Digital twinning-based digital supply chain collaborative scheduling method and system

Abstract

The invention relates to the field of collaborative management scheduling, in particular to a digital supply chain collaborative scheduling method and system based on digital twinning, and the method comprises the steps: calculating the estimated stock of any production area; calculating an influence coefficient of any production area; obtaining a plurality of optional transportation paths of any production area, calculating the transportation risk of any production area in any optional transportation path based on the influence coefficient and the transportation loss rate, and obtaining the optimal transportation path of any production area; and obtaining an estimated supply quantity of any production area according to the estimated stock quantity, calculating a first product of the estimated supply quantity and the normalized transportation risk, taking an accumulated value of the first products of all the production areas as a transportation cost function, responding to a maximum value of the transportation cost function, obtaining an actual supply quantity of any production area, and completing collaborative scheduling. According to the technical scheme of the invention, the efficiency and precision of cooperative scheduling can be improved.

Description

Technical Field

[0001] This invention relates to the field of collaborative management and scheduling. In particular, it relates to a method and system for collaborative scheduling of digital supply chains based on digital twins. Background Technology

[0002] Agricultural products, as a special type of commodity, have a short shelf life, are perishable, and their quality is highly sensitive to environmental conditions such as temperature and humidity. Due to these characteristics, the agricultural product supply chain must have efficient scheduling capabilities to ensure that the entire circulation process from the production site to the consumer can proceed under optimal conditions, avoiding product quality decline or even spoilage due to untimely scheduling or poor coordination.

[0003] In traditional agricultural supply chain management, supply chain coordination and scheduling often rely on manual decision-making and information transmission, which has significant limitations in terms of efficiency and response speed. With globalization and informatization, the complexity of agricultural supply chains is increasing daily. Factors such as diverse product types, fluctuating demand, and uncertain production cycles present unprecedented challenges to supply chain management. Especially in addressing the timeliness and quality assurance of agricultural products, traditional supply chain scheduling methods cannot meet the demands for rapid response and efficient collaboration, leading to problems such as inventory backlog, resource waste, and information asymmetry. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention provides solutions in the following aspects.

[0005] In the first aspect, the digital supply chain collaborative scheduling method based on digital twins includes: segmenting a target historical period to obtain several sub-periods, where the target historical period is any historical period; obtaining the inventory curve of any production area in each sub-period of the target historical period; calculating the estimated inventory of any production area based on the inventory curve; obtaining the transportation loss rate curve of any production area in each sub-period of the target historical period; calculating the impact coefficient of any production area based on the transportation loss rate curve; obtaining several optional transportation routes for any production area; calculating the transportation risk of any production area on any optional transportation route based on the impact coefficient and the transportation loss rate; obtaining the optimal transportation route for any production area based on the transportation risk and the obtained transportation time; obtaining the estimated supply volume of any production area based on the estimated inventory; calculating the first product of the estimated supply volume and the normalized transportation risk; and using the cumulative value of the first product of all production areas as the transportation cost function. The actual supply volume of any production area is obtained when the value of the transportation cost function is maximized, thus completing the collaborative scheduling.

[0006] Preferably, the step of segmenting the target historical time period to obtain several sub-time periods includes: obtaining the demand for each sampling time in the target historical time period, using DBSCAN to cluster the demand and the corresponding sampling time to obtain several clusters, and taking the sampling time within a cluster as a sub-time period.

[0007] Preferably, the step of calculating the estimated inventory level of any production area based on the inventory level curve includes: obtaining the historical corresponding time of the current moment, where the historical corresponding time is a historical moment within any sub-period of the target historical period, and taking the sub-period to which the historical moment belongs as a reference segment; for any production area, calculating the mean inventory level and standard deviation of the inventory level of the reference segment, iterating through the reference segment to obtain the standard deviation of the inventory level of each historical period, and normalizing and summing all the standard deviations of the inventory level, with the result being the sum of the standard deviations; taking the product of the mean inventory level of the target historical period and the normalized standard deviation of the inventory level as the numerator, calculating the ratio of the numerator to the sum of the standard deviations, and using the cumulative value of the ratio of all historical periods as the estimated inventory level.

[0008] Preferably, the step of calculating the impact coefficient of any production area based on the transportation loss rate curve includes: obtaining the historical corresponding time of the current moment, where the historical corresponding time is a historical moment within any sub-period in the target historical period, and taking the sub-period to which the historical moment belongs as a reference segment; for any production area, obtaining the transportation loss rate of each historical moment in the reference segment, and obtaining the maximum value of the transportation loss rate of the reference segment; obtaining the transportation loss rate entropy value of the reference segment in the target historical period, calculating the second product of the transportation loss rate entropy value of the reference segment in the target historical period and the maximum value of the transportation loss rate, and calculating the average of the second products of all historical periods as the impact coefficient.

[0009] Preferably, obtaining the transportation risk includes: obtaining the number of times any optional transportation route is used, and normalizing the number of times all optional transportation routes are used; for the same optional transportation route in the same production area, calculating the mean and standard deviation of the transportation loss rate for all transportation processes, and using the ratio of the mean to the standard deviation as the coefficient of variation; and using the product of the influence coefficient and the coefficient of variation as the transportation risk.

[0010] Preferably, obtaining the transportation risk further includes: calculating the third product of the transportation loss rate and the normalized number of uses in any transportation process of any available transportation route, and calculating the mean of the third product of all transportation processes; and taking the product of the impact coefficient, the coefficient of variation and the mean of the third product as the transportation risk.

[0011] Preferably, obtaining the optimal transportation route for any production area based on transportation risk and acquired transportation time includes: for any optional transportation route within the same production area, obtaining the transportation time for any transportation process; normalizing the transportation times of all transportation processes and calculating the average as the time mean; calculating the mapping value of the product of the time mean and transportation risk through a negative correlation mapping; constructing a judgment function, wherein the value of the judgment function is 0 in response to the first difference between the delivery time and the maximum value of the transportation time of any transportation process being less than 0; and the value of the judgment function is the first difference in response to the first difference between the delivery time and the maximum value of the transportation time of any transportation process being not less than 0; the product of the value of the judgment function and the mapping value is used as the selection probability of any optional transportation route, and the optional transportation route corresponding to the maximum selection probability is taken as the optimal transportation route.

[0012] Secondly, a digital supply chain collaborative scheduling system based on digital twins includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the digital supply chain collaborative scheduling method based on digital twins described in any one of the claims is implemented.

[0013] The present invention has the following effects: This invention optimizes logistics scheduling between production areas by calculating inventory and transportation risks and combining multiple dimensions such as estimated inventory levels, transportation loss rates, and delivery times. It not only accurately predicts inventory levels and supply capacity in each production area but also ensures orders are completed in the shortest possible time with the lowest possible transportation risk by comprehensively assessing the risks and efficiency of transportation routes.

[0014] This invention can avoid potential losses caused by unstable transportation routes or delivery delays, thereby improving the overall operational efficiency and responsiveness of the supply chain. Furthermore, by precisely controlling the matching of transportation risks and supply volume, it can minimize transportation costs and logistics risks while ensuring order demand is met, and improve the efficiency of collaborative scheduling. Attached Figure Description

[0015] Figure 1 This is a flowchart of a digital supply chain collaborative scheduling method based on digital twins, according to an embodiment of the present invention. Detailed Implementation

[0016] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0017] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0018] Reference Figure 1The digital supply chain collaborative scheduling method based on digital twins includes steps S1-S3, as detailed below: S1: Divide the target historical period into several sub-periods. The target historical period is any historical period. Obtain the inventory curve of any production area in each sub-period of the target historical period. Calculate the estimated inventory of any production area based on the inventory curve. Obtain the transportation loss rate curve of any production area in each sub-period of the target historical period. Calculate the impact coefficient of any production area based on the transportation loss rate curve.

[0019] In one embodiment, the process of segmenting a target historical period to obtain several sub-periods is as follows: First, the demand data corresponding to each sampling time point within the historical period is acquired, forming a two-dimensional dataset composed of timestamps and demand quantities. Then, DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is used, with the timestamps and corresponding demand quantities as feature inputs, to perform cluster analysis on these data points. DBSCAN can automatically identify high-density areas and divide them into independent clusters, while effectively eliminating noise points. Finally, all sampling time points belonging to the same cluster are grouped into a continuous or non-continuous sub-period with similar demand characteristics. The target historical period is any historical period. For example, a historical period is one year.

[0020] To obtain the inventory curve for each sub-period of a given production area within a target historical period and to calculate its estimated inventory, the specific method is as follows: First, based on the segmentation results of the target historical period (i.e., several sub-periods), the inventory data for each sampling moment within each sub-period of the production area is extracted to form the corresponding inventory time series, i.e., the inventory curve. Next, the corresponding moment in history is determined, i.e., the historical moment with the same time attribute as the current moment within the target historical period. For example, the current moment is April, and the corresponding historical moment is April of the target historical period. The sub-period to which the historical moment belongs is used as a reference segment. For this reference segment, the mean and standard deviation of the inventory within the corresponding reference segment in all historical years are calculated. Subsequently, the standard deviation of the reference segment in all historical periods is normalized, and the sum of the standard deviations is obtained. Furthermore, the average inventory level of the reference period in the target historical period is multiplied by the normalized standard deviation to obtain the numerator. The numerator is then divided by the sum of the standard deviations to obtain a weighted ratio. Finally, such ratios calculated in all historical periods are summed up, and the sum is the estimated inventory level of the production area at the current moment.

[0021] Thus, the estimated inventory level can be obtained, and the estimated supply level for the production area can be determined based on the estimated inventory level. That is, the estimated supply level ranges from 0 to the estimated inventory level.

[0022] To calculate the impact coefficient of transportation loss rate for any production area, it is first necessary to obtain the historical time corresponding to the target period from historical data, and then divide the historical time into several sub-periods as reference segments. Next, the transportation loss rate for each historical time in each reference segment is calculated, and the maximum transportation loss rate value for that segment is found. Then, based on the distribution of the transportation loss rate in the reference segment, its entropy value is calculated; the entropy value reflects the degree of change in the transportation loss rate of that segment. The second product of the entropy value of the transportation loss rate in the target historical period and the maximum transportation loss rate value reveals the degree of impact of the transportation loss rate at the current moment. By calculating the average of the second products of all historical periods, the impact coefficient of the current production area can be obtained, thus quantifying the transportation loss rate at the current moment.

[0023] It should be noted that when calculating inventory levels, the impact of product maturity in the production area on the transportation loss rate must also be considered. Products with higher maturity typically have higher transportation loss rates because they are more prone to damage or other losses during transportation. Since the maturity of products produced in different production areas varies at different times, it is necessary to combine historical data from the production area and segment the historical transportation loss rate curve according to maturity stages. The mean of each segment reflects the transportation loss rate level at that maturity stage. Based on the trend of transportation loss rate changes between the current moment and the corresponding historical period, the influence coefficient of the current product maturity in the production area on the transportation loss rate during transportation is analyzed and quantified.

[0024] S2: Obtain several optional transportation routes for any production area, calculate the transportation risk of any production area on any optional transportation route based on the impact coefficient and transportation loss rate, and obtain the optimal transportation route for any production area based on the transportation risk and the obtained transportation time.

[0025] It is important to note that when selecting the optimal transportation route for each production area, a comprehensive balance must be struck between maximizing revenue and minimizing losses. The transportation loss rate is not only affected by product maturity but also significantly depends on road conditions. Worse or more volatile road conditions often lead to higher damage rates. Historical transportation loss rate data can indirectly reflect the actual road condition performance of each route at different times. Therefore, route evaluation should not only calculate the historical average transportation loss rate for each available route but also analyze its standard deviation or entropy value to measure road condition stability: higher stability, i.e., smaller fluctuations in the transportation loss rate, indicates stronger predictability and reliability of the route. Simultaneously, transportation time is also a key constraint, especially in scenarios with tight delivery windows. Routes with shorter travel times must be prioritized to avoid defaults or market value losses due to delays.

[0026] In one embodiment, it is necessary to obtain the usage frequency of each transportation route and normalize the usage frequency of all transportation routes to eliminate the impact of differences in usage frequency between different routes on risk assessment. Next, for each transportation route within the same production area, the mean and standard deviation of the transportation loss rate generated during all transportation processes are calculated. The mean transportation loss rate reflects the average level of loss during transportation, while the standard deviation reflects the volatility of the loss level. To further analyze the variability of transportation risk, the ratio of the mean transportation loss rate to the standard deviation of the transportation loss rate is used as the coefficient of variation. A higher coefficient of variation indicates greater volatility in the transportation risk of that route, and thus a higher risk. Finally, the transportation risk of each route is obtained by multiplying the route's impact coefficient by the coefficient of variation. This transportation risk comprehensively considers the route's usage frequency, the stability of losses, and their volatility, thereby providing a basis for subsequent risk assessment and decision-making.

[0027] In another embodiment, the third product of the transportation loss rate and the normalized number of uses in any transportation process on any available transportation route is calculated, and the mean of the third product of all transportation processes is calculated; the product of the impact coefficient, the coefficient of variation, and the mean of the third product is taken as the transportation risk.

[0028] This embodiment takes into account the number of times the optional transportation route is used. The more times it is used, the greater the probability that the optional transportation route will be selected in the future.

[0029] It's important to note that after calculating the transportation risks of all available transportation routes for any production area, the next step is to comprehensively consider delivery time requirements to select the optimal route. To this end, delivery time is introduced as a constraint, and penalties are applied to transportation routes based on the degree to which delivery time is exceeded. Specifically, routes exceeding delivery time will incur an additional penalty term, the intensity of which can be adjusted according to the extent of the delay. Then, transportation risk, delivery time, and the corresponding penalty term are combined to establish an objective function aimed at maximizing the overall benefits of the transportation process. Optimizing this objective function will consider factors such as the risk, transportation cost, delivery time, and delay penalties for each route, thereby ensuring that the transportation route with the lowest risk and highest benefit is selected while meeting delivery time requirements.

[0030] Specifically, for any available transportation route within the same production area, the transportation time for any transportation process is obtained. The transportation times for all transportation processes are normalized and averaged to obtain the time mean. The product of the time mean and the transportation risk is calculated using a negative correlation mapping. A judgment function is constructed: the value of the judgment function is 0 if the first difference between the delivery time and the maximum value of the transportation time for any transportation process is less than 0; the value of the judgment function is the first difference if the first difference between the delivery time and the maximum value of the transportation time for any transportation process is not less than 0. The product of the judgment function value and the mapping value is used as the selection probability of any available transportation route, and the available transportation route corresponding to the maximum selection probability is taken as the optimal transportation route.

[0031] It should be explained that if the first difference between the delivery time and the maximum transportation time of any transportation process is less than 0, it means that the delivery time has been exceeded, and the probability of choosing this transportation option is 0. If the first difference between the delivery time and the maximum transportation time of any transportation process is not less than 0, it means that the required transportation time is within the delivery time range. In this case, the transportation option with the lowest transportation risk can be selected while meeting the delivery time requirement. This ultimately determines the optimal transportation route for any production area.

[0032] S3: Based on the estimated inventory, obtain the estimated supply volume of any production area, calculate the first product of the estimated supply volume and the normalized transportation risk, and use the cumulative value of the first product of all production areas as the transportation cost function. When the value of the transportation cost function is maximized, obtain the actual supply volume of any production area and complete the collaborative scheduling.

[0033] In one embodiment, the transportation cost function satisfies the following relationship: , Represents the transportation cost function. Indicates production area The transportation risk of the optimal transportation route Indicates production area The estimated supply volume Indicates the total number of production areas. This indicates the quantity of orders required.

[0034] This restricts the sum of estimated supply from all production areas to be no less than the order demand, meaning orders must be fulfilled. When the transportation cost function is maximized, the required quantity of goods from each production area can be determined, enabling coordinated scheduling.

[0035] The system includes a processor and a memory, the memory storing computer program instructions, which, when executed by the processor, implement the digital supply chain collaborative scheduling method based on digital twins according to the first aspect of the present invention.

[0036] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0037] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A digital supply chain collaborative scheduling method based on digital twins, characterized in that, include: The target historical period is divided into segments to obtain several sub-periods. The target historical period is any historical period. The inventory curve of any production area in each sub-period of the target historical period is obtained. The estimated inventory of any production area is calculated based on the inventory curve. The transportation loss rate curve of any production area in each sub-period of the target historical period is obtained. The impact coefficient of any production area is calculated based on the transportation loss rate curve. Obtain several possible transportation routes for any production area, calculate the transportation risk of any production area on any possible transportation route based on the impact coefficient and transportation loss rate, and obtain the optimal transportation route for any production area based on the transportation risk and the obtained transportation time. Based on the estimated inventory level, obtain the estimated supply volume of any production area, calculate the first product of the estimated supply volume and the normalized transportation risk, and use the sum of the first products of all production areas as the transportation cost function. When the value of the transportation cost function is maximized, obtain the actual supply volume of any production area and complete the coordinated scheduling.

2. The digital supply chain collaborative scheduling method based on digital twins according to claim 1, characterized in that, The process of segmenting the target historical period to obtain several sub-periods includes: Obtain the demand for each sampling time in the target historical period, use DBSCAN to cluster the demand and the corresponding sampling time to obtain several clusters, and take the sampling time in a cluster as a sub-period.

3. The digital supply chain collaborative scheduling method based on digital twins according to claim 1, characterized in that, The calculation of the estimated inventory level for any production area based on the inventory level curve includes: Obtain the historical corresponding time of the current moment. The historical corresponding time is the historical time within any sub-period of the target historical period. The sub-period to which the historical time belongs is used as the reference segment. For any production area, calculate the mean and standard deviation of inventory in the reference segment, iterate through the reference segment to obtain the standard deviation of inventory in each historical period, and normalize and sum all the standard deviations of inventory, and use the result as the sum of standard deviations. The product of the mean inventory level for the target historical period and the normalized standard deviation of the inventory level is used as the numerator. The ratio of the numerator to the sum of the standard deviations is calculated, and the sum of the ratios for all historical periods is used as the estimated inventory level.

4. The digital supply chain collaborative scheduling method based on digital twins according to claim 1, characterized in that, The calculation of the impact coefficient of any production area based on the transportation loss rate curve includes: Obtain the historical corresponding time of the current moment. The historical corresponding time is the historical time within any sub-period of the target historical period. The sub-period to which the historical time belongs is used as the reference segment. For any production area, obtain the transportation loss rate at each historical moment in the reference segment, and obtain the maximum value of the transportation loss rate of the reference segment; Obtain the entropy value of the transportation loss rate of the reference segment in the target historical period, calculate the second product of the entropy value of the transportation loss rate of the reference segment in the target historical period and the maximum value of the transportation loss rate, and calculate the average of the second products of all historical periods as the influence coefficient.

5. The digital supply chain collaborative scheduling method based on digital twins according to claim 1, characterized in that, The acquisition of the aforementioned transportation risks includes: Get the usage count of any available transportation route and normalize the usage count of all available transportation routes; For the same transportation route within the same production area, calculate the mean and standard deviation of the transportation loss rate for all transportation processes, and use the ratio of the mean to the standard deviation as the coefficient of variation. The product of the impact coefficient and the coefficient of variation is used as the transportation risk.

6. The digital supply chain collaborative scheduling method based on digital twins according to claim 5, characterized in that, Acquiring the aforementioned transportation risks also includes: Calculate the third product of the transportation loss rate and the normalized number of uses in any transportation process for any available transportation route, and calculate the mean of the third product for all transportation processes; The product of the impact coefficient, the coefficient of variation, and the mean of the third product is used as the transportation risk.

7. The digital supply chain collaborative scheduling method based on digital twins according to claim 1, characterized in that, The process of determining the optimal transportation route for any production area based on transportation risk and available transportation time includes: For any available transportation route within the same production area, obtain the transportation time for any transportation process, normalize the transportation time of all transportation processes, and calculate the average as the time mean; calculate the mapping value of the product of the time mean and transportation risk through negative correlation mapping. Construct a judgment function. The value of the judgment function is 0 if the first difference between the delivery time and the maximum value of the transportation time of any transportation process is less than 0; the value of the judgment function is the first difference if the first difference between the delivery time and the maximum value of the transportation time of any transportation process is not less than 0. The product of the judgment function value and the mapping value is used as the selection probability of any transportation option path, and the transportation option path corresponding to the maximum selection probability is taken as the optimal transportation path.

8. A digital supply chain collaborative scheduling system based on digital twins, characterized in that: include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the digital supply chain collaborative scheduling method based on digital twins according to any one of claims 1-7.

Images