Supply chain multi-objective decision-making method and system for medical industry

Abstract

The invention discloses a supply chain multi-objective decision-making method and system for the medical industry, and relates to the technical field of medical material management. The method comprises the following steps: firstly, according to single-day demand quantity time sequence data of a target drug, estimating to obtain a single-day demand quantity predicted value of the drug in each day in a current next purchasing cycle; and then, according to a prediction result, constructing a multi-objective function comprising a total cost minimization objective function, a stock turnover rate maximization objective function, a close period proportion minimization objective function and a capacity utilization rate maximization objective function, and then, applying an NSGA-II algorithm to solve a multi-objective optimization problem established based on the multi-objective function. And finally, selecting any Pareto optimal solution conforming to the preset storage strategy from the solution set, and pushing the Pareto optimal solution as an optimal drug inventory scheme to the downstream side of the supply chain for implementation, so that the drug inventory management capability and efficiency of the downstream side of the supply chain in the medical industry can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical supplies management, and particularly relates to a multi-objective decision-making method and system for the supply chain in the medical industry. Background Art

[0002] Multi-objective decision-making in the supply chain refers to the situation in supply chain management where decision-makers need to make choices among multiple conflicting objectives to optimize the operation of the supply chain. These objectives may include improving the overall efficiency of the supply chain, reducing costs, enhancing service quality, delivering products on time, and lowering inventory levels. Since the supply chain involves multiple links and stakeholders, the decision-making process needs to consider the balance between long-term and short-term goals, the needs and interests of different stakeholders, the dynamics and uncertainties of the supply chain, and the factors of sustainable development.

[0003] Currently, drug inventory management is one of the important links in multi-objective decision-making for the medical industry supply chain. That is, on the downstream side of the medical industry supply chain such as hospitals or clinics, since it is necessary to ensure the sufficiency of key drugs while also ensuring that costs and waste are not increased due to excessive inventory, especially for those drugs with expiration dates, it is necessary to find a balance between inventory costs and stock-out risks, and possibly also consider inventory turnover rate, the proportion of drugs approaching the expiration date, and inventory capacity utilization rate. That is, how to provide a new multi-objective decision-making scheme for the medical industry supply chain to achieve multi-objective balance among minimizing inventory costs of drugs, minimizing stock-out risks, maximizing inventory turnover rate, minimizing the proportion of drugs approaching the expiration date, and maximizing capacity utilization rate, and improving the drug inventory management ability and efficiency on the downstream side of the medical industry supply chain is an urgent research topic for those skilled in the art. Summary of the Invention

[0004] The objective of the present invention is to provide a multi-objective decision-making method, system, computer-readable storage medium, and computer program product for the medical industry supply chain, so as to solve the problem that the existing drug inventory management solutions are limited in drug inventory management ability and efficiency on the downstream side of the medical industry supply chain due to the lack of multi-objective balancing means.

[0005] To achieve the above objective, the present invention adopts the following technical solutions:

[0006] In the first aspect, a multi-objective decision-making method for the medical industry supply chain is provided, including:

[0007] Obtaining the time-series data of the daily demand of the target drug on the downstream side of the medical industry supply chain, where the time-series data of the daily demand includes the most recent consecutive historical days in sequence and the daily demand of the target drug on each of the most recent consecutive historical days on the downstream side of the medical industry supply chain;

[0008] Based on the time series data of the daily demand for the target drug, the predicted value of the daily demand for the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain is estimated;

[0009] According to the predicted values of the daily demand for the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain, the following multi-objective function is constructed: the total cost minimization objective function f1(x), the inventory turnover rate maximization objective function f2(x), the near-expiry ratio minimization objective function f3(x), and the capacity utilization rate maximization objective function f4(x), where x represents the drug inventory plan for the target drug in the current next procurement cycle and on the downstream side of the medical industry supply chain and is represented by a Y-dimensional vector, and the y-th element c in the Y-dimensional vector y represents the storage quantity of the y-th production batch of the target drug, y represents a positive integer less than or equal to Y, and Y represents the total number of production batches of the target drug;

[0010] Apply the NSGA-II algorithm to solve the multi-objective optimization problem established based on the multi-objective function to obtain the Pareto front solution set, where the Pareto front solution set contains at least one Pareto optimal solution for the drug inventory plan x;

[0011] Select any Pareto optimal solution that meets the preset storage strategy from the Pareto front solution set, and push the any Pareto optimal solution as the optimal drug inventory plan to the downstream side of the medical industry supply chain for implementation.

[0012] Based on the above invention content, a new multi-objective decision-making scheme for the supply chain applicable to drug inventory management is provided, that is, first, based on the time series data of the daily demand for the target drug, the predicted value of the daily demand for the drug on each day in the current next procurement cycle is estimated, and then a multi-objective function including the total cost minimization objective function, the inventory turnover rate maximization objective function, the near-expiry ratio minimization objective function, and the capacity utilization rate maximization objective function is constructed according to the prediction results. Then, the NSGA-II algorithm is applied to solve the multi-objective optimization problem established based on the multi-objective function to obtain the Pareto front solution set. Finally, any Pareto optimal solution that meets the preset storage strategy is selected from the solution set and used as the optimal drug inventory plan to be pushed to the downstream side of the supply chain for implementation. In this way, a multi-objective balance can be achieved among minimizing the inventory cost of drugs, minimizing the out-of-stock risk, maximizing the inventory turnover rate, minimizing the near-expiry ratio, and maximizing the capacity utilization rate, improving the drug inventory management ability and efficiency on the downstream side of the medical industry supply chain, and facilitating practical application and promotion.

[0013] In a possible design, based on the time series data of the daily demand volume of the target drug, the predicted values of the daily demand volume of the target drug for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain are estimated, including:

[0014] Use the STL algorithm to decompose the time series data of the daily demand volume of the target drug to obtain the time series data of the daily demand trend item component of the target drug. Among them, the time series data of the daily demand trend item component includes the most recent historical days and the daily demand trend item components of the target drug for each day in the most recent historical days and on the downstream side of the medical industry supply chain;

[0015] Based on the time series data of the daily demand trend item component, calculate the predicted value of the first daily demand trend item component of the target drug for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain by using the dynamic exponential weighting method. Among them, the dynamic exponential weighting method assigns weights to the daily demand trend item components in different periods according to the following rules: assign larger weights to the daily demand trend item components in later periods and smaller weights to the daily demand trend item components in earlier periods;

[0016] Based on the time series data of the daily demand trend item component, train a prediction model based on the Prophet model, and use the trained prediction model to predict the predicted value of the second daily demand trend item component, the predicted value of the daily demand seasonal effect item component, the predicted value of the daily demand holiday effect item component, and the predicted value of the daily demand error item component of the target drug for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain;

[0017] For each day in the current next procurement cycle, calculate the predicted value sp1 of the first daily demand volume of the target drug for the corresponding day and on the downstream side of the medical industry supply chain according to the following formula:

[0018] sp1 = w s × g sp,1 + w p × g sp,2 + s prophet + h prophet + ε prophet

[0019] In the formula, g sp,1 represents the predicted value of the first daily demand trend item component, g sp,2 represents the predicted value of the second daily demand trend item component, s prophet represents the predicted value of the daily demand seasonal effect item component, h prophet represents the predicted value of the daily demand holiday effect item component, εprophet represents the predicted value of the single-day demand error term component, w s and w p respectively represent preset weight coefficients and w s +w p = 1.

[0020] In a possible design, when the current most recent multi-day period is all days of the current most recent consecutive multi-weeks in time sequence and the current next procurement cycle is all days of the current next week, according to the single-day demand trend term component time series data, based on the dynamic exponential weighting method, the predicted value of the first single-day demand trend term component of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain is calculated, including:

[0021] Divide the single-day demand trend term component time series data into multiple arrays corresponding one by one to the current most recent multi-weeks, and aggregate the multiple arrays to obtain the following matrix G:

[0022]

[0023] In the formula, n represents the total number of weeks of the current most recent multi-weeks, n' represents a positive integer less than or equal to n, m' represents a positive integer less than or equal to 7, and g n′,m′ represents the single-day demand trend term component of the target drug on the m'-th day in the n'-th week of the current most recent multi-weeks and on the downstream side of the medical industry supply chain;

[0024] Use the following recurrence formula to calculate the predicted value of the first single-day demand trend term component of the target drug in each day of the current next week and on the downstream side of the medical industry supply chain:

[0025]

[0026] In the formula, k represents a positive integer less than or equal to n, When k is less than n, represents the predicted value of the first single-day demand trend term component of the target drug on the m'-th day in the (k + 1)-th week of the current most recent multi-weeks and on the downstream side of the medical industry supply chain. When k is equal to n, represents the predicted value of the first single-day demand trend term component of the target drug on the m'-th day in the current next week and on the downstream side of the medical industry supply chain. α represents a preset smoothing factor and α ∈ (0, 1).

[0027] In a possible design, the total cost in the minimum total cost objective function f1(x) is equal to the sum of the procurement cost, storage cost, expiration loss cost, and shortage loss cost. Among them, the procurement cost is related to the y-th element c yand related to the remaining shelf life of the current drugs in the y-th production batch, the storage cost is related to the y-th element c y and related to the predicted daily demand values of the target drug for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain, the expired loss cost is related to the y-th element c y related to the y-th production batch of the remaining shelf life of the current drugs, and the predicted daily demand values of the target drug for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain, the shortage loss cost is related to the y-th element c y related to the y-th production batch of the remaining shelf life of the current drugs, and the predicted daily demand values of the target drug for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain;

[0028] And / or, the inventory turnover rate in the maximized inventory turnover rate objective function f2(x) is equal to the sum of the predicted daily demand values of the target drug for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain divided by

[0029] And / or, the near-expiry ratio in the minimized near-expiry ratio objective function f3(x) is equal to the storage quantity of the target drug with the remaining shelf life of the current drug less than the preset duration in the drug inventory plan divided by

[0030] And / or, the capacity utilization rate in the maximized capacity utilization rate objective function f4(x) is equal to divided by the maximum allowable storage quantity of the target drug on the downstream side of the medical industry supply chain.

[0031] In a possible design, selecting any Pareto optimal solution that meets the preset storage strategy from the Pareto front solution set includes:

[0032] For each Pareto optimal solution in the Pareto front solution set, determine the corresponding total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate according to the corresponding solution and the predicted daily demand values of the target drug for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain;

[0033] For each of the Pareto optimal solutions, calculate the corresponding total cost normalization value, inventory turnover rate normalization value, near-expiry ratio normalization value, and capacity utilization rate normalization value according to the corresponding total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate, and the total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate of all the Pareto optimal solutions, and calculate the corresponding recommended index value according to the following formula:

[0034] q z =-w1×q z,1 +w2×q z,2 -w3×q z,3 +w4×q z,4

[0035] Wherein, z represents a positive integer, and q z represents the recommended index value of the z-th Pareto optimal solution in the Pareto front solution set, and q z,1 represents the total cost normalization value corresponding to the z-th Pareto optimal solution, and q z,2 represents the inventory turnover rate normalization value corresponding to the z-th Pareto optimal solution, and q z,3 represents the near-expiry ratio normalization value corresponding to the z-th Pareto optimal solution, and q z,4 represents the capacity utilization rate normalization value corresponding to the z-th Pareto optimal solution. w1, w2, w3, and w4 respectively represent preset weight coefficients and w1 + w2 + w3 + w4 = 1;

[0036] Select a certain Pareto optimal solution with the maximum recommended index value from the Pareto front solution set as any Pareto optimal solution that meets the preset storage strategy.

[0037] In a possible design, for each of the Pareto optimal solutions, according to the corresponding total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate, as well as the total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate of all the Pareto optimal solutions, calculate the corresponding total cost normalization value, inventory turnover rate normalization value, near-expiry ratio normalization value, and capacity utilization rate normalization value, including:

[0038] For each of the Pareto optimal solutions, according to the corresponding total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate, as well as the total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate of all the Pareto optimal solutions, calculate the corresponding total cost normalization value, inventory turnover rate normalization value, near-expiry ratio normalization value, and capacity utilization rate normalization value according to the following formula:

[0039]

[0040] Wherein, z represents a positive integer, and q z,0 represents the calculated normalization value of the z-th Pareto optimal solution in the Pareto front solution set, and Q z,0 represents the original value before calculation of the z-th Pareto optimal solution, and Q min,0 represents the minimum value among the original values before calculation of all the Pareto optimal solutions, and Q max,0represents the maximum value among the original values before the calculation of all the Pareto optimal solutions. The calculated normalized value refers to the total cost normalized value, the inventory turnover rate normalized value, the proportion of near-expiry products normalized value, or the capacity utilization rate normalized value. The original value before the calculation refers to the total cost, the inventory turnover rate, the proportion of near-expiry products, or the capacity utilization rate, and is consistent with the calculated normalized value.

[0041] In a second aspect, a multi-objective decision-making system for the medical industry supply chain is provided, including a time-series data acquisition unit, a drug demand forecasting unit, an objective function construction unit, an optimization problem solving unit, and an optimal solution pushing unit that are sequentially communicatively connected;

[0042] The time-series data acquisition unit is used to acquire the time-series data of the single-day demand quantity of the target drug on the downstream side of the medical industry supply chain. Among them, the time-series data of the single-day demand quantity includes the current recent multiple consecutive days in time sequence and the single-day demand quantity of the target drug on each of the current recent multiple days on the downstream side of the medical industry supply chain;

[0043] The drug demand forecasting unit is used to estimate the predicted value of the single-day demand quantity of the target drug on each day in the current next procurement cycle on the downstream side of the medical industry supply chain according to the time-series data of the single-day demand quantity of the target drug;

[0044] The objective function construction unit is used to construct the following multi-objective function according to the predicted value of the single-day demand quantity of the target drug on each day in the current next procurement cycle on the downstream side of the medical industry supply chain: the total cost minimization objective function f1(x), the inventory turnover rate maximization objective function f2(x), the proportion of near-expiry products minimization objective function f3(x), and the capacity utilization rate maximization objective function f4(x). Among them, x represents the drug inventory plan of the target drug in the current next procurement cycle on the downstream side of the medical industry supply chain and is represented by a Y-dimensional vector. The y-th element c in the Y-dimensional vector y represents the storage quantity of the y-th production batch of the target drug, y represents a positive integer less than or equal to Y, and Y represents the total number of production batches of the target drug;

[0045] The optimization problem solving unit is used to apply the NSGA-II algorithm to solve the multi-objective optimization problem established based on the multi-objective function to obtain the Pareto front solution set. Among them, the Pareto front solution set includes at least one Pareto optimal solution for the drug inventory plan x;

[0046] The optimal solution pushing unit is configured to select any Pareto optimal solution that conforms to a preset storage strategy from the Pareto front solution set, and push the any Pareto optimal solution as an optimal drug inventory plan to the downstream side of the medical industry supply chain for implementation.

[0047] In a third aspect, the present invention provides a computer system, which includes a memory, a processor, and a transceiver that are communicatively connected in sequence. Among them, the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the supply chain multi-objective decision-making method as described in the first aspect or any possible design in the first aspect.

[0048] In a fourth aspect, the present invention provides a computer-readable storage medium, on which instructions are stored. When the instructions are run on a computer, the supply chain multi-objective decision-making method as described in the first aspect or any possible design in the first aspect is executed.

[0049] In a fifth aspect, the present invention provides a computer program product, including a computer program or instructions. When the computer program or the instructions are executed by a computer, the supply chain multi-objective decision-making method as described in the first aspect or any possible design in the first aspect is implemented.

[0050] Beneficial effects of the above solution:

[0051] (1) The present invention provides a new supply chain multi-objective decision-making solution applicable to drug inventory management. First, based on the time series data of the daily demand of the target drug, the predicted value of the daily demand for each day in the current next procurement cycle of the drug is estimated. Then, a multi-objective function including an objective function of minimizing the total cost, an objective function of maximizing the inventory turnover rate, an objective function of minimizing the proportion of near-expiry drugs, and an objective function of maximizing the capacity utilization rate is constructed. Then, the NSGA-II algorithm is applied to solve the multi-objective optimization problem established based on the multi-objective function to obtain a Pareto front solution set. Finally, any Pareto optimal solution that conforms to the preset storage strategy is selected from the solution set and pushed as the optimal drug inventory plan to the downstream side of the supply chain for implementation. In this way, a multi-objective balance can be achieved among minimizing the inventory cost of drugs, minimizing the out-of-stock risk, maximizing the inventory turnover rate, minimizing the proportion of near-expiry drugs, and maximizing the capacity utilization rate, improving the drug inventory management ability and efficiency on the downstream side of the medical industry supply chain, and facilitating practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0053] Figure 1 It is a schematic flowchart of the multi-objective decision-making method for the supply chain in the medical industry provided by the embodiments of the present application.

[0054] Figure 2 It is a schematic structural diagram of the multi-objective decision-making system for the supply chain in the medical industry provided by the embodiments of the present application.

[0055] Figure 3 It is a schematic structural diagram of the computer system provided by the embodiments of the present application. Detailed implementation manners

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the present invention in combination with the accompanying drawings and the description of the embodiments or the prior art. Obviously, the following description of the structure of the accompanying drawings is only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other embodiments can be obtained based on these embodiments. It should be noted here that the description of these embodiment modes is used to help understand the present invention, but does not constitute a limitation to the present invention.

[0057] It should be understood that although terms such as first and second etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, the first object can be called the second object, and similarly, the second object can be called the first object, without departing from the scope of the exemplary embodiments of the present invention.

[0058] It should be understood that for the term "and / or" that may appear in this article, it is only a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, B exists alone, or A and B exist simultaneously, etc.; another example, A, B and / or C can represent any one of A, B and C or any combination of them; for the term " / and" that may appear in this article, it is a description of another association object relationship, indicating that two relationships can exist. For example, A / and B can represent: A exists alone or A and B exist simultaneously, etc.; in addition, for the character " / " that may appear in this article, generally, the front and back associated objects are in an "or" relationship.

[0059] Embodiment

[0060] As Figure 1 shown, the multi-objective decision-making method for the supply chain provided in the first aspect of this embodiment and used in the medical industry can be, but is not limited to, executed by a computer device with certain computing resources, such as a management server, a platform server, a personal computer (Personal Computer, PC, referring to a multi-purpose computer suitable for personal use in terms of size, price, and performance; desktop computers, laptops, small laptops, tablet computers, and ultrabooks, etc. all belong to personal computers), a smart phone, a personal digital assistant (Personal Digital Assistant, PDA), or a wearable device and other electronic devices. As Figure 1 shown, the multi-objective decision-making method for the supply chain can be, but is not limited to, including the following steps S1 to S5.

[0061] S1. Obtain the time-series data of the single-day demand of the target drug on the downstream side of the medical industry supply chain. Among them, the time-series data of the single-day demand includes, but is not limited to, the current recent consecutive multiple days in time sequence and the single-day demand of the target drug on each of these days on the downstream side of the medical industry supply chain.

[0062] In the step S1, the downstream side of the medical industry supply chain can be, but is not limited to, a drug-consuming unit such as a hospital or a clinic. The target drug is specifically, but not limited to, an emergency drug or an anti-cancer drug with high clinical importance, high sensitivity to the expiration date (for example, the expiration date is less than 6 months), and high single-piece storage cost (for example, the single-piece single-day refrigeration cost is greater than 5 yuan). Their corresponding time-series data of the single-day demand can be obtained through regular statistics based on the historical drug consumption data.

[0063] S2. Estimate the predicted value of the single-day demand of the target drug on each day in the current next procurement cycle on the downstream side of the medical industry supply chain according to the time-series data of the single-day demand of the target drug.

[0064] In the step S2, the procurement cycle can be, but is not limited to, in units of weeks, months, or quarters. For example, if the current recent consecutive multiple days are from January 6th to February 16th (that is, 6 consecutive weeks), then the current next procurement cycle can be from February 17th to February 23rd (that is, the next week after the 6 consecutive weeks). In order to improve the prediction accuracy of the single-day demand of the target drug in the future, preferably, according to the time-series data of the single-day demand of the target drug, estimating the predicted value of the single-day demand of the target drug on each day in the current next procurement cycle on the downstream side of the medical industry supply chain includes, but is not limited to, the following steps S21 to S24.

[0065] S21. Decompose the time series data of the single - day demand of the target drug using the STL algorithm to obtain the time series data of the single - day demand trend component of the target drug. Among them, the time series data of the single - day demand trend component includes the recent multiple days of history and the single - day demand trend component of each day of the target drug in the recent multiple days of history on the downstream side of the medical industry supply chain.

[0066] In step S21, the STL (Seasonal and Trend decomposition using LOESS) algorithm is a non - parametric time - series decomposition method based on locally weighted scatterplot smoothing (LOESS). It can decompose time - series data into three parts: trend, seasonality, and residuals. The STL algorithm is flexible and robust, and can adapt to the characteristics of various time series, including complex seasonal patterns and non - linear trends. Its core idea is to decompose the time series y stl (t) into a trend g stl (t), a seasonality s stl (t), and a residual ∈ t and other three parts:

[0067] y stl (t)=g stl (t)+s stl (t)+∈ t

[0068] The STL algorithm mainly uses LOESS to smooth the time series to extract the trend component. Among them, LOESS is a local regression method that fits a smooth curve by performing weighted regression on the data points near each time point, without relying on a pre - assumed trend model or seasonal model. Therefore, it can process various types of time - series data, including non - linear trends and complex seasonal patterns, etc. Specifically, using the STL algorithm to decompose the time series data of the single - day demand of the target drug to obtain the time series data of the single - day demand trend component of the target drug includes, but is not limited to, the following steps S211 - S214.

[0069] S211. Estimate and remove the seasonal component in the time series data y stl (t) of the single - day demand of the target drug based on the loop process in the STL algorithm to obtain new time - series data x stl (t), where t represents the time node.

[0070] In the step S211, the loop process in the STL algorithm is a prior art means and will not be elaborated herein. In addition, if the seasonal component is represented by s stl (t), then the new time series data x stl (t) = y stl (t) - s stl (t).

[0071] S212. Based on the cubic weight function W i (t), perform locally weighted least squares regression on the new time series data x stl (t) to obtain the following local quadratic polynomial z(t) by fitting:

[0072] z(t) = β0 + β1×(i - t) + β2×(i - t) 2

[0073] In the formula, β0, β1, and β2 respectively represent coefficients and are obtained through the following minimization formula:

[0074]

[0075] In the formula, d represents a preset smoothing parameter, and i ∈ [t - d, t + d].

[0076] In the step S212, W i (t) is the regression weight. The smoothing parameter d determines the width of the regression window and can be set as a positive integer in days. In addition, the specific solution process of the minimization formula is a prior mathematical process and will not be elaborated herein.

[0077] S213. For each day in the current recent historical multi - day period, use the value on the local quadratic polynomial z(t) at the corresponding time node t as the single - day demand trend item component of the target drug on the corresponding day and on the downstream side of the medical industry supply chain.

[0078] S214. Aggregate the single - day demand trend item components of the target drug on each day in the current recent historical multi - day period and on the downstream side of the medical industry supply chain to obtain the single - day demand trend item component time series data of the target drug.

[0079] S22. Based on the single - day demand trend item component time series data, calculate the predicted value of the first single - day demand trend item component of the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain using the dynamic exponential weighting method. Among them, the dynamic exponential weighting method assigns weights to the single - day demand trend item components in different periods according to the following rules: assign larger weights to the single - day demand trend item components in later periods and smaller weights to the single - day demand trend item components in earlier periods.

[0080] In step S22, specifically, when the current recent historical multi - day period is all the days of the current recent multi - weeks that are sequentially continuous in time series, and the current next procurement cycle is all the days of the current next week, based on the single - day demand trend item component time - series data, the first single - day demand trend item component prediction values of the target drug for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain are calculated based on the dynamic exponential weighting method, including but not limited to the following steps S221 - S222.

[0081] S221. Divide the single - day demand trend item component time - series data into multiple arrays corresponding one - to - one with the current recent multi - weeks, and aggregate the multiple arrays to obtain the following matrix G:

[0082]

[0083] In the formula, n represents the total number of weeks of the current recent multi - weeks, n′ represents a positive integer less than or equal to n, m′ represents a positive integer less than or equal to 7, and g n′,m′ represents the single - day demand trend item component of any target drug on the m′ - th day in the n′ - th week of the current recent multi - weeks and on the downstream side of the medical industry supply chain.

[0084] S222. Use the following recurrence formula to calculate the first single - day demand trend item component prediction values of the target drug for each day in the current next week and on the downstream side of the medical industry supply chain:

[0085]

[0086] In the formula, k represents a positive integer less than or equal to n, When k is less than n, represents the first single - day demand trend item component prediction value of the target drug on the m′ - th day in the (k + 1) - th week of the current recent multi - weeks and on the downstream side of the medical industry supply chain. When k is equal to n, represents the first single - day demand trend item component prediction value of the target drug on the m′ - th day in the current next week and on the downstream side of the medical industry supply chain. α represents a preset smoothing factor and α ∈ (0, 1).

[0087] In step S222, the above recurrence formula conforms to the following rule of the dynamic exponential weighting method for assigning weights to single - day demand trend item components in different periods: assign larger weights to the single - day demand trend item components in later periods and smaller weights to the single - day demand trend item components in earlier periods.

[0088] S23. Based on the single - day demand trend item component time - series data, train a prediction model using the Prophet model, and apply the trained prediction model to predict the predicted values of the second single - day demand trend item component, single - day demand seasonal effect item component, single - day demand holiday effect item component, and single - day demand error item component for each day in the current next procurement cycle of the target drug on the downstream side of the medical industry supply chain.

[0089] In step S23, the Prophet model is an existing data prediction tool based on Python and R languages, suitable for processing time - series data with strong seasonal patterns and historical trends. The Prophet model is constructed based on an additive model, consisting of trends, seasonality, holiday effects, and errors. The mathematical representation of the corresponding algorithm is:

[0090] P(t) = g(t)+s(t)+h(t)+∈ t

[0091] In the formula, g(t) represents the trend component of the time series, s(t) is used to capture the seasonal effect of the time series, h(t) is used to reflect the holiday effect, and ∈ t is the error term, representing the noise in the data. The Prophet algorithm specifically includes two trend methods: using a non - linear saturation growth method and a piece - wise linear method to represent the trend component. Among them, the mathematical representation of the non - linear saturation growth method is:

[0092]

[0093] This method is used for the case where the data shows a saturation growth pattern. C represents the carrying capacity, and k and m are respectively used to control the growth rate and the mid - point; the mathematical representation of the piece - wise linear method is:

[0094] g(t)=(k + a(t) T δ)t+(m + a(t) T γ)

[0095] This method assumes that the trend can be described by a linear model and the slope changes at certain time points (i.e., change points). k is the initial slope, m is the offset, a(t) is an indicator function indicating whether a change point has been passed, δ is the slope change at each change point, and γ is used to adjust the intercept at each change point. The seasonal component s(t) is used to capture the periodic fluctuations in the data. The Prophet model allows multiple seasonal components with different periods (such as weekly or daily) to exist simultaneously. That is, the seasonal effect can be represented by a Fourier series:

[0096]

[0097] Wherein, P is the seasonal period. In the 120 emergency call volume data, the weekly seasonality P = 7×24, and the daily seasonality P = 24, a n and b n are Fourier coefficients, and N represents the maximum number of terms of the Fourier series. The holiday effect h(t) is used to simulate the impact on the time series when a specific event or holiday occurs. It can be modeled using dummy variables to represent the occurrence of holidays and can include lead or lag effects to account for changes in behavior before and after the event. Its mathematical representation is:

[0098]

[0099] Wherein, D i (t) is the indicator function for holiday i, and λ i represents the magnitude of the holiday effect. The error term ∈ t is used to capture the data noise that cannot be explained by trends, seasonality, or holiday effects. It is usually assumed to follow a normal distribution with a mean of zero, i.e., ∈ t ~N(0, σ 2 (t)), and the variance may vary with time t. Therefore, the aforementioned prediction model training process and the specific process of obtaining the predicted values of the second single-day demand trend term component, single-day demand seasonal effect term component, single-day demand holiday effect term component, and single-day demand error term component for each day in the current next procurement cycle and on the downstream side of the medical industry supply chain through model application can all be routinely derived based on existing technical means and will not be elaborated here. In addition, considering that there is a certain correlation between the demand quantities of different two drugs, it is also possible to combine the single-day demand trend term component time series data and the single-day demand trend term component time series data of other drugs with a relatively high correlation with the target drug during the prediction model training and application processes.

[0100] S24. For each day in the current next procurement cycle, calculate the predicted value sp1 of the first single-day demand for the target drug on the corresponding day and on the downstream side of the medical industry supply chain according to the following formula:

[0101] sp1 = w s ×g sp,1 + w p ×g sp,2 + s prophet + h prophet + ε prophet

[0102] Wherein, g sp,1 represents the predicted value of the first single-day demand trend term component, g sp,2Denote the predicted value of the second single - day demand trend item component as s prophet Denote the predicted value of the single - day demand seasonal effect item component as h prophet Denote the predicted value of the single - day demand holiday effect item component as ε prophet Denote the predicted value of the single - day demand error item component as w s and w p respectively denote preset weight coefficients and there is w s +w p = 1.

[0103] In step S24, considering that the consumption of the target drug is closely related to medical data such as morbidity data and 120 emergency event data, and these medical data have significant daily and weekly seasonality, but the trend is usually affected by the interaction of multiple complex factors such as weather, urban traffic, and urban population growth. Therefore, although the Prophet model can quickly extract seasonal features and holiday effects from time - series data, in the extraction of the trend of time - series, only the piece - wise linear method is used, and it is difficult to accurately extract the complex trend features in medical data when the trend change point occurs. The STL trend prediction method based on exponential weighting uses LOESS to smooth the time - series to extract the trend component, fits a smooth curve by weighted regression of the data points near each time point, and finally aggregates the trend exponentially weighted for each past day, so that complex non - linear trend features can be extracted from time - series data. Therefore, based on the above steps S21 - S24, by combining the advantages of the Prophet in quickly extracting seasonal features and holiday effects and the advantages of the STL decomposition in quickly extracting non - linear trend features, the prediction accuracy can be effectively improved.

[0104] S3. According to the predicted value of the single - day demand of the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain, construct the following multi - objective functions: the total cost minimization objective function f1(x), the inventory turnover rate maximization objective function f2(x), the near - expiration ratio minimization objective function f3(x), and the capacity utilization rate maximization objective function f4(x), where x represents the drug inventory plan of the target drug in the current next procurement cycle and on the downstream side of the medical industry supply chain and is represented by a Y - dimensional vector, and the y - th element c y in the Y - dimensional vector represents the storage quantity of the y - th production batch of the target drug, y represents a positive integer less than or equal to Y, and Y represents the total number of production batches of the target drug.

[0105] In the step S3, the multi-objective function is used to achieve multi-objective balance among minimizing the inventory cost of drugs, minimizing the out-of-stock risk, maximizing the inventory turnover rate, minimizing the proportion of drugs approaching the expiration date, and maximizing the capacity utilization rate. Specifically, the total cost in the total cost minimization objective function f1(x) is equal to the sum of the procurement cost, storage cost, expiration loss cost, and out-of-stock loss cost. Among them, the procurement cost is related to the y-th element c y and the remaining shelf life of the current drugs in the y-th production batch (generally speaking, the shorter the remaining shelf life of the current drugs, the cheaper the selling price of the corresponding drugs, and the larger the procurement and storage volume, the cheaper the selling price of the corresponding drugs). The storage cost is related to the y-th element c y and the predicted daily demand value of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain (because the target drug will be consumed in each day of the current next procurement cycle, resulting in a daily decrease in the required storage volume and a corresponding decrease in the storage cost for each day). The expiration loss cost is related to the y-th element c y , the remaining shelf life of the current drugs in the y-th production batch, and the predicted daily demand value of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain (although the target drug will be consumed in each day of the current next procurement cycle, if the consumption speed is too slow and the remaining shelf life of the current drugs is too short, the expiration loss cost will be too large). The out-of-stock loss cost is related to the y-th element c y , the remaining shelf life of the current drugs in the y-th production batch, and the predicted daily demand value of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain (that is, if the consumption speed of the target drug with a remaining shelf life meeting the usage requirements is too fast, it will cause out-of-stock, and thus there will be a certain penalty cost). Thus, by constructing the aforementioned total cost minimization objective function f1(x), it can be used to achieve multi-objective balance in the dimensions of minimizing the inventory cost of drugs and minimizing the out-of-stock risk.

[0106] In the step S3, specifically, the inventory turnover rate in the inventory turnover rate maximization objective function f2(x) is equal to the sum of the predicted daily demand values of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain divided by Thus, by constructing the aforementioned inventory turnover rate maximization objective function f2(x), it can be used to achieve multi-objective balance in the dimension of maximizing the inventory turnover rate of drugs.

[0107] In the step S3, specifically, the expiration ratio in the expiration ratio minimization objective function f3(x) is equal to the storage quantity of the target drug with the remaining shelf life less than the preset duration in the drug inventory plan divided by Thus, by constructing the aforementioned expiration ratio minimization objective function f3(x), it can be used for multi-objective balance in terms of the expiration ratio minimization of drugs.

[0108] In the step S3, specifically, the capacity utilization rate in the capacity utilization rate maximization objective function f4(x) is equal to divided by the maximum allowable storage quantity of the target drug on the downstream side of the medical industry supply chain. Thus, by constructing the aforementioned capacity utilization rate maximization objective function f4(x), it can be used for multi-objective balance in terms of the capacity utilization rate maximization of drugs.

[0109] S4. Apply the NSGA-II algorithm to solve the multi-objective optimization problem established based on the multi-objective function, and obtain the Pareto front solution set, where the Pareto front solution set contains at least one Pareto optimal solution for the drug inventory plan x.

[0110] In the step S4, the NSGA-II (Nondominated Sorting Genetic Algorithm II) algorithm is an effective algorithm for solving multi-objective optimization problems, proposed by Deb et al. in 2002; with its fast non-dominated sorting method, crowding degree calculation strategy and elite retention mechanism, it performs excellently in dealing with multi-objective optimization problems and has received wide attention and application. The NSGA-II algorithm mainly includes the following steps: initializing the population, evaluating the fitness of the population individuals, non-dominated sorting, crowding degree calculation, selection operation, crossover and mutation operations, elite retention, and termination condition check. Finally, in the termination condition check link, if the maximum iteration number or other termination conditions are reached, the Pareto front solution set in the current population is output; otherwise, the aforementioned steps are repeated. Thus, by applying the NSGA-II algorithm, the Pareto front solution set for solving the multi-objective optimization problem can be obtained. In addition, the specific solution process of the multi-objective optimization problem can be routinely deduced based on the specific steps of the NSGA-II algorithm, and the Pareto front solution set and the Pareto optimal solution are both common terms in the NSGA-II algorithm, which will not be elaborated here.

[0111] S5. Select any Pareto optimal solution that meets the preset storage strategy from the Pareto front solution set, and push the any Pareto optimal solution as the optimal drug inventory plan to the downstream side of the medical industry supply chain for implementation.

[0112] In step S5, the preset storage strategy can be pre-designed by users on the downstream side of the medical industry supply chain according to actual needs, for example, by pre-designing the weight ratio of minimizing inventory cost dimension, minimizing out-of-stock risk dimension, maximizing inventory turnover dimension, minimizing near-expiration ratio dimension and maximizing capacity utilization dimension to reflect the preset storage strategy. Specifically, any Pareto optimal solution that meets the preset storage strategy is selected from the Pareto front solution set, including but not limited to the following steps S51 to S53.

[0113] S51. For each Pareto optimal solution in the Pareto front solution set, determine the corresponding total cost, inventory turnover rate, near-expiration ratio and capacity utilization rate based on the corresponding solution and the predicted daily demand value of the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain.

[0114] In step S51, since the Pareto optimal solution is a drug inventory plan x, the total cost, inventory turnover rate, near-expiry ratio and capacity utilization rate corresponding to each Pareto optimal solution can be conventionally determined based on the specific description of the total cost, inventory turnover rate, near-expiry ratio and capacity utilization rate in the aforementioned step S3.

[0115] S52. For each of the Pareto optimal solutions, the corresponding normalized total cost value, normalized inventory turnover rate, normalized near-effectiveness ratio, and capacity utilization rate are calculated based on the corresponding total cost, inventory turnover rate, near-effectiveness ratio, and capacity utilization rate, as well as the total cost, inventory turnover rate, near-effectiveness ratio, and capacity utilization rate of all the Pareto optimal solutions, and the corresponding recommended indicator value is calculated according to the following formula:

[0116] q z =-w1×q z,1 +w2×q z,2 -w3×q z,3 +w4×q z,4

[0117] In the formula, z represents a positive integer, q z represents the recommended index value of the zth Pareto optimal solution in the Pareto front solution set, q z,1 represents the normalized total cost corresponding to the zth Pareto optimal solution, q z,2 represents the normalized value of inventory turnover rate corresponding to the zth Pareto optimal solution, q z,3 represents the normalized value of the near-effective period ratio corresponding to the zth Pareto optimal solution, q z,4It represents the normalized value of the capacity utilization rate corresponding to the z-th Pareto optimal solution. w1, w2, w3, and w4 respectively represent preset weight coefficients and w1 + w2 + w3 + w4 = 1.

[0118] In the step S52, the weight coefficients w1, w2, w3, and w4 reflect the preset storage strategy. In addition, specifically, for each of the Pareto optimal solutions, according to the corresponding total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate, as well as the total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate of all the Pareto optimal solutions, the corresponding normalized values of the total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate are calculated, including but not limited to the following steps: For each of the Pareto optimal solutions, according to the corresponding total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate, as well as the total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate of all the Pareto optimal solutions, the corresponding normalized values of the total cost, inventory turnover rate, near-expiry ratio, and capacity utilization rate are calculated according to the following formula:

[0119]

[0120] In the formula, z represents a positive integer, q z,0 represents the calculated normalized value of the z-th Pareto optimal solution in the Pareto front solution set, Q z,0 represents the original value before calculation of the z-th Pareto optimal solution, Q min,0 represents the minimum value among the original values before calculation of all the Pareto optimal solutions, Q max,0 represents the maximum value among the original values before calculation of all the Pareto optimal solutions. The calculated normalized value refers to the normalized value of the total cost, inventory turnover rate, near-expiry ratio, or capacity utilization rate, and the original value before calculation refers to the total cost, inventory turnover rate, near-expiry ratio, or capacity utilization rate, and is consistent with the calculated normalized value.

[0121] S53. Select a certain Pareto optimal solution with the maximum recommended index value from the Pareto front solution set as any Pareto optimal solution that conforms to the preset storage strategy.

[0122] Based on the supply chain multi-objective decision-making method described in the foregoing steps S1 to S5, a new supply chain multi-objective decision-making solution applicable to drug inventory management is provided. That is, first, based on the time series data of the daily demand of the target drug, the predicted values of the daily demand of the drug in each day of the current next procurement cycle are estimated. Then, a multi-objective function is constructed according to the prediction results, including an objective function of minimizing the total cost, an objective function of maximizing the inventory turnover rate, an objective function of minimizing the proportion of near-expiry drugs, and an objective function of maximizing the capacity utilization rate. Then, the NSGA-II algorithm is applied to solve the multi-objective optimization problem established based on the multi-objective function to obtain the Pareto front solution set. Finally, any Pareto optimal solution that meets the preset storage strategy is selected from the solution set and pushed to the downstream side of the supply chain for implementation. In this way, a multi-objective balance can be achieved among minimizing the inventory cost of drugs, minimizing the out-of-stock risk, maximizing the inventory turnover rate, minimizing the proportion of near-expiry drugs, and maximizing the capacity utilization rate, improving the drug inventory management ability and efficiency on the downstream side of the medical industry supply chain, and facilitating practical application and promotion.

[0123] As Figure 2 shown, in the second aspect of this embodiment, a virtual system for implementing the supply chain multi-objective decision-making method described in the first aspect is provided, including but not limited to a time series data acquisition unit, a drug demand prediction unit, an objective function construction unit, an optimization problem solving unit, and an optimal solution pushing unit that are sequentially communicatively connected;

[0124] The time series data acquisition unit is used to acquire the time series data of the daily demand of the target drug on the downstream side of the medical industry supply chain. Among them, the time series data of the daily demand includes the current most recent consecutive multiple days in time series and the daily demand of the target drug on each of the current most recent consecutive multiple days on the downstream side of the medical industry supply chain;

[0125] The drug demand prediction unit is used to estimate the predicted values of the daily demand of the target drug on each day in the current next procurement cycle on the downstream side of the medical industry supply chain according to the time series data of the daily demand of the target drug;

[0126] The objective function construction unit is used to construct the following multi-objective function according to the predicted values of the daily demand of the target drug on each day in the current next procurement cycle on the downstream side of the medical industry supply chain: an objective function f1(x) of minimizing the total cost, an objective function f2(x) of maximizing the inventory turnover rate, an objective function f3(x) of minimizing the proportion of near-expiry drugs, and an objective function f4(x) of maximizing the capacity utilization rate. Among them, x represents the drug inventory plan of the target drug in the current next procurement cycle on the downstream side of the medical industry supply chain and is represented by a Y-dimensional vector. The y-th element c in the Y-dimensional vector yrepresents the storage quantity of the y-th production batch of the target drug, where y represents a positive integer less than or equal to Y, and Y represents the total number of production batches of the target drug;

[0127] The optimization problem solving unit is used to apply the NSGA-II algorithm to solve the multi-objective optimization problem established based on the multi-objective function, and obtain a Pareto front solution set, where the Pareto front solution set contains at least one Pareto optimal solution for the drug inventory plan x;

[0128] The optimal solution pushing unit is used to select any Pareto optimal solution that meets the preset storage strategy from the Pareto front solution set, and push the any Pareto optimal solution as the optimal drug inventory plan to the downstream side of the medical industry supply chain for implementation.

[0129] For the working process, working details and technical effects of the foregoing system provided in the second aspect of this embodiment, reference can be made to the supply chain multi-objective decision-making method described in the first aspect, which will not be elaborated here.

[0130] As Figure 3 shown, the third aspect of this embodiment provides a computer system for executing the supply chain multi-objective decision-making method described in the first aspect, including a memory, a processor and a transceiver that are sequentially communicatively connected, where the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the supply chain multi-objective decision-making method described in the first aspect. Specifically, for example, the memory may include, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a flash memory, a first input first output (FIFO) memory, and / or a first input last output (FILO) memory, etc.; the processor may include, but is not limited to, a microprocessor of the STM32F105 series. In addition, the computer device may also include, but is not limited to, a power module, a display screen and other necessary components.

[0131] For the working process, working details and technical effects of the foregoing computer system provided in the third aspect of this embodiment, reference can be made to the supply chain multi-objective decision-making method described in the first aspect, which will not be elaborated here.

[0132] The fourth aspect of this embodiment provides a computer-readable storage medium storing instructions for the multi-objective decision-making method for the supply chain as described in the first aspect, that is, instructions are stored on the computer-readable storage medium, and when the instructions run on a computer, the multi-objective decision-making method for the supply chain as described in the first aspect is executed. Among them, the computer-readable storage medium refers to a carrier for storing data, and may include, but is not limited to, computer-readable storage media such as floppy disks, optical discs, hard disks, flash memories, USB flash drives, and / or Memory Sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0133] For the working process, working details, and technical effects of the foregoing computer-readable storage medium provided in the fourth aspect of this embodiment, reference may be made to the multi-objective decision-making method for the supply chain as described in the first aspect, which will not be elaborated here.

[0134] The fifth aspect of this embodiment provides a computer program product, including a computer program or instructions, and when the computer program or the instructions are executed by a computer, the multi-objective decision-making method for the supply chain as described in the first aspect is implemented. Among them, the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0135] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A multi-objective decision-making method for the supply chain in the medical industry, characterized in that Including: Obtaining the time - series data of the single - day demand volume of a target drug on the downstream side of the medical industry supply chain. Among them, the time - series data of the single - day demand volume contains the current recent historical multi - days in sequence in time series and the single - day demand volume of the target drug on each day of the current recent historical multi - days and on the downstream side of the medical industry supply chain; Based on the time - series data of the single - day demand volume of the target drug, estimating the predicted value of the single - day demand volume of the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain; According to the predicted value of the single-day demand of the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain, construct the following multi-objective function: minimize the total cost objective function f1(x), maximize the inventory turnover rate objective function f2(x), minimize the proportion of near-expiry date objective function f3(x), and maximize the capacity utilization rate objective function f4(x), where x represents the drug inventory plan of the target drug in the current next procurement cycle and on the downstream side of the medical industry supply chain and is represented by a Y-dimensional vector, and the y-th element c in the Y-dimensional vector y represents the storage quantity of the y-th production batch of the target drug, y represents a positive integer less than or equal to Y, and Y represents the total number of production batches of the target drug; Applying the NSGA - II algorithm to solve the multi - objective optimization problem established based on the multi - objective function, and obtaining the Pareto front solution set. Among them, the Pareto front solution set contains at least one Pareto optimal solution for the drug inventory plan x; Selecting any Pareto optimal solution that meets the preset storage strategy from the Pareto front solution set, and pushing the any Pareto optimal solution as the optimal drug inventory plan to the downstream side of the medical industry supply chain for implementation.

2. The multi-objective decision-making method for the supply chain according to claim 1, wherein Based on the time - series data of the single - day demand volume of the target drug, estimating the predicted value of the single - day demand volume of the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain, including: Using the STL algorithm to decompose the time - series data of the single - day demand volume of the target drug, and obtaining the time - series data of the single - day demand trend item component of the target drug. Among them, the time - series data of the single - day demand trend item component contains the current recent historical multi - days and the single - day demand trend item component of the target drug on each day of the current recent historical multi - days and on the downstream side of the medical industry supply chain; Based on the time - series data of the single - day demand trend item component, calculating the predicted value of the first single - day demand trend item component of the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain by using the dynamic exponential weighting method. Among them, the dynamic exponential weighting method assigns weights to the single - day demand trend item components in different periods according to the following rules: assigning larger weights to the single - day demand trend item components in the later periods and smaller weights to the single - day demand trend item components in the earlier periods; Based on the time - series data of the single - day demand trend item component, training a prediction model based on the Prophet model, and using the trained prediction model to predict the predicted value of the second single - day demand trend item component, the predicted value of the single - day demand seasonal effect item component, the predicted value of the single - day demand holiday effect item component, and the predicted value of the single - day demand error item component of the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain; For each day in the current next procurement cycle, calculating the predicted value sp1 of the single - day demand volume of the target drug on the corresponding day and on the downstream side of the medical industry supply chain according to the following formula: sp1 = w s × g sp,1 + w p × g sp,2 + s prophet + h prophet + ε prophet Wherein, g sp,1 represents the predicted value of the first single-day demand trend item component, g sp,2 represents the predicted value of the second single-day demand trend item component, s prophet represents the predicted value of the single-day demand seasonal effect item component, h prophet represents the predicted value of the single-day demand holiday effect item component, ε prophet represents the predicted value of the single-day demand error item component, w s and w p respectively represent preset weight coefficients and w s + w p = 1.

3. The multi-objective decision-making method for a supply chain according to claim 2, wherein When the current recent historical multi - day period is all the days of the current recent multi - weeks that are sequentially continuous in time series and the current next procurement cycle is all the days in the current next week, based on the single - day demand trend item component time - series data, the first single - day demand trend item component prediction values of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain are calculated based on the dynamic exponential weighting method, including: Dividing the single - day demand trend item component time - series data into multiple arrays corresponding one - by - one to the current recent multi - weeks, and aggregating the multiple arrays to obtain the following matrix G: Wherein, n represents the total number of weeks in the current most recent multiple weeks, n' represents a positive integer less than or equal to n, m' represents a positive integer less than or equal to 7, and g n′,m′ represents the component of the single-day demand trend item of the target drug on the m'-th day in the n'-th week among the current most recent multiple weeks and on the downstream side of the medical industry supply chain; Using the following recurrence formula to calculate the first single - day demand trend item component prediction values of the target drug in each day of the current next week and on the downstream side of the medical industry supply chain: where k represents a positive integer less than or equal to n, When k is less than n, represents the predicted value of the first single-day demand trend item component of the target drug on the m'-th day in the (k + 1)-th week among the current most recent multiple weeks and on the downstream side of the medical industry supply chain. When k is equal to n, represents the predicted value of the first single-day demand trend item component of the target drug on the m'-th day in the next week of the current week and on the downstream side of the medical industry supply chain. α represents a preset smoothing factor and α ∈ (0, 1).

4. The multi-objective decision-making method for a supply chain according to claim 1, wherein The total cost in the minimum total cost objective function f1(x) is equal to the sum of the procurement cost, the storage cost, the expired loss cost, and the shortage loss cost. Among them, the procurement cost is related to the y-th element c y and the remaining shelf life of the current drugs in the y-th production batch. The storage cost is related to the y-th element c y and the predicted value of the daily demand of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain. The expired loss cost is related to the y-th element c y , the remaining shelf life of the current drugs in the y-th production batch, and the predicted value of the daily demand of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain. The shortage loss cost is related to the y-th element c y , the remaining shelf life of the current drugs in the y-th production batch, and the predicted value of the daily demand of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain; And / or, the inventory turnover rate in the maximized inventory turnover rate objective function f2(x) is equal to the sum of the predicted daily demand values of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain, divided by And / or, the expiration approaching ratio in the minimization of expiration approaching ratio objective function f3(x) is equal to the storage quantity of the target drug with the remaining shelf life of the current drug less than the preset duration in the drug inventory plan divided by And / or, the capacity utilization rate in the maximized capacity utilization objective function f4(x) is equal to Divided by the maximum allowable storage quantity of the target drug on the downstream side of the medical industry supply chain.

5. The multi-objective decision-making method for supply chain according to claim 1, wherein Selecting any Pareto - optimal solution that meets the preset storage strategy from the Pareto - front solution set, including: For each Pareto - optimal solution in the Pareto - front solution set, determining the corresponding total cost, inventory turnover rate, near - expiration date ratio, and capacity utilization rate according to the corresponding solution and the single - day demand prediction values of the target drug in each day of the current next procurement cycle and on the downstream side of the medical industry supply chain; For each of the Pareto - optimal solutions, calculating the corresponding total - cost normalization value, inventory - turnover - rate normalization value, near - expiration - date - ratio normalization value, and capacity - utilization - rate normalization value according to the corresponding total cost, inventory turnover rate, near - expiration date ratio, and capacity utilization rate and the total cost, inventory turnover rate, near - expiration date ratio, and capacity utilization rate of all the Pareto - optimal solutions, and calculating the corresponding recommended index value according to the following formula: q z = -w1×q z,1 +w2×q z,2 -w3×q z,3 +w4×q z,4 where z represents a positive integer, q z represents the recommended index value of the z-th Pareto optimal solution in the Pareto front solution set, q z,1 represents the total cost normalization value corresponding to the z-th Pareto optimal solution, q z,2 represents the inventory turnover rate normalization value corresponding to the z-th Pareto optimal solution, q z,3 represents the near-expiry ratio normalization value corresponding to the z-th Pareto optimal solution, q z,4 represents the capacity utilization rate normalization value corresponding to the z-th Pareto optimal solution, and w1, w2, w3, and w4 respectively represent preset weight coefficients and w1 + w2 + w3 + w4 = 1; Selecting a Pareto - optimal solution with the largest recommended index value from the Pareto - front solution set as any Pareto - optimal solution that meets the preset storage strategy.

6. The multi-objective decision-making method for a supply chain according to claim 5, wherein, For each of the Pareto - optimal solutions, calculating the corresponding total - cost normalization value, inventory - turnover - rate normalization value, near - expiration - date - ratio normalization value, and capacity - utilization - rate normalization value according to the corresponding total cost, inventory turnover rate, near - expiration date ratio, and capacity utilization rate and the total cost, inventory turnover rate, near - expiration date ratio, and capacity utilization rate of all the Pareto - optimal solutions, including: For each of the Pareto - optimal solutions, calculating the corresponding total - cost normalization value, inventory - turnover - rate normalization value, near - expiration - date - ratio normalization value, and capacity - utilization - rate normalization value according to the corresponding total cost, inventory turnover rate, near - expiration date ratio, and capacity utilization rate and the total cost, inventory turnover rate, near - expiration date ratio, and capacity utilization rate of all the Pareto - optimal solutions according to the following formula: Where z represents a positive integer, and q z,0 represents the calculated normalized value of the z-th Pareto optimal solution in the Pareto front solution set, Q z,0 represents the original value before calculation of the z-th Pareto optimal solution, Q min,0 represents the minimum value among the original values before calculation of all the Pareto optimal solutions, Q max,0 represents the maximum value among the original values before calculation of all the Pareto optimal solutions. The calculated normalized value refers to the total cost normalized value, inventory turnover normalized value, near-expiry ratio normalized value or capacity utilization rate normalized value. The original value before calculation refers to the total cost, inventory turnover, near-expiry ratio or capacity utilization rate, and is consistent with the calculated normalized value.

7. A multi-objective decision-making system for the supply chain in the medical industry, characterized in that, Including a time - series data acquisition unit, a drug demand prediction unit, a target - function construction unit, an optimization - problem solving unit, and an optimal - solution push unit that are sequentially communicatively connected; The time series data acquisition unit is configured to acquire the time series data of the single-day demand of the target drug on the downstream side of the medical industry supply chain. The single-day demand time series data includes the current most recent consecutive multi-days in time series and the single-day demand of the target drug on each of the current most recent consecutive multi-days on the downstream side of the medical industry supply chain; The drug demand forecasting unit is configured to estimate the predicted value of the single-day demand of the target drug on each day in the current next procurement cycle on the downstream side of the medical industry supply chain according to the single-day demand time series data of the target drug; The objective function construction unit is configured to construct the following multi-objective functions according to the predicted values of the single-day demand of the target drug on each day in the current next procurement cycle and on the downstream side of the medical industry supply chain: minimizing the total cost objective function f1(x), maximizing the inventory turnover rate objective function f2(x), minimizing the proportion of near-expiry date objective function f3(x), and maximizing the capacity utilization rate objective function f4(x), where x represents the drug inventory plan of the target drug in the current next procurement cycle and on the downstream side of the medical industry supply chain and is represented by a Y-dimensional vector, and the y-th element c in the Y-dimensional vector y represents the storage quantity of the y-th production batch of the target drug, y represents a positive integer less than or equal to Y, and Y represents the total number of production batches of the target drug; The optimization problem solving unit is configured to apply the NSGA-II algorithm to solve the multi-objective optimization problem established based on the multi-objective function, and obtain the Pareto front solution set. The Pareto front solution set includes at least one Pareto optimal solution for the drug inventory plan x; The optimal solution pushing unit is configured to select any Pareto optimal solution that meets the preset storage strategy from the Pareto front solution set, and push the any Pareto optimal solution as the optimal drug inventory plan to the downstream side of the medical industry supply chain for implementation.

8. A computer system, characterized in that, It includes a memory, a processor, and a transceiver that are communicatively connected in sequence. The memory is configured to store computer programs, the transceiver is configured to send and receive messages, and the processor is configured to read the computer programs and execute the supply chain multi-objective decision-making method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that , An instruction is stored on the computer-readable storage medium. When the instruction runs on the computer, the supply chain multi-objective decision-making method according to any one of claims 1 to 6 is executed.

10. A computer program product, comprising a computer program or instructions, characterized in that, The computer program or the instruction, when executed by the computer, implements the supply chain multi-objective decision-making method according to any one of claims 1 to 6.

Images