Enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping

Abstract

The invention provides an enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping, and belongs to the technical field of intelligent manufacturing. Comprising the following steps: receiving a resource configuration request input by a user, and performing natural language analysis on the resource configuration request to generate structured resource configuration demand information; according to the structured resource configuration demand information, matching and calling a preset resource configuration template; based on the called resource configuration template, in combination with real-time resource state data in the ERP system, performing analog computation of resource configuration, and generating at least one resource configuration scheme; and feeding back the at least one resource configuration scheme to the user, and receiving feedback information of the user on the resource configuration scheme. User requirements are analyzed through intelligent interaction, an optimization strategy is matched through template mapping, real-time data simulation calculation is combined, feedback is received, and intelligentization, high efficiency and continuous optimization of enterprise manufacturing resource configuration are achieved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent manufacturing, and particularly relates to an enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping. BACKGROUND

[0002] In a modern enterprise manufacturing environment, the efficiency and accuracy of resource configuration are directly related to production cost, delivery cycle and overall operational efficiency.

[0003] Traditional resource configuration mainly relies on manual experience or systems based on fixed rules, which not only consumes time and effort, but also is difficult to cope with complex and variable production demands and dynamic changes in resource status (such as equipment failure, material shortage, personnel change, etc.). With the expansion of enterprise scale and the increase of product customization, the complexity of resource configuration decision-making has increased dramatically, and traditional methods have shown obvious limitations in response speed, optimization accuracy and knowledge reuse. SUMMARY

[0004] In order to solve at least one aspect of the technical problems in the background art, the application provides an enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping, which realizes the intelligentization, high efficiency and continuous optimization of enterprise manufacturing resource configuration by analyzing user demand through intelligent interaction, matching optimization strategy through template mapping, combining real-time data simulation calculation and receiving feedback.

[0005] The technical solution adopted by the application is: The first aspect of the application provides an enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping, comprising: receiving a resource configuration request input by a user, and performing natural language analysis on the resource configuration request to generate structured resource configuration demand information; matching and calling a preset resource configuration template according to the structured resource configuration demand information; based on the called resource configuration template, combining real-time resource state data in an ERP system to perform simulation calculation of resource configuration, and generating at least one resource configuration scheme; feeding back the at least one resource configuration scheme to the user, and receiving feedback information of the user on the resource configuration scheme.

[0006] According to one embodiment of the application, the receiving of the resource configuration request input by the user and the natural language analysis of the resource configuration request to generate the structured resource configuration demand information are specifically: receiving a resource configuration request input by a user in natural language form; The pre-trained natural language processing model is used to perform word segmentation, part-of-speech tagging, and named entity recognition on the resource configuration request, and to extract a production order number, a product model, a material code, a device type, a station number, a required quantity, and a planned completion time; An operation intention in the resource configuration request is identified based on a semantic understanding algorithm, and the operation intention includes adding resource allocation, changing an existing resource plan, or querying a resource occupation state; The extracted production order number, product model, material code, device type, station number, required quantity, planned completion time, and identified operation intention are structurally combined to generate structured resource configuration requirement information including “production order-product model-material-device-station-quantity-time-intention” fields.

[0007] According to an embodiment of the present application, the structured resource configuration requirement information is matched with and calls a preset resource configuration template, specifically: The structured resource configuration requirement information is converted into a first feature vector; Second feature vectors of each resource configuration template are obtained from a preset resource configuration template library, and the second feature vectors include production types, product categories, process routes, device combinations, and resource constraint conditions applicable to the templates; The semantic similarity between the first feature vector and each second feature vector is calculated; The resource configuration template with the highest semantic similarity is called as a target template; If the highest semantic similarity is lower than a preset threshold, a template mismatch prompt is generated, and a general resource configuration strategy based on a rule engine is started as a backup solution.

[0008] According to an embodiment of the present application, based on the called resource configuration template, real-time resource state data in an ERP system is combined to perform simulation calculation of resource configuration, and at least one resource configuration scheme is generated, specifically: Resource state data related to the target template is obtained from the ERP system in real time, and the resource state data includes a current running state of a device, personnel scheduling information, a material inventory level, a work-in-process flow progress, and a station occupation state; The resource state data is injected into the called target resource configuration template to construct a resource configuration optimization model; Based on the resource configuration optimization model, a linear programming or mixed integer programming algorithm is used to minimize at least one of a production cycle, a device utilization rate, or a resource scheduling cost as an optimization target to perform multi-scenario simulation calculation; A plurality of candidate resource configuration schemes satisfying the constraint conditions are generated, and each scheme is comprehensively scored, and at least one resource configuration scheme with the highest score is output.

[0009] According to an embodiment of the present application, the at least one resource configuration scheme is fed back to the user, and feedback information of the user on the resource configuration scheme is received, specifically: The at least one resource configuration scheme is presented to the user in the form of a Gantt chart, a resource load chart, or a three-dimensional visual layout chart; A scheme comparison function is provided in the interactive interface, allowing the user to compare the production cycle, resource cost, equipment load balancing degree, and material set rate indicators of different resource configuration schemes horizontally; Receiving the user's confirmation, rejection or modification operation on any resource configuration scheme through clicking, dragging or natural language instruction input, the modification operation includes adjusting the resource allocation object, changing the job time window or replacing the standby material; The user's confirmation, rejection or modification operation is recorded as user feedback information, and the optimization preference features thereof are extracted for subsequent template self-learning update.

[0010] According to an embodiment of the present application, the method further comprises: If the user performs a modification operation on the generated resource configuration scheme, the implicit rules embodied by the modification operation are analyzed, and a template parameter adjustment suggestion is generated; Compare the template parameter adjustment suggestion with the historical optimization record to identify repetitive or trend optimization requirements; When the trigger frequency of the same type of parameter adjustment suggestion exceeds a preset threshold, the parameter configuration or constraint condition of the corresponding template in the resource configuration template library is automatically updated; The updated resource configuration template is marked as "optimized" state, and its calling priority is improved in the next matching process.

[0011] According to an embodiment of the present application, the method further comprises: The real-time resource state data and the called resource configuration template version when the rejected scheme is generated are called to build a scheme generation context; Based on the preset rule engine and machine learning model, the potential reasons for the rejection of the scheme are analyzed, including material inventory data delay, device failure unsynchronization, process constraint condition absence or template applicable scene misjudgment; Generate a diagnostic report containing problem type, impact range and correction suggestion, and push it to the system administrator and related business personnel; According to the correction suggestion in the diagnostic report, an ERP system data verification task or a resource configuration template retraining process is automatically triggered.

[0012] A second embodiment of the present application provides an enterprise manufacturing resource configuration optimization device based on intelligent interaction and template mapping, comprising: The parsing module is adapted to receive a resource configuration request input by a user, and perform natural language parsing on the resource configuration request to generate structured resource configuration requirement information; A calling module, adapted to match and call a preset resource configuration template according to the structured resource configuration requirement information; A calculation module adapted to perform a simulation calculation of resource allocation based on the called resource allocation template and in combination with real-time resource status data in the ERP system, and generate at least one resource allocation plan; The information feedback module is adapted to feed back the at least one resource configuration scheme to the user and receive feedback information of the user on the resource configuration scheme.

[0013] An embodiment of the third aspect of the present application provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping as described in any embodiment of the first aspect described above is implemented.

[0014] The present application also provides a non-volatile computer storage medium having computer executable instructions stored thereon. When the computer executable instructions are executed by a processor, the enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping as described in any embodiment of the first aspect as described above can be implemented.

[0015] Due to the adoption of the above technical solution, the beneficial effects achieved by this application are as follows: According to the method for optimizing enterprise manufacturing resource configuration based on intelligent interaction and template mapping provided in the first aspect of the present application, firstly, the resource configuration request input by the user in the form of natural language is received, and the natural language processing technology is used to deeply analyze it, so as to accurately extract key information such as production order, product model, material, equipment, work station, quantity, time and operation intention, and generate structured resource configuration demand information, which greatly reduces the use threshold of the user and improves the friendliness and efficiency of human-computer interaction; then, the system intelligently matches the structured demand information with the preset resource configuration template library, automatically calls the most suitable resource configuration template through feature vector similarity calculation, and realizes the rapid reuse and scenario application of the optimization strategy; on this basis, the system deeply integrates the real-time resource state data (such as equipment running state, material inventory, personnel scheduling, etc.) in the ERP system, constructs an optimization model based on the called template, performs multi-scenario simulation calculation by using an advanced mathematical programming algorithm, and generates multiple high-quality candidate resource configuration schemes, thereby ensuring the scientificity and optimality of the resource configuration decision; finally, the system feeds back the generated scheme to the user in the form of Gantt chart, resource load chart and other visual forms, supports the user to perform interactive confirmation, rejection or modification, and collects the user preference through the closed-loop feedback mechanism, thereby providing data support for the continuous self-learning and optimization of the template. Through the closed-loop process of “intelligent interaction-template mapping-data driven-feedback learning”, the method effectively solves the problems of traditional resource configuration methods, such as dependence on manual experience, slow response, poor flexibility, difficulty in knowledge precipitation, and the like, and significantly improves the intelligent level, decision efficiency and accuracy of resource configuration. BRIEF DESCRIPTION OF DRAWINGS

[0016] The accompanying drawings, which are included to provide a further understanding of the present application, form a part of the present application and illustrate the illustrative embodiments of the present application and effects achieved by the same, and do not limit the present application. In the drawings: Figure 1 A flowchart of the method for optimizing enterprise manufacturing resource configuration based on intelligent interaction and template mapping provided in the embodiments of the present application; Figure 2 A structural diagram of the device for optimizing enterprise manufacturing resource configuration based on intelligent interaction and template mapping provided in the embodiments of the present application; Figure 3 A structural diagram of the electronic device provided in the embodiments of the present application.

[0017] Reference signs: 110, analysis module; 120, calling module; 130, calculation module; 140, information feedback module; 810, processor; 820, communication interface; 830, memory; 840, communication bus. DETAILED DESCRIPTION

[0018] In order to more clearly illustrate the overall concept of the present application, the following will be described in detail with reference to the accompanying drawings.

[0019] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be apparent to one skilled in the art that the present application can be practiced without the specific details and other implementations can be employed. In other instances, well-known methods have not been described in detail in order to avoid obscuring the present application. It will be appreciated that embodiments of the present application can be used in combination with each other unless specifically stated otherwise, and that features of the various embodiments can be combined with each other, unless specifically stated otherwise.

[0020] In the present application, unless otherwise explicitly specified and limited, the first feature is "on" or "under" the second feature, which can be direct contact between the first and second features, or indirect contact between the first and second features through an intermediate medium. In the description of the present application, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples.

[0021] As shown in Figure 1 The first aspect of the present application provides a method for optimizing enterprise manufacturing resource configuration based on intelligent interaction and template mapping, which comprises: Step 100, receiving a resource configuration request input by a user, and performing natural language analysis on the resource configuration request to generate structured resource configuration requirement information.

[0022] Step 200, matching and calling a preset resource configuration template according to the structured resource configuration requirement information.

[0023] Step 300, based on the called resource configuration template, combining real-time resource state data in the ERP system to perform simulation calculation of resource configuration, and generating at least one resource configuration scheme.

[0024] Step 400, feeding back at least one resource configuration scheme to the user, and receiving feedback information of the user on the resource configuration scheme.

[0025] In step 100, the system receives a resource configuration request input by a user in natural language through an intelligent interaction interface, such as "allocate the numerical control machine tool and assembly station required for producing B-type motors for order A123, and complete within three days"; a pre-trained natural language processing model is used to perform word segmentation, named entity recognition, and intent understanding on the request, and extract key entities (such as order number A123, product model B-type motor, resource type numerical control machine tool / assembly station, time requirement three days) and operation intent (add resource allocation); the extracted information is structured into standardized data containing fields such as "order-product-resource type-quantity-time node-operation intent", facilitating subsequent system processing, thereby realizing low-threshold and high-accuracy human-computer interaction input.

[0026] In step 200, the system converts the structured resource configuration requirement information generated in step 100 into a first feature vector, and performs similarity calculation with second feature vectors of various resource configuration templates pre-stored in a template library, the template feature vectors cover applicable product categories, process flows, equipment combinations, resource constraints, etc.; the most suitable resource configuration template is determined and called through a semantic matching algorithm (such as cosine similarity); if the matching degree is below a threshold, a general resource configuration strategy is started to ensure that the system still has basic decision-making ability when facing new scenarios, realizing efficient reuse and flexible adaptation of optimized knowledge.

[0027] In step 300, the system obtains resource state data related to the current task from the ERP system in real time, including device availability, material inventory, personnel scheduling, work-in-process progress, etc., and injects it into the target resource configuration template called in step 200 to build a resource configuration optimization model; based on the model, linear programming or mixed integer programming optimization algorithms are used to minimize production cycle, maximize equipment utilization, or minimize scheduling cost as the goal, and perform multi-scheme simulation calculation; multiple candidate resource configuration schemes that meet the constraints of capacity, process, and delivery date are generated, and each scheme is scored and ranked, outputting the optimal or multiple alternative schemes, ensuring the scientificity and optimality of resource configuration decisions.

[0028] In step 400, the system presents one or more generated resource configuration schemes to the user through a graphical interface (such as a Gantt chart, resource load chart, or three-dimensional layout view), and provides a scheme comparison function to support the user in evaluating production cycle, resource cost, load balancing degree, etc. of different schemes; the user can provide feedback on the scheme by clicking, dragging, or natural language instructions, including accepting, rejecting, or partially modifying; the system receives and records this feedback information for subsequent analysis of user preferences, diagnosis of decision bias, and driving of resource configuration template self-learning optimization, forming a closed-loop mechanism of "decision-feedback-evolution" to continuously improve the system's intelligence level.

[0029] According to the method for optimizing enterprise manufacturing resource configuration based on intelligent interaction and template mapping provided by the first aspect of the application, first, the resource configuration request input by the user in the form of natural language is received, and natural language processing technology is used for in-depth analysis to accurately extract key information such as production order, product model, material, equipment, work station, quantity, time and operation intention, and generate structured resource configuration demand information, which greatly reduces the use threshold of the user and improves the friendliness and efficiency of human-computer interaction. Then, the system intelligently matches the structured demand information with a preset resource configuration template library, automatically calls the most suitable resource configuration template through feature vector similarity calculation, and realizes rapid reuse and scenario application of the optimization strategy. On this basis, the system deeply integrates real-time resource state data (such as equipment running state, material inventory, personnel scheduling, etc.) in the ERP system, constructs an optimization model based on the called template, performs multi-scenario simulation calculation by using an advanced mathematical programming algorithm, and generates multiple high-quality candidate resource configuration schemes to ensure the scientificity and optimality of the resource configuration decision. Finally, the system feeds back the generated scheme to the user in the form of Gantt chart, resource load chart and other visual forms, supports interactive confirmation, veto or modification of the user, and collects user preferences through the closed-loop feedback mechanism to provide data support for continuous self-learning and optimization of the template. Through the closed-loop process of “intelligent interaction-template mapping-data driven-feedback learning”, the method effectively solves the problems of traditional resource configuration methods, such as dependence on manual experience, slow response, poor flexibility, and difficulty in knowledge sedimentation, and significantly improves the intelligent level, decision efficiency and accuracy of resource configuration.

[0030] In some embodiments of the application, a resource configuration request input by a user is received, and natural language analysis is performed on the resource configuration request to generate structured resource configuration demand information, specifically: receiving a resource configuration request input by a user in the form of natural language; using a pre-trained natural language processing model to perform word segmentation, part-of-speech tagging and named entity recognition on the resource configuration request to extract production order number, product model, material code, equipment type, work station number, quantity required and planned completion time; identifying the operation intention in the resource configuration request based on a semantic understanding algorithm, the operation intention including adding resource allocation, changing existing resource plan or querying resource occupation state; structurally combining the extracted production order number, product model, material code, equipment type, work station number, quantity required, planned completion time and identified operation intention to generate structured resource configuration demand information containing the fields of “production order-product model-material-equipment-work station-quantity-time-intention”.

[0031] The system first receives the resource allocation request expressed in daily language through an intelligent interaction interface such as a chat window or voice input. This stage takes unstructured text as the original input, preparing for subsequent analysis. Tokenization refers to the process of cutting continuous text (e.g., "Allocate 2 CNC500 devices for order P20240910, complete by September 15th") into independent word or symbol units ("for / order / P20240910 / allocate / 2 / devices / CNC500 / , / September / 15th / complete").

[0032] Part-of-speech tagging refers to the process of tagging each tokenized result with its grammatical role (e.g., noun, verb, numeral, time word, etc.), helping the system understand sentence structure.

[0033] Named Entity Recognition (NER) is a key step. The system uses NLP models pre-trained on a large amount of manufacturing domain text (such as BERT-based sequence labeling models) to accurately identify and classify entities with specific meanings in the text. For example, the model can identify "P20240910" as "production order number", "CNC500" as "device type", "2" as "quantity required", and "September 15th" as "planned completion time". This relies on the model's deep learning capabilities for domain-specific vocabulary and contextual semantics.

[0034] After extracting the entities, the system further uses semantic understanding algorithms (such as intent recognition based on classification models) to analyze the context and verbs (such as "allocate", "adjust", "query") of the entire request to determine the user's operation intent. For example, "allocate" usually corresponds to "new resource allocation", "adjust" corresponds to "change existing resource plan", and "query" corresponds to "query resource occupation status". This ensures that the system can accurately understand the user's true needs.

[0035] Finally, the system integrates all the entity information (production order number, product model, etc.) and operation intent identified in the previous steps according to the predefined data structure (such as JSON, database table structure) to generate a structured data package containing fields such as "production order - product model - material - device - station - quantity - time - intent". This data package is the input that can be directly read and processed by subsequent processes (such as template matching, simulation calculation).

[0036] Users do not need to learn complex system operations or fill out tedious forms, but only need to describe their needs in daily language, greatly reducing the threshold for use, especially for front-line production managers to quickly initiate resource allocation requests, improving operational efficiency. Compared with manual input, automated parsing based on NLP can more accurately and consistently extract key information from text, effectively avoiding human input errors (such as order number input error, quantity misreading), and ensuring the data quality of subsequent decision-making. Different users may use different expressions to describe the same requirement (such as "give me two CNC machines" vs "apply for 2 CNC500 devices"). NLP models can understand these semantically equivalent expressions and map them to the same structured fields, enabling standardized processing of diverse inputs and laying the foundation for subsequent automated decision-making.

[0037] In some embodiments of the present application, according to the structured resource allocation requirement information, a preset resource allocation template is matched and called, specifically: The structured resource allocation requirement information is converted into a first feature vector; The second feature vector of each resource allocation template is obtained from the preset resource allocation template library, and the second feature vector includes the production type, product category, process route, device combination and resource constraint condition applicable to the template; The semantic similarity between the first feature vector and each second feature vector is calculated; The resource allocation template with the highest semantic similarity is called as the target template; If the highest semantic similarity is lower than a preset threshold, a template mismatch prompt is generated, and a general resource allocation strategy based on a rule engine is started as a backup solution.

[0038] The system encodes the generated structured resource allocation requirement information (such as production order, product model, device type, quantity, time, etc. fields) into a high-dimensional numerical vector (i.e. "first feature vector"). This process usually uses One-Hot Encoding, Word Embedding or direct numerical mapping, etc. so that each requirement instance can have a unique coordinate point in a multi-dimensional space. For example, "product model = B motor" may be mapped to a specific dimension in the vector, and "quantity required = 50" corresponds to another specific value of the dimension.

[0039] The system maintains a "resource configuration template library", each template of which is pre-defined with its applicable scenario characteristics. These characteristics (such as production type-discrete manufacturing, product category-motor series, process route-machining + assembly, equipment combination-CNC + robot, resource constraint-maximum waiting time 2 hours, etc.) are also encoded as high-dimensional numerical vectors, i.e. "second characteristic vectors". These vectors are essentially "knowledge fingerprints" after historical optimization or expert experience solidification.

[0040] The system calculates the distance or similarity between the "first characteristic vector" and all "second characteristic vectors" in the template library. Common algorithms include cosine similarity (measuring directional consistency), Euclidean distance (measuring spatial distance), or more complex neural network similarity models. The higher the similarity, the closer the current demand is to the applicable scenario of a certain template.

[0041] The system selects the template with the highest similarity as the "target template" for calling to guide subsequent resource configuration simulation. To deal with new scenarios or abnormal requests (i.e. no high matching template), the system sets a similarity threshold. If the highest similarity is lower than the threshold, it is determined as "template mismatch", at which point the general strategy based on the rule engine is started (such as rough allocation according to first-come-first-served, minimum load priority, etc.), ensuring that the system always has basic decision-making ability and avoids process interruption.

[0042] The enterprise's internal expert experience or historical successful cases are solidified as "resource configuration templates" and stored in a vectorized manner, making implicit knowledge explicit and computable. When encountering similar scenarios, the system can automatically call the optimal strategy, avoiding repeated exploration and significantly improving decision-making efficiency and quality. The matching mechanism based on semantic similarity can accurately identify the nature of the business scenario behind the demand, rather than simple keyword matching. For example, it can distinguish between "emergency insertion" and "regular production" and ensure that the most suitable optimization template is called, improving the rationality of the resource configuration scheme. The dual mechanism of "template mismatch prompt + general strategy backup" enables the system to pursue the optimal solution (call specialized templates) and maintain basic functionality in unknown scenarios, effectively dealing with uncertainties in the production environment and ensuring stable operation of the system.

[0043] In some embodiments of the present application, based on the called resource configuration template, combined with real-time resource state data in the ERP system, the simulation calculation of resource configuration is performed to generate at least one resource configuration scheme, specifically: Real-time resource state data related to the target template is obtained from the ERP system, including device current running state, personnel scheduling information, material inventory level, work-in-process flow progress, and workstation occupation situation; Inject resource state data into the target resource configuration template of the call to build a resource configuration optimization model; Based on the resource configuration optimization model, linear programming or mixed integer programming algorithm is used to minimize at least one of the production cycle, maximize the equipment utilization or minimize the resource scheduling cost as the optimization target, and perform multi-scenario simulation calculation; Generate multiple candidate resource configuration schemes that meet the constraint conditions, and score each scheme comprehensively to output at least one resource configuration scheme with the highest score.

[0044] The system pulls real-time dynamic data closely related to the current resource configuration task from the ERP (Enterprise Resource Planning) system through API interface or middleware. These data include: Current device running state (idle, processing, failure, maintenance); Staff scheduling information (on-duty personnel, skill level, work duration); Material inventory level (raw material, semi-finished product, finished product inventory); Work-in-process flow progress (completion of each process, queuing queue); Station occupancy (whether occupied by other orders, estimated release time); These data constitute the "realistic constraints" of resource configuration decision-making, ensuring that simulation calculations are based on real production environment.

[0045] Inject the real-time data obtained in the previous step into the target resource configuration template that has been called. The resource configuration template is essentially a predefined "optimization framework" that includes the objective function (such as minimizing the cycle), decision variables (such as assigning which device, when to start) and constraint conditions (such as process sequence, device capacity, material complete set) in this type of scene. Real-time data provides specific parameter values for these variables and constraints, thereby instantiating the abstract template into a specific, solvable mathematical model (such as linear programming or mixed integer programming model).

[0046] The system uses mature mathematical optimization algorithms (such as simplex method, branch and bound method, etc.) to solve the model. In order to improve the robustness and selectivity of the scheme, the system can set multiple optimization targets (such as considering both cycle and cost), and run multiple simulation calculations to generate multiple candidate resource configuration schemes that meet all hard constraints (such as not exceeding delivery date, not exceeding equipment load). For example, one scheme may focus on the shortest cycle, and another scheme may focus on the lowest energy consumption.

[0047] For the generated multiple candidate solutions, the system conducts comprehensive evaluation according to the preset scoring system (such as weighted scoring method). The scoring dimensions can include production cycle, resource utilization rate, scheduling cost, load balancing degree, risk coefficient, etc. Finally, the system outputs one or more solutions with the highest comprehensive score for user reference, realizing the optimal recommendation under the multi-objective trade-off.

[0048] The optimization algorithm based on mathematical programming can search for the optimal or near-optimal resource allocation scheme in the global feasible solution space, far exceeding the optimization level that can be achieved by artificial experience, significantly improving resource utilization efficiency and production performance. The fusion of real-time data of the ERP system enables the resource allocation scheme to be generated based on the latest production status at all times, enabling a quick response to equipment failures, material delays, order changes, and other sudden situations, enhancing the agility and flexibility of the manufacturing system. By generating multiple candidate solutions and conducting comprehensive scoring, the system can balance the conflicting goals of "fast", "economical", "stable", etc., providing a more comprehensive decision-making basis for managers and avoiding suboptimal results caused by single-objective optimization.

[0049] In some embodiments of the present application, at least one resource allocation scheme is fed back to the user, and feedback information of the user on the resource allocation scheme is received, specifically: The at least one resource allocation scheme is presented to the user in the form of a Gantt chart, a resource load chart, or a three-dimensional visual layout chart; A scheme comparison function is provided in the interactive interface, allowing the user to compare the production cycle, resource cost, equipment load balancing degree, and material set rate indicators of different resource allocation schemes horizontally; Receiving the user's confirmation, rejection, or modification operation on any resource allocation scheme through clicking, dragging, or natural language instruction input, the modification operation including adjusting the resource allocation object, changing the job time window, or replacing the standby material; The user's confirmation, rejection, or modification operation is recorded as user feedback information, and the optimization preference features are extracted for subsequent template self-learning update.

[0050] The core principle of this step is to build a human-machine collaborative closed-loop feedback system, realizing the deep integration of machine intelligence and human experience through intuitive visualization, convenient interaction, and intelligent data analysis, and its technical basis lies in the circulation mechanism of "visual presentation-interactive decision-feedback learning": Multi-dimensional visual presentation: The system converts abstract resource allocation schemes (such as task allocation, time scheduling) into graphical interfaces that are easy for users to understand. Specific forms include: Gantt chart: clearly shows the scheduling, dependency relationship, and resource occupation of each production task on the time axis, facilitating the evaluation of the overall production cycle.

[0051] Resource load chart: Displays the load rate of equipment, workstations, or personnel in the form of a bar chart or heat map to help identify bottleneck resources and idle periods.

[0052] 3D visualization layout diagram: Dynamically simulates material flow, equipment operation, and personnel work in a virtual factory environment, providing immersive spatial perception.

[0053] This multi-dimensional and multi-modal presentation method greatly reduces the cognitive burden on users to understand complex scheduling plans.

[0054] Key performance indicators (KPIs) for multiple candidate solutions are displayed side by side on the same interface, such as production cycle, resource cost, equipment load balance, and material consistency rate. Users can compare them horizontally by sliding and switching views, intuitively weighing the pros and cons of each solution and making more informed choices.

[0055] The system supports multiple user input methods to enhance interaction flexibility: Click / Drag: Users can drag the task bar directly on the Gantt chart to adjust the start time, or click the resource icon to reassign it.

[0056] Natural language instructions: Users can input instructions such as "advance task A to tomorrow morning" or "switch to a spare CNC machine tool", and the system will automatically execute the modification after NLP analysis.

[0057] Confirm / Reject: The user clicks "Confirm" for a satisfactory solution and "Reject" for an unsatisfactory solution.

[0058] These operations break the rigid processes of traditional systems and give users a high degree of autonomous decision-making power.

[0059] The system records every user confirmation, rejection, or modification action as "user feedback." Using data mining and machine learning algorithms (such as association rule analysis and preference learning models), the system extracts "optimization preference features" from this feedback. For example, if a user repeatedly rejects a high-load solution, the system learns that the user prefers "load balancing." If a user frequently replaces a certain type of material, the system learns that this material presents a supply risk. These features are then used to adjust optimization template parameters and perform self-learning updates.

[0060] The visualization interface makes the complex resource allocation logic "visible, tangible, and understandable", and users no longer face "black box" outputs, thereby enhancing the willingness to trust and adopt the system's recommended solutions. The first-line managers' field experience, emergency handling capabilities, and judgments on non-quantitative factors (such as interpersonal relationships and potential risks) are integrated into the system through a feedback mechanism, making up for the limitations of pure algorithmic models and achieving intelligent optimization with "people in the loop". Real-time feedback from users enables the system to quickly correct erroneous decisions caused by data delays, model biases, or unexpected situations, improving the practical feasibility of resource allocation solutions and the ability to respond to complex environments.

[0061] In some embodiments of the present application, the method further comprises: If the user performs a modification operation on the generated resource allocation solution, the implicit rules embodied in the modification operation are analyzed, and template parameter adjustment suggestions are generated; Compare the template parameter adjustment suggestions with the historical optimization records to identify repetitive or trend optimization requirements; When the trigger frequency of the same type of parameter adjustment suggestion exceeds a preset threshold, automatically update the parameter configuration or constraint conditions of the corresponding template in the resource allocation template library; Mark the updated resource allocation template as "optimized" and increase its call priority in the next matching process.

[0062] When the user performs a modification operation on the generated resource allocation solution, the system attempts to analyze the implicit rules embodied in these operations. For example, if the user repeatedly extends the time window of a certain type of task or tends to use a specific resource combination, it may reflect certain business priorities or limitations that are not captured by the current model. Based on this analysis, the system can automatically generate template parameter adjustment suggestions, such as adding time buffers, redefining resource allocation strategies, etc.

[0063] By comparing the generated template parameter adjustment suggestions with the historical optimization records, the system can identify which adjustments are repeatedly occurring or there is a certain trend of change. For example, the resource load of a certain project continues to exceed expectations, resulting in frequent adjustments to load balancing strategies, which indicates that the existing template may have deficiencies and needs to be improved.

[0064] Once it is found that the trigger frequency of a certain type of parameter adjustment suggestion exceeds a preset threshold (which can be flexibly set according to actual business conditions), the system will automatically update the parameter configuration or constraint conditions of the corresponding template in the resource allocation template library. This not only corrects known problems in a timely manner, but also prevents similar problems from occurring again in future projects, improving resource allocation efficiency and effectiveness.

[0065] To maximize the utilization of these proven optimization measures, the system will mark the updated resource configuration template as "optimized" and prioritize its invocation in the next matching process. This means that when facing similar types of resource configuration needs, the "optimized" template will be given priority, thereby speeding up the decision-making process and ensuring higher quality results.

[0066] In some embodiments of the present application, the method further comprises: Retrieving real-time resource state data and the invoked resource configuration template version relied upon when the rejected solution was generated, to build a solution generation context; Analyzing the potential reasons for the rejection of the solution based on a pre-set rule engine and machine learning model, including material inventory data delay, equipment failure unsynchronization, missing process constraint conditions, or template applicable scenario misjudgment; Generating a diagnostic report containing problem type, impact scope, and correction suggestions, and pushing it to system administrators and relevant business personnel; According to the correction suggestions in the diagnostic report, automatically triggering ERP system data verification tasks or resource configuration template retraining processes.

[0067] When a user rejects a resource configuration solution, the system immediately retrieves all input data relied upon when the solution was generated, including: Real-time resource state data: such as the recorded material inventory, equipment operating status, workstation occupancy, etc. at that time; Invoked resource configuration template version: including template ID, parameter configuration, constraint conditions, etc. metadata.

[0068] These data collectively constitute a "historical snapshot" or "decision context" of solution generation, providing complete factual basis for subsequent analysis, ensuring that diagnosis is based on real scenarios rather than speculation.

[0069] The system combines a pre-set rule engine and machine learning model to conduct multi-dimensional diagnosis of the rejection event: Rule engine: based on expert experience to set hard rules for investigation. For example, "if the material inventory used in the solution > the current inventory in the ERP system, it may be a data delay"; "if the equipment assigned in the solution is currently in a 'failure' state, there is a state synchronization problem".

[0070] Machine learning model: use historical rejection cases to train classification or anomaly detection models to identify more complex patterns. For example, the model can learn that "when a certain type of template is used for non-standard products, it is often rejected due to missing process constraints", thereby judging it as "template applicable scenario misjudgment".

[0071] Through the synergy of rules and models, the system can accurately locate the root cause of the problem, such as data asynchronization, template inapplicability, and incomplete constraint conditions.

[0072] The system structures the analysis results into a diagnostic report, including: Problem type (such as "data delay", "template misuse"); Impact scope (such as "affecting the machining sequence scheduling of order P20240910"); Correction suggestions (such as "checking the ERP material inventory interface", "supplementing the process constraint rules for C-class products").

[0073] The report is automatically pushed to system administrators, data engineers, or process engineers through email, messaging systems, or management dashboards, achieving accurate problem alerting and responsibility assignment.

[0074] The system not only "diagnoses", but also "treats". According to the correction suggestions in the diagnostic report, the corresponding repair tasks are automatically triggered: If it is a data problem (such as inventory delay), automatically start the ERP system data verification and synchronization task; If it is a template problem (such as constraint missing), automatically trigger the retraining or parameter update process of the resource configuration template, which may include re-labeling data, adjusting model weights, or updating rule libraries.

[0075] This realizes the automatic connection from "finding problems" to "solving problems", significantly shortening the self-repair cycle of the system.

[0076] Through active analysis of rejection reasons and automatic repair, the system can quickly correct internal defects such as data errors and template deviations, reduce decision failures caused by system problems, and enhance the stability of overall operation. Each rejection becomes an opportunity for system learning and improvement. Through structured diagnosis and automatic repair, enterprises can systematically accumulate "pit-avoiding" experience, avoid the repeated occurrence of similar errors, and promote the evolution of the system from "passive response" to "active prevention". In the traditional mode, problem discovery and repair rely on manual investigation, which is time-consuming and costly. This method greatly reduces the dependence on maintenance personnel, shortens the problem response time, and improves the efficiency of system maintenance.

[0077] As shown in Figure 2 The second aspect embodiment of the present application provides an enterprise manufacturing resource configuration optimization device based on intelligent interaction and template mapping, which comprises: The analysis module 110 is adapted to receive the resource configuration request input by the user, and performs natural language analysis on the resource configuration request to generate structured resource configuration demand information; The calling module 120 is adapted to match and call the preset resource configuration template according to the structured resource configuration requirement information. The computing module 130 is adapted to perform simulation calculation of resource configuration based on the called resource configuration template and in combination with real-time resource state data in the ERP system, to generate at least one resource configuration scheme. The information feedback module 140 is adapted to feed back the at least one resource configuration scheme to the user and receive feedback information of the user on the resource configuration scheme.

[0078] The enterprise manufacturing resource configuration optimization apparatus based on intelligent interaction and template mapping provided by the second aspect embodiment of the present application can implement the enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping in any of the first aspect embodiments, and thus can implement any of the technical effects of the enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping, which will not be repeated here.

[0079] The third aspect embodiment of the present application provides an electronic device, including a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping in any of the first aspect embodiments when executing the program.

[0080] Figure 3 An example of an entity structure diagram of an electronic device is shown in FIG. 8. Figure 3 As shown in FIG. 8, the electronic device can include a processor 810, a communications interface 820, a memory 830 and a communications bus 840, wherein the processor 810, the communications interface 820 and the memory 830 complete mutual communication through the communications bus 840. The processor 810 can call logical instructions in the memory 830 to execute the enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping in any of the first aspect embodiments, which includes: Step 100, receiving a resource configuration request input by a user and performing natural language analysis on the resource configuration request to generate structured resource configuration requirement information.

[0081] Step 200, matching and calling a preset resource configuration template according to the structured resource configuration requirement information.

[0082] Step 300, performing simulation calculation of resource configuration based on the called resource configuration template and in combination with real-time resource state data in the ERP system, to generate at least one resource configuration scheme.

[0083] Step 400: Feedback at least one resource configuration plan to the user, and receive user feedback on the resource configuration plan.

[0084] Furthermore, the logic instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0085] In another aspect, the present invention further provides a computer program product, comprising a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping provided by the above methods, the method comprising: Step 100: Receive a resource configuration request input by a user, perform natural language parsing on the resource configuration request, and generate structured resource configuration requirement information.

[0086] Step 200: Match and call a preset resource configuration template according to the structured resource configuration requirement information.

[0087] Step 300: Based on the called resource configuration template and in combination with the real-time resource status data in the ERP system, simulate the resource configuration and generate at least one resource configuration plan.

[0088] Step 400: Feedback at least one resource configuration plan to the user, and receive user feedback on the resource configuration plan.

[0089] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for optimizing enterprise manufacturing resource configuration based on intelligent interaction and template mapping provided by the above methods is implemented. The method comprises: Step 100: Receive a resource configuration request input by a user, perform natural language parsing on the resource configuration request, and generate structured resource configuration requirement information.

[0090] Step 200, according to the structured resource configuration requirement information, matching and calling the preset resource configuration template.

[0091] Step 300, based on the called resource configuration template, combining the real-time resource state data in the ERP system, simulating the resource configuration, and generating at least one resource configuration scheme.

[0092] Step 400, feeding back at least one resource configuration scheme to the user, and receiving the feedback information of the user on the resource configuration scheme.

[0093] Finally, the application also provides a non-volatile computer storage medium, which stores computer executable instructions, and the computer executable instructions are executed by a processor to realize the enterprise manufacturing resource configuration optimization method based on intelligent interaction and template mapping provided by the above method, and the method comprises the following steps: Step 100, receiving the resource configuration request input by the user, and performing natural language analysis on the resource configuration request to generate structured resource configuration requirement information.

[0094] Step 200, according to the structured resource configuration requirement information, matching and calling the preset resource configuration template.

[0095] Step 300, based on the called resource configuration template, combining the real-time resource state data in the ERP system, simulating the resource configuration, and generating at least one resource configuration scheme.

[0096] Step 400, feeding back at least one resource configuration scheme to the user, and receiving the feedback information of the user on the resource configuration scheme.

[0097] The places not mentioned in the application can be realized by using or referring to the existing technology.

[0098] Each embodiment in the specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other, and each embodiment mainly describes the difference from other embodiments.

[0099] The above only describes the embodiments of the application and does not limit the application. Those skilled in the art can make various modifications and changes to the application. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the application shall be included in the protection scope of the application.

Claims

1. A method for intelligent interaction and template mapping based optimization of enterprise manufacturing resource allocation, comprising: The method comprises the following steps: receiving a resource configuration request input by a user and performing natural language parsing on the resource configuration request to generate structured resource configuration requirement information; matching and calling a preset resource configuration template according to the structured resource configuration requirement information; based on the called resource configuration template, combining real-time resource state data in an ERP system to perform simulation calculation of resource configuration and generate at least one resource configuration scheme; feeding back the at least one resource configuration scheme to the user and receiving feedback information of the user on the resource configuration scheme.

2. The smart interaction and template mapping based enterprise manufacturing resource configuration optimization method of claim 1, wherein, The receiving of the resource configuration request input by the user and the natural language parsing of the resource configuration request to generate the structured resource configuration requirement information specifically comprises: receiving a resource configuration request input by a user in a natural language form; using a pre-trained natural language processing model to perform word segmentation, part-of-speech tagging and named entity recognition on the resource configuration request to extract a production order number, a product model, a material code, a device type, a workstation number, a required quantity and a planned completion time; identifying an operation intention in the resource configuration request based on a semantic understanding algorithm, the operation intention including adding resource allocation, changing an existing resource plan or querying a resource occupation state; structurally combining the extracted production order number, product model, material code, device type, workstation number, required quantity, planned completion time and identified operation intention to generate structured resource configuration requirement information containing "production order-product model-material-device-workstation-quantity-time-intention" fields.

3. The smart interaction and template mapping based enterprise manufacturing resource configuration optimization method of claim 1, wherein, The matching and calling of the preset resource configuration template according to the structured resource configuration requirement information specifically comprises: converting the structured resource configuration requirement information into a first feature vector; obtaining a second feature vector of each resource configuration template from a preset resource configuration template library, the second feature vector containing a production type, a product category, a process route, a device combination and a resource constraint condition applicable to the template; calculating semantic similarity between the first feature vector and each second feature vector; calling a resource configuration template with the highest semantic similarity as a target template; if the highest semantic similarity is lower than a preset threshold, generating a template mismatch prompt and starting a general resource configuration strategy based on a rule engine as a backup scheme.

4. The smart interaction and template mapping based enterprise manufacturing resource configuration optimization method of claim 3, wherein, The simulation calculation of resource configuration based on the called resource configuration template and the real-time resource state data in the ERP system to generate at least one resource configuration scheme specifically comprises: real-time acquisition of resource state data related to the target template from the ERP system, the resource state data including a current running state of a device, personnel scheduling information, a material inventory level, a work-in-process flow progress and a workstation occupation state; injecting the resource state data into the called target resource configuration template to build a resource configuration optimization model; based on the resource configuration optimization model, using a linear programming or mixed integer programming algorithm to perform multi-scenario simulation calculation with at least one of minimizing a production cycle, maximizing device utilization or minimizing resource scheduling cost as an optimization target; A plurality of candidate resource configuration schemes satisfying the constraint conditions are generated, and each scheme is comprehensively scored, and at least one resource configuration scheme with the highest score is output.

5. The smart interaction and template mapping based enterprise manufacturing resource configuration optimization method of claim 4, wherein, The at least one resource configuration scheme is fed back to the user, and feedback information of the user on the resource configuration scheme is received, specifically: The at least one resource configuration scheme is presented to the user in the form of a Gantt chart, a resource load chart, or a three-dimensional visual layout chart; A scheme comparison function is provided in the interactive interface, allowing the user to compare the production cycle, resource cost, equipment load balancing, and material set rate indicators of different resource configuration schemes horizontally; Receive the user's confirmation, veto or modification operation on any resource configuration scheme through clicking, dragging or natural language instruction input, and the modification operation includes adjusting the resource allocation object, changing the job time window or replacing the standby material; Record the user's confirmation, veto or modification operation as user feedback information, and extract the optimization preference features for subsequent template self-learning update.

6. The smart interaction and template mapping based enterprise manufacturing resource configuration optimization method according to any one of claims 1 to 5, characterized in that, The method further comprises: If the user performs a modification operation on the generated resource configuration scheme, analyze the implicit rules embodied in the modification operation, and generate template parameter adjustment suggestions; Compare the template parameter adjustment suggestions with historical optimization records to identify repetitive or trend optimization requirements; When the trigger frequency of the same type of parameter adjustment suggestion exceeds the preset threshold, automatically update the parameter configuration or constraint condition of the corresponding template in the resource configuration template library; Mark the updated resource configuration template as "optimized" state, and improve its calling priority in the next matching process.

7. The smart interaction and template mapping based enterprise manufacturing resource configuration optimization method according to any one of claims 1 to 5, characterized in that, The method further comprises: Call the real-time resource state data and the called resource configuration template version when the rejected scheme is generated to build the scheme generation context; Based on the preset rule engine and machine learning model, analyze the potential reasons for the rejection of the scheme, including material inventory data delay, device failure unsynchronization, process constraint condition missing or template application scenario misjudgment; Generate a diagnostic report containing problem type, impact range and correction suggestion, and push it to the system administrator and related business personnel; According to the correction suggestion in the diagnostic report, automatically trigger the ERP system data verification task or the retraining process of the resource configuration template.

8. An apparatus for optimization of enterprise manufacturing resource planning based on intelligent interaction and template mapping, comprising: It includes: The analysis module is adapted to receive the resource configuration request input by the user, and perform natural language analysis on the resource configuration request to generate structured resource configuration requirement information; The calling module is adapted to match and call the preset resource configuration template according to the structured resource configuration requirement information; The calculation module is adapted to perform simulation calculation of resource configuration based on the called resource configuration template and the real-time resource state data in the ERP system, and generate at least one resource configuration scheme; The information feedback module is adapted to feed back the at least one resource configuration scheme to the user, and receive the feedback information of the user on the resource configuration scheme.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor implements the intelligent interaction and template mapping based enterprise manufacturing resource configuration optimization method of any one of claims 1-7 when executing the program. The processor implements the intelligent interaction and template mapping based enterprise manufacturing resource configuration optimization method of any one of claims 1-7 when executing the program.

10. A non-transitory computer storage medium having stored thereon computer- executable instructions which, when executed by a computer, cause the computer to perform: The computer executable instructions, when executed by the processor, implement the method for intelligent interaction and template mapping based enterprise manufacturing resource configuration optimization as claimed in any one of claims 1 to 7.

Images