Automobile industry supply chain coordination method based on knowledge graph

Abstract

The invention relates to the technical field of data processing, in particular to an automobile industry supply chain coordination method based on a knowledge graph. The method comprises the following steps: determining a sudden change moment of a warehouse-out amount of a commodity according to the warehouse-out amount of the commodity of the automobile within a preset time period before a historical moment; matching the sudden change moments of the first commodity and the second commodity, and determining a correlation value between the first commodity and the second commodity according to a matching result; inputting the ex-warehouse amount of the commodity at the historical moment into a pre-constructed isolated forest model, and obtaining an abnormal score value corresponding to the ex-warehouse amount; and according to the correlation value between the target commodity and the other commodities and the abnormal score values of the target commodity and the other commodities at the historical moment, obtaining the predicted warehouse-out quantity of the target commodity at the next moment so as to coordinate the supply quantity of the target commodity from the supply chain according to the predicted warehouse-out quantity. According to the technical scheme, management of the supply chain of the automobile commodities can be effectively realized.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method for coordinating supply chains in the automotive industry based on a knowledge graph. Background Art

[0002] Sellers or stockists of auto parts can store auto products from the supply chain and ship the auto products to their customers' locations based on the purchase orders they receive from their customers. By managing the inventory of auto products from the supply chain, such as the inventory of auto engines, transmissions, tires, brake pads and other parts, they can effectively reduce their own inventory costs.

[0003] In order to achieve supply chain management, a Chinese patent application document with publication number CN111523841A provides an automobile parts supply chain management system, including: damaged automobile parts generated during the production process are transported to the maintenance department for maintenance; relevant information about automobile parts stored in the warehouse is uploaded to the database; relevant information about automobile parts in the database can be queried by the sales department, the sales department contacts dealers, and sells automobile parts in the warehouse to dealers. The dealers feed back information about the automobile parts they need to purchase to the sales department, the sales department places an order in the database, and the database sends the order information to the warehouse; the warehouse transports the automobile parts purchased by the dealer to the dealer's receiving location through logistics; the maintenance plant purchases the automobile parts needed to repair the vehicle from the dealer for automobile maintenance; the assembly plant sends the required automobile parts information to the warehouse and places an order to purchase automobile parts with the warehouse; the warehouse transports the automobile parts purchased by the assembly plant to the assembly plant through logistics.

[0004] In the related art, it is difficult to ensure effective management of the inventory of automobile products such as automobile parts by preparing parts according to received orders, and therefore it is difficult to achieve effective management of the supply chain of automobile products. Summary of the invention

[0005] In order to overcome the problem that it is difficult to effectively manage the supply chain of automobile products in related technologies, the present application provides an automobile industry supply chain coordination method based on a knowledge graph, including: obtaining the outbound quantity of automobile products from a pre-established knowledge graph, determining the mutation value of the outbound quantity of the products at the historical moment according to the outbound quantity of the products in a preset time period before the historical moment, and taking the historical moment when the mutation value is greater than the preset mutation threshold as the mutation moment of the product; matching the mutation moments of the first product and the second product, and determining the correlation value between the first product and the second product according to the matching result; the first product and the second product are any two commodities among the various commodities of automobiles; input the shipment quantity of the commodities at a historical moment into a pre-built isolation forest model to obtain an abnormality score value corresponding to the shipment quantity output by the isolation forest model; the abnormality score value is used to characterize the degree of isolation of the shipment quantity; for a target commodity among the various commodities of automobiles, according to the correlation value between the target commodity and other commodities, the abnormality score value of the target commodity at a historical moment and the abnormality score value of other commodities at a historical moment, obtain the predicted shipment quantity of the target commodity at the next moment; according to the predicted shipment quantity of the target commodity at the next moment, coordinate the supply quantity of the target commodity from the supply chain.

[0006] In this way, the mutation moment of the goods is determined according to the shipment quantity of automobile goods at historical moments, and the correlation value between the goods is determined through the correlation between different goods at the mutation moment. The shipment quantity of automobile goods can be predicted according to the correlation between different goods. Therefore, a more accurate prediction of the shipment quantity can be obtained, thereby more accurately coordinating the supply quantity of goods from the supply chain.

[0007] Optionally, the sudden change value of the product's outbound volume at a historical moment is determined by the following method: , where Q is the sudden change in the quantity of goods shipped out at a historical moment, norm is the normalization function, and K is the number of moments in a preset time period. is the quantity of goods shipped out at the historical moment, is the shipment quantity of the product at the jth moment within the preset time period, It is the average quantity of goods shipped out during the preset time period.

[0008] In this way, the shipment volume of goods at historical moments is compared with the shipment volume of the same goods at different moments within a preset time period; and the shipment volume of goods at different moments within a preset time period is compared. The mutation value obtained can more objectively evaluate the degree of mutation in the sales volume of goods at historical moments.

[0009] Optionally, the correlation value between the first product and the second product is determined in the following manner: , where A is the correlation value between the first product and the second product, norm is the normalization function, P is the Pearson correlation coefficient between the outbound quantity of the first product and the outbound quantity of the second product, and n is the number of mutation moments that match the first product and the second product. is the time sequence of the mth mutation moment in the mutation moment of the first product that matches the second product, is the time sequence of the mth mutation moment of the second product that matches the first product, and N is the number of mutation moments that the first product does not match the second product.

[0010] In this way, by the time interval of the matched mutation moments and the number of the unmatched mutation moments, the matched mutation moments and the unmatched mutation moments in the mutation moments of the first commodity and the second commodity can be fully considered, and the obtained correlation value can better characterize the degree of correlation between the two commodities.

[0011] Optionally, the predicted outbound quantity of the target product at the next moment is determined by: , where R is the predicted outbound quantity of the target product at the next moment, M is the preset quantity, and norm is the normalization function. is the abnormal score value of the target product at the kth historical moment before the current moment, is the quantity of other products except the target product, is the correlation value between the target product and the ath other product, is the abnormal score value of the a-th other commodity at the k-th historical moment, is the outbound quantity of the target product at the kth historical moment.

[0012] In this way, through the shipment quantity of the target product of the automobile at multiple moments before the current moment, as well as the correlation value between the shipment quantity of other products and the target product, and considering the abnormal score values ​​of the target product and other products at different historical moments, the predicted shipment quantity can better meet the demand for the target product of the automobile.

[0013] Optionally, the supply of the target commodity from the supply chain is coordinated according to the predicted delivery quantity of the target commodity at the next moment, including: obtaining the inventory of the target commodity at the next moment, and coordinating the supply of the target commodity from the supply chain according to the inventory of the target commodity at the next moment and the predicted delivery quantity.

[0014] In this way, the supply quantity of the target goods from the supply chain is coordinated to ensure that the inventory quantity of the target goods for automobiles matches the predicted delivery quantity, thereby ensuring the supply of the target goods for automobiles.

[0015] Optionally, the supply of the target commodity from the supply chain is coordinated based on the inventory quantity of the target commodity at the next moment and the predicted delivery quantity, including: when it is determined that the inventory quantity of the target commodity at the next moment does not meet the predicted delivery quantity, purchasing the target commodity from the supply chain so that the inventory quantity of the target commodity at the next moment meets the predicted delivery quantity of the target commodity at the next moment.

[0016] Optionally, the supply of the target commodity from the supply chain is coordinated based on the inventory level of the target commodity at the next moment and the predicted delivery level, including: when it is determined that the inventory level of the target commodity at the next moment does not meet the predicted delivery level, determining the demand for the target commodity, and obtaining alternative commodities for the target commodity from the knowledge graph, so as to use the demand for the target commodity as the demand for the alternative commodity; and purchasing alternative commodities from the supply chain based on the demand for the alternative commodities.

[0017] Optionally, matching the mutation moments of the first product and the second product includes: for a first moment among multiple mutation moments to be matched of the first product, determining a second moment having the smallest time interval with the first moment from multiple mutation moments to be matched of the second product; forming the first moment and the second moment into a matching mutation moment pair to complete matching of the mutation moments in the first product and the second product, and treating the mutation moments of the first product or the second product that do not form a mutation moment pair as unmatched mutation moments.

[0018] Optionally, the method also includes: obtaining supply capacity information of the upstream supply chain for different models of automobile products; adjusting the priority order of multiple orders to be shipped corresponding to different models of automobile products based on the supply capacity information for different models of automobile products; and controlling the operation progress information of the multiple orders to be shipped based on the adjusted priority order of the multiple orders to be shipped.

[0019] Optionally, the automotive products include at least one of the following: automotive parts, automotive components, finished automotive products, and automotive supplies.

[0020] The technical solution provided by the embodiments of the present application may include the following beneficial effects: determining the mutation moment of goods according to the shipment quantity of automobile goods at historical moments, and determining the correlation value between goods through the correlation between different goods at the mutation moment, and being able to predict the shipment quantity of automobile goods according to the correlation between different goods, thereby being able to obtain a more accurate prediction of the shipment quantity, thereby more accurately coordinating the supply quantity of goods from the supply chain.

[0021] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0023] Figure 1 It is a flowchart of a method for coordinating supply chains in the automotive industry based on a knowledge graph according to an exemplary embodiment. DETAILED DESCRIPTION

[0024] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0025] First, the application scenario of the embodiment of the present application is briefly introduced. In the application scenario of the present application, the supply quantity of automobile products from the upstream supply chain can be determined based on the purchase quantity indicated by the purchase order for automobile products from the downstream and the inventory quantity of automobile products, making it difficult to meet the supply of automobile products. Therefore, the relevant technology is difficult to achieve effective management of the supply chain of automobile products.

[0026] In view of the above technical problems, the present application embodiment provides a method for coordinating the supply chain of the automotive industry based on a knowledge graph. Figure 1 is a flowchart of a method for coordinating a supply chain in the automotive industry based on a knowledge graph according to an exemplary embodiment. Figure 1 As shown, the method includes the following steps.

[0027] In step S101, the outbound quantity of automobile products is obtained from a pre-established knowledge graph, and the mutation value of the outbound quantity of the products at the historical moment is determined based on the outbound quantity of the products in a preset time period before the historical moment, and the historical moment when the mutation value is greater than the preset mutation threshold is used as the mutation moment of the product.

[0028] Automobile products may include at least one of the following: automobile parts, automobile components, automobile finished products, and automobile supplies. For example, automobile parts may be engine parts such as pistons, piston rings, valves, connecting rods, and camshafts, transmission system parts such as clutch pressure plates, transmission gears, transmission synchronizers, and drive shafts, and brake system parts such as brake pads.

[0029] Automobile parts can be engine assemblies, gearbox assemblies, brake system assemblies, etc., and automobile parts can also be electrical system assemblies; the automobile products in the embodiments of the present application can also be complete automobile products, and can also be automobile supplies such as seat covers, floor mats, steering wheel cover ornaments, navigators, driving recorders, reversing radars, and car refrigerators that are used in conjunction with the automobile.

[0030] The inventory of automobile products can record the outbound information of the products according to the product model, outbound quantity, outbound time and the warehouse location of the products, and can also record the inbound information of the products according to the product model, inbound quantity, inbound time and the warehouse location of the products.

[0031] Based on the goods' inbound and outbound information, the inventory party can construct a knowledge graph according to actual needs, and periodically update the inbound and outbound information recorded in the knowledge graph, so that they or other personnel can obtain the inbound and outbound information from the knowledge graph; technical personnel in this field can construct a knowledge graph according to actual needs, and the embodiments of the present application do not limit the way the knowledge graph is constructed.

[0032] According to the shipment volume of the goods in a preset time period before the historical moment, the mutation value of the shipment volume of the goods at the historical moment is determined; the mutation value is used to characterize the degree of mutation of the shipment volume of the goods at the historical moment, and the greater the mutation value of the shipment volume of the goods at a certain historical moment, the greater the mutation degree of the shipment volume of the goods at the historical moment.

[0033] For at least two related commodities of automobiles, there is a correlation in the shipment quantity of at least two related commodities, so that the shipment quantity of the related commodities usually has a mutation at the same or close time nodes, and two or more commodities among the multiple commodities of automobiles are usually purchased in a certain proportion within the same time period, so that the shipment quantity of the two or more related commodities all has a mutation within the same time period. Therefore, determining the mutation degree value of the shipment quantity of automobile commodities at a historical moment is helpful to determine the commodities with correlated shipment quantities among the multiple commodities of automobiles.

[0034] When the mutation value of the shipment volume of a commodity at a historical moment is greater than the preset mutation threshold, it means that the mutation degree of the shipment volume of the commodity at the historical moment is relatively large, and temporary or sudden shipment orders for the commodity may be received. The moment with a larger mutation value can be used as the mutation moment, so as to determine the correlation between different commodities based on the correlation between different commodities at the mutation moments.

[0035] The preset mutation threshold can be set according to actual needs. For example, when the mutation value ranges from 0 to 1, the preset mutation threshold can be set between 0.8 and 0.9.

[0036] In one embodiment, the sudden change value of the product shipment quantity at a historical moment is determined in the following manner: , where Q is the sudden change in the quantity of goods shipped out at a historical moment, norm is the normalization function, and K is the number of moments in a preset time period. is the quantity of goods shipped out at the historical moment, is the shipment quantity of the product at the jth moment within the preset time period, It is the average quantity of goods shipped out during the preset time period.

[0037] The numerator of the calculation formula for the mutation value of the commodity's shipment volume at a historical moment reflects the difference between the commodity's shipment volume at a historical moment and the shipment volume of the same commodity at different moments within a preset time period, thereby reflecting the degree of mutation in the commodity's shipment volume at a historical moment.

[0038] The denominator of the calculation formula for the mutation value of the product's outbound volume at a historical moment reflects the degree of difference in the product's outbound volume at different moments within a preset time period, thereby reflecting the overall fluctuation degree of the product's outbound volume within the preset time period.

[0039] For example, the shipment volume of the same commodity within a preset time period before a historical moment can provide a good reference for the mutation of the shipment volume of the commodity at the historical moment; for the historical moment W1, the preset time period before the historical moment may include moments W2, W3, W4, W5 and W6. By comparing the shipment volume of the commodity at the historical moment W1 with moments W2, W3, W4, W5 and W6, the difference between the shipment volume of the commodity at the historical moment and the shipment volume before the historical moment can be reflected.

[0040] By comparing the differences in the shipment quantities of goods in the preset time periods W2, W3, W4, W5 and W6 before the historical moment, the volatility of the sales volume of the goods itself can be reflected; for example, if the difference in the sales volume of a goods in the preset time period before the historical moment is large, it means that the fluctuation degree of the sales volume of the goods itself at different moments is large; or, if the difference in the sales volume of a goods in the preset time period before the historical moment is small, it means that the fluctuation degree of the sales volume of a certain goods itself at different moments is small.

[0041] The preset time period can be set according to actual needs, for example, the preset time period can be 30 natural days; different moments within the preset time period correspond to different time stages of the same time interval, for example, different moments within the preset time period can correspond to different natural days within 30 natural days; or, different moments within the preset time period can correspond to different natural weeks within 5 natural weeks; the embodiment of the present application does not constitute a limitation on the specific time interval between different moments and the specific duration of the preset time period.

[0042] In this way, the shipment volume of goods at historical moments is compared with the shipment volume of the same goods at different moments within a preset time period; and the shipment volume of goods at different moments within a preset time period is compared. The mutation value obtained can more objectively evaluate the degree of mutation in the sales volume of goods at historical moments.

[0043] In step S102, the mutation times of the first product and the second product are matched, and the correlation value between the first product and the second product is determined according to the matching result.

[0044] The first commodity and the second commodity are any two commodities among a variety of commodities such as automobiles. Compared with the shipment quantities of the first commodity and the second commodity at all times, the moment when there is a sudden change in the shipment quantity of the commodities can more obviously reflect the fluctuation in the shipment quantity of the commodities. Matching the sudden change moments of the first commodity and the second commodity can better determine the correlation value between the first commodity and the second commodity, so as to characterize the correlation degree between the shipment quantities of the two commodities through the correlation value.

[0045] In one embodiment, matching the mutation moments of a first product and a second product includes: for a first moment among multiple mutation moments to be matched of the first product, determining a second moment having a minimum time interval with the first moment from multiple mutation moments to be matched of the second product; forming a matching mutation moment pair with the first moment and the second moment to complete matching of the mutation moments in the first product and the second product, and treating the mutation moments of the first product or the second product that do not form a mutation moment pair as unmatched mutation moments.

[0046] By finding the moment with the smallest time interval, we can more accurately analyze the correlation between the shipment quantities of the two commodities in the time series, thereby discovering the correlation between the shipment quantities of the two commodities; by forming a matching mutation moment pair with the first moment and the second moment, we can avoid wasting resources caused by re-matching the already matched moments.

[0047] In the same historical time period, the number of mutation moments of the first product and the number of mutation moments of the second product may be different. After matching the mutation moments of the first product and the mutation moments of the second product, the matched mutation moments and the unmatched mutation moments may be considered to determine the correlation value between the first product and the second product.

[0048] In this way, a first moment among multiple mutation moments to be matched of the first product and a second moment among multiple mutation moments to be matched of the second product are combined to form a matching mutation moment pair, where the second moment is the moment with the shortest time interval between the multiple mutation moments to be matched of the second product and the first moment. This can better achieve the matching between the mutation moments to be matched of the two products and avoid repeated matching of the same mutation moment to be matched.

[0049] In one embodiment, the association value between the first product and the second product is determined in the following manner: , where A is the correlation value between the first product and the second product, norm is the normalization function, P is the Pearson correlation coefficient between the outbound quantity of the first product and the outbound quantity of the second product, and n is the number of mutation moments that match the first product and the second product. is the time sequence of the mth mutation moment in the mutation moment of the first product that matches the second product, is the time sequence of the mth mutation moment of the second product that matches the first product, and N is the number of mutation moments that the first product does not match the second product.

[0050] The Pearson correlation coefficient between the shipment quantity of the first commodity and the shipment quantity of the second commodity can be used to measure the degree of linear correlation between the shipment quantities of the first commodity and the second commodity. The Pearson correlation coefficient between the shipment quantities of the first commodity and the second commodity can be determined based on the shipment quantities of the first commodity and the second commodity before the current moment, which will not be repeated here.

[0051] For example, the first standard deviation of the shipment quantity of the first commodity within the target time period can be determined, and the second standard deviation of the shipment quantity of the second commodity within the target time period can be determined; the covariance of the shipment quantity of the first commodity and the shipment quantity of the second commodity within the target time period can be determined; the product of the first standard deviation and the second standard deviation can be determined, and the ratio of the covariance to the product can be used as the Pearson correlation coefficient between the first commodity and the second commodity.

[0052] The first product and the second product may be different automobile parts, automobile components, automobile finished products or automobile supplies; for example, the first product may be a transmission, and the second product may be a brake pad.

[0053] The larger the Pearson correlation coefficient between the shipment quantity of the first commodity and the shipment quantity of the second commodity, the greater the correlation between the shipment quantity of the first commodity and the shipment quantity of the second commodity, the greater the probability that the first commodity and the second commodity are shipped out at the same time, and the more likely the shipment of the first commodity will promote the shipment of the second commodity.

[0054] The smaller the Pearson correlation coefficient between the shipment quantity of the first commodity and the shipment quantity of the second commodity, the smaller the correlation between the shipment quantity of the first commodity and the shipment quantity of the second commodity, the smaller the probability that the first commodity and the second commodity are shipped out at the same time, the smaller the impact of the shipment of the first commodity on the shipment of the second commodity, and even the shipment of the first commodity is more likely to restrict the shipment of the second commodity.

[0055] For the mutation moments that match the first commodity and the second commodity, the smaller the time interval between the two matching mutation moments, the greater the correlation between the two commodities in terms of shipment quantity.

[0056] For the mutation moments at which the first commodity and the second commodity do not match, the smaller the number of the mutation moments that do not match, the greater the correlation between the first commodity and the second commodity in terms of the shipment quantity.

[0057] In this way, by the time interval of the matched mutation moments and the number of the unmatched mutation moments, the matched mutation moments and the unmatched mutation moments in the mutation moments of the first commodity and the second commodity can be fully considered, and the obtained correlation value can better characterize the degree of correlation between the two commodities.

[0058] In step S103, the shipment quantity of the goods at the historical moment is input into the pre-built isolation forest model to obtain the abnormal score value corresponding to the shipment quantity output by the isolation forest model.

[0059] The abnormal score value corresponding to the outbound volume is used to characterize the degree of isolation of the outbound volume; the larger the abnormal score value corresponding to the outbound volume, the greater the degree of isolation of the outbound volume; conversely, the smaller the abnormal score value corresponding to the outbound volume, the smaller the degree of isolation of the outbound volume.

[0060] Isolation Forest is an unsupervised learning algorithm for anomaly detection. The core idea of ​​Isolation Forest is that anomalies are usually few and unique in the feature space, so anomalies are easier to isolate than normal points.

[0061] The isolation forest algorithm constructs multiple binary trees by randomly selecting features and then randomly selecting split values. These binary trees can be used to isolate data points. When constructing each binary tree, the isolation forest algorithm can randomly select a feature point and then randomly select a split value within the value range of the feature point to split the data, so as to complete the segmentation of all data points in the data set and obtain the constructed isolation forest.

[0062] The anomaly score of a data point can be determined based on the path length of the data point in the tree in the isolation forest model; the shorter the path length of the data point in the tree, the larger the anomaly score of the data point; in the isolation forest algorithm, the process of determining the anomaly score of a data point based on the path length of the data point is common knowledge to those skilled in the art and will not be elaborated here.

[0063] Among them, an isolation forest model composed of multiple isolated trees can be constructed according to the historical outbound quantities at historical moments; for the target outbound quantities among multiple outbound quantities, the path length of the target outbound quantities in the same isolated tree can be determined; the average value of the path length of the target outbound quantities in different isolated trees can be obtained; the abnormal score value of the target outbound quantity in the isolation forest model is negatively correlated with the average value of the path length of the target outbound quantity in different isolated trees.

[0064] For a certain product among the multiple commodities of automobiles, the smaller the abnormal score value of the product's shipment at a certain historical moment, the more common the shipment of the product at the historical moment is, the higher the probability that the inventory of the product at the historical moment can meet the demand for shipment, and the lower the probability that the shipment of the product at the historical moment cannot meet the demand.

[0065] On the contrary, the larger the abnormal score value of the product's outbound quantity at a certain historical moment, the rarer the product's outbound quantity at the historical moment, and the lower the probability that the product's inventory at the historical moment can adapt to this rare outbound quantity. Alternatively, the smaller the safety stock that the product's inventory at the historical moment has after meeting this rare outbound quantity, the remaining safety stock will be difficult to meet the subsequent demand for the product, and the higher the probability that the product's outbound quantity at the historical moment cannot meet the demand.

[0066] Here, by determining the abnormal score value of the product's outbound volume at a historical moment, it is helpful to predict the product's outbound volume so that the inventory of automobile products can better meet demand.

[0067] In step S104, for a target product among multiple automobile products, the predicted shipment quantity of the target product at the next moment is obtained according to the correlation value between the target product and other products, the abnormal score value of the target product at a historical moment, and the abnormal score values ​​of other products at a historical moment.

[0068] The greater the correlation value between the target product and other products, the greater the probability or degree that the shipment quantity of the target product is affected by the shipment quantity of other products.

[0069] The larger the abnormal score value of the target product at a historical moment, the higher the probability that the target product's shipment quantity at a historical moment will be unable to meet the demand. In order to meet the demand for the target product, a larger predicted shipment quantity can be set for the target product so that more target products can be purchased to meet the corresponding demand.

[0070] The larger the abnormal score value of other products at a historical moment, the higher the probability that the shipment of other products at that historical moment will be unable to meet the demand, and the shipment of other products may affect the shipment of the target product; for example, the brake pads and shock absorbers of a vehicle may be purchased at the same time.

[0071] When the shipment of other goods is difficult to meet the demand, since the shipment of other goods is related to the shipment of the target product, the shipment of other goods that is difficult to meet their own demand may cause or aggravate the insufficient inventory of the target product. In order to adapt to the changes in demand, the contribution of other goods with larger abnormal score values ​​to the predicted shipment of the target product can be increased.

[0072] In one embodiment, the predicted outbound quantity of the target commodity at the next moment is determined by: , where R is the predicted outbound quantity of the target product at the next moment, M is the preset quantity, and norm is the normalization function. is the abnormal score value of the target product at the kth historical moment before the current moment, is the quantity of other products except the target product, is the correlation value between the target product and the ath other product, is the abnormal score value of the a-th other commodity at the k-th historical moment, is the outbound quantity of the target product at the kth historical moment.

[0073] The greater the correlation value between other automobile products and the shipment quantity of the target product, the greater the influence of the shipment quantity of other automobile products on the shipment quantity of the target product. Therefore, the contribution of other products with greater correlation values ​​with the shipment quantity of the target product to the predicted shipment quantity can be increased.

[0074] The quantity of the target commodity required at the next moment has a high probability of continuing the trend in multiple moments before the current moment; the predicted outbound quantity of the target commodity at the next moment is determined based on the outbound quantities of the target commodity at multiple historical moments before the current moment. Therefore, the larger the outbound quantities of the target commodity at multiple historical moments before the current moment, a larger predicted outbound quantity can be set for the target commodity of automobiles to ensure the supply of the target commodity of automobiles at the next moment.

[0075] In the calculation formula for the predicted shipment volume of the target commodity at the next moment, the larger the abnormal score value of the shipment volume of the target commodity of automobiles at a certain historical moment, the more likely the shipment volume of the target commodity at the historical moment will be affected by factors such as promotions, holidays or temporary market turmoil. In order to ensure the supply of the target commodity of automobiles; for example, to ensure the supply of the target commodity to the company's own downstream customers, the contribution of the shipment volume at the historical moment with a larger abnormal score value to the predicted shipment volume at the next moment can be increased.

[0076] In this way, through the shipment quantity of the target product of the automobile at multiple moments before the current moment, as well as the correlation value between the shipment quantity of other products and the target product, and considering the abnormal score values ​​of the target product and other products at different historical moments, the predicted shipment quantity can better meet the demand for the target product of the automobile.

[0077] In step S105 , the supply quantity of the target commodity from the supply chain is coordinated according to the predicted shipment quantity of the target commodity at the next moment.

[0078] During the course of production and operation, companies in the automotive industry usually need to process or assemble raw materials from the upstream supply chain to obtain automotive parts and other commodities, and then ship automotive parts and other commodities to downstream customers. Downstream customers can be consumers or other companies that need to further process parts and other commodities.

[0079] When an enterprise's own production capacity is temporarily unable to meet demand, the enterprise can also directly purchase finished products such as automobile parts and other commodities from the upstream supply chain, and ship the purchased finished products such as automobile parts and other commodities to downstream customers or other enterprises that submit demand orders to the enterprise.

[0080] After determining the predicted outbound volume of the target commodity at the next moment, the enterprise can choose to adjust the purchase quantity of raw materials for the target commodity based on its actual inventory, thereby adjusting the production volume of the target commodity; alternatively, the enterprise can adjust the purchase quantity of finished products of the target commodity to adapt to changes in demand for the target commodity.

[0081] In one embodiment, the supply of the target commodity from the supply chain is coordinated according to the predicted delivery quantity of the target commodity at the next moment, including: obtaining the inventory of the target commodity at the next moment, and coordinating the supply of the target commodity from the supply chain according to the inventory of the target commodity at the next moment and the predicted delivery quantity.

[0082] By accurately predicting the target product’s outbound shipment volume at the next moment and obtaining the target product’s inventory volume at the next moment, it is possible to avoid over-purchasing and inventory backlogs of the target product’s finished products, or to avoid over-purchasing and inventory backlogs of the target product’s production raw materials, thereby reducing storage costs and the risk of product obsolescence.

[0083] In this way, the supply quantity of the target goods from the supply chain is coordinated to ensure that the inventory quantity of the target goods for automobiles matches the predicted delivery quantity, thereby ensuring the supply of the target goods for automobiles.

[0084] In one embodiment, the supply of the target commodity from the supply chain is coordinated based on the inventory quantity of the target commodity at the next moment and the predicted delivery quantity, including: when it is determined that the inventory quantity of the target commodity at the next moment does not meet the predicted delivery quantity, purchasing the target commodity from the supply chain so that the inventory quantity of the target commodity at the next moment meets the predicted delivery quantity of the target commodity at the next moment.

[0085] By monitoring the inventory level of the target product at the next moment and evaluating the relationship between the inventory level and the predicted delivery level, the supply of the target product from the supply chain can be negotiated to ensure that the inventory level matches the predicted delivery level and the supply of the target product for automobiles is guaranteed.

[0086] In one embodiment, the supply of the target commodity from the supply chain is coordinated based on the inventory of the target commodity at the next moment and the predicted delivery quantity, including: when it is determined that the inventory of the target commodity at the next moment does not meet the predicted delivery quantity, the demand for the target commodity is determined, and alternative commodities for the target commodity are obtained from the knowledge graph, so that the demand for the target commodity is used as the demand for the alternative commodity; and the alternative commodity is purchased from the supply chain according to the demand for the alternative commodity.

[0087] The knowledge graph may pre-store alternative products that can achieve the same functions as the target product. The target product and the alternative product may be products of the same type that differ only in the manufacturer. For example, the target product and the alternative product may be brake pads with the same composition and size.

[0088] In this way, when the inventory of the target product at the next moment does not meet the predicted delivery quantity, alternative products can be purchased from the supply chain, so as to meet the demand for the target product with the alternative products.

[0089] In one embodiment, the supply quantity of the target commodity from the supply chain is coordinated based on the inventory quantity of the target commodity at the next moment and the predicted delivery quantity, including: when it is determined that the inventory quantity of the target commodity at the next moment does not meet the predicted delivery quantity, coordination information can also be sent to the terminal device where the upstream supply chain is located, and the coordination information is used to coordinate at least one of the supply quantity and delivery time of the target commodity.

[0090] For example, Company A may purchase finished products of the target commodity automobile from Company B for production and operation needs. After Company B delivers the finished products of the target commodity automobile to Company A, Company A may respond to receiving a demand order for the target commodity from Company C and dispatch the finished products of the target commodity automobile to Company C.

[0091] When it is determined that the inventory quantity of the target commodity at the next moment does not meet the predicted delivery quantity, Company A can send coordination information to the terminal device of Company B through the terminal device. The coordination information can be, for example, "Please deliver the target commodity 3 days in advance based on the pre-agreed delivery period", or "When delivering the target commodity next time, please provide an additional 20% delivery quantity based on the pre-agreed delivery quantity" and other information.

[0092] In this way, coordination information can be sent to the terminal device where the upstream supply chain is located to coordinate at least one of the supply quantity and supply time of the target commodity to ensure the supply of the demand for the target commodity of automobiles.

[0093] Through the automobile industry supply chain coordination method based on knowledge graph provided in the embodiment of the present application, the mutation moment of the goods is determined according to the outbound quantity of automobile goods at historical moments, and the correlation value between the goods is determined through the correlation between different goods at the mutation moment. The outbound quantity of automobile goods can be predicted according to the correlation between different goods. Therefore, a more accurate predicted outbound quantity can be obtained, thereby more accurately coordinating the supply quantity of goods from the supply chain.

[0094] In one embodiment, the supply capacity information of the upstream supply chain for different models of automobile products can also be obtained; based on the supply capacity information for different models of automobile products, the priority order of multiple orders to be shipped corresponding to different models of automobile products is adjusted; based on the adjusted priority order of the multiple orders to be shipped, the operation progress information of the multiple orders to be shipped is controlled.

[0095] The target enterprise can use automobile products from the upstream supply chain to process orders to be shipped; for example, the target enterprise can use automobile parts from the upstream supply chain to produce finished automobiles, thereby processing orders for finished automobiles to be shipped.

[0096] The target enterprise can receive supply capacity information for automobile products from other enterprises in the upstream supply chain. For example, the target enterprise and other enterprises in the upstream supply chain can communicate through terminal devices or servers, so that other enterprises in the upstream supply chain send supply capacity information for automobile products to the target enterprise. The supply capacity information is used to characterize the supply capacity of the upstream supply chain for automobile products.

[0097] Adjusting the priority order of multiple to-be-shipped orders corresponding to different models of automobile products based on supply capacity information for different models of automobile products may include: determining the priority corresponding to the model of automobile product based on the size of the supply capacity represented by the supply capacity information of the automobile products; and determining the priority order of multiple to-be-shipped orders corresponding to different models of automobile products based on the order of the priorities corresponding to the different models of automobile products.

[0098] For example, when the supply capacity information of the target model of automobile products indicates that the upstream supply chain has insufficient supply capacity for the target model of automobile products, in order to ensure the smooth delivery of the pending orders of the corresponding target model of automobile products, the priority of the pending orders of the corresponding target model of automobile products can be increased.

[0099] Alternatively, when the supply capacity information of the target model of automobile products indicates that the upstream supply chain has sufficient supply capacity for the target model of automobile products, in order to ensure the smooth delivery of the pending orders for other models of automobile products with tighter supply capacity, the priority value of the pending orders for the corresponding target model of automobile products can be lowered.

[0100] Controlling the operation progress information of the multiple orders to be shipped according to the adjusted priority order of the multiple orders to be shipped may include: executing the operation of the orders to be shipped according to the adjusted priority order of the orders to be shipped.

[0101] For example, the adjusted orders with higher priority can be displayed first on the terminal bound to the pickers or assemblers, so that the pickers can pick the items in the orders with higher priority first, or the assemblers can assemble the items in the orders with higher priority first.

[0102] Among them, the pickers can be humans, or picking robots or other equipment used for picking; the assemblers can be humans, or assembling robots or other equipment used for assembling.

[0103] Controlling the operation progress information of multiple orders to be shipped based on the adjusted priority order of multiple orders to be shipped can include: improving the operation progress of orders to be shipped with higher priority to give priority to ensuring the delivery of orders to be shipped with higher priority; for example, priority can be given to ensuring the assembly of automobile parts with insufficient supply capacity of the upstream supply chain to obtain assembled finished automobiles; or, priority can be given to ensuring the assembly of automobile parts with insufficient supply capacity of the upstream supply chain to obtain assembled automobile parts.

[0104] It should be understood that the features of some embodiments of the various applications described herein may be combined with each other, unless specifically stated otherwise.

[0105] Although terms such as "first", "second", and "third" may be used herein to describe various components, parts, regions, layers, or sections, these components, parts, regions, layers, or sections are not limited to these terms. Instead, these terms are only used to distinguish one component, part, region, layer, or section from another component, part, region, layer, or section. Therefore, without departing from the teachings of the examples described herein, the first component, part, region, layer, or section mentioned in the examples may also be referred to as the second component, part, region, layer, or section.

[0106] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description herein, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0107] In addition, the word "exemplary" is used herein to mean serving as an example, instance, or diagram. Any aspect or design described herein as "exemplary" is not necessarily to be understood as being advantageous over other aspects or designs. Instead, the use of the word exemplary is intended to present concepts in a concrete manner. As used herein, the term "or" is intended to mean an inclusive "or" rather than an exclusive "or".

[0108] Likewise, although the present application has been shown and described with respect to one or more implementations, equivalent variations and modifications will occur to those skilled in the art upon reading and understanding this specification and the accompanying drawings. With particular regard to the various functions performed by the components (e.g., elements, resources, etc.) described above, unless otherwise noted, the terms used to describe such components are intended to correspond to any component (functionally equivalent) that performs the specific functions of the described components, even if not structurally equivalent to the disclosed structures.

[0109] Additionally, while particular features of the present application may have been disclosed with respect to only one of several implementations, such features may be combined with one or more other features of other implementations as may be desired and advantageous for any given or particular application.

[0110] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary technical means in the art that are not disclosed in the present application, and the specification and embodiments are only considered as exemplary.

[0111] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims

1. A method for coordinating the supply chain in the automotive industry based on knowledge graph, characterized in that: include: The outbound quantity of automobile products is obtained from the pre-established knowledge graph, and the mutation value of the outbound quantity of the products at the historical moment is determined according to the outbound quantity of the products in the preset time period before the historical moment, and the historical moment when the mutation value is greater than the preset mutation threshold is taken as the mutation moment of the products; Matching the mutation moments of the first product and the second product, and determining the correlation value between the first product and the second product according to the matching result; the first product and the second product are any two products among the multiple products of automobiles; Input the product shipment quantity at a historical moment into the pre-built isolation forest to obtain the abnormal score value corresponding to the shipment quantity output by the isolation forest model; The abnormal score value is used to characterize the isolation degree of the outbound quantity; For a target product among multiple products of automobiles, the predicted outbound quantity of the target product at the next moment is obtained according to the correlation value between the target product and other products, the abnormal score value of the target product at a historical moment, and the abnormal score value of other products at a historical moment; Coordinate the supply of target products from the supply chain based on the predicted shipment volume of the target products at the next moment.

2. The automotive industry supply chain coordination method based on knowledge graph according to claim 1 is characterized in that: The sudden change in the product's outbound volume at a historical moment is determined in the following way: , where Q is the sudden change in the quantity of goods shipped out at a historical moment, norm is the normalization function, and K is the number of moments in a preset time period. is the quantity of goods shipped out at the historical moment, is the shipment quantity of the product at the jth moment within the preset time period, It is the average quantity of goods shipped out during the preset time period.

3. The automotive industry supply chain coordination method based on knowledge graph according to claim 1 is characterized in that: The relevance value between the first product and the second product is determined in the following way: , where A is the correlation value between the first product and the second product, norm is the normalization function, P is the Pearson correlation coefficient between the outbound quantity of the first product and the outbound quantity of the second product, and n is the number of mutation moments that match the first product and the second product. is the time sequence of the mth mutation moment in the mutation moment of the first product that matches the second product, is the time sequence of the mth mutation moment of the second product that matches the first product, and N is the number of mutation moments that the first product does not match the second product.

4. The automotive industry supply chain coordination method based on knowledge graph according to claim 1 is characterized in that: The predicted outbound quantity of the target product at the next moment is determined by the following method: , where R is the predicted outbound quantity of the target product at the next moment, M is the preset quantity, and norm is the normalization function. is the abnormal score value of the target product at the kth historical moment before the current moment, is the quantity of other products except the target product, is the correlation value between the target product and the ath other product, is the abnormal score value of the a-th other product at the k-th historical moment, is the outbound quantity of the target product at the kth historical moment.

5. The automotive industry supply chain coordination method based on knowledge graph according to claim 1 is characterized in that: Coordinate the supply of target goods from the supply chain based on the predicted outbound volume of the target goods at the next moment, including: Obtain the inventory of the target product at the next moment, and coordinate the supply of the target product from the supply chain based on the inventory of the target product at the next moment and the predicted delivery quantity.

6. The automotive industry supply chain coordination method based on knowledge graph according to claim 5 is characterized in that: Coordinate the supply of target products from the supply chain based on the target product's inventory at the next moment and the predicted outbound volume, including: When it is determined that the inventory quantity of the target commodity at the next moment does not meet the predicted delivery quantity, the target commodity is purchased from the supply chain so that the inventory quantity of the target commodity at the next moment meets the predicted delivery quantity of the target commodity at the next moment.

7. The automotive industry supply chain coordination method based on knowledge graph according to claim 5 is characterized in that: Coordinate the supply of target products from the supply chain based on the target product's inventory at the next moment and the predicted outbound volume, including: When it is determined that the inventory of the target product at the next moment does not meet the predicted delivery quantity, the demand for the target product is determined, and a substitute product of the target product is obtained from the knowledge graph, so that the demand for the target product is used as the demand for the substitute product; Purchase substitute goods from the supply chain based on the demand for substitute goods.

8. The automotive industry supply chain coordination method based on knowledge graph according to claim 1 is characterized in that: Matching the mutation moments of the first product and the second product includes: For a first moment among multiple mutation moments to be matched of the first product, determine a second moment having the shortest time interval with the first moment from multiple mutation moments to be matched of the second product; The first moment and the second moment are combined into a matching mutation moment pair to complete the matching of the mutation moments in the first product and the second product, and the mutation moments in the mutation moment of the first product or the mutation moment of the second product that do not form a mutation moment pair are regarded as unmatched mutation moments.

9. The automotive industry supply chain coordination method based on knowledge graph according to claim 1 is characterized in that: The method further comprises: Obtain information on the upstream supply chain's supply capacity for different models of automobile products; According to the supply capacity information of different models of automobile products, the priority order of multiple to-be-shipped orders corresponding to different models of automobile products is adjusted; According to the adjusted priority order of the multiple orders to be shipped, the operation progress information of the multiple orders to be shipped is controlled.

10. The automotive industry supply chain coordination method based on knowledge graph according to claim 1 is characterized in that: The automobile product includes at least one of the following: Auto parts, auto components, finished auto products and auto supplies.

Images