Ceramic supply chain quality traceability data management system based on block chain

Abstract

The invention relates to the technical field of data processing, in particular to a ceramic supply chain quality traceability data management system based on a block chain, which comprises a data acquisition and processing module, a dynamic identifier generation module, an identifier embedding control module, a cross-chain data synchronization module, a multi-modal verification module and a redundancy recovery module. The data acquisition and processing module integrates firing temperature, forming pressure parameters and microscopic image data through Internet of Things equipment and edge computing nodes to generate a structured data set, the system integrated resource scheduling module optimizes node computing resource allocation, and anti-quantum encryption and IPFS distributed storage guarantee data security. And the terminal equipment integration module simplifies the multi-modal data acquisition process, and finally collaborative optimization of the ceramic quality tracing reliability, the operation efficiency and the artistic integrity is realized.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a ceramic supply chain quality traceability data management system based on blockchain. Background Art

[0002] In the production process of ceramic products, the preparation data and image information of each processing link are encrypted and stored through blockchain technology, and product traceability is achieved through blockchain identification attached to the back of ceramic packaging or products. However, the physical carrier of blockchain identification in the prior art has significant defects: when the identification is set on the ceramic packaging, damage to the packaging during transportation and storage will directly lead to the loss of traceability information; if the identification is placed on the back of the ceramic product, the user needs to flip the product to read it, which is not convenient to operate; and placing the identification on the surface of the ceramic product improves readability, but it will destroy the integrity and artistic beauty of the ceramic surface. The above technical defects reduce the reliability of the ceramic supply chain traceability system and affect the user's efficient acquisition of traceability information, making it difficult to meet the dual requirements of the modern ceramic industry for quality traceability and user experience. Summary of the invention

[0003] In view of the shortcomings of the existing technology, the present invention provides a ceramic supply chain quality traceability data management system based on blockchain, which is used to solve the technical problems of low reliability of blockchain identification physical carriers in the existing ceramic supply chain, insufficient operational convenience and difficulty in balancing the surface integrity of ceramics.

[0004] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows: The ceramic supply chain quality traceability data management system based on blockchain provided by the present invention includes: The data acquisition and processing module is deployed in the ceramic production line, and is used to collect process parameters and image data through IoT devices, perform format conversion and outlier cleaning through edge computing nodes to generate a structured data set, and transmit the structured data set to the dynamic identification generation module; A dynamic identification generation module receives the structured data set, is used to perform a hash operation on the structured data set to generate a hash summary, combines the production batch number, the timestamp and the physical characteristics of the ceramic to generate a blockchain identification of a variable length, and outputs the characteristic parameters of the blockchain identification to the identification embedding control module; The logo embedding control module receives the characteristic parameters of the blockchain logo, and is used to convert the dot density and code length of the blockchain logo into the pulse frequency and engraving depth parameters of the laser device, drive the laser micro-engraving module to engrave invisible dots inside the ceramic body or in the transparent coating, and collect the engraving quality data through the optical detection device and feed it back to the blockchain node; A cross-chain data synchronization module receives the engraving quality data and ceramic production data, is used to write the production data into the private chain and synchronize it to the alliance chain node through the cross-chain protocol, triggers the spatiotemporal data verification logic of the smart contract based on the logistics path, and freezes the batch data update permission when the logistics track deviates from the preset path; The multimodal verification module is deployed on the user terminal to parse the hash value of the ceramic identification and initiate a verification request to the blockchain node. When the identification cannot be read, it triggers the multi-source data fusion process of light source reflection feature analysis, logistics path backtracking calculation and image feature similarity comparison, generates a comprehensive traceability conclusion and outputs it to the redundancy recovery module; The redundant recovery module receives the comprehensive traceability conclusion and is used to call the 3D contour data, element distribution map and light source reflection characteristics in the ceramic multi-dimensional feature library when the physical identification part is damaged, match the local identification fragments through the dynamic time warping algorithm, verify the global feature consistency in combination with the cosine similarity analysis, reconstruct the complete identification hash chain and update the blockchain node storage data.

[0005] Furthermore, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the data acquisition and processing module includes: A sensor data interface unit, used to receive firing temperature, molding pressure parameters and high-resolution images collected by the IoT device, and transmit the parameters and images to the edge computing unit; an edge computing unit, receiving the parameters and the image, for performing JSON / XML format conversion, filtering outliers based on a preset threshold to generate cleaned data, and outputting the cleaned data to a data alignment unit; The data alignment unit receives the cleaned data and is used to synchronize the process parameters with the surface micro-morphology data according to the timestamp, generate a structured data set and transmit it to the dynamic identification generation module.

[0006] Furthermore, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the dynamic identification generation module includes: A hash summary generating unit receives the structured data set transmitted by the data acquisition and processing module and generates a hash summary using a SHA-3 algorithm; An identification parameter binding unit receives the hash summary and is used to associate the summary with the ceramic size and surface texture characteristics to generate a blockchain identification of variable length; The index association unit receives the blockchain identifier, is used to establish a mapping relationship between the identifier and the multi-source feature data, and outputs the mapped identifier parameters to the identifier embedding control module.

[0007] Furthermore, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the identification embedding control module includes: A process parameter mapping unit receives the blockchain identifier output by the dynamic identifier generation module and is used to convert the dot density and code length of the identifier into laser pulse frequency and engraving depth parameters; The laser micro-engraving execution unit receives the laser pulse frequency and engraving depth parameters, drives the laser device to engrave the internal dot matrix before firing the ceramic body according to the control instructions, and collects the engraving quality data through the optical detection device and feeds it back to the blockchain node; The coating identification adapter unit receives the thermal expansion coefficient data of the ceramic substrate, is used to dynamically adjust the composition ratio of the transparent zirconium oxide coating, and drives the ultraviolet laser to generate a refractive index difference type identification.

[0008] Furthermore, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the cross-chain data synchronization module includes: A private chain writing unit receives the engraving quality data fed back by the identification embedding control module and is used to write the identification data, process parameters and engraving quality data into the private chain; A cross-chain gateway unit receives the production batches and quality inspection results stored in the private chain writing unit, and synchronizes them to the alliance chain node through the cross-chain protocol; The smart contract execution unit receives the logistics trajectory data updated by the alliance chain node to verify the spatiotemporal consistency of the logistics path with the preset rules. When the path deviates, it freezes the batch data update authority and generates an early warning event.

[0009] Furthermore, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the multimodal verification module includes: A trigger judgment unit is used to monitor NFC or optical reading failure events and start the coordinated work of the light source reflection analysis unit, the path backtracking unit and the image feature comparison unit; A light source reflection analysis unit receives instructions from the trigger judgment unit and is used to match the reflection spectrum data collected by the user terminal with the glaze firing characteristics stored in the blockchain node; A path tracing unit receives the ceramic geographic location data uploaded by the user, is used to filter the production batch range based on the logistics path tracing algorithm, and outputs the screening results to the image feature comparison unit; The image feature comparison unit receives the screening result of the path tracing unit, is used to extract the ceramic surface image features based on the ResNet-50 model, calculate the similarity with the original production image stored in the blockchain, generate the traceability conclusion and output it to the redundancy recovery module.

[0010] Furthermore, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the redundant recovery module includes: A feature library construction unit for storing ceramic 3D profile data, X-ray fluorescence element distribution maps and light source reflection characteristics; A local matching unit receives the damaged identification fragment output by the multimodal verification module, calculates its local similarity with the complete blockchain identification using a dynamic time warping algorithm, screens potential matches and outputs them to a global verification unit; A global verification unit receives the potential matches and is used to perform cosine similarity analysis on the feature vectors to verify the consistency of the ceramic identity; The data reconstruction unit receives the verified matches and is used to retrieve the original data copy stored redundantly in the blockchain, reconstruct the complete identification hash chain and update the smart contract status.

[0011] Furthermore, the blockchain-based ceramic supply chain quality traceability data management system of the present invention also includes: The resource scheduling module connects the cross-chain data synchronization module and the multimodal verification module, is used to receive the logistics verification data of the cross-chain data synchronization module and the traceability request of the multimodal verification module, dynamically allocates the computing resources of the blockchain nodes through the Q-learning algorithm, and gives priority to high-priority requests; The terminal device integrated module integrates NFC, optical recognition and light source reflection devices, and is used to transmit the collected identification data, reflection spectrum data and image data to the blockchain node through a unified communication protocol and receive verification results.

[0012] Furthermore, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the resource scheduling module includes: Load monitoring unit, used to collect real-time data on the request processing volume and response time of blockchain nodes; A priority determination unit, receiving the data of the load monitoring unit and the query confidence parameter of the multimodal verification module, for classifying the request levels; A resource allocation unit receives the request level, allocates computing resources of blockchain nodes according to the level, and gives priority to high-confidence traceability requests.

[0013] Furthermore, the blockchain-based ceramic supply chain quality traceability data management system of the present invention also includes: A quantum-resistant encryption module, connected to the dynamic identification generation module and the redundancy recovery module, configured to perform quantum-resistant signature algorithm encryption on the feature vector output by the dynamic identification generation module, and output the encrypted data to a distributed storage unit; The distributed storage unit receives the encrypted data, is used to store the encrypted feature vector through the IPFS protocol, and establishes an index association with the blockchain node to support data calling of the redundant recovery module.

[0014] Beneficial effects of the present invention: The beneficial effect of the present invention is to systematically solve the core problems in the quality traceability of the ceramic supply chain through a modular collaborative technical solution: the data acquisition and processing module integrates multi-source heterogeneous data through edge computing and timestamp synchronization technology, provides highly consistent input for dynamic identification generation, and combines the blockchain identification generated by hash algorithm and physical feature binding to improve the data immutability; the identification embedding control module adopts laser micro-engraving and transparent coating adaptation technology to engrave invisible identification inside the ceramic body or in the coating to avoid destroying the surface artistic integrity, and at the same time realizes real-time verification of engraving quality and intelligent verification of logistics path through optical detection and cross-chain data synchronization; the multimodal verification module can be used when the physical identification is unreadable. , a comprehensive traceability conclusion is generated through light source reflection feature matching, logistics path backtracking and ResNet-50 image feature comparison. The dynamic time warping algorithm of the redundant recovery module and the cosine similarity hierarchical verification mechanism are combined to reconstruct the damaged identification hash chain to solve the defect of insufficient reliability of the physical carrier; the resource scheduling module dynamically optimizes the node computing resource allocation based on the Q-learning algorithm, and the terminal equipment integration module reduces the complexity of user operations through multi-hardware collaboration and unified protocol. Quantum-resistant encryption and IPFS distributed storage technology ensure the long-term security and recoverability of data, and finally achieve the coordinated optimization of the accuracy of ceramic supply chain quality traceability, operational efficiency and artistic protection needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on the drawings without paying any creative labor.

[0016] Figure 1 A system architecture diagram of a blockchain-based ceramic supply chain quality traceability data management system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The technical solutions provided by the embodiments of the present invention are described in detail below in conjunction with the drawings. In order to better understand the purpose of the present invention, the present invention is further described in detail below.

[0018] See also Figure 1The ceramic supply chain quality traceability data management system based on blockchain provided by the present invention includes: The data acquisition and processing module is deployed in the ceramic production line, and is used to collect process parameters and image data through IoT devices, perform format conversion and outlier cleaning through edge computing nodes to generate a structured data set, and transmit the structured data set to the dynamic identification generation module; The technical solution of the data acquisition and processing module is specifically implemented as follows: The sensor data interface unit is connected to the firing furnace, pressure sensor and high-resolution camera of the ceramic production line through the Internet of Things device to receive the real-time collected firing temperature values, molding pressure parameters and surface microscopic image data. The unit uses the MQTT protocol to transmit the received raw data to the edge computing unit to complete the initial data aggregation. The edge computing unit performs JSON format standardization conversion on the input firing temperature and pressure parameters, and encapsulates the high-resolution image data into XML metadata. Based on the preset temperature fluctuation threshold and pressure range threshold, the edge computing unit filters abnormal data points that exceed the reasonable range and generates a cleaned process parameter set and image data set. The cleaned data is output to the data alignment unit to complete the preprocessing process.

[0019] The data alignment unit receives the cleaned data transmitted by the edge computing unit, extracts the timestamp information in the process parameters and the acquisition time code of the surface micro-morphology image. The firing temperature and molding pressure parameters are aligned with the micro-morphology image data at the corresponding time through the timestamp synchronization algorithm to eliminate the data timing offset caused by the difference in sensor acquisition frequency. The aligned process parameters and image data are associated and integrated according to the batch number to generate a structured data set containing the time dimension, process dimension and image dimension. This data set is transmitted to the dynamic identification generation module through the data bus to provide multi-dimensional data input for blockchain identification generation.

[0020] The data processing flow between each unit forms a closed loop: the sensor data interface unit realizes the collection and transmission of raw data, the edge computing unit completes the format conversion and abnormal cleaning, and the data alignment unit builds cross-modal data association through timestamp synchronization. This technical solution solves the problem of real-time integration of multi-source heterogeneous data in the ceramic production process through the collaboration of edge computing and data alignment, and provides a highly consistent data foundation for subsequent identification generation.

[0021] A dynamic identification generation module receives the structured data set, is used to perform a hash operation on the structured data set to generate a hash summary, combines the production batch number, the timestamp and the physical characteristics of the ceramic to generate a blockchain identification of a variable length, and outputs the characteristic parameters of the blockchain identification to the identification embedding control module; The technical solution of the dynamic identification generation module is specifically implemented as follows: The hash summary generation unit receives the structured data set transmitted by the data acquisition and processing module, and parses the firing temperature, molding pressure parameters and surface micromorphology data. The SHA-3 algorithm is used to perform hash operations on the structured data set, inputs the binary concatenation results of the process parameters and image data, and outputs a hash summary string of fixed length. The generated hash summary is transmitted to the identification parameter binding unit as the tamper-proof data core of the blockchain identification.

[0022] The identification parameter binding unit associates the hash summary with the physical characteristics of the ceramic (including size parameters and surface texture feature vectors). The physical characteristics are converted into a numerical sequence through the feature encoding algorithm and spliced ​​with the hash summary to generate a blockchain identification code of variable length. The identification length is dynamically adjusted according to the size of the ceramic: a short code is used for smaller sizes to adapt to the engraving space limitations, and a long code is used for larger sizes to carry more feature information. The generated identification contains the production batch number, timestamp and physical feature index, which is output to the index association unit.

[0023] The index association unit receives the blockchain identification code and parses the physical feature index and hash summary information it contains. The mapping relationship between the identification code and multi-source feature data (including firing process parameters, microscopic image fragments, and logistics path data) is established through a distributed key-value pair database to form an identification-feature bidirectional index table. The mapped identification parameters are transmitted to the identification embedding control module through the data interface to drive the laser engraving or coating identification generation process, completing the physical binding of the blockchain identification and the ceramic entity and the association of the on-chain data.

[0024] The data interaction between each unit forms a closed loop: the hash summary generation unit provides the data integrity basis for identification, the identification parameter binding unit realizes the dynamic adaptation of physical characteristics and blockchain data, and the index association unit supports cross-chain traceability through distributed indexes. This technical solution solves the problems of data credibility, physical adaptability and scalability in ceramic identification generation through the collaboration of hash algorithm and feature coding, and provides a highly reliable identification carrier for supply chain quality traceability.

[0025] The logo embedding control module receives the characteristic parameters of the blockchain logo, and is used to convert the dot density and code length of the blockchain logo into the pulse frequency and engraving depth parameters of the laser device, drive the laser micro-engraving module to engrave invisible dots inside the ceramic body or in the transparent coating, and collect the engraving quality data through the optical detection device and feed it back to the blockchain node; The technical solution of embedding the logo into the control module is specifically implemented as follows: The process parameter mapping unit receives the blockchain identification characteristic parameters transmitted by the dynamic identification generation module and analyzes the dot density and code length of the identification. The dot density is converted into the pulse frequency parameter of the laser device through the preset mapping rule library, and the code length is converted into the engraving depth parameter to generate a laser control instruction set. The converted pulse frequency and engraving depth parameters are output to the laser micro-engraving execution unit to adapt to the physical properties of different ceramic blanks.

[0026] The laser micro-engraving execution unit drives the high-precision laser equipment according to the received control instructions to engrave the internal dot pattern before the ceramic body is fired. During the engraving process, the optical detection device collects the geometric accuracy (such as dot spacing, depth uniformity) and surface integrity data of the dot matrix in real time to generate an engraving quality assessment report. The report is fed back to the blockchain node through the data interface and compared with the original identification parameters for verification; if the deviation is detected to exceed the preset threshold, the identification reconstruction process is triggered to regenerate the engraving instructions.

[0027] The coating identification adapter unit receives the thermal expansion coefficient data of the ceramic matrix for the fired finished ceramics. The composition ratio of the transparent zirconia coating is dynamically adjusted through the material calculation model to match the thermal expansion coefficients of the coating and the matrix, avoiding deformation or shedding of the mark due to thermal stress during the firing and cooling process. The adjusted coating parameters drive the ultraviolet laser equipment to generate a refractive index difference mark inside the coating, and use the difference in the optical properties of the coating material to achieve the coordination of the invisibility and high readability of the mark. The optical detection device synchronously collects the refractive index distribution data of the coating mark and feeds it back to the blockchain node.

[0028] The technical connection between each unit forms a closed loop: the process parameter mapping unit completes the conversion of identification parameters to physical control instructions, the laser micro-engraving execution unit realizes the precise engraving and quality feedback of the internal identification of the blank, and the coating identification adaptation unit solves the problem of embedding the identification of the finished ceramic through material adaptation. This technical solution solves the problem of reliable embedding of blockchain identification on ceramic carriers through the cross-domain integration of laser control and material science, taking into account both artistic integrity and data immutability.

[0029] A cross-chain data synchronization module receives the engraving quality data and ceramic production data, is used to write the production data into the private chain and synchronize it to the alliance chain node through the cross-chain protocol, triggers the spatiotemporal data verification logic of the smart contract based on the logistics path, and freezes the batch data update permission when the logistics track deviates from the preset path; The technical solution of the cross-chain data synchronization module is specifically implemented as follows: The private chain writing unit receives the engraving quality data fed back by the identification embedding control module, and parses the laser engraving dot matrix parameters, process parameters and engraving depth error data collected by the optical detection device. The unit classifies the above data according to the ceramic production batch number, and writes it into a specific data block of the ceramic enterprise private chain through distributed ledger technology to form an unalterable engraving quality record. During the writing process, the Merkle tree structure is used to hash and aggregate the data to generate a block header hash value, which improves the efficiency of data verification and provides a traceable data anchor point for cross-chain synchronization.

[0030] The cross-chain gateway unit monitors the data update events of the private chain write unit and extracts the stored production batch number, quality inspection results and corresponding block hash value. Through the cross-chain interoperability protocol (such as relay chain technology or hash time lock mechanism), the key data packet is synchronized to the alliance chain node composed of dealers, logistics providers and testing agencies. During the synchronization process, the cross-chain gateway unit digitally signs and encrypts the data packet, and completes the data chain after the alliance chain node verifies the legitimacy of the signature, realizing the trusted data intercommunication between the private chain and the alliance chain.

[0031] The smart contract execution unit is deployed on the alliance chain node, and monitors the updated logistics trajectory data in the alliance chain in real time (including geographic location coordinates, transportation timestamps, and environmental temperature and humidity parameters). The contract has a built-in time and space verification rule library, which presets the logistics path coordinate range, the maximum transportation time threshold, and the temperature and humidity allowable range. When it is detected that the logistics trajectory deviates from the preset path or the environmental parameters exceed the threshold, the smart contract automatically triggers the execution logic: freeze the data update permissions of the batch of ceramics in the alliance chain to prevent abnormal data from continuing to be written; at the same time, an event log containing deviation details is generated, and the relevant participants are notified through the message queue for manual verification. The freezing state lasts until the data validity is confirmed by manual intervention to ensure the authenticity and traceability of the supply chain data.

[0032] The technical connection between each unit forms a closed loop: the private chain writing unit provides a data source for cross-chain, the cross-chain gateway unit realizes the trusted data synchronization between heterogeneous chains, and the smart contract execution unit performs automatic verification based on the synchronized data. This technical solution solves the problem of inter-chain isolation of production data and logistics data in the ceramic supply chain through the collaboration of cross-chain protocols and smart contracts, and improves the transparency and consistency of data in the entire chain.

[0033] The multimodal verification module is deployed on the user terminal to parse the hash value of the ceramic identification and initiate a verification request to the blockchain node. When the identification cannot be read, it triggers the multi-source data fusion process of light source reflection feature analysis, logistics path backtracking calculation and image feature similarity comparison, generates a comprehensive traceability conclusion and outputs it to the redundancy recovery module; The comprehensive tracing conclusions mainly include the following contents: Light source reflection analysis results: Based on the similarity score between the reflection spectrum and the glaze firing characteristics (such as crystal morphology and refractive index) stored in the blockchain; Logistics path backtracking screening scope: a list of candidate production batches reversely derived from geographic location and alliance chain logistics data; Image feature similarity score: cosine similarity ranking between the ceramic surface image features extracted by the ResNet-50 model and the original image; Weighted integration conclusion: The weight distribution results of the above multi-source data are combined to generate the credibility score and the final batch matching results; Recovery trigger instruction: When the conclusion does not meet the threshold, the redundancy recovery module is triggered to execute the identification reconstruction process.

[0034] The comprehensive traceability conclusion is the verification output of multimodal data fusion, which supports subsequent data recovery or user traceability query.

[0035] The technical solution of the multimodal verification module is specifically implemented as follows: The trigger judgment unit continuously monitors the NFC reading request or optical scanning operation initiated by the user terminal. When it detects that the number of consecutive reading failures exceeds the preset threshold, it is determined as an unreadable physical identification event, triggering the multi-source feature fusion process. The unit sends a collaborative start instruction to the light source reflection analysis unit, the path backtracking unit, and the image feature comparison unit, and synchronously allocates computing resources to start parallel processing.

[0036] After receiving the trigger command, the light source reflection analysis unit controls the built-in 650nm wavelength light source of the user terminal to irradiate the ceramic surface, collect reflection spectrum data and perform normalization processing. The processed spectrum data is compared with the glaze firing feature database stored in the blockchain node, which contains the glaze crystal morphology, pore distribution and refractive index parameters of each production batch. The similarity score between the current spectrum and the database record is calculated through the spectrum matching algorithm, and the preliminary production batch inference result is generated and output to the path tracing unit.

[0037] The path tracing unit receives the coordinates of the current ceramic location uploaded by the user and the batch inference results of the light source reflection analysis unit, and calls the logistics path tracing algorithm. The algorithm is based on the logistics node data stored in the alliance chain, and reversely traces the possible circulation path of the batch of ceramics to filter out the range of production batches that match the current location. The screening results include the identification code, production time and quality inspection records of the candidate batches, which are output to the image feature comparison unit to narrow the feature comparison range.

[0038] The image feature comparison unit receives the candidate batch data provided by the path tracing unit and controls the user terminal camera to collect ceramic surface images. The image feature vector is extracted based on the ResNet-50 convolutional neural network model, and the cosine similarity is calculated with the original production image features stored in the blockchain. During the comparison process, the image data corresponding to the candidate batch is loaded first, and a similarity score sorting list of each batch is generated. The comprehensive traceability conclusion is generated by weighted analysis of the light source reflection analysis results, the path tracing screening range and the image similarity score and transmitted to the redundant recovery module to complete the verification process of multimodal data fusion.

[0039] The technical connection between each unit forms a closed loop: the trigger judgment unit starts the multi-source analysis process, the light source reflection unit and the path backtracking unit narrow the verification scope from the physical characteristics and logistics data dimensions respectively, and the image comparison unit performs precise matching based on the previous results. This technical solution solves the traceability problem in the scenario of physical identification damage through the collaborative verification of multimodal data, and improves the robustness of the system and the convenience of user operation.

[0040] The redundant recovery module receives the comprehensive traceability conclusion and is used to call the 3D contour data, element distribution map and light source reflection characteristics in the ceramic multi-dimensional feature library when the physical identification part is damaged, match the local identification fragments through the dynamic time warping algorithm, verify the global feature consistency in combination with the cosine similarity analysis, reconstruct the complete identification hash chain and update the blockchain node storage data.

[0041] The technical solution of the redundancy recovery module is specifically implemented as follows: the feature library construction unit collects 3D contour data of ceramic products through a 3D laser scanner, generates an element distribution map in combination with an X-ray fluorescence spectrometer, and records the reflection spectrum characteristics under the irradiation of a specific wavelength light source. The above multi-dimensional feature data is classified and stored in a distributed feature library according to the production batch number, forming a physical feature file associated with the blockchain identifier, providing data support for the recovery of damaged identifiers.

[0042] The local matching unit receives the damaged identification fragment transmitted by the multimodal verification module, and parses the dot matrix coding fragment and local spectral features contained therein. The dynamic time warping algorithm is used to align the time coding sequence of the damaged fragment and the complete blockchain identification, and the similarity score of the local dot matrix density and coding interval is calculated. Candidate complete identifications with similarity higher than the threshold are screened, and potential matches containing matching positions and confidence parameters are output to the global verification unit.

[0043] The global verification unit performs multi-dimensional feature fusion verification on the candidate identifications in the potential matches: extracts the 3D contour feature vector, element distribution vector and reflectance spectrum vector associated with the candidate identification, and performs cosine similarity analysis with the feature data currently collected from the broken ceramic. When the comprehensive similarity score exceeds the preset verification threshold, the ceramic identity consistency is determined to be established, and a verification pass instruction is generated and transmitted to the data reconstruction unit.

[0044] The data reconstruction unit retrieves the original identification data copy of the corresponding batch in the blockchain redundant storage node according to the verification instruction, including the complete dot matrix code, process parameter hash chain and logistics path record. The association between the identification code fragment and the original data is reorganized through the hash tree structure to reconstruct the complete blockchain identification hash chain. The updated hash chain triggers the change of the smart contract status, marks the ceramic status as recovered, and synchronously updates the traceability information of the alliance chain node.

[0045] The technical connection between each unit forms a closed loop: the feature library provides a multi-source data basis for recovery, local matching and global verification realize hierarchical verification, and data reconstruction completes the synchronization of on-chain and off-chain data. This technical solution solves the data recovery problem in the scenario of physical identification damage through multi-dimensional feature fusion and hierarchical verification mechanism, and improves the fault tolerance and data integrity of the supply chain traceability system.

[0046] The technical solution of the ceramic supply chain quality traceability data management system based on blockchain provided by the present invention is specifically implemented as follows: The data acquisition and processing module is deployed on the ceramic production line. It collects process parameters such as firing temperature and molding pressure and high-resolution image data in real time through IoT devices, and transmits the original data to the edge computing node. The edge computing node performs JSON or XML format standardization processing on the original data, filters sensor noise and outliers based on preset thresholds, and generates cleaned data. The cleaned process parameters and surface micromorphology data are synchronously aligned by timestamp through the data alignment unit to form a structured data set and transmit it to the dynamic identification generation module, providing standardized data input for subsequent identification generation.

[0047] After receiving the structured data set, the dynamic identification generation module uses the SHA-3 hash algorithm to generate a unique hash summary for the data set. The module combines the ceramic production batch number, timestamp and physical characteristics (such as size and surface texture) to convert the hash summary into a blockchain identification with variable length. The identification length is adaptively adjusted according to the physical characteristics of the ceramic, and an index association relationship with multi-source feature data is established. The generated feature parameters are output to the identification embedding control module to achieve dynamic binding of data and physical objects.

[0048] The logo embedding control module converts the received blockchain logo characteristic parameters (such as dot density and code length) into the pulse frequency and engraving depth control instructions of the laser equipment. The laser micro-engraving execution unit engraves the invisible dot pattern before firing the ceramic body according to the instruction parameters, and collects the engraving quality data in real time through the optical detection device and feeds it back to the blockchain node to verify the integrity of the logo. For the fired finished ceramics, the coating logo adapter unit dynamically adjusts the proportion of the transparent zirconia coating based on the thermal expansion coefficient of the ceramic matrix, and drives the ultraviolet laser to generate a refractive index difference invisible logo inside the coating to avoid deformation of the logo caused by high-temperature firing.

[0049] The cross-chain data synchronization module writes ceramic production data and engraving quality data into the private chain, and synchronizes key information such as production batches and quality inspection results to the alliance chain nodes composed of dealers and testing agencies through the cross-chain gateway. The smart contract execution unit receives the logistics track data updated by the alliance chain node and verifies the spatiotemporal consistency of the logistics path and the preset rules. When it is detected that the logistics track deviates from the preset path, the smart contract automatically freezes the update permission of the batch data and generates an early warning event to achieve data authenticity and supply chain traceability.

[0050] The multimodal verification module is deployed on the user terminal to parse the hash value of the ceramic identification and initiate a verification request to the blockchain node. When the physical identification cannot be read due to wear, the judgment unit is triggered to start the multi-source feature fusion process: the light source reflection analysis unit illuminates the ceramic surface with 650nm visible light, matches the reflection spectrum data with the glaze firing characteristics stored in the blockchain; the path backtracking unit combines the user's geographic location data and screens the production batch range based on the logistics path backtracking algorithm; the image feature comparison unit uses the ResNet-50 model to extract the image features of the ceramic surface and calculates the similarity with the original blockchain image. The above multi-source data is weighted and integrated to generate a comprehensive traceability conclusion, which is output to the redundancy recovery module.

[0051] When the physical identification part is damaged, the redundant recovery module calls the 3D contour data, X-ray fluorescence element distribution map and light source reflection characteristics in the ceramic multi-dimensional feature library to generate a composite feature vector. The local matching unit calculates the local similarity between the damaged identification fragment and the complete blockchain identification through the dynamic time warping (DTW) algorithm to screen potential matches; the global verification unit performs cosine similarity analysis on the screening results to verify the consistency of the ceramic identity. After meeting the matching threshold, the data reconstruction unit calls the original data copy stored redundantly in the blockchain, reconstructs the complete identification hash chain and updates the smart contract status to achieve the recovery of traceability information and synchronization of on-chain data.

[0052] The modules form a closed-loop management through data flow: the output of the data acquisition and processing module drives the generation of dynamic identification, the generated identification parameters control the embedding process, the embedded data and production data are synchronized to the blockchain node through the cross-chain module, and the multi-modal verification and redundant recovery mechanism ensure data availability. This technical solution solves the problem of collaborative optimization of ceramic identification reliability, concealment and traceability efficiency through the interdisciplinary integration of material science, cryptography and machine learning technology, and meets the dual needs of the ceramic industry for quality traceability and user experience.

[0053] Specifically, the data collection and processing module of the ceramic supply chain quality traceability data management system based on blockchain of the present invention includes: A sensor data interface unit, used to receive firing temperature, molding pressure parameters and high-resolution images collected by the IoT device, and transmit the parameters and images to the edge computing unit; an edge computing unit, receiving the parameters and the image, for performing JSON / XML format conversion, filtering outliers based on a preset threshold to generate cleaned data, and outputting the cleaned data to a data alignment unit; The data alignment unit receives the cleaned data and is used to synchronize the process parameters with the surface micro-morphology data according to the timestamp, generate a structured data set and transmit it to the dynamic identification generation module.

[0054] The technical solution of the data acquisition and processing module is specifically implemented as follows: The sensor data interface unit is connected to the firing furnace, pressure sensor and high-resolution camera of the ceramic production line through the Internet of Things device to receive the real-time collected firing temperature values, molding pressure parameters and surface microscopic image data. The unit uses the MQTT protocol to transmit the received raw data to the edge computing unit to complete the initial data aggregation. The edge computing unit performs JSON format standardization conversion on the input firing temperature and pressure parameters, and encapsulates the high-resolution image data into XML metadata. Based on the preset temperature fluctuation threshold and pressure range threshold, the edge computing unit filters abnormal data points that exceed the reasonable range and generates a cleaned process parameter set and image data set. The cleaned data is output to the data alignment unit to complete the preprocessing process.

[0055] The data alignment unit receives the cleaned data transmitted by the edge computing unit, extracts the timestamp information in the process parameters and the acquisition time code of the surface micro-morphology image. The firing temperature and molding pressure parameters are aligned with the micro-morphology image data at the corresponding time through the timestamp synchronization algorithm to eliminate the data timing offset caused by the difference in sensor acquisition frequency. The aligned process parameters and image data are associated and integrated according to the batch number to generate a structured data set containing the time dimension, process dimension and image dimension. This data set is transmitted to the dynamic identification generation module through the data bus to provide multi-dimensional data input for blockchain identification generation.

[0056] The data processing flow between each unit forms a closed loop: the sensor data interface unit realizes the collection and transmission of raw data, the edge computing unit completes the format conversion and abnormal cleaning, and the data alignment unit builds cross-modal data association through timestamp synchronization. This technical solution solves the problem of real-time integration of multi-source heterogeneous data in the ceramic production process through the collaboration of edge computing and data alignment, and provides a highly consistent data foundation for subsequent identification generation.

[0057] Specifically, in the ceramic supply chain quality traceability data management system based on blockchain described in the present invention, the dynamic identification generation module includes: A hash summary generating unit receives the structured data set transmitted by the data acquisition and processing module and generates a hash summary using a SHA-3 algorithm; An identification parameter binding unit receives the hash summary and is used to associate the summary with the ceramic size and surface texture characteristics to generate a blockchain identification of variable length; The index association unit receives the blockchain identifier, is used to establish a mapping relationship between the identifier and the multi-source feature data, and outputs the mapped identifier parameters to the identifier embedding control module.

[0058] The technical solution of the dynamic identification generation module is specifically implemented as follows: The hash summary generation unit receives the structured data set transmitted by the data acquisition and processing module, and parses the firing temperature, molding pressure parameters and surface micromorphology data. The SHA-3 algorithm is used to perform hash operations on the structured data set to generate a unique hash summary, where the algorithm input includes the binary concatenation results of all fields in the data set, and the output is a summary string of fixed length. The generated hash summary is transmitted to the identification parameter binding unit as the core data basis of the blockchain identification.

[0059] The identification parameter binding unit associates the hash summary with the physical features of the ceramic (including size parameters and surface texture feature vectors). The physical features are converted into a numerical sequence through a feature encoding algorithm and concatenated with the hash summary to generate a blockchain identification code of variable length. The identification length is dynamically adjusted according to the size of the ceramic: smaller ceramics use short codes to adapt to the engraving space limitations, and larger ceramics use long codes to carry more feature information. The generated blockchain identification contains the production batch number, timestamp and physical feature index, which is output to the index association unit.

[0060] The index association unit receives the blockchain identification code and parses the physical feature index and hash summary information it contains. The mapping relationship between the identification code and multi-source feature data (including firing process parameters, microscopic image fragments, and logistics path data) is established through a distributed key-value pair database to form an identification-feature bidirectional index table. The mapped identification parameters are transmitted to the identification embedding control module through the data interface to drive the subsequent laser engraving or coating identification generation process, completing the physical binding of the blockchain identification and the ceramic entity and the association of the on-chain data.

[0061] The data interaction between each unit forms a closed loop: the hash summary generation unit provides a cryptographic basis for the identification that the data cannot be tampered with, the identification parameter binding unit realizes the dynamic association between physical characteristics and blockchain data, and the index association unit supports cross-chain data traceability through distributed indexes. This technical solution solves the technical problems of data integrity, physical adaptability and scalability in ceramic identification generation through the collaboration of hash algorithm and feature coding, and provides a highly reliable identification carrier for supply chain quality traceability.

[0062] Specifically, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the identification embedding control module includes: A process parameter mapping unit receives the blockchain identifier output by the dynamic identifier generation module and is used to convert the dot density and code length of the identifier into laser pulse frequency and engraving depth parameters; The laser micro-engraving execution unit receives the laser pulse frequency and engraving depth parameters, drives the laser device to engrave the internal dot matrix before firing the ceramic body according to the control instructions, and collects the engraving quality data through the optical detection device and feeds it back to the blockchain node; The coating identification adapter unit receives the thermal expansion coefficient data of the ceramic substrate, is used to dynamically adjust the composition ratio of the transparent zirconium oxide coating, and drives the ultraviolet laser to generate a refractive index difference type identification.

[0063] The process parameter mapping unit receives the blockchain identification code transmitted by the dynamic identification generation module and parses the dot density and code length parameters of the identification. The dot density is converted into the pulse frequency parameter of the laser device through the preset mapping rule library, and the code length is converted into the engraving depth parameter to generate a laser control instruction set. The converted pulse frequency and engraving depth parameters are output to the laser micro-engraving execution unit to adapt to the physical properties of different ceramic blanks.

[0064] The laser micro-engraving execution unit drives the high-precision laser equipment to write the internal dot matrix before the ceramic body is fired according to the received pulse frequency and engraving depth parameters. During the engraving process, the optical detection device collects the geometric accuracy and depth data of the dot matrix in real time and generates an engraving quality assessment report. The report is fed back to the blockchain node through the data interface and compared with the original identification code for verification. If the consistency is lower than the threshold, the identification reconstruction process is triggered.

[0065] The coating identification adapter unit receives the thermal expansion coefficient data of the ceramic matrix for the fired finished ceramics. The composition ratio of the transparent zirconium oxide coating is dynamically adjusted through the material calculation model to match the thermal expansion coefficient of the coating with that of the ceramic matrix, avoiding deformation of the mark due to thermal stress during the firing and cooling process. The adjusted coating parameters drive the ultraviolet laser to generate a refractive index difference type mark inside the coating, and the difference in the optical properties of the coating material is used to achieve the coordination of the invisibility and high readability of the mark.

[0066] The technical logic between each unit forms a closed loop: the process parameter mapping unit completes the conversion of identification parameters to physical control instructions, the laser micro-engraving execution unit realizes the precise engraving and quality feedback of the internal identification of the blank, and the coating identification adaptation unit solves the problem of embedding the identification of the finished ceramic. This technical solution realizes the reliable embedding of blockchain identification on the ceramic carrier through the cross-domain integration of laser control and material science, taking into account the artistic integrity and the non-tamperability of traceability data.

[0067] Specifically, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the cross-chain data synchronization module includes: A private chain writing unit receives the engraving quality data fed back by the identification embedding control module and is used to write the identification data, process parameters and engraving quality data into the private chain; A cross-chain gateway unit receives the production batches and quality inspection results stored in the private chain writing unit, and synchronizes them to the alliance chain node through the cross-chain protocol; The smart contract execution unit receives the logistics trajectory data updated by the alliance chain node to verify the spatiotemporal consistency of the logistics path with the preset rules. When the path deviates, it freezes the batch data update authority and generates an early warning event.

[0068] The technical solution of the cross-chain data synchronization module is specifically implemented as follows: The private chain writing unit receives the engraving quality data fed back by the identification embedding control module, and parses the laser engraving dot matrix parameters, process parameters and engraving depth error data collected by the optical detection device. The unit classifies the above data by ceramic production batch number, and writes it into a specific data block of the ceramic enterprise private chain through distributed ledger technology to form an unalterable engraving quality record. During the writing process, the Merkle tree structure is used to hash and aggregate the data to improve the efficiency of data verification and provide a traceable data anchor point for subsequent cross-chain synchronization.

[0069] The cross-chain gateway unit monitors the data update events of the private chain write unit and extracts the stored production batch number, quality inspection results and corresponding block hash value. Through the cross-chain interoperability protocol (such as relay chain technology or hash time lock mechanism), the key data packet is synchronized to the alliance chain node composed of dealers, logistics providers and testing agencies. During the synchronization process, the cross-chain gateway unit signs and encrypts the data packet, and completes the data chain after the alliance chain node verifies the legitimacy of the signature, realizing the trusted data intercommunication between the private chain and the alliance chain.

[0070] The smart contract execution unit is deployed on the alliance chain node, and monitors the updated logistics trajectory data (including geographic location, transportation timestamp and environmental parameters) in the alliance chain in real time. The contract has a built-in time and space verification rule library, such as the preset logistics path coordinate range, the maximum transportation time threshold and the temperature and humidity allowable range. When it is detected that the logistics trajectory deviates from the preset path or exceeds the threshold, the smart contract automatically triggers the execution logic: freeze the data update permission of the batch of ceramics in the alliance chain to prevent abnormal data from being written; at the same time, an event log containing deviation details is generated, and the relevant participants are notified through the message queue for manual verification. The freezing state lasts until the data validity is confirmed by manual intervention to ensure the authenticity of the supply chain data.

[0071] The technical connection between each unit forms a closed loop: the private chain writing unit provides a data source for cross-chain, the cross-chain gateway unit realizes the trusted data synchronization between heterogeneous chains, and the smart contract execution unit performs automatic verification based on the synchronized data. This technical solution solves the problem of inter-chain isolation of production data and logistics data in the ceramic supply chain through the collaboration of cross-chain protocols and smart contracts, and improves the transparency and consistency of data in the entire chain.

[0072] Specifically, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the multimodal verification module includes: A trigger judgment unit is used to monitor NFC or optical reading failure events and start the coordinated work of the light source reflection analysis unit, the path backtracking unit and the image feature comparison unit; A light source reflection analysis unit receives instructions from the trigger judgment unit and is used to match the reflection spectrum data collected by the user terminal with the glaze firing characteristics stored in the blockchain node; A path tracing unit receives the ceramic geographic location data uploaded by the user, is used to filter the production batch range based on the logistics path tracing algorithm, and outputs the screening results to the image feature comparison unit; The image feature comparison unit receives the screening result of the path tracing unit, is used to extract the ceramic surface image features based on the ResNet-50 model, calculate the similarity with the original production image stored in the blockchain, generate the traceability conclusion and output it to the redundancy recovery module.

[0073] The technical solution of the multimodal verification module is specifically implemented as follows: the trigger judgment unit continuously monitors the NFC reading request or optical scanning operation initiated by the user terminal. When it detects that the number of consecutive reading failures exceeds the preset threshold, it is determined as an unreadable physical identification event, triggering the multi-source feature fusion process. The unit sends a collaborative start instruction to the light source reflection analysis unit, the path backtracking unit, and the image feature comparison unit, and synchronously allocates computing resources to start parallel processing.

[0074] After receiving the trigger command, the light source reflection analysis unit controls the built-in 650nm wavelength light source of the user terminal to irradiate the ceramic surface, collect reflection spectrum data and perform normalization processing. The processed spectrum data is compared with the glaze firing feature database stored in the blockchain node, which contains the glaze crystal morphology, pore distribution and refractive index parameters of each production batch. The similarity score between the current spectrum and the database record is calculated through the spectrum matching algorithm, and the preliminary production batch inference result is generated and output to the path tracing unit.

[0075] The path tracing unit receives the coordinates of the current ceramic location uploaded by the user and the batch inference results of the light source reflection analysis unit, and calls the logistics path tracing algorithm. The algorithm is based on the logistics node data stored in the alliance chain, and reversely traces the possible circulation path of the batch of ceramics to filter out the range of production batches that match the current location. The screening results include the identification code, production time and quality inspection records of the candidate batches, which are output to the image feature comparison unit to narrow the feature comparison range.

[0076] The image feature comparison unit receives the candidate batch data provided by the path tracing unit and controls the user terminal camera to collect ceramic surface images. The image feature vector is extracted based on the ResNet-50 convolutional neural network model, and the cosine similarity is calculated with the original production image features stored in the blockchain. During the comparison process, the image data corresponding to the candidate batch is loaded first, and a similarity score sorting list of each batch is generated. The comprehensive traceability conclusion is generated by weighted analysis of the light source reflection analysis results, the path tracing screening range and the image similarity score and transmitted to the redundant recovery module to complete the verification process of multimodal data fusion.

[0077] The technical connection between each unit forms a closed loop: the trigger judgment unit starts the multi-source analysis process, the light source reflection unit and the path backtracking unit narrow the verification scope from the physical characteristics and logistics data dimensions respectively, and the image comparison unit performs precise matching based on the previous results. This technical solution solves the traceability problem in the scenario of physical identification damage through the collaborative verification of multimodal data, and improves the robustness of the system and the convenience of user operation.

[0078] Specifically, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the redundant recovery module includes: A feature library construction unit for storing ceramic 3D profile data, X-ray fluorescence element distribution maps and light source reflection characteristics; A local matching unit receives the damaged identification fragment output by the multimodal verification module, calculates its local similarity with the complete blockchain identification using a dynamic time warping algorithm, screens potential matches and outputs them to a global verification unit; A global verification unit receives the potential matches and is used to perform cosine similarity analysis on the feature vectors to verify the consistency of the ceramic identity; The data reconstruction unit receives the verified matches and is used to retrieve the original data copy stored redundantly in the blockchain, reconstruct the complete identification hash chain and update the smart contract status.

[0079] The technical solution of the redundancy recovery module is specifically implemented as follows: the feature library construction unit collects 3D contour data of ceramic products through a 3D laser scanner, generates an element distribution map in combination with an X-ray fluorescence spectrometer, and records the reflection spectrum characteristics under the irradiation of a specific light source. The above multi-dimensional feature data is classified and stored in a distributed feature library according to the production batch number, forming a physical feature file associated with the blockchain identifier, providing data support for the recovery of damaged identifiers.

[0080] The local matching unit receives the damaged identification fragment transmitted by the multimodal verification module, and parses the dot matrix coding fragment and local spectral features contained therein. The dynamic time warping algorithm is used to align the time coding sequence of the damaged fragment and the complete blockchain identification, and the similarity score of the local dot matrix density and coding interval is calculated. Candidate complete identifications with similarity higher than the threshold are screened, and potential matches containing matching positions and confidence parameters are output to the global verification unit.

[0081] The global verification unit performs multi-dimensional feature fusion verification on the candidate identifications in the potential matches: extracts the 3D contour feature vector, element distribution vector and reflectance spectrum vector associated with the candidate identification, and performs cosine similarity analysis with the feature data currently collected from the broken ceramic. When the comprehensive similarity score exceeds the preset verification threshold, the ceramic identity consistency is determined to be established, and a verification pass instruction is generated and transmitted to the data reconstruction unit.

[0082] The data reconstruction unit retrieves the original identification data copy of the corresponding batch in the blockchain redundant storage node according to the verification instruction, including the complete dot matrix code, process parameter hash chain and logistics path record. The association between the identification code fragment and the original data is reorganized through the hash tree structure to reconstruct the complete blockchain identification hash chain. The updated hash chain triggers the change of the smart contract status, marks the ceramic status as recovered, and synchronously updates the traceability information of the alliance chain node.

[0083] The technical connection between each unit forms a closed loop: the feature library provides a multi-source data basis for recovery, local matching and global verification realize hierarchical verification, and data reconstruction completes the synchronization of on-chain and off-chain data. This technical solution solves the data recovery problem in the scenario of physical identification damage through multi-dimensional feature fusion and hierarchical verification mechanism, and improves the fault tolerance and data integrity of the supply chain traceability system.

[0084] Specifically, the blockchain-based ceramic supply chain quality traceability data management system of the present invention also includes: The resource scheduling module connects the cross-chain data synchronization module and the multimodal verification module, is used to receive the logistics verification data of the cross-chain data synchronization module and the traceability request of the multimodal verification module, dynamically allocates the computing resources of the blockchain nodes through the Q-learning algorithm, and gives priority to high-priority requests; The terminal device integrated module integrates NFC, optical recognition and light source reflection devices, and is used to transmit the collected identification data, reflection spectrum data and image data to the blockchain node through a unified communication protocol and receive verification results.

[0085] The technical solution of the resource scheduling module is specifically implemented as follows: The resource scheduling module collects the request processing volume, CPU utilization and network delay data of the blockchain node in real time through the load monitoring unit, and constructs the node load status matrix. The priority determination unit receives the logistics verification data of the cross-chain data synchronization module and the traceability request of the multimodal verification module, parses the batch urgency label and user query confidence parameter in the request, and divides the request priority into high, medium and low levels based on the preset rule base. The resource allocation unit adopts the Q-learning reinforcement learning model to generate a dynamic scheduling strategy based on the node load status and request priority, and allocates high-priority requests to idle nodes or low-load nodes for processing, while limiting the number of concurrent threads of low-priority requests to maintain the overall response efficiency of the system. The scheduling strategy is pushed to each blockchain node for execution through a distributed message queue, and the node load status matrix is ​​updated in real time for closed-loop optimization.

[0086] The technical solution of the terminal device integration module is specifically implemented as follows: The terminal device integration module integrates the NFC card reader, optical scanning head and 650nm wavelength light source reflection sensor at the hardware layer, and encapsulates the collected identification code data, reflection spectrum waveform and ceramic surface image data through a unified communication protocol stack (compatible with HTTP / 2 and MQTT). The protocol adapter unit reorganizes the frame structure and aligns the timestamp of the multimodal data, generates a standardized data packet and transmits it to the blockchain node through an encrypted channel. After the verification result returned by the node is parsed by the protocol, it drives the terminal display screen to output the traceability conclusion or trigger an audio and light prompt. The module has a built-in power management unit, which dynamically adjusts the sensor sampling rate and the communication module sleep cycle according to the data transmission frequency to extend the battery life of the terminal device.

[0087] The overall logical connection of the technical solution is as follows: the resource scheduling module dynamically allocates computing resources based on node load and request priority to ensure the real-time processing capability of high-confidence traceability requests; the terminal device integration module reduces the complexity of user operations and improves the reliability of data collection through multi-hardware collaboration and protocol optimization. The two modules support the large-scale commercial use of the system from the aspects of computing resource management and terminal interaction, forming a full-link efficiency optimization from data collection to on-chain verification.

[0088] Specifically, in the blockchain-based ceramic supply chain quality traceability data management system of the present invention, the resource scheduling module includes: Load monitoring unit, used to collect real-time data on the request processing volume and response time of blockchain nodes; A priority determination unit, receiving the data of the load monitoring unit and the query confidence parameter of the multimodal verification module, for classifying the request levels; A resource allocation unit receives the request level, allocates computing resources of blockchain nodes according to the level, and gives priority to high-confidence traceability requests.

[0089] The technical solution of the resource scheduling module is implemented as follows: the load monitoring unit is deployed on each blockchain node through a distributed probe to collect the number of concurrent requests, CPU occupancy rate and request response delay time data of the node in real time. The collected data is normalized to generate a node load status table, which contains node ID, real-time load factor and network topology information, providing dynamic data input for priority determination. The load status table is transmitted to the priority determination unit through a lightweight message protocol, and the update cycle is synchronized with the blockchain block generation interval.

[0090] The priority determination unit receives the node load status table transmitted by the load monitoring unit and the query confidence parameter provided by the multimodal verification module. The query confidence parameter is derived from the similarity score of the multimodal data fusion process, and the requests are divided into three levels: high, medium, and low through the preset confidence-priority mapping rule base. High-confidence requests correspond to emergency traceability queries with similarity scores exceeding the threshold, medium-level requests are for routine quality inspection data synchronization requests, and low-level requests are for historical data statistics. The divided request level labels are associated with the load status table to generate a resource allocation instruction set.

[0091] The resource allocation unit parses the resource allocation instruction set and dynamically allocates computing resources based on the node load factor. High-confidence requests are preferentially allocated to idle nodes or low-load node clusters for processing. Medium-level requests are allocated to medium-load nodes using a polling strategy, and low-level requests are delayed through batch queues. The allocation strategy is pushed to each node through the distributed task scheduler, and the thread pool size and memory allocation ratio are adjusted in real time. The execution results are fed back to the load monitoring unit to form a closed loop for updating the load status table and optimizing subsequent resource allocation decisions.

[0092] The data flow between each unit forms a closed-loop control: load monitoring provides real-time node status, priority determination establishes request classification logic, and resource allocation performs dynamic scheduling and feedback on load changes. This technical solution solves the resource contention problem of blockchain nodes in high-concurrency scenarios through a load-aware and priority-driven resource allocation mechanism, improving the processing efficiency of high-value traceability requests and the overall stability of the system.

[0093] Specifically, the blockchain-based ceramic supply chain quality traceability data management system of the present invention also includes: A quantum-resistant encryption module, connected to the dynamic identification generation module and the redundancy recovery module, configured to perform quantum-resistant signature algorithm encryption on the feature vector output by the dynamic identification generation module, and output the encrypted data to a distributed storage unit; The distributed storage unit receives the encrypted data, is used to store the encrypted feature vector through the IPFS protocol, and establishes an index association with the blockchain node to support data calling of the redundant recovery module.

[0094] The technical solution of the quantum-resistant encryption module is specifically implemented as follows: The quantum-resistant encryption module receives the ceramic feature vector data transmitted by the dynamic identification generation module and parses the hash summary, physical feature code and timestamp information contained therein. The feature vector is encrypted using a quantum-resistant signature algorithm based on lattice cryptography to generate an encrypted data packet and attach a digital signature. During the encryption process, the algorithm key management unit dynamically generates a one-time session key and generates key derivation parameters based on the ceramic batch number to improve the forward security of the encryption process. The encrypted data packet is transmitted to the distributed storage unit through a secure channel, and the key derivation parameters are written into the key management contract of the blockchain node for the authorized party to call when the data is recovered.

[0095] After receiving the encrypted data packet, the distributed storage unit stores the data shards in the distributed node network through the IPFS protocol to generate a unique content identifier (CID). During the storage process, the index association unit establishes a mapping relationship between the CID and the blockchain identification code and writes it into the index smart contract of the alliance chain. The contract records the storage node address, data shard location and access permission policy, and supports the redundant recovery module to quickly locate and call encrypted data when needed. When the redundant recovery process is triggered, the index smart contract verifies the requester's authority and returns the CID and storage node information to complete the distributed retrieval and decryption process of the encrypted data.

[0096] The technical connection between the two modules forms a closed loop: the anti-quantum encryption module provides security protection against quantum attacks for the feature vector, and the distributed storage unit achieves high-availability data storage through IPFS and blockchain indexing mechanism. This technical solution solves the security and recoverability issues of ceramic traceability data in long-term storage through the collaboration of encryption algorithms and distributed storage, and meets the privacy protection and efficient access needs of supply chain participants for sensitive data.

[0097] The technical features of the technical solution of the present invention are explained as follows: Data acquisition and processing module: Sensor data interface unit: Connects to IoT devices (such as temperature sensors, pressure sensors, and cameras) on the ceramic production line, receives firing temperature, molding pressure parameters, and high-resolution image data, and transmits them to the edge computing unit via the MQTT protocol.

[0098] Edge computing unit: converts raw data into a standardized format (such as JSON / XML), filters out abnormal values ​​based on preset thresholds (for example, removes temperature fluctuation data that exceeds a reasonable range), and generates post-cleaning process parameters and image data sets.

[0099] Data alignment unit: Synchronizes firing parameters and surface micro-morphology image data according to timestamps, eliminates data timing misalignment caused by differences in sensor acquisition frequencies, and generates a structured data set containing time, process, and image dimensions.

[0100] Dynamic logo generation module: Hash summary generation unit: Use the SHA-3 algorithm to generate a unique hash summary for the structured data set, which serves as the tamper-proof data basis for blockchain identification.

[0101] Identification parameter binding unit: Combines the hash summary with the physical characteristics of the ceramic (such as size and surface texture) to generate a blockchain identification with variable length to adapt to the engraving space limitations of ceramics of different sizes.

[0102] Index association unit: Establish a mapping relationship between blockchain identification and multi-source data (process parameters, images, logistics paths), and support cross-chain data traceability.

[0103] Logo embedded in the control module: Process parameter mapping unit: converts the dot density and code length of the blockchain logo into laser pulse frequency and engraving depth parameters to drive the engraving equipment.

[0104] Laser micro-engraving execution unit: engraves invisible dots before firing the ceramic body, and feeds back the engraving quality data to the blockchain node through an optical detection device.

[0105] Coating marking adapter unit: Dynamically adjust the transparent zirconium oxide coating composition according to the thermal expansion coefficient of the ceramic substrate, and use ultraviolet laser to generate a refractive index difference mark to avoid high temperature deformation.

[0106] Cross-chain data synchronization module: Private chain writing unit: writes identification data, process parameters and engraving quality data into the ceramic enterprise private chain, and uses the Merkle tree structure to aggregate data hash values.

[0107] Cross-chain gateway unit: Synchronize production batches and quality inspection results to the alliance chain through cross-chain protocols (such as relay chains), realizing data interoperability between private chains and alliance chains.

[0108] Smart contract execution unit: Verify the spatiotemporal consistency of the logistics route, freeze data update permissions when it deviates from the preset route, and generate warning events.

[0109] Multimodal Verification Module: Trigger judgment unit: monitors NFC or optical reading failure events and starts the multi-source verification process.

[0110] Light source reflection analysis unit: matches reflection spectrum data with glaze firing characteristics (such as crystal morphology and pore distribution) stored in blockchain.

[0111] Path tracing unit: reverse screening of production batch range based on logistics node data.

[0112] Image feature comparison unit: extracts surface image features through the ResNet-50 model and calculates the similarity with the original production image.

[0113] Redundant recovery module: Feature library construction unit: stores multi-dimensional feature data such as ceramic 3D profile, element distribution map and reflectance spectrum.

[0114] Local matching unit: The dynamic time warping (DTW) algorithm is used to align the coding sequence of the broken logo fragment and the complete logo to screen potential matches.

[0115] Global verification unit: Verify the multi-dimensional feature consistency of candidate identifications through cosine similarity analysis.

[0116] Data reconstruction unit: retrieves the original data copy stored redundantly in the blockchain, reconstructs the complete hash chain and updates the smart contract status.

[0117] Resource scheduling module: Load monitoring unit: collects load data such as request processing volume and response delay of blockchain nodes in real time.

[0118] Priority determination unit: classifies requests into high, medium, and low levels based on query confidence (such as multimodal similarity scores).

[0119] Resource allocation unit: uses the Q-learning algorithm to dynamically allocate computing resources and give priority to high-confidence traceability requests.

[0120] Terminal device integration module: Integrates NFC, optical scanning and light source reflection sensors, transmits multimodal data to blockchain nodes through a unified communication protocol (HTTP / 2 / MQTT), and simplifies user operations.

[0121] Quantum-resistant encryption module: uses a quantum-resistant signature algorithm based on lattice cryptography to encrypt feature vectors, combined with a one-time session key to improve forward security.

[0122] Distributed storage unit: Encrypted data is stored in shards through the IPFS protocol and associated with the blockchain index smart contract to support efficient retrieval and decryption.

[0123] ResNet-50 Model: Function and application: ResNet-50 is a deep residual convolutional neural network model used to extract high-dimensional feature vectors from ceramic surface images. Its residual structure solves the gradient vanishing problem in deep network training through skip connections, and can effectively capture microscopic features such as glaze texture and cracks.

[0124] Technical function: In the multimodal verification module, the model extracts features from the ceramic surface images collected by the user terminal, and calculates cosine similarity with the original production image features stored in the blockchain, to assist in generating high-precision traceability conclusions.

[0125] Synergy: After combining with the path tracing unit to narrow down the candidate batches, the ResNet-50 model focuses on the image comparison of a specific batch, improving verification efficiency.

[0126] Dynamic Time Warping (DTW) algorithm: Function and application: The DTW algorithm is used to align time series data of different lengths or speeds and find the minimum distance path between sequences through dynamic programming.

[0127] Technical function: In the redundancy recovery module, the DTW algorithm calculates the local similarity between the damaged identification fragment (such as the dot matrix coding sequence) and the complete blockchain identification, screens potential matches, and solves the data discontinuity problem caused by local damage to the physical identification.

[0128] Collaborative relationship: Cooperate with the cosine similarity analysis of the global verification unit to realize the hierarchical recovery logic of "local matching to global verification".

[0129] Q-learning algorithm: Function and application: Q-learning is a model-free reinforcement learning algorithm that optimizes decision-making strategies through exploration and exploitation mechanisms.

[0130] Technical role: In the resource scheduling module, Q-learning dynamically allocates computing resources based on the real-time load status of blockchain nodes (such as CPU utilization, request processing volume) and request priority (high / medium / low). For example, high-confidence traceability requests are preferentially allocated to idle nodes to improve system response speed.

[0131] Collaborative relationship: Linking with the data of the load monitoring unit forms a dynamic scheduling closed loop to ensure the processing efficiency of high-value requests.

[0132] Cosine similarity analysis: Function and application: By calculating the cosine value of the angle between two vectors, the directional consistency of the two vectors is measured. The value range is [-1,1]. The larger the value, the higher the similarity.

[0133] Technical function: In the global verification stage of the redundant recovery module, cosine similarity analysis is performed on the 3D contour, element distribution and reflectance spectrum feature vector of the candidate mark to verify its consistency with the broken ceramic and avoid mismatching.

[0134] Collaborative relationship: Combined with the local matching results of the DTW algorithm, the ceramic identity is comprehensively judged to ensure the reliability of the recovered data.

[0135] Quantum-resistant signature algorithm (based on lattice cryptography): Functions and applications: The digital signature scheme built using lattice cryptography can resist quantum computer attacks and ensure the long-term security of data.

[0136] Technical function: In the quantum-resistant encryption module, the feature vector (such as hash summary, physical feature coding) output by the dynamic identification generation module is encrypted to prevent future quantum computing from cracking traditional encryption algorithms.

[0137] Collaborative relationship: The encrypted data is stored in shards through distributed storage units (IPFS protocol) and associated with blockchain indexes, supporting secure data calls by redundant recovery modules.

[0138] Model interoperability and technical problem solving: ResNet-50 and multimodal verification: Improve traceability accuracy through image feature comparison and solve verification problems when physical labels are unreadable.

[0139] DTW and cosine similarity analysis: The hierarchical verification mechanism (local to global) solves the matching error problem caused by identifier fragmentation and improves the recovery success rate.

[0140] Q-learning and resource scheduling: Dynamic resource allocation optimizes system throughput in high-concurrency scenarios and solves the response delay problem caused by traditional static scheduling.

[0141] Anti-quantum signatures and distributed storage: Ensure data's ability to resist attacks during long-term storage and transmission, and address security risks under the threat of future quantum computing.

[0142] IPFS (InterPlanetary File System) is a distributed file storage protocol based on content addressing. Its core principle is to locate data through the unique hash value (CID, content identifier) ​​of the file content, rather than the traditional storage method based on the server address. In this invention, the IPFS protocol is used to store encrypted ceramic multi-dimensional feature data (such as 3D contours, element distribution maps and light source reflection characteristics). The specific technical logic is as follows: The encrypted feature vector data output by the quantum-resistant encryption module is divided into multiple data blocks through the IPFS protocol and distributedly stored in the global node network. Each data block generates a unique CID hash identifier to avoid single point failure risks and improve the reliability and anti-attack capabilities of data storage.

[0143] The CID of each data block stored in IPFS is mapped to the blockchain identification code, and the storage node address and access permission policy are recorded through the index smart contract of the alliance chain. When the redundant recovery module needs to call data, it quickly locates the storage node according to the CID and retrieves the encrypted data shards to achieve efficient data recovery.

[0144] The IPFS network uses a point-to-point transmission mechanism. After data is encrypted, it is stored in multiple independent nodes in pieces. Even if some nodes fail or are attacked, the data can still be fully restored through the remaining nodes. Combined with the anti-quantum signature algorithm, the encryption management of CID and access rights prevents unauthorized access or data tampering.

[0145] In the application scenario of the present invention, IPFS distributed storage and blockchain technology work together to solve the capacity bottleneck and single-point dependence problems in the long-term storage of ceramic traceability data. At the same time, through hash addressing and encrypted sharding mechanisms, high availability, security and retrieval efficiency of data are guaranteed, providing reliable data underlying support for the redundant recovery module.

[0146] NFC (Near Field Communication) technology explained: NFC (Near Field Communication) is a short-range wireless communication technology based on radio frequency identification (RFID), with an operating frequency of 13.56MHz and a communication distance of usually less than 10cm, suitable for contactless data exchange between devices. In the technical solution of the present invention, NFC technology is integrated into the user terminal device to quickly read the blockchain identification information embedded on the surface or inside of ceramic products. The specific application logic is as follows: Ceramic products are embedded with NFC tags through laser micro-engraving or transparent coating technology, and the hash value of the blockchain logo and the production batch number are stored in the tag. When the user terminal device (such as a mobile phone or a dedicated scanner) is close to the ceramic product, the tag data is automatically read through NFC sensing, without the need for physical contact or complex operations, which improves the convenience of user interaction.

[0147] Triggering the multimodal verification process: When NFC reading fails (such as tag damage or signal interference), the trigger judgment unit monitors continuous reading failure events and automatically starts the multi-source feature fusion process (such as light source reflection analysis, logistics path backtracking and image feature comparison) to avoid traceability interruption due to failure of a single verification method.

[0148] NFC tag data is stored and transmitted in encrypted form to prevent unauthorized access or data tampering. At the same time, the identification hash value read by NFC is compared with the original data stored in the blockchain node in real time to verify the authenticity of the ceramic identity and ensure the credibility of the data source.

[0149] In the present invention, NFC technology is used as the physical interface between the user terminal and the ceramic tag to achieve fast and low-power data collection, and cooperates with optical recognition, light source reflection and other modules to form a multi-modal verification system. Its short-range communication characteristics not only ensure the convenience of operation, but also enhance the fault tolerance of the system through redundant design (such as multi-source verification), ultimately supporting the efficiency and reliability of quality traceability in the ceramic supply chain.

[0150] JSON (JavaScript Object Notation) is a lightweight data exchange format that uses a key-value structure to represent data. It is concise and easy to parse. In the data acquisition and processing module of the present invention, the edge computing unit performs standardized conversion of process parameters such as firing temperature and molding pressure in JSON format. For example, the real-time temperature value collected by the sensor is converted into structured data of {"temperature": 1250, "unit": "℃"}, which facilitates the subsequent modules to quickly extract key parameters and improve data processing efficiency.

[0151] XML (eXtensible Markup Language) is an extensible markup language that describes the hierarchical structure and meta-information of data through custom tags. In the present invention, high-resolution image data (such as surface microscopic morphology images) encapsulates metadata in XML format.

[0152] Ceramic 3D contour data: The surface and internal structure of ceramic products are non-contact scanned by a 3D laser scanner to generate a high-precision 3D point cloud model to record the geometric shape, surface features and micro-texture of the ceramic. The data is stored in a distributed feature library in a gridded format (such as STL or OBJ) and is used in the redundant recovery module to match the local contour of the damaged mark with the dynamic time warping (DTW) algorithm to solve the problem of missing structural information caused by the fragmentation of physical marks. For example, by comparing the curvature, volume distribution and other parameters of the damaged area with the complete 3D model, potential matching batches can be screened.

[0153] X-ray fluorescence element distribution map: Use X-ray fluorescence spectrometer to perform non-destructive testing on ceramic materials to obtain its chemical elements (such as SiO 2 、Al 2 O 3 ) to generate a two-dimensional or three-dimensional element concentration map. The map is visualized through pseudo-color coding technology and stored in association with the production batch number. In the identity verification stage, the element distribution vector of the current ceramic is compared with the original map stored in the blockchain in combination with cosine similarity analysis to verify the consistency of material composition and prevent counterfeit products from entering the supply chain.

[0154] Light source reflection characteristics: The ceramic surface is illuminated by a specific wavelength (such as 650nm visible light or near-infrared light) to measure its reflection spectrum intensity, polarization characteristics and scattering angle distribution data. The reflection characteristics are strongly related to the glaze firing process. For example, differences in glaze thickness and crystal morphology can cause changes in the position or intensity of the reflection peak. In the multimodal verification module, the reflection spectrum data is spectrally matched with the firing feature database stored in the blockchain to generate a similarity score and assist in the generation of traceability conclusions.

[0155] The specific implementation method of the present invention is as follows: The data acquisition and processing module collects firing temperature, molding pressure parameters and high-resolution image data in real time through the Internet of Things devices deployed on the ceramic production line. The edge computing node converts the original data into JSON / XML format, and generates a structured data set after filtering outliers. The data set is synchronized with the process parameters and surface micromorphology data by timestamp and transmitted to the dynamic identification generation module. The dynamic identification generation module uses the SHA-3 algorithm to generate a hash summary, combines the production batch number, timestamp, and ceramic size and surface texture features to generate a variable-length blockchain identification, establishes an index association relationship with multi-source feature data, and outputs it to the identification embedding control module.

[0156] The identification embedding control module converts the dot density and code length of the blockchain identification into laser pulse frequency and engraving depth parameters, and drives the laser equipment to engrave invisible dots before the ceramic body is fired; for finished ceramics, the transparent zirconia coating composition ratio is dynamically adjusted based on the thermal expansion coefficient of the matrix, and the ultraviolet laser generates a refractive index difference type identification inside the coating. The optical detection device collects the engraving quality data in real time and feeds it back to the blockchain node to verify the integrity of the identification. The cross-chain data synchronization module writes the production data into the private chain, and synchronizes the key information to the alliance chain node through the cross-chain protocol. The smart contract verifies the time and space consistency based on the logistics track data, and freezes the data update authority when it deviates from the preset path.

[0157] When NFC or optical reading fails, the multimodal verification module triggers the light source reflection analysis unit to match the reflection spectrum with the glaze firing characteristics, the path tracing unit selects the production batches based on the geographic location, and the image feature comparison unit calculates the surface image similarity through the ResNet-50 model to generate a comprehensive traceability conclusion. The redundancy recovery module calls the 3D contour, element distribution map and reflection characteristics, uses the dynamic time warping algorithm to locally match the damaged mark, and reconstructs the complete hash chain and updates the smart contract status after verifying the global consistency through the cosine similarity analysis. The resource scheduling module dynamically allocates node computing resources based on the Q-learning algorithm, giving priority to high-confidence requests; the terminal device integration module transmits multimodal data to the blockchain node through a unified protocol to simplify the user operation process. The anti-quantum encryption module performs lattice cryptography encryption on the feature vector, and the distributed storage unit stores encrypted data through the IPFS protocol and associates it with the blockchain index to support secure data recovery.

[0158] The above implementation method systematically solves the problems of low reliability, cumbersome operation and damaged surface integrity of ceramic identification carriers through a closed-loop process of data collection, identification generation, embedded control, cross-chain synchronization, multimodal verification and redundant recovery, and meets the accuracy, efficiency and artistic protection requirements of supply chain quality traceability.

[0159] The present invention solves the technical defects of physical carriers of blockchain identification in the ceramic supply chain through a multi-module collaborative technical solution: Improved reliability: The redundant recovery module calls the ceramic multi-dimensional feature library (3D profile, element distribution map, light source reflection characteristics), combines the dynamic time warping algorithm with cosine similarity analysis, and performs local matching and global verification of damaged identification fragments. When the physical identification is damaged, the data reconstruction unit retrieves the original data copy stored redundantly in the blockchain, reconstructs the complete hash chain and updates the smart contract status, realizing self-healing recovery of identification data and avoiding the loss of traceability information due to carrier damage.

[0160] Optimized operation convenience: The terminal device integration module integrates NFC, optical recognition and light source reflection devices, and transmits multimodal data (identification code, reflection spectrum, surface image) to the blockchain node through a unified communication protocol. Users only need a single operation to trigger the multi-source verification process. The resource scheduling module dynamically allocates node computing resources based on the Q-learning algorithm, prioritizes high-confidence requests, reduces user waiting time, and simplifies interaction steps.

[0161] Surface integrity assurance: The logo embedding control module uses laser micro-engraving technology to write invisible dots inside the ceramic body, or generates a refractive index difference logo on the transparent zirconium oxide coating through ultraviolet laser. The coating composition ratio is dynamically adjusted based on the thermal expansion coefficient of the ceramic matrix to avoid logo deformation due to thermal stress during the firing process. This technology realizes the hidden embedding of blockchain logos without destroying the artistic integrity of the ceramic surface, taking into account both aesthetic value and data readability.

[0162] The above solution combines cross-chain data synchronization, multimodal verification and quantum-resistant encryption technology to form a full-link closed loop from data collection, identification generation to redundant recovery, systematically solving the collaborative problems of physical carrier reliability, operational efficiency and artistic integrity in the ceramic supply chain.

Claims

1. The ceramic supply chain quality traceability data management system based on blockchain is characterized by: include: The data acquisition and processing module is used to collect process parameters and image data, and perform format conversion and outlier cleaning through the edge computing node to generate a structured data set; A dynamic identification generation module receives the structured data set, is used to perform a hash operation on the structured data set to generate a hash summary, combines the production batch number, the timestamp and the physical characteristics of the ceramic to generate a blockchain identification of a variable length, and outputs the characteristic parameters of the blockchain identification to the identification embedding control module; The logo embedding control module is used to convert the dot density and code length of the blockchain logo into the pulse frequency and engraving depth parameters of the laser device, drive the laser micro-engraving module to engrave invisible dots inside the ceramic body or in the transparent coating, and collect the engraving quality data through the optical detection device and feed it back to the blockchain node; A cross-chain data synchronization module receives the engraving quality data and ceramic production data, is used to write the production data into the private chain and synchronize it to the alliance chain node through the cross-chain protocol, triggers the spatiotemporal data verification logic of the smart contract based on the logistics path, and freezes the batch data update permission when the logistics track deviates from the preset path; The multimodal verification module is used to parse the hash value of the ceramic identification and initiate a verification request to the blockchain node. When the identification cannot be read, it triggers the multi-source data fusion process of light source reflection feature analysis, logistics path backtracking calculation and image feature similarity comparison to generate a comprehensive traceability conclusion; The redundant recovery module is used to call the 3D contour data, element distribution map and light source reflection characteristics in the ceramic multi-dimensional feature library when the physical identification is partially damaged, match the local identification fragments through the dynamic time warping algorithm, and verify the global feature consistency in combination with the cosine similarity analysis, reconstruct the complete identification hash chain and update the blockchain node storage data.

2. The ceramic supply chain quality traceability data management system based on blockchain as claimed in claim 1, characterized in that: The data acquisition and processing module comprises: A sensor data interface unit, used to receive firing temperature, molding pressure parameters and high-resolution images collected by the IoT device, and transmit the parameters and images to the edge computing unit; an edge computing unit, receiving the parameters and the image, for performing JSON / XML format conversion, filtering outliers based on a preset threshold to generate cleaned data, and outputting the cleaned data to a data alignment unit; The data alignment unit receives the cleaned data and is used to synchronize the process parameters with the surface micro-morphology data according to the timestamp, generate a structured data set and transmit it to the dynamic identification generation module.

3. The ceramic supply chain quality traceability data management system based on blockchain as claimed in claim 1 is characterized in that: The dynamic identification generation module comprises: A hash summary generating unit receives the structured data set transmitted by the data acquisition and processing module and generates a hash summary using a SHA-3 algorithm; An identification parameter binding unit receives the hash summary and is used to associate the summary with the ceramic size and surface texture characteristics to generate a blockchain identification of variable length; The index association unit receives the blockchain identifier, is used to establish a mapping relationship between the identifier and the multi-source feature data, and outputs the mapped identifier parameters to the identifier embedding control module.

4. The ceramic supply chain quality traceability data management system based on blockchain as claimed in claim 1, characterized in that: The identification embedding control module includes: A process parameter mapping unit receives the blockchain identifier output by the dynamic identifier generation module and is used to convert the dot density and code length of the identifier into laser pulse frequency and engraving depth parameters; The laser micro-engraving execution unit receives the laser pulse frequency and engraving depth parameters, drives the laser device to engrave the internal dot matrix before firing the ceramic body according to the control instructions, and collects the engraving quality data through the optical detection device and feeds it back to the blockchain node; The coating identification adapter unit receives the thermal expansion coefficient data of the ceramic substrate, is used to dynamically adjust the composition ratio of the transparent zirconium oxide coating, and drives the ultraviolet laser to generate a refractive index difference type identification.

5. The ceramic supply chain quality traceability data management system based on blockchain as claimed in claim 1, characterized in that: The cross-chain data synchronization module includes: A private chain writing unit receives the engraving quality data fed back by the identification embedding control module and is used to write the identification data, process parameters and engraving quality data into the private chain; A cross-chain gateway unit receives the production batches and quality inspection results stored in the private chain writing unit, and synchronizes them to the alliance chain node through the cross-chain protocol; The smart contract execution unit receives the logistics trajectory data updated by the alliance chain node to verify the spatiotemporal consistency of the logistics path with the preset rules. When the path deviates, it freezes the batch data update authority and generates an early warning event.

6. The blockchain-based ceramic supply chain quality traceability data management system according to claim 1, characterized in that: The multimodal verification module comprises: A trigger judgment unit is used to monitor NFC or optical reading failure events and start the coordinated work of the light source reflection analysis unit, the path backtracking unit and the image feature comparison unit; A light source reflection analysis unit receives instructions from the trigger judgment unit and is used to match the reflection spectrum data collected by the user terminal with the glaze firing characteristics stored in the blockchain node; A path tracing unit receives the ceramic geographic location data uploaded by the user, is used to filter the production batch range based on the logistics path tracing algorithm, and outputs the screening results to the image feature comparison unit; The image feature comparison unit receives the screening result of the path tracing unit, is used to extract the ceramic surface image features based on the ResNet-50 model, calculate the similarity with the original production image stored in the blockchain, generate the traceability conclusion and output it to the redundancy recovery module.

7. The blockchain-based ceramic supply chain quality traceability data management system according to claim 1, characterized in that: The redundancy recovery module comprises: A feature library construction unit for storing ceramic 3D profile data, X-ray fluorescence element distribution maps and light source reflection characteristics; A local matching unit receives the damaged identification fragment output by the multimodal verification module, calculates its local similarity with the complete blockchain identification using a dynamic time warping algorithm, screens potential matches and outputs them to a global verification unit; A global verification unit receives the potential matches and is used to perform cosine similarity analysis on the feature vectors to verify the consistency of the ceramic identity; The data reconstruction unit receives the verified matches and is used to retrieve the original data copy stored redundantly in the blockchain, reconstruct the complete identification hash chain and update the smart contract status.

8. The blockchain-based ceramic supply chain quality traceability data management system according to claim 1, characterized in that: Also includes: The resource scheduling module connects the cross-chain data synchronization module and the multimodal verification module, is used to receive the logistics verification data of the cross-chain data synchronization module and the traceability request of the multimodal verification module, dynamically allocates the computing resources of the blockchain nodes through the Q-learning algorithm, and gives priority to high-priority requests; The terminal device integrated module integrates NFC, optical recognition and light source reflection devices, and is used to transmit the collected identification data, reflection spectrum data and image data to the blockchain node through a unified communication protocol and receive verification results.

9. The blockchain-based ceramic supply chain quality traceability data management system according to claim 8, characterized in that: The resource scheduling module includes: Load monitoring unit, used to collect real-time data on the request processing volume and response time of blockchain nodes; A priority determination unit, receiving the data of the load monitoring unit and the query confidence parameter of the multimodal verification module, for classifying the request levels; A resource allocation unit receives the request level, allocates computing resources of blockchain nodes according to the level, and gives priority to high-confidence traceability requests.

10. The ceramic supply chain quality traceability data management system based on blockchain as claimed in claim 1, characterized in that: Also includes: A quantum-resistant encryption module, connected to the dynamic identification generation module and the redundancy recovery module, configured to perform quantum-resistant signature algorithm encryption on the feature vector output by the dynamic identification generation module, and output the encrypted data to a distributed storage unit; The distributed storage unit receives the encrypted data, is used to store the encrypted feature vector through the IPFS protocol, and establishes an index association with the blockchain node to support data calling of the redundant recovery module.

Images