Part intelligent precision machining pricing method and system based on large language model

Abstract

The application discloses a part intelligent precision machining quotation method and system based on a large language model, belongs to the technical field of intelligent manufacturing and artificial intelligence, and extracts key manufacturing process parameters by performing semantic analysis on unstructured manufacturing demand information through the large language model; the demand integrity is checked, and if the demand is incomplete, a structured process instruction list is output through multiple rounds of clarification inquiry; after analyzing a three-dimensional part model, machining features are identified, and when there are to-be-confirmed features, a machining feature list is generated through a man-machine collaborative confirmation mechanism; technical costs are calculated according to the structured process instruction list and the machining feature list; non-technical factor information is obtained through the large language model, and the non-technical factor information is checked and adjusted to be compliant, and non-technical costs are calculated; total quotation is obtained by integrating the two types of costs, and the total quotation and cost composition are visually displayed, and a quotation sheet is output after confirmation; the problems that the existing quotation system has difficulty in analyzing natural language demand, feature identification is unreliable, and lacks an interactive confirmation mechanism are solved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of intelligent manufacturing and artificial intelligence, and specifically relates to a part intelligent precision machining pricing method and system based on a large language model. BACKGROUND

[0002] In the current context of increasingly fierce global manufacturing competition, fast and accurate pricing capability has become a key link for manufacturing enterprises, especially small and medium-sized enterprises, to win orders and improve market competitiveness. Pricing is not only the accounting of product costs, but also a comprehensive reflection of the manufacturing capacity and market responsiveness of enterprises. Traditional manufacturing pricing processes are highly dependent on human experience. Process personnel need to repeatedly communicate and interpret the vague requirements described in natural language by customers (such as "milling slot here, depth about 5mm, surface finish"), and then combine three-dimensional CAD (Computer-Aided Design) models to decompose processes, estimate working hours, and calculate costs. This process often requires cross-departmental collaboration, takes from several hours to several days, is inefficient, and is prone to pricing deviations due to information asymmetry or experience differences.

[0003] In terms of technical application, existing solutions mainly include two categories: one is a computer-aided pricing (CAQ, Computer-Aided Quoting) system based on structured data input. Although this type of system can achieve a certain degree of automation, it cannot effectively understand and process unstructured natural language inputs, resulting in a semantic gap between requirement communication and cost estimation. The second is an integrated environment relying on professional CAD / CAM (Computer-Aided Manufacturing) platforms (such as Siemens NX (Siemens Computer-Aided Design / Manufacturing / Engineering Integrated Software), PTC Creo (Parametric Technology Corporation Creo, 3D design software for computer-aided design / manufacturing / engineering integration)). These platforms are powerful, but their high licensing fees and closed ecosystems exclude small and micro enterprises with limited resources. In addition, the above systems lack a mechanism for interactive confirmation with users at key decision points (such as when there is uncertainty in the recognition results of machining features), resulting in a one-way, "black box" pricing process whose accuracy depends entirely on the recognition accuracy of the background algorithm, and cannot cope with the challenges posed by complex models or vague requirements.

[0004] In recent years, with the rapid development of artificial intelligence technology, large language models (LLM) have shown great capabilities in natural language understanding and reasoning, providing a new technical path for intelligent decision-making in manufacturing. Existing research attempts to introduce semantic technology (such as ontology and knowledge systems) into the cost estimation field, conducting semantic reasoning through pre-defined manufacturing rules. However, the ability of such methods is limited by the pre-constructed knowledge base, making it difficult to handle open-domain unstructured natural language input and lacking dynamic interaction capabilities with users. Meanwhile, in terms of three-dimensional model processing, graph-based feature recognition methods (such as attributed adjacency graphs (AAG)) have been researched for many years, but their accuracy and reliability in identifying complex features (such as curved surfaces and intersecting features) still need to be improved. Moreover, there is no research that deeply integrates these methods with the semantic interaction capabilities of LLM to build an end-to-end intelligent pricing system.

[0005] Figure 7 The existing pricing system information selection diagram, where CNC represents Computer Numerical Control, ISO is International Organization for Standardization, and manufacturing processes and additional requirements can only be selected according to pre-set options. Obviously, the existing pricing system lacks an intelligent human-machine interaction interface based on natural language and cannot understand users' fuzzy processing requirements. The existing model-driven pricing system has a one-way and non-interactive processing flow, which cannot initiate collaborative confirmation with users at key nodes (such as when feature recognition is uncertain), resulting in limited accuracy of the pricing system. Existing technical solutions are mostly single-function tools (such as pure feature recognition software or pure cost calculation software), and fail to build a complete system that integrates "demand interaction understanding, model intelligent analysis, human-machine collaborative decision-making, and accurate pricing generation".

[0006] Therefore, the manufacturing industry, especially small and micro enterprises, urgently needs a new type of intelligent pricing solution. This solution should be able to understand fuzzy requirements described in natural language using LLM technology, clarify key process parameters through intelligent interaction with users, have high-reliability processing feature recognition capabilities, and initiate human-machine collaborative confirmation for recognition uncertainty. Finally, it should form a complete system that integrates "demand understanding, model analysis, interactive confirmation, and cost calculation" to provide accurate, fast, and reliable pricing services for enterprises at low cost and easy deployment. SUMMARY

[0007] The purpose of the present application is to overcome the problems of existing offer systems in natural language demand analysis difficulty, unreliable feature recognition and lack of interactive confirmation mechanism, and to propose a part intelligent precision machining offer method and system based on a large language model.

[0008] To achieve the above purpose, the present application adopts the following technical solutions:

[0009] In the first aspect, the present application provides a part intelligent precision machining offer method based on a large language model, comprising:

[0010] Unstructured manufacturing demand information and a three-dimensional part model are obtained, the unstructured manufacturing demand information is semantically analyzed by using a large language model, and key manufacturing process parameters are extracted; whether the unstructured manufacturing demand information is perfect is judged in combination with a preset manufacturing field knowledge rule base, if not, the large language model generates a clarification inquiry and interacts with the user for multiple rounds until the unstructured manufacturing demand information is perfect, and a structured process instruction list with key manufacturing process parameters is output;

[0011] After the three-dimensional part model is analyzed, a feature recognition algorithm is used to recognize the machining features, and whether there are to-be-confirmed features is judged, if not, the machining feature list is directly obtained, if there are to-be-confirmed features, after the human-computer collaborative confirmation mechanism is executed, the machining feature list is obtained;

[0012] The preset manufacturing field knowledge rule base and the preset real-time price database are called, and the technical cost is calculated according to the structured process instruction list with key manufacturing process parameters and the machining feature list; the non-technical factor information is obtained by the large language model interacting with the user, the user demand is judged to be reasonable or not in combination with the manufacturing field knowledge rule base and the actual execution feasibility, if not, the unreasonable user demand is prompted and adjusted through interaction negotiation until the user demand is reasonable, the non-technical user demand is obtained, and the non-technical cost is calculated according to the non-technical user demand;

[0013] The total offer is obtained by integrating the technical cost and the non-technical cost, the total offer and the cost composition are visually displayed, and the offer sheet is output after being confirmed by the user.

[0014] Further, the input form of the unstructured manufacturing demand information includes voice input and text input;

[0015] The format of the three-dimensional part model is the product model data exchange standard STEP format;

[0016] The key manufacturing process parameters include material type, machining precision, surface treatment method, special process requirement and delivery time;

[0017] The unstructured manufacturing demand information imperfect conditions include information missing, ambiguity or conflict.

[0018] Further, when the three-dimensional part model is parsed, a boundary representation B-Rep based geometry processing kernel is adopted;

[0019] The feature recognition algorithm adopts an enhanced feature recognition algorithm;

[0020] The machining features include holes, grooves, cavities and chamfers.

[0021] Further, if there is a feature to be confirmed, a human-computer collaborative confirmation mechanism is performed, including:

[0022] Marking the area of the feature to be confirmed in the visualization interface of the three-dimensional part model to obtain a to-be-confirmed area;

[0023] For the to-be-confirmed area, a large language model is used to generate multiple-choice or question-and-answer confirmation prompts to interact with the user and obtain a confirmation result;

[0024] According to the confirmation result, the machining feature parameters are corrected to obtain a collaborative confirmation machining feature list;

[0025] Output the collaborative confirmation machining feature list.

[0026] Further, the manufacturing field knowledge rule base includes cutting parameters, work time quota and material density;

[0027] The real-time price database includes material unit price and outsourcing processing quotation;

[0028] The technical cost includes material cost, processing cost, heat treatment cost and surface treatment cost;

[0029] The non-technical factor information includes delivery location, logistics mode, expected construction period;

[0030] The non-technical cost includes logistics cost and urgent cost.

[0031] Further, the visualization display includes a cost composition chart, the quotation sheet includes itemized fees, total fees and delivery period, and the output format of the quotation sheet is PDF format.

[0032] In a second aspect, the present application provides a part intelligent precision machining quotation system based on a large language model, including:

[0033] The demand intelligent analysis and interaction clarification module is used to obtain unstructured manufacturing demand information and a three-dimensional part model, utilize a large language model to perform semantic analysis on the unstructured manufacturing demand information, and extract key manufacturing process parameters; the manufacturing field knowledge rule library is combined to determine whether the unstructured manufacturing demand information is perfect, if not, the large language model generates a clarification inquiry and interacts with the user for multiple rounds until the unstructured manufacturing demand information is perfect, and a structured process instruction list with key manufacturing process parameters is output;

[0034] The feature recognition and man-machine collaborative confirmation module is used to analyze the three-dimensional part model, adopt a feature recognition algorithm to recognize machining features, and determine whether there is a feature to be confirmed, if not, a machining feature list is directly obtained, if there is a feature to be confirmed, a man-machine collaborative confirmation mechanism is executed, and a machining feature list is obtained;

[0035] The cost layering calculation and negotiation module is used to call the preset manufacturing field knowledge rule library and the preset real-time price database, calculate the technical cost according to the structured process instruction list with key manufacturing process parameters and the machining feature list; the large language model is interacted with the user to obtain non-technical factor information, the manufacturing field knowledge rule library and the actual execution feasibility are combined to determine whether the user demand is reasonable, if not, the unreasonable user demand is prompted and interactively negotiated and adjusted until the user demand is reasonable, the non-technical user demand is obtained, and the non-technical cost is calculated according to the non-technical user demand;

[0036] The quotation output module is used to integrate the technical cost and the non-technical cost to obtain a total quotation, visually display the total quotation and the cost composition, and output a quotation sheet after the user confirms.

[0037] Further, it further comprises:

[0038] The user interaction layer is used to provide a three-dimensional model visualization window, a voice or text input box, an intelligent dialogue flow interface, and a quotation result display panel, receive user input, render a model and a dialogue, and initiate a voice recognition operation;

[0039] The intelligent service layer is used to be constructed based on a micro-service architecture, deployed on a cloud or a local server, and perform large language model interaction services, geometric processing services, cost calculation services, and business logic services;

[0040] The data and knowledge layer is used to store structured data including material parameters, machining time, and equipment rate by utilizing a manufacturing field knowledge rule library, form a real-time price database by accessing dynamic data including market material prices and outsourcing quotations, and store user historical projects, model files, and quotation sheets through a project and quotation database;

[0041] A computing support layer is configured to provide hardware and basic software support required for running the user interaction layer, the intelligent service layer, and the data and knowledge layer;

[0042] A voice input device is configured to perform real-time conversion of input voice into text, and transmit the converted text to a large language model for interactive service;

[0043] A three-dimensional model preview device is configured to integrate a part visualization window and highlight specific features;

[0044] A dialogue interaction device is configured to simulate an instant messaging software interaction form to present a dialogue between a user and an artificial intelligence (AI) assistant.

[0045] In a third aspect, the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for intelligent and accurate pricing of part machining based on a large language model when executing the computer program.

[0046] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program, and the computer program is executable on a processor to implement the method for intelligent and accurate pricing of part machining based on a large language model.

[0047] Compared with the prior art, the present application has the following beneficial technical effects:

[0048] The method for intelligent and accurate pricing of part machining based on a large language model can accurately understand and efficiently analyze unstructured natural language requirements, avoid the problems of difficulty in analyzing natural language requirements, unreliable feature recognition, and lack of interactive confirmation mechanism in existing pricing systems, and fundamentally solve the problem of low efficiency of requirement communication. By introducing a large language model as the core of intelligent interaction, the system can directly understand the user's fuzzy or colloquial machining requirements and automatically extract key process parameters, completely changing the traditional interactive mode of menu selection and structured input. This converts the time-consuming requirement clarification process into a minute-level or even second-level automated interaction, greatly improving the efficiency and accuracy of requirement communication and effectively solving the semantic gap problem existing in traditional methods. By introducing a human-machine collaborative confirmation mechanism, the accuracy of the quotation is ensured. For the uncertainty in feature recognition, a human-machine interaction confirmation process is proposed, and user collaborative decision-making is performed, so that human intelligence is integrated into the automated process. This completely avoids the risk of outputting incorrect results due to algorithm limitations in traditional one-way processing systems, so that the accuracy of the quotation is no longer completely dependent on the recognition rate of the background algorithm, but is based on the reliable foundation of human-machine collaboration, fundamentally solving the problem of inaccurate quotation caused by false detection and missed detection. BRIEF DESCRIPTION OF DRAWINGS

[0049] The drawings described herein are for purposes of illustration only and are not intended to limit the scope of the present disclosure in any way. Additionally, the shapes and proportions of the various components depicted in the drawings are not intended to be specific, but rather are provided as illustrative examples for the purpose of understanding the present disclosure.

[0050] In the drawings:

[0051] Figure 1 Flow chart of the part intelligent precision machining pricing method based on large language model of the present application.

[0052] Figure 2 Structure diagram of the part intelligent precision machining pricing system based on large language model of the present application.

[0053] Figure 3 Electronic device schematic diagram of the part intelligent precision machining pricing method based on large language model of the present application.

[0054] Figure 4 Flow chart of the intelligent pricing method in the embodiment of the present application.

[0055] Figure 5 System overall architecture diagram in the embodiment of the present application.

[0056] Figure 6 Pricing sheet schematic diagram in the embodiment of the present application.

[0057] Figure 7 Existing pricing system information selection schematic diagram. DETAILED DESCRIPTION

[0058] In order for those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

[0059] Embodiment one

[0060] Referring to Figure 1 The part intelligent precision machining pricing method based on large language model, comprising:

[0061] The unstructured manufacturing demand information and the three-dimensional part model are acquired, the unstructured manufacturing demand information is semantically parsed by using a large language model, and key manufacturing process parameters are extracted; whether the unstructured manufacturing demand information is perfect is judged in combination with a preset manufacturing field knowledge rule base, if not, the large language model generates a clarification inquiry and interacts with the user in multiple rounds until the unstructured manufacturing demand information is perfect, and a structured process instruction list with key manufacturing process parameters is output; after the three-dimensional part model is parsed, a feature recognition algorithm is used to recognize machining features, and whether there are to-be-confirmed features is judged, if not, a machining feature list is directly obtained, if there are to-be-confirmed features, a human-computer collaborative confirmation mechanism is executed, and a machining feature list is obtained; a preset manufacturing field knowledge rule base and a preset real-time price database are called, and technical cost is calculated according to the structured process instruction list with key manufacturing process parameters and the machining feature list; non-technical factor information is obtained by the large language model interacting with the user, whether the user demand is reasonable is judged in combination with the manufacturing field knowledge rule base and the actual execution feasibility, if not, the unreasonable user demand is prompted and interactive negotiation is adjusted until the user demand is reasonable, non-technical user demand is obtained, and non-technical cost is calculated according to the non-technical user demand; total quotation is obtained by integrating the technical cost and the non-technical cost, and the total quotation and cost composition are visually displayed, and the quotation sheet is output after being confirmed by the user.

[0062] The large language model semantically parses unstructured demand in this embodiment, can accurately extract key process parameters, complete the demand in combination with the manufacturing knowledge rule base, and avoid information bias; three-dimensional model feature recognition is combined with human-computer collaborative confirmation to ensure that there is no omission of machining features, and then the real-time price database is linked to calculate technical cost, and non-technical factors are interactively verified, and the total quotation fits the actual scene. Automatic analysis replaces manual information screening, multiple rounds of clarification inquiries shorten the demand confirmation period, feature recognition and cost calculation process are standardized, which greatly reduces the time-consuming of manual operation and improves the response speed of quotation. The human-computer collaborative mechanism solves complex problems such as feature confirmation and demand rationality judgment, visualized quotation display reduces user understanding cost, multiple rounds of interaction enhance the consensus of supply and demand, reduce subsequent consultation and rework, and balance technical rigor and user experience.

[0063] Embodiment two

[0064] Referring to Figure 2 , the part intelligent precise machining quotation system based on the large language model comprises:

[0065] The demand intelligent analysis and interaction clarification module is used to obtain unstructured manufacturing demand information and a three-dimensional part model, utilize a large language model to perform semantic analysis on the unstructured manufacturing demand information, and extract key manufacturing process parameters; the manufacturing field knowledge rule library is combined to determine whether the unstructured manufacturing demand information is perfect, if not, the large language model generates a clarification inquiry and interacts with the user for multiple rounds until the unstructured manufacturing demand information is perfect, and a structured process instruction list with key manufacturing process parameters is output;

[0066] The feature recognition and man-machine collaborative confirmation module is used to analyze the three-dimensional part model, adopt a feature recognition algorithm to recognize machining features, and determine whether there is a feature to be confirmed, if not, a machining feature list is directly obtained, if there is a feature to be confirmed, a man-machine collaborative confirmation mechanism is executed, and a machining feature list is obtained;

[0067] The cost layering calculation and negotiation module is used to call the preset manufacturing field knowledge rule library and the preset real-time price database, calculate the technical cost according to the structured process instruction list with key manufacturing process parameters and the machining feature list; the large language model is interacted with the user to obtain non-technical factor information, the manufacturing field knowledge rule library and the actual execution feasibility are combined to determine whether the user demand is reasonable, if not, the unreasonable user demand is prompted and interactively adjusted until the user demand is reasonable, non-technical user demand is obtained, and the non-technical cost is calculated according to the non-technical user demand;

[0068] The quotation output module is used to integrate the technical cost and the non-technical cost to obtain a total quotation, visually display the total quotation and the cost composition, and output a quotation sheet after user confirmation.

[0069] The user interaction layer is used to provide a three-dimensional model visualization window, a voice or text input box, an intelligent dialogue flow interface, and a quotation result display panel, receive user input, render models and dialogues, and initiate voice recognition operations;

[0070] The intelligent service layer is used to be constructed based on a micro-service architecture, deployed on a cloud or a local server, and perform large language model interaction services, geometry processing services, cost calculation services, and business logic services;

[0071] The data and knowledge layer is used to store structured data including material parameters, processing time and equipment rate by utilizing a manufacturing field knowledge rule library, form a real-time price database by accessing dynamic data including market material prices and outsourcing quotations, and store user historical projects, model files and quotation sheets through a project and quotation database;

[0072] The computing support layer is used to provide hardware and basic software support required for the user interaction layer, the intelligent service layer, and the data and knowledge layer.

[0073] A voice input device for real-time conversion of input voice into text, and the converted text is transmitted to a large language model for interactive services;

[0074] A three-dimensional model preview device for integrating a part visualization window and highlighting specific features;

[0075] A dialogue interaction device for simulating the interactive form of instant messaging software to present a user's dialogue with an artificial intelligence (AI) assistant.

[0076] Embodiment three

[0077] Referring to Figure 3 An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements a part intelligent precision machining pricing method based on a large language model when executing the computer program: obtaining unstructured manufacturing demand information and a three-dimensional part model, using a large language model to perform semantic analysis on the unstructured manufacturing demand information, and extracting key manufacturing process parameters; combining a pre-set manufacturing field knowledge rule base to determine whether the unstructured manufacturing demand information is complete, if not, the large language model generates a clarification inquiry and interacts with the user for multiple rounds until the unstructured manufacturing demand information is complete, and outputs a structured process instruction list with key manufacturing process parameters; after analyzing the three-dimensional part model, a feature recognition algorithm is used to recognize machining features, and it is determined whether there are features to be confirmed, if there are no features to be confirmed, a machining feature list is directly obtained, if there are features to be confirmed, a human-computer collaborative confirmation mechanism is executed, and a machining feature list is obtained; a pre-set manufacturing field knowledge rule base and a pre-set real-time price database are called, and based on the structured process instruction list with key manufacturing process parameters and the machining feature list, a technical cost is calculated; non-technical factor information is obtained by interacting with the user through the large language model, and it is determined whether the user's demand is reasonable in combination with the manufacturing field knowledge rule base and the actual execution feasibility, if not, the unreasonable user demand is prompted and interactive negotiation is adjusted until the user demand is reasonable, non-technical user demand is obtained, and a non-technical cost is calculated based on the non-technical user demand; the technical cost and the non-technical cost are integrated to obtain a total price, the total price and the cost composition are visually displayed, and after the user confirms, a pricing sheet is output.

[0078] Embodiment four

[0079] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements a part intelligent precision machining pricing method based on a large language model: obtaining unstructured manufacturing demand information and a three-dimensional part model, using a large language model to perform semantic analysis on the unstructured manufacturing demand information, and extracting key manufacturing process parameters; determine whether the unstructured manufacturing demand information is complete by combining the pre-set manufacturing field knowledge rule base, if not, the large language model generates a clarification inquiry and interacts with the user for multiple rounds until the unstructured manufacturing demand information is complete, and outputs a structured process instruction list with key manufacturing process parameters; after analyzing the three-dimensional part model, a feature recognition algorithm is used to identify the machining features, and it is determined whether there are to-be-confirmed features, if there are no to-be-confirmed features, the machining feature list is directly obtained, if there are to-be-confirmed features, after executing the man-machine collaborative confirmation mechanism, the machining feature list is obtained; call the pre-set manufacturing field knowledge rule base and the pre-set real-time price database, according to the structured process instruction list with key manufacturing process parameters and the machining feature list, the technical cost is calculated; obtain non-technical factor information by interacting with the user through the large language model, and determine whether the user demand is reasonable in combination with the manufacturing field knowledge rule base and the actual execution feasibility, if not, prompt the unreasonable user demand and interact to adjust until the user demand is reasonable, obtain the non-technical user demand, and calculate the non-technical cost according to the non-technical user demand; integrate the technical cost and the non-technical cost to obtain the total price, and visually display the total price and the cost composition, and output the price list after the user confirms.

[0080] Embodiment five

[0081] Referring to Figure 4 , where N is determined to be no, Y is determined to be yes, the part intelligent precision machining pricing method based on a large language model, interactive manufacturing intelligent pricing based on a large language model, the core is to build a complete closed loop of "demand intelligent analysis-model feature recognition and man-machine collaborative confirmation-cost accurate calculation". Through multi-level human-computer interaction, the whole process from fuzzy intention to accurate pricing is reliable, transparent and efficient. The following is described in detail from three aspects of method, device structure and integrated system.

[0082] The core of the intelligent pricing method proposed in this embodiment is to build a complete man-machine collaborative closed loop of "demand intelligent analysis-model feature collaborative identification-cost accurate calculation-result visualization feedback". The method is based on a large language model (LLM, Large Language Model), geometric model processing, intelligent decision-making and other technologies, see Figure 4 , which is specifically divided into the following four stages:

[0083] Step 1, multi-modal intelligent analysis and interactive clarification of manufacturing demand based on LLM

[0084] Step 101, input receiving: the system supports users to input unstructured natural language processing requirements through voice or text (for example: "process 50 aluminum alloy joints, mill a groove in the middle, oxidize the surface black, and deliver next Wednesday") and supports users to upload three-dimensional part models in STEP (Standard for the Exchange of Product Model Data) format. For example, in this embodiment, the system prompts "please upload your part model file", the user inputs a three-dimensional part model in STEP format and "want to make 20 parts for the medical field, the processing precision is 0.1 mm, the part is a stepped shaft, there are threaded segments at both ends, there are key grooves in the middle, it is used for gear transmission, the material is aluminum magnesium silicon alloy 6061, and the surface standard treatment" specific requirements, and starts interactive requirement confirmation.

[0085] Step 102, semantic understanding and key information extraction: the LLM module integrated in the system performs deep semantic analysis on the input text, identifies and extracts key manufacturing process parameters, including material type, processing precision, surface treatment method, special process requirement, delivery time, etc. These key manufacturing process parameters are used in the pricing link to evaluate the processing quotation according to the manufacturing process parameters required by the user.

[0086] Step 103, multi-round interactive clarification mechanism: the LLM module integrates manufacturing field knowledge rule base, automatically judges the completeness and rationality of user input. If it is found that the information is missing (such as not specifying the material), ambiguous (such as "the surface is a little bit bright"), or there is potential conflict (such as unreasonable roughness of medical part labeling), a clarification type natural language inquiry (for example: "please confirm whether the surface treatment requires anodizing black?") is automatically generated, and the user is prompted to confirm or supplement through the front-end dialog box. The process is iterated until the requirement is clear.

[0087] Step 104, output: the LLM module outputs a structured process instruction list as the input of the downstream module.

[0088] Step 2, feature recognition and human-computer collaborative confirmation

[0089] Step 201, model analysis and feature recognition: the system receives the STEP three-dimensional part model uploaded by the user, uses a boundary representation (B-Rep) based geometry processing kernel (such as PythonOCC (Python Open CASCADE Technology Bindings, a Python language based binding library for the open source three-dimensional geometry kernel Open CASCADE Technology)) for analysis, and uses an enhanced feature recognition algorithm to recognize machining features (such as holes, grooves, cavities, chamfers, etc.).

[0090] Step 202, uncertainty processing: the user determines whether there are any "to-be-confirmed" features that need to be modified, and if there are, proceeds to step 203 to perform a human-machine collaborative confirmation mechanism and interact with the LLM to re-determine the recognized machining features, and if there are not, proceeds to step 3 to start calculating the cost.

[0091] Step 203, human-machine collaborative confirmation mechanism: for "to-be-confirmed" features, the system automatically triggers an interactive process:

[0092] Step 2031, the front-end interface highlights the area of the feature in the three-dimensional model;

[0093] Step 2032, the LLM generates a multiple-choice or question-and-answer confirmation prompt (for example: "The highlighted area may be a 'key groove' or a 'tool withdrawal groove', please select the correct type?"); For example, the system features recognition human-computer interaction to analyze the features of the three-dimensional model, and displays the recognition progress and time used of the feature analysis, when it finds the feature that needs to be confirmed, it makes a prompt and points out the specific feature that needs to be confirmed, such as in this embodiment, the system finds that the feature that needs to be confirmed is a cavity (needs to be confirmed): based on the EWV-AAG algorithm and the cavity feature template matching, it is found that the concave / through area but the shape has limited matching degree with the single template; prompt the user to confirm the feature. Wherein, EW in the EW-AAG algorithm is Enhanced Weighted (enhanced weighted), AAG is Attributed Adjacency Graph (attributed adjacency graph), and the EW-AAG algorithm is an enhanced weighted attributed adjacency graph commonly used in image processing and model feature recognition fields;

[0094] Step 2033, the user confirms through point selection or voice, and the confirmation result is fed back to the system in real time for correcting the feature parameter list.

[0095] Step 2034, output: generate a precise machining feature list and its geometric parameters that have been confirmed by the user in collaboration.

[0096] Step 3, hierarchical calculation and negotiation of technical and non-technical costs

[0097] Step 301, technical cost calculation: the cost calculation engine receives the confirmed feature list and process instructions, calls the integrated manufacturing domain knowledge rule base (including cutting parameters, work time quota, material density, etc.) and real-time price database (including material unit price, outsourcing processing quotation, etc.), automatically calculates material cost, processing cost, heat treatment cost, surface treatment cost, etc., and dynamically updates and displays on the front-end interface.

[0098] Part processing cost:

[0099]

[0100] wherein, is the part processing cost, in yuan, is the part processing time, in minutes, is the machine tool hourly rate, in yuan per hour;

[0101] Part material cost:

[0102]

[0103] wherein, is the part material cost, in yuan, is the volume of the blank material, in cubic millimeters, is the density of the material, in grams per cubic centimeter, is the unit price of the blank material, in yuan per kilogram;

[0104] Surface treatment cost:

[0105]

[0106] wherein, is the surface treatment cost, is the surface area of the part, in square millimeters, is the unit price of ordinary surface treatment, in yuan per square meter;

[0107] Heat treatment cost:

[0108]

[0109] wherein, is the volume of the part, in cubic millimeters, is the unit price of the heat treatment process, in yuan per kilogram.

[0110] For example, the technical cost of manufacturing 20 parts for the medical field is calculated in this embodiment, and the details are shown, with a total of 2650 yuan, including material cost 500 yuan, processing cost 1200 yuan, surface treatment 300 yuan, feature processing 450 yuan, and quality control 200 yuan.

[0111] Step 302, non-technical cost interaction and calculation: LLM module initiates a dialogue, asking about the delivery location, logistics mode, expected delivery time, and other non-technical factors. LLM has the ability to make reasonable judgments. If the user's requirements are unreasonable (such as "deliver to a certain place within a day"), for example, the time required for transportation, the system will make a friendly suggestion and propose a feasible solution based on the interfaces of express service companies such as SF Express and Jingdong, etc. to determine the time between the destination and the sending location (according to the express delivery requirements, ordinary, SF Express, etc.). will be friendly and suggest a feasible solution until an agreement is reached. Subsequently, the system calculates the logistics, rush delivery, etc. based on the results of the negotiation, and calls the interfaces of express service companies such as SF Express to estimate the cost. For example, in this embodiment, when the user requires "express delivery, preferably tomorrow, the delivery address is a certain high-tech park, use reinforced packaging, protect it a little better, need value-added tax special invoice, no other special requirements", LLM analyzes the user's requirements and gives a friendly suggestion: "I must be very honest and tell you that it is impossible to receive the 20 parts processed tomorrow from a production point of view. A precision part also needs to go through the following necessary steps: 1. Programming and machine adjustment: write CNC machining program according to drawing and debug machine tool. 2. Processing: CNC turning (shaft), milling (keyway), tapping (thread). 3. Post-processing: surface treatment (such as anodizing, usually takes several hours). 4. Inspection and packaging: full inspection of size, reinforced packaging. This process takes 3-5 working days even if it is urgent."

[0112] Step 4, generation, visualization and export of quotation

[0113] Step 401, quotation integration: the system integrates all technical and non-technical costs to generate a total quotation.

[0114] Step 402, result visualization: the front-end interface dynamically generates a detailed quotation sheet (including itemized costs, total, delivery period, etc.), see Figure 6 For example, the quotation sheet indicates the part name, part type, quotation item and summary, and can switch to view the technical cost composition pie chart. The three-dimensional model view is continuously displayed and can be highlighted in association with the features.

[0115] Step 403, interactive confirmation and output: the LLM informs the user in a dialogue form that the quotation has been completed, and asks "whether to generate the final quotation sheet?" After the user confirms, the system provides a one-key export PDF (Portable Document Format) quotation sheet function, completing the entire intelligent quotation process.

[0116] To realize the above method, the part intelligent precision machining quotation method based on the large language model in the embodiment is also designed, and a part intelligent precision machining quotation system based on the large language model is designed, which is based on cloud edge end cooperation, and the hardware and software architecture is as follows:

[0117] I. The overall architecture of the system adopts hierarchical design, please refer to Figure 5 , which is divided into four layers in general:

[0118] First, the user interaction layer (front end): based on Web (World Wide Web) technology to develop responsive interface, running in the user's browser. Provide a three-dimensional model visualization window, voice / text input box, intelligent dialogue flow interface, quotation result display panel. This layer is responsible for receiving user input, rendering model and dialogue, and initiating voice recognition and other operations.

[0119] Second, the intelligent service layer (back end): based on micro-service architecture, deployed on the cloud or local server, including the following core services:

[0120] LLM interaction service: encapsulate large language model, responsible for demand analysis, dialogue management, clarification inquiry generation.

[0121] Geometric processing service: based on PythonOCC (Python Open CASCADE Technology Bindings, a binding library based on Python language to open source three-dimensional geometric kernel Open CASCADE Technology), responsible for STEP (Standard for the Exchange of Product Model Data) file analysis, feature recognition.

[0122] Cost calculation service: integrate knowledge base and database, execute cost calculation logic.

[0123] Business logic service: coordinate service calls, manage business processes and status.

[0124] Third, the data and knowledge layer: includes the following databases:

[0125] Manufacturing process knowledge base: stores structured data such as material parameters, processing time, equipment rates, etc. It is a structured and standardized collection of data related to material parameters and processing data used in the manufacturing process, serving as a processing quotation service.

[0126] Real-time price database: access market material prices, outsourcing quotes, and other dynamic data through API (Application Programming Interface) interface.

[0127] Project and quotation database: stores user historical projects, model files, and quotation sheets.

[0128] Fourth, computing support layer: provides hardware and basic software support required for system operation. When deployed, the system can run on a general workstation equipped with a modern CPU (Central Processing Unit), 16GB of memory, and LLM inference can be completed on a consumer-grade GPU (Graphics Processing Unit), demonstrating its friendliness to small and micro enterprises.

[0129] II. Core interaction devices and interfaces:

[0130] 1) Voice input device: the system supports user voice input through a microphone. The front end integrates Web Speech API (Web Speech Application Programming Interface) to realize real-time conversion of voice to text, and the converted text is automatically filled into the input box and transmitted to the backend LLM service in real time through WebSocket (a full-duplex communication protocol based on Transmission Control Protocol) protocol.

[0131] 2) Three-dimensional model preview device: the system front end integrates a part visualization window based on PythonOCC, providing model visualization functions and supporting highlighting of specific features according to backend instructions.

[0132] 3) Dialog interaction device: the front-end interface simulates the interaction form of instant messaging software, and the conversation between the user and the AI (Artificial Intelligence) assistant is presented in the form of bubbles, with a head icon, clearly distinguishing the conversation subject, providing a natural and smooth interaction experience.

[0133] Based on the above methods and architecture, the part intelligent precision machining quotation system based on large language model designed in this embodiment has the following workflow:

[0134] Requirement input and analysis: the user logs in to the system through the Web interface, uploads the part model and describes the requirements in natural language. The LLM service performs analysis and multi-round interaction, and outputs structured process instructions.

[0135] Model analysis and collaborative confirmation: The geometry processing service analyzes the model and identifies features. The system confirms the identification results with the user through the front-end interface, forming the final feature list.

[0136] Cost calculation and negotiation: The cost calculation service calculates the technical cost based on the feature list and process instructions. The LLM service negotiates non-technical terms with the user and calculates the corresponding fees.

[0137] Quotation generation and delivery: The system generates a detailed quotation and a visual chart. After user confirmation, the final quotation (in PDF format) is output, completing the service.

[0138] This embodiment scheme first realizes the intelligentization, interaction, and collaboration of the whole process of manufacturing quotation, significantly improving the efficiency and accuracy of quotation, especially suitable for small and medium-sized manufacturing enterprises that lack professional process personnel.

[0139] In this embodiment, the intelligent analysis of manufacturing quotation requirements and multi-round interaction use LLM as the intelligent hub to directly understand the user's fuzzy and unstructured processing requirements described in natural language, and through the built-in manufacturing field knowledge rule base, actively judge the completeness and rationality of the requirements, and then generate clarifying natural language inquiries and interact with the user in multiple rounds until the machine executable and structured process instructions are output. This method first applies the deep semantic understanding and reasoning ability of LLM to the field of manufacturing quotation, realizing a fundamental change from "man adapting to machine" to "machine understanding man", and significantly reducing the use threshold of professional quotation software.

[0140] In this embodiment, the feature recognition result of the quotation accuracy-oriented human-machine collaborative confirmation mechanism introduces a human-computer interaction confirmation process. After the system identifies the machining features of the three-dimensional CAD model, if the user believes that the identified machining features still need to be modified, the collaborative confirmation mechanism is triggered, the feature area is highlighted through the front-end interface, and easy-to-understand options or questions are generated by LLM to interact with the user for final decision-making. This method intelligently integrates human judgment into the automated process, ensuring 100% accuracy of the geometric features on which the quotation is based, effectively solving the inherent risk when dealing with uncertainty by pure algorithm.

[0141] The embodiment is based on a part intelligent precision machining quotation system of a large language model, provides a cloud collaborative manufacturing quotation system integrated with intelligent interaction, feature processing and cost calculation, and constructs a complete micro-service architecture including an intelligent interaction front end, an LLM service, a geometric processing service and a cost calculation service. The system seamlessly integrates the above-mentioned intelligent analysis of manufacturing quotation requirements and multi-round interaction using LLM as an intelligent center, a feature recognition result human-machine collaborative confirmation mechanism for quotation accuracy, real-time cost calculation, non-technical factor negotiation, quotation sheet generation and other functions, forming an end-to-end intelligent solution. The overall architecture of the system, the collaborative data flow between services (especially the transmission and processing logic of natural language, structured instructions, feature information and user confirmation signals in the system), and finally realizes the one-stop intelligent workflow of "demand input-interaction confirmation-quotation output".

[0142] Low-cost, easy-to-deploy system implementation solution. The core of this solution is to build the system based on an open source technology stack (such as Dify (Define Modify, an open source large language model application development platform based on the concept of definition and continuous improvement), PythonOCC), so that it can run stably without relying on expensive commercial software licenses or high-end hardware resources. The application of this specific open source technology combination in implementing the cloud collaborative manufacturing intelligent quotation system, and the significant economic and social benefits it brings, such as "making it affordable for small and micro enterprises".

[0143] Compared with traditional manual quotation or existing automated quotation software, the embodiment can bring the following significant effects:

[0144] The embodiment realizes accurate understanding and efficient analysis of unstructured natural language requirements, fundamentally solving the problem of low efficiency of demand communication: by introducing a large language model (LLM) as the core of intelligent interaction, the system can directly understand fuzzy or colloquial machining requirements such as "milling a groove, making the surface a little brighter", and automatically extract key process parameters, completely changing the traditional system's interaction mode of relying on menu selection and structured input. This converts the time-consuming demand clarification process that originally required repeated communication by process personnel into minutes or even seconds of automated interaction, greatly improving the efficiency and accuracy of demand communication, and effectively solving the semantic gap problem existing in traditional methods.

[0145] The embodiment introduces a man-machine cooperative confirmation mechanism to ensure the absolute accuracy of the quotation basis. For the uncertainty that may exist in feature recognition, a man-machine interactive confirmation process is proposed. User cooperative decision-making integrates human intelligence into the automated process. This completely avoids the risk of outputting incorrect results due to algorithm limitations in traditional one-way processing systems, making the accuracy of the quotation no longer completely dependent on the recognition rate of the background algorithm, but rather on the reliable foundation of man-machine cooperation, fundamentally solving the problem of inaccurate quotation caused by false detection and missed detection.

[0146] The embodiment provides a highly integrated, low-cost, and easy-to-deploy one-stop solution.

[0147] The embodiment constructs a complete system integrating "demand intelligent analysis - model cooperative recognition - cost accurate calculation" and is implemented through a fully open source technology stack. The system does not require expensive commercial software licenses and can run on ordinary computing devices, greatly reducing the hardware and software investment threshold and use cost of enterprises. At the same time, the natural language interaction mode greatly reduces the operation difficulty, so that small and micro enterprises can also enjoy the dividends brought by cutting-edge AI technology, effectively solving the contradiction between the high threshold and high cost of existing professional software and the single function of open source tools, and promoting the popularization of intelligent manufacturing technology.

[0148] Those skilled in the art will appreciate that embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, etc.) containing computer-usable program code.

[0149] The present application is described with reference to flowcharts and / or block diagrams according to the method, device (system), and computer program product of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The device that implements the functions specified in one or more flows and / or blocks. Figure 1 The device that implements the functions specified in one or more flows and / or blocks.

[0150] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0151] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions executed on the computer or other programmable apparatus provide steps for implementing the flow Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0152] Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limiting the same. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or equivalent replacements without departing from the spirit and scope of the present application, and any modifications or equivalent replacements shall be included in the protection scope of the present application.

Claims

1. A method for intelligent and precise machining quotation of parts based on a large language model, characterized in that, include: Obtain unstructured manufacturing requirement information and 3D part models, and use a large language model to perform semantic parsing on the unstructured manufacturing requirement information to extract key manufacturing process parameters. The system uses a pre-defined knowledge rule base in the manufacturing domain to determine whether the unstructured manufacturing requirement information is complete. If it is incomplete, the large language model generates a clarifying query and interacts with the user in multiple rounds until the unstructured manufacturing requirement information is complete. Then, it outputs a list of structured process instructions with key manufacturing process parameters. After analyzing the three-dimensional part model, a feature recognition algorithm is used to identify the processing features and determine whether there are any features to be confirmed. If there are no features to be confirmed, the processing feature list is obtained directly. If there are features to be confirmed, a human-machine collaborative confirmation mechanism is executed to obtain the processing feature list. The system invokes the preset manufacturing domain knowledge rule base and the preset real-time price database, and calculates the technical cost based on the structured process instruction list with key manufacturing process parameters and the processing feature list. It then interacts with the user through the large language model to obtain non-technical factor information, and combines this with the manufacturing domain knowledge rule base and actual execution feasibility to determine whether the user's demand is reasonable. If unreasonable, it provides a prompt and engages in interactive negotiation to adjust the demand until it becomes reasonable, thus obtaining the non-technical user demand. Based on this non-technical user demand, the non-technical cost is calculated. The total price is obtained by integrating the technical and non-technical costs. The total price and cost composition are then visualized and displayed. After user confirmation, a price list is generated.

2. The intelligent and precise part machining quotation method based on a large language model according to claim 1, characterized in that, The input formats for the unstructured manufacturing requirements information include voice input and text input; The format of the three-dimensional part model is the STEP format, the product model data exchange standard. The key manufacturing process parameters include material type, machining accuracy, surface treatment method, special process requirements, and delivery time; The incompleteness of unstructured manufacturing demand information includes situations where information is missing, ambiguous, or conflicting.

3. The intelligent and precise part machining quotation method based on a large language model according to claim 1, characterized in that, When parsing the three-dimensional part model, a geometric processing kernel based on boundary representation B-Rep is used. The feature recognition algorithm adopts an enhanced feature recognition algorithm; The machining features include holes, grooves, cavities, and chamfers.

4. The intelligent and precise machining quotation method for parts based on a large language model according to claim 1, characterized in that, If there are features to be confirmed, a human-machine collaborative confirmation mechanism is executed, including: Mark the areas of features to be confirmed in the visualization interface of the three-dimensional part model to obtain the areas to be confirmed; For the area to be confirmed, the large language model generates multiple-choice or question-and-answer confirmation prompts to interact with the user and obtain the confirmation result. Based on the confirmation results, the processing feature parameters are corrected to obtain a collaboratively confirmed list of processing features; Output a list of collaboratively confirmed processing features.

5. The intelligent and precise part machining quotation method based on a large language model according to claim 1, characterized in that, The manufacturing domain knowledge rule base includes cutting parameters, time quotas, and material density; The preset real-time price database includes material unit prices and outsourcing processing quotations; The technical costs include material costs, processing costs, heat treatment costs, and surface treatment costs; The non-technical factors include delivery location, logistics method, and expected delivery time; The non-technical costs include logistics fees and expedited processing fees.

6. The intelligent and precise machining quotation method for parts based on a large language model according to claim 1, characterized in that, The visualization includes a cost breakdown chart, and the quotation includes itemized costs, total costs, and delivery dates. The quotation is output in PDF format.

7. A parts intelligent precision machining quotation system based on a large language model, characterized in that, include: The intelligent demand analysis and interactive clarification module is used to acquire unstructured manufacturing demand information and 3D part models. It uses a large language model to perform semantic analysis on the unstructured manufacturing demand information and extract key manufacturing process parameters. It combines a preset manufacturing domain knowledge rule base to determine whether the unstructured manufacturing demand information is complete. If it is incomplete, the large language model generates clarification queries and interacts with the user in multiple rounds until the unstructured manufacturing demand information is complete. Finally, it outputs a list of structured process instructions with key manufacturing process parameters. The feature recognition and human-machine collaborative confirmation module is used to analyze the three-dimensional part model, identify the processing features using a feature recognition algorithm, and determine whether there are any features to be confirmed. If there are no features to be confirmed, the processing feature list is obtained directly. If there are features to be confirmed, the human-machine collaborative confirmation mechanism is executed to obtain the processing feature list. The cost tiered calculation and negotiation module is used to call the preset manufacturing domain knowledge rule base and the preset real-time price database, and calculate the technical cost based on the structured process instruction list with key manufacturing process parameters and the processing feature list; it also uses the large language model to interact with the user to obtain non-technical factor information, and combines the manufacturing domain knowledge rule base and actual execution feasibility to judge whether the user's demand is reasonable. If it is unreasonable, it prompts the unreasonable user demand and negotiates and adjusts it interactively until the user demand is reasonable, thus obtaining the non-technical user demand, and calculating the non-technical cost based on the non-technical user demand. The quotation output module is used to integrate the technical and non-technical costs to obtain a total quotation, visualize the total quotation and cost composition, and output a quotation sheet after user confirmation.

8. The intelligent and precise machining quotation system for parts based on a large language model according to claim 7, characterized in that, Also includes: The user interaction layer is used to provide a 3D model visualization window, voice or text input boxes, intelligent dialogue flow interface, quotation result display panel, receive user input, render models and dialogues, and initiate voice recognition operations. The intelligent service layer is built on a microservice architecture and deployed in the cloud or on a local server to provide services such as large language model interaction, geometry processing, cost calculation, and business logic. The data and knowledge layer is used to store structured data, including material parameters, processing time and equipment rates, using a knowledge rule base in the manufacturing field. By accessing dynamic data, including market material prices and outsourcing quotations, a real-time price database is formed. The project and quotation database stores users' historical projects, model files and quotations. The computing support layer provides the hardware and basic software support required for the operation of the user interaction layer, intelligent service layer, and data and knowledge layer. A voice input device is used to convert input voice into text in real time, and the converted text is transmitted to the large language model for interactive services. A 3D model preview device is used to integrate a part visualization window and highlight specific features; A conversational interactive device used to simulate the interaction format of instant messaging software, presenting a dialogue between a user and an artificial intelligence (AI) assistant.

9. An electronic device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the intelligent and precise machining quotation method for parts based on a large language model as described in any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the intelligent and precise part machining quotation method based on a large language model as described in any one of claims 1-6.

Images