Real-time monitoring and management system for perishable product supply chain based on digital twin technology

Through real-time monitoring and management of perishable product supply chains through digital twin technology and artificial intelligence, the problems of insufficient monitoring, inaccurate analysis and low transparency in the existing technology are solved, and efficient supply chain management and quality assurance are achieved.

CN118246839BActive Publication Date: 2024-11-22DEZHOU UNIV
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410259190.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-03-07
Publication Date
2024-11-22
Estimated Expiration
2044-03-07

AI Technical Summary

Technical Problem

The existing technology has insufficient real-time monitoring capabilities, inaccurate data analysis, low transparency and low management efficiency in the perishable product supply chain, which makes it difficult to ensure quality and safety, and it is difficult to quickly respond to market changes and emergencies.

Method used

Using a system based on digital twin technology, data is collected in real time through IoT sensors, virtual copies of perishable products and their supply chain environment are created, and supply chain management strategies are optimized in combination with data analysis and artificial intelligence algorithms to achieve real-time monitoring and management.

Benefits of technology

It improves the transparency and management efficiency of the supply chain, reduces product losses, ensures quality and safety, can quickly respond to market changes and emergencies, and enhances consumer trust and corporate efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118246839B_ABST
    Figure CN118246839B_ABST
Patent Text Reader

Abstract

The present invention discloses a real-time monitoring and management method for a perishable product supply chain based on digital twin technology, including S1, deploying IoT sensors at key nodes of the perishable product supply chain; S2, sending the key data to a central processing unit in real time through network transmission; S3, analyzing and simulating the created digital twin model in real time; S4, analyzing the collected data using an artificial intelligence algorithm to optimize the supply chain management strategy; S5, the system automatically adjusts the supply chain plan according to the analysis results to minimize losses and maximize efficiency, and automatically takes predetermined response measures when problems in the supply chain are identified; S6, providing a user interface so that supply chain managers and consumers can access the status information of perishable products in real time. The present invention has the advantages of improving supply chain efficiency, increasing transparency, enhancing consumer trust, and optimizing product quality and safety.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of supply chain technology, and in particular to a real-time monitoring and management system for a perishable product supply chain based on digital twin technology. Background Art

[0002] With the rapid development of global trade and information technology, supply chain management has become a key link in corporate operations, especially in the field of perishable products such as food, medicine and agricultural products, whose quality and safety are directly related to public health and corporate reputation. However, in a highly complex and dynamically changing market environment, ensuring the quality and safety of perishable products throughout the supply chain while improving supply chain efficiency and transparency has always been a major challenge in the field of supply chain management.

[0003] In the context of existing technologies, traditional supply chain management methods mainly rely on manual monitoring and static data processing systems, which are inadequate in the face of the complexity and uncertainty of modern supply chains. Specifically, existing technologies have obvious defects in the following aspects:

[0004] 1. Insufficient real-time monitoring capabilities: Lack of effective real-time data collection and monitoring methods make it impossible to monitor the status of perishable products in each link of the supply chain in real time, resulting in delayed response to problems and difficulty in taking effective measures to ensure product quality and safety.

[0005] 2. Inaccurate data analysis and decision support: Existing data processing and analysis tools are not advanced enough, making it difficult to accurately extract valuable information from massive amounts of data, which affects the accuracy and effectiveness of supply chain management decisions.

[0006] 3. Low supply chain transparency and consumer trust: Due to the lack of effective information sharing and transparency mechanisms, it is difficult for consumers to obtain detailed information about products, such as origin, temperature and humidity conditions during transportation, etc., which in turn affects consumers' trust in products and brands.

[0007] 4. Low efficiency and flexibility in supply chain management: Traditional supply chain management methods lack sufficient flexibility and are unable to quickly adapt to emergencies such as changes in market demand and supply chain disruptions, affecting the efficiency and response speed of the entire supply chain.

[0008] Therefore, how to provide a real-time monitoring and management system for perishable product supply chains based on digital twin technology is an urgent problem that technical personnel in this field need to solve. Summary of the invention

[0009] One object of the present invention is to propose a real-time monitoring and management system for perishable product supply chains based on digital twin technology. The present invention makes full use of the Internet of Things, data analysis, artificial intelligence algorithms and digital twin technology, and describes in detail how to achieve real-time monitoring and management of perishable products throughout the supply chain process. By deploying sensors at key nodes to collect data in real time, using digital twin technology to create virtual copies of perishable products and their supply chain environment, and combining data analysis with artificial intelligence algorithms to optimize supply chain management strategies, the present invention has the advantages of improving supply chain efficiency, increasing transparency, enhancing consumer trust, and optimizing product quality and safety.

[0010] According to an embodiment of the present invention, a real-time monitoring and management system for a perishable product supply chain based on digital twin technology includes the following steps:

[0011] S1. Deploy IoT sensors at key nodes in the perishable product supply chain to collect key data on raw material procurement, processing, packaging, transportation and sales of perishable products in real time;

[0012] S2. The key data is transmitted to the central processing unit in real time via the network. The central processing unit uses the digital twin technology to create a virtual copy of the perishable product and its supply chain environment based on the received data to simulate the actual physical state and environmental conditions;

[0013] S3. Perform real-time analysis and simulation on the created digital twin model to assess the current status of perishable products in the supply chain and predict potential risks or problems;

[0014] S4. Use artificial intelligence algorithms to analyze the collected data and optimize supply chain management strategies, including product demand forecasting, transportation route optimization, and storage condition adjustment;

[0015] S5. The system automatically adjusts the supply chain plan based on the analysis results to minimize losses and maximize efficiency, and automatically takes predetermined response measures when problems in the supply chain are identified;

[0016] S6. Provide a user interface that enables supply chain managers and consumers to access real-time status information of perishable products, including origin, current condition, and expected arrival time, to improve supply chain transparency and consumer trust;

[0017] The S1 specifically includes:

[0018] S11. Key nodes in the perishable product supply chain k It is the raw material purchasing point P, processing center C, packaging station B, logistics transfer station L and sales terminal S;

[0019] S12. At each key node N kInstall IoT devices equipped with temperature T, humidity H and location GPS sensors to monitor and record key parameters of perishable products in real time;

[0020] S13, each IoT device sends the collected data, including temperature T, humidity H and location information GPS, to the central processing unit CPU in real time via the wireless network W;

[0021] S14. Set the data collection frequency F so that the IoT sensor can effectively capture the real-time changes in the status of perishable products. The frequency F is set according to the product characteristics and supply chain requirements, and the unit is times / hour. -1 ;

[0022] S15. Each IoT device has a unique ID to ensure the accuracy of data transmission and traceability of its source;

[0023] The S2 specifically includes:

[0024] S21. Establish a central processing unit CPU to receive and process data from key nodes N in the supply chain. k The data collected by IoT devices include temperature T, humidity H and location information GPS;

[0025] S22, whenever the IoT device sends data through the wireless network W, the central processing unit CPU receives the data through a receiving module R;

[0026] S23, the central processing unit CPU has a built-in data pre-processing module P to format and clean the received data:

[0027] The temperature and humidity based Z-score method is used to treat all Z-scores with an absolute value greater than 3 as outliers and remove them from the data set:

[0028]

[0029] Where X represents the measurement value of a single sensor, μ and σ represent the mean and standard deviation of the historical data of the sensor, respectively;

[0030] For the missing data of temperature T and humidity H, linear interpolation is used to fill in the missing data:

[0031]

[0032] Among them, X miss is the missing value to be filled, X n and X n+1 is the measured value before and after the missing value, t miss ,t n ,t n+1represent the missing values ​​and the time points corresponding to the two measurements, respectively;

[0033] Convert the temperature and humidity data to a common scale:

[0034]

[0035] Among them, X norm is the normalized value, X min and X max are the minimum and maximum values ​​within the measurement range of the sensor respectively;

[0036] S24. After data preprocessing, digital twin technology is used to create a virtual copy of the perishable product and its supply chain environment on the central processing unit CPU, including each key node N k The virtual representation of the product and the simulation of the product's state changes at each node. The model is based on the pre-processed temperature T, humidity H and location GPS data, and uses the linear regression method to simulate the behavior and state changes of perishable products in the supply chain;

[0037] Use a physics-based model to estimate the loss of product mass under given environmental conditions:

[0038] Q new =Q old -Δt·f(T,H);

[0039] Among them, Q new and Q old Represent the new and original product quality indicators respectively, Δt is the time interval, and f(T,H) is a function based on temperature T and humidity H;

[0040] Using location data GPS, combined with logistics and transportation models, the expected arrival time of products from one node to another is calculated:

[0041]

[0042] Among them, T arrival and T depart are the expected arrival time and departure time, D is the distance between the two nodes, and V is the transportation speed;

[0043] S25. After creating a virtual copy of the perishable product and its supply chain environment, use integrated analytical tools and algorithms to perform real-time analysis and simulation on the virtual copy:

[0044] The decision tree model is used for risk assessment and prediction. The decision tree model is expressed as a product quality prediction model P(Q|T1,H1), where Q is the change in product quality, T1 and H1 are temperature and humidity conditions respectively;

[0045] Optimize the transportation and storage processes in the supply chain using a linear programming model to minimize the total cost:

[0046] minZ=C transport +C loss ;

[0047] Where Z is the total cost, C transport is the transportation cost, C loss is the cost due to product loss;

[0048] Use time series models to predict product demand based on data:

[0049] Y t =φ1Y t-1 +φ2Y t-2 +...+θ n ε t-n +ε t ;

[0050] Among them, Y t is the forecast demand at time t, φ and θ are model parameters, ε t is the error term;

[0051] S26. Based on the above analysis and simulation results, the central processing unit CPU performs decision support to automatically adjust and optimize supply chain operations:

[0052] Implement dynamic adjustment strategies to deal with predicted risks and problems. According to the output of the product quality prediction model P(Q|T1,H1), when the predicted quality result of a batch of products is lower than the threshold Q threshold , the system will automatically adjust the relevant links to reduce losses:

[0053] P(Q|T1,H1) threshold ;

[0054] Using the optimization analysis results, the optimal transportation and storage strategies obtained by automatically replanning transportation routes or adjusting inventory levels will be automatically applied to supply chain operations;

[0055] Based on the product demand forecast results, when the forecast demand Y t Exceeding the sum of inventory levels plus production capacity, automatically adjust production and supply plans, using demand forecasts obtained from time series analysis Y t , combined with the inventory level L inventory and production capacity C production , calculate the production adjustment D production :

[0056] D production =Y t -(L​inventory +C production );

[0057] The S3 specifically includes:

[0058] S31. Perform real-time analysis and simulation on a virtual copy of the perishable product and its supply chain environment created using digital twin technology on the central processing unit CPU to assess the current status of the perishable product in the supply chain and predict potential risks and problems;

[0059] S32. A dynamic system model DSM is used to simulate the behavior and state changes of perishable products in the supply chain. The dynamic system model considers the influence of temperature T, humidity H, time t and location GPS factors. The dynamic system model DSM formula is:

[0060]

[0061] Among them, S represents the state of perishable products;

[0062] S33. Use the random forest algorithm to analyze the data collected from IoT devices to identify and predict possible problems in the supply chain. The MLA formula of the random forest algorithm is:

[0063] P(event|T,H,GPS)=learn(data);

[0064] Among them, P(event|T,H,GPS) is the probability of a specific event occurring under given temperature T, humidity H and location GPS;

[0065] S34. Based on the output of the digital twin model and the random forest algorithm, the decision support system DSS in the central processing unit CPU will evaluate the current status of perishable products and generate corresponding risk reports and warnings, which will automatically trigger the warning mechanism. The report format is:

[0066] Report: {product ID, risk level, recommended action, timestamp};

[0067] Among them, the risk level is classified based on the severity of potential risks, and the recommended measures are proposed based on the predicted results and best practices;

[0068] S35. Using the output of the forecasting model and the decision support system DSS, the central processing unit CPU will dynamically adjust the supply chain management strategy to prevent potential problems and mitigate their impact, adjust transportation routes, change storage conditions or modify supply chain plans;

[0069] The S4 specifically includes:

[0070] S41, using the collected temperature T, humidity H, location information GPS, and time t of perishable products to perform data analysis and optimize supply chain management strategies;

[0071] S42. Use the time series analysis method TSA to predict the demand for perishable products based on historical data:

[0072] D t+1 =αD t +βT t +γH t +∈;

[0073] Among them, D t+1 represents the forecast demand for the next time period, D t is the demand in the current time period, T t and H t Respectively represent the average temperature and humidity in the corresponding time period, α, β, γ are model parameters, and ∈ is the error term;

[0074] S43. Use linear programming LP to optimize transportation routes and storage conditions to minimize the total cost C:

[0075]

[0076] Among them, c i represents the cost of the i-th transportation route, x i is a decision variable, indicating whether to select (1) or not select (0) the ith transport route, s j represents the cost of the jth storage condition, y j is a decision variable, indicating whether to select (1) or not select (0) the jth storage condition;

[0077] S44. Apply artificial intelligence algorithms AI, neural networks or deep learning to conduct in-depth analysis of data in the supply chain to identify improvement points and optimization opportunities. The algorithm automatically proposes improvement suggestions by analyzing the relationship between temperature T, humidity H, location GPS data and supply chain efficiency;

[0078] S44. Use a multi-layer perceptron (MLP) neural network model to conduct in-depth analysis of data in the supply chain, especially temperature T, humidity H, and location GPS. The model structure includes an input layer, at least one hidden layer, and an output layer. The training process aims to minimize the prediction error. Through training, MLP can discover nonlinear relationships in the data and provide data-driven decision support for supply chain management:

[0079]

[0080] Where E is the mean error, n is the number of samples, and Y i is the true value of the i-th sample, is the predicted value of the model.

[0081] Optionally, the S5 specifically includes:

[0082] S51. Use the decision support algorithm DSA to automatically generate a supply chain adjustment plan based on the data analysis results. The decision support algorithm DSA formula is:

[0083] Optimize(SC parameters )→SC adjustments ;

[0084] Among them, SC parameters Represents a set of key parameters of the supply chain, SC adjustments It is a set of supply chain adjustment measures determined according to the optimization goal;

[0085] S52. Implement dynamic inventory management (DIM) to adjust inventory levels based on real-time data and demand forecasts to reduce the risk of oversupply and out-of-stock:

[0086] L new =L current +P incoming -D forecasted ;

[0087] Among them, L new is the adjusted inventory level, L current is the current inventory level, P incoming is the expected purchase quantity, D forecasted is the predicted demand;

[0088] S53, optimize logistics and transportation strategy OTS, select the best transportation mode and route according to the characteristics of perishable products and real-time monitoring data. The transportation strategy OTS formula is:

[0089] Minimize(T cost +T risk );

[0090] Among them, T cost is the transportation cost, T risk It is the risk cost of product loss or quality degradation during transportation;

[0091] S54. Apply advanced scheduling algorithm ASA to automatically adjust production and distribution plans, respond to demand changes and supply chain disruption risks in real time, and maximize customer satisfaction. The scheduling algorithm ASA formula is:

[0092] Minimize(D delay -C satisfaction );

[0093] Among them, Ddelay is the production or delivery delay, C satisfaction is a cost measure of customer satisfaction, with negative values ​​indicating increased satisfaction.

[0094] Optionally, the S6 specifically includes:

[0095] S61. Provide a user interface to enable supply chain managers and consumers to access real-time status information of perishable products;

[0096] S62. Implement the data visualization module DVM to dynamically display the real-time data and historical trends of perishable products on the user interface, including temperature T, humidity H, location GPS, and supply chain efficiency indicators. The formula of the data visualization module DVM is:

[0097] Visualize(T,H,GPS,SC metrics );

[0098] Among them, SC metrics represents supply chain efficiency metrics, including shipping times, inventory levels, and order fill rates;

[0099] S63, integrated feedback mechanism FM, allowing users to provide feedback and evaluation of perishable product status information through the user interface UI. These feedbacks will be automatically recorded and analyzed by the system to further optimize supply chain management and improve user experience. The feedback mechanism FM formula is:

[0100] CollectFeedback(UserID,Feedback)→SC improvements ;

[0101] Among them, UserID is the user ID providing feedback, Feedback is the specific feedback content provided by the user, and SC improvements are supply chain improvement measures taken based on user feedback;

[0102] S64. Deploy the intelligent recommendation system IRS to recommend optimized purchase and consumption suggestions of related products to users based on user behavior and preferences and real-time status information of perishable products. The formula of the intelligent recommendation system IRS is:

[0103] Recommend(UserID,SC data )→Recommendations;

[0104] Among them, SC data Including real-time status data and supply chain information of perishable products, Recommendations are based on the behavior and preferences of the user UserID, and SC data Generated recommendation list;

[0105] S65. Using blockchain technology BCT, by recording and verifying information at each transaction node, consumers can be provided with unalterable proof of product origin and history. The blockchain technology BCT formula is:

[0106] Secure(SC transactions )→Trust;

[0107] Among them, SC transactions Representing transaction records in the supply chain, Trust is the consumer trust established by ensuring that data cannot be tampered with.

[0108] The beneficial effects of the present invention are:

[0109] (1) The present invention realizes real-time monitoring and intelligent management of the entire supply chain of perishable products by integrating IoT sensing technology, digital twin technology, data analysis and artificial intelligence algorithms. This comprehensive application enables the system to collect and process key data in the supply chain, such as temperature, humidity and location information, in real time to ensure the quality and safety of perishable products during transportation and storage. By building a digital twin model of the supply chain, the system can simulate the situation in the physical world in a virtual environment, predict and solve potential problems, thereby effectively improving the transparency and management efficiency of the supply chain, reducing losses and ensuring product quality.

[0110] (2) The user interface and data visualization module of the present invention provide a platform for real-time data access and information sharing for supply chain managers and consumers, enhancing supply chain transparency and improving consumers' trust in products and brands. Through real-time monitoring and data analysis, consumers can easily obtain detailed information about products, such as origin, current status, and expected arrival time, which is difficult to achieve in traditional supply chain management methods.

[0111] (3) The present invention uses advanced scheduling algorithms and decision support systems to dynamically adjust supply chain plans, such as production scheduling, inventory management, and transportation routes, based on real-time data and market demand. This dynamic management capability enables the supply chain to quickly adapt to market changes and unexpected events such as supply chain disruptions, improving the response speed and flexibility of the entire supply chain while maximizing customer satisfaction and corporate benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0112] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0113] Figure 1 This is a schematic diagram of the overall framework of a real-time monitoring and management system for a perishable product supply chain based on digital twin technology proposed by the present invention;

[0114] Figure 2 A schematic diagram of the deployment of IoT sensors at key nodes of a perishable product supply chain real-time monitoring and management system based on digital twin technology proposed by the present invention;

[0115] Figure 3 The present invention proposes a structural diagram of a real-time monitoring and management system for a perishable product supply chain based on digital twin technology, using blockchain technology to ensure data security and non-tamperability of supply chain information. DETAILED DESCRIPTION

[0116] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0117] refer to Figure 1-Figure 3 , a real-time monitoring and management system for perishable product supply chain based on digital twin technology includes the following steps:

[0118] S1. Deploy IoT sensors at key nodes in the perishable product supply chain to collect key data on raw material procurement, processing, packaging, transportation and sales of perishable products in real time;

[0119] S2. Send key data to the central processing unit in real time through network transmission. Based on the received data, the central processing unit uses digital twin technology to create a virtual copy of the perishable product and its supply chain environment, simulating the actual physical state and environmental conditions;

[0120] S3. Perform real-time analysis and simulation on the created digital twin model to assess the current status of perishable products in the supply chain and predict potential risks or problems;

[0121] S4. Use artificial intelligence algorithms to analyze the collected data and optimize supply chain management strategies, including product demand forecasting, transportation route optimization, and storage condition adjustment;

[0122] S5. The system automatically adjusts the supply chain plan based on the analysis results to minimize losses and maximize efficiency, and automatically takes predetermined response measures when problems in the supply chain are identified;

[0123] S6. Provide a user interface that enables supply chain managers and consumers to access real-time status information of perishable products, including origin, current condition, and expected arrival time, to improve supply chain transparency and consumer trust;

[0124] S1 specifically includes:

[0125] S11. Key nodes in the perishable product supply chaink It is the raw material purchasing point P, processing center C, packaging station B, logistics transfer station L and sales terminal S;

[0126] S12. At each key node N k Install IoT devices equipped with temperature T, humidity H and location GPS sensors to monitor and record key parameters of perishable products in real time;

[0127] S13, each IoT device sends the collected data, including temperature T, humidity H and location information GPS, to the central processing unit CPU in real time via the wireless network W;

[0128] S14. Set the data collection frequency F so that the IoT sensor can effectively capture the real-time changes in the status of perishable products. The frequency F is set according to the product characteristics and supply chain requirements, and the unit is times / hour. -1 ;

[0129] S15. Each IoT device has a unique ID to ensure the accuracy of data transmission and traceability of its source;

[0130] S2 specifically includes:

[0131] S21. Establish a central processing unit CPU to receive and process data from key nodes N in the supply chain. k The data collected by IoT devices include temperature T, humidity H and location information GPS;

[0132] S22, whenever the IoT device sends data through the wireless network W, the central processing unit CPU receives the data through a receiving module R;

[0133] S23, the central processing unit CPU has a built-in data pre-processing module P to format and clean the received data:

[0134] The temperature and humidity based Z-score method is used to treat all Z-scores with an absolute value greater than 3 as outliers and remove them from the data set:

[0135]

[0136] Where X represents the measurement value of a single sensor, μ and σ represent the mean and standard deviation of the historical data of the sensor, respectively;

[0137] For the missing data of temperature T and humidity H, linear interpolation is used to fill in the missing data:

[0138]

[0139] Among them, X miss is the missing value to be filled, Xn and X n+1 is the measured value before and after the missing value, t miss ,t n ,t n+1 represent the missing values ​​and the time points corresponding to the two measurements, respectively;

[0140] Convert the temperature and humidity data to a common scale:

[0141]

[0142] Among them, X norm is the normalized value, X min and X max are the minimum and maximum values ​​within the measurement range of the sensor respectively;

[0143] S24. After data preprocessing, digital twin technology is used to create a virtual copy of the perishable product and its supply chain environment on the central processing unit CPU, including each key node N k The virtual representation of the product and the simulation of the product's state changes at each node. The model is based on the pre-processed temperature T, humidity H and location GPS data, and uses the linear regression method to simulate the behavior and state changes of perishable products in the supply chain;

[0144] Use a physics-based model to estimate the loss of product mass under given environmental conditions:

[0145] Q new =Q old -Δt·f(T,H);

[0146] Among them, Q new and Q old Represent the new and original product quality indicators respectively, Δt is the time interval, and f(T,H) is a function based on temperature T and humidity H;

[0147] Using location data GPS, combined with logistics and transportation models, the expected arrival time of products from one node to another is calculated:

[0148]

[0149] Among them, T arrival and T depart are the expected arrival time and departure time, D is the distance between the two nodes, and V is the transportation speed;

[0150] S25. After creating a virtual copy of the perishable product and its supply chain environment, use integrated analytical tools and algorithms to perform real-time analysis and simulation on the virtual copy:

[0151] The decision tree model is used for risk assessment and prediction. The decision tree model is expressed as a product quality prediction model P(Q|T1,H1), where Q is the change in product quality, T1 and H1 are temperature and humidity conditions respectively;

[0152] Optimize the transportation and storage processes in the supply chain using a linear programming model to minimize the total cost:

[0153] minZ=C transport +C loss ;

[0154] Where Z is the total cost, C transport is the transportation cost, C loss is the cost due to product loss;

[0155] Use time series models to predict product demand based on data:

[0156] Y t =φ1Y t-1 +φ2Y t-2 +...+θ n ε t-n +ε t ;

[0157] Among them, Y t is the forecast demand at time t, φ and θ are model parameters, ε t is the error term;

[0158] S26. Based on the above analysis and simulation results, the central processing unit CPU performs decision support to automatically adjust and optimize supply chain operations:

[0159] Implement dynamic adjustment strategies to deal with predicted risks and problems. According to the output of the product quality prediction model P(Q|T1,H1), when the predicted quality result of a batch of products is lower than the threshold Q threshold , the system will automatically adjust the relevant links to reduce losses:

[0160] P(Q|T1,H1) threshold ;

[0161] Using the optimization analysis results, the optimal transportation and storage strategies obtained by automatically replanning transportation routes or adjusting inventory levels will be automatically applied to supply chain operations;

[0162] Based on the product demand forecast results, when the forecast demand Y t Exceeding the sum of inventory levels plus production capacity, automatically adjust production and supply plans, using demand forecasts obtained from time series analysis Y t , combined with the inventory level L inventory and production capacity C​production , calculate the production adjustment D production :

[0163] D production =Y t -(L inventory +C production );

[0164] S3 specifically includes:

[0165] S31. Perform real-time analysis and simulation on a virtual copy of the perishable product and its supply chain environment created using digital twin technology on the central processing unit CPU to assess the current status of the perishable product in the supply chain and predict potential risks and problems;

[0166] S32. The dynamic system model DSM is used to simulate the behavior and state changes of perishable products in the supply chain. The dynamic system model considers the influence of temperature T, humidity H, time t and location GPS factors. The dynamic system model DSM formula is:

[0167]

[0168] Among them, S represents the state of perishable products;

[0169] S33. Use the random forest algorithm to analyze the data collected from IoT devices to identify and predict possible problems in the supply chain. The MLA formula of the random forest algorithm is:

[0170] P(event|T,H,GPS)=learn(data);

[0171] Among them, P(event|T,H,GPS) is the probability of a specific event occurring under given temperature T, humidity H and location GPS;

[0172] S34. Based on the output of the digital twin model and the random forest algorithm, the decision support system DSS in the central processing unit CPU will evaluate the current status of perishable products and generate corresponding risk reports and warnings, which will automatically trigger the warning mechanism. The report format is:

[0173] Report: {product ID, risk level, recommended action, timestamp};

[0174] Among them, the risk level is classified based on the severity of potential risks, and the recommended measures are proposed based on the predicted results and best practices;

[0175] S35. Using the output of the forecasting model and the decision support system DSS, the central processing unit CPU will dynamically adjust the supply chain management strategy to prevent potential problems and mitigate their impact, adjust transportation routes, change storage conditions or modify supply chain plans;

[0176] S4 specifically includes:

[0177] S41, using the collected temperature T, humidity H, location information GPS, and time t of perishable products to perform data analysis and optimize supply chain management strategies;

[0178] S42. Use the time series analysis method TSA to predict the demand for perishable products based on historical data:

[0179] D t+1 =αD t +βT t +γH t +∈;

[0180] Among them, D t+1 represents the forecast demand for the next time period, D t is the demand in the current time period, T t and H t Respectively represent the average temperature and humidity in the corresponding time period, α, β, γ are model parameters, and ∈ is the error term;

[0181] S43. Use linear programming LP to optimize transportation routes and storage conditions to minimize the total cost C:

[0182]

[0183] Among them, c i represents the cost of the i-th transportation route, x i is a decision variable, indicating whether to select (1) or not select (0) the ith transport route, s j represents the cost of the jth storage condition, y j is a decision variable, indicating whether to select (1) or not select (0) the jth storage condition;

[0184] S44. Apply artificial intelligence algorithms AI, neural networks or deep learning to conduct in-depth analysis of data in the supply chain to identify improvement points and optimization opportunities. The algorithm automatically proposes improvement suggestions by analyzing the relationship between temperature T, humidity H, location GPS data and supply chain efficiency;

[0185] S44. Use a multi-layer perceptron (MLP) neural network model to conduct in-depth analysis of data in the supply chain, especially temperature T, humidity H, and location GPS. The model structure includes an input layer, at least one hidden layer, and an output layer. The training process aims to minimize the prediction error. Through training, MLP can discover nonlinear relationships in the data and provide data-driven decision support for supply chain management:

[0186]

[0187] Where E is the mean error, n is the number of samples, and Y i is the true value of the i-th sample, is the predicted value of the model.

[0188] In this implementation, S5 specifically includes:

[0189] S51. Use the decision support algorithm DSA to automatically generate a supply chain adjustment plan based on the data analysis results. The decision support algorithm DSA formula is:

[0190] Optimize(SC parameters )→SC adjustments ;

[0191] Among them, SC parameters Represents a set of key parameters of the supply chain, SC adjustments It is a set of supply chain adjustment measures determined according to the optimization goal;

[0192] S52. Implement dynamic inventory management (DIM) to adjust inventory levels based on real-time data and demand forecasts to reduce the risk of oversupply and out-of-stock:

[0193] L new =L current +P incoming -D forecasted ;

[0194] Among them, L new is the adjusted inventory level, L current is the current inventory level, P incoming is the expected purchase quantity, D forecasted is the predicted demand;

[0195] S53, optimize logistics and transportation strategy OTS, select the best transportation mode and route according to the characteristics of perishable products and real-time monitoring data. The transportation strategy OTS formula is:

[0196] Minimize(T cost +T risk );

[0197] Among them, T cost is the transportation cost, T risk It is the risk cost of product loss or quality degradation during transportation;

[0198] S54. Apply advanced scheduling algorithm ASA to automatically adjust production and distribution plans, respond to demand changes and supply chain disruption risks in real time, and maximize customer satisfaction. The scheduling algorithm ASA formula is:

[0199] Minimize(Ddelay -C satisfaction );

[0200] Among them, D delay is the production or delivery delay, C satisfaction is a cost measure of customer satisfaction, with negative values ​​indicating increased satisfaction.

[0201] In this implementation, S6 specifically includes:

[0202] S61. Provide a user interface to enable supply chain managers and consumers to access real-time status information of perishable products;

[0203] S62. Implement the data visualization module DVM to dynamically display the real-time data and historical trends of perishable products on the user interface, including temperature T, humidity H, location GPS, and supply chain efficiency indicators. The formula of the data visualization module DVM is:

[0204] Visualize(T,H,GPS,SC metrics );

[0205] Among them, SC metrics represents supply chain efficiency metrics, including shipping times, inventory levels, and order fill rates;

[0206] S63, integrated feedback mechanism FM, allowing users to provide feedback and evaluation of perishable product status information through the user interface UI. These feedbacks will be automatically recorded and analyzed by the system to further optimize supply chain management and improve user experience. The feedback mechanism FM formula is:

[0207] CollectFeedback(UserID,Feedback)→SC improvements ;

[0208] Among them, UserID is the user ID providing feedback, Feedback is the specific feedback content provided by the user, and SC improvements are supply chain improvement measures taken based on user feedback;

[0209] S64. Deploy the intelligent recommendation system IRS to recommend optimized purchase and consumption suggestions of related products to users based on user behavior and preferences and real-time status information of perishable products. The formula of the intelligent recommendation system IRS is:

[0210] Recommend(UserID,SC data )→Recommendations;

[0211] Among them, SC dataIncluding real-time status data and supply chain information of perishable products, Recommendations are based on the behavior and preferences of the user UserID, and SC data Generated recommendation list;

[0212] S65. Using blockchain technology BCT, by recording and verifying information at each transaction node, consumers can be provided with unalterable proof of product origin and history. The blockchain technology BCT formula is:

[0213] Secure(SC transactions )→Trust;

[0214] Among them, SC transactions Representing transaction records in the supply chain, Trust is the consumer trust established by ensuring that data cannot be tampered with.

[0215] Embodiment 1:

[0216] In the supply chain management of a large fruit and vegetable distribution company, due to the perishable characteristics of products and the complex supply chain structure, ensuring that products maintain the best quality from the place of production to the hands of consumers has always been a major challenge for the company. Especially in the hot summer season, temperature control during transportation and storage is essential to maintain the freshness of products. Traditional supply chain management methods rely on regular manual inspections and historical data analysis, which is not only time-consuming and labor-intensive, but also difficult to achieve real-time monitoring and instant decision-making of the entire supply chain, resulting in high product loss rates, inventory backlogs, and reduced consumer satisfaction.

[0217] To solve these problems, the company decided to deploy the real-time monitoring and management system for perishable product supply chains based on digital twin technology proposed in this paper. The system first deployed IoT sensors at key nodes of the supply chain, such as farm collection points, processing plants, warehouses, and transportation vehicles, to collect data on temperature, humidity, and product location in real time. These data are transmitted to the central processing unit in real time through the network, and a virtual copy of the supply chain is created using digital twin technology, achieving an accurate simulation of the actual supply chain environment.

[0218] Through real-time monitoring data and digital twin models, the system can accurately predict quality problems that products may encounter during transportation and storage, and take measures in advance, such as adjusting the temperature of refrigerated trucks or changing transportation routes. In addition, the system also uses artificial intelligence algorithms to analyze historical data and real-time data to optimize supply chain management strategies, such as adjusting product purchases, optimizing inventory levels, and predicting market demand.

[0219] After implementing the system, the company conducted a three-month pilot project in the summer of 2019. During the project, the system successfully reduced product loss from 12% to 5%; at the same time, logistics costs were reduced by 15% by optimizing transportation routes and warehouse management. More importantly, the system improves supply chain transparency, and consumers can query the transportation status and storage conditions of products in real time through the company's website, which significantly improves consumer trust and satisfaction. The specific data is shown in the following table:

[0220] Table 1: Comparison before and after the implementation of the perishable product supply chain management system

[0221] Data indicators Before implementation After implementation After implementation Product loss rate 12% slow Reduced by 7 percentage points Logistics costs - 15% reduction - Customer Satisfaction 75% 92% An increase of 17 percentage points Supply Chain Transparency Low high - Ability to respond to market changes slow quick Significant improvement

[0222] In addition, the deployment of the system has also helped the company achieve dynamic management of the supply chain, enabling rapid response to emergencies such as market changes and supply chain disruptions, ensuring the quality and safety of perishable products and improving the efficiency and response speed of the entire supply chain.

[0223] In summary, the implementation of the present invention not only solves the monitoring and management problems of perishable products in the supply chain process, but also brings significant economic and social benefits by improving the transparency of the supply chain and optimizing management strategies, providing a new solution for the supply chain management of perishable products.

[0224] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. Real-time monitoring and management system for perishable product supply chain based on digital twin technology, characterized by: The following steps are included: S1. Deploy IoT sensors at key nodes in the perishable product supply chain to collect key data on raw material procurement, processing, packaging, transportation and sales of perishable products in real time; S2. The key data is transmitted to the central processing unit in real time via the network. The central processing unit uses the digital twin technology to create a virtual copy of the perishable product and its supply chain environment based on the received data to simulate the actual physical state and environmental conditions; S3. Perform real-time analysis and simulation on the created digital twin model to assess the current status of perishable products in the supply chain and predict potential risks or problems; S4. Use artificial intelligence algorithms to analyze the collected data and optimize supply chain management strategies, including product demand forecasting, transportation route optimization, and storage condition adjustment; S5. The system automatically adjusts the supply chain plan based on the analysis results to minimize losses and maximize efficiency, and automatically takes predetermined response measures when problems in the supply chain are identified; S6. Provide a user interface that enables supply chain managers and consumers to access real-time status information of perishable products, including origin, current condition, and expected arrival time, to improve supply chain transparency and consumer trust; The S1 specifically includes: S11. Key nodes in the perishable product supply chain k It is the raw material purchasing point P, processing center C, packaging station B, logistics transfer station L and sales terminal S; S12. At each key node N k Install IoT devices equipped with temperature T, humidity H and location GPS sensors to monitor and record key parameters of perishable products in real time; S13, each IoT device sends the collected data, including temperature T, humidity H and location information GPS, to the central processing unit CPU in real time via the wireless network W; S14. Set the data collection frequency F so that the IoT sensor can effectively capture the real-time changes in the status of perishable products. The frequency F is set according to the product characteristics and supply chain requirements, and the unit is times / hour. -1 ; S15. Each IoT device has a unique ID to ensure the accuracy of data transmission and traceability of its source; The S2 specifically includes: S21. Establish a central processing unit CPU to receive and process data from key nodes N in the supply chain. k The data collected by IoT devices include temperature T, humidity H and location information GPS; S22, whenever the IoT device sends data through the wireless network W, the central processing unit CPU receives the data through a receiving module R; S23, the central processing unit CPU has a built-in data pre-processing module P to format and clean the received data: The temperature and humidity based Z-score method is used to treat all Z-scores with an absolute value greater than 3 as outliers and remove them from the data set: Where X represents the measurement value of a single sensor, μ and σ represent the mean and standard deviation of the historical data of the sensor, respectively; For the missing data of temperature T and humidity H, linear interpolation is used to fill in the missing data: Among them, X miss is the missing value to be filled, X n and X n+1 is the measured value before and after the missing value, t miss ,t n ,t n+1 represent the missing values ​​and the time points corresponding to the two measurements, respectively; Convert the temperature and humidity data to a common scale: Among them, X norm is the normalized value, X min and X max are the minimum and maximum values ​​within the measurement range of the sensor respectively; S24. After data preprocessing, digital twin technology is used to create a virtual copy of the perishable product and its supply chain environment on the central processing unit CPU, including each key node N k The virtual representation of the product and the simulation of the product's state changes at each node. The model is based on the pre-processed temperature T, humidity H and location GPS data, and uses the linear regression method to simulate the behavior and state changes of perishable products in the supply chain; Use a physics-based model to estimate the loss of product mass under given environmental conditions: Q new =Q old -Δt·f(T,H); Among them, Q new and Q old Represent the new and original product quality indicators respectively, Δt is the time interval, and f(T,H) is a function based on temperature T and humidity H; Using location data GPS, combined with logistics and transportation models, the expected arrival time of products from one node to another is calculated: Among them, T arrival and T depart are the expected arrival time and departure time, D is the distance between the two nodes, and V is the transportation speed; S25. After creating a virtual copy of the perishable product and its supply chain environment, use integrated analytical tools and algorithms to perform real-time analysis and simulation on the virtual copy: The decision tree model is used for risk assessment and prediction. The decision tree model is expressed as a product quality prediction model P(Q|T1,H1), where Q is the change in product quality, T1 and H1 are temperature and humidity conditions respectively; Optimize the transportation and storage processes in the supply chain using a linear programming model to minimize the total cost: minZ=C transport +C loss ; Where Z is the total cost, C transport is the transportation cost, C loss is the cost due to product loss; Use time series models to predict product demand based on data: Y t =φ1Y t-1 +φ2Y t-2 +...+θ n e t-n +e t ; Among them, Y t is the forecast demand at time t, φ and θ are model parameters, ε t is the error term; S26. Based on the above analysis and simulation results, the central processing unit CPU performs decision support to automatically adjust and optimize supply chain operations: Implement dynamic adjustment strategies to deal with predicted risks and problems. According to the output of the product quality prediction model P(Q|T1,H1), when the predicted quality result of a batch of products is lower than the threshold Q threshold , the system will automatically adjust the relevant links to reduce losses: P(Q|T1,H1) threshold ;​ Using the optimization analysis results, the optimal transportation and storage strategies obtained by automatically replanning transportation routes or adjusting inventory levels will be automatically applied to supply chain operations; Based on the product demand forecast results, when the forecast demand Y t Exceeding the sum of inventory levels plus production capacity, automatically adjust production and supply plans, using demand forecasts obtained from time series analysis Y t , combined with the inventory level L inventory and production capacity C production , calculate the production adjustment D production : D production =Y t -(L inventory +C production ); The S3 specifically includes: S31. Perform real-time analysis and simulation on a virtual copy of the perishable product and its supply chain environment created using digital twin technology on the central processing unit CPU to assess the current status of the perishable product in the supply chain and predict potential risks and problems; S32. A dynamic system model DSM is used to simulate the behavior and state changes of perishable products in the supply chain. The dynamic system model considers the influence of temperature T, humidity H, time t and location GPS factors. The dynamic system model DSM formula is: Among them, S represents the state of perishable products; S33. Use the random forest algorithm to analyze the data collected from IoT devices to identify and predict possible problems in the supply chain. The MLA formula of the random forest algorithm is: P(event|T,H,GPS)=learn(data); Among them, P(event|T,H,GPS) is the probability of a specific event occurring under given temperature T, humidity H and location GPS; S34. Based on the output of the digital twin model and the random forest algorithm, the decision support system DSS in the central processing unit CPU will evaluate the current status of perishable products and generate corresponding risk reports and warnings, which will automatically trigger the warning mechanism. The report format is: Report: {product ID, risk level, recommended action, timestamp}; Among them, the risk level is classified based on the severity of potential risks, and the recommended measures are proposed based on the predicted results and best practices; S35. Using the output of the forecasting model and the decision support system DSS, the central processing unit CPU will dynamically adjust the supply chain management strategy to prevent potential problems and mitigate their impact, adjust transportation routes, change storage conditions or modify supply chain plans; The S4 specifically includes: S41, using the collected temperature T, humidity H, location information GPS, and time t of perishable products to perform data analysis and optimize supply chain management strategies; S42. Use the time series analysis method TSA to predict the demand for perishable products based on historical data: <h2 style=";text-align:left;direction:ltr">D<h2 style=";text-align:left;direction:ltr"> t+1 <h2 style=";text-align:left;direction:ltr"> =αD<h2 style=";text-align:left;direction:ltr"> t <h2 style=";text-align:left;direction:ltr"> +βT<h2 style=";text-align:left;direction:ltr"> t <h2 style=";text-align:left;direction:ltr"> +γH<h2 style=";text-align:left;direction:ltr"> t <h2 style=";text-align:left;direction:ltr"> +∈; Among them, D t+1 represents the forecast demand for the next time period, D t is the demand in the current time period, T t and H t Respectively represent the average temperature and humidity in the corresponding time period, α, β, γ are model parameters, and ∈ is the error term; S43. Use linear programming LP to optimize transportation routes and storage conditions to minimize the total cost C: Among them, c i represents the cost of the i-th transportation route, x i is a decision variable, indicating whether to select (1) or not select (0) the ith transport route, s j represents the cost of the jth storage condition, y j is a decision variable, indicating whether to select (1) or not select (0) the jth storage condition; S44. Apply artificial intelligence algorithms AI, neural networks or deep learning to conduct in-depth analysis of data in the supply chain to identify improvement points and optimization opportunities. The algorithm automatically proposes improvement suggestions by analyzing the relationship between temperature T, humidity H, location GPS data and supply chain efficiency; S44. Use a multi-layer perceptron (MLP) neural network model to conduct in-depth analysis of data in the supply chain, especially temperature T, humidity H, and location GPS. The model structure includes an input layer, at least one hidden layer, and an output layer. The training process aims to minimize the prediction error. Through training, MLP can discover nonlinear relationships in the data and provide data-driven decision support for supply chain management: Where E is the mean error, n is the number of samples, and Y i is the true value of the i-th sample, is the predicted value of the model.

2. The real-time monitoring and management system for perishable product supply chain based on digital twin technology according to claim 1 is characterized in that: The S5 specifically includes: S51. Use the decision support algorithm DSA to automatically generate a supply chain adjustment plan based on the data analysis results. The decision support algorithm DSA formula is: Optimize(SC parameters )→SC adjustments ; Among them, SC parameters Represents a set of key parameters of the supply chain, SC adjustments It is a set of supply chain adjustment measures determined according to the optimization goal; S52. Implement dynamic inventory management (DIM) to adjust inventory levels based on real-time data and demand forecasts to reduce the risk of oversupply and out-of-stock: L new =L current +P incoming -D forecasted ; Among them, L new is the adjusted inventory level, L current is the current inventory level, P incoming is the expected purchase quantity, D forecasted is the predicted demand; S53, optimize logistics and transportation strategy OTS, select the best transportation mode and route according to the characteristics of perishable products and real-time monitoring data. The transportation strategy OTS formula is: Minimize(T cost +T risk ); Among them, T cost is the transportation cost, T risk It is the risk cost of product loss or quality degradation during transportation; S54. Apply advanced scheduling algorithm ASA to automatically adjust production and distribution plans, respond to demand changes and supply chain disruption risks in real time, and maximize customer satisfaction. The scheduling algorithm ASA formula is: Minimize(D delay -C satisfaction ); Among them, D delay is the production or delivery delay, C satisfaction is a cost measure of customer satisfaction, with negative values ​​indicating increased satisfaction.

3. The real-time monitoring and management system for perishable product supply chain based on digital twin technology according to claim 1 is characterized in that: The S6 specifically includes: S61. Provide a user interface to enable supply chain managers and consumers to access real-time status information of perishable products; S62. Implement the data visualization module DVM to dynamically display the real-time data and historical trends of perishable products on the user interface, including temperature T, humidity H, location GPS, and supply chain efficiency indicators. The formula of the data visualization module DVM is: Visualize(T,H,GPS,SC metrics ); Among them, SC metrics represents supply chain efficiency metrics, including shipping times, inventory levels, and order fill rates; S63, integrated feedback mechanism FM, allowing users to provide feedback and evaluation of perishable product status information through the user interface UI. These feedbacks will be automatically recorded and analyzed by the system to further optimize supply chain management and improve user experience. The feedback mechanism FM formula is: CollectFeedback(UserID,Feedback)→SC improvements ; UserID is the user ID that provides feedback, Feedback is the specific feedback content provided by the user, and SC improvements are supply chain improvement measures taken based on user feedback; S64. Deploy the intelligent recommendation system IRS to recommend optimized purchase and consumption suggestions of related products to users based on user behavior and preferences and real-time status information of perishable products. The formula of the intelligent recommendation system IRS is: Recommend(UserID,SC data )→Recommendations; Among them, SC data Including real-time status data and supply chain information of perishable products, Recommendations are based on the behavior and preferences of the user UserID, and SC data Generated recommendation list; S65. Using blockchain technology BCT, by recording and verifying information at each transaction node, consumers can be provided with unalterable proof of product origin and history. The blockchain technology BCT formula is: Secure(SC transactions )→Trust; Among them, SC transactions Representing transaction records in the supply chain, Trust is the consumer trust established by ensuring that data cannot be tampered with.

Citation Information

Cited By

  • Product supply chain real-time monitoring and management system based on digital twin technology

    CN121766704A