Supply chain collaborative prediction method and system

Abstract

The invention relates to the technical field of supply chain collaborative prediction, in particular to a supply chain collaborative prediction method and system. The method comprises the steps of obtaining a supply chain data fluctuation degree at a collection moment through a difference between supply chain data of each dimension at the collection moment and historical supply chain data; obtaining a transportation end point position corresponding to the ex-warehouse data in the supply chain data, and obtaining the credibility of the change of the supply chain data at the acquisition moment according to the Euclidean distance between the transportation end points of the ex-warehouse data at the acquisition moment; calculating the importance degree of the collection moment; in the process of acquiring the supply chain data prediction value at the next moment of the acquisition moment by the EWMA algorithm, the weight of each acquisition moment is corrected by using the importance of each acquisition moment to obtain the supply chain collaborative prediction result at the next moment, so that the accuracy of supply chain collaborative prediction is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of supply chain collaborative forecasting, and particularly to a supply chain collaborative forecasting method and system. Background Art

[0002] The inbound volume, outbound volume, and inventory volume of a supply chain warehouse can intuitively reflect the real-time status of the inventory. By monitoring and analyzing the inventory data, the quantity, location, and status of the inventory can be grasped in real time. Each department can intuitively understand the work progress and requirements of other departments, thereby breaking down departmental barriers, ensuring the transparency of the supply chain, and enhancing the collaborative effect among departments.

[0003] In order to monitor the supply chain collaboration effect, the patent application document with the publication number CN119250740A in the prior art discloses a hierarchical collaborative control and warning method, device, equipment, and medium based on supply chain integration. This application monitors and collects the inventory data and contract performance data of the supply chain in real time; obtains the procurement data of the supply chain, combines the inventory data and demand data, and establishes a data analysis model based on machine learning to generate a warning evaluation coefficient; compares and analyzes the warning evaluation coefficient with a pre-set warning evaluation threshold, determines the warning level, issues a warning prompt, and performs maintenance according to the warning prompt.

[0004] Obviously, the above solution mainly identifies anomalies based on the ex-post analysis of data such as inventory, delivery, and procurement, lacks the direct prediction ability for sudden demand changes such as seasonality, and cannot effectively avoid potential anomalies. Moreover, the above solution establishes a data analysis model based on machine learning to generate a warning evaluation coefficient, but does not mention the training data and algorithm details of the model. If the training data cannot include various situations of sudden demand changes such as seasonality, the model may not be able to accurately identify relevant anomalies, and the accuracy is relatively low.

[0005] Therefore, it is urgent to solve the problem of how to accurately predict the supply chain collaboration degree at future moments when there are sudden demand changes such as seasonality in product supply. Summary of the Invention

[0006] To solve the technical problem of how to accurately predict the supply chain collaboration degree at future moments when there are sudden demand changes such as seasonality in product supply, the present invention provides a supply chain collaborative forecasting method and system.

[0007] In a first aspect, the present invention provides a supply chain collaborative forecasting method, adopting the following technical solution: A supply chain collaborative forecasting method includes the steps: Obtain the supply chain data of each dimension of the product at each collection moment; obtain the degree of supply chain data fluctuation at this collection moment through the difference between the supply chain data of each dimension at the collection moment and the historical supply chain data; obtain the transport end position corresponding to the outbound data in the supply chain data, and record the Euclidean distance between the transport ends of two outbound data in the historical supply chain data at any collection moment as a position index at this collection moment. According to the difference between each position index at the collection moment and the mean value of the position index, obtain the credibility of the supply chain data change at this collection moment; calculate the importance at the collection moment, and the importance is positively correlated with the degree of supply chain data fluctuation and the credibility of supply chain data change at this collection moment; in the process of obtaining the predicted value of the supply chain data at the next moment by the EWMA algorithm, use the importance at each collection moment to correct the weight of each collection moment to obtain the supply chain collaborative prediction result at the next moment.

[0008] The present invention can predict various factors including seasonal or sudden demands by predicting the collaborative degree of the supply chain, so as to effectively identify potential abnormal situations of the collaborative degree and make timely handling. In this process, the present invention considers that when the EWMA algorithm obtains the predicted value of the supply chain at the next moment and sets corresponding weights according to the chronological order of the collection times at the historical collection moments, it may ignore the influence of seasonal changes at some collection moments; based on this, the present invention sets weights for each collection moment by quantifying the possibility of seasonal changes at each collection moment, so that the supply chain collaborative prediction result at the next moment can continue the seasonal characteristics, and thus accurately obtain the predicted value of the supply chain data at the next moment to evaluate the supply chain collaborative prediction result. On this basis, the present invention also considers that the characteristics of occasional orders are relatively close to the seasonal characteristics and will affect the prediction accuracy at the next moment; based on this, the present invention corrects the possibility of seasonal changes at the collection moment by obtaining the distribution of the outbound endpoints of the product, which can effectively distinguish the data fluctuations driven by real market demands from the fluctuations caused by occasional factors, accurately obtain the corrected weight values at each collection moment, and thus effectively improve the accuracy of the supply chain collaborative prediction result at the next moment.

[0009] According to a supply chain collaborative prediction method provided by the present invention, the obtaining the supply chain data of each dimension of the product at each collection moment includes: obtaining the inbound quantity, outbound quantity and inventory quantity of the warehouse, and the outbound data of the warehouse each time at each collection moment, and obtaining the supply chain data of each dimension after preprocessing.

[0010] The present invention considers that there may be problems such as inconsistent dimensions and missing data among the supply chain data of different dimensions. Therefore, the data quality is improved through preprocessing to prepare for supply chain collaborative prediction.

[0011] A supply chain collaborative forecasting method provided by the present invention, obtaining the degree of fluctuation of supply chain data at the collection moment includes: presetting the length of historical supply chain data at each collection moment; recording the average value of the absolute values of the differences between the supply chain data of each dimension at the collection moment and the historical supply chain data as the fluctuation index of this dimension; and performing cumulative sum normalization processing on the fluctuation indexes of all dimensions at this collection moment to obtain the degree of fluctuation of supply chain data at this collection moment.

[0012] By analyzing the fluctuation of supply chain data in the short term, the present invention can accurately obtain the degree of fluctuation of supply chain data at the collection moment. The higher the degree of fluctuation of supply chain data at the collection moment, the more significantly the supply chain is affected by external factors, the higher the mutability of product flow in the warehouse, and the higher the possibility of seasonal changes.

[0013] A supply chain collaborative forecasting method provided by the present invention takes the longitude and latitude of the transportation destination corresponding to the outbound data as the location of this transportation destination.

[0014] A supply chain collaborative forecasting method provided by the present invention, the credibility of the change of supply chain data at this collection moment satisfies the relational expression: ; is the credibility of the change of supply chain data at the collection moment, is the number of outbound data in the historical supply chain data at the collection moment, is the Euclidean distance between the th and the th transportation destinations of outbound data in the historical supply chain data at the collection moment, is the average value of the position indexes in the historical supply chain data at the collection moment, is the exponential function with base e.

[0015] The present invention provides an accurate calculation formula for the credibility of the change of supply chain data at the collection moment. By analyzing the Euclidean distance between the outbound destination positions, the distribution of outbound data can be accurately obtained. Thus, the authenticity of the supply chain data fluctuation can be verified according to the closeness between the distribution of outbound data and the actual market sales network layout, and the degree of change of supply chain data can be accurately obtained.

[0016] A supply chain collaborative forecasting method provided by the present invention, calculating the importance degree of the acquisition moment includes: recording the product of the fluctuation degree of the supply chain data at each acquisition moment and the credibility of the change of the supply chain data as the degree of data change at this moment; normalizing the ratio of the degree of data change at the acquisition moment to the degree of data change at the previous acquisition moment to obtain the importance degree of this acquisition moment.

[0017] A supply chain collaborative forecasting method provided by the present invention, in the process of obtaining the predicted value of the supply chain data at the next moment by the EWMA algorithm, using the importance degree of each acquisition moment to correct the weight of each acquisition moment includes: taking the product of the weight of each acquisition moment and the importance degree as the weight index of this acquisition moment; taking the ratio of the weight index of the acquisition moment to the sum value of the weight indexes of the historical supply chain data acquisition moments as the corrected weight value of this acquisition moment.

[0018] A supply chain collaborative forecasting method provided by the present invention, using the importance degree of each acquisition moment to correct the weight of each acquisition moment to obtain the supply chain collaborative forecasting result at the next moment includes: using the corrected weight value of the acquisition moment to perform weighted average on the supply chain data of each dimension corresponding to the historical supply chain data of this acquisition moment to obtain the predicted value of the supply chain data of each dimension at the next moment; obtaining the supply chain collaborative forecasting result at the next moment according to the predicted value of the supply chain data of each dimension at the next moment.

[0019] A supply chain collaborative forecasting method provided by the present invention, obtaining the supply chain collaborative forecasting result at the next moment includes: after obtaining the sum value of the predicted value of the incoming stock quantity and the predicted value of the inventory quantity at the next moment, taking the absolute value obtained by subtracting the predicted value of the outgoing stock quantity from the sum value as the supply chain collaborative forecasting result at the next moment.

[0020] The present invention takes into account that when the degree of collaboration is relatively high, the difference between the sum value of the incoming stock quantity and the inventory quantity of the supply chain and the outgoing stock quantity is relatively small. Therefore, according to the difference between the sum value of the predicted value of the incoming stock quantity and the predicted value of the inventory quantity at the next moment and the predicted value of the outgoing stock quantity, the supply chain collaborative forecasting result at the next moment can be accurately obtained.

[0021] In the second aspect, the present invention provides a supply chain collaborative forecasting system, adopting the following technical solution: A supply chain collaborative forecasting system includes: a processor and a memory, and the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned supply chain collaborative forecasting method is implemented.

[0022] By adopting the above technical solution, generating a computer program for the above-mentioned supply chain collaborative forecasting method and storing it in the memory to be loaded and executed by the processor, thereby manufacturing a terminal device according to the memory and the processor, which is convenient to use.

[0023] The present invention has the following technical effects: Based on the above technical solution, for a supply chain collaborative prediction method and system provided by the present invention, when obtaining the supply chain collaborative prediction result at the next moment, by predicting the collaborative degree of the supply chain, sudden demands such as seasonal demands can be predicted, so as to effectively identify potential abnormal collaborative degrees. In this process, the present invention sets weights for each acquisition moment by quantifying the possibility that each acquisition moment may be a seasonal change, so that the supply chain collaborative prediction result at the next moment can continue the seasonal characteristics, thereby accurately obtaining the predicted value of the supply chain data at the next moment to evaluate the supply chain collaborative prediction result. On this basis, the present invention also corrects the possibility of seasonal change at the acquisition moment by obtaining the distribution of the end points of product outbound, which can effectively distinguish the data fluctuations driven by real market demands from the fluctuations caused by accidental factors, accurately obtain the corrected weight values of each acquisition moment, and thus effectively improve the accuracy of the supply chain collaborative prediction result at the next moment. Brief Description of the Drawings

[0024] Figure 1 It is a schematic flow chart in a supply chain collaborative prediction method provided by an embodiment of the present invention. Detailed Embodiments

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention.

[0026] In order to solve the technical problem of how to accurately predict the collaborative degree of the supply chain at a future moment when there are sudden demand changes such as seasonality in product supply, an embodiment of the present invention discloses a supply chain collaborative prediction method. The method analyzes the changes in supply chain data in the product supply chain to obtain its fluctuation degree, and determines its weight by calculating the credibility of the fluctuation degree after correction, which can effectively reduce the impact of sudden demand changes such as seasonality on the supply chain data prediction result, improve the accuracy of supply chain collaborative prediction, and facilitate enterprises to coordinate in a timely manner.

[0027] Specifically, please refer to Figure 1 as shown in Figure 1 It is a schematic flow chart in a supply chain collaborative prediction method provided by an embodiment of the present invention. The method specifically includes the following steps: S1: Obtain the supply chain data of each dimension of the product at each acquisition moment.

[0028] Among them, the dimensions of the product supply chain data may include the inbound volume, outbound volume, and inventory volume of the warehouse.

[0029] It should be noted that the inbound quantity, outbound quantity, and inventory quantity in the product supply chain data can directly reflect the changes in the product supply and demand relationship. Before supply chain collaborative forecasting, all supply chain data can be separately entered into the ERP system, and databases for different types of supply chain data can be established in the ERP system. After entering the supply chain data of the product, the ERP system will associate and store the ID of the product with the supply chain data of the product.

[0030] For the sake of easy understanding, in the embodiment of the present invention, the supply chain collaboration degree of any product in a warehouse is predicted, but it does not mean that the embodiment of the present invention is only limited to this.

[0031] Specifically, when obtaining the supply chain data of each dimension of the product, the warehouse terminal can periodically access the enterprise's ERP system through methods such as SQL query and API connection, and send an authentication information request to the inventory query API of the ERP system at each collection moment. The request includes at least the ID authentication, access token, access target, etc. that are uniquely corresponding to each type of product. After receiving the authentication information request, the ERP system verifies the authentication information, and after successful verification, returns the corresponding supply chain data according to the access target, that is, the inbound quantity, outbound quantity, and inventory quantity of the product in the warehouse.

[0032] Among them, the period for obtaining the supply chain data can be set to once per hour; the period can be specifically set according to actual needs, and the embodiment of the present invention does not limit it too much here.

[0033] Exemplarily, in the embodiment of the present invention, the supply chain data of each dimension of the product is obtained at each collection moment, including: obtaining the inbound quantity, outbound quantity, and inventory quantity of the warehouse at each collection moment, and preprocessing to obtain the supply chain data of each dimension.

[0034] Among them, the preprocessing can be missing data interpolation, data standardization processing to eliminate the dimension, etc., which can be specifically set according to actual needs.

[0035] It can be understood that the inbound quantity, outbound quantity, and inventory quantity of the warehouse are the dimensions of the supply chain data. Obtaining the supply chain data of each dimension of the product is to obtain the inbound quantity, outbound quantity, and inventory quantity of the product. The inbound quantity, outbound quantity, and inventory quantity obtained at each collection moment are the supply chain data at that collection moment.

[0036] After obtaining the supply chain data of the product at the current warehouse terminal, the supply chain collaboration status of the products in the current warehouse can be predicted by analyzing the supply chain data of the product.

[0037] The Exponentially Weighted Moving Average (EWMA) is a time series forecasting method. This algorithm performs a weighted average by assigning different weights to historical data to achieve forecasting, and it has high real-time performance and adaptability. Therefore, in the embodiments of the present invention, EWMA is used to predict the degree of supply chain collaboration.

[0038] S2: Obtain the degree of supply chain data fluctuation at the collection moment by calculating the difference between the supply chain data of each dimension at the collection moment and the historical supply chain data.

[0039] It should be noted that during the normal product supply process of an enterprise, the degree of collaboration in the product flow within the warehouse is usually high, and all links of the supply chain can operate efficiently and stably, that is, the data of each dimension is in a relatively stable state during local time periods. When there are sudden product demands such as seasonal demands, the data of each dimension will show obvious fluctuations in the short term. For example, during the peak season of seasonal demand (such as the peak demand season for heating equipment in winter), the outbound volume will increase sharply. At the same time, in order to meet the demand, the inbound volume and inventory will also change significantly, and the stable state will be changed, thus affecting the degree of supply chain collaboration.

[0040] Based on this, before predicting the supply chain forecast value at the next moment of the current collection moment through the EWMA algorithm in the embodiments of the present invention, it is possible to first obtain the data fluctuation situation at each collection moment, and obtain the possibility of being affected by external factors such as seasonal changes at each collection moment, so that the weight for calculating the supply chain forecast value at the next moment can be obtained based on the data fluctuation situation at the collection moment.

[0041] It can be understood that the demand forecast of the supply chain has requirements for agility and needs to adjust the replenishment strategy in real time according to the collaborative forecast result. Therefore, the embodiments of the present invention adopt short-term fluctuation analysis, which can capture real market demand signals more accurately and quickly.

[0042] Exemplarily, in the embodiments of the present invention, obtaining the degree of supply chain data fluctuation at the collection moment includes: presetting the length of the historical supply chain data at each collection moment; recording the average absolute value of the difference between the supply chain data of each dimension at the collection moment and the historical supply chain data as the fluctuation index of this dimension; and normalizing the cumulative sum of the fluctuation indexes of all dimensions at the collection moment to obtain the degree of supply chain data fluctuation at the collection moment.

[0043] Among them, the length of the historical supply chain data at each collection moment is the number of historical collection moments at each collection moment; the length of the historical supply chain data at each collection moment can be set to 10, and can be specifically set according to actual needs.

[0044] Exemplarily, when obtaining the historical supply chain data at the current collection moment, the supply chain data at the current collection moment can be used as the first historical collection moment at the current collection moment, and other collection moments can be obtained according to the interval from the current collection moment, so as to obtain the historical collection moments at the current moment composed of the current collection moment and other collection moments, and the supply chain data corresponding to the historical collection moments is used as the historical supply chain data at the current collection moment.

[0045] For the convenience of understanding, an embodiment of the present invention provides a relational expression for calculating the fluctuation degree of the supply chain data at the collection moment: ; is the fluctuation degree of the supply chain data at the collection moment, is the number of dimensions of the supply chain data, is the number of historical supply chain data at the collection moment, is the data value of the th dimension at the collection moment, is the data value of the th dimension at the collection moment,

[0046] In the above formula, the fluctuation degree of the supply chain data at the collection moment is the fluctuation degree of the historical supply chain data at that collection moment. The greater the fluctuation degree of the historical supply chain data at the collection moment, the lower the operating stability in the historical period of the collection moment, and the more unstable the changes of the data in each dimension.

[0047] represents the fluctuation index of the th dimension at the collection moment. The larger this value, the greater the change in the data of the th dimension at the collection moment relative to the short-term historical collection period corresponding to the historical supply chain data at the th dimension. The greater the possibility that the supply chain is affected by external factors such as seasonality, the worse the operating state stability, and the more obvious the changes in the data of each link due to the impact on the coordinated operation of each link.

[0048] Based on the above steps, the degree of supply chain data fluctuation at each collection moment can be obtained. The degree of supply chain data fluctuation at each collection moment can reflect the possibility of being impacted by external factors such as seasonality at each collection moment. External factors such as seasonality may continue for a relatively long period in the future. Therefore, when predicting the collaboration degree in the future period based on the EWMA algorithm, corresponding weights need to be set according to the possibility of being impacted by external factors such as seasonality at each collection moment, so as to accurately obtain the collaborative prediction result for the future period.

[0049] However, in addition to seasonal demand, the production activities of enterprises may also cause large fluctuations in supply chain data in the short term. For example, a large number of orders are purchased for company activities, small merchants stockpile goods, etc. Obviously, this situation does not have sustainability in the future period. Therefore, setting weights for each collection moment only based on the fluctuation of historical short-term data may cause the final collaborative prediction result to tend to accidental events without sustainability, reducing the accuracy of supply chain collaborative prediction.

[0050] Based on this, the embodiments of the present invention further verify the authenticity of the degree of supply chain data fluctuation at the collection moment by continuing to execute the following steps.

[0051] S3: Obtain the transportation end positions corresponding to the outbound data in the supply chain data, and obtain the credibility of the change of the supply chain data at this collection moment through the Euclidean distance between the transportation end points of the outbound data in the historical supply chain data at the collection moment.

[0052] It should be noted that the outbound end point in the outbound data of each product can reflect the distribution of product demand. In the case of a large degree of fluctuation, if the position distribution of the outbound end points is relatively scattered, it indicates that the greater the possibility that the degree of supply chain data fluctuation at the collection moment is due to real market demand, and the product demand in multiple regions increases; on the contrary, if the position distribution of the outbound end points is relatively concentrated, it indicates that the greater the possibility that the degree of supply chain data fluctuation at the collection moment is due to local stockpiling or large orders.

[0053] Based on this, the embodiments of the present invention can verify the degree of supply chain data fluctuation at each collection moment through the outbound end point positions of product outbound data, so as to accurately obtain the credibility that the degree of fluctuation at each collection moment is due to seasonal changes.

[0054] Exemplarily, in the embodiments of the present invention, the outbound data when the product is shipped each time can be obtained through the ERP system, where the outbound data at least includes the ID of the product and the outbound transportation end point when placing an order.

[0055] Among them, the longitude and latitude of the transportation end point corresponding to the outbound data can be used as the transportation end point position of this outbound data.

[0056] It is understandable that the longitude and latitude of the transportation destination are dimensionless data, and the outbound data is directly related to the user's order. Therefore, when analyzing the outbound data, the number of outbound data in the historical supply chain data at the current collection moment is the number of outbound data in the historical period corresponding to the historical supply chain data at the current collection moment, and each outbound data corresponds to an outbound transportation destination.

[0057] Exemplarily, in the embodiments of the present invention, when calculating the credibility of the change in the supply chain data at the collection moment, the transportation destination position corresponding to the outbound data in the supply chain data can be obtained, and the Euclidean distance between the transportation destinations of two outbound data in the historical supply chain data at any collection moment is recorded as a position index at this collection moment. Based on the difference between each position index at the collection moment and the mean value of the position indexes, the credibility of the change in the supply chain data at this collection moment is obtained.

[0058] It is understandable that the Euclidean distance between every two outbound data transportation destinations is one of the position indexes at this collection moment, and finally the mean value of the position indexes can be obtained based on multiple position indexes at this collection moment. When analyzing the Euclidean distance between the outbound data transportation destinations, only the Euclidean distance between the transportation destination of each outbound data and the transportation destinations of other outbound data can be calculated.

[0059] For the sake of easy understanding, the embodiments of the present invention provide a relational expression for the credibility of the change in the supply chain data at the collection moment, which is specifically as follows: ; is the credibility of the change in the supply chain data at the collection moment, is the number of outbound data in the historical supply chain data at the collection moment, is the Euclidean distance between the th and th outbound data transportation destinations in the historical supply chain data at the collection moment, is the mean value of the position indexes in the historical supply chain data at the collection moment, is the exponential function with e as the base.

[0060] In the above formula, represents one of the position indexes at the collection moment.

[0061] represents the difference between one of the position indexes at the collection moment and the mean value of the position indexes. is the The standard deviation of the Euclidean distance between the shipping destinations of the outbound data in the historical supply chain data at the collection time. The larger this value is, it indicates that in the historical collection period corresponding to the historical supply chain data at the collection time, the position distribution of the shipping destinations of the outbound data is more dispersed, that is, the shipping destinations are widely distributed in different regions. Therefore, the higher the credibility that the change in the supply chain data at the collection time is driven by the real market demand, and the lower the possibility caused by accidental factors.

[0062] After obtaining the credibility of the change in the supply chain data at each collection time based on the above steps, the fluctuation degree at the collection time can be corrected according to the credibility of the change in the supply chain data at each collection time, so as to accurately obtain the possibility that the fluctuation at each collection time is a seasonal change. The higher the possibility that the fluctuation at the collection time is a seasonal change, the higher the corresponding importance.

[0063] S4: Calculate the importance of the collection time; use the importance of each collection time to correct the weight of each collection time to obtain the corrected weight value of each collection time.

[0064] Among them, the importance is positively correlated with the fluctuation degree of the supply chain data at this collection time and the credibility of the change in the supply chain data.

[0065] It should be noted that there are accidental factors such as merchants' stockpiling and large one-time orders in the supply chain, which cause the data fluctuation degree not to truly reflect the state of the supply chain. If the prediction model biases towards this kind of change, it will lead to the prediction result deviating from the actual situation. Based on this, the embodiment of the present invention can use the credibility of the change in the supply chain data to correct the degree of data change, and can filter out the unreliable fluctuations caused by accidental factors.

[0066] Exemplarily, in the embodiment of the present invention, calculating the importance of the collection time includes: recording the product of the fluctuation degree of the supply chain data at each collection time and the credibility of the change in the supply chain data as the degree of data change at this time; normalizing the ratio of the degree of data change at the collection time to the degree of data change at its previous collection time to obtain the importance of this collection time.

[0067] Among them, the larger the ratio of the degree of data change at the current collection time to the degree of data change at its previous collection time, it indicates that the degree of change at the current collection time is greater than that at its previous collection time, the change in the supply chain data at the current collection time is more significant, there are large changes in the market demand or supply chain links, and the corresponding importance is also higher.

[0068] It can be understood that the higher the importance of the current collection moment, the greater the possibility that the supply chain demand at the current collection moment will continue to occur at the next moment in the future of the current collection moment, and the higher the importance in the supply chain collaborative prediction at the next moment. Based on this, the corrected weight value of each collection moment can be accurately obtained, so that the predicted value of the next moment can be accurately obtained based on the corrected weight values of each collection moment in the EWMA algorithm.

[0069] Exemplarily, in the embodiment of the present invention, the weights of each collection moment are corrected by using the importance of each collection moment, including: taking the product of the weight of each collection moment and the importance as the weight index of this collection moment; taking the ratio of the weight index of the collection moment to the sum of the weight indexes of the historical supply chain data collection moments as the corrected weight value of this collection moment.

[0070] Among them, the weight of each collection moment is a preset value in the EWMA algorithm. Generally, the closer the collection moment is to the current collection moment, the greater the weight. Specifically, it can be set according to actual needs, and the embodiment of the present invention does not limit it too much here.

[0071] S5: In the process of the EWMA algorithm obtaining the predicted value of the supply chain data at the next moment of the collection moment, the corrected weight values of each collection moment are used to obtain the supply chain collaborative prediction result at the next moment.

[0072] Exemplarily, in the embodiment of the present invention, the corrected weight values of each collection moment are used to obtain the supply chain collaborative prediction result at the next moment, including: using the corrected weight values of the collection moment to perform weighted averaging on the supply chain data of each dimension corresponding to the historical supply chain data of this collection moment to obtain the predicted values of the supply chain data of each dimension at the next moment; obtaining the supply chain collaborative prediction result at the next moment according to the predicted values of the supply chain data of each dimension at the next moment.

[0073] Specifically, when obtaining the predicted value of the supply chain data of any dimension at the next moment of the current collection moment, the corrected weight value of the collection moment corresponding to the historical supply chain data of the current collection moment and the data of this dimension can be used in the EWMA algorithm to calculate the predicted value of the supply chain data of this dimension at the next moment.

[0074] An example is given to illustrate the calculation of the predicted value of the incoming stock volume at the next moment of the current collection moment: When the current collection moment is the 4th collection moment and calculating the predicted value of the incoming stock volume at the 5th collection moment, the incoming stock volume at the 1st collection moment in the historical supply chain data at the current collection moment can be used as the smoothed value at the 1st collection moment. Respectively obtain the product of the corrected weight value at the 2nd collection moment and the incoming stock volume, and the product of the corrected weight value at the 1st collection moment and the smoothed value. The sum of the two products is used as the smoothed value at the 2nd collection moment; and so on. Finally, obtain the product of the corrected weight value at the 4th collection moment and the incoming stock volume, and the product of the corrected weight value at the 3rd collection moment and the smoothed value. The sum of the two products is used as the smoothed value at the 4th collection moment, and the smoothed value at the 4th collection moment is the predicted value of the incoming stock volume at the 5th collection moment.

[0075] After obtaining the predicted values of the supply chain data for each dimension at the next moment based on the above steps, the supply chain collaborative prediction result at the next moment can be obtained according to the predicted values of the supply chain data for each dimension.

[0076] It should be noted that the dynamic balance of the product inventory, incoming stock volume and outgoing stock volume in the warehouse can measure the supply chain collaborative effect. When the degree of collaboration in the product flow process is relatively high, the sum of the inventory volume and the incoming stock volume is relatively close to the outgoing stock volume. At this time, the products in the current warehouse will neither cause supply interruption due to inventory shortage nor occupy a large amount of funds and warehouse space due to excessive inventory backlog, thus increasing costs in vain.

[0077] Exemplarily, in the embodiment of the present invention, obtaining the supply chain collaborative prediction result at the next moment includes: after obtaining the sum of the predicted value of the incoming stock volume and the predicted value of the inventory volume at the next moment, taking the absolute value obtained by subtracting the predicted value of the outgoing stock volume from the sum as the supply chain collaborative prediction result at the next moment.

[0078] It can be understood that if the supply chain collaborative prediction result at the next moment is too large, it indicates that there may be inventory shortage or excessive backlog of products in the current warehouse. At this time, a prompt needs to be sent in time for the staff to handle.

[0079] Exemplarily, in the embodiment of the present invention, after obtaining the supply chain collaborative prediction result at the next moment, it can also: after the supply chain collaborative prediction result at the next moment is greater than the preset maximum allowable threshold, send out a prompt externally.

[0080] Among them, the maximum allowable threshold can be set to 100; the maximum allowable threshold can be specifically set according to actual needs, and the embodiment of the present invention does not limit it too much here.

[0081] It can be seen that in the embodiments of the present invention, when obtaining the supply chain collaborative prediction result, the supply chain data of each dimension of the product can be obtained at each collection moment; the degree of fluctuation of the supply chain data at the collection moment can be obtained through the difference between the supply chain data of each dimension at the collection moment and the historical supply chain data; the transportation end position corresponding to the outbound data in the supply chain data is obtained, and the Euclidean distance between the transportation ends of two outbound data in the historical supply chain data at any collection moment is recorded as a position index at this collection moment. According to the difference between each position index at the collection moment and the average value of the position index, the credibility of the change of the supply chain data at this collection moment is obtained; the importance degree at the collection moment is calculated, and the importance degree is positively correlated with the degree of fluctuation of the supply chain data at this collection moment and the credibility of the change of the supply chain data; in the process of obtaining the predicted value of the supply chain data at the next moment by the EWMA algorithm, the importance degree of each collection moment is used to correct the weight of each collection moment, so as to obtain the supply chain collaborative prediction result at the next moment, effectively improving the accuracy of the supply chain collaborative prediction.

[0082] The embodiments of the present invention also disclose a supply chain collaborative prediction system, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a supply chain collaborative prediction method provided by the present invention is implemented.

[0083] The above system also includes other components well-known to those skilled in the art, such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be elaborated here.

[0084] In the present invention, the foregoing memory can be any tangible medium that contains or stores a program, and the program can be used by or in combination with an instruction execution system, apparatus, or device.

[0085] The above are all the preferred embodiments of the present invention. The protection scope of the present invention is not limited by this. Therefore, all equivalent changes made according to the structure, shape, and principle of the present invention should be covered within the protection scope of the present invention.

Claims

1. A supply chain collaborative forecasting method, characterized in that: include: Obtain supply chain data for each dimension of the product at each collection moment; The fluctuation degree of the supply chain data at the time of collection is obtained by comparing the difference between the supply chain data of each dimension at the time of collection and the historical supply chain data; Obtain the transport destination location corresponding to the outbound data in the supply chain data, record the Euclidean distance between the transport destinations of two outbound data in the historical supply chain data at any collection time as a location index at the collection time, and obtain the credibility of the supply chain data change at the collection time based on the difference between each location index at the collection time and the location index mean; Calculate the importance of the collection moment, which is positively correlated with the fluctuation degree of supply chain data at the collection moment and the credibility of supply chain data changes; In the process of obtaining the supply chain data forecast value at the next moment after the collection moment using the EWMA algorithm, the weight of each collection moment is corrected using the importance of each collection moment to obtain the supply chain collaborative forecast result at the next moment.

2. A supply chain collaborative forecasting method according to claim 1, characterized in that: The supply chain data of each dimension of the product is obtained at each collection moment, including: At each collection moment, the warehouse's incoming, outgoing and inventory quantities, as well as the warehouse's outgoing data each time, are obtained, and after preprocessing, supply chain data of various dimensions are obtained.

3. A supply chain collaborative forecasting method according to claim 1, characterized in that: The fluctuation degree of the supply chain data at the collection time is obtained, including: The length of historical supply chain data at each collection moment is preset; the absolute value mean of the difference between the supply chain data of each dimension at the collection moment and the historical supply chain data is recorded as the volatility index of the dimension; the volatility indexes of all dimensions at the collection moment are accumulated and normalized to obtain the volatility degree of the supply chain data at the collection moment.

4. A supply chain collaborative forecasting method according to claim 1, characterized in that: The longitude and latitude of the transport destination corresponding to the outbound data are used as the transport destination location.

5. A supply chain collaborative forecasting method according to claim 1, characterized in that: The credibility of the supply chain data change at the time of collection satisfies the relationship: ； For the The credibility of changes in supply chain data at the time of collection, For the The number of outbound data in the historical supply chain data at the time of collection, For the The first historical supply chain data at the time of collection and The Euclidean distance between the transport destinations of outbound data, For the The mean value of the location indicator in the historical supply chain data at the time of collection, is an exponential function with base e.

6. A supply chain collaborative forecasting method according to claim 1, characterized in that: The calculating the importance of the collection time comprises: The product of the fluctuation degree of supply chain data at each collection moment and the credibility of supply chain data change is recorded as the data change degree at that moment; the ratio of the data change degree of the collection moment and its previous collection moment is normalized to obtain the importance of the collection moment.

7. A supply chain collaborative prediction method according to claim 1, characterized in that In the process of obtaining the supply chain data forecast value at the next moment after the collection moment using the EWMA algorithm, the weight of each collection moment is corrected using the importance of each collection moment, including: The product of the weight and importance of each collection moment is taken as the weight index of the collection moment; the ratio of the weight index of the collection moment to the weight index and value of the historical supply chain data collection moment is taken as the corrected weight value of the collection moment.

8. A supply chain collaborative forecasting method according to claim 7, characterized in that: The weight of each collection moment is modified by using the importance of each collection moment to obtain the supply chain collaborative prediction result at the next moment, including: The supply chain data of each dimension corresponding to the historical supply chain data at the collection time is weighted averaged using the revised weight value at the collection time to obtain the supply chain data forecast value of each dimension at the next time; The supply chain collaborative prediction result at the next moment is obtained based on the supply chain data prediction values ​​of each dimension at the next moment.

9. A supply chain collaborative forecasting method according to claim 2, characterized in that: The supply chain collaborative forecasting result obtained at the next moment includes: After obtaining the sum of the incoming quantity forecast value and the inventory quantity forecast value at the next moment, the absolute value of the outgoing quantity forecast value is subtracted from the sum value as the supply chain collaborative forecast result at the next moment.

10. A supply chain collaborative forecasting system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a supply chain collaborative forecasting method according to any one of claims 1-9 is implemented.

Images