A method, apparatus, electronic device, and storage medium for determining a project schedule.

Abstract

This application provides a method, apparatus, electronic device, and storage medium for determining a schedule. The method includes: acquiring pending demand, supply information, and material information; determining material constraints, capacity constraints, and date constraints based on the material information and supply information; inputting the pending demand, the supply of each supplier, and the capacity into a pre-trained knowledge reasoning model to obtain bottleneck materials and bottleneck suppliers; grouping and sorting the pending demand based on the scores corresponding to the bottleneck suppliers to obtain multiple sorted demand subgroups; determining the weight corresponding to each demand in the current subgroup based on the bottleneck material; using material constraints, capacity constraints, and date constraints as constraints; using the weighted demand achievement rate as the optimization objective; and performing heuristic solution for the supply plan corresponding to each demand based on the supply information of each supplier to obtain the supply plan corresponding to the pending demand.

Description

Technical Field

[0001] This application relates to the field of intelligent manufacturing technology, and in particular to a method, apparatus, electronic device, and storage medium for determining planning and scheduling. Background Technology

[0002] In the face of increasingly fierce market competition, enterprises should be able to respond quickly to changes in market demand and achieve effective planning and scheduling. Planning and scheduling refers to the process of generating an executable supply plan that meets various demands and achieves supply-demand balance under certain resource constraints.

[0003] Current scheduling systems often focus on production scheduling at the execution level, with research typically centered on efficient and dynamic shop floor scheduling or production scheduling. Optimization goals are usually minimum cost or minimum completion time, while research on planning-level scheduling is relatively limited. The few existing planning systems are also ineffective, exhibiting low feasibility and low requirement fulfillment rates. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, electronic device, and storage medium for determining project schedules, so as to achieve project schedules with high executability and high achievement rate. The specific technical solution is as follows:

[0005] In a first aspect, embodiments of this application provide a method for determining a planned schedule, the method comprising:

[0006] Obtain pending demand, supply information, and material information, wherein the supply information includes at least the supply, capacity, cycle time, and supply time information of each supply organization, the cycle time is used to represent the working hours required for the supply organization to provide each material, and / or, is used to represent the quantity of each material provided by the supply organization per unit time.

[0007] Based on the demand transmission relationship between the materials identified by the material information, material constraints are determined; based on the capacity and the cycle time, capacity constraints are determined; and based on the supply time relationship between the materials identified by the supply time information, date constraints are determined. Here, the material constraints are used to represent the dependency relationship between the materials, the capacity constraints are used to represent the ability constraints of each supply organization to provide each material, and the date constraints are used to represent the time limits of each supply organization to provide each material.

[0008] The pending demand, the supply and capacity of each supplying organization are input into a pre-trained knowledge reasoning model to obtain the bottleneck material and bottleneck supplying organization corresponding to the pending demand. The knowledge reasoning model constructs material nodes and supplying organization nodes based on the material information included in historical planning and scheduling data and the relationship between each supplying organization and the material identified by the cycle time. The material nodes, supplying organization nodes and the edges between the nodes are represented by the demand, supply and capacity included in the historical planning and scheduling data, which is used to predict the bottleneck material and bottleneck supplying organization corresponding to the input demand.

[0009] Based on the score of the supply organization corresponding to the pending demand hitting the bottleneck supply organization, the pending demand is grouped and sorted to obtain multiple sorted demand subgroups. The bottleneck supply organization is used to characterize the bottleneck in the supply organization's material production capacity.

[0010] Each demand subgroup is obtained in sorting order. For each demand included in the current subgroup, the weight corresponding to the demand is determined based on the number of times the material corresponding to the demand hits the bottleneck material and the score corresponding to the bottleneck material.

[0011] Using the material constraints, capacity constraints, and date constraints as constraints, and the demand achievement rate with the weights as the optimization objective, the supply plan corresponding to each demand in the current subgroup is heuristically solved based on the supply information of each supply institution, to obtain the supply plan corresponding to the demand to be processed.

[0012] Optionally, the knowledge reasoning model can be constructed in the following ways:

[0013] Obtain historical planning and scheduling data, historical bottleneck materials, and historical bottleneck suppliers;

[0014] Based on the material information included in the historical planning and scheduling data, and the relationship between each supply unit and material identified by the cycle time, material nodes and supply unit nodes are constructed.

[0015] The material nodes, supply mechanism nodes, and edges between nodes are characterized using the historical planning and scheduling data, including demand, supply, and capacity, to obtain a heterogeneous directed graph. In the heterogeneous directed graph, the edges between material nodes represent the constraint relationships between the materials corresponding to the material nodes, and the edges between material nodes and supply mechanism nodes represent that the materials corresponding to the material nodes can be produced by the supply mechanism corresponding to the supply mechanism nodes.

[0016] Based on the historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations, the heterogeneous directed graph is trained to obtain the knowledge reasoning model.

[0017] Optionally, the step of training the heterogeneous directed graph based on the historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations to obtain the knowledge reasoning model includes:

[0018] The nodes corresponding to historical bottleneck materials and historical bottleneck supply organizations in the heterogeneous directed graph are identified as positive samples, and the nodes other than the positive samples are identified as negative samples.

[0019] The historical planning and scheduling data includes demand and supply as the initial vector of material nodes, capacity as the initial vector of supply mechanism nodes, the relationship between material information as the initial vector of the edge between material nodes, and cycle time as the initial vector of the edge between material nodes and supply mechanism nodes.

[0020] The heterogeneous directed graph is trained based on the positive samples, the negative samples, and the initial vectors to obtain the knowledge reasoning model.

[0021] Optionally, the step of grouping and sorting the pending demands based on the score of the supply organization corresponding to the bottleneck supply organization, to obtain multiple sorted demand subgroups, includes:

[0022] For each demand included in each demand group, the supply organization corresponding to the demand is determined based on the material information corresponding to the demand and the cycle time. The demand group is obtained by grouping the demands to be processed according to priority.

[0023] The demand that is the same as that of the supply institution and the bottleneck supply institution is divided into a group of demand to obtain a demand subgroup;

[0024] The demand subgroups are sorted according to the scores of the corresponding bottleneck supply institutions to obtain multiple sorted demand subgroups.

[0025] Optionally, the step of using the material constraints, capacity constraints, and date constraints as constraints, and the demand fulfillment rate with the weights as the optimization objective, and based on the supply information of each supply organization, heuristically solving the supply plan corresponding to each demand in the current subgroup to obtain the supply plan corresponding to the demand to be processed includes:

[0026] Get the highest priority set of requirements as the current group of requirements;

[0027] For the demand subgroups included in the current demand group, with the material constraints, the capacity constraints, and the date constraints as constraints, and with the demand achievement rate with the weights corresponding to the demand subgroups included in the current demand group as the optimization objective, the supply plan corresponding to the demand subgroups included in the current demand group is heuristically solved based on the current supply information of each supply institution to obtain the supply plan corresponding to the current demand group.

[0028] Subtracting the supply information of each supply institution required for the supply plan, the current supply information of each supply institution is obtained;

[0029] The next set of demands is retrieved in descending order of priority as the current set of demands. The subgroups of demands included in the current set of demands are returned. The material constraints, capacity constraints, and date constraints are used as constraints. The target is to optimize the achievement rate of the demand subgroups of the current set of demands with the weights. Based on the current supply information of each supply organization, the supply plan corresponding to the subgroups of demands included in the current set of demands is heuristically solved to obtain the supply plan corresponding to the current set of demands. This process continues until all sets of demands have been processed.

[0030] Optionally, the step of using the material constraints, capacity constraints, and date constraints as constraints for the demand subgroups included in the current demand group, and using the demand achievement rate with the weights corresponding to the demand subgroups included in the current demand group as the optimization objective, and heuristically solving the supply plan corresponding to the demand subgroups included in the current demand group based on the current supply information of each supply institution, to obtain the supply plan corresponding to the current demand group, includes:

[0031] For the current group's requirements, the first requirement subgroup is obtained according to the sorting order and is taken as the current subgroup;

[0032] For the current subgroup, with the material constraints, capacity constraints, and date constraints as constraints, and with the weighted demand achievement rate corresponding to the current subgroup as the optimization objective, the supply plan corresponding to the current subgroup is heuristically solved based on the current supply information of each supply institution to obtain the supply plan corresponding to the current subgroup.

[0033] Subtracting the supply information of each supply institution required for the supply plan, the current supply information of each supply institution is obtained;

[0034] The process involves obtaining the next demand subgroup in the specified order, using it as the current subgroup, and returning the steps described above. For the current subgroup, the process is as follows: using the material constraints, capacity constraints, and date constraints as constraints, and the weighted demand fulfillment rate corresponding to the current subgroup as the optimization objective. Based on the current supply information of each supply institution, the process is heuristically solved to obtain the supply plan corresponding to the current subgroup. This process continues until all demand subgroups have been processed.

[0035] Optionally, the demand to be processed includes predicted demand;

[0036] Prior to the step of obtaining the demand to be processed, supply information, and material information, the method further includes:

[0037] Obtain historical demand data and switching data for materials, wherein the switching data refers to material data that changes according to business needs;

[0038] Based on the time sequence corresponding to the historical demand data and the switching data, a demand sequence for the material is constructed;

[0039] Feature extraction is performed on the demand sequence to obtain the sequence features of the demand sequence, and the materials are classified based on the sequence features to obtain multiple categories of materials;

[0040] For each category of materials, a prediction method adapted to the sequence characteristics of the materials in that category is used to predict the demand for the materials in that category, thus obtaining the predicted demand for the materials in that category.

[0041] Optionally, the supply from each supplying entity includes the revised procurement supply;

[0042] Prior to the step of obtaining the demand to be processed, supply information, and material information, the method further includes:

[0043] Obtain historical supply data from supply organizations, wherein the historical supply data includes the supply date and quantity of each material supplied by each supply organization;

[0044] Feature extraction is performed on the historical supply data to obtain the features of the historical supply data, and the materials are classified based on the features to obtain multiple categories of materials;

[0045] For each category of materials, a forecasting method adapted to the characteristics of that category of materials is used to predict the supply date and supply quantity of that category of materials, thus obtaining the corrected procurement supply for that category of materials.

[0046] Optionally, the supply plan includes allocation strategies for existing supply and / or new supply plans resulting from capacity allocation;

[0047] The existing supply allocation strategy includes whether to use the existing supply, the allocation date, the supply selection on the allocation date, and the supply quantity selection on the allocation date.

[0048] The new supply plan resulting from the capacity allocation includes the type of planned supply, the date of planned supply, the supplying agency of planned supply, and the quantity of planned supply.

[0049] Secondly, embodiments of this application provide a device for determining a planned schedule, the device comprising:

[0050] The information acquisition module is used to acquire demand, supply information and material information to be processed. The supply information includes at least the supply, capacity, cycle time and supply time information of each supply unit. The cycle time is used to represent the working hours required for the supply unit to provide each material, and / or to represent the quantity of each material provided by the supply unit per unit time.

[0051] The constraint determination module is used to determine material constraints based on the demand transmission relationship between the materials identified by the material information, to determine capacity constraints based on the capacity and the cycle time, and to determine date constraints based on the supply time relationship between the materials identified by the supply time information. The material constraints are used to represent the dependency relationship between the materials, the capacity constraints are used to represent the ability constraints of each supply organization to provide each material, and the date constraints are used to represent the time limits of each supply organization to provide each material.

[0052] The bottleneck prediction module is used to input the demand to be processed, the supply and capacity of each supply unit into a pre-trained knowledge reasoning model to obtain the bottleneck material and bottleneck supply unit corresponding to the demand to be processed. The knowledge reasoning model constructs material nodes and supply unit nodes based on the material information included in the historical planning and scheduling data and the relationship between each supply unit and the material identified by the cycle time. The demand, supply and capacity included in the historical planning and scheduling data are used to characterize the material nodes, supply unit nodes and the edges between the nodes, in order to predict the bottleneck material and bottleneck supply unit corresponding to the input demand.

[0053] The grouping and sorting module is used to group and sort the pending demands based on the score of the supply organization corresponding to the bottleneck supply organization. This results in multiple sorted demand subgroups, where the bottleneck supply organization is used to characterize the bottleneck in the supply organization's material production capacity.

[0054] The weight calculation module is used to obtain each demand subgroup in sorting order, and for each demand included in the current subgroup, determine the weight corresponding to the demand based on the number of times the material corresponding to the demand hits the bottleneck material and the score corresponding to the bottleneck material.

[0055] The supply plan acquisition module is used to heuristically solve the supply plan corresponding to each demand in the current subgroup based on the supply information of each supply institution, using the material constraints, the capacity constraints, and the date constraints as constraints and the demand achievement rate with the weights as the optimization objective, so as to obtain the supply plan corresponding to the demand to be processed.

[0056] Thirdly, embodiments of this application provide an electronic device, including:

[0057] Memory, used to store computer programs;

[0058] When a processor executes a program stored in memory, it implements any of the methods described in the first aspect above.

[0059] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements any of the methods described in the first aspect above.

[0060] Beneficial effects of the embodiments in this application:

[0061] The solution provided in this application, when heuristically solving the supply plan corresponding to the demand to be processed, not only considers the dependencies between materials, the capacity constraints of each supplier in providing each material, and the time limits for each supplier in providing each material, but also uses a pre-trained knowledge reasoning model to predict potential bottleneck suppliers and bottleneck materials. Based on the scores corresponding to the bottleneck suppliers, the demands to be processed are grouped and sorted, each demand subgroup is obtained according to the sorting order, and the weight of each demand is determined based on the bottleneck material. The weighted demand achievement rate is used as the optimization objective. Therefore, a reasonable supply plan can be obtained for both demands constrained by bottleneck materials and bottleneck suppliers, and demands not constrained by bottleneck materials and bottleneck suppliers, enabling each supplier to supply reasonably, achieving supply and demand balance, and thus achieving a planning schedule with high executability and high achievement rate. Of course, implementing any product or method of this application does not necessarily require achieving all the advantages described above simultaneously. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other embodiments can be obtained based on these drawings.

[0063] Figure 1 A flowchart illustrating a method for determining a project schedule as provided in an embodiment of this application;

[0064] Figure 2 for Figure 1 A specific flowchart of step S104 in the illustrated embodiment;

[0065] Figure 3 for Figure 1 A specific flowchart of step S106 in the illustrated embodiment;

[0066] Figure 4 for Figure 3 A specific flowchart of step S302 in the illustrated embodiment;

[0067] Figure 5 For based on Figure 1 A specific flowchart for constructing a knowledge reasoning model in the illustrated embodiment;

[0068] Figure 6 For based on Figure 5 A schematic diagram of constructing a knowledge reasoning model in the embodiment shown;

[0069] Figure 7 for Figure 5 A specific flowchart of step S503 in the illustrated embodiment;

[0070] Figure 8 This is a specific flowchart for obtaining predicted demand provided in an embodiment of this application;

[0071] Figure 9 For based on Figure 8 A schematic diagram illustrating the acquisition of predicted demand in the embodiment shown;

[0072] Figure 10 A specific flowchart for obtaining the modified procurement supply provided in an embodiment of this application;

[0073] Figure 11 For based on Figure 10 The illustrated embodiment is a schematic diagram of obtaining the modified procurement supply.

[0074] Figure 12 A schematic diagram illustrating a method for determining a schedule provided in an embodiment of this application;

[0075] Figure 13 For based on Figure 12 A schematic diagram of the mechanism model of the embodiment shown;

[0076] Figure 14 For based on Figure 12 A schematic diagram of the planning and scheduling of the embodiment shown;

[0077] Figure 15 A schematic diagram of a scheduling determination device provided in an embodiment of this application;

[0078] Figure 16 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0079] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art based on this application are within the scope of protection of this application.

[0080] To improve the executability and achievement rate of project scheduling, embodiments of this application provide a method, apparatus, electronic device, computer-readable storage medium, and computer program product for determining project scheduling. The method for determining project scheduling provided by embodiments of this application will be described first below.

[0081] The method for determining a schedule provided in this application can be applied to any device that needs to determine a schedule, such as a processor, server, desktop computer, laptop computer, etc., without specific limitations. For clarity, it will be referred to as an electronic device.

[0082] like Figure 1 As shown, a method for determining a project schedule includes:

[0083] S101, Obtain pending demand, supply information, and material information;

[0084] The supply information includes at least the supply, capacity, cycle time and supply time information of each supplying organization. The cycle time is used to represent the working hours required for the supplying organization to provide each material, and / or to represent the quantity of each material provided by the supplying organization per unit time.

[0085] S102, determine material constraints based on the demand transmission relationship between the materials identified by the material information, determine capacity constraints based on the capacity and the cycle time, and determine date constraints based on the supply time relationship between the materials identified by the supply time information.

[0086] The material constraints are used to represent the dependencies between materials, the capacity constraints are used to represent the ability constraints of each supplying organization to provide each material, and the date constraints are used to represent the time limits for each supplying organization to provide each material.

[0087] S103, input the demand to be processed, the supply of each supply institution and the production capacity into the pre-trained knowledge reasoning model to obtain the bottleneck material and bottleneck supply institution corresponding to the demand to be processed.

[0088] The knowledge reasoning model constructs material nodes and supply organization nodes based on the material information included in the historical planning and scheduling data and the relationship between each supply organization and material identified by the cycle time. The material nodes, supply organization nodes and the edges between the nodes are characterized by the demand, supply and capacity included in the historical planning and scheduling data, in order to predict the bottleneck material and bottleneck supply organization corresponding to the input demand.

[0089] S104, based on the score of the supply organization corresponding to the demand to be processed hitting the bottleneck supply organization, the demand to be processed is grouped and sorted to obtain multiple subgroups of demand after sorting.

[0090] The bottleneck supply mechanism is used to characterize the bottleneck in the supply mechanism's material production capacity;

[0091] S105, obtain each demand subgroup according to the sorting order, and for each demand included in the current subgroup, determine the weight corresponding to the demand based on the number of times the material corresponding to the demand hits the bottleneck material and the score corresponding to the bottleneck material.

[0092] S106, using the material constraints, capacity constraints, and date constraints as constraints, and the demand achievement rate with the weights as the optimization objective, based on the supply information of each supply organization, the supply plan corresponding to each demand in the current subgroup is heuristically solved to obtain the supply plan corresponding to the demand to be processed.

[0093] As can be seen, in the solution provided in this application embodiment, when heuristically solving the supply plan corresponding to the demand to be processed, not only are the dependencies between materials, the capacity constraints of each supplying organization to provide each material, and the time constraints of each supplying organization to provide each material considered, but also the potential bottleneck supplying organizations and bottleneck materials are predicted by the pre-trained knowledge reasoning model. Based on the scores corresponding to the bottleneck supplying organizations, the demand to be processed is grouped and sorted, and each demand subgroup is obtained according to the sorting order. The weight of each demand is determined based on the bottleneck material, and the weighted demand achievement rate is used as the optimization objective. Therefore, a reasonable supply plan can be obtained for both demand restricted by bottleneck materials and bottleneck supplying organizations and demand not restricted by bottleneck materials and bottleneck supplying organizations, so that each supplying organization can supply reasonably, achieve supply and demand balance, and thus achieve a plan scheduling with high executability and high achievement rate.

[0094] In step S101, the electronic device can acquire pending demands, supply information, and material information. Pending demands may include multiple demands, and the types of demands can be delivery demands, reserved demands, sales demands, allocation demands, forecast demands, etc., without specific limitations. Different types of demands can be assigned different priorities according to business logic.

[0095] Supply information includes at least the supply, capacity, cycle time, and supply time information of each supplying entity. Capacity can represent the maximum quantity of each material that a supplying entity can provide within a certain time period, or the maximum time limit for providing each material. Cycle time represents the working hours required for a supplying entity to provide each material; specifically, it can be the working hours required for a supplying entity to provide one unit of material, and / or the quantity of each material that a supplying entity provides per unit of time. For example, cycle time is the time required for supplying entity A to produce one monitor. The type of supply can be inventory, procurement, etc., without specific limitations. The supplying entity is the supplier, which can be a supplier, factory, work center, etc., without specific limitations.

[0096] In one implementation, the electronic device can obtain pending demand, supply information, and material information from a business system. The business system stores data related to business needs, such as supply organization, demand, supply, capacity, cycle time, BOM (bill of material), master data (demand master data and material master data), and work calendar data. The master data can represent the basic attributes of demand and materials.

[0097] For example, the business system stores demand 1, demand 2 and demand 3. Suppliers A, B and C provide the materials required for each demand. Electronic devices can obtain demand 1, demand 2 and demand 3 from the business system, and obtain information on the supply, capacity, cycle time and supply time of materials from suppliers A, B and C.

[0098] In step S102, the electronic device can determine material constraints based on the demand transfer relationships between the materials identified in the material information. Material constraints represent the dependencies between materials. For example, the production of upper-level materials depends on lower-level materials. As one implementation, the electronic device can determine material constraints based on the relationships between materials in the BOM. The BOM can show which components and raw materials constitute a product, and can show the temporal and quantitative relationships between these components.

[0099] For example, a Bill of Materials (BOM) might show that product A consists of component 1, component 2, and component 3. Component 1 is composed of material a and material b, component 2 is composed of material b, component 3 is composed of material b and material c, and material c is composed of material e and material f. Thus, electronic devices can determine, based on the demand transfer relationships between these materials, that the production of component 1 depends on material a and material b, the production of component 2 depends on material b, the production of component 3 depends on material b and material c, and the production of material c depends on material e and material f. Therefore, there are material constraints between these materials.

[0100] Electronic equipment can determine capacity constraints based on production capacity and cycle time. Capacity constraints represent the ability of each supplier to provide each material. In other words, electronic equipment can determine the time and quantity required for each supplier to provide each material. The materials provided by each supplier can be from inventory or purchased; no specific limitations are made here. If the materials provided by the supplier are for production, then capacity constraints may also include the number of workers, etc., which are also not specifically limited here.

[0101] For example, the production of product A requires materials a, b, and c. Supply mechanism 1 provides materials a and b, supply mechanism 2 provides materials b and c, and supply mechanism 3 provides materials a, b, and c. Supply mechanism 1 can provide n1 units of material a and n2 units of material b per day, supply mechanism 2 can provide n3 units of material b and n4 units of material c within a week, and supply mechanism 3 can provide n3 units of material a, n5 units of material b, and n6 units of material c within three days. Electronic equipment can thus determine the quantities of materials a and b that supply mechanism 1 can provide within a certain timeframe. Similarly, it can determine the quantities of materials b and c that supply mechanism 2 and supply mechanism 3 can provide within a certain timeframe, thus determining the capacity constraints.

[0102] Electronic devices can determine date constraints based on the supply time relationships of each material identified by the supply time information. Date constraints represent the time limits for each supplier to provide each material. That is, the supply times of each supplier may differ; therefore, electronic devices can determine the supply times of each supplier to establish date constraints. The supply times of each supplier may include working days or may not; both of these are considered date constraints.

[0103] For example, a Bill of Materials (BOM) might include materials a, b, and c. The production of material a depends on materials d and e. Supplying agency 1 provides materials a and b, supplying agency 2 provides materials b and c, and supplying agency 3 provides material c. Since the production of material a depends on materials d and e, the production time of material a is constrained by the supply time of materials d and e. The composition of the product is constrained by the time constraints imposed by supplying agencies 1 (materials a and b), 2 (materials b and c), and 3 (material c).

[0104] For example, assembling a computer requires a monitor, a host, a keyboard, and a mouse. If it takes 3 days for supplier 1 to produce the monitor, 6 days for supplier 2 to produce the host, and 2 days for supplier 3 to purchase the keyboard and mouse, then the electronic device can determine the date constraints based on the supply time relationship of the monitor, host, keyboard, and mouse, and thus determine the time when the computer assembly is completed.

[0105] Since potential bottleneck suppliers and bottleneck capacity may affect the supply plan for the demand to be processed, the electronic device can consider the bottleneck suppliers and bottleneck materials in the demand to be processed when solving the supply plan corresponding to the demand to be processed. In step S103, the electronic device can input the demand to be processed, the supply and capacity of each supplier into a pre-trained knowledge reasoning model to obtain the bottleneck materials and bottleneck suppliers corresponding to the demand to be processed, and can obtain the scores corresponding to the bottleneck materials and the bottleneck suppliers.

[0106] Among them, the knowledge reasoning model constructs material nodes and supply organization nodes based on the material information included in the historical planning and scheduling data and the relationship between each supply organization and material identified by the cycle time. The demand, supply and capacity included in the historical planning and scheduling data are used to characterize the material nodes, supply organization nodes and the edges between nodes, which are used to predict the bottleneck materials and bottleneck supply organizations corresponding to the input demand.

[0107] A bottleneck supplier is used to characterize a supplier whose production capacity for materials is limited, meaning that the supplier's ability to produce materials is constrained. For example, when a supplier's production quantity is limited due to factors such as equipment, supply chain issues, or site constraints, it becomes a bottleneck supplier. Historical planning and scheduling data can include demand, supply, capacity, cycle time, supply timing information, and material information for each supplier.

[0108] For example, the pending demands include demand 1, demand 2, and demand 3. The materials required for demand 1 include material a, material b, material c, and material d; the materials required for demand 2 include material a and material c; and the materials required for demand 3 include material c and material d. Supply organization 1 provides material a and material b, supply organization 2 provides material a and material c, and supply organization 3 provides material a, material c, and material d.

[0109] The electronic device inputs the current group of demands, including demand 1, demand 2 and demand 3, supply mechanism 1, supply mechanism 2 and the supply of supply mechanism, and the production capacity of materials a, b, c and d into a pre-trained knowledge reasoning model, and can obtain that the bottleneck material corresponding to the demand to be processed is material b and the bottleneck supply mechanism is supply mechanism 1.

[0110] After obtaining the bottleneck supply organization corresponding to the demand to be processed and the score corresponding to the bottleneck supply organization, the electronic device can group and sort the demand to be processed based on the score of the bottleneck supply organization corresponding to the supply organization corresponding to the demand to be processed, and obtain multiple subgroups of demand after sorting, i.e., execute step S104.

[0111] In one implementation, the electronic device can divide the demand to be processed into multiple demand subgroups based on the bottleneck supply organization hit by the supply organization corresponding to the demand to be processed, and then sort the multiple demand subgroups according to the score of hitting the bottleneck supply organization.

[0112] For example, the demand to be processed includes demand 1 to demand 6. The supply institutions corresponding to demand 1 and demand 6 are supply institution 1 and supply institution 4, respectively. The supply institution corresponding to demand 2 and demand 5 is supply institution 3, the supply institution corresponding to demand 3 is supply institution 1 and supply institution 4, and the supply institution corresponding to demand 4 is supply institution 1 and supply institution 2. The bottleneck supply institutions are supply institution 1, supply institution 2, supply institution 3, and supply institution 4, with corresponding scores of 8, 5, 3, and 1, respectively.

[0113] The electronic device can divide requirements 1, 3, and 6 into a first subgroup, requirements 2 and 5 into a second subgroup, and requirement 4 into a third subgroup. The first subgroup has a score of 9, the second subgroup has a score of 3, and the third subgroup has a score of 13. Therefore, the electronic device can obtain multiple sorted requirement subgroups, namely the third subgroup, the first subgroup, and the second subgroup.

[0114] The division of demand subgroups takes into account bottleneck supply mechanisms. For each demand included in a demand subgroup, the bottleneck materials included in that demand can be further considered. After obtaining multiple sorted demand subgroups, the electronic device can obtain each demand subgroup in the sorting order. For each demand included in the current subgroup, the weight corresponding to that demand is determined based on the number of times the material corresponding to that demand hits the bottleneck material and the score corresponding to the bottleneck material, i.e., step S105 is executed. Among them, the higher the score of hitting the bottleneck material, the greater the weight, which means that there are more bottleneck materials in that demand, which may lead to a low demand fulfillment rate.

[0115] In one implementation, the electronic device can record the number of times each material in the lower layer hits the bottleneck material and the sum of the scores corresponding to the hits on the bottleneck material, based on the bill of materials for each demand, and then determine the weight corresponding to each demand based on the sum of the scores.

[0116] For example, the current subgroup includes demand 1 and demand 2. The bill of materials (BOM) for material a required by demand 1 includes materials b and c, and the BOM for material b includes materials d and e. The BOM for material f required by demand 2 includes materials b and g, and the BOM for material f includes materials d and e. Materials c, d, and e are bottleneck materials, with corresponding scores of 5, 3, and 1 respectively. For demand 1, materials c, d, and e are bottleneck materials, and the electronic equipment determines that demand 1 has a score of 9. For demand 2, materials d and e are bottleneck materials, and the electronic equipment determines that demand 2 has a score of 8. Weights can be determined based on the scores of demand 1 and demand 2, i.e., demand 1 has a weight of 1, and demand 2 has a weight of 0.89.

[0117] After determining the weight of each demand in the current subgroup, in step S106, the electronic device can use material constraints, capacity constraints, and date constraints as constraints, and the weighted demand achievement rate as the optimization objective, and based on the supply information of each supply organization, perform heuristic solution for the supply plan corresponding to each demand in the current subgroup to obtain the supply plan corresponding to the demand to be processed.

[0118] In other words, electronic devices can rationally allocate the supply of each supplier according to the material requirements of each demand, based on the dependencies between materials, the capacity constraints of each supplier in providing each material, and the time constraints of each supplier in providing each material, with the weighted demand fulfillment rate as the optimization objective, to obtain the supply plan corresponding to the demand to be processed. Here, the weight is the weight corresponding to each demand included in the current subgroup.

[0119] Heuristic algorithms are proposed in contrast to optimization algorithms. An optimization algorithm for a problem finds the optimal solution for every instance of the problem. Heuristic algorithms can include ant colony optimization, simulated annealing, genetic algorithms, etc. The encoding for heuristic solutions differs depending on the heuristic algorithm; therefore, the corresponding encoding needs to be executed based on the heuristic algorithm.

[0120] For example, genetic algorithms, as a typical heuristic algorithm, solve optimization problems by following the steps of encoding, generating an initial population, selection, crossover, mutation, and fitness updates. Specifically, based on the optimization problem (i.e., demand fulfillment rate), the optimization variable (i.e., supply plan) is encoded to form genes and chromosomes; an initial population is obtained using a certain method (such as randomness), containing multiple individuals with different genes; in the algorithm's iterative phase, in each iteration, some individuals are selected from the population as parents, and their genes are crossovered and mutated to obtain offspring individuals. The fitness values ​​of the parents and offspring individuals are calculated, and the better individuals are selected from the parents and offspring individuals according to a certain method (such as elite retention) and added to the population to form a new population; the above iterative process is repeated until the algorithm converges or the specified number of iterations is reached, at which point the algorithm terminates. Finally, the obtained optimal individuals are decoded to obtain the optimal solution corresponding to the optimization problem (i.e., the supply plan corresponding to each demand) and the optimal objective function (i.e., the optimal demand fulfillment rate).

[0121] For example, after determining that the weight corresponding to demand 1 is 1 and the weight corresponding to demand 2 is 0.89, the electronic equipment can, for the demands 1 and 2 included in the current subgroup, use material constraints, capacity constraints, and date constraints as constraints, and the achievement rate of the weighted demands 1 and 2 as the optimization objective. Based on the supply information of each supply unit, it can heuristically solve for the supply plan corresponding to demands 1 and 2 included in the current subgroup, and obtain the supply plan corresponding to demands 1 and 2. This allows for prioritizing demands with higher weights, enabling the rational allocation of materials and capacity among the supply units.

[0122] For example, the current subgroup includes demand 1, with a weight of 0.92. Demand 1 requires 3 units of material a, 5 units of material b, and 3 units of material c. Supply agency 1 provides materials a and b. The production of material a in supply agency 1 depends on materials d and e. Supply agency 2 provides materials b and c. Supply agency 3 provides materials a, b, and c. Material a provided by supply agency 3 is procured and can be supplied on time. Supply agency 1 can provide 1 unit of material a and 2 units of material b per day when materials d and e are sufficient. Supply agency 2 can provide 6 units of material b and 10 units of material c within five days. Supply agency 3 can provide 3 units of material a, 3 units of material b, and 1 unit of material c within three days.

[0123] Since supplying material a by supplier 1 is constrained by the supply times of materials d and e, material a can be supplied by supplier 3. Supplier 1 can supply 6 units of material b within three days, which is less time than supplier 2 to supply 6 units of material b, meaning that supplier 1 supplies material b earlier than supplier 2. Similarly, supplier 2 can supply material c earlier than supplier 3. Therefore, under the constraints of material constraints, capacity constraints, and date constraints, and with the weighted achievement rate of demand 1 as the optimization objective, the electronic device heuristically solves the supply plan corresponding to demand 1 based on the supply information of each supplier, obtaining a supply plan for demand 1 where supplier 1 supplies material b, supplier 2 supplies material c, and supplier 3 supplies material a.

[0124] In one implementation, after the electronic device receives the supply plan corresponding to the demand to be processed, it can modify the supply plan according to the business needs to obtain the modified supply plan, and then store the modified supply plan in the business system to realize the modified supply plan.

[0125] In this embodiment, when heuristically solving the supply plan corresponding to the demand to be processed, not only are the dependencies between materials, the capacity constraints of each supplying organization to provide each material, and the time limits for each supplying organization to provide each material considered, but also the potential bottleneck supplying organizations and bottleneck materials are predicted by a pre-trained knowledge reasoning model. Based on the scores corresponding to the bottleneck supplying organizations, the demand to be processed is grouped and sorted, and each demand subgroup is obtained according to the sorting order. The weight of each demand is determined based on the bottleneck material, and the weighted demand achievement rate is used as the optimization objective. Therefore, a reasonable supply plan can be obtained for both demand restricted by bottleneck materials and bottleneck supplying organizations and demand not restricted by bottleneck materials and bottleneck supplying organizations, so that each supplying organization can supply reasonably, achieve supply and demand balance, and thus achieve a planning schedule with high executability and high achievement rate.

[0126] As one implementation method of this application, such as Figure 2 As shown, the step of grouping and sorting the unprocessed demands based on the scores of the supply organizations corresponding to the bottleneck supply organizations, to obtain multiple sorted demand subgroups, may include:

[0127] S201, for each demand included in each group of demands, determine the supply organization corresponding to the demand based on the material information corresponding to the demand and the cycle time;

[0128] Each group of requirements is obtained by grouping the requirements to be processed according to priority.

[0129] Different types of requirements have different priorities. The requirements to be processed can include requirements with different priorities. Electronic devices can group requirements with the same priority into a group and sort the requirements in each group in order of priority from high to low or from low to high. In this way, the requirements in each group have different priorities, while the requirements in each group have the same priority.

[0130] Since the cycle time is used to represent the time required for a supply organization to provide each material, and / or to represent the quantity of each material provided by the supply organization per unit time, i.e., there is a correspondence between supply organizations and materials, electronic equipment can determine the supply organization that provides the material based on the material. For each demand included in each demand group, electronic equipment can determine the supply organization corresponding to that demand based on the material corresponding to the demand and the cycle time of each supply organization included in the supply information.

[0131] For example, for demand 1 in the current group of demand, the materials required for demand 1 are material a and material b. According to the cycle time, it can be determined that supply mechanism 1 provides material a and supply mechanism 2 provides material b. That is, the supply mechanism corresponding to demand 1 is supply mechanism 1 and supply mechanism 2.

[0132] For example, regarding demand 1 and demand 2 in the current demand group, the material required for demand 1 is material a, and the material required for demand 2 is material b. According to the cycle time, it can be determined that supply organization 1 provides material a and supply organization 2 provides material b. That is, the supply organization corresponding to demand 1 is supply organization 1, and the supply organization corresponding to demand 2 is supply organization 2.

[0133] S202, the same demand from the supplying institution and the bottleneck supplying institution is divided into a group of demand to obtain a demand subgroup;

[0134] In one implementation, for each group of demand, after the electronic device determines the supply mechanism corresponding to each demand included in the group of demand, the demands with the same supply mechanism as the bottleneck supply mechanism can be divided into one group, and other demands can be divided into another group, thus obtaining multiple demand subgroups.

[0135] For example, the current demand group includes demands 1 to 10. Demands 1 and 2 require material a, demands 3 and 6 require material b, demand 5 requires material c, demands 4 and 7 require material d, and demands 8 to 10 require material e. Supply mechanism 1 provides material c, supply mechanism 2 provides material e, supply mechanism 3 provides material b, and supply mechanism 4 provides materials a and d. The bottleneck supply mechanisms are supply mechanisms 2 and 4. Electronic equipment can group demands with the same supply mechanisms as the bottleneck supply mechanisms into groups: demand 1, demand 2, demand 4, and demand 7 are grouped together; demand 8 to 10 are grouped together; and demand 3, demand 5, and demand 6 are grouped together.

[0136] In another implementation, for the current group of demand, after the electronic device determines the supply organization corresponding to each demand, the demands with the same supply organization can be divided into a group to obtain multiple demand subgroups.

[0137] For example, the current demand group includes demands 1 to 10. Demands 1 and 6 require material a, demands 2 and 5 require material b, demands 3 and 9 require material c, demands 4 and 7 require material d, demand 8 requires material e, and demand 10 requires material f. Supply organization 1 provides materials a and c, supply organization 2 provides materials b and d, supply organization 3 provides material e, and supply organization 4 provides material f. Electronic equipment can be grouped together based on the demand from the same supply organization; that is, demands 1, 3, 6, and 9 are grouped together; demands 2, 4, 5, and 7 are grouped together; demand 8 is grouped together; and demand 10 is grouped together.

[0138] S203, sort the demand subgroups according to the scores of the corresponding bottleneck supply institutions to obtain multiple sorted demand subgroups.

[0139] Different bottleneck suppliers indicate the degree to which different suppliers constitute a bottleneck. In one implementation, the electronic device can assign different scores to different bottleneck suppliers, and sort the demand subgroups according to the scores of the corresponding bottleneck suppliers from high to low, resulting in multiple sorted demand subgroups. The first demand subgroup obtained by the electronic device is the demand subgroup with the highest score.

[0140] For example, an electronic device receives multiple demand subgroups: demand subgroups including demand 1, demand 2, demand 4, and demand 7; demand subgroups including demand 8 to demand 10; and demand subgroups including demand 3, demand 5, and demand 6. Supply mechanism 2 and supply mechanism 4 have scores of 5 and 3 respectively, while the score for demand subgroups excluding bottleneck supply mechanisms is 0. The electronic device can determine the score of the bottleneck supply mechanism corresponding to each demand subgroup: the score for demand subgroups including demand 1, demand 2, demand 4, and demand 7 is 5; the score for demand subgroups including demand 8 to demand 10 is 3; and the score for demand subgroups including demand 6 and demand 7 is 0. The electronic device then sorts these three demand subgroups according to the scores of their corresponding bottleneck supply mechanisms, resulting in multiple sorted demand subgroups.

[0141] In another implementation, the electronic device can sort the demand subgroups from highest to lowest according to the scores of the different supply institutions that become bottleneck supply institutions, thus obtaining multiple sorted demand subgroups.

[0142] For example, an electronic device receives multiple demand subgroups: subgroups containing demands 1, 3, 6, and 9; subgroups containing demands 2, 4, 5, and 7; subgroups containing demand 8; and subgroups containing demand 10. Supply agencies 1-4 are assigned bottleneck supply agency scores of 10, 8, 5, and 3, respectively. The electronic device can sort these demand subgroups based on their corresponding bottleneck supply agency scores, resulting in multiple sorted demand subgroups.

[0143] As can be seen, in this embodiment, for each demand within each demand group, the electronic device can determine the corresponding supply mechanism based on the material information and cycle time of that demand. Demands with the same supply mechanism as the bottleneck supply mechanism are grouped into a demand subgroup. The demand subgroups are then sorted according to the scores of their corresponding bottleneck supply mechanisms, resulting in multiple sorted demand subgroups. Since the bottleneck supply mechanism is considered when grouping and sorting the demand subgroups, the demand subgroups with higher scores from the bottleneck supply mechanism can be processed first, thereby enabling each supply mechanism to achieve a reasonable allocation of materials and production capacity.

[0144] As one implementation method of this application, such as Figure 3 As shown, the above steps, which use the material constraints, capacity constraints, and date constraints as constraints, and the demand fulfillment rate with the weights as the optimization objective, and heuristically solve for the supply plan corresponding to each demand in the current subgroup based on the supply information of each supply institution, to obtain the supply plan corresponding to the demand to be processed, may include:

[0145] S301, Obtain the highest priority set of requirements as the current set of requirements;

[0146] In one implementation, to process high-priority requests promptly, the electronic device can select the highest-priority group of requests as the current group of requests. For example, if there are three groups of requests to be processed, with a highest priority of 1, the electronic device can select the group of requests with priority 1 as the current group of requests.

[0147] S302, for the demand subgroups included in the current demand group, with the material constraints, the capacity constraints and the date constraints as constraints, and with the demand achievement rate with the weights corresponding to the demand subgroups included in the current demand group as the optimization objective, based on the current supply information of each supply institution, the supply plan corresponding to the demand subgroups included in the current demand group is heuristically solved to obtain the supply plan corresponding to the current demand group.

[0148] After obtaining the current group demand, the electronic device can, based on the demand subgroups included in the current group demand, and according to the dependencies between the materials in the current group demand, the capacity constraints of each supplier to provide each material, and the time limits of each supplier to provide each material, optimize the supply plan corresponding to the current group demand by taking the demand achievement rate corresponding to the weighted demand subgroups included in the current group demand as the optimization target and allocating the supply of each supplier according to the materials required for the current group demand.

[0149] For example, the current demand group includes demand subgroup 1, demand subgroup 2, and demand subgroup 3. Demand subgroup 1 includes demand 1, demand subgroup 2 includes demand 2, and demand subgroup 3 includes demand 3. The weights of demand 1, demand 2, and demand 3 are 1, 0.89, and 0.95, respectively. The materials required by demand subgroup 1 include material a, material b, material c, and material d. The materials required by demand subgroup 2 include material a and material c, and the materials required by demand subgroup 3 include material c and material d. The production of material b depends on material d. Supply mechanism 1 provides material a and material b, supply mechanism 2 provides material b and material c, and supply mechanism 3 provides material a, material c, and material d.

[0150] The electronic device can, for the current demand group including demand subgroups 1, demand subgroup 2, and demand subgroup 3, based on the dependencies between materials a, b, c, and d, and considering the capacity constraints and time limits of supplying materials a, b, c, and d by supplying agencies 1, 2, and 3, optimize the supply plan for the current demand group by taking the weighted demand achievement rate of demand subgroups 1, 2, and 3 as the optimization objective. According to the materials required by demand subgroups 1, 2, and 3, the supply of supplying agencies 1, 2, and 3 is rationally allocated to obtain the supply plan for the current demand group. That is, the supply plan for demand 1 is that supplying agency 2 provides material b, and supplying agency 3 provides materials a, c, and d; the supply plan for demand 2 is that supplying agencies 1 and 3 provide material a, and supplying agency 2 provides material c; and the supply plan for demand 3 is that supplying agency 3 provides materials c and d.

[0151] S303, subtract the supply information of each supply institution required by the supply plan to obtain the current supply information of each supply institution;

[0152] After the electronic device rationally allocates the supply of each supply unit according to the materials required by the demand subgroups included in the current demand, the supply of each supply unit will decrease accordingly. The electronic device can then deduct the supply information of each supply unit required by the supply plan to obtain the current supply information of each supply unit.

[0153] For example, the current demand group includes demand subgroup 1, demand subgroup 2, and demand subgroup 3. Demand subgroup 1 includes demand 1, demand subgroup 2 includes demand 2, and demand subgroup 3 includes demand 3. The materials required for demand 1 include n1 units of material a, n2 units of material b, n3 units of material c, and n2 units of material d. The materials required for demand 2 include n4 units of material a and n5 units of material c. The materials required for demand 3 include n1 units of material c and n2 units of material d. Supply mechanism 1 can provide N1 units of material a and N2 units of material b, supply mechanism 2 can provide N3 units of material b and N4 units of material c, and supply mechanism 3 can provide N5 units of material a, N6 units of material c, and N7 units of material d.

[0154] The supply plan corresponding to demand 1 is that supply institution 2 provides material b, and supply institution 3 provides material a, material c, and material d. The supply plan corresponding to demand 2 is that supply institution 1 provides material a, and supply institution 2 provides material c. The supply plan corresponding to demand 3 is that supply institution 3 provides material c and material d.

[0155] After the electronic device rationally allocates the supply of supply mechanism 1, supply mechanism 2 and supply mechanism 3 according to the materials required by demand subgroup 1, demand subgroup 2 and demand subgroup 3, the electronic device can deduct the supply information of supply mechanism 1, supply mechanism 2 and supply mechanism 3 required by the supply plan to obtain the current supply information of each supply mechanism. That is, supply mechanism 1 can provide N1-n4 material a and N2 material b, supply mechanism 2 can provide N3-n2 material b and N4-n5 material c, and supply mechanism 3 can provide N5-n1 material a, N6-n3-n1 material c and N7-n2-n2 material d.

[0156] S304. Obtain the next group of demands in descending order of priority as the current group of demands, and return the demand subgroups included in the current group of demands. Using the material constraints, capacity constraints, and date constraints as constraints, and the demand achievement rate with the weights corresponding to the demand subgroups included in the current group of demands as the optimization objective, heuristically solve the supply plan corresponding to the demand subgroups included in the current group of demands based on the current supply information of each supply institution, and obtain the supply plan corresponding to the current group of demands. This process continues until all groups of demands have been processed.

[0157] After the electronic equipment finishes processing the current group of demands, it can continue to process the next group of demands. That is, it obtains the next group of demands in descending order of priority and uses it as the current group of demands. Then, for the demand subgroups included in the current group of demands, it uses material constraints, capacity constraints, and date constraints as constraints, and the weighted demand achievement rate of the demand subgroups included in the current group of demands as the optimization objective. Based on the current supply information of each supply institution, it uses heuristic solution to solve the supply plan corresponding to the demand subgroups included in the current group of demands, and obtains the steps of the supply plan corresponding to the current group of demands until all groups of demands have been processed.

[0158] For example, the pending demands include three groups of demands. After the electronic device processes the demand with priority 1, it can acquire the demand with priority 2, which is then taken as the current demand group. The current demand group includes demand subgroup 4 and demand subgroup 5. For this demand group, considering the dependencies between materials in demand subgroups 4 and 5, the capacity constraints of each supply organization in providing each material, and the time limits for each supply organization in providing each material, and using the weighted demand fulfillment rate corresponding to demand subgroups 4 and 5 as the optimization objective, the supply plan corresponding to demand subgroups 4 and 5 is heuristically solved based on the current supply information of each supply organization to obtain the supply plan corresponding to the current demand group, until all three demand groups are processed.

[0159] As can be seen, in this embodiment, the electronic device can process each group of demands according to their priority, from highest to lowest. For the demand subgroups included in the current demand group, material constraints, capacity constraints, and date constraints are used as constraints. The optimization objective is the weighted demand achievement rate corresponding to the demand subgroups included in the current demand group. Based on the current supply information of each supply organization, a heuristic solution is performed on the supply plan corresponding to the demand subgroups included in the current demand group to obtain the supply plan for the current demand group. This process continues until multiple demand groups have been processed. In this way, for the planning domain, highly executable supply plans are output for various demands, enabling each supply organization to supply reasonably, achieving supply and demand balance, and thus improving the executability and achievement rate of the planning schedule.

[0160] As one implementation method of this application, such as Figure 4 As shown, the steps described above, for the demand subgroups included in the current demand group, using the material constraints, capacity constraints, and date constraints as constraints, and taking the demand achievement rate corresponding to the demand subgroups included in the current demand group with the weighted value as the optimization objective, and based on the current supply information of each supply institution, to heuristically solve the supply plan corresponding to the demand subgroups included in the current demand group, to obtain the supply plan corresponding to the current demand group, may include:

[0161] S401, for the current group requirements, obtain the first requirement subgroup according to the sorting order, and use it as the current subgroup;

[0162] To rationally allocate supply from various supply entities, priority can be given to the demand from bottleneck supply entities with higher scores. For the current group of demand, electronic devices can obtain the first demand subgroup according to the sorting order, using it as the current subgroup; that is, priority is given to processing the demand subgroup with higher bottleneck supply entity scores.

[0163] For example, the current group of requirements includes requirement 1, requirement 2, requirement 3, and requirement 4. The current group of requirements comprises three subgroups: requirement 1 and requirement 2 form the first subgroup, requirement 3 the second subgroup, and requirement 4 the third subgroup. The electronic device can sequentially acquire the first subgroup of requirements, i.e., acquire requirement 1 and requirement 2, and then use requirement 1 and requirement 2 as the current subgroup.

[0164] S402, for the current subgroup, with the material constraints, the capacity constraints and the date constraints as constraints, and with the weighted demand achievement rate corresponding to the current subgroup as the optimization objective, based on the current supply information of each supply institution, the supply plan corresponding to the current subgroup is heuristically solved to obtain the supply plan corresponding to the current subgroup.

[0165] For example, the current subgroup includes demand 1 and demand 2, with weights of 1 and 0.89 respectively. The materials required for demand 1 include material a, material b and material c, and the materials required for demand 2 include material c and material d. The production of material a depends on materials b and c. Supply organization 1 provides material a and material b, supply organization 2 provides material c, and supply organization 3 provides material b and material d.

[0166] The electronic device can, for the current subgroup including demand 1 and demand 2, based on the dependencies between materials a, b, c, and d, and considering the capacity constraints and time limits of supply agencies 1, 2, and 3 in providing the corresponding materials a, b, c, and d, with the weighted achievement rate of demand 1 and demand 2 as the optimization objective, rationally allocate the supply of materials from supply agencies 1, 2, and 3 according to the materials required for demand 1 and demand 2, to obtain the supply plan corresponding to demand 1 and demand 2. That is, the supply plan for demand 1 is that supply agency 1 provides materials a and b, and supply agency 2 provides materials c; the supply plan for demand 2 is that supply agency 2 provides materials c, and supply agency 3 provides materials d.

[0167] For example, the current subgroup includes demand 1, demand 2, demand 3, and demand 4, and we obtain the weights of demand 1, demand 2, demand 3, and demand 4. The material required for demand 1 is material a, and the production of material a depends on materials b and c. The material required for demand 2 is material d, the material required for demand 3 is material e, and the production of material e depends on materials b and f. The material required for demand 4 is material g. Supply organization 1 provides materials a, b, and g; supply organization 2 provides materials c, d, and e; and supply organization 3 provides material f.

[0168] The electronic equipment can, for the current subgroup containing demand 1, demand 2, demand 3, and demand 4, provide the corresponding materials a, b, c, d, e, f, and g based on the dependencies between materials a, b, c, d, e, f, and g. It can also determine the time constraints for supply mechanisms 1, 2, and 3 to provide the corresponding materials a, b, c, d, e, f, and g, using weighted demand 1, demand 2, demand 3, and demand 4. The achievement rate is the optimization target. According to the materials required by demand 1, demand 2, demand 3, and demand 4, the supply of supply institutions 1, 2, and 3 is reasonably allocated to obtain the supply plan for demand 1, demand 2, demand 3, and demand 4. That is, the supply plan for demand 1 is provided by supply institution 1 with materials a and b, and supply institution 2 with materials c. The supply plan for demand 2 is provided by supply institution 2 with materials d. The supply plan for demand 3 is provided by supply institution 2 with materials e, supply institution 1 with materials b, and supply institution 3 with materials f. The supply plan for demand 4 is provided by supply institution f.

[0169] S403, subtract the supply information of each supply institution required by the supply plan to obtain the current supply information of each supply institution;

[0170] After the electronic device rationally allocates the supply of each supply unit according to the materials required by the current subgroup, the supply of each supply unit is reduced accordingly. The electronic device can deduct the supply of each supply unit required by the supply plan to obtain the current supply of each supply unit.

[0171] For example, the current subgroup includes demand 1 and demand 2. Demand 1 requires n1 units of material a, n2 units of material b, and n3 units of material c. Demand 2 requires n4 units of material c and n5 units of material d. Supply agency 1 can provide N1 units of material a and N2 units of material b, supply agency 2 can provide N3 units of material c, and supply agency 3 can provide N4 units of material b and N5 units of material d. The supply plan for demand 1 is that supply agency 1 provides material a and material b, and supply agency 2 provides material c. The supply plan for demand 2 is that supply agency 2 provides material c, and supply agency 3 provides material d.

[0172] After the electronic device rationally allocates the supply of supply institutions 1, 2 and 3 according to demand 1 and demand 2, the electronic device can deduct the supply information of supply institutions 1, 2 and 3 required for the supply plan to obtain the current supply information of each supply institution. That is, supply institution 1 can provide N1-n1 material a and N2-n2 material b, supply institution 2 can provide N3-n3-n4 material c, and supply institution 3 can provide N4 material b and N5-n5 material d.

[0173] For example, the current subgroup includes demand 1, demand 2, demand 3, and demand 4. The materials needed to fulfill demand 1 include n1 material a, n2 material b, and n3 material c; the materials needed to fulfill demand 2 include n4 material d; the materials needed to fulfill demand 3 include n5 material e, n6 material b, and n7 material f; and the materials needed to fulfill demand 4 include n8 material g. Supply organization 1 can provide N1 material a, N2 material b, and N3 material g; supply organization 2 can provide N4 material c, N5 material d, and N6 material e; and supply organization 3 can provide N7 material f. The supply plan for demand 1 is as follows: supply organization 1 provides materials a and b; supply organization 2 provides material c; the supply plan for demand 2 is as follows: supply organization 2 provides material d; the supply plan for demand 3 is as follows: supply organization 2 provides material e; supply organization 1 provides material b; supply organization 3 provides material f; and the supply plan for demand 4 is as follows: supply organization 1 provides material f.

[0174] After the electronic device rationally allocates the supply of supply mechanism 1, supply mechanism 2 and supply mechanism 3 according to demand 1, demand 2, demand 3 and demand 4, the electronic device can deduct the supply information of supply mechanism 1, supply mechanism 2 and supply mechanism 3 required for the supply plan to obtain the current supply information of each supply mechanism. That is, supply mechanism 1 can provide N1-n1 material a, N2-n2-n6 material b, N3-n8 material g, supply mechanism 2 can provide N4-n3 material c, N5-n4 material d, N6-n5 material e, and supply mechanism 3 can provide N7-n7 material f.

[0175] S404, obtain the next demand subgroup in the order described, take it as the current subgroup, and return the steps of taking the material constraints, the capacity constraints, and the date constraints as constraints, taking the weighted demand achievement rate corresponding to the current subgroup as the optimization objective, and performing heuristic solution on the supply plan corresponding to the current subgroup based on the current supply information of each supply institution to obtain the supply plan corresponding to the current subgroup, until multiple demand subgroups have been processed.

[0176] After the electronic device finishes processing the current subgroup, it can continue to process the next subgroup. That is, it obtains the next subgroup in sequence as the current subgroup, and then, for this subgroup, it uses material constraints, capacity constraints, and date constraints as constraints, and the weighted demand achievement rate corresponding to the current subgroup as the optimization objective. Based on the current supply information of each supply institution, it performs a heuristic solution for the supply plan corresponding to the current subgroup, and obtains the steps of the supply plan corresponding to the current subgroup, until all demand subgroups have been processed.

[0177] For example, the current demand group includes three demand subgroups. After the electronic device processes the first demand subgroup, it can obtain the second demand subgroup in the order of demand, and use this demand subgroup as the current subgroup. The current subgroup includes demand 3. For this demand subgroup, considering the dependencies between materials in demand 3, the capacity constraints of each supplier in providing each material, and the time limits for each supplier in providing each material, and using the weighted achievement rate of demand 3 as the optimization objective, a heuristic solution is performed on the supply plan corresponding to demand 3 based on the current supply information of each supplier, to obtain the supply plan corresponding to demand 3, until all three demand subgroups have been processed.

[0178] As can be seen, in this embodiment, the electronic device can process each demand subgroup according to its arrangement order. For the current subgroup, material constraints, capacity constraints, and date constraints are used as constraints, and the weighted demand achievement rate corresponding to the current subgroup is used as the optimization objective. Based on the current supply information of each supply organization, the supply plan corresponding to the current subgroup is heuristically solved to obtain the supply plan corresponding to the current subgroup, until multiple demand subgroups have been processed. In this way, for the planning domain, highly executable supply plans are output for various demands, enabling each supply organization to supply reasonably, achieving supply and demand balance, and thus improving the executability and achievement rate of the planning schedule.

[0179] As one implementation method of this application, such as Figure 5 As shown, the construction methods of the above knowledge reasoning model can include:

[0180] S501, obtain historical planning and scheduling data, historical bottleneck materials, and historical bottleneck suppliers;

[0181] S502, Based on the material information included in the historical planning and scheduling data and the relationship between each supply unit and the material identified by the cycle time, construct material nodes and supply unit nodes;

[0182] To construct a knowledge reasoning model, electronic devices can acquire historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply providers. The historical planning and scheduling data can include demand, supply from each supplier, capacity, cycle time, supply timing information, and material information, etc., without specific limitations here.

[0183] In this way, electronic devices can construct two types of nodes in a heterogeneous directed graph based on the material information included in historical planning and scheduling data and the relationship between each supply unit and material identified by the cycle time. These nodes are material nodes and supply unit nodes.

[0184] S503, using the demand, supply and capacity included in the historical planning and scheduling data to characterize the material nodes, supply organization nodes and the edges between nodes, a heterogeneous directed graph is obtained.

[0185] In the heterogeneous directed graph, the edges between material nodes represent the constraint relationships between the materials corresponding to the material nodes, and the edges between material nodes and supply mechanism nodes represent that the materials corresponding to the material nodes can be produced by the supply mechanism corresponding to the supply mechanism nodes.

[0186] In one implementation, the electronic device uses historical planning and scheduling data, including demand, supply, and capacity, to characterize material nodes, supply mechanism nodes, and the edges between nodes, resulting in a directed heterogeneous graph. This heterogeneous directed graph includes two types of edges: edges between material nodes and edges between material nodes and supply mechanism nodes. Edges between material nodes represent the constraints between the materials corresponding to those nodes, and the direction of the edges can represent the dependencies between materials. For example, in a bill of materials, if a node corresponding to an upper-level material points to a node corresponding to a lower-level material, it indicates that the upper-level material depends on the lower-level material.

[0187] The edge between the material node and the supply organization node indicates that the material corresponding to the material node can be produced by the supply organization corresponding to the supply organization node. The direction of the edge is from the material node to the supply organization node, which is the time transfer relationship, that is, the supply organization corresponding to the supply organization node produces the material corresponding to the material node.

[0188] For example, such as Figure 6As shown, on the graph construction side, the historical planning and scheduling data includes two Bills of Materials (BOMs). The first BOM includes materials A, B, C, D, and E. The production of material A depends on materials B and C, and the production of material B depends on materials D and E. The second BOM includes materials F, B, G, D, and E. The production of material F depends on materials B and G, and the production of material B depends on materials D and E. The cycle time indicator shows that supply mechanism W1 can produce material A, supply mechanism W2 can produce both material A and material F, and supply mechanism W3 can produce material F.

[0189] Electronic devices can construct a directed heterogeneous graph of material nodes and supply facility nodes based on historical planning and scheduling data, including the Bill of Materials (BOM) and the relationships between supply facilities and materials as identified by the cycle time. The material nodes, supply facility nodes, and edges between nodes are represented using historical planning and scheduling data, including demand, supply, and capacity, resulting in a heterogeneous directed graph 601. The material nodes are F, A, G, B, C, D, and E, and the supply facility nodes are W1, W2, and W3. Directed arrows between material nodes represent dependencies between materials, and directed arrows between material nodes and supply facility nodes represent materials produced by the supply facility.

[0190] S504, Based on the historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations, the heterogeneous directed graph is trained to obtain the knowledge reasoning model.

[0191] After constructing the heterogeneous directed graph, electronic devices can train the heterogeneous directed graph based on historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations, thereby obtaining a knowledge reasoning model.

[0192] For example, such as Figure 6 As shown, on the training and knowledge reasoning side, the electronic device can input the demand, supply, capacity, historical bottleneck materials, and historical bottleneck supply institutions included in the historical planning and scheduling data into the heterogeneous directed graph 601 until the loss function converges, thus obtaining the knowledge reasoning model. After inputting the demand to be processed, the supply of each supply institution, and the capacity into this knowledge reasoning model, the electronic device can obtain the potential bottleneck materials and potential bottleneck supply institutions corresponding to the demand to be processed, and can obtain the scores corresponding to the bottleneck materials and bottleneck supply institutions.

[0193] As can be seen, in this embodiment, historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations are acquired. Based on the material information included in the historical planning and scheduling data and the relationship between each supply organization and material identified by the cycle time, material nodes and supply organization nodes are constructed. The material nodes, supply organization nodes, and edges between nodes are represented by the demand, supply, and capacity included in the historical planning and scheduling data to obtain a heterogeneous directed graph. Based on the historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations, the heterogeneous directed graph is trained to obtain a knowledge reasoning model. In this way, the knowledge reasoning model can be used to predict potential bottleneck materials and bottleneck supply organizations, and the scores corresponding to bottleneck materials and bottleneck supply organizations can be obtained. Thus, the information on bottleneck materials and bottleneck supply organizations can be fully utilized to achieve a reasonable allocation of materials and capacity for each supply organization.

[0194] As one implementation method of this application, such as Figure 7 As shown, the steps described above for training the heterogeneous directed graph based on the historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations to obtain the knowledge reasoning model may include:

[0195] S701, the nodes corresponding to historical bottleneck materials and historical bottleneck supply organizations in the heterogeneous directed graph are determined as positive samples, and the nodes other than the positive samples are determined as negative samples.

[0196] After constructing a heterogeneous directed graph, the electronic device can construct positive and negative samples for the material nodes or supply organization nodes of the heterogeneous directed graph based on historical bottleneck materials and historical bottleneck supply organizations. In one embodiment, the electronic device can determine the nodes corresponding to historical bottleneck materials and historical bottleneck supply organizations in the heterogeneous directed graph as positive samples, and determine the nodes other than the positive samples as negative samples.

[0197] For example, the historical bottleneck materials are material B and material A, and the historical bottleneck supply mechanism is supply mechanism W2. Figure 6 As shown, on the training and knowledge reasoning side, the electronic device can identify node B corresponding to material B, node A corresponding to material A, and node W2 corresponding to supply mechanism W2 in the heterogeneous directed graph 601 as positive samples, and identify material nodes F, G, C, D, E, and supply mechanism nodes W1 and W3 as negative samples.

[0198] S702, the historical planning and scheduling data includes demand and supply as the initial vector of material nodes, capacity as the initial vector of supply mechanism nodes, the relationship between material information as the initial vector of the edge between material nodes, and the cycle time as the initial vector of the edge between material nodes and supply mechanism nodes.

[0199] Electronic devices can determine the initial vector of a heterogeneous directed graph based on historical planning and scheduling data, including demand, supply, capacity, cycle time, and material information. Specifically, demand and supply are used as the initial vectors of material nodes, capacity is used as the initial vector of supply mechanism nodes, the relationships between material information are used as the initial vectors of the edges between material nodes, and cycle time is used as the initial vector of the edges between material nodes and supply mechanism nodes.

[0200] S703, Based on the positive samples, the negative samples, and each initial vector, the heterogeneous directed graph is trained to obtain the knowledge reasoning model.

[0201] In one implementation, after the electronic device determines the positive samples, negative samples, and initial vectors, the positive samples, negative samples, and initial vectors can be input into a heterogeneous directed graph. The heterogeneous directed graph is then trained until the loss function converges to obtain a knowledge reasoning model. This model can then be used to predict bottleneck materials and bottleneck supply mechanisms, and obtain the corresponding scores for bottleneck materials and bottleneck supply mechanisms.

[0202] As can be seen, in this embodiment, the electronic device can identify the nodes corresponding to historical bottleneck materials and historical bottleneck supply organizations in the heterogeneous directed graph as positive samples, and identify the nodes other than the positive samples as negative samples. The demand and supply included in the historical planning and scheduling data are used as the initial vectors of the material nodes, the production capacity is used as the initial vector of the supply organization nodes, the relationship between material information is used as the initial vector of the edge between material nodes, and the cycle time is used as the initial vector of the edge between material nodes and supply organization nodes. Based on the positive samples, negative samples, and each initial vector, the heterogeneous directed graph is trained to obtain a knowledge reasoning model, which can then be used to predict potential bottleneck materials and bottleneck supply organizations.

[0203] As one implementation of this application, the above-mentioned demand to be processed includes predicted demand;

[0204] like Figure 8 As shown, prior to the step of obtaining the demand to be processed, supply information, and material information, the above method may further include:

[0205] S801, acquire historical demand data and switching data for materials;

[0206] The switching data refers to material data that changes according to business needs.

[0207] The pending demand can include forecasted demand, which is the prediction of potential future material needs. Forecasted demand can be considered a type of demand. When a supply plan is obtained based on the forecasted demand, staff can replenish materials according to the plan, ensuring timely fulfillment of the forecasted demand when it materializes.

[0208] Electronic devices can acquire historical demand data and switching data for materials. Switching data refers to material data that changes based on business needs. Historical demand data can be data stored in the business system. For example, for product 'a', the market includes the original model A and the new model B. Electronic devices can acquire historical demand data for materials required by the original model A and material data for the new model B, i.e., switching data.

[0209] In one implementation, after the electronic device acquires historical demand data and switching data of materials, the historical demand data and switching data can be cleaned and organized, such as processing outliers.

[0210] S802, construct the demand sequence of the material based on the time sequence corresponding to the historical demand data and the switching data;

[0211] After acquiring historical demand data and switching data, electronic devices can construct material demand sequences based on the time sequence corresponding to the historical demand data and switching data. For example, based on historical demand data, the demand data for materials a, b, c, d, and e are obtained; based on historical switching data, the switching time t1 for materials a and c, and the switching time t2 for materials d and e are obtained. Then, the material demand sequences are determined as demand sequence 1, demand sequence 2, and demand sequence 3. Demand sequence 1 includes material c, and the demand quantity is the sum of the demand quantities of materials a and c at the same time point. Demand sequence 2 includes material b, and demand sequence 3 includes material e, and the demand quantity is the sum of the demand quantities of materials d and e at the same time point.

[0212] S803, extract features from the demand sequence to obtain the sequence features of the demand sequence, and classify the materials based on the sequence features to obtain multiple categories of materials;

[0213] In one implementation, after constructing the demand sequence of materials, the electronic device can extract features from the demand sequence to obtain the sequence features of the demand sequence, and classify the materials using a clustering algorithm based on the sequence features to obtain multiple categories of materials.

[0214] For example, following the example in step S802, after the electronic device constructs demand sequence 1, demand sequence 2 and demand sequence 3, it can extract features from demand sequence 1, demand sequence 2 and demand sequence 3 respectively to obtain the sequence features of each demand sequence. Based on the sequence features, a clustering algorithm is used to classify materials b, c and e, thereby obtaining two categories of materials. Category 1 includes materials b and e, and category 2 includes materials c.

[0215] S804: For each category of materials, a prediction method adapted to the sequence characteristics of the materials in that category is used to predict the demand for the materials in that category, thus obtaining the predicted demand for the materials in that category.

[0216] For different categories of materials, electronic devices can employ different prediction algorithms. In one implementation, for each category of materials, the electronic device can use a prediction method adapted to the sequence characteristics of that category of materials to predict the demand for that category of materials, thereby obtaining the predicted demand for that category of materials. The prediction method can be machine learning algorithms, deep learning algorithms, etc., and is not specifically limited here.

[0217] For example, for material b in category 1, electronic devices can use a deep learning algorithm adapted to the sequence characteristics of materials in that category to predict the demand for material b. If the predicted demand for material b increases, appropriate replenishment can be made to meet the demand for material b in a timely manner when it arises.

[0218] In one implementation, the electronic device can pre-build a demand forecasting model, and when historical demand data and switching data of materials are input into the demand forecasting model, it can output the predicted demand for each category of materials. For example... Figure 9 As shown, the demand forecasting model comprises three parts: feature engineering, material clustering, and machine learning / deep learning algorithm training and inference. Feature engineering extracts features from the demand sequence, material clustering classifies materials, and machine learning / deep learning algorithm training and inference uses a forecasting method adapted to the sequence features of each material category to predict the demand for each category. When electronic devices input historical demand data and material switching data into this demand forecasting model, it can output the predicted demand for each category of materials.

[0219] As can be seen, in this embodiment, the electronic device can acquire historical demand data and switching data of materials. Based on the time sequence corresponding to the historical demand data and switching data, a demand sequence of materials is constructed. Features are extracted from the demand sequence to obtain the sequence features of the demand sequence. Based on the sequence features, materials are classified to obtain multiple categories of materials. For each category of materials, a prediction method adapted to the sequence features of that category of materials is used to predict the demand of that category of materials, thus obtaining the predicted demand of that category of materials. In this way, when rationally allocating materials and capacity of various supply institutions, predicted demand can be considered, and staff can be guided to replenish materials related to the supply plan according to the predicted demand. When the predicted demand occurs, the materials required for the predicted demand can be provided in a timely manner, thereby improving the feasibility and achievement rate of the planning schedule.

[0220] As one implementation of the present application, the supply of each of the above-mentioned supply institutions includes the modified procurement supply;

[0221] like Figure 10 As shown, prior to the step of obtaining the demand to be processed, supply information, and material information, the above method may further include:

[0222] S1001, Obtain historical supply data from supply institutions;

[0223] The historical supply data includes the supply date and quantity of each material supplied by each supplying organization.

[0224] Since the supply capacity of the selected suppliers is not fixed, it is possible to assess the supply capacity of the selected suppliers. In one implementation, the electronic device can assess the accuracy of the supply date and quantity of the procurement-related supplies provided by each supplier based on the supplier's historical supply data, that is, to correct the supply date and quantity when the supply date and quantity of materials change.

[0225] Electronic devices can acquire historical supply data from supply organizations. This historical supply data includes the supply dates and quantities of each material from each supply organization. For example, if the selected supply organizations include supply organization 1, supply organization 2, and supply organization 3, the electronic device can acquire the historical supply data of supply organizations 1, 2, and 3. The historical supply data includes the supply dates and quantities of materials a and b from supply organization 1, the supply dates and quantities of materials b and c from supply organization 2, and the supply dates and quantities of material d from supply organization 3.

[0226] In one implementation, after the electronic device acquires the historical supply data of the supply organization, the historical supply data can be cleaned and organized, such as processing outliers.

[0227] S1002, extract features from the historical supply data to obtain the features of the historical supply data, and classify each material based on the features to obtain multiple categories of materials;

[0228] After the electronic device acquires historical supply data, it can extract features from the historical supply data to obtain the characteristics of the historical supply data. Based on the extracted features, a clustering method is used to classify each material, thereby obtaining multiple categories of materials. For example, following the example in step S1001, the electronic device can extract features from the historical supply data to obtain the characteristics of the historical supply data. Based on these characteristics, materials a, b, c, and d are classified into three categories: Category 1 includes materials a and b provided by supply organization 1 and material b provided by supply organization 2; Category 2 includes material c provided by supply organization 2; and Category 3 includes material d provided by supply organization 3.

[0229] S1003: For each category of materials, a forecasting method adapted to the characteristics of that category of materials is used to forecast the supply date and supply quantity of that category of materials, thus obtaining the corrected procurement supply for that category of materials.

[0230] For different categories of materials, electronic devices can employ different prediction algorithms. In one implementation, for each category of materials, the electronic device can use a prediction method adapted to the characteristics of that category to predict the supply date and quantity of the materials, thereby obtaining a revised procurement supply for that category of materials. The prediction method can be machine learning algorithms, deep learning algorithms, etc., and is not specifically limited here.

[0231] For example, for material d provided by supply organization 3 in category 3, the electronic device can use a deep learning algorithm adapted to the characteristics of the category to predict the supply date and supply quantity of material d. If the supply date and supply quantity of material d provided by supply organization 3 are different from those in the historical supply data of supply organization 3, the supply date and supply quantity of material d by supply organization 3 shall be corrected.

[0232] In one implementation, the electronic device can pre-build a supply organization capacity assessment model. When historical supply data from the supply organization is input into this model, it can output revised procurement supply estimates for each category of materials. For example... Figure 11As shown, the supply organization capacity assessment model comprises three parts: feature engineering, material clustering, and machine learning / deep learning algorithm training and inference. Feature engineering extracts features from historical supply data, material clustering categorizes materials, and machine learning / deep learning algorithm training and inference uses prediction methods adapted to the characteristics of each material category to predict the supply date and quantity for each category. When the electronic device inputs historical supply data from the supply organization into the supply organization capacity assessment model, it can output the supply date and quantity of each category of materials provided by the supply organization.

[0233] As can be seen, in this embodiment, the electronic device can acquire historical supply data from the supply organization, extract features from the historical supply data to obtain its characteristics, and classify the materials based on these characteristics to obtain multiple categories of materials. For each category of materials, a prediction method adapted to the characteristics of that category is used to predict the supply date and quantity of the materials in that category, resulting in a revised procurement supply for that category of materials. This modification of the supply date and quantity for each category of materials improves the feasibility and achievement rate of the planning and scheduling.

[0234] As one implementation of this application, the above supply plan may include an existing supply allocation strategy and / or a new supply plan generated by capacity allocation;

[0235] The existing supply allocation strategy includes whether to use the existing supply, the allocation date, the supply selection under the allocation date, and the supply quantity selection under the allocation date; the new supply plan generated by the capacity allocation includes the type of planned supply, the date of planned supply, the supply institution of planned supply, and the quantity of planned supply.

[0236] The supply plan for the demand to be processed by electronic devices may include the allocation strategy of existing supply and / or the new supply plan generated by capacity allocation. That is, the supply plan may include how to allocate the existing supply of each supply unit and / or how to allocate the capacity of each supply unit to generate new planned supply.

[0237] In one implementation, the allocation strategy for existing supply may include whether to use existing supply, the allocation date, the supply selection on the allocation date, and the supply quantity selection on the allocation date. The new supply plan resulting from capacity allocation may include the type of planned supply, the date of planned supply, the supply provider for planned supply, and the quantity of planned supply. The type of planned supply may include procurement or inventory.

[0238] For example, the supply plan for electronic equipment demand 1 includes the allocation strategy of existing supply and the new supply plan generated by capacity allocation. The allocation strategy of existing supply includes using existing supply, an allocation date of 2023-10-09, selecting supplier 1 for material a on the allocation date, and selecting quantity n1 for the supply quantity on the allocation date. The new supply plan generated by capacity allocation may include the planned supply type as procurement, the planned supply date as 2023-10-10, the planned supply supplier as supplier 2, and the planned supply quantity as n2.

[0239] As can be seen, in this embodiment, the existing supply allocation strategy may include whether to use the existing supply, the allocation date, the supply selection under the allocation date, and the supply quantity selection under the allocation date. The new supply plan generated by capacity allocation may include the type of planned supply, the date of planned supply, the supply organization of planned supply, and the quantity of planned supply. In this way, after obtaining the supply plan for the demand to be processed, the staff can use the supply plan as auxiliary information for various decisions, effectively assisting the staff in making decisions, reducing workload, and improving the quality of demand fulfillment and delivery.

[0240] Figure 12 This is a schematic diagram illustrating a method for determining a schedule provided in an embodiment of this application. The following is in conjunction with... Figure 12 The method for determining the schedule provided in the embodiments of this application will be illustrated with examples. For instance... Figure 12 As shown, the method for determining the schedule provided in this application is designed with the interaction framework between the agent and the environment as the core. This framework includes the following parts: mechanism model, prediction unit, knowledge reasoning and schedule.

[0241] The environment comprises a mechanistic model and a prediction unit. The mechanistic model identifies material constraints, capacity constraints, and date constraints. The prediction unit performs demand forecasting and supply chain assessment. The results from the mechanistic model and prediction unit are input into the environment and transmitted to the intelligent agent for processing, thus realizing the relevant state input. The intelligent agent includes knowledge reasoning and planning / scheduling. Knowledge reasoning identifies bottleneck materials and bottleneck supply chains, while planning / scheduling assigns demand grouping weights, allocates supply, and allocates capacity. After the intelligent agent processes the relevant states, a decision-making action is obtained, thus realizing the decision-making action output.

[0242] The mechanistic model can organize various basic data according to business logic to provide a unified data foundation for other components. This data foundation is an infrastructure platform with data storage, management, computation, and distribution capabilities. In this way, electronic devices can obtain pending demand, supply information, and material information from the mechanistic model.

[0243] like Figure 13As shown, the mechanism model can extract data such as supply structure, demand, supply, capacity, takt time, BOM, master data, and work calendar from the business system. The demand side includes various types of demand in the business system (such as shipping demand, reserved demand, sales demand, and allocation demand), as well as potential forecast demand output by the forecasting unit through demand forecasting. The supply side includes various types of supply in the business system (such as inventory and production orders), as well as modified procurement-related supply (such as purchase orders) output by the forecasting unit through supply structure capacity assessment. Master data includes demand master data and material master data, which can represent the basic attributes of demand and materials. The BOM is used to form material constraints, takt time and capacity are used to guide the formation of capacity constraints, and the work calendar is used to guide the formation of date constraints. Supply structures can share a global work calendar, or different supply structures can have separate work calendars. The BOM, takt time, and capacity can also serve as inputs to the knowledge reasoning model, which can output bottleneck materials and bottleneck supply structures.

[0244] Planning and scheduling can target pending demands, using material constraints, capacity constraints, and date constraints as constraints, and demand fulfillment rate as the optimization objective. Based on the supply information of each supply institution, the supply plan corresponding to the pending demands can be heuristically solved to obtain the supply plan corresponding to the pending demands, thus realizing planned supply.

[0245] like Figure 14 As shown, in the process of implementing the planning and scheduling, the demands to be processed can include multiple groups of demands with priorities, from the highest priority demand group to the lowest priority demand group. For the highest priority demand group, this demand group includes demand 1, demand 2, demand 3, and demand 4. Knowledge reasoning outputs the bottleneck supply organization and its score, and the bottleneck material and its score. The electronic device can group and sort the demands according to the bottleneck supply organization and its score, dividing demand 1 and demand 3 into one demand subgroup, demand 2 into another demand subgroup, and demand 4 into yet another demand subgroup. The electronic device can assign weights to the demands based on the bottleneck material and its score.

[0246] For the first demand subgroup, namely the demand subgroup consisting of demand 1 and demand 3, the electronic equipment can use material constraints, capacity constraints and date constraints as constraints, and the weighted demand achievement rate as the optimization objective. Based on the supply information of each supply institution, the supply plan corresponding to demand 1 and demand 3 included in this demand subgroup can be heuristically solved to obtain the supply plan corresponding to demand 1 and demand 3.

[0247] Supply planning can include supply plans based on existing supply allocation and supply plans generated from capacity allocation. Supply plans based on supply allocation can include 1) date decisions, 2) supply decisions, and 3) quantity decisions. For demand 1, this can generate existing supply 1, existing supply 2, and planned supply 1. Existing supply 2 is based on the results of supply facility capacity assessments. Supply plans based on capacity allocation include 1) date decisions, 2) supply decisions, and 3) quantity decisions. For demand 1, this can generate planned supply 2 and planned supply 3. Planned supply 1, planned supply 2, and planned supply 3 can achieve planned output. Similarly, for other demands, corresponding supply plans can be obtained.

[0248] As can be seen, in the solution provided in this application embodiment, when heuristically solving the supply plan corresponding to the demand to be processed, not only are the dependencies between materials, the capacity constraints of each supplying organization to provide each material, and the time constraints of each supplying organization to provide each material considered, but also the potential bottleneck supplying organizations and bottleneck materials are predicted by the pre-trained knowledge reasoning model. Based on the scores corresponding to the bottleneck supplying organizations, the demand to be processed is grouped and sorted, and each demand subgroup is obtained according to the sorting order. The weight of each demand is determined based on the bottleneck material, and the weighted demand achievement rate is used as the optimization objective. Therefore, a reasonable supply plan can be obtained for both demand restricted by bottleneck materials and bottleneck supplying organizations and demand not restricted by bottleneck materials and bottleneck supplying organizations, so that each supplying organization can supply reasonably, achieve supply and demand balance, and thus achieve a plan scheduling with high executability and high achievement rate.

[0249] Corresponding to the above-described method for determining a schedule, this application also provides a device for determining a schedule. The following describes the device for determining a schedule provided by this application.

[0250] like Figure 15 As shown, a device for determining a planned schedule includes:

[0251] The information acquisition module 1510 is used to acquire demand to be processed, supply information and material information. The supply information includes at least the supply, capacity, cycle time and supply time information of each supply organization. The cycle time is used to represent the working hours required for the supply organization to provide each material, and / or to represent the quantity of each material provided by the supply organization per unit time.

[0252] The constraint determination module 1520 is used to determine material constraints based on the demand transmission relationship between the materials identified by the material information, to determine capacity constraints based on the capacity and the cycle time, and to determine date constraints based on the supply time relationship between the materials identified by the supply time information. The material constraints are used to represent the dependency relationship between the materials, the capacity constraints are used to represent the ability constraints of each supply organization to provide each material, and the date constraints are used to represent the time limits of each supply organization to provide each material.

[0253] The bottleneck prediction module 1530 is used to input the pending demand, the supply and capacity of each supply unit into a pre-trained knowledge reasoning model to obtain the bottleneck material and bottleneck supply unit corresponding to the pending demand. The knowledge reasoning model constructs material nodes and supply unit nodes based on the material information included in the historical planning and scheduling data and the relationship between each supply unit and the material identified by the cycle time. The demand, supply and capacity included in the historical planning and scheduling data are used to characterize the material nodes, supply unit nodes and the edges between the nodes, and to predict the bottleneck material and bottleneck supply unit corresponding to the input demand. The bottleneck supply unit is used to characterize the bottleneck in the capacity of the supply unit for the material.

[0254] The grouping and sorting module 1540 is used to group and sort the demand to be processed based on the score of the supply organization corresponding to the bottleneck supply organization that the demand to be processed corresponds to, and obtain multiple subgroups of demand after sorting.

[0255] The weight calculation module 1550 is used to obtain each demand subgroup in sorting order, and for each demand included in the current subgroup, determine the weight corresponding to the demand based on the number of times the material corresponding to the demand hits the bottleneck material and the score corresponding to the bottleneck material.

[0256] The supply plan acquisition module 1560 is used to heuristically solve the supply plan corresponding to each demand in the current subgroup based on the supply information of each supply institution, using the material constraints, the capacity constraints, and the date constraints as constraints and the demand achievement rate with the weights as the optimization objective, so as to obtain the supply plan corresponding to the demand to be processed.

[0257] As can be seen, in the solution provided in this application embodiment, when heuristically solving the supply plan corresponding to the demand to be processed, not only are the dependencies between materials, the capacity constraints of each supplying organization to provide each material, and the time constraints of each supplying organization to provide each material considered, but also the potential bottleneck supplying organizations and bottleneck materials are predicted by the pre-trained knowledge reasoning model. Based on the scores corresponding to the bottleneck supplying organizations, the demand to be processed is grouped and sorted, and each demand subgroup is obtained according to the sorting order. The weight of each demand is determined based on the bottleneck material, and the weighted demand achievement rate is used as the optimization objective. Therefore, a reasonable supply plan can be obtained for both demand restricted by bottleneck materials and bottleneck supplying organizations and demand not restricted by bottleneck materials and bottleneck supplying organizations, so that each supplying organization can supply reasonably, achieve supply and demand balance, and thus achieve a plan scheduling with high executability and high achievement rate.

[0258] As one embodiment of this application, the above-described apparatus may further include a knowledge reasoning model construction module, which is used to construct a knowledge reasoning model and may include:

[0259] The first acquisition submodule is used to acquire historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations.

[0260] The node construction submodule is used to construct material nodes and supply organization nodes based on the material information included in the historical planning and scheduling data and the relationship between each supply organization and material identified by the cycle time.

[0261] The heterogeneous directed graph acquisition submodule is used to characterize the material nodes, supply mechanism nodes and edges between nodes using the demand, supply and capacity included in the historical planning and scheduling data to obtain a heterogeneous directed graph. In the heterogeneous directed graph, the edges between material nodes represent the constraint relationship between the materials corresponding to the material nodes, and the edges between material nodes and supply mechanism nodes represent that the materials corresponding to the material nodes can be produced by the supply mechanism corresponding to the supply mechanism nodes.

[0262] The second acquisition submodule is used to train the heterogeneous directed graph based on the historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations to obtain the knowledge reasoning model.

[0263] As one embodiment of this application, the second acquisition submodule described above may include:

[0264] The positive and negative sample determination unit is used to determine the nodes corresponding to historical bottleneck materials and the nodes corresponding to historical bottleneck supply institutions in the heterogeneous directed graph as positive samples, and to determine the nodes other than the positive samples as negative samples.

[0265] The initial vector determination unit is used to take the demand and supply included in the historical planning and scheduling data as the initial vector of the material node, the capacity as the initial vector of the supply mechanism node, the relationship between material information as the initial vector of the edge between material nodes, and the cycle time as the initial vector of the edge between the material node and the supply mechanism node.

[0266] The knowledge reasoning model acquisition unit is used to train the heterogeneous directed graph based on the positive samples, the negative samples, and each initial vector to obtain the knowledge reasoning model.

[0267] As one embodiment of this application, the grouping and sorting module 1540 may include:

[0268] The supply organization determination submodule is used to determine the supply organization corresponding to each demand in each demand group based on the material information corresponding to the demand and the cycle time. The demand group is obtained by grouping the demands to be processed according to priority.

[0269] The demand subgroup acquisition submodule is used to divide the same demand from the supply organization and the bottleneck supply organization into a group of demand to obtain a demand subgroup.

[0270] The demand subgroup acquisition submodule is used to sort the demand subgroups according to the scores of the corresponding bottleneck supply institutions to obtain multiple sorted demand subgroups.

[0271] As one embodiment of this application, the supply plan acquisition module 1560 described above may include:

[0272] The Current Group Requirements Acquisition Submodule is used to retrieve the highest priority group of requirements as the current group requirements.

[0273] The supply plan acquisition submodule is used to perform heuristic solution on the supply plan corresponding to the demand subgroups included in the current demand group, with the material constraints, the capacity constraints and the date constraints as constraints, and the demand achievement rate with the weights corresponding to the demand subgroups included in the current demand group as the optimization objective, based on the current supply information of each supply institution, to obtain the supply plan corresponding to the current demand group.

[0274] The current supply information acquisition submodule is used to deduct the supply information of each supply institution required by the supply plan to obtain the current supply information of each supply institution;

[0275] The multi-group demand processing completion submodule is used to obtain the next group of demands in descending order of priority as the current group of demands, and return the demand subgroups included in the current group of demands. With the material constraints, capacity constraints, and date constraints as constraints, and the demand achievement rate with the weights corresponding to the demand subgroups included in the current group of demands as the optimization objective, the module performs heuristic solution on the supply plan corresponding to the demand subgroups included in the current group of demands based on the current supply information of each supply institution, and obtains the supply plan corresponding to the current group of demands. This process continues until all multiple groups of demands have been processed.

[0276] As one embodiment of this application, the above-mentioned supply planning acquisition submodule may include:

[0277] The current subgroup determination unit is used to obtain the first requirement subgroup according to the sorting order for the current group requirements, and use it as the current subgroup;

[0278] The supply plan acquisition unit is used to perform heuristic solution on the supply plan corresponding to the current subgroup, based on the current supply information of each supply institution, with the material constraints, the capacity constraints and the date constraints as constraints, the weighted demand achievement rate corresponding to the current subgroup as the optimization objective, and the current supply information of each supply institution.

[0279] The current supply information acquisition unit is used to subtract the supply information of each supply institution required by the supply plan to obtain the current supply information of each supply institution;

[0280] The demand subgroup processing completion unit is used to obtain the next demand subgroup in the order described above, as the current subgroup, and return the steps of using the material constraints, capacity constraints, and date constraints as constraints, the weighted demand achievement rate corresponding to the current subgroup as the optimization objective, and performing heuristic solution on the supply plan corresponding to the current subgroup based on the current supply information of each supply institution to obtain the supply plan corresponding to the current subgroup, until multiple demand subgroups have been processed.

[0281] As one implementation of this application, the above-mentioned demand to be processed includes predicted demand;

[0282] The above-mentioned device may further include:

[0283] The data acquisition module is used to acquire historical demand data and switching data of materials before the step of acquiring the demand to be processed, supply information and material information, wherein the switching data is material data that changes according to business needs;

[0284] The demand sequence construction module is used to construct the demand sequence of the material based on the time sequence corresponding to the historical demand data and the switching data;

[0285] The first acquisition module is used to extract features from the demand sequence to obtain the sequence features of the demand sequence, and classify the materials based on the sequence features to obtain multiple categories of materials;

[0286] The demand forecasting module is used to predict the demand for each category of materials by using a forecasting method that is adapted to the sequence characteristics of the materials in that category, and thus obtain the predicted demand for the materials in that category.

[0287] As one implementation of this application, the supply of each of the above-mentioned supply organizations includes the modified procurement supply;

[0288] The above-mentioned device may further include:

[0289] The historical supply data acquisition module is used to acquire historical supply data of the supply organizations before the steps of acquiring the demand to be processed, supply information and material information. The historical supply data includes the supply date and supply quantity of each material by each supply organization.

[0290] The second acquisition module is used to extract features from the historical supply data to obtain the features of the historical supply data, and classify each material based on the features to obtain multiple categories of materials.

[0291] The revised procurement supply acquisition module is used to predict the supply date and quantity of materials for each category using a forecasting method adapted to the characteristics of that category, thus obtaining the revised procurement supply for that category of materials.

[0292] As one embodiment of this application, the supply plan includes an existing supply allocation strategy and / or a new supply plan generated from capacity allocation;

[0293] The existing supply allocation strategy includes whether to use the existing supply, the allocation date, the supply selection on the allocation date, and the supply quantity selection on the allocation date.

[0294] The new supply plan resulting from the capacity allocation includes the type of planned supply, the date of planned supply, the supplying agency of planned supply, and the quantity of planned supply.

[0295] This application also provides an electronic device, such as... Figure 16 As shown, it includes:

[0296] Memory 1601 is used to store computer programs;

[0297] When the processor 1602 executes the program stored in the memory 1601, it implements the steps of the method for determining the schedule as described in any of the above embodiments.

[0298] Furthermore, the aforementioned electronic device may also include a communication bus and / or a communication interface, with the processor 1602, the communication interface, and the memory 1601 communicating with each other via the communication bus.

[0299] As can be seen, in the solution provided in this application embodiment, when heuristically solving the supply plan corresponding to the demand to be processed, not only are the dependencies between materials, the capacity constraints of each supplying organization to provide each material, and the time constraints of each supplying organization to provide each material considered, but also the potential bottleneck supplying organizations and bottleneck materials are predicted by the pre-trained knowledge reasoning model. Based on the scores corresponding to the bottleneck supplying organizations, the demand to be processed is grouped and sorted, and each demand subgroup is obtained according to the sorting order. The weight of each demand is determined based on the bottleneck material, and the weighted demand achievement rate is used as the optimization objective. Therefore, a reasonable supply plan can be obtained for both demand restricted by bottleneck materials and bottleneck supplying organizations and demand not restricted by bottleneck materials and bottleneck supplying organizations, so that each supplying organization can supply reasonably, achieve supply and demand balance, and thus achieve a plan scheduling with high executability and high achievement rate.

[0300] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0301] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0302] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0303] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0304] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of any of the above-described methods for determining a schedule.

[0305] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the scheduling determination methods described in the above embodiments.

[0306] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a solid-state drive (SSD), etc.

[0307] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0308] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, electronic devices, computer-readable storage media, and computer program products are basically similar to the method embodiments, and therefore the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0309] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.

Claims

1. A method for determining a project schedule, characterized in that, The method includes: Obtain pending demand, supply information, and material information, wherein the supply information includes at least the supply, capacity, cycle time, and supply time information of each supply organization, the cycle time is used to represent the working hours required for the supply organization to provide each material, and / or, is used to represent the quantity of each material provided by the supply organization per unit time. Based on the demand transmission relationship between the materials identified by the material information, material constraints are determined; based on the capacity and the cycle time, capacity constraints are determined; and based on the supply time relationship between the materials identified by the supply time information, date constraints are determined. Here, the material constraints are used to represent the dependency relationship between the materials, the capacity constraints are used to represent the ability constraints of each supply organization to provide each material, and the date constraints are used to represent the time limits of each supply organization to provide each material. The pending demand, the supply and capacity of each supplying organization are input into a pre-trained knowledge reasoning model to obtain the bottleneck material and bottleneck supplying organization corresponding to the pending demand. The knowledge reasoning model constructs material nodes and supplying organization nodes based on the material information included in historical planning and scheduling data and the relationship between each supplying organization and the material identified by the cycle time. The material nodes, supplying organization nodes and the edges between the nodes are represented by the demand, supply and capacity included in the historical planning and scheduling data, which is used to predict the bottleneck material and bottleneck supplying organization corresponding to the input demand. Based on the score of the supply organization corresponding to the pending demand hitting the bottleneck supply organization, the pending demand is grouped and sorted to obtain multiple sorted demand subgroups. The bottleneck supply organization is used to characterize the bottleneck in the supply organization's material production capacity. Each demand subgroup is obtained in sorting order. For each demand included in the current subgroup, the weight corresponding to the demand is determined based on the number of times the material corresponding to the demand hits the bottleneck material and the score corresponding to the bottleneck material. Using the material constraints, capacity constraints, and date constraints as constraints, and the demand achievement rate with the weights as the optimization objective, the supply plan corresponding to each demand in the current subgroup is heuristically solved based on the supply information of each supply institution, to obtain the supply plan corresponding to the demand to be processed.

2. The method according to claim 1, characterized in that, The construction method of the knowledge reasoning model includes: Obtain historical planning and scheduling data, historical bottleneck materials, and historical bottleneck suppliers; Based on the material information included in the historical planning and scheduling data, and the relationship between each supply unit and material identified by the cycle time, material nodes and supply unit nodes are constructed. The material nodes, supply mechanism nodes, and edges between nodes are characterized using the historical planning and scheduling data, including demand, supply, and capacity, to obtain a heterogeneous directed graph. In the heterogeneous directed graph, the edges between material nodes represent the constraint relationships between the materials corresponding to the material nodes, and the edges between material nodes and supply mechanism nodes represent that the materials corresponding to the material nodes can be produced by the supply mechanism corresponding to the supply mechanism nodes. Based on the historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations, the heterogeneous directed graph is trained to obtain the knowledge reasoning model.

3. The method according to claim 2, characterized in that, The step of training the heterogeneous directed graph based on the historical planning and scheduling data, historical bottleneck materials, and historical bottleneck supply organizations to obtain the knowledge reasoning model includes: The nodes corresponding to historical bottleneck materials and historical bottleneck supply organizations in the heterogeneous directed graph are identified as positive samples, and the nodes other than the positive samples are identified as negative samples. The historical planning and scheduling data includes demand and supply as the initial vector of material nodes, capacity as the initial vector of supply mechanism nodes, the relationship between material information as the initial vector of the edge between material nodes, and cycle time as the initial vector of the edge between material nodes and supply mechanism nodes. The heterogeneous directed graph is trained based on the positive samples, the negative samples, and the initial vectors to obtain the knowledge reasoning model.

4. The method according to claim 1, characterized in that, The step of grouping and sorting the pending demands based on the score of the supply organization corresponding to the bottleneck supply organization, to obtain multiple sorted demand subgroups, includes: For each demand included in each demand group, the supply organization corresponding to the demand is determined based on the material information corresponding to the demand and the cycle time. The demand group is obtained by grouping the demands to be processed according to priority. The demand that is the same as that of the supply institution and the bottleneck supply institution is divided into a group of demand to obtain a demand subgroup; The demand subgroups are sorted according to the scores of the corresponding bottleneck supply institutions to obtain multiple sorted demand subgroups.

5. The method according to claim 4, characterized in that, The step of using the material constraints, capacity constraints, and date constraints as constraints, and the demand fulfillment rate with the weights as the optimization objective, and based on the supply information of each supply organization, heuristically solving the supply plan corresponding to each demand in the current subgroup to obtain the supply plan corresponding to the demand to be processed includes: Get the highest priority set of requirements as the current group of requirements; For the demand subgroups included in the current demand group, with the material constraints, the capacity constraints, and the date constraints as constraints, and with the demand achievement rate with the weights corresponding to the demand subgroups included in the current demand group as the optimization objective, the supply plan corresponding to the demand subgroups included in the current demand group is heuristically solved based on the current supply information of each supply institution to obtain the supply plan corresponding to the current demand group. Subtracting the supply information of each supply institution required for the supply plan, the current supply information of each supply institution is obtained; The next set of demands is retrieved in descending order of priority as the current set of demands. The subgroups of demands included in the current set of demands are returned. The material constraints, capacity constraints, and date constraints are used as constraints. The target is to optimize the achievement rate of the demand subgroups of the current set of demands with the weights. Based on the current supply information of each supply organization, the supply plan corresponding to the subgroups of demands included in the current set of demands is heuristically solved to obtain the supply plan corresponding to the current set of demands. This process continues until all sets of demands have been processed.

6. The method according to claim 5, characterized in that, The step of taking the material constraints, capacity constraints, and date constraints as constraints for the demand subgroups included in the current demand group, and using the weighted demand achievement rate corresponding to the demand subgroups included in the current demand group as the optimization objective, and heuristically solving the supply plan corresponding to the demand subgroups included in the current demand group based on the current supply information of each supply institution, to obtain the supply plan corresponding to the current demand group, includes: For the current group's requirements, the first requirement subgroup is obtained according to the sorting order and is taken as the current subgroup; For the current subgroup, with the material constraints, capacity constraints, and date constraints as constraints, and with the weighted demand achievement rate corresponding to the current subgroup as the optimization objective, the supply plan corresponding to the current subgroup is heuristically solved based on the current supply information of each supply institution to obtain the supply plan corresponding to the current subgroup. By deducting the supply information of each supply institution required for the supply plan, the current supply information of each supply institution is obtained; The process involves obtaining the next demand subgroup in the specified order, using it as the current subgroup, and returning the steps described above. For the current subgroup, the process is as follows: using the material constraints, capacity constraints, and date constraints as constraints, and the weighted demand fulfillment rate corresponding to the current subgroup as the optimization objective. Based on the current supply information of each supply institution, the process is heuristically solved to obtain the supply plan corresponding to the current subgroup. This process continues until all demand subgroups have been processed.

7. The method according to any one of claims 1-6, characterized in that, The unprocessed demand includes predicted demand; Prior to the step of obtaining the demand to be processed, supply information, and material information, the method further includes: Obtain historical demand data and switching data for materials, wherein the switching data refers to material data that changes according to business needs; Based on the time sequence corresponding to the historical demand data and the switching data, a demand sequence for the material is constructed; Feature extraction is performed on the demand sequence to obtain the sequence features of the demand sequence, and the materials are classified based on the sequence features to obtain multiple categories of materials; For each category of materials, a prediction method adapted to the sequence characteristics of the materials in that category is used to predict the demand for the materials in that category, thus obtaining the predicted demand for the materials in that category.

8. The method according to any one of claims 1-6, characterized in that, The supply from each supplying entity includes the revised procurement supply; Prior to the step of obtaining the demand to be processed, supply information, and material information, the method further includes: Obtain historical supply data from supply organizations, wherein the historical supply data includes the supply date and quantity of each material supplied by each supply organization; Feature extraction is performed on the historical supply data to obtain the features of the historical supply data, and the materials are classified based on the features to obtain multiple categories of materials; For each category of materials, a forecasting method adapted to the characteristics of that category of materials is used to predict the supply date and supply quantity of that category of materials, thus obtaining the corrected procurement supply for that category of materials.

9. The method according to any one of claims 1-6, characterized in that, The supply plan includes the allocation strategy for existing supply and / or the new supply plan resulting from capacity allocation; The existing supply allocation strategy includes whether to use the existing supply, the allocation date, the supply selection on the allocation date, and the supply quantity selection on the allocation date. The new supply plan resulting from the capacity allocation includes the type of planned supply, the date of planned supply, the supplying agency of planned supply, and the quantity of planned supply.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-9.

Images