CN117131349B - Gynostemma pentaphylla processing method and system based on fresh root cleaning - Google Patents
Gynostemma pentaphylla processing method and system based on fresh root cleaning Download PDFInfo
- Publication number
- CN117131349B CN117131349B CN202311403134.3A CN202311403134A CN117131349B CN 117131349 B CN117131349 B CN 117131349B CN 202311403134 A CN202311403134 A CN 202311403134A CN 117131349 B CN117131349 B CN 117131349B
- Authority
- CN
- China
- Prior art keywords
- raw material
- processing
- feature
- gynostemma pentaphylla
- group
- 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
- 241001065361 Gynostemma Species 0.000 title claims abstract description 167
- 235000002956 Gynostemma pentaphyllum Nutrition 0.000 title claims abstract description 167
- 238000004140 cleaning Methods 0.000 title claims abstract description 50
- 238000003672 processing method Methods 0.000 title claims abstract description 17
- 238000012545 processing Methods 0.000 claims abstract description 248
- 239000002994 raw material Substances 0.000 claims abstract description 234
- 238000000605 extraction Methods 0.000 claims abstract description 51
- 230000011218 segmentation Effects 0.000 claims abstract description 31
- 238000012549 training Methods 0.000 claims abstract description 23
- 238000010183 spectrum analysis Methods 0.000 claims abstract description 13
- 238000001914 filtration Methods 0.000 claims description 81
- 238000000034 method Methods 0.000 claims description 71
- 238000004422 calculation algorithm Methods 0.000 claims description 53
- 238000001819 mass spectrum Methods 0.000 claims description 49
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 26
- 238000012216 screening Methods 0.000 claims description 22
- 238000009826 distribution Methods 0.000 claims description 17
- 230000008485 antagonism Effects 0.000 claims description 15
- 238000000926 separation method Methods 0.000 claims description 13
- 230000004927 fusion Effects 0.000 claims description 10
- 238000013507 mapping Methods 0.000 claims description 9
- 238000011176 pooling Methods 0.000 claims description 9
- 230000000903 blocking effect Effects 0.000 claims description 7
- 238000001514 detection method Methods 0.000 claims description 7
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 230000005856 abnormality Effects 0.000 claims description 6
- 230000003321 amplification Effects 0.000 claims description 6
- 239000003814 drug Substances 0.000 claims description 6
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 6
- 238000013375 chromatographic separation Methods 0.000 claims description 5
- 238000002922 simulated annealing Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 4
- 230000035772 mutation Effects 0.000 claims description 3
- 238000005192 partition Methods 0.000 claims description 3
- 230000004931 aggregating effect Effects 0.000 claims description 2
- 239000000463 material Substances 0.000 abstract description 5
- 239000000047 product Substances 0.000 description 78
- 230000008569 process Effects 0.000 description 32
- 230000006870 function Effects 0.000 description 17
- 239000007789 gas Substances 0.000 description 9
- 239000013598 vector Substances 0.000 description 5
- 238000013473 artificial intelligence Methods 0.000 description 4
- 239000012467 final product Substances 0.000 description 3
- 241000196324 Embryophyta Species 0.000 description 2
- 150000001875 compounds Chemical class 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 229940079593 drug Drugs 0.000 description 2
- 238000012804 iterative process Methods 0.000 description 2
- 238000004949 mass spectrometry Methods 0.000 description 2
- 238000007637 random forest analysis Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 244000025254 Cannabis sativa Species 0.000 description 1
- 241000219104 Cucurbitaceae Species 0.000 description 1
- 241000219071 Malvaceae Species 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000012159 carrier gas Substances 0.000 description 1
- 230000009194 climbing Effects 0.000 description 1
- 238000004440 column chromatography Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002330 electrospray ionisation mass spectrometry Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 241000411851 herbal medicine Species 0.000 description 1
- 238000010813 internal standard method Methods 0.000 description 1
- 238000000534 ion trap mass spectrometry Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000004885 tandem mass spectrometry Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
- A61K36/00—Medicinal preparations of undetermined constitution containing material from algae, lichens, fungi or plants, or derivatives thereof, e.g. traditional herbal medicines
- A61K36/18—Magnoliophyta (angiosperms)
- A61K36/185—Magnoliopsida (dicotyledons)
- A61K36/42—Cucurbitaceae (Cucumber family)
- A61K36/424—Gynostemma
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
- A61K2236/00—Isolation or extraction methods of medicinal preparations of undetermined constitution containing material from algae, lichens, fungi or plants, or derivatives thereof, e.g. traditional herbal medicine
- A61K2236/10—Preparation or pretreatment of starting material
- A61K2236/13—Preparation or pretreatment of starting material involving cleaning, e.g. washing or peeling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Natural Medicines & Medicinal Plants (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biotechnology (AREA)
- Mycology (AREA)
- Chemical & Material Sciences (AREA)
- Alternative & Traditional Medicine (AREA)
- Quality & Reliability (AREA)
- Botany (AREA)
- Medical Informatics (AREA)
- Medicinal Chemistry (AREA)
- Microbiology (AREA)
- Multimedia (AREA)
- Pharmacology & Pharmacy (AREA)
- Epidemiology (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the field of medicinal material processing, and relates to a gynostemma pentaphylla processing method based on fresh root cleaning, which comprises the following steps of: sequentially carrying out multidimensional feature extraction and feature standardization operation on the processing flow data set to obtain a processing flow feature set; sequentially carrying out root detail segmentation and global feature extraction operation on the raw material atlas to obtain a raw material feature set; carrying out gas spectrum analysis and component matching operation on the finished product sample set in sequence to obtain a finished product feature set; performing iterative performance training on a preset gynostemma pentaphylla processing model by using the raw material characteristic set, the processing flow characteristic set and the finished product characteristic set to obtain a gynostemma pentaphylla prediction model; and obtaining a real-time raw material picture of the target gynostemma pentaphylla raw material to be processed, calculating the standard flow characteristics corresponding to the real-time raw material picture by utilizing the gynostemma pentaphylla prediction model, and processing. The invention also provides a gynostemma pentaphylla processing system based on fresh root cleaning. The invention can improve the processing efficiency of gynostemma pentaphylla.
Description
Technical Field
The invention relates to the technical field of medicinal material processing, in particular to a gynostemma pentaphylla processing method and system based on fresh root cleaning.
Background
Gynostemma pentaphylla is a common herb plant, belongs to the mallow family, is widely distributed on the whole world, and is particularly widely applied to the field of medicines in dry and semiarid regions, so that the gynostemma pentaphylla needs to be processed.
The existing gynostemma pentaphylla processing method is mainly based on a manual regulation and control processing method, according to long-term herbal medicine processing experience, the temperature, the time length, the humidity, the pressure and the like of drying are manually adjusted, in practical application, the manual regulation and control processing method needs pharmaceutical personnel to unify the quality of medicinal materials, the processed medicinal materials are single in variety, for example, under the requirement of root medicine, a great amount of time is consumed for cleaning fresh roots, the quality of finished products of the processed medicinal materials cannot be accurately controlled, and the processing efficiency of gynostemma pentaphylla is low.
Disclosure of Invention
The invention provides a gynostemma pentaphylla processing method and system based on fresh root cleaning, and mainly aims to solve the problem of low efficiency in processing gynostemma pentaphylla.
In order to achieve the above purpose, the invention provides a gynostemma pentaphylla processing method based on fresh root cleaning, which comprises the following steps:
Sequentially carrying out data splitting and data cleaning operations on the pre-acquired historical gynostemma pentaphylla processing data to obtain a processing flow data set, and sequentially carrying out multidimensional feature extraction and feature standardization operations on the processing flow data set to obtain a processing flow feature set;
acquiring a raw material atlas and a finished product sample set corresponding to the processing flow data set, sequentially performing root detail segmentation and global feature extraction operations on the raw material atlas to obtain a raw material feature set, wherein the sequentially performing root detail segmentation and global feature extraction operations on the raw material atlas to obtain the raw material feature set comprises the following steps: selecting raw material pictures in the raw material picture set one by one as target raw material pictures, and sequentially carrying out picture denoising and size stretching operation on the target raw material pictures to obtain specification raw material pictures; and performing primary feature extraction on the specification raw material picture by using the following linear filtering algorithm to obtain specification raw material features:
；
wherein,means that the coordinates in the raw material picture with the specification are +.>Is the specification raw material characteristic obtained by filtering the filtering convolution kernel center of the linear filtering algorithm, < ->The abscissa of the pixel being the center of the filter convolution kernel, >Is the ordinate of the pixel in the center of the filter convolution kernel,>is a preset amplification factor, < >>Is an exponential function, ++>As cosine function +.>Is the filtering direction of the linear filtering algorithm, < >>Is a sine function +.>Means the filtering wavelength of said linear filtering algorithm, < > or->Is the standard deviation of the Gaussian distribution, +.>Is the circumference rate, < >>Means the filtering wavelength of said linear filtering algorithm, < > or->Means the filtering phase of the linear filtering algorithm, < >>Is a preset offset coefficient; performing picture blocking operation on the specification raw material pictures to obtain raw material block groups; extracting a root block group from the raw material block group, and extracting a root feature group from the root block group by using the linear filtering formula; performing global fusion operation on the root feature group and the specification raw material features to obtain raw material features, and collecting all the raw material features into a raw material feature set;
carrying out gas spectrum analysis and component matching operation on the finished product sample set in sequence to obtain a finished product feature set;
performing iterative performance training on a preset gynostemma pentaphylla processing model by using the raw material characteristic set, the processing flow characteristic set and the finished product characteristic set to obtain a gynostemma pentaphylla prediction model;
And obtaining a real-time raw material picture of a target gynostemma pentaphylla raw material to be processed, calculating a standard flow characteristic corresponding to the real-time raw material picture by using the gynostemma pentaphylla prediction model, and processing the target gynostemma pentaphylla raw material according to the standard flow characteristic.
Optionally, the sequentially performing data splitting and data cleaning operations on the pre-acquired historical gynostemma pentaphylla processing data to obtain a processing flow data set, including:
extracting a processing time stamp set from the historical gynostemma pentaphylla processing data, and splitting the historical gynostemma pentaphylla processing data into a primary processing data set according to the processing time stamp set;
performing unit formatting operation on the primary processing data set to obtain a secondary processing data set;
performing minimum hash coding on the secondary processing data set to obtain a processing data hash coding set;
performing data deduplication on the secondary processing data set by using the processing data hash code set to obtain a deduplication processing data set;
generating a recursive segmentation tree according to the de-duplication processing data set, and calculating the abnormality degree of each node on the recursive segmentation tree;
and extracting abnormal nodes from the recursion partition tree according to the degree of abnormality, and screening the duplicate removal processing data corresponding to the abnormal nodes from the duplicate removal processing data to obtain a processing flow data set.
Optionally, the sequentially performing multidimensional feature extraction and feature standardization on the processing flow data set to obtain a processing flow feature set, including:
respectively extracting a water yield characteristic set, a water pressure characteristic set, a rotation characteristic set, a temperature characteristic set, a humidity characteristic set and a timely long characteristic set from the processing flow data set;
aggregating the water volume feature set, the water pressure feature set, the cycle feature set, the temperature feature set, the humidity feature set, and the duration feature set into a primary flow feature set group;
performing correlation screening on the primary flow characteristic set to obtain a secondary flow characteristic set;
and global pooling of the secondary flow characteristic set group into a standard flow characteristic set group, and performing dimension splicing operation on the standard flow characteristic set group to obtain a processing flow characteristic set.
Optionally, the performing correlation screening on the primary flow feature set to obtain a secondary flow feature set includes:
rank ordering is carried out on each primary flow characteristic set in the primary flow characteristic set group, so that a rank order flow characteristic set group is obtained;
calculating the antagonism correlation among each rank order flow characteristic set in the rank order flow characteristic set group by using the following antagonism correlation algorithm to obtain the antagonism correlation set:
；
Wherein,means that the rank order procedure feature set in the rank order procedure feature set group +.>And rank order flow feature setCorrelation of the antagonism between->Refer to the feature index, ++>Refers to the total number of features of the rank order procedure feature set,/for>、/>Is a dimension index>Is the characteristic total dimension of each rank order flow characteristic in the rank order flow characteristic set,/the characteristic total dimension is the characteristic total dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics>Characteristic value of dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics>Characteristic value of dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics>Characteristic value of dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics>Characteristic values of dimensions;
threshold screening is carried out on the antagonism correlation set to obtain a standard correlation set;
and screening a secondary flow characteristic set group from the primary flow characteristic set group according to the standard correlation set.
Optionally, the performing global fusion operation on the root feature set and the specification raw material feature to obtain raw material features includes:
performing position coding on the root feature set to obtain a standard root feature set;
Position coding is carried out on the specification raw material characteristics to obtain standard raw material characteristics;
global pooling of the standard root feature set into a pooled root feature set and global pooling of the standard raw material features into pooled raw material features, respectively;
and carrying out multistage full-connection operation on the pooled root feature group and the pooled raw material feature to obtain the raw material feature.
Optionally, the performing gas spectrum analysis and component matching on the finished product sample set sequentially to obtain a finished product feature set includes:
performing chromatographic separation and column temperature separation on the finished sample set in sequence to obtain a separation layer sample set;
performing mass spectrum detection on the separation layer sample set to obtain a sample mass spectrum atlas;
and carrying out feature extraction and mapping matching operation on the sample mass spectrum atlas in sequence to obtain a finished product feature set.
Optionally, the sequentially performing feature extraction and mapping matching operations on the sample mass spectrum atlas to obtain a finished product feature set, including:
selecting sample mass spectrum pictures in the sample mass spectrogram set one by one as target sample mass spectrum pictures, and respectively extracting sample peak characteristics, sample integral characteristics and sample frequency domain coefficient characteristics from the target sample mass spectrum pictures;
Collecting the sample peak characteristics, the sample integral characteristics and the sample frequency domain coefficient characteristics into sample mass spectrum characteristics;
sequentially performing full connection and normalization operation on the sample mass spectrum characteristics to obtain standard mass spectrum characteristics;
performing mass spectrum matching on the standard mass spectrum characteristics to obtain a sample composition group;
quantitatively mapping the standard mass spectrum characteristics according to the sample component group to obtain a component distribution group;
vectorizing the component distribution group into finished product features, and collecting all the finished product features into a finished product feature set.
Optionally, performing iterative performance training on a preset gynostemma pentaphylla processing model by using the raw material feature set, the processing flow feature set and the finished product feature set to obtain a gynostemma pentaphylla prediction model, including:
randomly generating an initial parameter group, and updating a preset gynostemma pentaphylla processing model into an initial processing model group by utilizing the initial parameter group;
calculating an initial finished product feature cluster corresponding to the raw material feature set and the processing flow feature set by using the initial processing model cluster;
calculating the initial product feature cluster and an error performance cluster corresponding to the product feature cluster;
Screening a parent parameter group from the initial parameter group according to the error performance group, and carrying out cross mutation on the parent parameter group to obtain a variant parameter group;
and iteratively updating the variation parameter group into standard parameters by using the initial parameter group, and updating the gynostemma pentaphylla processing model into a gynostemma pentaphylla prediction model by using the standard parameters.
Optionally, the calculating, by using the gynostemma pentaphylla prediction model, the standard flow characteristic corresponding to the real-time raw material picture includes:
sequentially carrying out root detail segmentation and global feature extraction operation on the real-time raw material picture to obtain real-time raw material features;
randomly generating a real-time flow characteristic group, and matching and combining the real-time raw material characteristic and the real-time flow characteristic group into a real-time gynostemma pentaphylla characteristic group;
calculating a predicted finished product feature set corresponding to the real-time gynostemma pentaphylla feature set by using the gynostemma pentaphylla prediction model;
performing iterative updating on the predicted finished product feature set by using a simulated annealing algorithm to obtain standard finished product features;
and screening out standard flow characteristics from the real-time flow characteristic group according to the standard finished product characteristics.
In order to solve the problems, the invention also provides a gynostemma pentaphylla processing system based on fresh root cleaning, which comprises:
The feature extraction module is used for sequentially carrying out data splitting and data cleaning operations on the pre-acquired historical gynostemma pentaphylla processing data to obtain a processing flow data set, and sequentially carrying out multidimensional feature extraction and feature standardization operations on the processing flow data set to obtain a processing flow feature set;
the detail segmentation module is used for acquiring a raw material atlas and a finished product sample set corresponding to the processing flow data set, sequentially carrying out root detail segmentation and global feature extraction operation on the raw material atlas to obtain a raw material feature set, wherein the sequentially carrying out root detail segmentation and global feature extraction operation on the raw material atlas to obtain the raw material feature set comprises the following steps: selecting raw material pictures in the raw material picture set one by one as target raw material pictures, and sequentially carrying out picture denoising and size stretching operation on the target raw material pictures to obtain specification raw material pictures; and performing primary feature extraction on the specification raw material picture by using the following linear filtering algorithm to obtain specification raw material features:
；
wherein,means that the coordinates in the raw material picture with the specification are +.>Is the specification raw material characteristic obtained by filtering the filtering convolution kernel center of the linear filtering algorithm, < - >The abscissa of the pixel being the center of the filter convolution kernel,>is the ordinate of the pixel in the center of the filter convolution kernel,>is a preset amplification factor, < >>Is an exponential function, ++>As cosine function +.>Is the filtering direction of the linear filtering algorithm, < >>Is a sine function +.>Means the filtering wavelength of said linear filtering algorithm, < > or->Is the standard deviation of the Gaussian distribution, +.>Is the circumference rate, < >>Means the filtering wavelength of said linear filtering algorithm, < > or->Means the filtering phase of the linear filtering algorithm, < >>Is a preset offset coefficient; performing picture blocking operation on the specification raw material pictures to obtain raw material block groups; extracting a root block group from the raw material block group, and extracting a root feature group from the root block group by using the linear filtering formula; performing global fusion operation on the root feature group and the specification raw material features to obtain raw material features, and collecting all the raw material features into a raw material feature set;
the component detection module is used for sequentially carrying out gas spectrum analysis and component matching operation on the finished product sample set to obtain a finished product feature set;
the model training module is used for carrying out iterative performance training on a preset gynostemma pentaphylla processing model by utilizing the raw material characteristic set, the processing flow characteristic set and the finished product characteristic set to obtain a gynostemma pentaphylla prediction model;
And the medicine processing module is used for acquiring a real-time raw material picture of a target gynostemma pentaphylla raw material to be processed, calculating standard flow characteristics corresponding to the real-time raw material picture by utilizing the gynostemma pentaphylla prediction model, and processing the target gynostemma pentaphylla raw material according to the standard flow characteristics.
According to the embodiment of the invention, the processing flow characteristic set is obtained by carrying out data splitting, data cleaning, multidimensional characteristic extraction and characteristic standardization operation on historical gynostemma pentaphylla processing data, the accuracy of the data set can be improved, the standardization of the data set is realized, meanwhile, relevant important characteristics are extracted, the accuracy of subsequent model training is improved, the image characteristic and root detail characteristic of gynostemma pentaphylla raw materials before each processing can be obtained by utilizing the image pickup device of processing equipment through carrying out root detail segmentation and global characteristic extraction operation on the raw material atlas, the accuracy of the prediction of a gynostemma pentaphylla finished product is further improved, the component distribution of a gynostemma pentaphylla finished product obtained after each gynostemma pentaphylla processing can be determined through carrying out gas spectrum analysis and component matching operation on the finished product sample set in sequence, the visual expression is carried out on the quality of the finished product after the processing, and the flexibility of the gynostemma pentaphylla processing is further improved.
By carrying out iterative performance training on a preset gynostemma pentaphylla processing model, a local optimal solution can be jumped out by utilizing a variation step in an iterative process, a search space is increased, the model training efficiency and the accuracy of the follow-up gynostemma pentaphylla prediction model in prediction are improved, the follow-up processing efficiency is further improved, and the standard flow characteristics corresponding to the real-time raw material picture are calculated by utilizing the gynostemma pentaphylla prediction model, so that the processing flow meeting the requirement of a preset finished product index can be determined according to the picture of the gynostemma pentaphylla raw material to be processed, and therefore, various parameters in the processing process are automatically configured, and the flexibility of gynostemma pentaphylla processing and the processing efficiency are improved. Therefore, the gynostemma pentaphylla processing method and system based on fresh root cleaning can solve the problem of low efficiency in processing gynostemma pentaphylla.
Drawings
FIG. 1 is a schematic flow chart of a method for processing gynostemma pentaphylla based on fresh root cleaning according to an embodiment of the present invention;
FIG. 2 is a flow chart of extracting a feature set of a processing flow according to an embodiment of the present invention;
FIG. 3 is a flow chart of extracting a final feature set according to an embodiment of the present invention;
FIG. 4 is a functional block diagram of a system for processing gynostemma pentaphylla based on fresh root cleaning according to an embodiment of the present invention;
the achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The embodiment of the application provides a gynostemma pentaphylla processing method based on fresh root cleaning. The main execution body of the gynostemma pentaphylla processing method based on fresh root cleaning comprises at least one of a server side, a terminal and the like which can be configured to execute the method provided by the embodiment of the application. In other words, the gynostemma pentaphylla processing method based on fresh root cleaning can be executed by software or hardware installed in terminal equipment or server equipment, and the software can be a blockchain platform. The service end includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like. The server may be an independent server, or may be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.
Referring to fig. 1, a flow chart of a processing method of gynostemma pentaphylla based on fresh root cleaning according to an embodiment of the present invention is shown. In this embodiment, the method for processing gynostemma pentaphylla based on fresh root cleaning includes:
s1, sequentially carrying out data splitting and data cleaning operations on historical gynostemma pentaphylla processing data acquired in advance to obtain a processing flow data set, and sequentially carrying out multidimensional feature extraction and feature standardization operations on the processing flow data set to obtain a processing flow feature set.
In the embodiment of the invention, the historical processing data of the gynostemma pentaphylla refer to data recorded when the gynostemma pentaphylla is processed by using processing equipment of the gynostemma pentaphylla in a past time period, the gynostemma pentaphylla is a cucurbitaceae and gynostemma pentaphylla grass climbing plant, and each processing flow data in the processing flow data set corresponds to flow data recorded when the gynostemma pentaphylla is processed once, such as water consumption, water pressure, washing cycle times, slicing processing temperature, humidity, processing time and the like.
In the embodiment of the present invention, the sequentially performing data splitting and data cleaning operations on the pre-acquired historical gynostemma pentaphylla processing data to obtain a processing flow data set includes:
Extracting a processing time stamp set from the historical gynostemma pentaphylla processing data, and splitting the historical gynostemma pentaphylla processing data into a primary processing data set according to the processing time stamp set;
performing unit formatting operation on the primary processing data set to obtain a secondary processing data set;
performing minimum hash coding on the secondary processing data set to obtain a processing data hash coding set;
performing data deduplication on the secondary processing data set by using the processing data hash code set to obtain a deduplication processing data set;
generating a recursive segmentation tree according to the de-duplication processing data set, and calculating the abnormality degree of each node on the recursive segmentation tree;
and extracting abnormal nodes from the recursion partition tree according to the degree of abnormality, and screening the duplicate removal processing data corresponding to the abnormal nodes from the duplicate removal processing data to obtain a processing flow data set.
Specifically, the processing timestamp set is a timestamp of processing record for each gynostemma pentaphylla, and includes a flow start timestamp and a flow end timestamp, and the unit formatting operation is performed on the primary processing data set to obtain a secondary processing data set, which means that different types of data in each primary processing data in the primary processing data set are arranged according to a fixed sequence, and the data of the same attribute type are unified into the same data unit.
In detail, a consistency group algorithm may be used to generate a recursive segmentation tree according to the deduplication processing dataset, a standard deviation algorithm or an average algorithm may be used to calculate a path length of each node on the recursive segmentation tree, and the path length is used as the anomaly degree of the corresponding node.
In detail, referring to fig. 2, the steps of sequentially performing multidimensional feature extraction and feature normalization on the process flow data set to obtain a process flow feature set include:
s21, respectively extracting a water yield characteristic set, a water pressure characteristic set, a rotation characteristic set, a temperature characteristic set, a humidity characteristic set and a timely long characteristic set from the processing flow data set;
s22, collecting the water quantity feature set, the water pressure feature set, the rotation feature set, the temperature feature set, the humidity feature set and the duration feature set into a primary flow feature set group;
s23, carrying out correlation screening on the primary flow characteristic set to obtain a secondary flow characteristic set;
s24, the secondary flow characteristic set group is globally pooled into a standard flow characteristic set group, and dimension splicing operation is carried out on the standard flow characteristic set group to obtain a processing flow characteristic set.
Specifically, each water volume feature in the water volume feature set, each water pressure feature in the water pressure feature set, each cycle feature in the cycle feature set, each temperature feature in the temperature feature set, each humidity feature in the humidity feature set corresponds to one of the process flow data sets, each time duration feature in the time duration feature set corresponds to one of the process flow data sets, wherein, the water quantity characteristic is the water quantity characteristic in the processing cleaning process, the water pressure characteristic is the water pressure intensity characteristic in the processing cleaning process, the rotation characteristic is the rotation speed, the frequency and the frequency characteristic of each cleaning in the processing cleaning process, the temperature characteristic is the temperature characteristic in the processing process, the humidity characteristic is the humidity characteristic in the processing process, and the time length characteristic is the time length characteristic of the corresponding step of each processing cleaning.
In detail, each primary process feature set in the primary process feature set group corresponds to the water volume feature set, the water pressure feature set, the rotation feature set, the temperature feature set, the humidity feature set and the duration feature set, the dimension splicing operation is performed on the standard process feature set group, the processing process feature set is obtained by splitting the standard process feature set group into a plurality of standard process feature sets according to a corresponding relation, each standard process feature in each standard process feature set is spliced into one processing process feature as a feature vector of one dimension, and all the processing process features are integrated into the processing process feature set.
In detail, the performing correlation screening on the primary flow feature set to obtain a secondary flow feature set includes:
rank ordering is carried out on each primary flow characteristic set in the primary flow characteristic set group, so that a rank order flow characteristic set group is obtained;
calculating the antagonism correlation among each rank order flow characteristic set in the rank order flow characteristic set group by using the following antagonism correlation algorithm to obtain the antagonism correlation set:
；
wherein,means that the rank order procedure feature set in the rank order procedure feature set group +.>And rank order flow feature setCorrelation of the antagonism between->Refer to the feature index, ++>Refers to the total number of features of the rank order procedure feature set,/for>、/>Is a dimension index>Is the characteristic total dimension of each rank order flow characteristic in the rank order flow characteristic set,/the characteristic total dimension is the characteristic total dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics>Characteristic value of dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics>Characteristic value of dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics>Characteristic value of dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics >Characteristic values of dimensions;
threshold screening is carried out on the antagonism correlation set to obtain a standard correlation set;
and screening a secondary flow characteristic set group from the primary flow characteristic set group according to the standard correlation set.
Specifically, the rank ordering of each primary flow feature set in the primary flow feature set group means that feature ranks of each primary flow feature in each primary flow feature set are calculated, the feature ranks of each primary flow feature are ordered according to the order from big to small or from small to big, a rank order flow feature set is obtained, and all the rank order flow feature sets are collected into a rank order flow feature set group.
In detail, by calculating the countermeasure correlation using the countermeasure correlation algorithm, the correlation of the feature sets can be determined according to the correlations of the feature flows and feature distributions among the feature sets of the respective primary flows, and the accuracy of the correlation calculation can be further improved.
According to the embodiment of the invention, the processing flow characteristic set is obtained by carrying out data splitting, data cleaning, multidimensional characteristic extraction and characteristic standardization operation on the historical gynostemma pentaphylla processing data, so that the accuracy of the data set can be improved, the standardization of the data set is realized, and meanwhile, relevant important characteristics are extracted, so that the accuracy of subsequent model training is improved.
S2, acquiring a raw material drawing set and a finished product sample set corresponding to the processing flow data set, and sequentially carrying out root detail segmentation and global feature extraction operation on the raw material drawing set to obtain a raw material feature set.
In the embodiment of the invention, each raw material picture in the raw material picture set and each finished product sample in the finished product sample set correspond to one piece of processing flow data in the processing flow data set, wherein the raw material picture refers to a picture obtained by photographing raw gynostemma pentaphylla before processing, and the finished product sample refers to a sample obtained by collecting finished gynostemma pentaphylla after processing.
In the embodiment of the present invention, root detail segmentation and global feature extraction operations are sequentially performed on the raw material atlas to obtain a raw material feature set, including:
selecting raw material pictures in the raw material picture set one by one as target raw material pictures, and sequentially carrying out picture denoising and size stretching operation on the target raw material pictures to obtain specification raw material pictures;
and performing primary feature extraction on the specification raw material picture by using the following linear filtering algorithm to obtain specification raw material features:
；
Wherein,means that the coordinates in the raw material picture with the specification are +.>Is the specification raw material characteristic obtained by filtering the filtering convolution kernel center of the linear filtering algorithm, < ->The abscissa of the pixel being the center of the filter convolution kernel,>is the ordinate of the pixel in the center of the filter convolution kernel,>is a preset amplification factor, < >>Is an exponential function, ++>As cosine function +.>Is the filtering direction of the linear filtering algorithm, < >>Is a sine function +.>Means the filtering wavelength of said linear filtering algorithm, < > or->Is the standard deviation of the Gaussian distribution, +.>Is the circumference rate, < >>Means the filtering wavelength of said linear filtering algorithm, < > or->Means the filtering phase of the linear filtering algorithm, < >>Is a preset offset coefficient;
performing picture blocking operation on the specification raw material pictures to obtain raw material block groups;
extracting a root block group from the raw material block group, and extracting a root feature group from the root block group by using the linear filtering formula;
and performing global fusion operation on the root feature group and the specification raw material features to obtain raw material features, and collecting all the raw material features into a raw material feature set.
In detail, the target raw material pictures can be subjected to picture denoising by using a median filtering algorithm or a gaussian filtering algorithm, and the size stretching refers to stretching the picture sizes of all the target raw material pictures to the same size.
Specifically, the linear filtering algorithm is utilized to perform primary feature extraction on the standard raw material picture, so that multi-amplitude and multi-directional feature extraction can be realized, the feature extraction flexibility is improved, the texture, detail and edge features of the picture are further highlighted, and the picture blocking refers to uniformly splitting the standard raw material picture into raw material picture block groups formed by a plurality of raw material picture blocks according to a fixed picture size.
In detail, the performing global fusion operation on the root feature group and the specification raw material feature to obtain raw material features includes:
performing position coding on the root feature set to obtain a standard root feature set;
position coding is carried out on the specification raw material characteristics to obtain standard raw material characteristics;
global pooling of the standard root feature set into a pooled root feature set and global pooling of the standard raw material features into pooled raw material features, respectively;
and carrying out multistage full-connection operation on the pooled root feature group and the pooled raw material feature to obtain the raw material feature.
In detail, the position coding means that corresponding position vectors are configured according to the position relation of each root feature in the root feature group in the corresponding specification raw material picture, and the position vectors are used as one dimension feature of the corresponding root feature to be fused into standard root features.
Specifically, the global pooling refers to unifying feature dimensions and feature sizes of each standard root feature in the standard root feature group and the standard raw material feature, so that subsequent feature fusion operation is facilitated, and multi-level full-connection operation can be performed on the pooled root feature group and the pooled raw material feature by using a multi-layer perceptron (Multilayer Perceptron, abbreviated as MLP) to obtain the raw material feature.
In the embodiment of the invention, the image device of the processing equipment can be utilized to acquire the picture characteristics and the root detail characteristics of the gynostemma pentaphylla raw material before each processing by carrying out root detail segmentation and global characteristic extraction operation on the raw material atlas, so that the accuracy of the prediction of the gynostemma pentaphylla finished product of the subsequent model is improved.
S3, carrying out gas spectrum analysis and component matching operation on the finished product sample set in sequence to obtain a finished product feature set.
In the embodiment of the invention, each finished product feature in the finished product feature set corresponds to a physicochemical feature of one finished product sample in the finished product sample set, and the finished product feature can be used for measuring the quality of the finished product of the gynostemma pentaphylla after processing.
In the embodiment of the present invention, the steps of sequentially performing gas spectrum analysis and component matching on the final product sample set to obtain a final product feature set include:
Performing chromatographic separation and column temperature separation on the finished sample set in sequence to obtain a separation layer sample set;
performing mass spectrum detection on the separation layer sample set to obtain a sample mass spectrum atlas;
and carrying out feature extraction and mapping matching operation on the sample mass spectrum atlas in sequence to obtain a finished product feature set.
Specifically, the chromatographic separation refers to performing affinity separation on each finished product sample in the finished product sample set by using a preset chromatographic column, and performing column chromatography on the compound after affinity separation by using carrier gas of the chromatographic column, and the column temperature separation refers to controlling the chromatographic column to perform chromatographic separation within a preset column temperature.
In detail, electron bombardment can be performed on the separation layer sample set by using a mass spectrometer to obtain an ion sample set, and mass spectrometry is performed on the ion sample set to obtain a sample mass spectrum chart set, wherein mass spectrometry can be performed on the ion sample set by using a method of tandem mass spectrometry, electrospray mass spectrometry or ion trap mass spectrometry.
Specifically, referring to fig. 3, the sequentially performing feature extraction and mapping matching operations on the sample mass spectrum atlas to obtain a finished product feature set includes:
s31, selecting sample mass spectrum pictures in the sample mass spectrogram set one by one as target sample mass spectrum pictures, and respectively extracting sample peak characteristics, sample integral characteristics and sample frequency domain coefficient characteristics from the target sample mass spectrum pictures;
S32, collecting the sample peak characteristics, the sample integral characteristics and the sample frequency domain coefficient characteristics into sample mass spectrum characteristics;
s33, sequentially performing full connection and normalization operation on the sample mass spectrum characteristics to obtain standard mass spectrum characteristics;
s34, performing mass spectrum matching on the standard mass spectrum characteristics to obtain a sample composition group;
s35, quantitatively mapping the standard mass spectrum characteristics according to the sample component group to obtain a component distribution group;
s36, vectorizing the component distribution group into finished product features, and collecting all the finished product features into a finished product feature set.
Specifically, the sample peak characteristic refers to a peak characteristic of the target sample mass spectrum picture, the sample integral characteristic refers to an area characteristic of each peak of the target sample mass spectrum picture, and the sample frequency domain coefficient characteristic refers to a wavelet coefficient characteristic of the target sample mass spectrum picture after wavelet transformation.
In detail, the standard mass spectrum characteristics may be subjected to mass spectrum matching by using a known mass spectrum library such as NIST mass spectrum library, so as to obtain a sample component group, wherein the sample component group is a combination of each compound component corresponding to the standard mass spectrum characteristics, and the standard mass spectrum characteristics may be quantitatively mapped according to the sample component group by using a standard curve method or an internal standard method, so as to obtain a component distribution group.
According to the embodiment of the invention, the gas spectrum analysis and the component matching operation are sequentially carried out on the finished product sample set to obtain the finished product feature set, so that the component distribution of the finished product of gynostemma pentaphylla obtained after each gynostemma pentaphylla processing can be determined, the quality of the finished product after processing is intuitively expressed, the subsequent targeted raw material processing is facilitated, and the flexibility of gynostemma pentaphylla processing is improved.
And S4, performing iterative performance training on a preset gynostemma pentaphylla processing model by using the raw material characteristic set, the processing flow characteristic set and the finished product characteristic set to obtain a gynostemma pentaphylla prediction model.
In the embodiment of the invention, the gynostemma pentaphylla processing model can be a machine learning regression model such as a polynomial regression model (Polynomial Regression, PR for short), a random forest model (Random Forest Regression, RFR for short) or a support vector regression model (Support Vector Regression, SVR for short).
In the embodiment of the present invention, performing iterative performance training on a preset gynostemma pentaphylla processing model by using the raw material feature set, the processing flow feature set and the finished product feature set to obtain a gynostemma pentaphylla prediction model includes:
randomly generating an initial parameter group, and updating a preset gynostemma pentaphylla processing model into an initial processing model group by utilizing the initial parameter group;
Calculating an initial finished product feature cluster corresponding to the raw material feature set and the processing flow feature set by using the initial processing model cluster;
calculating the initial product feature cluster and an error performance cluster corresponding to the product feature cluster;
screening a parent parameter group from the initial parameter group according to the error performance group, and carrying out cross mutation on the parent parameter group to obtain a variant parameter group;
and iteratively updating the variation parameter group into standard parameters by using the initial parameter group, and updating the gynostemma pentaphylla processing model into a gynostemma pentaphylla prediction model by using the standard parameters.
In detail, each initial parameter in the initial parameter group corresponds to a group of model parameters in the gynostemma pentaphylla processing model, and updating the preset gynostemma pentaphylla processing model into the initial processing model group by using the initial parameter group refers to replacing the model parameters in the gynostemma pentaphylla processing model by using each initial parameter in the initial parameter group to obtain an initial processing model group formed by a plurality of initial processing models.
Specifically, a mean square error method (Mean Squared Error, abbreviated as MSE) may be used to calculate error performance of each initial product feature set in the initial product feature set and corresponding to the final product feature set, and all error performance may be collected into an error performance group.
In detail, the step of selecting a parent parameter group from the initial parameter groups according to the error performance groups refers to selecting initial parameters corresponding to a plurality of error performances with the smallest value in the error performance groups to form a parent parameter group, the step of iteratively updating the variation parameter groups into standard parameters by using the initial parameter groups refers to replacing the initial parameter groups with the variation parameter groups, and the step of updating the preset gynostemma pentaphylla processing model into an initial processing model group by using the initial parameter groups is returned to iterate until the preset iteration times are reached, wherein the variation parameter of the smallest error performance in the error performance groups at the moment is used as the standard parameter.
According to the embodiment of the invention, the iteration performance training is carried out on the preset gynostemma pentaphylla processing model, so that the local optimal solution can be jumped out by utilizing the variation step in the iteration process, the search space is increased, the model training efficiency and the accuracy in the prediction of the follow-up gynostemma pentaphylla prediction model are improved, and the follow-up processing efficiency is further improved.
S5, acquiring a real-time raw material picture of a target gynostemma pentaphylla raw material to be processed, calculating standard flow characteristics corresponding to the real-time raw material picture by using the gynostemma pentaphylla prediction model, and processing the target gynostemma pentaphylla raw material according to the standard flow characteristics.
In the embodiment of the invention, the real-time raw material picture is a picture obtained by shooting the target gynostemma pentaphylla raw material to be processed, and the processing of the target gynostemma pentaphylla raw material according to the standard flow characteristics is to extract corresponding flow parameters in the process of configuring and processing the water quantity characteristics, the water pressure characteristics, the rotation characteristics, the temperature characteristics and the humidity characteristics according to the standard flow characteristics, so as to process the target gynostemma pentaphylla raw material.
In the embodiment of the present invention, the calculating, by using the gynostemma pentaphylla prediction model, the standard flow characteristics corresponding to the real-time raw material picture includes:
sequentially carrying out root detail segmentation and global feature extraction operation on the real-time raw material picture to obtain real-time raw material features;
randomly generating a real-time flow characteristic group, and matching and combining the real-time raw material characteristic and the real-time flow characteristic group into a real-time gynostemma pentaphylla characteristic group;
calculating a predicted finished product feature set corresponding to the real-time gynostemma pentaphylla feature set by using the gynostemma pentaphylla prediction model;
performing iterative updating on the predicted finished product feature set by using a simulated annealing algorithm to obtain standard finished product features;
and screening out standard flow characteristics from the real-time flow characteristic group according to the standard finished product characteristics.
In detail, each real-time process feature in the real-time process feature set corresponds to a randomly generated process feature, and the step of using the simulated annealing algorithm to update the predicted finished product feature set in an iterative manner refers to determining the iteration times by using the simulated annealing algorithm, returning to the step of randomly generating the real-time process feature set for iteration, and screening predicted finished product features meeting preset component indexes from all the predicted finished product feature sets generated in the iteration times to serve as standard finished product features.
According to the embodiment of the invention, the standard flow characteristics corresponding to the real-time raw material picture are calculated by utilizing the gynostemma pentaphylla prediction model, so that the processing flow meeting the requirement of the preset finished product index can be determined according to the picture of the gynostemma pentaphylla raw material to be processed, and therefore, various parameters in the processing process are automatically configured, and the flexibility and the processing efficiency of gynostemma pentaphylla processing are improved.
According to the embodiment of the invention, the processing flow characteristic set is obtained by carrying out data splitting, data cleaning, multidimensional characteristic extraction and characteristic standardization operation on historical gynostemma pentaphylla processing data, the accuracy of the data set can be improved, the standardization of the data set is realized, meanwhile, relevant important characteristics are extracted, the accuracy of subsequent model training is improved, the image characteristic and root detail characteristic of gynostemma pentaphylla raw materials before each processing can be obtained by utilizing the image pickup device of processing equipment through carrying out root detail segmentation and global characteristic extraction operation on the raw material atlas, the accuracy of the prediction of a gynostemma pentaphylla finished product is further improved, the component distribution of a gynostemma pentaphylla finished product obtained after each gynostemma pentaphylla processing can be determined through carrying out gas spectrum analysis and component matching operation on the finished product sample set in sequence, the visual expression is carried out on the quality of the finished product after the processing, and the flexibility of the gynostemma pentaphylla processing is further improved.
By carrying out iterative performance training on a preset gynostemma pentaphylla processing model, a local optimal solution can be jumped out by utilizing a variation step in an iterative process, a search space is increased, the model training efficiency and the accuracy of the follow-up gynostemma pentaphylla prediction model in prediction are improved, the follow-up processing efficiency is further improved, and the standard flow characteristics corresponding to the real-time raw material picture are calculated by utilizing the gynostemma pentaphylla prediction model, so that the processing flow meeting the requirement of a preset finished product index can be determined according to the picture of the gynostemma pentaphylla raw material to be processed, and therefore, various parameters in the processing process are automatically configured, and the flexibility of gynostemma pentaphylla processing and the processing efficiency are improved. Therefore, the gynostemma pentaphylla processing method based on fresh root cleaning can solve the problem of low efficiency in processing gynostemma pentaphylla.
Fig. 4 is a functional block diagram of a gynostemma pentaphylla processing system based on fresh root cleaning according to an embodiment of the present invention.
The gynostemma pentaphylla processing system 100 based on fresh root cleaning of the present invention may be installed in an electronic device. Depending on the functions implemented, the gynostemma pentaphylla processing system 100 based on fresh root cleaning may include a feature extraction module 101, a detail segmentation module 102, a component detection module 103, a model training module 104 and a drug processing module 105. The module of the invention, which may also be referred to as a unit, refers to a series of computer program segments, which are stored in the memory of the electronic device, capable of being executed by the processor of the electronic device and of performing a fixed function.
In the present embodiment, the functions concerning the respective modules/units are as follows:
the feature extraction module 101 is configured to sequentially perform data splitting and data cleaning operations on the pre-acquired historical gynostemma pentaphylla processing data to obtain a processing flow data set, and sequentially perform multidimensional feature extraction and feature standardization operations on the processing flow data set to obtain a processing flow feature set;
the detail segmentation module 102 is configured to obtain a raw material atlas and a finished product sample set corresponding to the processing flow data set, sequentially perform root detail segmentation and global feature extraction operations on the raw material atlas to obtain a raw material feature set, wherein sequentially perform root detail segmentation and global feature extraction operations on the raw material atlas to obtain the raw material feature set, and includes: selecting raw material pictures in the raw material picture set one by one as target raw material pictures, and sequentially carrying out picture denoising and size stretching operation on the target raw material pictures to obtain specification raw material pictures; and performing primary feature extraction on the specification raw material picture by using the following linear filtering algorithm to obtain specification raw material features:
；
wherein,means that the coordinates in the raw material picture with the specification are +. >Is the specification raw material characteristic obtained by filtering the filtering convolution kernel center of the linear filtering algorithm, < ->The abscissa of the pixel being the center of the filter convolution kernel,>is the ordinate of the pixel in the center of the filter convolution kernel,>is a preset amplification factor, < >>Is an exponential function, ++>As cosine function +.>Is the filtering direction of the linear filtering algorithm, < >>Is a sine function +.>Means the filtering wavelength of said linear filtering algorithm, < > or->Is the standard deviation of the Gaussian distribution, +.>Is the circumference rate, < >>Means the filtering wavelength of said linear filtering algorithm, < > or->Means the filtering phase of the linear filtering algorithm, < >>Is a preset offset coefficient; performing picture blocking operation on the specification raw material pictures to obtain raw material block groups; extracting a root block group from the raw material block group, and extracting a root feature group from the root block group by using the linear filtering formula; performing global fusion operation on the root feature group and the specification raw material features to obtain raw material features, and collecting all the raw material features into a raw material feature set;
the component detection module 103 is configured to perform gas spectrum analysis and component matching operations on the finished product sample set in sequence, so as to obtain a finished product feature set;
The model training module 104 is configured to perform iterative performance training on a preset gynostemma pentaphylla processing model by using the raw material feature set, the processing flow feature set and the finished product feature set to obtain a gynostemma pentaphylla prediction model;
the medicine processing module 105 is configured to obtain a real-time raw material picture of a target gynostemma pentaphylla raw material to be processed, calculate a standard flow characteristic corresponding to the real-time raw material picture by using the gynostemma pentaphylla prediction model, and process the target gynostemma pentaphylla raw material according to the standard flow characteristic.
In detail, each module in the gynostemma pentaphylla processing system 100 based on fresh root cleaning in the embodiment of the present invention adopts the same technical means as the gynostemma pentaphylla processing method based on fresh root cleaning described in the above-mentioned fig. 1 to 3 when in use, and can produce the same technical effects, and the details are not repeated here.
In the several embodiments provided by the present invention, it should be understood that the disclosed apparatus, system and method may be implemented in other manners. For example, the system embodiments described above are merely illustrative, e.g., the division of the modules is merely a logical function division, and other manners of division may be implemented in practice.
The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, may be located in one place, or may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional module in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units can be realized in a form of hardware or a form of hardware and a form of software functional modules.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof.
The embodiment of the application can acquire and process the related data based on the artificial intelligence technology. Among these, artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use knowledge to obtain optimal results.
Furthermore, it is evident that the word "comprising" does not exclude other elements or steps, and that the singular does not exclude a plurality. Multiple units or systems set forth in the system embodiments may also be implemented by one unit or system in software or hardware. The terms first, second, etc. are used to denote a name, but not any particular order.
Finally, it should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention.
Claims (10)
1. A processing method of gynostemma pentaphylla based on fresh root cleaning is characterized by comprising the following steps:
sequentially carrying out data splitting and data cleaning operations on the pre-acquired historical gynostemma pentaphylla processing data to obtain a processing flow data set, and sequentially carrying out multidimensional feature extraction and feature standardization operations on the processing flow data set to obtain a processing flow feature set;
acquiring a raw material atlas and a finished product sample set corresponding to the processing flow data set, sequentially performing root detail segmentation and global feature extraction operations on the raw material atlas to obtain a raw material feature set, wherein the sequentially performing root detail segmentation and global feature extraction operations on the raw material atlas to obtain the raw material feature set comprises the following steps: selecting raw material pictures in the raw material picture set one by one as target raw material pictures, and sequentially carrying out picture denoising and size stretching operation on the target raw material pictures to obtain specification raw material pictures; and performing primary feature extraction on the specification raw material picture by using the following linear filtering algorithm to obtain specification raw material features:
；
Wherein,means that the coordinates in the raw material picture with the specification are +.>Is the specification raw material characteristic obtained by filtering the filtering convolution kernel center of the linear filtering algorithm, < ->The abscissa of the pixel being the center of the filter convolution kernel,>is the ordinate of the pixel in the center of the filter convolution kernel,>is a preset amplification factor, < >>Is an exponential function of the number of times,as cosine function +.>Is the filtering direction of the linear filtering algorithm, < >>Is a sine function +.>Means the filtering wavelength of said linear filtering algorithm, < > or->Is the standard deviation of the Gaussian distribution, +.>Is the circumference rate, < >>Means the filtering wavelength of said linear filtering algorithm, < > or->Means the filtering phase of the linear filtering algorithm, < >>Is a preset offset coefficient; performing picture blocking operation on the specification raw material pictures to obtain raw material block groups; extracting a root block group from the raw material block group, and extracting a root feature group from the root block group by using the linear filtering algorithm; performing global fusion operation on the root feature group and the specification raw material features to obtain raw material features, and collecting all the raw material features into a raw material feature set;
carrying out gas spectrum analysis and component matching operation on the finished product sample set in sequence to obtain a finished product feature set;
Performing iterative performance training on a preset gynostemma pentaphylla processing model by using the raw material characteristic set, the processing flow characteristic set and the finished product characteristic set to obtain a gynostemma pentaphylla prediction model;
and obtaining a real-time raw material picture of a target gynostemma pentaphylla raw material to be processed, calculating a standard flow characteristic corresponding to the real-time raw material picture by using the gynostemma pentaphylla prediction model, and processing the target gynostemma pentaphylla raw material according to the standard flow characteristic.
2. The method for processing gynostemma pentaphylla based on fresh root cleaning according to claim 1, wherein the sequentially performing data splitting and data cleaning operations on the pre-acquired historical gynostemma pentaphylla processing data to obtain a processing flow data set comprises:
extracting a processing time stamp set from the historical gynostemma pentaphylla processing data, and splitting the historical gynostemma pentaphylla processing data into a primary processing data set according to the processing time stamp set;
performing unit formatting operation on the primary processing data set to obtain a secondary processing data set;
performing minimum hash coding on the secondary processing data set to obtain a processing data hash coding set;
performing data deduplication on the secondary processing data set by using the processing data hash code set to obtain a deduplication processing data set;
Generating a recursive segmentation tree according to the de-duplication processing data set, and calculating the abnormality degree of each node on the recursive segmentation tree;
and extracting abnormal nodes from the recursion partition tree according to the degree of abnormality, and screening the duplicate removal processing data corresponding to the abnormal nodes from the duplicate removal processing data to obtain a processing flow data set.
3. The method for processing gynostemma pentaphylla based on fresh root cleaning according to claim 1, wherein the steps of sequentially performing multidimensional feature extraction and feature standardization on the processing flow data set to obtain the processing flow feature set comprise:
respectively extracting a water yield characteristic set, a water pressure characteristic set, a rotation characteristic set, a temperature characteristic set, a humidity characteristic set and a timely long characteristic set from the processing flow data set;
aggregating the water volume feature set, the water pressure feature set, the cycle feature set, the temperature feature set, the humidity feature set, and the duration feature set into a primary flow feature set group;
performing correlation screening on the primary flow characteristic set to obtain a secondary flow characteristic set;
and global pooling of the secondary flow characteristic set group into a standard flow characteristic set group, and performing dimension splicing operation on the standard flow characteristic set group to obtain a processing flow characteristic set.
4. The method for processing gynostemma pentaphylla based on fresh root cleaning of claim 3, wherein the performing correlation screening on the primary flow characteristic set to obtain a secondary flow characteristic set comprises the following steps:
rank ordering is carried out on each primary flow characteristic set in the primary flow characteristic set group, so that a rank order flow characteristic set group is obtained;
calculating the antagonism correlation among each rank order flow characteristic set in the rank order flow characteristic set group by using the following antagonism correlation algorithm to obtain the antagonism correlation set:
；
wherein,means that the rank order procedure feature set in the rank order procedure feature set group +.>And rank order procedure feature set +.>Correlation of the antagonism between->Refer to the feature index, ++>Refers to the total number of features of the rank order procedure feature set,/for>、/>Is the index of the dimensions of the device,is the characteristic total dimension of each rank order flow characteristic in the rank order flow characteristic set,/the characteristic total dimension is the characteristic total dimension>Refers to rank order streamProgram feature set->The%>No. of the individual rank order flow characteristics>Characteristic value of dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics>Characteristic value of dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics >Characteristic value of dimension>Refers to the rank order procedure feature set +.>The%>No. of the individual rank order flow characteristics>Characteristic values of dimensions;
threshold screening is carried out on the antagonism correlation set to obtain a standard correlation set;
and screening a secondary flow characteristic set group from the primary flow characteristic set group according to the standard correlation set.
5. The method for processing gynostemma pentaphylla based on fresh root cleaning of claim 1, wherein the performing global fusion operation on the root feature set and the specification raw material feature to obtain raw material features comprises:
performing position coding on the root feature set to obtain a standard root feature set;
position coding is carried out on the specification raw material characteristics to obtain standard raw material characteristics;
global pooling of the standard root feature set into a pooled root feature set and global pooling of the standard raw material features into pooled raw material features, respectively;
and carrying out multistage full-connection operation on the pooled root feature group and the pooled raw material feature to obtain the raw material feature.
6. The method for processing gynostemma pentaphylla based on fresh root cleaning according to claim 1, wherein the steps of sequentially performing gas spectrum analysis and component matching operation on the finished product sample set to obtain a finished product feature set comprise the steps of:
Performing chromatographic separation and column temperature separation on the finished sample set in sequence to obtain a separation layer sample set;
performing mass spectrum detection on the separation layer sample set to obtain a sample mass spectrum atlas;
and carrying out feature extraction and mapping matching operation on the sample mass spectrum atlas in sequence to obtain a finished product feature set.
7. The method for processing gynostemma pentaphylla based on fresh root cleaning according to claim 6, wherein the sequentially performing feature extraction and mapping matching operations on the sample mass spectrum atlas to obtain a finished product feature set comprises:
selecting sample mass spectrum pictures in the sample mass spectrogram set one by one as target sample mass spectrum pictures, and respectively extracting sample peak characteristics, sample integral characteristics and sample frequency domain coefficient characteristics from the target sample mass spectrum pictures;
collecting the sample peak characteristics, the sample integral characteristics and the sample frequency domain coefficient characteristics into sample mass spectrum characteristics;
sequentially performing full connection and normalization operation on the sample mass spectrum characteristics to obtain standard mass spectrum characteristics;
performing mass spectrum matching on the standard mass spectrum characteristics to obtain a sample composition group;
quantitatively mapping the standard mass spectrum characteristics according to the sample component group to obtain a component distribution group;
Vectorizing the component distribution group into finished product features, and collecting all the finished product features into a finished product feature set.
8. The method for processing gynostemma pentaphylla based on fresh root cleaning according to claim 1, wherein the iterative performance training is performed on a preset gynostemma pentaphylla processing model by using the raw material characteristic set, the processing flow characteristic set and the finished product characteristic set to obtain a gynostemma pentaphylla prediction model, and the method comprises the following steps:
randomly generating an initial parameter group, and updating a preset gynostemma pentaphylla processing model into an initial processing model group by utilizing the initial parameter group;
calculating an initial finished product feature cluster corresponding to the raw material feature set and the processing flow feature set by using the initial processing model cluster;
calculating the initial product feature cluster and an error performance cluster corresponding to the product feature cluster;
screening a parent parameter group from the initial parameter group according to the error performance group, and carrying out cross mutation on the parent parameter group to obtain a variant parameter group;
and iteratively updating the variation parameter group into standard parameters by using the initial parameter group, and updating the gynostemma pentaphylla processing model into a gynostemma pentaphylla prediction model by using the standard parameters.
9. The method for processing gynostemma pentaphylla based on fresh root cleaning according to claim 1, wherein the calculating the standard flow characteristics corresponding to the real-time raw material picture by using the gynostemma pentaphylla prediction model comprises the following steps:
sequentially carrying out root detail segmentation and global feature extraction operation on the real-time raw material picture to obtain real-time raw material features;
randomly generating a real-time flow characteristic group, and matching and combining the real-time raw material characteristic and the real-time flow characteristic group into a real-time gynostemma pentaphylla characteristic group;
calculating a predicted finished product feature set corresponding to the real-time gynostemma pentaphylla feature set by using the gynostemma pentaphylla prediction model;
performing iterative updating on the predicted finished product feature set by using a simulated annealing algorithm to obtain standard finished product features;
and screening out standard flow characteristics from the real-time flow characteristic group according to the standard finished product characteristics.
10. A gynostemma pentaphylla processing system based on fresh root cleaning, the system comprising:
the feature extraction module is used for sequentially carrying out data splitting and data cleaning operations on the pre-acquired historical gynostemma pentaphylla processing data to obtain a processing flow data set, and sequentially carrying out multidimensional feature extraction and feature standardization operations on the processing flow data set to obtain a processing flow feature set;
The detail segmentation module is used for acquiring a raw material atlas and a finished product sample set corresponding to the processing flow data set, sequentially carrying out root detail segmentation and global feature extraction operation on the raw material atlas to obtain a raw material feature set, wherein the sequentially carrying out root detail segmentation and global feature extraction operation on the raw material atlas to obtain the raw material feature set comprises the following steps: selecting raw material pictures in the raw material picture set one by one as target raw material pictures, and sequentially carrying out picture denoising and size stretching operation on the target raw material pictures to obtain specification raw material pictures; and performing primary feature extraction on the specification raw material picture by using the following linear filtering algorithm to obtain specification raw material features:
；
wherein,means that the coordinates in the raw material picture with the specification are +.>Is the specification raw material characteristic obtained by filtering the filtering convolution kernel center of the linear filtering algorithm, < ->The abscissa of the pixel being the center of the filter convolution kernel,>is the ordinate of the pixel in the center of the filter convolution kernel,>is a preset amplification factor, < >>Is an exponential function of the number of times,as cosine function +.>Is said to be linearThe filtering direction of the filtering algorithm, < >>Is a sine function +. >Means the filtering wavelength of said linear filtering algorithm, < > or->Is the standard deviation of the Gaussian distribution, +.>Is the circumference rate, < >>Means the filtering wavelength of said linear filtering algorithm, < > or->Means the filtering phase of the linear filtering algorithm, < >>Is a preset offset coefficient; performing picture blocking operation on the specification raw material pictures to obtain raw material block groups; extracting a root block group from the raw material block group, and extracting a root feature group from the root block group by using the linear filtering algorithm; performing global fusion operation on the root feature group and the specification raw material features to obtain raw material features, and collecting all the raw material features into a raw material feature set;
the component detection module is used for sequentially carrying out gas spectrum analysis and component matching operation on the finished product sample set to obtain a finished product feature set;
the model training module is used for carrying out iterative performance training on a preset gynostemma pentaphylla processing model by utilizing the raw material characteristic set, the processing flow characteristic set and the finished product characteristic set to obtain a gynostemma pentaphylla prediction model;
and the medicine processing module is used for acquiring a real-time raw material picture of a target gynostemma pentaphylla raw material to be processed, calculating standard flow characteristics corresponding to the real-time raw material picture by utilizing the gynostemma pentaphylla prediction model, and processing the target gynostemma pentaphylla raw material according to the standard flow characteristics.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311403134.3A CN117131349B (en) | 2023-10-27 | 2023-10-27 | Gynostemma pentaphylla processing method and system based on fresh root cleaning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311403134.3A CN117131349B (en) | 2023-10-27 | 2023-10-27 | Gynostemma pentaphylla processing method and system based on fresh root cleaning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117131349A CN117131349A (en) | 2023-11-28 |
| CN117131349B true CN117131349B (en) | 2023-12-22 |
Family
ID=88860422
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311403134.3A Active CN117131349B (en) | 2023-10-27 | 2023-10-27 | Gynostemma pentaphylla processing method and system based on fresh root cleaning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117131349B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105248654A (en) * | 2015-10-18 | 2016-01-20 | 张建新 | Preparation method of gynostemma pentaphyllum containing nutritional bean milk |
| CN205016630U (en) * | 2015-10-12 | 2016-02-03 | 株洲盈定自动化设备科技有限公司 | In semi -automatic freezing into scouring machine of big close battery |
| CN106727867A (en) * | 2016-10-27 | 2017-05-31 | 广西金秀香料香精有限责任公司 | A kind of extracting method of gypenoside |
| CN109646477A (en) * | 2017-10-10 | 2019-04-19 | 张巧娜 | A kind of gynostemma pentaphylla processing method |
| CN115643956A (en) * | 2022-11-03 | 2023-01-31 | 咸阳职业技术学院 | High-efficiency comprehensive prevention and control method for gynostemma pentaphylla nematode |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180228823A1 (en) * | 2016-05-04 | 2018-08-16 | Intelligent Synthetic Biology Center | Anti-diabetic effects of gypenoside 75 |
-
2023
- 2023-10-27 CN CN202311403134.3A patent/CN117131349B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN205016630U (en) * | 2015-10-12 | 2016-02-03 | 株洲盈定自动化设备科技有限公司 | In semi -automatic freezing into scouring machine of big close battery |
| CN105248654A (en) * | 2015-10-18 | 2016-01-20 | 张建新 | Preparation method of gynostemma pentaphyllum containing nutritional bean milk |
| CN106727867A (en) * | 2016-10-27 | 2017-05-31 | 广西金秀香料香精有限责任公司 | A kind of extracting method of gypenoside |
| CN109646477A (en) * | 2017-10-10 | 2019-04-19 | 张巧娜 | A kind of gynostemma pentaphylla processing method |
| CN115643956A (en) * | 2022-11-03 | 2023-01-31 | 咸阳职业技术学院 | High-efficiency comprehensive prevention and control method for gynostemma pentaphylla nematode |
Non-Patent Citations (1)
| Title |
|---|
| 鲜样前处理方式对绞股蓝总皂甙提取率的影响;张涛;张顺萍;袁弟顺;;河南农业大学学报(第06期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117131349A (en) | 2023-11-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106777038B (en) | A kind of ultralow complexity image search method retaining Hash based on sequence | |
| CN111291139B (en) | Knowledge graph long-tail relation completion method based on attention mechanism | |
| Li et al. | Lightweight detection networks for tea bud on complex agricultural environment via improved YOLO v4 | |
| US8266149B2 (en) | Clustering with similarity-adjusted entropy | |
| CN108288270B (en) | Target detection method based on channel pruning and full convolution deep learning | |
| CN116012720B (en) | Crop growth condition monitoring method, device and medium based on high-resolution remote sensing image | |
| Yang et al. | A scalable data chunk similarity based compression approach for efficient big sensing data processing on cloud | |
| CN108519347A (en) | A kind of infrared spectrum Wavelength selecting method based on binary system dragonfly algorithm | |
| CN106294219A (en) | A kind of equipment identification, data processing method, Apparatus and system | |
| CN114428937A (en) | Time sequence data prediction method based on space-time diagram neural network | |
| CN117151353B (en) | Intelligent litchi pest identification and ecological regulation method, system and medium | |
| CN117131349B (en) | Gynostemma pentaphylla processing method and system based on fresh root cleaning | |
| CN117016517A (en) | Pesticide spraying and drift preventing method | |
| CN115222044A (en) | Model training method, graph data processing method, device, equipment and storage medium | |
| Li et al. | Filter pruning via probabilistic model-based optimization for accelerating deep convolutional neural networks | |
| Wei et al. | YOLO_MRC: A fast and lightweight model for real-time detection and individual counting of Tephritidae pests | |
| CN114577967B (en) | Chromatographic analysis method for traditional Chinese medicine compound sample based on artificial neural network and difference spectrum | |
| CN113705589A (en) | Data processing method, device and equipment | |
| Rong et al. | Soft Taylor pruning for accelerating deep convolutional neural networks | |
| Kumar et al. | [Retracted] Secure Data Deduplication System with Cyber Security Multikey Management in Cloud Storage | |
| San Woo et al. | End-to-end lightweight berry number prediction for supporting table grape cultivation | |
| CN109583490B (en) | Large-scale image matching method | |
| CN114530197A (en) | Medicine target point prediction method and system based on matrix completion | |
| YANG et al. | Plant disease and pest detection based on visiual attention enhancement | |
| Gunes et al. | Classification of hazelnuts with CNN based deep learning system |
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 |