Manufacturing decision-making system-oriented functional architecture and collaboration mechanism thereof

Abstract

The invention provides a manufacturing decision system-oriented modular functional architecture and a collaborative mechanism thereof, and the structural organization and reusable collaboration of a manufacturing decision process are realized by constructing a multi-layer capability system, a functional module system and a standardized behavior interface. The architecture comprises a data interface module, a decision-making target module, a disturbance response module and a fusion decision-making module, and defines a data flow, a control flow and a trigger relationship among the modules. The system completes target expression, constraint analysis, disturbance identification and collaborative decision generation through standardized access of multi-source data such as orders, productivity, materials and processes, and forms an interpretable and executable decision link. According to the invention, modularization, interpretability and intelligent cooperation of the manufacturing decision-making system are realized, and multi-scene and multi-industry manufacturing decision-making application is supported.

Description

Technical Field

[0001] This invention relates to the fields of intelligent manufacturing and manufacturing decision-making technology, specifically to a functional architecture and its collaborative mechanism for manufacturing decision-making systems, belonging to the interdisciplinary technical field of manufacturing decision-making systems engineering, system architecture design, and decision-making methods. This invention focuses on the functional module division of manufacturing decision-making systems at the engineering level, the interaction relationships between modules, and their standardized collaborative methods. Background Technology

[0002] As the scale and complexity of the manufacturing industry continue to increase, enterprises face significant uncertainties in order delivery, capacity utilization, material availability, and equipment stability. Manufacturing decision-making processes require continuous high-quality decisions amidst multi-source heterogeneous data, complex process constraints, dynamic disturbances, and multi-objective trade-offs. However, existing manufacturing decision-making systems are generally built around a single algorithm or functional module, lacking a unified functional structure, leading to the following problems:

[0003] 1. Blurred functional boundaries: Functions such as data processing, target representation, decision generation, and anomaly response are often coupled in the same module, making them difficult to extend, reuse, or maintain;

[0004] 2. Insufficient module collaboration: There is a lack of standardized data and control flows between different functions, and the system behavior depends on the internal logic of the modules, resulting in unpredictable and unexplainable decision-making processes;

[0005] 3. Weak disturbance response capability: Disturbances such as order insertion, material shortage, and equipment failure occur frequently, but traditional systems are unable to quickly trigger local or global decision updates and have insufficient dynamic scheduling capabilities.

[0006] 4. Fragmentation of intelligent capabilities: Human experience, operations research (OR), and artificial intelligence (AI) often exist independently, lacking a unified integration mechanism, making it difficult to form collaborative decision-making capabilities;

[0007] 5. Lack of systematic architecture: Existing technologies mostly focus on algorithm implementation or scenario application, lacking a systematic definition and structured organization of manufacturing decision-making systems at the functional level.

[0008] The inventors previously proposed a capability architecture for a manufacturing decision-making system in the parent application (application number 2026100451795), defining the system's capability components and decision-making links at the abstract level and constructing a multi-layered capability foundation. In subsequent applications (application numbers 2025113006004, 2025116134466, 2025119141404, and 2026100085765), the inventors provided implementation schemes for functional modules such as data interfaces, target construction, disturbance response, and fusion decision-making. However, neither the parent nor the subsequent applications systematically defined the collaborative mechanism between the functional modules of the manufacturing decision-making system nor the standardized action system.

[0009] Therefore, it is necessary to propose a functional architecture and its collaborative mechanism for manufacturing decision-making systems, in order to build a unified structure and behavioral interface for the application layer, so that the various functional modules of the capability base can work collaboratively in the scenario, thereby forming a complete patent system for manufacturing decision-making systems. Summary of the Invention

[0010] This invention pertains to the functional architecture layer of a manufacturing decision-making system, and belongs to a different technical level from the capability architecture proposed by the inventor in patent application number 2026100451795. The aforementioned capability architecture defines the capability components and decision-making chain of a manufacturing decision-making system from an abstract level, while this invention structurally divides the system's functional modules from an engineering implementation perspective, including a data interface module, a decision target module, a disturbance response module, and a fusion decision module, and further defines the data flow, control flow, and collaborative mechanisms between the modules.

[0011] The inventors previously described specific implementation methods for the aforementioned functional modules in sub-applications with application numbers 2025113006004, 2025116134466, 2025119141404, and 2026100085765, which belong to the implementation-level technical solutions. For example... Figure 1 As shown, this invention, as a functional architecture layer, forms a three-layer complementary structure with the capability architecture layer (parent design) and the functional module implementation layer (existing sub-designs). Independent of capability definition and specific implementation, this invention focuses on the structured organization of functional modules and their collaborative mechanisms, thereby constituting a complete technical system for manufacturing decision-making.

[0012] I. Technological Positioning and Innovation Points

[0013] In existing manufacturing decision-making systems, each functional module (such as data interface, target construction, disturbance response, decision generation, etc.) usually operates independently, lacking a unified structured organization and standardized data flow, control flow, and behavioral logic. This makes it difficult for the system to form a consistent decision-making link in real manufacturing scenarios and to achieve cross-module collaborative operation.

[0014] This invention proposes a functional architecture and its collaborative mechanism for manufacturing decision-making systems. By defining functional modules in a structured manner and introducing standard actions as a unified behavioral interface, it achieves unified encapsulation, action-based invocation, and scenario-based orchestration of module capabilities, enabling the system to form continuous, interpretable, and scalable decision-making behaviors in real manufacturing scenarios.

[0015] This invention is not a simple superposition of the contents of the sub-cases, but an integration and collaborative design of various functional modules, which is the key exit for the manufacturing decision-making system from capability realization to practical application.

[0016] II. Functional Module Structure

[0017] To support the decision-making needs of typical manufacturing scenarios, this invention provides a structured definition of the basic functions of a manufacturing decision-making system, including:

[0018] 1. Data Interface Function: Used for standardized modeling, access and synchronization of multi-source data such as orders, processes, capacity, materials, and equipment status, providing a consistent, complete and traceable data foundation for the decision-making process.

[0019] 2. Target building function: Used to express the multi-target requirements of the manufacturing system, including delivery time, cost, equipment utilization, work-in-process control, energy consumption, etc., and supports target priority setting and dynamic adjustment. It is the intent expression layer of the system.

[0020] 3. Disturbance response function: It is used to identify and handle disturbance events such as order insertion, material shortage, equipment failure, and process change in the manufacturing process, trigger decision updates and maintain the executability of the plan, which is the key to the system to achieve dynamic scheduling.

[0021] 4. Integrated Decision-Making Function: This function is used to build a collaborative decision-making link between human experience, operations research optimization models, and artificial intelligence models, enabling solution generation, constraint verification, result verification, and interpretable output. It is the core decision execution layer of the system.

[0022] III. Standard Movement System

[0023] This invention constructs a set of 12 standard actions for manufacturing decision-making systems, serving as a unified behavioral language for the system. This system abstracts the complex manufacturing decision-making process into composable, reusable, and interpretable action units, enabling the system to construct decision-making links for various manufacturing scenarios through action sequences, achieving modular collaboration and scenario-based execution.

[0024] The standard actions are divided into four categories: input, target, decision, and response, which correspond to the data interface module, target construction module, fusion decision module, and disturbance response module, respectively.

[0025] 1. Input actions

[0026] Input actions are responsible for transforming external environmental information, production status information, and constraints into structured inputs that the system can understand, and are the starting point of the decision-making chain.

[0027] 1) Data Acquisition: Obtaining real-time or batch data from devices, MES, ERP, sensors, or external systems to form the basic dataset required for decision-making;

[0028] 2) Constraint Parsing: Performs structured parsing of process constraints, resource constraints, order constraints, strategy constraints, etc., to generate constraint models that can be called by the decision-making module;

[0029] Input actions ensure that the system has complete, accurate, and structured input information before making decisions, providing a data foundation for subsequent goal construction and solution generation.

[0030] 2. Target-based actions

[0031] Goal-oriented actions are used to build a decision-making goal system, transforming business requirements into quantifiable and optimizable goal models.

[0032] 1) Goal Formulation: Based on information such as order requirements, production strategies, resource status, and empirical rules, construct a multi-objective model including delivery time, cost, load balancing, and energy consumption.

[0033] By using goal-oriented actions, business requirements are formally expressed as optimization goals, enabling the system to execute subsequent decision-making actions within a unified goal framework.

[0034] 3. Decision-making actions

[0035] Decision-making actions constitute the core decision-making chain of the system, enabling executable manufacturing decisions through solution generation, evaluation, selection, and interpretation.

[0036] 1) Plan Generation: Based on the input data, constraint model, and target model, the fusion decision module is invoked to generate one or more feasible plans.

[0037] 2) Plan Evaluation: Candidate plans are evaluated from multiple dimensions, including goal achievement, resource utilization, and risk indicators.

[0038] 3) Plan Selection: Based on the evaluation results, select the optimal or second-best plan to form the final decision output;

[0039] 4) Decision Explanation: Explaining the basis, key constraints, and objective trade-offs for the selection of solutions to improve the understandability and auditability of the system;

[0040] 5) Execution Dispatch: The final decision plan is distributed to the execution system (such as MES, scheduling system, equipment control system) to trigger actual production behavior;

[0041] The system achieves a complete decision-making loop from solution generation to implementation through decision-making actions, and enhances system transparency through an explanation mechanism.

[0042] 4. Responsive actions

[0043] Response actions are used to handle disturbances during the manufacturing process, ensuring that the system has the ability to respond in real time and adjust dynamically.

[0044] 1) Exception Detection: Identifies disruptive events such as equipment malfunctions, material shortages, and order changes;

[0045] 2) Impact Analysis: Analyzes the scope and severity of the impact of the anomaly on the current plan, resource status, and goal achievement.

[0046] 3) Local Adjustment: Quickly adjust local resources, tasks, or time windows without refactoring the overall plan;

[0047] 4) Collaborative Reconfiguration: When local adjustments fail to meet the objectives, cross-module collaboration is triggered to regenerate or reconfigure the overall solution.

[0048] By using responsive actions, the system gains adaptive capabilities, enabling it to maintain the feasibility of solutions and the stability of objectives in dynamic environments.

[0049] Overall effect of the system

[0050] The standard action system serves as a unified behavioral language for manufacturing decision-making systems, enabling the system to construct decision-making chains across various manufacturing scenarios through action sequences, including but not limited to:

[0051] 1) Initial production scheduling;

[0052] 2) Order processing;

[0053] 3) Equipment fault response;

[0054] 4) Material shortage adjustment;

[0055] 5) Real-time rolling scheduling;

[0056] 6) Multi-factory collaborative optimization;

[0057] By combining and choreographing standard actions, the system achieves the technical effects of reusable modules, scalable scenarios, interpretable decisions, and traceable execution.

[0058] IV. Module Collaboration Mechanism Based on Standard Actions

[0059] This invention enables collaborative operation between functional modules through standardized actions. These standardized actions serve as application interfaces, exposing the capabilities of functional modules in a unified manner, allowing the system to invoke capabilities using actions as the smallest unit of behavior.

[0060] The system operation process is as follows:

[0061] 1. Data Acquisition: Invoke the data acquisition action to obtain basic data, status data, and relevant business information from the data interface module to form decision input;

[0062] 2. Constraint Resolution: Obtain and resolve process constraints, resource constraints, and strategy constraints from the data interface module to form a structured constraint model;

[0063] 3. Target Construction: Invoke the target construction action to obtain a multi-target model from the target construction module, including optimization targets such as delivery time, cost, load balancing, and energy consumption, as well as custom optimization targets;

[0064] 4. Solution Generation → Solution Evaluation → Solution Selection: The decision generation, evaluation, and selection actions are called sequentially to form an initial executable solution;

[0065] 5. Anomaly Identification → Impact Analysis → Local Adjustment / Collaborative Reconstruction: When a disturbance event occurs, the disturbance identification and analysis action is invoked, and local optimization or global collaborative reconstruction action is selected according to the scope of impact to achieve dynamic optimization of the solution;

[0066] 6. Execution and Issuance → Decision Interpretation: The execution and issuance action is invoked to issue the final solution to the execution system, and the decision interpretation action generates interpretable information to achieve transparency and closed-loop management of the decision-making process.

[0067] V. Three-layer structural system

[0068] This invention constructs a three-layer manufacturing decision-making system consisting of a "capability layer, module layer, and action layer":

[0069] 1. Capability Layer: Provides the foundational capabilities for the manufacturing decision-making system and is the unified source of capabilities for the system;

[0070] 2. Module Layer: The basic capabilities of the capability layer are encapsulated in an engineered manner to form functional modules such as data interface, target management, disturbance response, and fusion decision-making, providing callable module capabilities for the action layer;

[0071] 3. Action Layer: Abstracting module capabilities into standard actions to form a unified behavior interface, and constructing the decision-making chain of the manufacturing scenario through action sequences.

[0072] This three-layer structure enables hierarchical organization of capabilities, unified encapsulation of module capabilities, and standardized expression of action interfaces. This allows the system to reuse capabilities and modules in the form of action sequences in different manufacturing scenarios, forming an interpretable, scalable, and reusable scenario-based decision-making system.

[0073] VI. Multi-layered Capability Structure

[0074] This invention further constructs a multi-layered capability structure for the manufacturing decision-making system, used for hierarchical organization and structured expression of system capabilities. This capability structure includes the following levels from bottom to top:

[0075] 1. Algorithm layer, acceleration layer, and model / operator layer

[0076] These three layers form the computational and modeling foundation of the system. The algorithm layer provides optimization and inference algorithms, the acceleration layer provides parallel computing and operator acceleration capabilities, and the model / operator layer encapsulates reusable model components and operator resources.

[0077] 2. Data layer, target layer, response layer, and collaboration layer

[0078] These four layers constitute the decision input and decision objective system. The data layer describes the manufacturing system status, the objective layer expresses the decision objectives, the response layer describes the disturbance identification, impact analysis capabilities, and disturbance response process, and the collaboration layer is used for cross-module collaboration and decision objective integration.

[0079] 3. Template layer

[0080] It is used to provide industry knowledge, rule templates and experience models to enable knowledge reuse across scenarios.

[0081] 4. Smart Layer

[0082] This layer provides intelligent capabilities for manufacturing decision-making, including capacity forecasting, demand forecasting, risk forecasting, and intervention strategy generation. By calling upon the data, model, and algorithm capabilities of the lower layers, the intelligent layer performs core decision-making actions such as solution generation, solution evaluation, and solution selection. It is a key path for the system to integrate artificial intelligence technology and achieve a leap from traditional to intelligent decision-making.

[0083] The aforementioned multi-layered capabilities form a complete capability system through input-output relationships and dependencies, providing a unified and reusable source of capabilities for standard actions. Figure 2 is a schematic diagram of the multi-layered capability structure proposed in this invention, used to illustrate the organization of capability layers and their collaborative relationships.

[0084] VII. Mapping Relationship between Ability Layer and Action Layer

[0085] This invention presents the mapping relationship between multi-layer capabilities and standard actions, as shown in Table 1, and further shows the inverse dependency relationship between standard actions and multi-layer capabilities, as shown in Table 2. This mapping relationship clarifies how the capability layer supports the action layer and how the action layer invokes the capability layer, forming the basis of the decision-making system of this invention.

[0086]

[0087] Table 1. Mapping Relationship between Ability Levels and Standard Actions

[0088]

[0089] Table 2. Back-mapping of standard actions to ability levels

[0090] VIII. Scene choreography mechanism for standard actions

[0091] Building upon the standard action layer, this invention proposes a scenario orchestration mechanism based on action sequences. The system selects a set of actions according to scenario requirements and constructs action sequences based on dependencies and triggering conditions, thereby forming executable decision-making links for scenarios such as production scheduling, order insertion, material shortages, and fault response.

[0092] Specifically, the system selects several actions from a set of standard actions based on the business needs of the manufacturing scenario, and forms an action sequence according to preset action dependencies and triggering conditions. During the execution of the action sequence, the capability model of the capability layer is called in sequence to realize behaviors such as data collection, constraint resolution, target construction, solution generation, solution evaluation, solution selection, anomaly identification, impact analysis, local adjustment, collaborative reconstruction and execution, thereby forming a decision-making process for scenarios such as production scheduling, order insertion, fault response and material shortage handling.

[0093] Through the above-mentioned action arrangement mechanism, this invention realizes the structured expression of manufacturing scenarios, which separates the scenario logic from the module and forms a reusable and scalable scenario construction method.

[0094] IX. Technical Effects

[0095] This invention constructs a manufacturing decision-making system consisting of a capability layer, a module layer, and an action layer, and systematically designs functional modules, standard actions, and capability structures. This significantly improves the manufacturing decision-making system in terms of structure, interpretability, dynamism, and intelligence. The specific technical effects are as follows:

[0096] 1. Clear structure and strong scalability: This invention modularizes the manufacturing decision-making system and clarifies module boundaries through standardized data flow and control flow, enabling the system to have good scalability, facilitating function enhancement, module replacement and cross-industry adaptation.

[0097] 2. The decision-making process is interpretable and highly consistent: This invention abstracts complex decision-making behaviors into standard actions, enabling the decision-making process to be presented in the form of an action chain. It also ensures the consistency and traceability of decisions in different scenarios through a unified goal expression, constraint resolution, and decision verification mechanism.

[0098] 3. Achieving integrated decision-making based on human experience, operations research optimization, and artificial intelligence: This invention constructs an integrated decision-making link of human × OR × AI through an integrated decision-making mechanism, enabling the system to simultaneously utilize empirical rules, optimization models, and intelligent predictions to achieve higher quality and more robust decision results.

[0099] 4. Forming a unified decision-making closed loop and improving execution consistency: This invention forms a complete closed loop through actions such as data collection, scheme generation, scheme evaluation, execution and decision interpretation, enabling the system to have continuous iteration and self-optimization capabilities, thereby improving the consistency of decision execution.

[0100] 5. Standardized and highly reusable scenario construction: This invention constructs multiple scenarios such as production scheduling, order insertion, and fault response through standard action sequences, which decouples scenario logic from the module, making it highly reusable and significantly reducing scenario development costs. Attached Figure Description

[0101] Figure 1 is a schematic diagram of the patent system architecture provided by the present invention, which is used to illustrate the overall relationship between the capability layer, functional architecture, behavior interface layer and scene orchestration mechanism proposed by the present invention.

[0102] Figure 2 is a schematic diagram of the multi-layer capability structure provided by the present invention, which is used to illustrate that the capability system of the manufacturing decision system is composed of multiple capability layers, and to show the organization of each capability layer and its hierarchical relationship.

[0103] Figure 3 is a schematic diagram of the functional architecture of the manufacturing decision system provided by the present invention, which is used to show the structural relationship between the data interface module, the decision target module, the disturbance response module and the fusion decision module, as well as their data flow and control flow interaction methods.

[0104] Figure 4 is a schematic diagram of the action chain for the initial production scheduling provided by the present invention, which is used to illustrate that in the initial production scheduling scenario, the system forms an executable decision chain by arranging the sequence of standard actions. Detailed Implementation

[0105] The technical solution of the present invention will be described in detail below with reference to specific embodiments, so that those skilled in the art can more clearly understand the concept, technical features and beneficial effects of the present invention. It should be understood that these embodiments are only used to illustrate the basic principles of the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0106] The technical solutions of this invention can be modified and extended in various ways according to actual application needs. As long as they do not depart from the core idea of ​​this invention and the scope defined by the claims, they should be regarded as the objects of protection of this invention.

[0107] Without causing technical conflicts, the technical features or sub-modules in the following embodiments can be combined arbitrarily to adapt to different application scenarios and system architectures.

[0108] The illustrations are for illustrative purposes only. The processes, module divisions, and interaction relationships shown can be adjusted and optimized according to actual application scenarios, and should not be construed as the sole limitation of the collaborative decision-making optimization system and method of this invention. Those skilled in the art can adapt and extend the method steps, system structure, or information interaction methods according to specific needs without departing from the core ideas of this invention.

[0109] To enhance the clarity of the technical expression, terms such as "first" and "second" used in the specification are only used to distinguish similar components or steps and do not indicate their execution order or priority; terms such as "comprising" and "having" are open-ended expressions intended to cover other technical features not explicitly listed; terms "and / or" and " / " indicate a parallel or selective relationship, and their specific meaning should be understood in conjunction with the context.

[0110] Example 1: Overall Description of Functional Architecture

[0111] This embodiment provides a functional architecture for manufacturing decision-making systems, such as... Figure 3 As shown, the system is divided into four basic functional modules: a data interface module, a decision target module, a disturbance response module, and a fusion decision module. These modules interact through standardized data and control flows, forming a clearly structured, scalable, and interpretable functional system, providing module-level capabilities for subsequent action-oriented invocation and scenario-based orchestration.

[0112] 1. Data Interface Module

[0113] It is used to access multi-source data such as orders, processes, capacity, materials, and equipment status, and to complete data cleaning, verification, and standardized modeling to form a consistent, complete, and traceable data foundation, providing reliable input for target construction and solution generation.

[0114] 2. Decision-making objective module

[0115] It is used to express the multi-objective requirements of the manufacturing system, including delivery time, cost, equipment utilization, work-in-process control, etc., and supports the setting of objective priorities, handling of objective conflicts and dynamic adjustment, forming an optimizable objective model.

[0116] 3. Disturbance Response Module

[0117] It is used to identify, classify and handle disturbance events in the manufacturing process, such as order insertion, material shortage, equipment failure, process change, etc., and trigger local adjustments or global reconstruction according to the disturbance range to realize the dynamic response capability of the system.

[0118] 4. Integrated Decision Module

[0119] It is used to build a collaborative decision-making link between human experience, operations research and optimization models and artificial intelligence models, and realizes solution generation, solution evaluation, solution selection and interpretable output. It is the core module of the system to form executable decisions.

[0120] Through the above module division, the present invention constructs the basic functional structure of the manufacturing decision-making system, enabling the system to have a modular, scalable and interpretable engineering architecture.

[0121] Example 2: Module Collaboration Mechanism

[0122] like Figure 3 As shown, this embodiment provides data flow, control flow, and standard action mechanisms between functional modules to enable collaborative operation between modules and form a reusable and composable decision-making link.

[0123] I. Data flow includes the following directions:

[0124] 1. Data Interface Module → Decision Objective Module: Provides basic data such as orders, capacity, materials, and equipment status required for objective expression, used to build multi-objective models.

[0125] 2. Data Interface Module → Fusion Decision Module: Provides standardized data input required for decision reasoning.

[0126] 3. Disturbance Response Module → Fusion Decision Module: Provides disturbance event information and triggers local or global decision updates.

[0127] 4. Fusion Decision Module → Data Interface Module: Requests data refresh or verifies data consistency, used to verify the feasibility of the solution and the validity of the data.

[0128] II. Control flow is used to define the triggering relationships between modules, including:

[0129] 1. Target change triggers decision update: When the target priority, target weight or business strategy changes, the decision target module sends a control signal to the fusion decision module to trigger the recalculation of the solution.

[0130] 2. Disturbance-triggered decision update: When disturbance events such as order insertion, material shortage, or equipment failure are detected, the disturbance response module triggers local or global reordering to achieve dynamic response.

[0131] 3. Decision verification triggers data refresh: When verifying the feasibility of a solution, the fusion decision module can request the data interface module to refresh the data or verify consistency.

[0132] 4. Decision release triggers closed-loop feedback: After a decision is released, a feedback loop is triggered to update the system status, record the execution results, and support subsequent optimization.

[0133] Third, this invention abstracts module capabilities into standard actions, including: data acquisition actions, constraint parsing actions, target construction actions, solution generation actions, solution evaluation actions, solution selection actions, anomaly identification actions, impact analysis actions, local adjustment actions, collaborative reconstruction actions, execution and distribution actions, and decision interpretation actions. These standard actions serve as a unified behavioral interface, enabling unified encapsulation, action-based invocation, and scenario-based orchestration of module capabilities.

[0134] Example 3: Typical Scenario Example - Initial Production Scheduling

[0135] This embodiment is used to illustrate the application of the functional architecture of the present invention in a typical manufacturing scenario, and does not involve specific algorithms or implementation details.

[0136] During the initial production scheduling, the system needs to generate an executable scheduling plan based on the current production status, process constraints, and business objectives. The system drives the collaborative operation of functional modules through standard actions, and the process is as follows:

[0137] 1. Data Acquisition: The system obtains real-time or batch data such as equipment status, work order information, inventory quantity, and process parameters from the data interface module to form the structured input required for production scheduling.

[0138] 2. Constraint Resolution: The system calls the constraint modeling module to resolve constraints such as process routes, capacity limits, shift rules, and resource conflicts, generating a constraint model that can be called by the decision-making module.

[0139] 3. Goal Construction: The system obtains the production scheduling goal model from the goal management module, including business goals such as delivery date priority, cost priority, load balancing, and minimizing the number of switching operations, and forms a quantifiable expression of optimization goals.

[0140] 4. Solution Generation: The system calls the decision generation module to generate one or more production scheduling candidate solutions based on the input data, constraint model and target model.

[0141] 5. Solution Evaluation: The system calls the evaluation module to evaluate candidate solutions from multiple dimensions, including goal achievement, resource utilization, and constraint satisfaction.

[0142] 6. Solution selection: The system selects the optimal or second-best solution based on the evaluation results to form the final production scheduling result.

[0143] 7. Execution and Distribution: The system distributes the final production schedule to the execution system, such as MES, scheduling system or equipment control system, to trigger actual production behavior.

[0144] As shown in Figure 4, the initial production scheduling action chain is as follows: Data acquisition → Constraint resolution → Target construction → Solution generation → Solution evaluation → Solution selection → Execution and issuance. This chain demonstrates how the action mechanism of this invention can structure the capability layer, functional architecture, and behavioral units to achieve interpretability and reusability of the production scheduling process.

[0145] To further illustrate the coverage of action calls in the initial production scheduling scenario, this embodiment uses the standard action-scenario mapping relationship of the present invention, as shown in Table 3:

[0146]

[0147] Table 3. Examples of Standard Actions and Scene Coverage Relationships

[0148] This embodiment demonstrates how the standard action system of the present invention drives the coordinated operation of modules in an initial production scheduling scenario. Through the structured combination of input, target, decision, and response actions, the system can:

[0149] 1. Construct an interpretable production scheduling decision-making process;

[0150] 2. Supports modular reuse;

[0151] 3. Adaptable to different manufacturing scenarios;

[0152] 4. Maintain transparency and consistency in the decision-making process;

[0153] This enables a reusable, scalable, and interpretable manufacturing decision-making method and system.

Claims

1. A manufacturing decision-making method, characterized in that, include: 1) Construct a multi-layered capability system for manufacturing decision-making; 2) Construct the functional architecture of the manufacturing decision-making system based on the aforementioned capability system; 3) Abstract the functional capabilities in the aforementioned functional architecture into a set of standardized behavioral units; 4) Construct a decision-making chain for the manufacturing scenario based on the aforementioned behavioral units; 5) The decision-making process for the manufacturing scenario is realized by executing the aforementioned decision-making chain; The multi-layered capability system, functional architecture, standardized behavioral units, and decision-making links are defined independently and form a reusable and scalable manufacturing decision-making system through preset dependencies and invocation relationships.

2. The method according to claim 1, wherein, The multi-layered capability system includes a data capability layer, a target capability layer, a constraint capability layer, a collaborative capability layer, a response capability layer, an intelligent capability layer, a model / operator capability layer, an algorithm capability layer, and a template / acceleration capability layer.

3. The method according to claim 1, wherein, The functional architecture includes functional units for data processing, target management, decision generation, or disturbance response.

4. The method according to claim 1, wherein, The standardized behavior unit is used to uniformly encapsulate the capabilities of the functional unit.

5. The method according to claim 1, wherein, The decision-making link consists of multiple behavioral units arranged according to a preset dependency relationship.

6. The method according to claim 1, wherein, The decision-making chain is used to construct scenarios for production scheduling, dispatching, order insertion processing, fault response, or material shortage handling.

7. A manufacturing decision-making system, characterized in that, include: 1) A multi-layered capability system to provide the basic capabilities required for manufacturing decisions; 2) Functional architecture, used to engineer the organization of the aforementioned capabilities; 3) Behavior interface layer, used to abstract the capabilities of the functional architecture into standardized behavioral units; The system performs scenario orchestration based on the behavior units to construct the decision-making chain of the manufacturing scenario. This decouples the capability layer from the scenario application through the behavior interface layer, enabling the manufacturing decision-making process to be combined and reused in the form of behavior units.

8. The system according to claim 7, wherein, The functional architecture includes a data interface module, a decision target module, a disturbance response module, and a fusion decision module.

9. The system according to claim 7, wherein, The behavior interface layer is used to expose the capabilities of the functional architecture in a unified interface manner.

10. The system according to claim 7, wherein, The scene orchestration is used to select behavioral units and form action sequences according to scene requirements.

11. The system according to claim 7, wherein, The system adopts a hierarchical structure of capability layer, function layer, and behavior layer.

12. The system according to claim 7, wherein, The behavioral units are used to achieve reusability, interpretability, and scalability of the manufacturing decision-making process.

Images