Supply chain demand prediction method and device, equipment, storage medium and program product
By using ant colony optimization algorithm to optimize the prediction model in the supply chain, the problem of insufficient processing capabilities of traditional methods for nonlinear relationships and multi-factor interactions in supply chain demand prediction is solved, and higher prediction accuracy and real-time performance are achieved .
Patent Information
- Application Number
- CN202410795834.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-19
- Publication Date
- 2024-11-12
AI Technical Summary
Traditional mathematical methods have problems in supply chain demand prediction, which are insufficient nonlinear relationship capture capabilities and insufficient processing capabilities for multi-factor interaction influence in complex environments, making it difficult to meet the requirements of accurate and real-time demand forecasting of modern supply chain management.
By obtaining the demand impact data of the supply chain, input it into the trained initial prediction model for prediction, and optimizing the model parameters in combination with the ant colony optimization algorithm to form a target prediction model to improve the accuracy of supply chain demand prediction.
Effectively capture the regularity and changing trends of supply chain demand, improve the accuracy of supply chain demand forecasting, and better handle complex nonlinear relationships and multi-factor interactions.
Smart Images

Figure CN118941322A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of supply chain management and artificial intelligence technology, and in particular to a supply chain demand forecasting method, apparatus, device, storage medium and program product. Background Art
[0002] In recent years, the continuous changes in the global market and the intensification of competition have made efficient supply chain management the key to maintaining competitive advantages for enterprises. With the rapid development of digital technology and the continuous expansion of the global supply chain network, enterprises are facing a more complex and dynamic market environment. In this context, supply chain demand forecasting has become one of the core links of supply chain management. Enterprises need to accurately grasp consumer purchasing behaviors and trends through more accurate demand forecasting in order to achieve efficient production and logistics operations in the supply chain. Through effective demand forecasting, enterprises can avoid overstocking or shortages, improve the flexibility and responsiveness of production plans, reduce inventory costs and logistics costs, and thus enhance the market competitiveness and profitability of enterprises.
[0003] Since supply chain demand is affected by many factors, such as fluctuations in market demand, seasonal changes, fluctuations in product prices, the impact of promotional activities, and the squeeze on market share by competition, it is highly uncertain and nonlinear. Traditional mathematical methods, such as time series analysis and regression analysis, have some limitations in supply chain demand forecasting, mainly reflected in the weak ability to capture nonlinear relationships, insufficient ability to handle the interaction of multiple factors in complex environments, and inability to process large-scale data. Especially in the face of rapid changes and diversification in market demand, the traditional way of predicting supply chain demand through mathematical methods is often difficult to meet the needs of modern supply chain management for accurate and real-time forecasting. Summary of the invention
[0004] Based on this, it is necessary to provide a supply chain demand forecasting method, device, equipment, storage medium and program product that can improve the accuracy of supply chain demand forecasting in response to the above technical problems.
[0005] In a first aspect, the present application provides a supply chain demand forecasting method, the method comprising:
[0006] Obtain demand impact data for the supply chain;
[0007] Inputting the demand impact data into a trained initial forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data through the initial forecasting model, and obtain an initial demand result for the supply chain;
[0008] By using an ant colony optimization algorithm, based on the demand impact data and the initial demand result, the model parameters of the initial prediction model are optimized to obtain a target prediction model;
[0009] The demand impact data is input into the target forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data through the target forecasting model to obtain a target demand result for the supply chain.
[0010] In a second aspect, the present application provides a supply chain demand forecasting device, the device comprising:
[0011] The acquisition module is used to obtain demand impact data of the supply chain;
[0012] A forecasting module, used for inputting the demand impact data into a trained initial forecasting model, so as to forecast the demand for the supply chain based on the demand impact data through the initial forecasting model, and obtain an initial demand result for the supply chain;
[0013] An optimization module, configured to optimize the model parameters of the initial prediction model based on the demand impact data and the initial demand result through an ant colony optimization algorithm to obtain a target prediction model;
[0014] The forecasting module is also used to input the demand impact data into the target forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data through the target forecasting model to obtain a target demand result for the supply chain.
[0015] In a third aspect, the present application provides a computer device including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps in the method embodiments of the present application are implemented.
[0016] In a fourth aspect, the present application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the method embodiments of the present application.
[0017] In a fifth aspect, the present application provides a computer program product, including a computer program, which implements the steps in the method embodiments of the present application when the computer program is executed by a processor.
[0018] The above-mentioned supply chain demand forecasting method, device, equipment, storage medium and program product obtains the demand impact data of the supply chain, inputs the demand impact data into the trained initial forecasting model, and performs demand forecasting for the supply chain based on the demand impact data through the initial forecasting model to obtain the initial demand results of the supply chain. Based on the demand impact data and the initial demand results, the model parameters of the initial forecasting model are optimized to obtain the target forecasting model, and the demand impact data is input into the target forecasting model to perform demand forecasting for the supply chain based on the demand impact data through the target forecasting model to obtain the target demand results of the supply chain. Compared with the traditional method of predicting supply chain demand through mathematical methods, this application combines the forecasting model with the ant colony optimization algorithm, which can not only mine the patterns and characteristics of the supply chain demand from the demand impact data of the supply chain, but also effectively learn the complex nonlinear relationships in the data, so as to better capture the regularity and changing trends of the supply chain demand. The parameters of the forecasting model can be optimized through the ant colony optimization algorithm, thereby improving the accuracy of the supply chain demand forecast. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 An application environment diagram of a supply chain demand forecasting method in one embodiment;
[0020] Figure 2 A schematic diagram of a flow chart of a supply chain demand forecasting method in one embodiment;
[0021] Figure 3 A schematic diagram of a model structure of a deep neural network model in one embodiment;
[0022] Figure 4 is a structural block diagram of a supply chain demand forecasting device in one embodiment;
[0023] Figure 5 It is a structural block diagram of a supply chain demand forecasting device in another embodiment;
[0024] Figure 6 is an internal structure diagram of a computer device in one embodiment;
[0025] Figure 7 FIG. 4 is a diagram showing the internal structure of a computer device in another embodiment. DETAILED DESCRIPTION
[0026] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0027] The supply chain demand forecasting method provided in this application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can be set up separately and can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other servers. Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptops, smart phones, tablet computers, vehicle-mounted terminals, intelligent voice interaction devices, aircraft, smart home appliances and portable wearable devices. Smart home appliances can be smart speakers, smart TVs and smart air conditioners. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, cloud security, host security and other network security services, CDN, and basic cloud computing services such as big data and artificial intelligence platforms. The terminal 102 and the server 104 can be directly or indirectly connected by wired or wireless communication, and this application is not limited here.
[0028] The server 104 may obtain the demand impact data of the supply chain, and input the demand impact data into the trained initial prediction model, so as to perform demand forecasting for the supply chain based on the demand impact data through the initial prediction model, and obtain the initial demand result of the supply chain. The server 104 may optimize the model parameters of the initial prediction model based on the demand impact data and the initial demand result through the ant colony optimization algorithm, and obtain the target prediction model. The server 104 may input the demand impact data into the target prediction model, so as to perform demand forecasting for the supply chain based on the demand impact data through the target prediction model, and obtain the target demand result of the supply chain.
[0029] It is understood that this embodiment does not limit this. It is understood that Figure 1 The application scenarios are only for illustration and are not limited to this.
[0030] In one embodiment, Figure 2 As shown, a supply chain demand forecasting method is provided, which can be applied to a computer device, which can be a terminal or a server, that is, the method can be executed by the terminal or the server alone, or can be implemented through interaction between the terminal and the server. This embodiment takes the method applied to a computer device as an example for explanation, and includes the following steps:
[0031] Step 202, obtaining supply chain demand impact data.
[0032] In one embodiment, the computer device may obtain supply chain demand impact data for the current time period, wherein the supply chain demand impact data includes at least one of sales volume data, price data, promotion data, seasonality data, and competition data.
[0033] Step 204, input the demand impact data into the trained initial forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data through the initial forecasting model, and obtain the initial demand result of the supply chain.
[0034] In one embodiment, the computer device may input the demand impact data of the current time period into a trained initial forecasting model to perform demand forecasting for the supply chain based on the demand impact data of the current time period through the initial forecasting model to obtain the initial demand results of the supply chain in the current time period.
[0035] In one embodiment, the trained initial prediction model may be a trained initial deep neural network model.
[0036] Step 206, using an ant colony optimization algorithm, based on the demand impact data and the initial demand results, optimizes the model parameters of the initial prediction model to obtain a target prediction model.
[0037] In one embodiment, the computer device can optimize the model parameters of the initial prediction model through the ant colony optimization algorithm based on the demand impact data and initial demand results of the current time period of the supply chain, and obtain the target prediction model for predicting the supply chain demand in the current time period.
[0038] Step 208, input the demand impact data into the target forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data through the target forecasting model, and obtain the target demand result of the supply chain.
[0039] In one embodiment, a computer device may input demand impact data of the current time period of the supply chain into an optimized target forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data of the current time period of the supply chain through the optimized target forecasting model, and obtain the target demand result of the current time period of the supply chain.
[0040] In the above-mentioned supply chain demand forecasting method, by obtaining the demand impact data of the supply chain, the demand impact data is input into the trained initial forecasting model, so as to forecast the demand for the supply chain based on the demand impact data through the initial forecasting model, and obtain the initial demand results of the supply chain. Based on the demand impact data and the initial demand results, the model parameters of the initial forecasting model are optimized to obtain the target forecasting model, and the demand impact data is input into the target forecasting model, so as to forecast the demand for the supply chain based on the demand impact data through the target forecasting model, and obtain the target demand results of the supply chain. Compared with the traditional method of predicting supply chain demand through mathematical methods, this application combines the forecasting model with the ant colony optimization algorithm, which can not only mine the patterns and characteristics of the supply chain demand from the demand impact data of the supply chain, but also effectively learn the complex nonlinear relationships in the data, so as to better capture the regularity and changing trends of the supply chain demand. The parameters of the forecasting model can be optimized through the ant colony optimization algorithm, thereby improving the accuracy of the supply chain demand forecast.
[0041] In one embodiment, the method also includes a model training step, which includes: obtaining historical demand impact data of the supply chain; determining a reference demand result of the supply chain based on the historical demand impact data; inputting the historical demand impact data into the prediction model to be trained, so as to perform demand forecasting for the supply chain based on the historical demand impact data through the prediction model to be trained, and obtain a predicted demand result of the supply chain; determining a loss value based on an error between the predicted demand result and the reference demand result, and training the prediction model to be trained based on the loss value to obtain a trained initial prediction model.
[0042] In one embodiment, the computer device may obtain historical demand impact data of the supply chain, and determine the reference demand result of the supply chain based on the historical demand impact data. The computer device may input the historical demand impact data into the prediction model to be trained, so as to perform demand forecasting for the supply chain based on the historical demand impact data through the prediction model to be trained, and obtain the predicted demand result of the supply chain. The computer device may determine the loss value based on the error between the predicted demand result and the reference demand result, and iteratively train the prediction model to be trained in the direction of reducing the loss value until the iteration stop condition is met, and obtain the trained initial prediction model.
[0043] In the above embodiment, the loss value is determined by the error between the predicted demand result and the reference demand result, and the prediction model to be trained is trained based on the loss value to obtain the trained initial prediction model. The accuracy of the trained initial prediction model in predicting supply chain demand can be improved, thereby further improving the accuracy of the final supply chain demand forecast.
[0044] In one embodiment, a reference demand result of a supply chain is determined based on historical demand impact data, including: inputting the historical demand impact data into a pre-established multivariate regression model to determine the reference demand result of the supply chain based on the historical demand impact data through the multivariate regression model.
[0045] In one embodiment, the computer device may input the historical demand impact data into a pre-established multiple regression model to determine and output a reference demand result of the supply chain based on the historical demand impact data through the multiple regression model. It is understood that the multiple regression model may digitize the historical demand impact data to obtain a reference demand result.
[0046] In the above embodiment, by inputting historical demand impact data into a pre-established multivariate regression model to determine the reference demand results of the supply chain based on the historical demand impact data through the multivariate regression model, the accuracy of the reference demand results can be improved, thereby further improving the accuracy of the final supply chain demand forecast.
[0047] In one embodiment, the multiple regression model is expressed as:
[0048] y t =β 0 +β 1 x t1 +β 2 x t2 +…+β n x tn ;
[0049] Where t represents the tth time period, t = 1, 2, ..., T; y t represents the reference demand result of the supply chain in the tth time period, y t ≤0;x t represents the historical demand impact data of the supply chain in the tth time period, x t =(x t1 , x t2 , ..., x tn ) T , x ti represents the i-th characteristic variable in the historical demand impact data, n represents the number of characteristic variables in the historical demand impact data; β n represents the model parameters of the multiple regression model.
[0050] In the above embodiment, by limiting the specific structure of the multivariate regression model, the accuracy of the reference demand results can be further improved, thereby further improving the accuracy of the final supply chain demand forecast.
[0051] In one embodiment, the prediction model is a deep neural network model; the deep neural network model is expressed as:
[0052] f(x t )=W L+1 σ L (W L σ L-1 (...σ 1 (W 1 x t +b 1 )...)+b L )+b L+1 ;
[0053] σ l (z) = max(0, z);
[0054]
[0055] Wherein, t represents the t-th time period, t = 1, 2, ..., T; x t W represents the historical demand impact data of the supply chain in the tth time period; l and b l represents the model parameters of the deep neural network model, l = 1, 2, ..., L + 1; σ l (z) represents the activation function of the deep neural network model; Represents the loss function, which is used to determine the loss value; y t represents the reference demand result of the supply chain in the tth time period; f(x t ) represents the forecast demand result.
[0056] In one embodiment, Figure 3 As shown, the model structure of the deep neural network model includes an output layer, a hidden layer and an output layer.
[0057] In the above embodiment, by limiting the specific structure of the deep neural network model, the accuracy of the predicted demand results can be further improved, thereby further improving the accuracy of the final supply chain demand forecast.
[0058] In one embodiment, the ant colony optimization algorithm is expressed as:
[0059] (1) Ant colony algorithm parameter initialization: randomly generate an initial path P of K ants k , and calculate the corresponding cost C k and fitness F k , initialize the pheromone τ in the search space ij and heuristic information η ij , set the pheromone importance α, the heuristic information importance β, the pheromone volatility coefficient ρ, the number of iterations N, the current number of iterations n = 0; establish an empty taboo table A for each ant k , Taboo Table Ak Used to record the city that ant k has currently passed through, k = 1, 2, ..., K;
[0060] (2) Conditional termination judgment: Repeat the following steps (3) to (7) until the maximum number of iterations n = N is reached or the termination condition is met, then the algorithm ends and step (8) is executed;
[0061] (3) Selection: For each ant k, the probability of transferring from city i to city j is Select a new value for each parameter to generate a new path P k ;
[0062]
[0063] Wherein, i=1, 2, ..., m, j=1, 2, ..., n, k=1, 2, ..., K;
[0064] (4) Evaluation: For each ant k, calculate the new path P k The corresponding cost C k and fitness F k , and compare it with the old path. If the new path is better, keep the new path, otherwise keep the old path, with a cost of C k and fitness F k The calculation formula is as follows;
[0065]
[0066] Among them, C k Represents the cost, used to characterize the path P selected by the kth ant k The corresponding prediction error; x t Indicates demand impact data; y t represents the reference demand result determined based on demand impact data; f(x t ) represents the initial demand result, k = 1, 2, ..., K; F k Represents the fitness, which is used to characterize the path P selected by the kth ant k the degree of quality;
[0067] (5) Update: For each value of each parameter, the pheromone update is performed when all K ants complete the traversal task, according to the pheromone increment Δτ ij , update pheromone τ ij , the update rules are as follows:
[0068] τ ij new =(1-p)τ ij old +Δτ ij ;
[0069]
[0070] Where ρ is the pheromone volatility coefficient, which indicates the attenuation degree of pheromone after each iteration, 0<ρ<1; Δτ ij represents the pheromone increment, which is used to characterize the pheromone increase of the jth value of the i-th parameter in the search space after each iteration; if the k-th ant selects the jth value of the i-th parameter, then according to the formula Calculate, otherwise Where i = 1, 2, ..., m, j = 1, 2, ..., n, k = 1, 2, ..., K;
[0071] (6) Record: Record the optimal path P in the current iteration * , the corresponding cost C * and fitness F * ;
[0072] (7) Iteration: n=n+1, used to indicate that the current iteration number n is increased by 1;
[0073] (8) Output: Output the final optimal path P * , the corresponding cost C * and fitness F * , and the corresponding target prediction model.
[0074] In the above embodiment, the accuracy of the obtained target prediction model can be further improved through the specific structure of the ant colony optimization algorithm, thereby further improving the accuracy of the final supply chain demand prediction.
[0075] It should be understood that, although each step in the flow chart of the above-mentioned embodiments is shown in order, these steps are not necessarily performed in order. Unless there is a clear explanation in this article, the execution of these steps does not have strict order restrictions, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the above-mentioned embodiments may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in order, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.
[0076] In one embodiment, Figure 4 As shown, a supply chain demand forecasting device 400 is provided, and the device specifically includes:
[0077] An acquisition module 402 is used to acquire supply chain demand impact data;
[0078] The prediction module 404 is used to input the demand impact data into the trained initial prediction model, so as to perform demand prediction for the supply chain based on the demand impact data through the initial prediction model, and obtain the initial demand result of the supply chain;
[0079] The optimization module 406 is used to optimize the model parameters of the initial prediction model based on the demand impact data and the initial demand results through an ant colony optimization algorithm to obtain a target prediction model;
[0080] The forecasting module 404 is also used to input the demand impact data into the target forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data through the target forecasting model to obtain the target demand result of the supply chain.
[0081] In one embodiment, Figure 5 As shown, the supply chain demand forecasting device 400 also includes:
[0082] The training module 408 is used to obtain the historical demand impact data of the supply chain; determine the reference demand results of the supply chain based on the historical demand impact data; input the historical demand impact data into the prediction model to be trained, so as to perform demand forecasting for the supply chain based on the historical demand impact data through the prediction model to be trained, and obtain the predicted demand results of the supply chain; determine the loss value based on the error between the predicted demand result and the reference demand result, and train the prediction model to be trained based on the loss value to obtain a trained initial prediction model.
[0083] In one embodiment, the training module 408 is further configured to input the historical demand impact data into a pre-established multiple regression model, so as to determine a reference demand result of the supply chain based on the historical demand impact data through the multiple regression model.
[0084] In one embodiment, the multiple regression model is expressed as:
[0085] y t =β 0 +β 1 x t1 +β 2 x t2 +…+β n x tn ;
[0086] Where t represents the tth time period, t = 1, 2, ..., T; y t represents the reference demand result of the supply chain in the tth time period, y t ≤0;x t represents the historical demand impact data of the supply chain in the tth time period, x t =(x t1 , xt2 , ..., x tn ) T , x ti represents the i-th characteristic variable in the historical demand impact data, n represents the number of characteristic variables in the historical demand impact data; β n represents the model parameters of the multiple regression model.
[0087] In one embodiment, the prediction model is a deep neural network model; the deep neural network model is expressed as:
[0088] f(x t )=W L+1 σ L (W L σ L-1 (...σ 1 (W 1 x t +b 1 )...)+b L )+b L+1 ;
[0089] σ l (z) = max(0, z);
[0090]
[0091] Wherein, t represents the t-th time period, t = 1, 2, ..., T; x t W represents the historical demand impact data of the supply chain in the tth time period; l and b l represents the model parameters of the deep neural network model, l = 1, 2, ..., L + 1; σ l (z) represents the activation function of the deep neural network model; Represents the loss function, which is used to determine the loss value; y t represents the reference demand result of the supply chain in the tth time period; f(x t ) represents the forecast demand result.
[0092] In one embodiment, the ant colony optimization algorithm is expressed as:
[0093] (1) Ant colony algorithm parameter initialization: randomly generate an initial path P of K ants k , and calculate the corresponding cost C k and fitness F k , initialize the pheromone τ in the search space ij and heuristic information η ij, set the pheromone importance α, the heuristic information importance β, the pheromone volatility coefficient ρ, the number of iterations N, the current number of iterations n = 0; establish an empty taboo table A for each ant k , Taboo Table A k Used to record the city that ant k has currently passed through, k = 1, 2, ..., K;
[0094] (2) Conditional termination judgment: Repeat the following steps (3) to (7) until the maximum number of iterations n = N is reached or the termination condition is met, then the algorithm ends and step (8) is executed;
[0095] (3) Selection: For each ant k, the probability of transferring from city i to city j is Select a new value for each parameter to generate a new path P k ;
[0096]
[0097] Wherein, i=1, 2, ..., m, j=1, 2, ..., n, k=1, 2, ..., K;
[0098] (4) Evaluation: For each ant k, calculate the new path P k The corresponding cost C k and fitness F k , and compare it with the old path. If the new path is better, keep the new path, otherwise keep the old path, with a cost of C k and fitness F k The calculation formula is as follows;
[0099]
[0100] Among them, C k Represents the cost, used to characterize the path P selected by the kth ant k The corresponding prediction error; x t Indicates demand impact data; y t represents the reference demand result determined based on demand impact data; f(x t ) represents the initial demand result, k = 1, 2, ..., K; F k Represents the fitness, which is used to characterize the path P selected by the kth ant k the degree of excellence;
[0101] (5) Update: For each value of each parameter, the pheromone update is performed when all K ants complete the traversal task, according to the pheromone increment Δτ ij , update pheromone τ ij , the update rules are as follows:
[0102] τ ij new =(1-p)τ ij old +Δτ ij ;
[0103]
[0104] Where ρ is the pheromone volatility coefficient, which indicates the attenuation degree of pheromone after each iteration, 0<ρ<1; Δτ ij represents the pheromone increment, which is used to characterize the pheromone increase of the jth value of the i-th parameter in the search space after each iteration; if the k-th ant selects the jth value of the i-th parameter, then according to the formula Calculate, otherwise Where i = 1, 2, ..., m, j = 1, 2, ..., n, k = 1, 2, ..., K;
[0105] (6) Record: Record the optimal path P in the current iteration * , the corresponding cost C * and fitness F * ;
[0106] (7) Iteration: n=n+1, used to indicate that the current iteration number n is increased by 1;
[0107] (8) Output: Output the final optimal path P * , the corresponding cost C * and fitness F * , and the corresponding target prediction model.
[0108] The above-mentioned supply chain demand forecasting device obtains the demand impact data of the supply chain, inputs the demand impact data into the trained initial forecasting model, and performs demand forecasting for the supply chain based on the demand impact data through the initial forecasting model to obtain the initial demand results of the supply chain. Based on the demand impact data and the initial demand results, the model parameters of the initial forecasting model are optimized to obtain the target forecasting model, and the demand impact data is input into the target forecasting model to perform demand forecasting for the supply chain based on the demand impact data through the target forecasting model to obtain the target demand results of the supply chain. Compared with the traditional method of predicting supply chain demand through mathematical methods, this application combines the forecasting model with the ant colony optimization algorithm, which can not only mine the patterns and characteristics of the supply chain demand from the demand impact data of the supply chain, but also effectively learn the complex nonlinear relationships in the data, so as to better capture the regularity and changing trends of the supply chain demand. The parameters of the forecasting model can be optimized through the ant colony optimization algorithm, thereby improving the accuracy of the supply chain demand forecast.
[0109] Each module in the above supply chain demand forecasting device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each module.
[0110] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a supply chain demand forecasting method is implemented.
[0111] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 7As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a supply chain demand forecasting method is implemented. The display unit of the computer device is used to form a visually visible image, and can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse, etc.
[0112] Those skilled in the art will understand that Figure 6 and Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0113] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.
[0114] In one embodiment, a computer-readable storage medium is provided, storing a computer program, which implements the steps in the above method embodiments when executed by a processor.
[0115] In one embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.
[0116] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.
[0117] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).
[0118] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0119] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.
Claims
1. A supply chain demand forecasting method, characterized in that: The method comprises: Obtain demand impact data for the supply chain; Inputting the demand impact data into a trained initial forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data through the initial forecasting model, and obtain an initial demand result for the supply chain; By using an ant colony optimization algorithm, based on the demand impact data and the initial demand result, the model parameters of the initial prediction model are optimized to obtain a target prediction model; The demand impact data is input into the target forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data through the target forecasting model to obtain a target demand result for the supply chain.
2. The method according to claim 1, characterized in that The method further comprises a model training step, wherein the model training step comprises: obtaining historical demand impact data for the supply chain; Determining a reference demand result of the supply chain according to the historical demand impact data; Inputting the historical demand impact data into the forecasting model to be trained, so as to perform demand forecasting for the supply chain based on the historical demand impact data through the forecasting model to be trained, and obtain a forecast demand result for the supply chain; A loss value is determined according to an error between the predicted demand result and the reference demand result, and the prediction model to be trained is trained according to the loss value to obtain a trained initial prediction model.
3. The method according to claim 2, characterized in that Determining the reference demand result of the supply chain according to the historical demand impact data includes: The historical demand impact data is input into a pre-established multiple regression model to determine a reference demand result of the supply chain based on the historical demand impact data through the multiple regression model.
4. The method according to claim 3, characterized in that The multivariate regression model is expressed as: y t =β0+β1x t1 +β2x t2 +…+b n x tn ; Wherein, t represents the t-th time period, t=1, 2, ..., T; y t represents the reference demand result of the supply chain in the tth time period, y t ≥0;x t represents the historical demand impact data of the supply chain in the tth time period, x t =(x t1 , x t2 , ..., x tn ) T , x ti represents the i-th characteristic variable in the historical demand impact data, and n represents the number of characteristic variables in the historical demand impact data; β n represents the model parameters of the multiple regression model.
5. The method according to claim 2, characterized in that: The prediction model is a deep neural network model; the deep neural network model is expressed as: f(x t )=W L+1 s L (W L s L-1 (...σ1(W1x t +b1)...)+b L )+b L+1 ; σ l (from)=max(O,from); Wherein, t represents the t-th time period, t=1, 2, ..., T; x t represents the historical demand impact data of the supply chain in the t-th time period; W l and b l represents the model parameters of the deep neural network model, l = 1, 2, ..., L + 1; σ l (z) represents the activation function of the deep neural network model; represents a loss function, which is used to determine the loss value; y t represents the reference demand result of the supply chain in the tth time period; f(x t ) represents the forecast demand result.
6. The method according to any one of claims 1 to 5, characterized in that The ant colony optimization algorithm is expressed as: (1) Ant colony algorithm parameter initialization: randomly generate an initial path P of K ants k , and calculate the corresponding cost C k and fitness F k , initialize the pheromone τ in the search space ij and heuristic information η ij , set the pheromone importance α, the heuristic information importance β, the pheromone volatility coefficient p, the number of iterations N, the current number of iterations n = 0; establish an empty taboo table A for each ant k , the taboo table A k Used to record the city that ant k has currently passed through, k = 1, 2, ..., K; (2) Conditional termination judgment: Repeat the following steps (3) to (7) until the maximum number of iterations n = N is reached or the termination condition is met, then the algorithm ends and step (8) is executed; (3) Selection: For each ant k, the probability of transferring from city i to city j is Select a new value for each parameter to generate a new path P k ; Wherein, i=1, 2, ..., m, j=1, 2, ..., n, k=1, 2, ..., K; (4) Evaluation: For each ant k, calculate the new path P k The corresponding cost C k and fitness F k , and compare it with the old path. If the new path is better, keep the new path, otherwise keep the old path, with a cost of C k and fitness F k The calculation formula is as follows; Among them, C k Represents the cost, used to characterize the path P selected by the kth ant k The corresponding prediction error; x t Indicates the demand impact data; y t represents the reference demand result determined based on the demand impact data; f(x t ) represents the initial demand result, k=1, 2, ..., K; F k Represents the fitness, which is used to characterize the path P selected by the kth ant k the degree of excellence; (5) Update: For each value of each parameter, the pheromone update is performed when all K ants complete the traversal task, according to the pheromone increment Δτ ij , update pheromone τ ij , the update rules are as follows: t ij new =(1-p)τ ij old +Δτ ij ; Where p is the pheromone volatility coefficient, which indicates the attenuation degree of the pheromone after each iteration, 0<p<1; Δτ ij represents the pheromone increment, which is used to characterize the pheromone increase of the jth value of the i-th parameter in the search space after each iteration; if the kth ant selects the i-th value of the i-th parameter, then according to the formula Calculate, otherwise Where i = 1, 2, ..., m, j = 1, 2, ..., n, k = 1, 2, ..., K; (6) Record: Record the optimal path P in the current iteration * , the corresponding cost C * and fitness F * ; (7) Iteration: n=n+1, used to indicate that the current iteration number n is increased by 1; (8) Output: Output the final optimal path P * , the corresponding cost C * and fitness F * , and the corresponding target prediction model.
7. A supply chain demand forecasting device, characterized in that: The device comprises: The acquisition module is used to obtain demand impact data of the supply chain; A forecasting module, used for inputting the demand impact data into a trained initial forecasting model, so as to forecast the demand for the supply chain based on the demand impact data through the initial forecasting model, and obtain an initial demand result for the supply chain; An optimization module, configured to optimize the model parameters of the initial prediction model based on the demand impact data and the initial demand result through an ant colony optimization algorithm to obtain a target prediction model; The forecasting module is also used to input the demand impact data into the target forecasting model, so as to perform demand forecasting for the supply chain based on the demand impact data through the target forecasting model to obtain a target demand result for the supply chain.
8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.
9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.
Citation Information
Patent Citations
Load prediction method based on ant colony neural network, system comprising load prediction method and memory comprising load prediction method
CN110188956A
Internet financial credit evaluation method based on PSO-BP neural network
CN112037012A
Credit risk prediction method for optimizing MLP neural network based on locust algorithm
CN112581264A
Module packaging injection molding process parameter prediction method based on machine learning
CN116151133A
Cited By
Fuel optimization scheduling method based on coal supply chain prediction
CN119904026A
Supply chain demand prediction system and method based on artificial intelligence
CN120338866A
A supply chain demand forecasting system and method based on artificial intelligence
CN120338866B
