Supply chain operation planning prediction method and system

Abstract

The invention relates to a supply chain operation planning prediction method and system, and belongs to the technical field of supply chain operation planning. The method comprises the following steps: obtaining related data of goods return, related data of demand quantity and data of demand quantity of the current week, and obtaining a related data sequence of goods return, a related data sequence of demand quantity and a sequence of demand quantity according to the related data of goods return, the related data of demand quantity and the data of demand quantity; according to the goods return related data, the demand quantity related data, the goods return related data sequence, the demand quantity related data sequence and the demand quantity sequence, preprocessing to obtain preprocessed data; and according to the goods return related preprocessing data, calculating to obtain a goods return predicted quantity through a BP neural network, according to the demand quantity related preprocessing data and the demand quantity preprocessing sequence, calculating to obtain a demand quantity predicted value through a prediction model, and according to the goods return predicted quantity and demand quantity prediction, calculating to obtain an inventory predicted quantity through an integration model. According to the invention, inventory prediction is carried out in combination with upstream and downstream data, and the supply chain operation can be planned conveniently.

Description

Technical Field

[0001] The present invention belongs to the technical field of supply chain operation planning, and in particular relates to a supply chain operation planning forecasting method and system. Background Art

[0002] In today's complex and ever-changing business environment, the efficiency and accuracy of supply chain operations planning are crucial to the sustainable development and market competitiveness of enterprises. Traditional supply chain operations planning mainly relies on historical sales data, seasonal trend analysis and management experience. These methods are often unable to cope with the rapid changes in market demand, uncertainty in consumer preferences and supply chain disruption risks, and it is difficult to provide timely and accurate forecast results.

[0003] Especially in terms of return management, with the booming development of e-commerce and changes in consumer shopping habits, the return volume has become one of the important factors affecting inventory management and cost control. Traditional return volume forecasting methods are mostly based on simple statistical models or proportion estimates, ignoring various influencing factors such as market fluctuations, promotional activities, product quality feedback, etc., resulting in inaccurate return forecasts, which in turn causes inventory backlogs or shortages, increasing the company's operating costs.

[0004] At the same time, in terms of demand forecasting, although a variety of forecasting models have been applied to supply chain management, such as time series analysis and regression analysis, these models have limitations in dealing with nonlinearity, long-term dependence, and the impact of emergencies. Especially when faced with massive data in a big data environment, the forecasting ability of a single model is often difficult to meet the needs of high accuracy and high efficiency.

[0005] Finally, inventory calculation is the core link in supply chain operation planning, which is directly related to the cost-effectiveness and service level of the enterprise. Traditional inventory calculation methods are often based on static inventory strategies, ignoring the changes in dynamic factors such as market demand, returns, and supply chain lead time. Therefore, how to dynamically adjust inventory levels based on key parameters such as return volume forecasts, demand forecasts, safety inventory, and supply chain lead time to achieve the dual goals of minimizing costs and maximizing service levels is an urgent problem to be solved in current supply chain management. Summary of the invention

[0006] In order to solve the above problems existing in the prior art, the present invention provides a supply chain operation planning forecasting method and system. The purpose of the present invention can be achieved by the following technical solutions: A supply chain operation planning forecasting method, characterized by comprising: Acquire the return-related data, demand-related data, and demand-related data of the week, add timestamps according to the return-related data, demand-related data, and demand-related data to obtain time-series return-related data, time-series demand-related data, and time-series demand-related data, and store the time-series return-related data, time-series demand-related data, and time-series demand data to obtain a return-related data sequence, a demand-related data sequence, and a demand sequence; Preprocessed data are obtained by calculating the preprocessed data through a preprocessing model according to the return-related data, the demand-related data, the return-related data sequence, the demand-related data sequence, and the demand sequence; the preprocessed data includes return-related preprocessed data, demand-related preprocessed data, and demand preprocessed sequence; The return forecast quantity is calculated through the BP neural network based on the return-related preprocessing data, the demand forecast value is calculated through the prediction model based on the demand-related preprocessing data and the demand preprocessing sequence, and the inventory forecast quantity is calculated through the integration model based on the return forecast quantity and the demand forecast value.

[0007] As a preferred technical solution of the present invention, the return-related data includes the return service praise rate, the price of the returned goods, the product production cycle, the product praise rate, the completeness of the product information, and the sales volume of the week.

[0008] Specifically, the demand-related data include: commodity prices, downstream inventory, and downstream commodity damage.

[0009] Specifically, the preprocessing model includes: A time window is preset, and the return-related data sequence, the demand-related data sequence, and the demand-related sequence are intercepted according to the time window to obtain a return-related data subsequence, a demand-related data subsequence, and a demand subsequence; According to the return-related data, the demand-related data, and the demand subsequence, outlier detection is performed by a box plot method to obtain outliers in the return data, outliers in the demand-related data, and outliers in the demand sequence; Eliminate the abnormal values ​​of the return data, the abnormal values ​​of the demand data, and the abnormal values ​​of the demand sequence to obtain the cleaning return data, the cleaning demand data, and the cleaning demand sequence; Calculate the average value of the return-related data, the demand-related data, and the demand subsequence to obtain the average value of the return-related data, the average value of the demand-related data, and the average value of the demand; According to the average value of the return data, the average value of the demand data, and the average value of the demand, the missing values ​​of the cleaning return data, the cleaning demand data, and the cleaning demand sequence are filled to obtain filled return data, filled demand data, and filled demand sequence; The data related to filling returns, the data related to filling demand quantities, and the sequence of filling demand quantities are normalized to obtain the pre-processed data related to returns, the pre-processed data related to demand quantities, and the pre-processed sequence of demand quantities.

[0010] Specifically, the prediction model includes: Obtaining a first demand forecast value by calculating the pre-processed data related to the demand through a BP neural network; A second demand forecast value is obtained by calculating the demand preprocessing sequence through an ARIMA model; A weight value is obtained, and the demand forecast value is obtained by calculation according to the weight value, the first demand forecast value, and the second demand forecast value.

[0011] Specifically, the weight value calculation process includes: Obtaining a real value of demand in a training set, calculating a first training demand forecast value based on the training set by using the BP neural network, and calculating a second training demand forecast value based on the training set by using the ARIMA model; The error value is calculated by using the MAPE function according to the actual value of the demand, the first training demand prediction value, and the second training demand prediction value, and the weight value is calculated by using the PSO optimization algorithm according to the error value.

[0012] Specifically, the integration model includes: A safety stock quantity and a lead time are preset, and the forecast inventory quantity is calculated based on the safety stock quantity, the lead time, the forecast return quantity, and the forecast demand quantity.

[0013] A supply chain operation planning and forecasting system, comprising: a data acquisition module, a preprocessing module, and a forecasting module; The data acquisition module is used to acquire the return-related data, demand-related data, and demand-related data of the week, add timestamps according to the return-related data, demand-related data, and demand-related data to obtain time-series return-related data, time-series demand-related data, and time-series demand-related data, and store the time-series return-related data, time-series demand-related data, and time-series demand data to obtain return-related data sequence, demand-related data sequence, and demand sequence; The preprocessing module is used to calculate the preprocessed data through the preprocessing model according to the return-related data, the demand-related data, the return-related data sequence, the demand-related data sequence, and the demand sequence; the preprocessed data includes the return-related preprocessed data, the demand-related preprocessed data, and the demand preprocessing sequence; The prediction module is used to calculate the return forecast amount through the BP neural network based on the return-related preprocessing data, calculate the demand forecast value through the prediction model based on the demand-related preprocessing data and the demand preprocessing sequence, and calculate the inventory forecast amount through the integration model based on the return forecast amount and the demand forecast value.

[0014] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the supply chain operation planning and forecasting method as described above is implemented.

[0015] A storage medium containing computer executable instructions, which are used to execute the supply chain operation planning and forecasting method as described above when executed by a computer processor.

[0016] The beneficial effects of the present invention are: (1) Obtaining various influencing factors related to returns, and predicting the return volume through BP neural network based on these factors, solving the problem of low accuracy caused by calculating the historical return volume through simple statistical models or proportion estimation. Among them, BP neural network can automatically learn and extract complex features in data, and has been proven to have excellent prediction performance in many fields. However, its application in return volume prediction is still rare, especially in comprehensive prediction combined with other key elements of the supply chain.

[0017] (2) Obtain data on influencing factors related to demand and historical data on demand, calculate the first demand forecast value through the BP neural network based on the data on influencing factors of demand, calculate the second demand forecast value through the ARIMA model based on the historical data of demand, and calculate the final demand forecast value based on the weight coefficients of the first demand forecast value and the second demand forecast value. Combine these two models, calculate the demand forecast value from the two aspects of influencing factors and short-term dependencies, and assign weights according to the reliability of each forecast result, which is expected to further improve the accuracy and robustness of demand forecasting.

[0018] (3) Inventory calculation is the core link in supply chain operation planning, which is directly related to the cost-effectiveness and service level of the enterprise. According to key parameters such as return volume forecast, demand forecast, safety inventory and supply chain lead time, the inventory level is adjusted dynamically, which solves the problem that traditional inventory calculation methods are often based on static inventory strategies and ignore the changes in dynamic factors such as market demand, return status and supply chain lead time. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0020] Figure 1 It is a flow chart of a supply chain operation planning and forecasting method of the present invention. DETAILED DESCRIPTION

[0021] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0022] See also Figure 1 , a forecasting method for supply chain operations planning, Acquire the return-related data, demand-related data, and demand-related data of the week, add timestamps according to the return-related data, demand-related data, and demand-related data to obtain time-series return-related data, time-series demand-related data, and time-series demand-related data, and store the time-series return-related data, time-series demand-related data, and time-series demand data to obtain a return-related data sequence, a demand-related data sequence, and a demand sequence; Preprocessed data are obtained by calculating the preprocessed data through a preprocessing model according to the return-related data, the demand-related data, the return-related data sequence, the demand-related data sequence, and the demand sequence; the preprocessed data includes return-related preprocessed data, demand-related preprocessed data, and demand preprocessed sequence; The return forecast quantity is calculated through the BP neural network based on the return-related preprocessing data, the demand forecast value is calculated through the prediction model based on the demand-related preprocessing data and the demand preprocessing sequence, and the inventory forecast quantity is calculated through the integration model based on the return forecast quantity and the demand forecast value.

[0023] Specifically, the return-related data includes the return service praise rate, the price of the returned goods, the product production cycle, the product praise rate, the completeness of the product information, and the sales volume of the week.

[0024] Specifically, the demand-related data include commodity prices, downstream inventory, and downstream commodity damage.

[0025] Specifically, the preprocessing model includes: A time window is preset, and the return-related data sequence, the demand-related data sequence, and the demand-related sequence are intercepted according to the time window to obtain a return-related data subsequence, a demand-related data subsequence, and a demand subsequence; According to the return-related data, the demand-related data, and the demand subsequence, outlier detection is performed by a box plot method to obtain outliers in the return data, outliers in the demand-related data, and outliers in the demand sequence; Eliminate the abnormal values ​​of the return data, the abnormal values ​​of the demand data, and the abnormal values ​​of the demand sequence to obtain the cleaning return data, the cleaning demand data, and the cleaning demand sequence; Calculate the average value of the return-related data, the demand-related data, and the demand subsequence to obtain the average value of the return-related data, the average value of the demand-related data, and the average value of the demand; According to the average value of the return data, the average value of the demand data, and the average value of the demand, the missing values ​​of the cleaning return data, the cleaning demand data, and the cleaning demand sequence are filled to obtain filled return data, filled demand data, and filled demand sequence; The data related to filling returns, the data related to filling demand quantities, and the sequence of filling demand quantities are normalized to obtain the pre-processed data related to returns, the pre-processed data related to demand quantities, and the pre-processed sequence of demand quantities.

[0026] Specifically, the prediction model includes Obtaining a first demand forecast value by calculating the pre-processed data related to the demand through a BP neural network; A second demand forecast value is obtained by calculating the demand preprocessing sequence through an ARIMA model; A weight value is obtained, and the demand forecast value is obtained by calculation according to the weight value, the first demand forecast value, and the second demand forecast value.

[0027] In this embodiment, data on influencing factors related to demand and historical data on demand are obtained, a first forecast value of demand is calculated through a BP neural network based on the data on influencing factors of demand, a second forecast value of demand is calculated through an ARIMA model based on the historical data of demand, and the final forecast value of demand is calculated based on the weight coefficients of the first forecast value of demand and the second forecast value of demand. These two models are combined to calculate the forecast value of demand from two aspects: influencing factors and short-term dependencies, and weights are assigned according to the reliability of each forecast result, which is expected to further improve the accuracy and robustness of demand forecasting.

[0028] Specifically, the weight value calculation process includes: Obtaining a real value of demand in a training set, calculating a first training demand forecast value based on the training set by using the BP neural network, and calculating a second training demand forecast value based on the training set by using the ARIMA model; An error value is calculated by using a MAPE function according to the actual value of the demand, the first training demand prediction value, and the second training demand prediction value, and a weight value is calculated by using a PSO optimization algorithm according to the error value; The MAPE function expression is: , in, Y is the error value, n is the number of training sets, At For the t Prediction values, Ft For the t True value, t For variable parameters.

[0029] Specifically, the integration model includes: Preset a safety stock quantity and a lead time, and calculate the forecast inventory quantity according to the safety stock quantity, the lead time, the forecast return quantity, and the forecast demand quantity; The inventory forecast calculation formula is:

[0030] in, Y is the inventory forecast, D is the demand forecast, L is the lead time, S For safety stock, R Forecast quantity for returns.

[0031] In this embodiment, the calculation of inventory is the core link in the supply chain operation planning, which is directly related to the cost-effectiveness and service level of the enterprise. According to the key parameters such as the return volume forecast value, the demand volume forecast value, the safety inventory volume and the supply chain lead time, the inventory level is dynamically adjusted, which solves the problem that the traditional inventory calculation method is often based on a static inventory strategy, ignoring the changes in dynamic factors such as market demand, return situation and supply chain lead time.

[0032] A supply chain operation planning and forecasting system, comprising: a data acquisition module, a preprocessing module, and a forecasting module; The data acquisition module is used to acquire the return-related data, demand-related data, and demand-related data of the week, add timestamps according to the return-related data, demand-related data, and demand-related data to obtain time-series return-related data, time-series demand-related data, and time-series demand-related data, and store the time-series return-related data, time-series demand-related data, and time-series demand data to obtain return-related data sequence, demand-related data sequence, and demand sequence; The preprocessing module is used to calculate the preprocessed data through the preprocessing model according to the return-related data, the demand-related data, the return-related data sequence, the demand-related data sequence, and the demand sequence; the preprocessed data includes the return-related preprocessed data, the demand-related preprocessed data, and the demand preprocessing sequence; The prediction module is used to calculate the return forecast amount through the BP neural network based on the return-related preprocessing data, calculate the demand forecast value through the prediction model based on the demand-related preprocessing data and the demand preprocessing sequence, and calculate the inventory forecast amount through the integration model based on the return forecast amount and the demand forecast value.

[0033] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the supply chain operation planning and forecasting method as described above is implemented.

[0034] A storage medium containing computer executable instructions, which are used to execute the supply chain operation planning and forecasting method as described above when executed by a computer processor.

[0035] The computer storage medium of the embodiment of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device.

[0036] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0037] The program code included on the computer readable medium can be transmitted with any appropriate medium, including but not limited to wireless, electric wire, optical cable, RF, etc., or any suitable combination of the above. The computer program code for performing the operation of the present invention can be written in one or more programming languages ​​or their combinations, and the programming language includes object-oriented programming languages-such as Java, Smalltalk, C++, and also includes conventional procedural programming languages-such as "C" language or similar programming languages. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on the remote computer, or completely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0038] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A supply chain operation planning forecasting method, characterized in that: include: Acquire the return-related data, demand-related data, and demand-related data of the week, add timestamps according to the return-related data, demand-related data, and demand-related data to obtain time-series return-related data, time-series demand-related data, and time-series demand-related data, and store the time-series return-related data, time-series demand-related data, and time-series demand data to obtain a return-related data sequence, a demand-related data sequence, and a demand sequence; Preprocessed data are obtained by calculating the preprocessed data through a preprocessing model according to the return-related data, the demand-related data, the return-related data sequence, the demand-related data sequence, and the demand sequence; the preprocessed data includes return-related preprocessed data, demand-related preprocessed data, and demand preprocessed sequence; The return forecast quantity is calculated through the BP neural network based on the return-related preprocessing data, the demand forecast value is calculated through the prediction model based on the demand-related preprocessing data and the demand preprocessing sequence, and the inventory forecast quantity is calculated through the integration model based on the return forecast quantity and the demand forecast value.

2. The supply chain operation planning and forecasting method according to claim 1, characterized in that: The return-related data include the return service praise rate, return product price, product production cycle, product praise rate, product information completeness, and weekly sales volume.

3. The supply chain operation planning and forecasting method according to claim 1, characterized in that: The demand-related data include commodity prices, downstream inventory, and downstream commodity damage.

4. The supply chain operation planning and forecasting method according to claim 1, characterized in that: The preprocessing model includes: A time window is preset, and the return-related data sequence, the demand-related data sequence, and the demand-related sequence are intercepted according to the time window to obtain a return-related data subsequence, a demand-related data subsequence, and a demand subsequence; According to the return-related data, the demand-related data, and the demand subsequence, outlier detection is performed by a box plot method to obtain outliers in the return data, outliers in the demand-related data, and outliers in the demand sequence; Eliminate the abnormal values ​​of the return data, the abnormal values ​​of the demand data, and the abnormal values ​​of the demand sequence to obtain the cleaning return data, the cleaning demand data, and the cleaning demand sequence; Calculate the average value of the return-related data, the demand-related data, and the demand subsequence to obtain the average value of the return-related data, the average value of the demand-related data, and the average value of the demand; According to the average value of the return data, the average value of the demand data, and the average value of the demand, the missing values ​​of the cleaning return data, the cleaning demand data, and the cleaning demand sequence are filled to obtain filled return data, filled demand data, and filled demand sequence; The data related to filling returns, the data related to filling demand quantities, and the sequence of filling demand quantities are normalized to obtain the pre-processed data related to returns, the pre-processed data related to demand quantities, and the pre-processed sequence of demand quantities.

5. The supply chain operation planning and forecasting method according to claim 1, characterized in that: The prediction model includes: Obtaining a first demand forecast value by calculating the pre-processed data related to the demand through a BP neural network; A second demand forecast value is obtained by calculating the demand preprocessing sequence through an ARIMA model; A weight value is obtained, and the demand forecast value is obtained by calculation according to the weight value, the first demand forecast value, and the second demand forecast value.

6. The supply chain operation planning and forecasting method according to claim 5, characterized in that: The weight value calculation process includes: Obtaining a real value of demand in a training set, calculating a first training demand forecast value based on the training set by using the BP neural network, and calculating a second training demand forecast value based on the training set by using the ARIMA model; The error value is calculated by using the MAPE function according to the actual value of the demand, the first training demand prediction value, and the second training demand prediction value, and the weight value is calculated by using the PSO optimization algorithm according to the error value.

7. The supply chain operation planning and forecasting method according to claim 1, characterized in that: The integrated model includes a preset safety stock and a lead time, and the inventory forecast is calculated based on the safety stock, the lead time, the return forecast, and the demand forecast.

8. A supply chain operation planning and forecasting system, characterized in that: include: Data acquisition module, preprocessing module, prediction module; The data acquisition module is used to acquire the return-related data, demand-related data, and demand-related data of the week, add timestamps according to the return-related data, demand-related data, and demand-related data to obtain time-series return-related data, time-series demand-related data, and time-series demand-related data, and store the time-series return-related data, time-series demand-related data, and time-series demand data to obtain return-related data sequence, demand-related data sequence, and demand sequence; The preprocessing module is used to calculate the preprocessed data through the preprocessing model according to the return-related data, the demand-related data, the return-related data sequence, the demand-related data sequence, and the demand sequence; the preprocessed data includes the return-related preprocessed data, the demand-related preprocessed data, and the demand preprocessing sequence; The prediction module is used to calculate the return forecast amount through the BP neural network based on the return-related preprocessing data, calculate the demand forecast value through the prediction model based on the demand-related preprocessing data and the demand preprocessing sequence, and calculate the inventory forecast amount through the integration model based on the return forecast amount and the demand forecast value.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the supply chain operation planning and forecasting method as described in any one of claims 1-7 is implemented.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the supply chain operation planning and forecasting method as described in any one of claims 1 to 7 when executed by a computer processor.

Images