AR and RFID-based mold repair knowledge base auxiliary system

Abstract

This invention discloses an AR and RFID-based mold repair knowledge base auxiliary system, relating to the field of big data analysis. It includes: an AR interaction module for information display and operational interaction; a fault analysis module that combines historical RFID data with on-site symptoms, calling upon cases and statistical information of similar molds in the repair knowledge base to generate a probabilistic fault diagnosis report; the system employs WiFi / 5G transmission technology to achieve bidirectional data synchronization between the glasses and the cloud-based knowledge base, updating mold repair records and statistical databases in real time; and a 3D modeling module that performs lightweight rendering of faulty parts and overlays them onto the real mold, providing visual repair guidance. The advantages of this invention are: the integration of RFID and augmented reality technologies enables accurate mold identification, intelligent fault analysis, visual repair guidance, and iterative knowledge updates, significantly improving the intelligence, efficiency, and standardization of industrial mold repair, and lowering the repair threshold.

Description

Technical Field

[0001] This invention relates to the field of big data analysis, and in particular to a knowledge base auxiliary system for mold repair based on AR and RFID. Background Technology

[0002] As a key tool in industrial production, the precision and lifespan of molds directly affect product quality and production efficiency. However, in the context of intelligent manufacturing transformation, the high costs of complex mold fault diagnosis, standardization of maintenance solutions, and personnel training have become pain points in the industry. With the development of industrial big data and artificial intelligence technologies, building a digital knowledge base system has become a solution—by integrating structured knowledge such as historical maintenance data, standard operating procedures, and typical fault cases, combined with intelligent retrieval, fault tree analysis, and expert system technologies, it enables the digital accumulation and rapid retrieval of maintenance experience.

[0003] In mold production and maintenance scenarios, mold repair is a crucial link in ensuring production continuity. However, traditional mold repair methods suffer from several technical bottlenecks: Fault diagnosis relies on experience, leading to low efficiency: Traditional repairs depend on the personal experience of repair personnel to determine the cause of mold failures, lacking standardized fault analysis criteria. For complex faults or new molds, misdiagnosis and omission are prone to occur, resulting in lengthy fault diagnosis times. Inconvenient access to repair knowledge and poor relevance: Mold repair knowledge bases are mostly in the form of paper manuals or standalone software. Repair personnel need to manually search for relevant fault solutions, which cannot be accurately linked to the currently repaired mold. Furthermore, knowledge updates are lagging, making it difficult to quickly obtain suitable repair guidance. Lack of intuitive repair instructions and high skill threshold: Existing repair instructions are mostly text descriptions or two-dimensional drawings, which do not clearly illustrate the internal structure of the mold and the installation positions of components (such as springs, bolts, and other small parts). Novice repair personnel find it difficult to quickly locate the repair parts, leading to a high error rate in repair operations. Lack of quantifiable fault references leads to highly arbitrary troubleshooting: Current solutions can only provide single or scattered fault suggestions, failing to provide a quantifiable fault probability ranking based on multi-dimensional data. Maintenance personnel must check each suspected fault individually, further extending fault location time. Long maintenance cycles and high production costs: Due to slow fault diagnosis and unfamiliarity with maintenance operations, the average maintenance time for traditional molds is as long as 1.5 hours, severely impacting production progress. Furthermore, operational errors may cause secondary damage, further increasing maintenance costs. While some current solutions introduce RFID for mold identification or AR technology for simple visualization, they fail to deeply integrate these with the maintenance knowledge base. They cannot achieve integrated maintenance assistance of "automatic fault association - accurate knowledge push - intuitive 3D guidance," thus failing to address the core pain points of traditional maintenance models. Summary of the Invention

[0004] To improve the existing system, an AR and RFID-based mold repair knowledge base auxiliary system is provided. This system integrates radio frequency identification and augmented reality technologies to achieve accurate mold identification, intelligent fault analysis, visual repair guidance, and iterative knowledge updates, which will greatly improve the intelligence, efficiency, and standardization of industrial mold repair and lower the repair threshold.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: An AR and RFID-based mold repair knowledge base auxiliary system includes: RFID identification module: Integrated into AR glasses, it includes an RFID reader and a signal receiving unit, used to read all the data stored in the RFID tag on the mold to be repaired and transmit it to the fault analysis module; AR Interaction Module: Using AR glasses as the hardware carrier, it has three core functions: data input, virtual information display, and real-time interaction. Fault Analysis Module: Receives historical mold data transmitted by the RFID identification module and fault symptoms entered by the AR interaction module, retrieves fault statistics data of the same type of mold, and obtains a fault probability ranking list marked with probability values ​​and key judgment criteria through quantitative calculation. Maintenance Knowledge Base Module: Based on a distributed database architecture, it stores basic information on all types of molds, fault cases, maintenance specifications, data on vulnerable parts, and statistics on faults of similar molds, and supports online updates and iterations of knowledge data; Data transmission and storage module: It adopts WiFi / 5G wireless communication technology to transmit all data between AR glasses and maintenance knowledge base in real time, and updates historical fault records in mold RFID tags; 3D modeling and rendering module: Receives pushed maintenance requests, performs lightweight processing and real-time rendering of the corresponding fault 3D model, and overlays the 3D guide model onto the corresponding physical position of the real mold. Similar mold data statistics module: Collects fault records and maintenance data of all networked molds in real time, calculates the probability of occurrence of various faults in similar molds according to mold model, usage scenario and running time, and synchronizes the statistical results to the fault analysis module in real time.

[0006] Preferably, the RFID identification module specifically includes: Reader unit: Equipped with a dedicated UHF RFID chip and a miniature ceramic patch antenna, it supports non-contact close-range scanning and accurately locates the RFID tags of the mold to be repaired in a multi-tag environment; Radio frequency signal receiving unit: integrates a low-noise amplifier and a bandpass filter circuit to filter out electromagnetic interference signals in the industrial environment and improve the sensitivity of radio frequency signal reception; Data verification and transmission unit: Based on the basic data verification logic, it performs integrity verification on the read tag data. After the verification is successful, it seamlessly connects with the AR glasses main control board and transmits the data to the fault analysis module in real time.

[0007] Preferably, the AR interaction module specifically includes: Voice interaction unit: Equipped with a dual-microphone array pickup module, integrating industrial-grade environmental noise reduction circuitry, and simultaneously broadcasting operation confirmation information; Gesture recognition unit: A miniature infrared motion sensor is deployed on the frame of the AR glasses to recognize basic hand gestures of maintenance personnel. Combined with a near-field motion capture algorithm, it enables contactless operation and is suitable for work scenarios where maintenance personnel wear gloves. Virtual display unit: It adopts a micro OLED high-definition display screen and a freeform optical projection module to form a high-definition virtual display area in the glasses window, which supports the superimposed display of 3D maintenance guidance and fault ranking information; Interactive command parsing unit: It interfaces with the voice and gesture recognition unit to collect data, parse commands and transmit them to the fault analysis, 3D modeling and rendering modules. At the same time, it receives feedback data from external modules and distributes it to the virtual display unit for visualization output.

[0008] Preferably, the fault analysis module specifically includes: Multi-source data access unit: Set up multi-channel standardized data interfaces to connect to RFID identification, AR interaction, and similar mold data statistics modules respectively, and perform parallel access to mold historical data, current fault symptoms, and similar fault statistics data; Data preprocessing unit: Cleans and standardizes the heterogeneous data received, removes invalid and abnormal data, extracts core fault feature factors, and matches them with the standard range of equipment parameters in the maintenance knowledge base; Fault probability quantification calculation unit: It has a built-in preset weighted calculation logic and fault feature matching engine. Based on multi-dimensional weights, it calculates the probability of various suspected faults and generates a fault probability ranking list with labeled probability values ​​and key judgment criteria. Fault knowledge matching output unit: Combines the maintenance knowledge base, matches the cause analysis, solutions and 3D model data corresponding to each fault in the ranking list, and pushes them to the AR interaction module after encapsulation and integration.

[0009] Preferably, the maintenance knowledge base module specifically includes: Multi-dimensional data storage unit: Independent sub-databases are split according to knowledge type to store basic mold information, fault cases, maintenance specifications, information on vulnerable parts, and statistical data of similar faults. Data sharding technology is used to build a multi-dimensional index. Intelligent retrieval and matching unit: It has a built-in fault feature association engine, establishes a data link with the fault analysis module, and retrieves and matches maintenance knowledge based on mold ID, fault symptoms, and equipment parameter deviations; Knowledge Iteration and Update Unit: Supports both automatic and manual update modes. It automatically receives maintenance records to complete knowledge supplementation, and manually supports administrators to enter new molds and new fault cases. It also supports OTA online synchronization updates. Access control and approval unit: Set hierarchical operation permissions. Ordinary terminals only have knowledge reading permissions, while administrators can approve and verify new knowledge and optimization solutions, and record all knowledge update operation logs.

[0010] Preferably, the data transmission and storage module specifically includes: Wireless communication transmission unit: Equipped with WiFi / 5G dual-mode communication module, integrated with anti-electromagnetic interference circuit, supports bidirectional data transmission, and connects to each module of AR glasses and cloud server respectively; The terminal-cloud hierarchical storage unit is divided into two layers of storage: terminal and cloud. The AR glasses terminal uses local temporary storage to cache real-time interactive data, while the cloud uses a distributed storage architecture for persistent storage, storing all data in categories such as fault information, maintenance records, and mold parameters. Data synchronization and update unit: Establishes a data link with RFID tags, maintenance knowledge base, and similar mold statistics module. After maintenance is completed, it automatically synchronizes the data to the mold RFID tag update history record, and updates the similar fault statistics and maintenance knowledge base. Transmission verification unit: Performs MD5 verification on various types of transmitted data and system upgrade files, compares the characteristic values ​​of the data at both ends, and triggers a retransmission command if the verification fails.

[0011] Preferably, the 3D modeling and rendering module specifically includes: Mold 3D Modeling Unit: Constructs 3D models of all types of molds, breaks them down to the smallest repairable component level, restores the component structure, installation position, and connection relationship, and stores the dimensions and assembly parameter association information of each part of the mold to form a standardized mold 3D model library; Lightweight Model Processing Unit: Based on the computing power and display characteristics of AR glasses terminals, it performs face reduction, texture compression, and geometric optimization on the completed 3D model; Real-time rendering and computing unit: Receives maintenance guidance requests from the fault analysis module, calls the corresponding mold 3D model, overlays disassembly sequence arrows, component annotations, tool parameters, torque prompts and maintenance guidance information to generate a targeted 3D guidance model; Spatial positioning overlay unit: used to calibrate the spatial coordinates of the virtual 3D guide model and the real mold, and to overlay the model at the physical location of the mold.

[0012] Preferably, the similar mold data statistics module specifically includes: Multi-source data acquisition unit: Connects to the data transmission and storage module to collect all networked molds' fault records, maintenance results, runtime, and usage scenarios in real time, and removes duplicate, invalid, and abnormal data; Mold Classification Clustering Unit: Establish a multi-dimensional classification system to cluster the collected mold data according to mold type, production process, usage scenario, and running time; Fault data statistical analysis unit: Statistically analyzes the fault types, frequency of occurrence, and causes of faults in similar molds, calculates the probability and proportion of each type of fault in the corresponding molds, forms a standardized statistical data set, and marks the statistical sample size and statistical time dimension. Statistical data output synchronization unit: Establishes a data link with the fault analysis module, pushes statistical data sets to the fault analysis module, supports dynamic data updates, and automatically iterates and updates the statistical results when new mold repair data is entered.

[0013] Compared with the prior art, the advantages of the present invention are: This AR and RFID-based mold repair knowledge base auxiliary system integrates RFID, augmented reality, big data analysis, and 3D modeling technologies to construct an industrial-grade repair support system that combines accurate identification, intelligent analysis, visual guidance, and knowledge iteration, offering significant advantages. The system uses an UHF RFID module to accurately lock molds in multi-tag environments, filtering out industrial electromagnetic interference and ensuring stable data reading and transmission. The AR interaction module adapts to gloved work scenarios, using voice and gesture contactless interaction combined with high-definition virtual display to present repair information in a real-world context. The fault analysis module relies on multi-source data fusion and quantitative calculation to accurately generate fault probability rankings and match optimal solutions. Combined with lightweight 3D modeling and spatial positioning rendering, it transforms abstract processes into intuitive visual guidance. The repair knowledge base supports automatic iteration and hierarchical management. The similar mold statistics module enables dynamic analysis of fault big data, and edge-cloud hierarchical storage and security verification ensure data reliability. The overall system breaks through the limitations of traditional repair relying on human experience, lowering the repair threshold and error rate, significantly improving the efficiency, standardization, and intelligence of mold repair, and is highly adaptable to complex industrial operation scenarios. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the system proposed in this invention; Figure 2 This is a diagram of the RFID identification module proposed in this invention; Figure 3 This is a diagram of the AR interaction module proposed in this invention; Figure 4 This is a diagram of the fault analysis module proposed in this invention; Figure 5 This is a module diagram of the maintenance knowledge base proposed in this invention; Figure 6 This is a diagram of the data transmission and storage module proposed in this invention; Figure 7 This is a diagram of the 3D modeling and rendering module proposed in this invention; Figure 8 This is a diagram of a data statistics module for similar molds proposed in this invention. Detailed Implementation

[0015] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0016] See Figure 1 As shown, the AR and RFID-based mold repair knowledge base auxiliary system includes: RFID identification module: Integrated into AR glasses, it includes an RFID reader and a signal receiving unit, used to read all the data stored in the RFID tag on the mold to be repaired and transmit it to the fault analysis module; AR Interaction Module: Using AR glasses as the hardware carrier, it has three core functions: data input, virtual information display, and real-time interaction. Fault Analysis Module: Receives historical mold data transmitted by the RFID identification module and fault symptoms entered by the AR interaction module, retrieves fault statistics data of the same type of mold, and obtains a fault probability ranking list marked with probability values ​​and key judgment criteria through quantitative calculation. Maintenance Knowledge Base Module: Based on a distributed database architecture, it stores basic information on all types of molds, fault cases, maintenance specifications, data on vulnerable parts, and statistics on faults of similar molds, and supports online updates and iterations of knowledge data; Data transmission and storage module: It adopts WiFi / 5G wireless communication technology to transmit all data between AR glasses and maintenance knowledge base in real time, and updates historical fault records in mold RFID tags; 3D modeling and rendering module: Receives pushed maintenance requests, performs lightweight processing and real-time rendering of the corresponding fault 3D model, and overlays the 3D guide model onto the corresponding physical position of the real mold. Similar mold data statistics module: Collects fault records and maintenance data of all networked molds in real time, calculates the probability of occurrence of various faults in similar molds according to mold model, usage scenario and running time, and synchronizes the statistical results to the fault analysis module in real time.

[0017] See Figure 2 As shown, the RFID identification module specifically includes: Reader unit: Equipped with a dedicated UHF RFID chip and a miniature ceramic patch antenna, it supports non-contact close-range scanning and accurately locates the RFID tags of the mold to be repaired in a multi-tag environment; Radio frequency signal receiving unit: integrates a low-noise amplifier and a bandpass filter circuit to filter out electromagnetic interference signals in the industrial environment and improve the sensitivity of radio frequency signal reception; Data verification and transmission unit: Based on the basic data verification logic, it performs integrity verification on the read tag data. After the verification is successful, it seamlessly connects with the AR glasses main control board and transmits the data to the fault analysis module in real time.

[0018] Specifically, based on CRC32 cyclic redundancy check logic, the tag data is verified from three dimensions: data length, field format, and feature code matching. If an anomaly is found, a reread instruction is immediately triggered. After the verification is passed, the AR glasses are connected via the SPI serial communication interface. The unit has a built-in data cache module that can temporarily store tag data to avoid data loss due to communication interruption. The data will be automatically resent after communication is restored.

[0019] See Figure 3 As shown, the AR interaction module specifically includes: Voice interaction unit: Equipped with a dual-microphone array pickup module, integrating industrial-grade environmental noise reduction circuitry, and simultaneously broadcasting operation confirmation information; Gesture recognition unit: A miniature infrared motion sensor is deployed on the frame of the AR glasses to recognize basic hand gestures of maintenance personnel. Combined with a near-field motion capture algorithm, it enables contactless operation and is suitable for work scenarios where maintenance personnel wear gloves. Virtual display unit: It adopts a micro OLED high-definition display screen and a freeform optical projection module to form a high-definition virtual display area in the glasses window, which supports the superimposed display of 3D maintenance guidance and fault ranking information; Interactive command parsing unit: It interfaces with the voice and gesture recognition unit to collect data, parse commands and transmit them to the fault analysis, 3D modeling and rendering modules. At the same time, it receives feedback data from external modules and distributes it to the virtual display unit for visualization output.

[0020] Specifically, the voice interaction unit integrates an environmental noise reduction circuit and adopts adaptive noise suppression technology, which can effectively filter out environmental noise such as motor operation and equipment noise in the workshop; at the same time, it supports real-time voice broadcast function, synchronously feeding back operation confirmation information, fault prompts, maintenance step guidance and other content. The broadcast volume can be automatically adjusted according to the environmental noise to avoid information omission due to noisy environment, and is suitable for maintenance scenarios such as one-handed operation and two-handed tool holding. To address the challenges faced by maintenance personnel who wear gloves and have oily hands, a miniature infrared motion sensor is deployed inside the AR glasses frame to accurately capture basic hand gestures. Combined with a near-field motion capture algorithm, the accuracy of gesture recognition is optimized for gloves, supporting four core gestures: tap, swipe, fist, and wave, corresponding to command confirmation, page switching, pause operation, and return to the homepage, respectively. This enables completely contactless operation, preventing oil and dust from contaminating the equipment. It adopts a micro-OLED high-definition display and a freeform optical projection module to form a virtual display area in the glasses window; it supports the superimposed display of information such as 3D maintenance guidance, fault probability ranking, and mold parameters. The virtual image is precisely aligned with the real mold scene, and maintenance personnel can intuitively view the component disassembly sequence, fault location markings, etc., without having to frequently look down to check the terminal equipment. It connects to the voice and gesture recognition unit to collect data, and quickly parses the operation intention through command feature matching technology. The parsed command is then transmitted to the fault analysis, 3D modeling and rendering module in real time. At the same time, it receives feedback data from various external modules and allocates it to the virtual display unit for visualization output according to priority.

[0021] See Figure 4 As shown, the fault analysis module specifically includes: Multi-source data access unit: Set up multi-channel standardized data interfaces to connect to RFID identification, AR interaction, and similar mold data statistics modules respectively, and perform parallel access to mold historical data, current fault symptoms, and similar fault statistics data; Data preprocessing unit: Cleans and standardizes the heterogeneous data received, removes invalid and abnormal data, extracts core fault feature factors, and matches them with the standard range of equipment parameters in the maintenance knowledge base; Fault probability quantification calculation unit: It has a built-in preset weighted calculation logic and fault feature matching engine. Based on multi-dimensional weights, it calculates the probability of various suspected faults and generates a fault probability ranking list with labeled probability values ​​and key judgment criteria. Fault knowledge matching output unit: Combines the maintenance knowledge base, matches the cause analysis, solutions and 3D model data corresponding to each fault in the ranking list, and pushes them to the AR interaction module after encapsulation and integration.

[0022] Specifically, the core of the fault probability quantification calculation unit is to prioritize faults. It has a built-in preset weighted calculation logic and fault feature matching engine. The weight dimensions cover four categories: historical fault frequency of the mold, probability of similar mold faults, current fault symptom matching degree, and mold running time. The probability value of each type of suspected fault is obtained through comprehensive weighted calculation. After the calculation is completed, a fault probability ranking list is generated. Each fault item is clearly marked with a probability value and key judgment criteria. For example, "Mold demolding abnormality (probability 72%), based on the fact that this fault accounts for 38% of similar molds, this mold has had 5 similar faults in the past, and the current symptom matches the knowledge base cases 85%", which intuitively presents the fault priority. The analysis results are linked with maintenance knowledge and connected to the maintenance knowledge base module in real time. For each type of fault in the fault ranking, the corresponding fault cause analysis, step-by-step maintenance solutions, precautions and associated 3D model data are accurately matched. After this information is encapsulated and integrated, it is pushed to the AR interaction module according to the fault probability priority, so that maintenance personnel can obtain fault judgment basis and maintenance guidance at the same time.

[0023] See Figure 5 As shown, the maintenance knowledge base module specifically includes: Multi-dimensional data storage unit: Independent sub-databases are split according to knowledge type to store basic mold information, fault cases, maintenance specifications, information on vulnerable parts, and statistical data of similar faults. Data sharding technology is used to build a multi-dimensional index. Intelligent retrieval and matching unit: It has a built-in fault feature association engine, establishes a data link with the fault analysis module, and retrieves and matches maintenance knowledge based on mold ID, fault symptoms, and equipment parameter deviations; Knowledge Iteration and Update Unit: Supports both automatic and manual update modes. It automatically receives maintenance records to complete knowledge supplementation, and manually supports administrators to enter new molds and new fault cases. It also supports OTA online synchronization updates. Access control and approval unit: Set hierarchical operation permissions. Ordinary terminals only have knowledge reading permissions, while administrators can approve and verify new knowledge and optimization solutions, and record all knowledge update operation logs.

[0024] Specifically, a categorized storage architecture is adopted, splitting the data into five independent sub-databases based on knowledge type. These sub-databases correspond to basic mold information, fault cases, maintenance specifications, information on vulnerable parts, and statistical data on similar faults. Each sub-database uses data sharding technology and establishes multi-dimensional indexes based on mold model, fault type, and usage scenario to improve data retrieval speed. The basic mold information sub-database stores static data such as mold model, size parameters, assembly drawings, and material specifications. The fault case sub-database stores complete fault scenarios, causes, solutions, and maintenance effect feedback. The vulnerable parts sub-database records vulnerable part models, replacement cycles, compatible molds, and supplier information. All data is stored in a distributed manner, deployed across multiple nodes, and has fault tolerance capabilities. The failure of a single node does not affect overall data access, and the storage capacity can be dynamically expanded according to needs. The intelligent retrieval and matching unit has a built-in fault feature association engine that establishes a real-time data link with the fault analysis module, supporting multi-dimensional retrieval and matching; it can accurately match the full knowledge of the corresponding mold based on the mold ID; it can perform feature association based on the fault symptoms entered by maintenance personnel through the AR interaction module, and fuzzy match similar fault cases and maintenance solutions; it can also support retrieval based on conditions such as equipment parameter deviation and fault location.

[0025] See Figure 6 As shown, the data transmission and storage module specifically includes: Wireless communication transmission unit: Equipped with WiFi / 5G dual-mode communication module, integrated with anti-electromagnetic interference circuit, supports bidirectional data transmission, and connects to each module of AR glasses and cloud server respectively; The terminal-cloud hierarchical storage unit is divided into two layers of storage: terminal and cloud. The AR glasses terminal uses local temporary storage to cache real-time interactive data, while the cloud uses a distributed storage architecture for persistent storage, storing all data in categories such as fault information, maintenance records, and mold parameters. Data synchronization and update unit: Establishes a data link with RFID tags, maintenance knowledge base, and similar mold statistics module. After maintenance is completed, it automatically synchronizes the data to the mold RFID tag update history record, and updates the similar fault statistics and maintenance knowledge base. Transmission verification unit: Performs MD5 verification on various types of transmitted data and system upgrade files, compares the characteristic values ​​of the data at both ends, and triggers a retransmission command if the verification fails.

[0026] Specifically, it is equipped with a WiFi / 5G dual-mode communication module, which automatically switches to the optimal communication link according to the workshop network environment to ensure continuous connection during mobile operations; it has a built-in dedicated anti-electromagnetic interference circuit, which resists electromagnetic interference from workshop motors, frequency converters and other equipment through shielding and filtering design, ensuring uninterrupted and packet-free communication in complex industrial environments. A dual-layer architecture of terminal temporary caching and cloud persistent storage is adopted to achieve efficient data management. The AR glasses are equipped with a high-speed flash memory chip as a local temporary storage unit to cache fault analysis results, real-time interactive data, and lightweight 3D models, ensuring the normal use of basic functions in the event of network outage. The cloud adopts a distributed storage architecture to classify and persistently store mold basic parameters, historical fault records, complete repair cases, and statistical data of similar molds. Data loss is avoided through multi-node redundant backup, and the storage capacity can be dynamically expanded as the number of molds and repair data increases to meet the data retention needs throughout the entire life cycle.

[0027] See Figure 7 As shown, the 3D modeling and rendering module specifically includes: Mold 3D Modeling Unit: Constructs 3D models of all types of molds, breaks them down to the smallest repairable component level, restores the component structure, installation position, and connection relationship, and stores the dimensions and assembly parameter association information of each part of the mold to form a standardized mold 3D model library; Lightweight Model Processing Unit: Based on the computing power and display characteristics of AR glasses terminals, it performs face reduction, texture compression, and geometric optimization on the completed 3D model; Real-time rendering and computing unit: Receives maintenance guidance requests from the fault analysis module, calls the corresponding mold 3D model, overlays disassembly sequence arrows, component annotations, tool parameters, torque prompts and maintenance guidance information to generate a targeted 3D guidance model; Spatial positioning overlay unit: used to calibrate the spatial coordinates of the virtual 3D guide model and the real mold, and to overlay the model at the physical location of the mold.

[0028] Specifically, a standardized 3D model library for all types of molds is constructed. The modeling process strictly follows the mold design drawings and physical scanning data to restore the overall structure and detailed features of the mold. During modeling, the mold is broken down to the smallest repairable component level, and each component is modeled separately and labeled with component name, size parameters, and assembly tolerances. At the same time, the installation position, connection relationship and movement trajectory of each component are accurately restored. After the modeling is completed, it is classified and archived according to mold model and component type to form a standardized model library that can be quickly accessed. To address the issues of limited computing power and insufficient display resources in AR glasses terminals, the completed 3D models are specifically optimized. Through face reduction technology, redundant and repetitive faces with no visual impact are eliminated from the model, reducing the face count of complex mold models to less than 30% of the original number, while retaining the detailed features of key repair parts without affecting the accuracy of repair guidance. Through geometric optimization algorithms, the model's topology is simplified, reducing the computing power consumption during rendering. The spatial positioning overlay unit achieves precise alignment between the virtual model and the real mold. It integrates the spatial positioning sensor and image recognition algorithm built into the AR glasses to capture the spatial coordinates and contour features of the real mold in real time and dynamically calibrate them with the coordinates of the virtual 3D model. This ensures that the virtual model is accurately overlaid on the corresponding physical position of the real mold. It also has a dynamic tracking function. When maintenance personnel move the work or adjust the position of the mold, the virtual model can follow synchronously and always maintain precise alignment with the real object, avoiding maintenance errors caused by positional deviation.

[0029] See Figure 8 As shown, the data statistics module for similar molds specifically includes: Multi-source data acquisition unit: Connects to the data transmission and storage module to collect all networked molds' fault records, maintenance results, runtime, and usage scenarios in real time, and removes duplicate, invalid, and abnormal data; Mold Classification Clustering Unit: Establish a multi-dimensional classification system to cluster the collected mold data according to mold type, production process, usage scenario, and running time; Fault data statistical analysis unit: Statistically analyzes the fault types, frequency of occurrence, and causes of faults in similar molds, calculates the probability and proportion of each type of fault in the corresponding molds, forms a standardized statistical data set, and marks the statistical sample size and statistical time dimension. Statistical data output synchronization unit: Establishes a data link with the fault analysis module, pushes statistical data sets to the fault analysis module, supports dynamic data updates, and automatically iterates and updates the statistical results when new mold repair data is entered.

[0030] Specifically, the mold classification and clustering unit constructs a multi-dimensional and refined mold classification system. According to key dimensions such as mold model, production process type, workshop usage scenario, and equipment running time, the collected mold data from the entire domain is standardized and clustered. Molds with the same structure, similar working conditions, and similar usage cycles are grouped into the same statistical group to avoid data confusion between molds of different types and working conditions. The classification rules support dynamic expansion, and newly added networked molds can be automatically matched to the corresponding group without manual classification, which greatly improves statistical efficiency. Specialized fault statistics are conducted for each cluster group. Information such as fault types, frequency of occurrence, main causes, and maintenance cycles of similar molds are summarized and sorted to form a standardized statistical data set. The statistical results clearly indicate the probability and proportion of each type of fault in the corresponding group, and also include key explanations such as the total number of statistical samples, statistical time period, and operating condition constraints, so as to intuitively reflect the fault distribution pattern under different usage conditions.

[0031] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0032] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0033] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A mold repair knowledge base auxiliary system based on AR and RFID, characterized in that, include: RFID identification module: Integrated into AR glasses, it includes an RFID reader and a signal receiving unit, used to read all the data stored in the RFID tag on the mold to be repaired and transmit it to the fault analysis module; AR Interaction Module: Using AR glasses as the hardware carrier, it has three core functions: data input, virtual information display, and real-time interaction. Fault Analysis Module: Receives historical mold data transmitted by the RFID identification module and fault symptoms entered by the AR interaction module, retrieves fault statistics data of the same type of mold, and obtains a fault probability ranking list marked with probability values ​​and key judgment criteria through quantitative calculation. Maintenance Knowledge Base Module: Based on a distributed database architecture, it stores basic information on all types of molds, fault cases, maintenance specifications, data on vulnerable parts, and statistics on faults of similar molds, and supports online updates and iterations of knowledge data; Data transmission and storage module: It adopts WiFi / 5G wireless communication technology to transmit all data between AR glasses and maintenance knowledge base in real time, and updates historical fault records in mold RFID tags; 3D modeling and rendering module: Receives pushed maintenance requests, performs lightweight processing and real-time rendering of the corresponding fault 3D model, and overlays the 3D guide model onto the corresponding physical position of the real mold. Similar mold data statistics module: Collects fault records and maintenance data of all networked molds in real time, calculates the probability of occurrence of various faults in similar molds according to mold model, usage scenario and running time, and synchronizes the statistical results to the fault analysis module in real time.

2. The AR and RFID-based mold repair knowledge base auxiliary system according to claim 1, characterized in that, The RFID identification module specifically includes: Reader unit: Equipped with a dedicated UHF RFID chip and a miniature ceramic patch antenna, it supports non-contact close-range scanning and accurately locates the RFID tags of the mold to be repaired in a multi-tag environment; Radio frequency signal receiving unit: integrates a low-noise amplifier and a bandpass filter circuit to filter out electromagnetic interference signals in the industrial environment and improve the sensitivity of radio frequency signal reception; Data verification and transmission unit: Based on the basic data verification logic, it performs integrity verification on the read tag data. After the verification is successful, it seamlessly connects with the AR glasses main control board and transmits the data to the fault analysis module in real time.

3. The AR and RFID-based mold repair knowledge base auxiliary system according to claim 1, characterized in that, The AR interaction module specifically includes: Voice interaction unit: Equipped with a dual-microphone array pickup module, integrating industrial-grade environmental noise reduction circuitry, and simultaneously broadcasting operation confirmation information; Gesture recognition unit: A miniature infrared motion sensor is deployed on the frame of the AR glasses to recognize basic hand gestures of maintenance personnel. Combined with a near-field motion capture algorithm, it enables contactless operation and is suitable for work scenarios where maintenance personnel wear gloves. Virtual display unit: It adopts a micro OLED high-definition display screen and a freeform optical projection module to form a high-definition virtual display area in the glasses window, which supports the superimposed display of 3D maintenance guidance and fault ranking information; Interactive command parsing unit: It interfaces with the voice and gesture recognition unit to collect data, parse commands and transmit them to the fault analysis, 3D modeling and rendering modules. At the same time, it receives feedback data from external modules and distributes it to the virtual display unit for visualization output.

4. The AR and RFID-based mold repair knowledge base auxiliary system according to claim 1, characterized in that, The fault analysis module specifically includes: Multi-source data access unit: Set up multi-channel standardized data interfaces to connect to RFID identification, AR interaction, and similar mold data statistics modules respectively, and perform parallel access to mold historical data, current fault symptoms, and similar fault statistics data; Data preprocessing unit: Cleans and standardizes the heterogeneous data received, removes invalid and abnormal data, extracts core fault feature factors, and matches them with the standard range of equipment parameters in the maintenance knowledge base; Fault probability quantification calculation unit: It has a built-in preset weighted calculation logic and fault feature matching engine. Based on multi-dimensional weights, it calculates the probability of various suspected faults and generates a fault probability ranking list with labeled probability values ​​and key judgment criteria. Fault knowledge matching output unit: Combines the maintenance knowledge base, matches the cause analysis, solutions and 3D model data corresponding to each fault in the ranking list, and pushes them to the AR interaction module after encapsulation and integration.

5. The AR and RFID-based mold repair knowledge base auxiliary system according to claim 1, characterized in that, The maintenance knowledge base module specifically includes: Multi-dimensional data storage unit: Independent sub-databases are split according to knowledge type to store basic mold information, fault cases, maintenance specifications, information on vulnerable parts, and statistical data of similar faults. Data sharding technology is used to build a multi-dimensional index. Intelligent retrieval and matching unit: It has a built-in fault feature association engine, establishes a data link with the fault analysis module, and retrieves and matches maintenance knowledge based on mold ID, fault symptoms, and equipment parameter deviations; Knowledge Iteration and Update Unit: Supports both automatic and manual update modes. It automatically receives maintenance records to complete knowledge supplementation, and manually supports administrators to enter new molds and new fault cases. It also supports OTA online synchronization updates. Access control and approval unit: Set hierarchical operation permissions. Ordinary terminals only have knowledge reading permissions, while administrators can approve and verify new knowledge and optimization solutions, and record all knowledge update operation logs.

6. The AR and RFID-based mold repair knowledge base auxiliary system according to claim 1, characterized in that, The data transmission and storage module specifically includes: Wireless communication transmission unit: Equipped with WiFi / 5G dual-mode communication module, integrated with anti-electromagnetic interference circuit, supports bidirectional data transmission, and connects to each module of AR glasses and cloud server respectively; The terminal-cloud hierarchical storage unit is divided into two layers of storage: terminal and cloud. The AR glasses terminal uses local temporary storage to cache real-time interactive data, while the cloud uses a distributed storage architecture for persistent storage, storing all data in categories such as fault information, maintenance records, and mold parameters. Data synchronization and update unit: Establishes a data link with RFID tags, maintenance knowledge base, and similar mold statistics module. After maintenance is completed, it automatically synchronizes the data to the mold RFID tag update history record, and updates the similar fault statistics and maintenance knowledge base. Transmission verification unit: Performs MD5 verification on various types of transmitted data and system upgrade files, compares the characteristic values ​​of the data at both ends, and triggers a retransmission command if the verification fails.

7. The AR and RFID-based mold repair knowledge base auxiliary system according to claim 1, characterized in that, The 3D modeling and rendering module specifically includes: Mold 3D Modeling Unit: Constructs 3D models of all types of molds, breaks them down to the smallest repairable component level, restores the component structure, installation position, and connection relationship, and stores the dimensions and assembly parameter association information of each part of the mold to form a standardized mold 3D model library; Lightweight Model Processing Unit: Based on the computing power and display characteristics of AR glasses terminals, it performs face reduction, texture compression, and geometric optimization on the completed 3D model; Real-time rendering and computing unit: Receives maintenance guidance requests from the fault analysis module, calls the corresponding mold 3D model, overlays disassembly sequence arrows, component annotations, tool parameters, torque prompts and maintenance guidance information to generate a targeted 3D guidance model; Spatial positioning overlay unit: used to calibrate the spatial coordinates of the virtual 3D guide model and the real mold, and to overlay the model at the physical location of the mold.

8. The AR and RFID-based mold repair knowledge base auxiliary system according to claim 1, characterized in that, The aforementioned similar mold data statistics module specifically includes: Multi-source data acquisition unit: Connects to the data transmission and storage module to collect all networked molds' fault records, maintenance results, runtime, and usage scenarios in real time, and removes duplicate, invalid, and abnormal data; Mold Classification Clustering Unit: Establish a multi-dimensional classification system to cluster the collected mold data according to mold type, production process, usage scenario, and running time; Fault data statistical analysis unit: Statistically analyzes the fault types, frequency of occurrence, and causes of faults in similar molds, calculates the probability and proportion of each type of fault in the corresponding molds, forms a standardized statistical data set, and marks the statistical sample size and statistical time dimension. Statistical data output synchronization unit: Establishes a data link with the fault analysis module, pushes statistical data sets to the fault analysis module, supports dynamic data updates, and automatically iterates and updates the statistical results when new mold repair data is entered.

Images