Fresh meat whole industry chain information fusion decision-making system based on digital twinning
By constructing a digital twin-based information fusion decision-making system for the entire fresh meat industry chain, we have achieved deep integration and real-time sharing of data across the entire chain, solving the problem of information fragmentation, improving the transparency and accuracy of the supply chain, and ensuring product quality and safety.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JIANGSU ACAD OF AGRI SCI
- Filing Date
- 2026-01-06
- Publication Date
- 2026-04-17
AI Technical Summary
The fragmentation of information across the entire fresh meat supply chain leads to a lack of holistic decision-making, difficulty in ensuring product quality and safety, low supply chain efficiency, and persistently high costs.
Construct a digital twin-based information fusion decision-making system for the entire fresh meat industry chain. Through communication connections between the platform and the links, achieve deep integration and real-time sharing of data across the entire chain. Utilize digital twin models for data analysis and simulation to provide precise decision support.
It has improved transparency and collaborative efficiency across the entire industry chain, enhanced the controllability of product quality and safety, reduced quality risks caused by information fragmentation, and improved the scientific nature and accuracy of decision-making.
Smart Images
Figure CN121882835A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of fresh meat management technology, specifically a fresh meat whole-industry chain information fusion decision-making system based on digital twins. Background Technology
[0002] In the management and operation of the entire fresh meat industry chain, ensuring product quality and safety, improving supply chain efficiency, and reducing costs have always been the core demands of the industry. Under the traditional model, there is a common problem of information fragmentation in all aspects of fresh meat production, from breeding, slaughtering and processing, cold chain logistics to terminal sales: the breeding stage relies on manual experience to record livestock health and growth data, making it difficult to analyze the impact on meat quality; the slaughtering and processing stage lacks precise implementation of meat grading standards, which can easily lead to fluctuations in yield due to differences in cutting processes; in cold chain logistics, the monitoring of key indicators such as transportation temperature and humidity relies on manual sampling, making it difficult to warn of chain breakage risks in advance, resulting in a high rate of meat spoilage and loss; in the terminal sales stage, due to the lag in updating inventory data, stockouts or backlogs often occur, further exacerbating cost pressures; especially for sales and procurement terminals, timely and accurate decision-making is required based on the breeding, processing, and logistics environments.
[0003] While existing technologies attempt to partially address these issues through the Internet of Things (IoT), big data, and other means, they still have significant limitations. For example, some systems focus only on a single link (such as cold chain monitoring) and fail to achieve data connectivity across the entire chain, resulting in a lack of holistic decision-making.
[0004] In order to solve the above problems, this invention provides a digital twin-based information fusion decision-making system for the entire fresh meat industry chain. Summary of the Invention
[0005] To address the problems of the above solutions, this invention provides a digital twin-based information fusion decision-making system for the entire fresh meat industry chain.
[0006] The objective of this invention can be achieved through the following technical solutions: A digital twin-based information fusion decision-making system for the entire fresh meat industry chain, including the platform end and the link end; The platform includes a process analysis module, a data analysis module, a user module, and a correlation simulation module; The process analysis module is used to perform process analysis, obtain supply chain information of fresh meat products, identify target processes at each stage based on the supply chain information, determine the process collection range and additional collection range at each stage based on the target processes, and send the process collection range and additional collection range to the corresponding stage.
[0007] Furthermore, the platform establishes communication connections with each of the various stages.
[0008] Furthermore, based on the target stage, the scope of data collection and additional data collection areas are determined, including: Identify the various basic link parties and the corresponding link collection scope of the basic link parties based on the information of the industrial chain; Identify the link parties at each stage, compare each basic link party with the link party, eliminate basic link parties that are duplicates of the link party, and determine the link acquisition range of each link party; perform additional judgments on the remaining basic link parties and each link party to obtain the additional basic link parties corresponding to each link party, and mark the link acquisition range of the basic link parties as the additional acquisition range; obtain the link acquisition range and additional acquisition range of the link party.
[0009] Furthermore, additional judgments are made on the remaining basic link parties and each link party, including: Additional standards are set for each basic link party, and feature collection is performed on the corresponding basic link party and link parties according to the additional standards to obtain additional feature data; An additional evaluation model is established. The additional evaluation model is used to analyze the additional characteristic data and additional standards of the corresponding basic link and link to obtain the additional evaluation value, which is 1 or 0. When the additional evaluation value is 1, the basic link party is used as the additional basic link party corresponding to the link party; When the additional evaluation value is 0, no corresponding processing is performed.
[0010] Furthermore, the expression for the additional evaluation model is: ; In the formula: (h, FB) are the input data, h represents the corresponding basic link and the corresponding additional feature data, FB is the additional standard of the corresponding basic link; h→FB means that the corresponding additional feature data meets the additional standard, and the output data is the additional evaluation value FP(h, FB).
[0011] The data analysis module is used to perform data analysis, obtain users' decision analysis needs, and establish a digital twin model of the entire fresh meat industry chain based on these needs. It also acquires comprehensive data from each stage in real time, allocates additional data from the comprehensive data to the corresponding basic stage participants, and obtains the stage data from each basic stage participant. Based on the digital twin, it performs decision simulation analysis on the stage data from each basic stage participant to obtain the decision analysis results for each decision analysis need, and then sends the decision analysis results to the user module.
[0012] The correlation simulation module is used to perform correlation simulation analysis on the data collected from each basic link, identify each data item corresponding to the data collected from each link, perform correlation analysis on each data item, and obtain correlation graphs and target verification schemes. By analyzing the data collected in each stage using digital twin models, target verification schemes, and correlation graphs, the true data results for the corresponding data items are obtained and then sent to the user module.
[0013] Furthermore, correlation analysis is performed on the various data items, including: The data items are combined to obtain several potential combinations. Based on the potential combinations and fresh meat information, relevant historical material data is collected. The combination effect of the potential combinations is obtained based on the historical material data. Based on the combination effect, each potential combination is filtered to obtain several target combinations and target verification schemes. Set corresponding graph units for each data item, add each graph unit to the initial graph, connect each graph unit in the initial graph according to each data item corresponding to each target combination, optimize the display of the initial graph, and obtain the associated graph.
[0014] Furthermore, the potential combinations are screened based on their combined effects, including: The data items that need to be verified are identified from the various data items and marked as data verification items. If the user does not make any settings, all data items can be used as data verification items, or the platform can preset the data items that need to be verified, such as marking data items that are fake or have a distortion rate greater than a preset value as data verification items. Each data validation item is treated as an analysis item, and potential validation methods that can be simulated to validate the analysis items are identified based on the combined effects. The analysis items that can be verified by each potential verification method are summarized to form the verification scope of the potential verification method. Based on the verification of all analysis items, the potential verification methods are combined to form several candidate verification schemes. Each candidate verification scheme is screened to obtain the target verification scheme, and each potential combination corresponding to the target verification scheme is marked as the target combination.
[0015] Furthermore, the data validation item is the data item whose distortion rate is greater than the threshold X1.
[0016] The user module is used to display the decision analysis results of each received decision analysis request and the actual data results of each data item to the user.
[0017] The link includes a data acquisition module; The data acquisition module is used to collect data according to the received stage acquisition range and additional acquisition range to obtain comprehensive acquisition data, which includes stage acquisition data and additional acquisition data; and to send the comprehensive acquisition data to the platform.
[0018] Compared with the prior art, the beneficial effects of the present invention are: By constructing a digital twin-based information fusion decision-making system for the entire fresh meat industry chain, the information barriers between various links in the traditional model have been effectively broken down, achieving deep integration and real-time sharing of data across the entire chain from farming to end-sales. This innovation not only improves the transparency and collaborative efficiency of the entire industry chain, enabling each link to make more accurate and scientific decisions based on global data, but also significantly enhances the controllability of product quality and safety, reducing quality risks caused by information fragmentation. Simultaneously, the system ensures the authenticity of data at each link through intelligent analysis and prediction, solving the problem of the inability to conduct unified analysis across different links in the past. Attached Figure Description
[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0020] Figure 1 This is a block diagram illustrating the principle of the present invention. Detailed Implementation
[0021] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.
[0022] like Figure 1 As shown, the information fusion and decision-making system for the entire fresh meat industry chain based on digital twins includes a platform and various links; the platform is connected to each link in communication.
[0023] The platform includes a process analysis module, a data analysis module, a user module, and a correlation simulation module; The process analysis module is used to perform process analysis, acquire information on the entire fresh meat industry chain, and mark it as industry chain information. This information includes detailed process data and related data for the entire fresh meat industry chain, such as information on each party involved in each process and their responsibilities. If privacy protection is required, code names or other methods can be used to de-anonymize the information. This information can be provided by each party involved in each process to form a complete and comprehensive industry chain information. Based on the industry chain information, the module identifies the target process corresponding to each process end. The target process is determined based on the scope of responsibility of the party involved in the entire fresh meat industry chain. Based on the target process, the module determines the process collection range and additional collection range for the corresponding process end. The process collection range is the collection range corresponding to the target process. The negative additional collection range targets a certain process party that does not use the process end but provides its process data to the process party for transmission. The process collection range and additional collection range are then sent to the corresponding process end.
[0024] In one embodiment, determining the scope of data collection and additional scope of data collection based on the target scope includes: Based on the supply chain information, we identify the basic link parties and the corresponding link collection scope of each basic link party. That is, we analyze the situation where each link party uses the link terminal to determine each link party and mark them as basic link parties. For example, we first classify them according to breeding, slaughtering, etc., and then determine the basic link parties according to the entities corresponding to each category. For example, if three companies are involved in transportation, then they correspond to three basic link parties. The link collection scope of each basic link party is determined according to the data collection requirements. Generally, it is determined according to the data analysis requirements, or supplemented and adjusted according to user needs, etc. Identify the corresponding link party for each link end, compare each basic link party with the link party, eliminate basic link parties that are duplicated with the link party, and determine the link collection range of each link party; perform additional judgment on the remaining basic link parties and each link party to obtain the additional basic link parties corresponding to each link party; if there are no additional basic link parties, the additional collection range is none; mark the link collection range of the basic link parties as the additional collection range.
[0025] In one embodiment, additional judgments are made on the remaining basic link parties and each link party, based on existing methods, such as judging whether they can obtain the link collection data of the basic link parties.
[0026] In one embodiment, additional judgments are made on the remaining basic link parties and each link party, including: Additional standards are set for each basic link party. The baseline standard is the ability to obtain the link data collected by the basic link party. Based on the baseline standard, targeted additional standards are generated for each basic link party. That is, what conditions must be met by the basic link party in order to obtain the link data collected by the basic link party, such as contractual relationship, employment relationship, etc. Based on the standard library of various preset situations, subsequent matching is performed according to the actual situation of the basic link party and industry chain information. Other methods can also be used to set up standards, such as intelligent analysis based on intelligent models such as machine learning and deep learning algorithms.
[0027] Based on the additional standards, feature collection is performed on the corresponding basic link parties and link parties to obtain additional feature data, that is, the relevant data between the two parties is collected according to the additional standards. An additional assessment model is established, and its expression is as follows: ; In the formula: (h, FB) are the input data, h represents the corresponding basic link and the corresponding additional feature data, FB is the additional standard of the corresponding basic link; h→FB means that the corresponding additional feature data meets the additional standard, and the output data is the additional evaluation value FP(h, FB), which is 1 or 0; training is performed using the historical data labeled training set. By analyzing the additional characteristic data and additional standards of the corresponding basic link and link, the additional evaluation value is obtained through additional evaluation model; When the additional evaluation value is 1, the basic link party is used as the additional basic link party corresponding to the link party; When the additional evaluation value is 0, no corresponding processing is performed.
[0028] The data analysis module is used to perform data analysis, obtain the user's decision analysis needs, and target the decision needs of the sales terminal. It manages supply, inventory, pricing, and other related aspects based on information from the entire industry chain. Based on these decision analysis needs and existing digital twin technology, it establishes a digital twin model of the entire fresh meat industry chain. This model can be built based on information from the entire fresh meat industry chain and can be set according to the differences in each stage, such as a breeding model (simulating livestock growth environment, health status, and feed conversion rate), a processing model (the efficiency of cutting various fresh meat products and meat grading under corresponding slaughtering processes), a logistics model (predicting the impact of transportation time and temperature fluctuations on meat quality), and a sales model (analyzing consumer demand, inventory turnover, and price fluctuations). In other words, it uses the digital twin model and subsequently aggregated data collected from each stage to simulate the optimal decisions for each decision analysis need; it acquires comprehensive data from each stage in real time; it obtains stage data from each basic stage in the entire industry chain based on the comprehensive data from each stage; it analyzes the stage data from each basic stage based on the digital twin to obtain the decision analysis results for each decision analysis need, and sends the decision analysis results to the user module.
[0029] In one embodiment, digital twin technology is used for simulation to determine the optimal decision, which is implemented in this invention using existing methods.
[0030] The correlation simulation module is used to perform correlation simulation analysis on the data collected from each basic link, identify the data items corresponding to the data collected from each link, and determine the data items based on the actual data collected from each link, such as the data items corresponding to the cold chain temperature data of meat products; perform correlation analysis on each data item to obtain correlation maps and target verification schemes; By analyzing the data collected in each stage using digital twin models, target verification schemes, and correlation graphs, the true data results for the corresponding data items are obtained and then sent to the user module.
[0031] That is, according to the correlation map and target verification scheme, the data collected in each link is simulated using a digital twin model to determine the corresponding simulation results. The simulation results are then compared with the real results in the data collected in each link to determine whether each data item is real.
[0032] In one embodiment, correlation analysis is performed on the various data items, including: The process involves combining various data items to obtain several potential combinations. For example, given three data items A, B, and C, the potential combinations are A, B, C, AB, AC, BC, and ABC. Fresh meat information, such as beef, is identified. Even though the cattle are live, the term "beef" can still be used to represent them during the breeding stage for consistent description. Historical data is collected based on the potential combinations and fresh meat information; this refers to the historical data corresponding to the specific fresh meat information for each potential combination. The combined effect of the corresponding potential combination is obtained from the historical data, i.e., the normal effect and impact of the combination within the context of the fresh meat information. This combined effect is statistically determined using the historical data. Finally, the potential combinations are filtered based on the combined effect to obtain several target combinations. Set corresponding graph units for each data item, such as circles containing the text of the data item, which can have multiple display forms; add each graph unit to the initial graph, which is a blank graph; connect the graph units in the initial graph according to the data items corresponding to each target combination, that is, connect the data items within the same target combination, and different target combinations can be connected with different forms of lines; optimize the display of the initial graph to obtain the associated graph.
[0033] Display optimization refers to adjusting the position and connections of each map unit to facilitate observation of the relationships between them; optimizing and adjusting based on existing methods; or directly dividing each target combination into independent regions, where each independent region only includes the relationships between the map units corresponding to that target combination; thus forming multiple independent regions.
[0034] In one embodiment, filtering potential combinations based on their combined effects includes: Users can select data items from various data items that need to be verified and mark them as data verification items. If users do not make any settings, all data items can be used as data verification items. Alternatively, the platform can preset the data items that need to be verified, such as marking data items that are fake or have a distortion rate greater than a preset value as data verification items. Each data validation item is treated as an analysis item. Based on the combined effects of each data item, potential validation methods that can be simulated to validate the data item are identified. A potential validation method is a way to determine whether the data is real by comparing the combined effects of one or more potential combinations. For example, if a combination effect is only affected by the data of that data item, the actual effect and the simulated effect can be compared. The difference in the combined effect determines whether the data of that data item is real. If another data item is also affected, the potential validation methods for that data item are eliminated one by one based on the differences between the combined effects. At the same time, a potential validation method may be applicable to the validation of multiple data items. The potential validation methods are intelligently determined using existing methods. Summarize the analytical items that can be verified by each potential verification method to form the verification scope of the potential verification method. Based on the verification of all analytical items, combine the potential verification methods to form several candidate verification schemes. That is, each candidate verification scheme can verify the data of all analytical items, and combine them based on this. Each candidate verification scheme is screened to obtain the target verification scheme, and each potential combination corresponding to the target verification scheme is marked as the target combination.
[0035] In one embodiment, the candidate verification schemes are screened based on cost, and the candidate verification scheme with the lowest cost is marked as the target verification scheme; the screening can also be based on verification accuracy, efficiency, etc.; or other existing technologies can be used for screening.
[0036] The user module is used to display the decision analysis results of each received decision analysis request and the actual data results of each data item to the user, and can perform corresponding processing based on the decision analysis results.
[0037] The aforementioned link is used by various parties in the entire fresh meat industry chain, such as breeding, transportation, slaughtering, and sales. The same link can have multiple parties in the same industry chain. For example, for a sales user, a certain product may be cooperated with multiple breeding and transportation parties; it includes a data collection module. The data acquisition module is used to collect data according to the received stage acquisition range and additional acquisition range to obtain comprehensive acquisition data, which includes stage acquisition data and additional acquisition data; and to send the comprehensive acquisition data to the platform.
[0038] The above formulas are all numerical calculations after removing dimensions. The formulas are obtained by software simulation based on a large amount of data and are closest to the real situation. The preset parameters and preset thresholds in the formulas are set by those skilled in the art according to the actual situation or obtained by simulation based on a large amount of data.
[0039] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.
Claims
1. A digital twin-based information fusion decision-making system for the entire fresh meat product supply chain, characterized in that: Including both the platform side and the process side; The platform includes a process analysis module, a data analysis module, a user module, and a correlation simulation module; the process includes a data acquisition module. The process analysis module is used to perform process analysis, obtain supply chain information of fresh meat products, identify target processes at each stage based on the supply chain information, determine the process collection range and additional collection range at each stage based on the target processes, and send the process collection range and additional collection range to the corresponding stage. The data analysis module is used to perform data analysis, obtain users' decision analysis needs, establish a digital twin model of the entire fresh meat industry chain based on the decision analysis needs, acquire comprehensive data from each link in real time, and allocate additional data from the comprehensive data to the corresponding basic link parties to obtain the link data from each basic link party. Based on the digital twin, decision simulation analysis is performed on the data collected from each benchmark link to obtain the decision analysis results for each decision analysis requirement, and the decision analysis results are sent to the user module. The correlation simulation module is used to perform correlation simulation analysis on the data collected from each basic link, identify each data item corresponding to the data collected from each link, perform correlation analysis on each data item, and obtain correlation graphs and target verification schemes. By analyzing the data collected in each stage through digital twin models, target verification schemes, and correlation graphs, the true data results of the corresponding data items are obtained and sent to the user module. The user module is used to display the decision analysis results of each received decision analysis request and the actual data results of each data item to the user; The data acquisition module is used to acquire data according to the received stage acquisition range and additional acquisition range to obtain comprehensive acquisition data, which includes stage acquisition data and additional acquisition data. The collected data is sent to the platform.
2. The information fusion and decision-making system for the entire fresh meat industry chain based on digital twins as described in claim 1, characterized in that, The platform is connected to each link in the communication system.
3. The information fusion and decision-making system for the entire fresh meat industry chain based on digital twins as described in claim 1, characterized in that, The scope of data collection for each stage and any additional data collection areas are determined based on the target stage, including: Identify the various basic link parties and the corresponding link collection scope of the basic link parties based on the information of the industrial chain; Identify the link parties at each stage, compare each basic link party with the link party, eliminate basic link parties that are duplicates of the link party, and determine the link acquisition range of each link party; perform additional judgments on the remaining basic link parties and each link party to obtain the additional basic link parties corresponding to each link party, and mark the link acquisition range of the basic link parties as the additional acquisition range; obtain the link acquisition range and additional acquisition range of the link party.
4. The information fusion and decision-making system for the entire fresh meat industry chain based on digital twins as described in claim 3, characterized in that, Additional judgments are made on the remaining basic link parties and each link party, including: Additional standards are set for each basic link party, and feature collection is performed on the corresponding basic link party and link parties according to the additional standards to obtain additional feature data; An additional evaluation model is established. The additional evaluation model is used to analyze the additional characteristic data and additional standards of the corresponding basic link and link to obtain the additional evaluation value, which is 1 or 0. When the additional evaluation value is 1, the basic link party is used as the additional basic link party corresponding to the link party; When the additional evaluation value is 0, no corresponding processing is performed.
5. The information fusion and decision-making system for the entire fresh meat industry chain based on digital twins as described in claim 4, characterized in that, The expression for the additional evaluation model is: ； In the formula: (h, FB) are the input data, h represents the corresponding basic link and the corresponding additional feature data, FB is the additional standard of the corresponding basic link; h→FB means that the corresponding additional feature data meets the additional standard, and the output data is the additional evaluation value FP(h, FB).
6. The information fusion and decision-making system for the entire fresh meat industry chain based on digital twins as described in claim 1, characterized in that, Perform correlation analysis on each data item, including: The data items are combined to obtain several potential combinations. Based on the potential combinations and fresh meat information, relevant historical material data is collected. The combination effect of the potential combinations is obtained based on the historical material data. Based on the combination effect, each potential combination is filtered to obtain several target combinations and target verification schemes. Set corresponding graph units for each data item, add each graph unit to the initial graph, connect each graph unit in the initial graph according to each data item corresponding to each target combination, optimize the display of the initial graph, and obtain the associated graph.
7. The information fusion and decision-making system for the entire fresh meat industry chain based on digital twins as described in claim 6, characterized in that, The potential combinations are screened based on their combined effects, including: Identify the data items that need to be validated from among the data items, and mark the data items as data validation items; Each data validation item is treated as an analysis item, and potential validation methods that can be simulated to validate the analysis items are identified based on the combined effects. The analysis items that can be verified by each potential verification method are summarized to form the verification scope of the potential verification method. Based on the verification of all analysis items, the potential verification methods are combined to form several candidate verification schemes. Each candidate verification scheme is screened to obtain the target verification scheme, and each potential combination corresponding to the target verification scheme is marked as the target combination.
8. The information fusion and decision-making system for the entire fresh meat industry chain based on digital twins as described in claim 7, characterized in that, Data validation items are data items whose distortion rate is greater than the threshold X1.