aspect, an embodiment of the present invention provides a method for identifying pork freshness, the method including:
s101, obtaining pork samples with different freshness grades;
s102, heating a pork sample in a water bath by using a water bath headspace device, and purging by using inert carrier gas to obtain a gas sample; ionizing the gas sample by using a vacuum ultraviolet lamp to obtain ions; capturing the ions by using an ion trap mass spectrum, and performing mass spectrum detection to obtain mass spectrum data;
s103, respectively carrying out mass spectrum detection on each pork sample through the step S102 to obtain mass spectrum data of all pork samples as a data set, and dividing the data set into a training set and a verification set;
s104, selecting mass spectrum data of N pork samples from a training set in a replaced sampling mode to form an intermediate data set, and then randomly selecting a plurality of features from a plurality of features corresponding to each mass spectrum data of the intermediate data set as the training data set, wherein N is a positive integer;
s105, repeating the step S104 until M training data sets are constructed, wherein M is a positive integer;
s106, establishing a completely split decision tree for each training data set to obtain the prediction results of the N pork samples in each decision tree, counting the prediction results of the N pork samples according to the prediction results of each decision tree, and respectively judging the freshness categories of the N pork samples according to the counting results;
s107, constructing a pork freshness distinguishing model based on a decision tree through the steps S104-S106, and carrying out validity verification on the pork freshness distinguishing model by using a verification set;
s108, when the pork is verified to be fresh, K pieces of biological marker information of the pork freshness are determined, wherein K is a positive integer;
s109, acquiring mass spectrum data of the pork sample to be detected, inputting the pork freshness discrimination model passing verification for identification, and outputting the freshness category.
Further, the mass spectrum data in step S102 and step S109 are both primary mass spectrum data.
Further, the K pieces of biomarker information for determining the freshness of pork include:
determining K ion fragments which have the largest influence on the output result of the pork freshness discrimination model;
determining a secondary mass spectrum of the K ion fragments;
and determining corresponding K pieces of biomarker information according to the secondary mass spectrum of the K pieces of ion fragments.
Further, the K biomarker information includes at least one of trimethylamine, trimethylamine hydrate, homocysteine, aspartic acid, glutamic acid, histamine, alanine, putrescine, methionine, dimethylpyrazine, chloramphenicol azide, and methyl alanine.
Further, the temperature range of the water bath in the water bath headspace device is 30-60 ℃, the time range for heating the pork sample in the water bath is 5-30 min, and the pressure range of the inert carrier gas is 0.5-1.5 MPa, so that the protein is degraded into trimethylamine in the pork decay and deterioration process.
Further, in step S102, the ion trap mass spectrum is scanned in the positive ion mode to capture ions, and the scanning range is determined according to the range of the ion fragments in the mass spectrogram of the pork sample.
Further, in step S107, when the pork freshness degree discrimination model based on the decision tree is constructed through steps S104 to S106, the number of decision trees of the pork freshness degree discrimination model is corrected so that the error of data outside the bag is lower than the threshold value.
In a second aspect, an embodiment of the present invention provides a pork freshness identification system, including:
the mass spectrum detection device is used for respectively carrying out mass spectrum detection on pork samples with different freshness grades to obtain mass spectrum data of all the pork samples as a data set, and the mass spectrum detection device comprises a water bath headspace device, a vacuum ultraviolet lamp and an ion trap mass spectrum, wherein the acquisition of the mass spectrum data of each pork sample comprises: heating the pork sample in a water bath by using a water bath headspace device, and purging by using inert carrier gas to obtain a gas sample; ionizing the gas sample by using a vacuum ultraviolet lamp to obtain ions; capturing the ions by using an ion trap mass spectrum, and performing mass spectrum detection to obtain mass spectrum data;
at least one processor;
at least one memory for storing at least one program;
the at least one processor executing the at least one program is in a unit of a system:
the dividing unit is used for dividing the data set into a training set and a verification set;
a training data set construction unit for constructing M training data sets, wherein the determination of each training data set comprises: selecting mass spectrum data of N pork samples from a training set in a replaced sampling mode to form an intermediate data set, and then randomly selecting a plurality of characteristics from a plurality of characteristics corresponding to each mass spectrum data of the intermediate data set as the training data set, wherein N is a positive integer;
the decision tree construction unit is used for establishing a completely split decision tree for each training data set to obtain the prediction result of the N pork samples in each decision tree, counting the prediction results of the N pork samples according to the prediction result of each decision tree, and respectively judging the freshness categories of the N pork samples according to the counting result;
the pork freshness degree distinguishing model building unit is used for building a pork freshness degree distinguishing model based on a decision tree through the training data set building unit and the decision tree building unit, and correcting the number of the decision trees of the pork freshness degree distinguishing model so as to enable the error of data outside the bag to be lower than a threshold value;
the verifying unit is used for verifying the validity of the pork freshness distinguishing model by using the verifying set;
and the identification unit is used for acquiring mass spectrum data of the pork sample to be detected, inputting the pork freshness identification model passing the verification for identification, and outputting the freshness category.
Further, the mass spectrum detection device further comprises an outer cover, the output end of the air outlet pipe penetrates through the outer cover to be located in the outer cover, the vacuum ultraviolet lamp is located in the outer cover, the outer cover is provided with an opening, and a mass spectrum port of the ion trap mass spectrum is placed in the outer cover through the opening so as to be connected with the outer cover in a sealing mode.
The method and the system for identifying the freshness of the pork provided by the embodiment of the invention at least have the following beneficial effects: the method for identifying the freshness of the pork can be used for directly obtaining mass spectrum data of the pork sample without preprocessing the sample, and then combining an integrated learning method, constructing a pork freshness identification model based on a decision tree, not only can quickly and intelligently identifying the freshness of the pork, but also can intelligently determine the biomarker information of the freshness of the pork, so that the biomarker information influencing the freshness of the pork can be more intuitively obtained.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the example serve to explain the principles of the invention and not to limit the invention.
FIG. 1 is a flow chart of a method for identifying pork freshness according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a mass spectrum detection apparatus according to an embodiment of the present invention;
FIG. 3(a), FIG. 3(b), FIG. 3(c), FIG. 3(d), FIG. 3(e) and FIG. 3(f) are primary mass spectra of a pork sample in positive ion mode at-8 ℃ for 0day, 1day, 2 days, 3 days, 7 days and 10 days, respectively;
FIG. 4(a) is a diagram illustrating a result of quantity optimization of a decision tree according to an embodiment of the present invention;
FIG. 4(b) is a diagram of the classification result of the pork freshness discrimination model provided by the embodiment of the present invention on the training set;
FIG. 5 is a schematic diagram of the first 12 variables of the importance ranking of the pork freshness identification model provided by the embodiment of the present invention;
FIG. 6 is a secondary mass spectrum of the pork freshness biomarker in positive ion mode according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
It should be noted that although functional block divisions are provided in the system drawings and logical orders are shown in the flowcharts, in some cases, the steps shown and described may be performed in different orders than the block divisions in the systems or in the flowcharts. The terms first, second and the like in the description and in the claims, and the drawings described above, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order.
Fig. 1 is a pork freshness identification method provided by an embodiment of the present invention, which includes the following steps:
s101, obtaining pork samples with different freshness grades;
specifically, each pork sample is 5g, and different freshness grades refer to pork samples preserved under refrigeration for different periods of time, including, for example, 0day, 1day, 2 days, 3 days, 7 days, 10 days, and the like.
S102, heating a pork sample in a water bath by using a water bath headspace device, and purging by using inert carrier gas to obtain a gas sample; ionizing the gas sample by using a vacuum ultraviolet lamp to obtain ions; capturing the ions by using an ion trap mass spectrum, and performing mass spectrum detection to obtain mass spectrum data;
specifically, as shown in fig. 2, a schematic structural diagram of a mass spectrometry detection device is provided, which comprises a water bath headspace device 1, a vacuum ultraviolet lamp 2 and an ion trap mass spectrum 3, wherein the water bath headspace device 1 comprises an air inlet pipe 11 and an air outlet pipe 12, a pork sample 4 is put into the water bath headspace device 1, sealed and heated in a water bath, and inert carrier gas, such as N, is used for heating in a water bath2Enters the water bath headspace device 1 through the air inlet pipe 11, sweeps the pork sample 4, carries the volatilized gas sample to be led into the front end lower part of the vacuum ultraviolet lamp 2 through the air outlet pipe 12, the optical axis of the vacuum ultraviolet lamp is vertical to the air outlet direction of the air outlet pipe 12, and the vacuum ultraviolet lamp 2 is used for gas sampleIonizing the product to obtain ions, capturing the ions by using an ion trap mass spectrum 3, and performing mass spectrum detection to obtain mass spectrum data.
The mass spectrum detection device further comprises an outer cover 5, the air outlet pipe 12 penetrates through the outer cover 5 at the output end to be located in the outer cover 5, the vacuum ultraviolet lamp is located in the outer cover 5, the outer cover 5 is provided with an opening, and a mass spectrum port of the ion trap mass spectrum is placed in the outer cover 5 through the opening, so that the ion trap mass spectrum 3 is connected with the outer cover 5 in a sealing mode. The setting of dustcoat 5 makes vacuum ultraviolet lamp 2 avoid external environment to disturb when carrying out the ionization to the gas sample, and the ionization effect is better to also make the ion better get into the ion trap mass spectrum through the mass spectrum mouth.
The mass spectrum data obtained in step S102 is primary mass spectrum data.
In one embodiment, mass spectrometry is performed using a compound having a mass to charge ratio between m/z 50-500 in positive ion mode as a representative fingerprint.
Orthogonal experiments are designed and optimized on parameters of water bath temperature, water bath time and carrier gas pressure, and a final product m/z60 (trimethylamine) of protein degradation in the pork spoilage process is used as an optimized index. The water bath temperature range in the water bath headspace device 1 is 30-60 ℃, the time range for heating the pork sample in the water bath is 5-30 min, the pressure range of the inert carrier gas is 0.5-1.5 MPa, so that the protein is degraded into trimethylamine in the pork decay and deterioration process, and the temperature range of an ion transmission tube for transmitting ions in the ion trap mass spectrum is 100-250 ℃.
In one embodiment, a pork sample is placed in a water bath headspace device 1, water bath heating is carried out with water bath temperature of 40 ℃, heating time is 20min, and N with carrier gas pressure of 0.5MPa is used2The pork samples were purged.
The mass spectrum detection device can directly obtain the mass spectrum data of the pork sample through the operation of the step S102, and the mass spectrum data is the mixed mass spectrum data of all the components of the sample; the traditional chromatography-mass spectrometry technology needs to apply a chromatography technology to separate the components of a sample one by one and collect mass spectrum data of a single chromatographic peak component; the mass spectrum detection device can quickly obtain the mixed mass spectrum data of all the components of the pork sample, and is simple to operate and high in efficiency.
S103, respectively carrying out mass spectrum detection on each pork sample through the step S102 to obtain mass spectrum data of all pork samples as a data set, and dividing the data set into a training set and a verification set;
specifically, the pork samples obtained in step S101 are sequentially detected in step S102 to obtain mass spectrum data of all the pork samples as a data set, the data set is divided into a training set and a verification set according to a certain proportion (set according to actual conditions), the pork freshness degree discrimination model is trained by using the training set, and the pork freshness degree discrimination model is verified by using the verification set.
FIGS. 3(a) -3 (f) are primary mass spectra of pork samples in positive ion mode at-8 ℃ for 0day, 1day, 2 days, 3 days, 7 days, and 10 days (as labeled as 0d, 1d, 2d, 3d, 7d, and 10d in FIGS. 3(a) -3 (f), respectively). The abscissa represents the mass-to-charge ratio (m/z) and the ordinate represents the percentage of relative abundance (relative abundance).
S104, selecting mass spectrum data of N pork samples from a training set in a replaced sampling mode to form an intermediate data set, and then randomly selecting a plurality of features from a plurality of features corresponding to each mass spectrum data of the intermediate data set as the training data set, wherein N is a positive integer;
specifically, the training data set includes a number of features corresponding to the mass spectral data for each of the selected N samples.
S105, repeating the step S104 until M training data sets are constructed, wherein M is a positive integer;
s106, establishing a completely split decision tree for each training data set to obtain the prediction results of the N pork samples in each decision tree, counting the prediction results of the N pork samples according to the prediction results of each decision tree, and respectively judging the freshness categories of the N pork samples according to the counting results;
specifically, the freshness classification is divided into different classes, and the number of classes is set according to actual needs, for example, class i is stored for 0-1 day, class ii is stored for 2-3 days, and class iii is stored for 7-10 days.
And each pork sample has M prediction results, the M prediction results are counted, and when the number of the prediction results with the same prediction results is more than a × M, the freshness category of the pork sample is judged to be the same as the prediction results with the number of the same prediction results more than a × M. a is a preset percentage. For example, a =90%, and if 95% of the M predicted results of the pork samples indicate that the freshness of the pork samples is class I, the freshness class of the pork samples is determined to be class I.
S107, constructing a pork freshness distinguishing model based on a decision tree through the steps S104-S106, and verifying the pork freshness distinguishing model by using a verification set;
when the pork freshness degree discrimination model is constructed by using the training set in the steps S104-S106, the pork freshness degree discrimination model is constructed by using the primary mass spectrum data of the pork sample, and the number of decision trees of the pork freshness degree discrimination model is optimized, so that the out-of-bag data error (OOB) is lower than a set threshold value; in one embodiment, the number of decision trees ranges from 100 to 500, and the threshold of OOB is set to 0.005.
The verification set specifically verifies the pork freshness discrimination model as follows: and inputting the verification set into a pork freshness judging model, and judging whether the classification of the pork freshness judging model is accurate or not.
FIG. 4(a) is a graph of the result of quantity optimization of decision Trees, where the abscissa is the Number of decision Trees (Number of growing Trees), and the ordinate represents the Out-of-Bag data Error (OOB), when the Number of decision Trees is 0-100, the Out-of-Bag data Error (OOB) tends to decrease with the increase of the Number of decision Trees, and the rate of decrease gradually slows down until the Number of decision Trees is 100-500, and the Out-of-Bag data Error (OOB) tends to a stable value. Thus, in one embodiment, the number of decision trees is chosen to be 100, and the error of the classification is less than 0.005, which is the optimal value. Modeling is performed on sample data of a training set in the optimal setting, and the result is a two-dimensional classification diagram of multi-scale analysis (MDS) as shown in fig. 4(b), wherein the principle of the multi-scale analysis is to construct a proper low-dimensional space by utilizing the similarity between paired samples, so that the distance of the samples in the space and the similarity between the samples in the high-dimensional space are kept consistent as much as possible, and visualization is realized. The abscissa and the ordinate are coordinate variables corresponding to the first two dimensions in the low-dimensional space. From the results, when pork was stored at-8 ℃ under refrigeration, the freshness grade of the sample was clearly classified into 3 categories, i category when stored for 0-1 day (0, 1 day), ii category when stored for 2-3 days (2, 3 day), and iii category when stored for 7-10 days (7, 10 day). In the figure 4(b), the three types of samples are distinguished obviously, have no overlap and have good identification effect.
S108, determining K pieces of biological marker information of pork freshness when the verification is passed;
specifically, when the accuracy of the verification result is greater than the accuracy threshold, the verification is considered to be passed.
Table 1 shows the discrimination results of the discrimination model for pork freshness for 75 samples in the verification set. The results show that the samples of the three freshness grades can be correctly distinguished, and the accuracy reaches 100%.
TABLE 1 discrimination results for validation set samples
The proteins of pork are slowly degraded during the cold storage process to generate other amino acids, such as m/z60 (trimethylamine), m/z90 (alanine), m/z106 (serine), m/z148 (glutamic acid), m/z180 (aminosugar), and the like. Therefore, the information on the content of these amino acids in the pork sample reflects its freshness.
In the high performance liquid chromatography and gas chromatography-mass spectrometry technology in the prior art, sample pretreatment is complex, detection time is long, and detection cost is high; techniques such as an electronic nose for simulating human senses have low accuracy and cannot obtain more specific biomarker information which influences pork freshness.
The identification method can well detect the ion peaks of the amino acids of the pork sample, and extract the characteristic information of the amino acid ion fragments from the mass spectrum data of the sample by using an ensemble learning method, so as to judge the freshness of the pork sample at different storage times. According to the modeling result of the pork freshness distinguishing model, the method can know the variable which greatly contributes to the model, and the corresponding ion fragments can be used as important biomarkers for evaluating the pork freshness. Fig. 5 is a schematic diagram of the variables at the top 12 of the Importance ranking of the pork freshness degree discrimination model, wherein the abscissa represents the mass-to-charge ratio (m/z), the ordinate represents the Importance weight of the Feature variable (Feature Importance), the 12 variables have a contribution rate to the model of 95% or more, and the m/z of the 12 variables are 60, 136, 78, 134, 148, 112, 90, 89, 150, 109, 296 and 104 in sequence. The secondary mass spectra of these ion fragments were collected for qualitative analysis, and the results are shown in FIG. 6, in which FIG. 6 is the secondary mass spectrum of the pork freshness biomarker in positive ion mode, the abscissa represents the mass-to-charge ratio (m/z), the ordinate represents the percentage of relative abundance (relative abundance), and FIG. 6 is composed of three rows and four columns of minimaps, wherein, the mark (a) is the minimap in the first row and the first column, the mark (b) is the minimap in the second row and the first column, the mark (c) is the minimap in the first row and the third column, the mark (d) is the minimap in the fourth row and the first column, the mark (e) is the minimap in the first row and the first column, the mark (f) is the minimap in the second row and the second column, the mark (g) is the minimap in the third row and the third column, the mark (h) is the minimap in the fourth row and the mark (i) is the minimap in the first row and the third column, the panels marked (k) and (l) in the third and fourth rows, respectively, are: m/z60 (trimethylamine), m/z78 (trimethylamine hydrate), m/z136 (homocysteine), m/z134 (aspartic acid), m/z148 (glutamic acid), m/z112 (histamine), m/z90 (alanine), m/z89 (putrescine), m/z150 (methionine), m/z109 (dimethylpyrazine), m/z296 (chloramphenicol azide), and m/z104 (cadaverine), as shown in the panels labeled (a), (b), (c), (d), (e), (f), (g), (h), (i), (j), (k), (l) in FIG. 6, respectively. After the model is established, the biological marker information of pork freshness is determined, so that the biological marker information influencing the pork freshness can be obtained more intuitively.
S109, acquiring mass spectrum data of the pork sample to be detected, inputting the pork freshness discrimination model passing verification for identification, and outputting the freshness category.
Specifically, mass spectrum data of the pork sample to be detected can be acquired through the mass spectrum detection device shown in fig. 2, the mass spectrum data is primary mass spectrum data, and the mass spectrum data is input into the pork freshness discrimination model for discrimination to obtain the freshness category. The mass spectrum data is directly extracted, the pretreatment of pork samples is not needed, and the identification of the freshness of the pork can be quickly realized.
The embodiment of the invention provides a pork freshness identification system, which comprises:
the mass spectrum detection device is used for respectively carrying out mass spectrum detection on pork samples with different freshness grades to obtain mass spectrum data of all the pork samples as a data set, and the mass spectrum detection device comprises a water bath headspace device, a vacuum ultraviolet lamp and an ion trap mass spectrum, wherein the acquisition of the mass spectrum data of each pork sample comprises: heating the pork sample in a water bath by using a water bath headspace device, and purging by using inert carrier gas to obtain a gas sample; ionizing the gas sample by using a vacuum ultraviolet lamp to obtain ions; capturing the ions by using an ion trap mass spectrum, and performing mass spectrum detection to obtain mass spectrum data;
at least one processor;
at least one memory for storing at least one program;
the at least one processor executing the at least one program is in a unit of a system:
the dividing unit is used for dividing the data set into a training set and a verification set;
a training data set construction unit for constructing M training data sets, wherein the determination of each training data set comprises: selecting mass spectrum data of N pork samples from a training set in a replaced sampling mode to form an intermediate data set, and then randomly selecting a plurality of characteristics from a plurality of characteristics corresponding to each mass spectrum data of the intermediate data set as the training data set, wherein N is a positive integer;
the decision tree construction unit is used for establishing a completely split decision tree for each training data set to obtain the prediction result of the N pork samples in each decision tree, counting the prediction results of the N pork samples according to the prediction result of each decision tree, and respectively judging the freshness categories of the N pork samples according to the counting result;
the pork freshness degree distinguishing model building unit is used for building a pork freshness degree distinguishing model based on a decision tree through the training data set building unit and the decision tree building unit, and correcting the number of the decision trees of the pork freshness degree distinguishing model so as to enable the error of data outside the bag to be lower than a threshold value;
the verifying unit is used for verifying the validity of the pork freshness distinguishing model by using the verifying set;
and the identification unit is used for acquiring mass spectrum data of the pork sample to be detected, inputting the pork freshness identification model passing the verification for identification, and outputting the freshness category.
Further, the mass spectrum detection device further comprises an outer cover, the output end of the air outlet pipe penetrates through the outer cover to be located in the outer cover, the vacuum ultraviolet lamp is located in the outer cover, the outer cover is provided with an opening, and a mass spectrum port of the ion trap mass spectrum is placed in the outer cover through the opening so as to be connected with the outer cover in a sealing mode.
One of ordinary skill in the art will appreciate that all or some of the steps, systems, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those of ordinary skill in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as known to those skilled in the art.
While the preferred embodiments of the present invention have been described in detail, it will be understood by those skilled in the art that the foregoing and various other changes, omissions and deviations in the form and detail thereof may be made without departing from the scope of this invention.
Claims (6)
1. A pork freshness identification method is characterized by comprising the following steps:
s101, obtaining pork samples with different freshness grades;
s102, heating a pork sample in a water bath by using a water bath headspace device, and purging by using inert carrier gas to obtain a gas sample; ionizing the gas sample by using a vacuum ultraviolet lamp to obtain ions; capturing the ions by using an ion trap mass spectrum, and carrying out mass spectrum detection to obtain mass spectrum data, wherein the water bath temperature range in the water bath headspace device is 30-60 ℃, the water bath heating time range for the pork sample is 5-30 min, and the inert carrier gas pressure range is 0.5-1.5 MPa;
s103, respectively carrying out mass spectrum detection on each pork sample through the step S102 to obtain mass spectrum data of all pork samples as a data set, and dividing the data set into a training set and a verification set;
s104, selecting mass spectrum data of N pork samples from a training set in a replaced sampling mode to form an intermediate data set, and then randomly selecting a plurality of features from a plurality of features corresponding to each mass spectrum data of the intermediate data set as the training data set, wherein N is a positive integer;
s105, repeating the step S104 until M training data sets are constructed, wherein M is a positive integer;
s106, establishing a completely split decision tree for each training data set to obtain the prediction results of the N pork samples in each decision tree, counting the prediction results of the N pork samples according to the prediction results of each decision tree, and respectively judging the freshness categories of the N pork samples according to the counting results;
s107, constructing a pork freshness distinguishing model based on a decision tree through the steps S104-S106, and carrying out validity verification on the pork freshness distinguishing model by using a verification set;
s108, when the pork is verified to be fresh, K pieces of biological marker information of the pork freshness are determined, wherein K is a positive integer; the K pieces of biomarker information for determining the freshness of the pork comprise:
determining K ion fragments which have the largest influence on the output result of the pork freshness discrimination model;
determining a secondary mass spectrum of the K ion fragments;
determining corresponding K pieces of biological marker information according to the secondary mass spectrum of the K pieces of ion fragments;
the K biomarker information includes trimethylamine, trimethylamine hydrate, homocysteine, aspartic acid, glutamic acid, histamine, alanine, putrescine, methionine, dimethyl pyrazine, chloramphenicol azide, and methyl alanine;
s109, acquiring mass spectrum data of the pork sample to be detected, inputting the pork freshness discrimination model passing verification for identification, and outputting freshness categories, wherein the freshness categories are divided into 3 categories.
2. The method for identifying pork freshness according to claim 1, wherein the mass spectrometry data in step S102 and step S109 are primary mass spectrometry data.
3. The method for identifying pork freshness according to claim 1, wherein in step S102, the ion trap mass spectrum is scanned in a positive ion mode to capture ions, and the scanning range is determined according to the range of ion fragments in the mass spectrum of the pork sample.
4. The method for identifying pork freshness according to claim 1, wherein in step S107, when the decision tree-based pork freshness determination model is constructed through steps S104 to S106, the number of decision trees of the pork freshness determination model is modified so that the error of data outside the bag is lower than a threshold value.
5. An identification system for pork freshness, comprising:
the mass spectrum detection device is used for respectively carrying out mass spectrum detection on pork samples with different freshness grades to obtain mass spectrum data of all the pork samples as a data set, and the mass spectrum detection device comprises a water bath headspace device, a vacuum ultraviolet lamp and an ion trap mass spectrum, wherein the acquisition of the mass spectrum data of each pork sample comprises: heating the pork sample in a water bath by using a water bath headspace device, and purging by using inert carrier gas to obtain a gas sample; ionizing the gas sample by using a vacuum ultraviolet lamp to obtain ions; capturing the ions by using an ion trap mass spectrum, and carrying out mass spectrum detection to obtain mass spectrum data, wherein the water bath temperature range in the water bath headspace device is 30-60 ℃, the water bath heating time range for the pork sample is 5-30 min, and the inert carrier gas pressure range is 0.5-1.5 MPa;
at least one processor;
at least one memory for storing at least one program;
the at least one processor executing the at least one program runs in the elements of the system:
the dividing unit is used for dividing the data set into a training set and a verification set;
a training data set construction unit for constructing M training data sets, wherein the determination of each training data set comprises: selecting mass spectrum data of N pork samples from a training set in a replaced sampling mode to form an intermediate data set, and then randomly selecting a plurality of characteristics from a plurality of characteristics corresponding to each mass spectrum data of the intermediate data set as the training data set, wherein N is a positive integer;
the decision tree construction unit is used for establishing a completely split decision tree for each training data set to obtain the prediction result of the N pork samples in each decision tree, counting the prediction results of the N pork samples according to the prediction result of each decision tree, and respectively judging the freshness categories of the N pork samples according to the counting result;
the pork freshness degree distinguishing model building unit is used for building a pork freshness degree distinguishing model based on a decision tree through the training data set building unit and the decision tree building unit, and correcting the number of the decision trees of the pork freshness degree distinguishing model so as to enable the error of data outside the bag to be lower than a threshold value;
the verification unit is used for verifying the validity of the pork freshness discrimination model by using the verification set, and when the verification is passed, K pieces of biological marker information of the pork freshness are determined, wherein the K pieces of biological marker information of the pork freshness comprise:
determining K ion fragments which have the largest influence on the output result of the pork freshness discrimination model;
determining a secondary mass spectrum of the K ion fragments;
determining corresponding K pieces of biological marker information according to the secondary mass spectrum of the K pieces of ion fragments;
the K biomarker information includes trimethylamine, trimethylamine hydrate, homocysteine, aspartic acid, glutamic acid, histamine, alanine, putrescine, methionine, dimethyl pyrazine, chloramphenicol azide, and methyl alanine;
and the identification unit is used for acquiring mass spectrum data of the pork sample to be detected, inputting the pork freshness identification model passing verification for identification, and outputting freshness categories, wherein the freshness categories are divided into 3 categories.
6. The pork freshness identification system according to claim 5, wherein the mass spectrometer device further comprises a housing, the output end of the air outlet pipe penetrates through the housing and is located in the housing, the vacuum ultraviolet lamp is located in the housing, the housing is provided with an opening, and the mass spectrometer port of the ion trap mass spectrometer is placed in the housing through the opening, so that the ion trap mass spectrometer is hermetically connected with the housing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110255589.XA CN112630293B (en) | 2021-03-09 | 2021-03-09 | Method and system for identifying freshness of pork |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110255589.XA CN112630293B (en) | 2021-03-09 | 2021-03-09 | Method and system for identifying freshness of pork |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112630293A CN112630293A (en) | 2021-04-09 |
| CN112630293B true CN112630293B (en) | 2021-06-01 |
Family
ID=75297780
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110255589.XA Active CN112630293B (en) | 2021-03-09 | 2021-03-09 | Method and system for identifying freshness of pork |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112630293B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114199989B (en) * | 2021-11-05 | 2023-11-28 | 佛山科学技术学院 | Method and system for identifying pericarpium citri reticulatae based on mass spectrum data fusion |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103424428A (en) * | 2013-06-05 | 2013-12-04 | 浙江工商大学 | Method for quickly detecting pork freshness based on electronic nose |
| CN108072692A (en) * | 2016-11-17 | 2018-05-25 | 中国科学院大连化学物理研究所 | A kind of portable drugs scene authentication equipment and method |
-
2021
- 2021-03-09 CN CN202110255589.XA patent/CN112630293B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN112630293A (en) | 2021-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11460453B2 (en) | Method for determining food-product quality and food-product quality determination device | |
| US7676329B2 (en) | Method and system for processing multi-dimensional measurement data | |
| US7462817B2 (en) | Mass spectrometer and method of analyzing isomers | |
| CN112578039B (en) | Method for analyzing and identifying adulterants in bird's nest | |
| CN108445094A (en) | Gaseous ion migration spectrum quickly differentiates method for building up and the application of age of Chinese rice wine | |
| US20080128607A1 (en) | Mass Spectrometry Analysis Method and System | |
| JP4857000B2 (en) | Mass spectrometry system | |
| CN111060642A (en) | Method for classifying and identifying tobacco leaves of same variety and different producing areas | |
| CN113138248B (en) | Characteristic spectrum extraction method and detection method for feed liquid preparation quality stability | |
| CN111044638A (en) | Method for classifying and identifying different varieties of flue-cured tobacco leaves | |
| CN109655529A (en) | A method of identifying genuine-fake cigarette | |
| EP1646866A2 (en) | Methods and systems for the annotation of biomolecule patterns in chromatography/mass-spectrometry analysis | |
| CN104713845A (en) | Mixture component identification method based on terahertz absorption spectrum processing | |
| CN112630293B (en) | Method and system for identifying freshness of pork | |
| CN111537659A (en) | Method for screening biomarkers | |
| KR101958900B1 (en) | Apparatus and method for quantification analysis of sample using liquid chromatograph-mass spectrometry | |
| CN111505174A (en) | Method for identifying true and false cigarettes | |
| CN109030611A (en) | Method for distinguishing is sentenced in a kind of quick, the lossless fingerprint analysis of green tea fragrance and odor type based on ion mobility spectrometry | |
| CN106526053A (en) | Method for quickly evaluating product quality conformity of tobacco by using HS-IMR-MS (headspace-ion molecule reaction-mass spectrometer) | |
| JP2009156722A (en) | Mass spectrometric spectral analysis method | |
| US10444206B2 (en) | Chromatography/mass spectrometry data processing device | |
| ES2767375B2 (en) | PROGRAM METHOD, SYSTEM AND PRODUCT FOR THE IDENTIFICATION OF COMPOUNDS IN COMPLEX BIOLOGICAL OR ENVIRONMENTAL SAMPLES | |
| CN113607850A (en) | Method for analyzing and identifying wheat varieties by utilizing volatile organic compounds | |
| CN113155946B (en) | Pork species identification method, pork mass spectrum detection method and device | |
| CN114199989A (en) | Method and system for identifying pericarpium citri reticulatae based on mass spectrum data fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |