Blockchain-based electronic component manufacturing production data traceability system

Abstract

This invention discloses a blockchain-based traceability system for electronic component manufacturing production data, belonging to the field of industrial big data technology. It solves the technical problem of low traceability reliability caused by insufficient uniqueness of characteristic information in electronic component production data in existing technologies. The system includes: a data acquisition module for acquiring raw characteristic data of electronic components; a feature generation module for converting the raw characteristic data from the spatial domain to the transform domain through signal transformation to generate a digital feature vector for the electronic component; a blockchain notarization module for calculating the hash value of the digital feature vector and its associated production process data, and writing the hash value and key metadata together into a blockchain distributed ledger for notarization, obtaining a digital identity certificate for the electronic component; and an application verification module for receiving traceability verification requests for the production data of the electronic component to be verified, comparing the associated data on-chain and off-chain, and returning the traceability information or verification result of the component to be verified.

Description

Technical Field

[0001] This invention belongs to the field of industrial big data and involves blockchain technology, specifically a blockchain-based traceability system for electronic component manufacturing production data. Background Technology

[0002] Electronic component manufacturing is the foundation of modern industry, and the authenticity of its production data and the traceability of product quality are crucial for ensuring supply chain security, fulfilling product quality responsibilities, and combating counterfeiting. Currently, the industry typically uses external identifiers for tracing electronic component production data, namely, affixing QR codes, barcodes, or RFID tags to component packages. These identifiers act as the component's "identity card," linking them to production data generated in the Manufacturing Execution System (MES) and Enterprise Resource Planning (ERP) systems. Users scan the identifier to access a centralized database to query traceability information. However, this method has weak anti-counterfeiting capabilities; external identifiers are easily copied, transferred, or tampered with, failing to truly bind them to the component itself and easily leading to vulnerabilities where counterfeit components contain genuine information. Summary of the Invention

[0003] This invention aims to solve at least one of the technical problems existing in the prior art; to this end, this invention proposes a blockchain-based electronic component manufacturing production data traceability system to solve the technical problems of low traceability credibility caused by easy tampering of electronic component production data, incomplete traceability chain, and insufficient uniqueness of feature information in the prior art.

[0004] To achieve the above objectives, a first aspect of the present invention provides a blockchain-based electronic component manufacturing production data traceability system, comprising: The data acquisition module is used to acquire raw feature data of electronic components; the raw feature data represents the physical field signal generated by the interaction between the high-energy beam and the microstructure of the component surface; The feature generation module is used to transform the original feature data from the spatial domain to the transform domain through signal transformation, thereby generating a digital feature vector of the electronic component. The blockchain evidence storage module is used to calculate the hash value of the digital feature vector and its associated production process data, and write the hash value and key metadata together into the blockchain distributed ledger for evidence storage, thereby obtaining the digital identity certificate of the electronic component; the key metadata represents the context data generated during the evidence storage process. The application verification module is used to receive production data traceability verification requests for electronic components to be verified, compare the associated data on-chain and off-chain, and return the traceability information or verification results of the components to be verified.

[0005] Based on the above technical solution, this application transforms the inherent, non-replicable physical microscopic characteristics of the component into a unique digital feature vector, solving the anti-counterfeiting problem at its physical root and eliminating the possibility of counterfeiting by copying external identifiers. Furthermore, by using blockchain technology to anchor this digital feature to key production data, it ensures the authenticity, immutability, and traceability of data throughout the entire supply chain, providing a data foundation for quality traceability, supply chain transparency, and precise recall.

[0006] Furthermore, the process of obtaining the original feature data includes: A laser beam of a specific wavelength is used and focused by a lens group to irradiate the microscopic feature area on the surface of an electronic component; An image sensor is used to receive and record the laser scattering field pattern generated by the microstructure of the component surface to obtain raw feature data.

[0007] Furthermore, the process of generating the digital feature vector includes: Perform a two-dimensional Fourier transform on the original feature data to convert it from a spatial domain image to a frequency domain spectrum; Bandpass filtering is performed on the frequency domain spectrum to retain the mid-to-high frequency components that characterize the microstructure of the component surface; Calculate the energy intensity and distribution entropy of the filtered spectrum in multiple concentric rings and sector regions; The calculated energy value, entropy value, and moment features of the spectrum are combined in a predetermined order to form a primary feature vector.

[0008] Furthermore, the method for generating digital feature vectors of electronic components also includes: Get truly random numbers generated by a true random number generator; The primary feature vector is mixed with the true random number; A cryptographic hash operation is performed on the mixed result, and the resulting hash value is used as the digital feature vector of the electronic component.

[0009] Furthermore, the workflow of the blockchain evidence storage module includes: The system receives the digital feature vector from the feature generation module and the hash value of the associated production process data; the production process data represents various quality data generated by the electronic components during the production process, including one or more of process parameter data, equipment status data, and operation record data. A comprehensive hash value is generated by combining the hash value of the production process data with the digital feature vector. The combined hash value, the digital feature vector, the hash value of the production process data, the timestamp, and the off-chain storage address index are used to construct a transaction. The transaction is packaged into a new block using a consensus algorithm, and after being verified by all nodes in the blockchain network, the new block is appended to the end of the blockchain, completing distributed storage.

[0010] Furthermore, the packaging into a new block means that the consensus nodes in the blockchain network verify the validity of the evidence-stored transactions. After the verification is successful, a batch of transactions are packaged together with the hash value of the previous block in chronological order to form a new data block. The new data block is then broadcast to all nodes in the entire blockchain network for storage through the consensus mechanism, resulting in a chain-like storage structure.

[0011] Furthermore, the process of obtaining the traceability information includes: Receive a user's verification request, the request containing the digital feature vector of the electronic component to be verified; Based on the identification information or the digital identity credential retrieved through feature matching, query the stored record from the blockchain distributed ledger; if the stored record cannot be found, a verification failure message is returned. Based on the off-chain storage address index in the evidence storage record, retrieve the associated complete production process data from the off-chain storage system; The on-chain notarized transaction information is linked and integrated with the complete off-chain production process data to generate traceability information in chronological order.

[0012] Furthermore, the digital identity credential retrieved through feature matching includes: Calculate the similarity between the digital feature vector to be verified and each digital feature vector already stored in the blockchain; The system filters out the most similar evidence records with a similarity exceeding a preset threshold and extracts the digital identity credential corresponding to the evidence record; otherwise, it determines that the retrieval has failed and generates a verification failure message.

[0013] Furthermore, the preset threshold is an adaptive threshold that is dynamically adjusted based on historical verification data, and the dynamic adjustment method includes: Periodically collect similarity scores from successfully matched historical verification records; Based on the distribution of the similarity values, a dynamic threshold that can maximize the distinction between genuine and counterfeit electronic components is calculated using a machine learning model. Update the dynamic threshold to the preset threshold used in the next verification.

[0014] Furthermore, the production data traceability system also includes a verification terminal module, the workflow of which includes: The built-in miniature laser emitter and image sensor scan a predetermined feature area of ​​the electronic component to be verified to obtain an image of the component to be verified. The same signal transformation and feature extraction process as the feature generation module is performed on the image to be verified to generate a digital feature vector to be verified. The digital feature vector to be verified is sent to the application verification module via a wireless network to initiate a verification request; Receive and parse the verification results returned by the application verification module, and intuitively display the verification conclusion to the user through a display screen or indicator light.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This application collects physical field signals generated by the microstructure of the component itself and converts them into frequency domain feature vectors. It uses the component's physical non-cloning characteristic as its unique identifier, offering greater security than traditional external identifiers such as QR codes and RFID tags. By combining blockchain technology with immutable storage of digital feature vectors and production process data, a trusted data chain spanning the entire component lifecycle is established, enabling refined management at the individual product level. This provides a reliable technical means for product quality traceability, supply chain transparency, and anti-counterfeiting. Furthermore, by introducing a threshold adjustment mechanism based on machine learning-based historical data, this application allows the system to self-optimize and continuously improve identification accuracy, effectively distinguishing genuine products from counterfeit ones, thus solving the anti-counterfeiting and security trust issues in electronic component traceability. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A schematic diagram of the framework of a blockchain-based electronic component manufacturing production data traceability system provided in an embodiment of the present invention; Figure 2 A flowchart illustrating the feature generation module in a blockchain-based electronic component manufacturing production data traceability system provided in this embodiment of the invention; Figure 3 A schematic diagram of the blockchain evidence storage module in the blockchain-based electronic component manufacturing production data traceability system provided in this embodiment of the invention; Figure 4 This is a flowchart illustrating the application of a verification module in a blockchain-based electronic component manufacturing production data traceability system provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] To address the technical problems of low traceability reliability caused by easily tampered electronic component production data, incomplete traceability chains, and insufficient uniqueness of feature information in existing technologies, this application provides a blockchain-based electronic component manufacturing production data traceability system, which includes: The data acquisition module is used to acquire raw characteristic data of electronic components; the raw characteristic data represents the physical field signal generated by the interaction between the high-energy beam and the microstructure of the component surface. The feature generation module is used to transform the original feature data from the spatial domain to the transform domain through signal transformation, thereby generating digital feature vectors for electronic components. The blockchain evidence storage module is used to calculate the hash value of the digital feature vector and its associated production process data, and write the hash value and key metadata together into the blockchain distributed ledger for evidence storage, thereby obtaining the digital identity certificate of the electronic component. The application verification module is used to receive production data traceability verification requests for electronic components to be verified, compare the associated data on-chain and off-chain, and return the traceability information or verification results of the components to be verified.

[0020] Based on this, this application combines the unique identifier of micro-features with the immutability of blockchain to achieve reliable traceability of electronic component data throughout its entire lifecycle, thereby improving the reliability and traceability efficiency of production data.

[0021] like Figure 1 As shown in the embodiment of this application, the blockchain-based electronic component manufacturing production data traceability system includes: a data acquisition module, a feature generation module, a blockchain evidence storage module, and an application verification module, which are connected in communication; wherein, The data acquisition module is used to collect raw characteristic data of electronic components.

[0022] Among them, the original feature data represents the physical field signal generated by the interaction between the high-energy beam and the microstructure of the component surface, including but not limited to the scattered field signal generated by the interaction between the laser beam and the component surface, the electromagnetic field signal generated by the electron beam bombarding the component surface, the acoustic field signal generated by the interaction between the ion beam and the component surface, and the thermal radiation field signal generated by the interaction between the infrared beam and the component surface.

[0023] In some implementations, different types of high-energy beam generating devices can be used, such as laser generators, electron beam generators, ion beam generators, and infrared beam emitters, in conjunction with corresponding signal acquisition devices, such as image sensors, spectrometers, magnetic field sensors, acoustic detectors, and thermal imagers, to act on and acquire signals from the microstructure of different regions on the surface of electronic components. During the acquisition process, the wavelength, intensity, and irradiation angle of the high-energy beam, as well as the sampling frequency and resolution of the signal acquisition devices, can be adjusted to meet the acquisition needs of electronic components with different materials and structures.

[0024] It should be noted that the collected original feature data must be able to reflect the inherent properties of the microstructure on the surface of the electronic component, and these properties must be non-replicable and stable, unaffected by minor wear and tear or changes in ambient temperature and humidity during the subsequent normal use of the electronic component, so as to ensure that the original feature data can serve as the basis for the unique identification of the electronic component.

[0025] For example, a 532nm green laser generator is used in conjunction with a 12-megapixel image sensor to focus the laser beam onto the ceramic package surface of a surface-mount resistor. The laser irradiation angle is adjusted to 45°, and the image sensor sampling frequency is set to 30fps. The scattered field image signal generated by the interaction between the laser and the micro-cracks and impurities on the resistor surface is collected as the original feature data of the resistor. Alternatively, an electron beam generator is used to generate an electron beam with an energy of 10keV to bombard the exposed silicon wafer surface of an integrated circuit chip. The electromagnetic field signal generated by the interaction between the electron beam and the lattice structure of the silicon wafer surface is collected by a magnetic field sensor as the original feature data of the integrated circuit.

[0026] The feature generation module is used to transform the original feature data from the spatial domain to the transform domain through signal transformation, thereby generating digital feature vectors for electronic components.

[0027] Among them, the digital feature vector contains the unique identification information of the surface microstructure of the electronic component and the key feature parameter information of the original feature data. It is the unique identity representation of the electronic component in the digital space, which can be used to distinguish different individual electronic components. It also has a fixed length and standardized format, which facilitates subsequent hash calculation and blockchain storage.

[0028] In some implementations, signal transformation includes, but is not limited to, Fourier transform, wavelet transform, Laplace transform, Hilbert transform, cosine transform, etc.; filtering and noise reduction processing can be combined during the transformation process; the generation of digital feature vectors can further include steps such as feature dimensionality reduction, feature normalization, and feature selection.

[0029] It should be noted that the algorithm for generating digital feature vectors must ensure stability. That is, the similarity of feature vectors generated by the same electronic component under different acquisition conditions must be higher than a preset threshold, while the similarity of feature vectors of different electronic components must be lower than a preset threshold. This is directly related to the accuracy of subsequent traceability verification. Insufficient stability will lead to the same component being misjudged as different individuals, while insufficient distinguishability will lead to different components being unable to be effectively distinguished. Both of these will undermine the reliability of the traceability system.

[0030] For example, wavelet transform is performed on the original light scattering signal to obtain time-frequency domain features. Principal component analysis is used to reduce the original 1024-dimensional features to 256-dimensional features. After Z-score normalization, the digital feature vector of the electronic component is obtained.

[0031] The blockchain evidence storage module is used to calculate the hash value of the digital feature vector and its associated production process data, and write the hash value and key metadata together into the blockchain distributed ledger for evidence storage, thereby obtaining the digital identity certificate of the electronic component.

[0032] Among them, key metadata represents the context data generated during the evidence preservation process, including but not limited to evidence preservation timestamps, evidence preservation node information, data source identifiers, basic information of electronic components, evidence preservation type identifiers, etc.; the associated production process data includes various types of data generated by electronic components in the manufacturing process, such as process parameter data, equipment status data, operation record data, raw material data, etc.

[0033] In some implementations, the key metadata includes, but is not limited to, the evidence storage timestamp and the evidence storage initiation time and block packaging time; the evidence storage node information includes, but is not limited to, the IP address of the production workshop server that initiated the evidence storage and the blockchain node ID; the data source identifier includes, but is not limited to, the equipment number that collected the original feature data and the production line number to which the production process data belongs; the basic information of the electronic components includes, but is not limited to, the component model, production batch, factory number, and material specifications; and the evidence storage type identifier includes, but is not limited to, the first production data evidence storage and the subsequent circulation data supplementary evidence storage.

[0034] In some implementations, the associated production process data includes, but is not limited to, welding temperature, encapsulation pressure, curing time, and etching depth; equipment status data includes but is not limited to production equipment operating speed, voltage, current, and fault records; operation record data includes but is not limited to operator ID, operation time, quality inspector ID, and quality inspection time; and raw material data includes but is not limited to raw material supplier information, raw material batch number, and raw material test report number.

[0035] It should be noted that the hash values ​​written to the blockchain can adopt a chained hash structure, that is, the hash value of the next link contains the hash value of the previous link, forming a complete traceability chain. This structure ensures that if the data in any link is tampered with, all subsequent hash values ​​will become invalid, thereby realizing data integrity verification throughout the entire process. The storage of key metadata can adopt the method of direct on-chain storage or off-chain storage (such as a distributed file system) associated with on-chain hashes. This can not only ensure the immutability of key information, but also avoid the storage pressure and performance loss caused by directly uploading a large amount of data to the chain.

[0036] For example, the SHA-256 algorithm is used to calculate the joint hash value of the digital feature vector with the raw material batch, welding temperature, and test results. This hash value, along with the evidence storage timestamp, production line number, and component serial number, is written into the blockchain through a consortium blockchain smart contract to generate a digital identity credential containing a block height of 10582 and a transaction ID of "0xabc...".

[0037] The application verification module is used to receive production data traceability verification requests for electronic components to be verified, compare the associated data on-chain and off-chain, and return the traceability information or verification results of the components to be verified.

[0038] The traceability information includes end-to-end data for electronic components, from raw material procurement to manufacturing, quality inspection, and final shipment. This includes information such as the name and qualifications of raw material suppliers, process parameters and operation records for each production stage, details of quality inspection reports, production equipment maintenance records, and shipment inspection dates and standards. Verification requests include, but are not limited to, the digital feature vector of the electronic component to be verified, the basic information of the component, the verification initiator information, and the verification request type. Verification results include successful and unsuccessful verification results. Verification request types include end-to-end traceability, data verification for specific stages, and authenticity verification.

[0039] In some implementations, the comparison methods include, but are not limited to, direct hash value comparison, which compares whether the hash value calculated on the data to be verified is consistent with the hash value stored on the chain; feature vector similarity comparison, which calculates the matching degree between the feature vector to be verified and the associated feature vector on the chain through cosine similarity, Euclidean distance, etc.; and metadata consistency verification, such as verifying the logical order of timestamps and the legality of device identifiers.

[0040] It should be noted that the application verification module can support verification requests with different permissions. For example, ordinary users can only query the verification results, while administrators can view detailed traceability data. This can not only meet the usage needs of different users, but also protect sensitive information in the production process through permission control. The comparison process can adopt a combination of offline calculation and on-chain query, which can reduce the frequency of direct interaction with the blockchain, reduce network transmission pressure, improve verification response speed, and ensure that the system can still run stably in large-scale verification scenarios.

[0041] For example, the system receives the digital feature vector and serial number of the element to be verified, calculates its hash value, compares it with the hash value corresponding to the serial number in the blockchain, and calculates the matching degree of the feature vector to be verified with the associated feature vector on the chain through cosine similarity. The system returns the verification result and a complete traceability path including the raw material supplier, processing time, and testing personnel.

[0042] Based on the above technical solutions, the blockchain-based electronic component manufacturing production data traceability system provided in this application ensures the uniqueness and non-replicability of the features by collecting physical field signals generated by the microscopic interactions of electronic components as the original features; it realizes the transformation from physical features to digital identifiers by generating digital feature vectors through signal transformation; it uses blockchain to store feature data and production process data to ensure the immutability and traceability of the data; and it achieves verification through comparison of on-chain and off-chain data, improving the credibility and efficiency of traceability and effectively solving the problems of data reliability and chain integrity in electronic component production data traceability.

[0043] In one possible implementation of this application embodiment, the data acquisition module can be specifically implemented through the following S101 and S102, which are described in detail below: S101. A laser beam of a specific wavelength is used and focused by a lens group to irradiate the microscopic feature area on the surface of the electronic component.

[0044] Among them, a specific wavelength laser beam refers to a laser wavelength selected based on the optical reflection and absorption characteristics of the surface material of electronic components, which can effectively interact with the microstructure of the component surface and generate a clear scattering signal; the lens group is used to focus the laser beam, reduce the laser spot diameter, ensure that the laser energy is concentrated on the micro-area of ​​the component surface, and improve the spatial resolution of the signal; the micro-feature area of ​​the electronic component surface refers to the micro-structure area of ​​the component surface that is naturally formed or generated during the production process and has uniqueness and cannot be replicated, such as the pin area, package surface area, chip exposed area, impurity distribution, crystal arrangement difference, etc. of electronic components, rather than artificially printed QR codes, labels and other external marking areas.

[0045] In some implementations, the choice of laser beam wavelength needs to be adapted to the material characteristics of the component surface: if the electronic component surface is made of metal, a green laser with a wavelength of 532nm or a red laser with a wavelength of 650nm can be selected. These wavelengths of laser have moderate reflectivity on the metal surface and can effectively highlight the microscopic features of the metal surface such as scratches and oxidation marks; if the component surface is made of ceramic, a near-infrared laser with a wavelength of 1064nm can be selected. Ceramics absorb less near-infrared laser and scatter stronger signals, making it easier to capture the tiny pores and impurity distribution on the ceramic surface; if the component surface is made of plastic packaging material, a purple laser with a wavelength of 405nm can be selected. Plastic has a more obvious scattering effect on short-wavelength lasers and can clearly present the injection molding texture and tiny depressions on the plastic surface. Meanwhile, the lens group can be a combination of 2-3 achromatic lenses, where the front lens is a plano-convex lens for initial laser focusing, and the rear lens is a biconvex lens to further reduce the spot diameter. The focal length of the lens group can be adjusted according to the size of the microscopic area to be collected, typically choosing a lens group with a focal length of 10mm-50mm to control the diameter of the focused laser spot within the range of 10μm-100μm, accurately covering the microscopic feature areas of the component surface. Furthermore, an adjustable aperture can be added between the lens group and the component. By adjusting the aperture size, the energy of the incident laser can be controlled, avoiding damage to the component surface microstructure due to excessive laser energy or weak scattered signals due to insufficient energy.

[0046] It is important to note that when determining the microscopic feature regions on the surface of electronic components, easily worn areas should be avoided, such as pin insertion / removal contact areas and casing edges. Instead, relatively stable areas on the component surface, less susceptible to physical changes from subsequent use or handling, such as the central area or non-functional areas of the component package, should be selected to ensure the long-term stability of the acquired microscopic features. Simultaneously, the laser beam irradiation angle must be kept stable, typically controlled between 30° and 60°. Within this angle range, the interaction between the laser and the microstructure is strongest, resulting in the richest scattered field signal. If the irradiation angle is too small (less than 15°), specular reflection of the laser is likely, resulting in a weak scattered signal; if the angle is too large (greater than 75°), the laser energy is easily absorbed by the component surface, also affecting signal quality.

[0047] For example, for a ceramic-packaged 0805 type surface mount resistor with an alumina ceramic surface material, a near-infrared laser generator with a wavelength of 1064nm is selected, paired with a lens group consisting of a plano-convex lens with a focal length of 20mm and a biconvex lens with a focal length of 15mm. The distance between the lens group and the resistor surface is adjusted to 35mm. An adjustable aperture with an aperture of 2mm is installed between the lens group and the resistor to control the laser energy at 5mW. The resistor is fixed by a displacement platform, and the displacement platform is adjusted to focus the laser beam on the central area of ​​the resistor surface. The angle between the laser beam and the resistor surface is set to 45°. The final focused laser spot diameter is approximately 30μm, accurately irradiating the ceramic microstructure area on the resistor surface, preparing for subsequent scattering signal acquisition.

[0048] S102. Use an image sensor to receive and record the laser scattering field pattern generated by the microstructure of the component surface to obtain the original feature data.

[0049] Among them, the image sensor is the core device used to convert light signals into electrical signals and then generate digital images; the laser scattering field pattern refers to the light field pattern with unique distribution characteristics formed after the laser beam shines on the microstructure of the component surface, due to the reflection, refraction and diffraction of the microstructure. This pattern corresponds one-to-one with the microstructure and is unique; while the original feature data is the digital image data output by the image sensor that contains all the information of the laser scattering field pattern. It is the basic data source for the subsequent generation of digital feature vectors of electronic components.

[0050] In some implementations, the type of image sensor can be selected based on the laser wavelength and the intensity of the scattered signal: if visible or near-infrared lasers are used, a complementary metal-oxide-semiconductor (CMOS) image sensor can be chosen, suitable for batch acquisition scenarios; if the laser wavelength is in the ultraviolet or mid-to-far-infrared band, or the scattered signal is weak, a charge-coupled device (CCD) image sensor can be selected, which can better capture weak scattered signals. Furthermore, to reduce interference from ambient light on the scattered field pattern, a narrowband filter matching the laser wavelength can be added in front of the image sensor lens to filter stray light from the environment; the storage format for the raw feature data can be RAW format, preserving all the original pixel information acquired by the image sensor; if a balance between storage capacity and data integrity is required, a lossless compression format can also be selected.

[0051] It should be noted that the image sensor's shooting angle and position need to be adjusted in conjunction with the laser beam's illumination angle. Typically, the optical axis of the image sensor should be aligned with the main scattering direction of the laser scattering field to ensure the sensor can receive the scattered signal to the maximum extent. For example, if the laser beam illuminates the element surface at a 45° angle, the image sensor can be positioned at a 45°-60° angle to the element surface, directly facing the main distribution area of ​​the scattering field. Furthermore, the original feature data must be validated for integrity. This validation includes checking for dead pixels, overexposed or underexposed areas, and whether the image resolution meets preset requirements. If the validation fails, steps S101 and S102 must be re-executed to prevent invalid data from entering subsequent processes.

[0052] For example, based on the acquisition scenario of the ceramic packaged surface mount resistor in S101, the specific operation of S102 is as follows: Select a 16-megapixel CMOS image sensor, set the sensor lens focal length to 50mm, and install a narrow-band filter with a center wavelength of 1064nm and a bandwidth of 8nm in front of the lens; adjust the sensor position so that the sensor optical axis forms a 50° angle with the resistor surface, and set the distance between the sensor and the resistor surface to 80mm; set the sampling frequency to 3fps, and perform dark current correction and white balance adjustment on the electrical signal output by the sensor; finally, the sensor outputs a 16-megapixel, 14-bit depth RAW format image, which is the original feature data of the surface mount resistor.

[0053] Based on the above technical solution, the raw feature data collected by the data acquisition module can truly reflect the inherent physical properties of the components, eliminating the problem of easy copying of external identifiers from the source, and laying a physical data foundation for the subsequent conversion of physical features into digital feature vectors and the realization of blockchain-based trusted traceability.

[0054] In one possible implementation of the embodiments of this application, combined with Figure 1 ,like Figure 2 As shown, the feature generation module described above can be implemented through the following steps S201, S202, S203, and S204, which are explained in detail below: S201. Perform a two-dimensional Fourier transform on the original feature data to convert it from a spatial domain image to a frequency domain spectrum.

[0055] Two-dimensional Fourier transform is a mathematical transformation method that converts image data based on pixel location in the spatial domain into data based on frequency distribution in the frequency domain. It can decompose the spatial features of an image into sinusoidal wave components of different frequencies. The frequency domain spectrum is the output of the two-dimensional Fourier transform. With frequency as the coordinate axis, the amplitude of each frequency point represents the contribution intensity of that frequency component in the original spatial image. The low-frequency components correspond to the overall contour information of the image, while the mid-to-high frequency components correspond to the detailed features of the image. These detailed features are the core manifestation of the uniqueness of the microstructure of the surface of electronic components.

[0056] In some implementations, the original spatial image needs to be preprocessed before performing the two-dimensional Fourier transform: first, the image is converted to grayscale, converting multi-channel pixel values ​​into single-channel grayscale values ​​to simplify subsequent calculations; second, zero-padding is performed on the grayscale image to expand the image size to an integer power of 2; finally, a Hanning window or Hamming window function can be applied to the padded image to suppress spectral leakage caused by signal discontinuities at the image edges and improve the accuracy of the frequency domain spectrum.

[0057] It should be noted that the core function of the two-dimensional Fourier transform is to realize the domain conversion from the spatial domain to the frequency domain. Its essence is to transform the spatial distribution features of the microstructure in the original image, which are difficult to quantify directly, into frequency features that can be quantitatively analyzed. Different microstructures on the surface of electronic components will result in significant differences in the amplitude and phase distribution of the mid-to-high frequency components in the corresponding frequency domain spectrum. This lays the foundation for the subsequent extraction of unique digital features.

[0058] For example, regarding the 16-megapixel RAW format ceramic resistivity scattering field spatial image acquired by the data acquisition module, the specific operation of S201 is as follows: First, convert the RAW format image to an 8-bit grayscale image, and then calculate the grayscale value using a weighted average method. Then, the grayscale image is zero-padded and expanded to 4096×4096 pixels, and the pixel value of the filled area is set to 0; Next, a Hamming window function is applied to suppress edge spectral leakage; Finally, a two-dimensional FFT is performed based on the Cooley-Tukey algorithm. First, a one-dimensional FFT is performed on each of the 4096 rows of the image, and then a one-dimensional FFT is performed on each of the 4096 columns to obtain a frequency domain spectrum of 4096×4096. By transforming the coordinates (u,v) of each pixel in the spectrum into (uN / 2,vM / 2), the spectrum is centered, and finally a frequency domain spectrum image with the DC component at the center and low-frequency to high-frequency components radiating outward is obtained.

[0059] S202. Bandpass filtering is performed on the frequency domain spectrum to retain the mid-to-high frequency components that characterize the microstructure of the component surface.

[0060] Among them, the mid-to-high frequency components refer to the frequency range components in the frequency domain spectrum that have frequencies higher than the DC component and can reflect the details of the microstructure on the surface of electronic components. These are key information for distinguishing different electronic components.

[0061] In some implementations, a Gaussian bandpass filter can be selected as the bandpass filter. The specific operation process is as follows: the frequency domain spectrum is multiplied pixel by pixel with the frequency response function of the Gaussian bandpass filter to obtain the filtered frequency domain spectrum. In this spectrum, the amplitudes of the mid-to-high frequency components in the passband remain unchanged, while the amplitudes of the low-frequency and high-frequency components in the stopband are attenuated to near 0.

[0062] For example, based on the 4096×4096 pixel centered frequency domain spectrum obtained in S201, the specific operation of S202 is as follows: Select a Gaussian bandpass filter and set the low cutoff frequency f. L =205 pixel units, high cutoff frequency f H =1638 pixels; the frequency response function of the Gaussian bandpass filter is Where D(u,v) is the Euclidean distance from pixel (u,v) to the center in the frequency domain, and D0=(f L +f H ) / 2 = 921.5 pixel units, which is the passband center frequency, σ = (f H -f L H(u,v) = 358.25 pixels, which is the passband width parameter. Multiply the amplitude of each pixel in the frequency domain spectrum by H(u,v) to obtain the filtered frequency domain spectrum. In this spectrum, the amplitude of low-frequency components with frequencies below 205 pixels is attenuated to less than 1% of the original amplitude, and the amplitude of high-frequency components with frequencies above 1638 pixels is attenuated to less than 5% of the original amplitude. Only the mid-to-high frequency components (corresponding to the tiny pores and crystal boundaries on the surface of the ceramic resistor) between 205 and 1638 pixels are retained.

[0063] S203. Calculate the energy intensity and distribution entropy of the filtered spectrum in multiple concentric rings and sector regions, and combine the calculated energy values, entropy values, and moment characteristics of the spectrum in a predetermined order to form a primary feature vector.

[0064] Among them, concentric ring zones refer to annular regions divided by different radii with the center of the frequency domain spectrum as the center, used to extract mid-to-high frequency features in different frequency ranges; fan-shaped regions refer to fan-shaped sub-regions that are evenly divided by angles from each concentric ring zone, used to extract features in different directions within the same frequency range; energy intensity refers to the sum of the squares of the amplitudes of all pixels in the region, reflecting the overall contribution of the frequency components in the region; distribution entropy value refers to the information entropy of the amplitude distribution of pixels in the region, reflecting the uniformity of the frequency component distribution in the region; moment feature refers to the statistical moment of the frequency domain spectrum; the primary feature vector is a multi-dimensional vector formed by arranging energy value, entropy value, and moment feature according to preset rules. It is the "prototype" of digital features of electronic components and contains the core quantitative information of the microstructure.

[0065] In some implementations, the number of concentric rings can be set to 5-8, with the radius increasing proportionally to ensure coverage of the entire mid-to-high frequency passband range; each concentric ring can be evenly divided into 8-12 sector regions according to the angle to capture the directional features of the microstructure.

[0066] In some implementations, energy intensity is calculated as follows: for each annular or sector-shaped region, all pixels within the region are traversed, the square of the amplitude of each pixel is calculated, and then the sum is obtained to obtain the energy value of the region; the distribution entropy is calculated as follows: the pixel amplitudes within the region are divided into 16-32 gray levels, the proportion of pixels at each gray level is counted, and then the information entropy formula H=-Σ(p i ×log2p i Calculate the entropy value; where p i Let represent the proportion of the i-th gray level. Then, calculate the first, second, and third moments of the spectrum, and define the combination order of the primary feature vectors as "ring region energy value → sector region entropy value → moment feature". For example, 5 ring region energy values ​​+ 40 sector region entropy values ​​(5 rings × 8 sectors) + 3 moment features form a 48-dimensional primary feature vector.

[0067] For example, based on the filtered frequency domain spectrum obtained in S202, the specific operation of S203 is as follows: 1. Divide the spectrum into regions: Using the center of the spectrum as the center, divide it into 5 concentric rings. Each ring is divided into 8 sector regions at a 45° angle, for a total of 5×8=40 sector regions; 2. Calculate the energy values: Calculate the energy values ​​for each of the 5 rings to obtain 5 energy values ​​from E1 to E5, and normalize them to the 0-1 interval; 3. Calculate the entropy value: Calculate the distribution entropy value for each of the 40 sector regions to obtain 40 entropy values ​​from H1 to H40, and normalize them to the interval of 0-5; 4. Calculate the moment characteristics: Calculate the first moment μ1 = 1200 pixels, the second moment μ2 = 350 pixels, and the third moment μ3 = -0.2 of the entire filtered spectrum, and normalize them to the 0-1 interval; 5. Combination vectors: Arrange them in the order of “E1-E5→H1-H40→μ1-μ3” to form a primary feature vector of 5+40+3=48 dimensions.

[0068] S204. Obtain a true random number from a true random number generator, mix the primary feature vector with the true random number, perform a cryptographic hash operation on the mixed result, and use the resulting hash value as the digital feature vector of the electronic component.

[0069] Among them, a true random number generator is a device that generates random numbers based on physical noise. The true random numbers it generates are unpredictable and non-repeatable, and are more secure than pseudo-random numbers based on algorithms. Cryptographic hashing is an algorithm that converts input data of arbitrary length into fixed-length, irreversible output. Digital feature vectors, which are the output results of hashing, are the unique identifiers of electronic components in digital space. They have a fixed length and can be directly used for blockchain notarization and subsequent verification.

[0070] In some implementations, the true random number generator can be a semiconductor thermal noise-based generator or a quantum random effect-based generator, and the length of the generated true random numbers must be consistent with the dimension of the primary eigenvector. The mixed operation can be an XOR operation or an element-wise weighted sum, as shown in the formula: Among them, V prim (i) is the i-th element of the primary feature vector, R(i) is the i-th truly random number, and V mix (i) represents the i-th element after mixing; or a weighted summation method can be used, with the formula as follows: α is the weighting coefficient, usually set to 0.6-0.8, to ensure the dominance of the primary feature vector.

[0071] For example, based on the 48-dimensional primary feature vector obtained in S203, the specific operation in S204 is as follows: 1. Obtaining truly random numbers: A truly random number generator is used to generate 48 32-bit truly random numbers; 2. Mixed operation: Weighted addition is selected, with weight coefficient α=0.7, resulting in a 48-dimensional mixed vector V. mix ; 3. Hash operation: Convert the 48-dimensional mixed vector V... mix Convert it to a binary data stream, input it into the SHA-256 algorithm for processing, and obtain a 256-bit hash value; 4. Integrity verification: Calculate the XOR sum of the first 8 bits and the last 8 bits of the hash value. If the XOR sum equals the preset value, the verification passes. This 256-bit hash value is the digital feature vector of the electronic component.

[0072] Based on the above technical solution, the feature generation module transforms the physical microscopic features of electronic components into digital feature vectors through domain transformation, filtering and purification, feature quantization, and random mixing and hashing operations. Specifically, steps S201-S203 extract quantified features that characterize the uniqueness of the microstructure from the original scattering field image through Fourier transform, bandpass filtering, and multi-dimensional feature calculation; step S204 further enhances the anti-counterfeiting and security of the digital feature vectors through true random number mixing and cryptographic hashing. The entire process preserves the uniqueness of the physical microscopic features of the electronic components while endowing them with immutability through digital technology, providing secure and unique core data for the subsequent blockchain-based evidence storage module to construct digital identity credentials for the electronic components.

[0073] In one possible implementation of the embodiments of this application, combined with Figure 1 ,like Figure 3 As shown, the aforementioned blockchain evidence storage module can be implemented through the following steps: S301, S302, S303, and S304. A detailed explanation follows: S301, Receive the digital feature vector output by the feature generation module, and the hash value of the production process data associated with the digital feature vector.

[0074] Among them, the hash value of the production process data is a fixed-length value obtained by hashing various quality data generated in the production process of electronic components, which is used to verify the integrity of the production process data; the association refers to the one-to-one correspondence between the digital feature vector and the production process data, ensuring that the digital identity of a certain electronic component is bound to its entire production process data.

[0075] In some implementations, the hash value of production process data can be generated using a "block hash + summary hash" method: first, the production process data is divided into blocks according to type or size, the hash value of each block is calculated, and then all the block hash values ​​are concatenated in a fixed order to calculate the summary hash.

[0076] For example, the feature generation module sends data to the blockchain evidence storage module via the gRPC interface: 1. The digital feature vector is a 256-bit hexadecimal value; 2. The hash value of the production process data is the SHA-256 summary hash: First, calculate the SHA-256 hash of the resistor's welding process parameters (220℃, 30s), equipment status (equipment EQ-202405, current 5A), and operation record (OP-123 operated on 2024-05-20 14:20) into 3 blocks. Then, concatenate the 3 block hashes to calculate the summary hash "0x9A2B3C4D5E6F7G8H...1J2K3L4M5N6O". After receiving the data, the blockchain evidence storage module verifies the length of the digital feature vector, the hash algorithm identifier of the production process data, and verifies the RSA digital signature of the feature generation module. Once the verification is successful, the data set is stored.

[0077] S302. Combine the hash value of the production process data with the digital feature vector according to preset rules to jointly calculate and generate a comprehensive hash value.

[0078] The comprehensive hash value is the result of hashing the combined data again. Its function is to further bind the digital identity of electronic components with the production process data to form an indivisible data unit. If either the digital feature vector or the hash value of the production process data is tampered with, the comprehensive hash value will change significantly, thereby quickly identifying data tampering behavior.

[0079] In some implementations, the default rule can be: "numerical feature vector + hash value of production process data". If the production data is highly sensitive, the numerical feature vector and the hash value of the production process data can be symmetrically encrypted separately before combination, and then the combined hash can be calculated on the encrypted results to further improve data security. In addition, a "salt value" (a randomly generated 128-bit string) can be introduced to participate in the combined hash calculation. The salt value is generated in real time with each calculation and stored together with the combined hash value to prevent attackers from cracking the original data through rainbow tables.

[0080] For example, the following operations are performed based on the data received in S301: 1. The preset combination rule is "digital feature vector + hash value of production process data", which, after concatenation, yields 512 bits of binary data; 2. Generate a 128-bit salt value in real time, and append the salt value to the end of the 512-bit data to form a 640-bit data; 3. The SHA-256 algorithm is used to perform a hash operation on the 640-bit data to obtain a 256-bit comprehensive hash value; 4. Associate the salt value with the comprehensive hash value and store it in the log. Record the following information: "Input: Numeric feature vector + production process hash, Algorithm: SHA-256, Salt value: 0x1234...4321, Result: 0x7E9F...4M5N".

[0081] S303, a blockchain transaction is constructed by combining the comprehensive hash value, digital feature vector, hash value of production process data, timestamp, and off-chain storage address index.

[0082] Among them, the off-chain storage address index is an address identifier pointing to the complete production process data in the off-chain storage system. Due to the high cost of blockchain storage, only the key hash and index are stored, and the complete data is stored off-chain. The complete data can be retrieved quickly through the index. The blockchain transaction is the basic unit of data transmission and storage in the blockchain network. It contains a fixed field structure to ensure that the data is identifiable and verifiable between nodes.

[0083] In some implementations, the field structure of a blockchain transaction typically includes: version number, input field, data field, metadata field, and length field. The input field records the data source; the data field stores the composite hash value, digital feature vector, and hash values ​​of the production process data; the metadata field stores the timestamp, off-chain storage address index, and the transaction initiator's digital signature; and the length field records the total number of bytes in the transaction.

[0084] For example, based on the comprehensive hash value and related data of S302, a transaction in Hyperledger Fabric format is constructed: 1. Version number: 0x01; 2. Input field: Feature generation module node ID "Node-Feature-05"; 3. Data fields: Comprehensive hash value "0x7E9F...4M5N", digital feature vector "0x5F3A...D1F2", production process data hash "0x9A2B...5N6O"; 4. Metadata fields: Timestamp "2024-05-2014:35:22.156", off-chain storage address index "http: / / 192.168.1.100 / production / R-20240520-0001.json", transaction initiator's ECC signature "0xAB12...CD34"; 5. Length field: Total number of bytes in the transaction "1024 Bytes"; After construction, the transaction is serialized in Protocol Buffers format for easy network transmission.

[0085] S304. The transaction is packaged into a new block using a consensus algorithm. After verification by all nodes in the blockchain network, the new block is appended to the end of the blockchain, completing the distributed storage.

[0086] Among them, consensus algorithm is the rule for all nodes in the blockchain network to achieve data consistency, ensuring that malicious nodes cannot forge or tamper with blocks; packaging into a new block means that consensus nodes combine multiple transactions that meet the conditions in chronological order with the hash value of the previous block to form a new data block; node verification means that all nodes in the blockchain network verify the legality of the new block, and only blocks that pass the verification of all nodes can be appended to the blockchain; distributed storage means that the new block is synchronously stored to all nodes in the blockchain network to achieve data redundancy and resistance to single points of failure.

[0087] In some implementations, node verification includes: verifying the validity of the transaction signature, verifying that the comprehensive hash value matches the calculated result of the digital feature vector and the hash of the production process data, and verifying that the hash of the previous block in the block header matches the hash of the last block on the chain. After successful verification, all nodes append the new block to their local blockchain replica, completing distributed storage. If a node fails verification, it needs to request the block data from other nodes for re-verification to ensure data consistency across the entire network.

[0088] For example, if a blockchain network consists of 5 enterprise nodes, the PBFT consensus algorithm can be used to perform the following operations: First, the consensus node selects a number of legitimate transactions from the transaction pool, arranges them in chronological order, and calculates the Merkle root of all transaction hashes. Then a new block is constructed, in which the block header contains the hash of the previous block, the Merkle root, and the block timestamp, and the block body contains 100 transactions; The proposer then broadcasts the new block to the other four nodes, each of which verifies the transaction signature, the overall hash value, and the hash of the previous block; if all four nodes return "verification passed", consensus is confirmed. Finally, all nodes append the new block to the end of their local blockchain, synchronously updating the chain structure and completing the distributed storage of the electronic component's evidence data.

[0089] Based on the above technical solution, the blockchain-based evidence storage module achieves trusted evidence storage of digital characteristics of electronic components and production process data. Specifically, S301 ensures the legality and correspondence of data sources, S302 strengthens data binding and anti-tampering capabilities, S303 optimizes data storage structure, and S304 achieves data immutability and network-wide consistency through consensus and distributed storage. The entire process deeply integrates the physical characteristics of electronic components, production data, and blockchain technology, forming a trusted link that runs through "physical characteristics - digital identity - production data - evidence storage block." This provides an immutable data foundation for the traceability and verification of subsequent application verification modules, while simultaneously balancing evidence storage efficiency, data security, and storage costs.

[0090] In one possible implementation of the embodiments of this application, combined with Figure 1 ,like Figure 4 As shown, the above application verification module can be implemented through the following steps S401, S402, and S403, which are explained in detail below: S401. Receive the production data traceability verification request of the electronic component to be verified, calculate the similarity between the digital feature vector to be verified and the digital feature vector already stored in the blockchain, and screen the stored records that meet the conditions to obtain digital identity credentials.

[0091] Among them, the production data traceability verification request is a query request initiated by the verification terminal. The core includes the digital feature vector to be verified, and can also include the basic information of the component to be verified for auxiliary screening. The "preset threshold" is the critical value to distinguish between genuine and counterfeit products. If the threshold is exceeded, the match is considered successful and the digital identity certificate in the corresponding evidence record can be extracted. If the threshold is not exceeded, the retrieval is judged to have failed.

[0092] In some implementations, the preset threshold is determined using a dynamic adjustment mechanism, which is obtained as follows: The system periodically collects similarity scores from successfully matched historical verification records. Machine learning models are used to analyze the distribution patterns of these scores, calculating a dynamic threshold that maximizes the distinction between genuine and counterfeit products. For example, if historical successful match similarities are concentrated between 90% and 98%, the model sets the threshold to 92% to reduce the false positive rate. This dynamic threshold is then automatically updated to the verification system. Furthermore, to improve retrieval efficiency, a subset of stored evidence vectors for the corresponding batch can be selected from the blockchain based on the basic information of the component to be verified. Similarity is then calculated within this subset, avoiding delays caused by traversing the entire set of vectors.

[0093] It should be noted that the process for generating the digital feature vector to be verified must be completely consistent with that of the feature generation module. Furthermore, when a retrieval fails, a clear message should be provided stating "No matching evidence found" or "Similarity below the threshold" to facilitate troubleshooting by the requesting party.

[0094] For example, a downstream electronic device manufacturer sends a verification request to the application verification module via an HTTPS API. After receiving the request, the application verification module first filters out the subset of stored evidence vectors for that batch in the blockchain by "model + batch". Then, it uses the cosine similarity algorithm to calculate the similarity between the vector to be verified and each vector in the subset, and obtains the highest similarity of 95.2%. Since the current dynamic threshold is 92%, 95.2% > 92%. Therefore, the stored evidence record corresponding to the highest similarity is selected, and the digital identity certificate is extracted from it.

[0095] S402. Query the evidence storage record from the blockchain distributed ledger based on the digital identity certificate. If the query fails, return the verification failure information. If the query succeeds, obtain the complete production process data through the off-chain storage address index in the evidence storage record.

[0096] Among them, complete production process data includes data from all stages of electronic component processing, from raw material processing to delivery, and is the core basis for generating traceability information.

[0097] In some implementations, blockchain queries can optimize efficiency through multi-level indexing: the primary index divides the evidence storage records into partitions based on component batches, and the secondary index establishes a hash index based on the unique ID in the digital identity certificate, reducing the full-chain traversal time; for frequently queried popular batches of data, a local cache can be established to reduce query latency from seconds to milliseconds. After obtaining off-chain data, integrity verification is required: the hash value of the obtained complete production process data is calculated and compared with the "production process data hash value" in the on-chain evidence storage record. If they match, it means that the off-chain data has not been tampered with; if they do not match, the data is deemed invalid, triggering a retry mechanism. If the retry fails after 3 attempts, a verification failure message of "off-chain data abnormal" is returned.

[0098] S403. Integrate on-chain evidence-based transaction information with off-chain production process data to generate traceability information in chronological order; if the data does not match, return a verification failure result and finally feed the result back to the requester.

[0099] Among them, the association integration is to associate on-chain evidence storage transaction information with off-chain production process data according to the "production link-time" dimension, forming a complete link of "physical characteristics → production link → on-chain evidence storage".

[0100] In some implementations, data integration can take the form of "structured + visual": the structured output includes fields such as "stage name, time, key data, and on-chain evidence identification", which facilitates subsequent data statistics and analysis; the visual output displays the production and evidence storage chain in a graphical way, marking the key nodes of each stage, reducing the understanding cost for the requester.

[0101] For example, the application verification module integrates the following data: on-chain notarized transaction information (notarization time 2024-05-20 14:35:22, transaction ID "TX-20240520-12345") and off-chain production process data (raw material input 2024-05-20 14:00:00, welding 2024-05-20 14:10:00, packaging 2024-05-20 14:20:00, quality inspection 2024-05-20 14:30:00), generates a visual timeline in chronological order, marks the process parameters and operators of each stage, and outputs the structured data in an Excel spreadsheet. Since all data matches, a "verification passed" result is generated, along with traceability information, which is then fed back to the downstream electronic equipment manufacturer. If the off-chain welding time (2024-05-20 15:00:00) is found to be later than the on-chain evidence storage time (2024-05-20 14:35:22), a "verification failed" result is returned, and the contradictory time points are marked.

[0102] Based on the above technical solution, the application verification module achieves reliable traceability and authenticity verification of electronic component production data through request reception and feature matching, on-chain querying and off-chain data retrieval, as well as data integration and result feedback. The entire process, centered on "on-chain hash verification + off-chain data verification," effectively solves problems such as difficulty in identifying counterfeit products, low data reliability, and incomplete traceability information in electronic component traceability, providing reliable technical support for transparent supply chain management and precise recall of products with quality issues.

[0103] Furthermore, the electronic component manufacturing production data traceability system provided in this application embodiment also includes a verification terminal module, used to acquire the original image of the electronic component to be verified, generate the digital feature vector to be verified and initiate a verification request to the application verification module, and simultaneously receive the verification result returned by the application verification module and display it intuitively to the user, realizing terminal-side data acquisition and result interaction for electronic component traceability verification. Specifically, this can be achieved through the following steps: S501, S502, S503, and S504. S501. The built-in miniature laser emitter and image sensor scan the predetermined feature area of ​​the electronic component to be verified to acquire the image to be verified. The predetermined feature area refers to the microstructure area on the surface of the electronic component to be verified that is completely consistent with the original feature data collected during the production stage.

[0104] S502. Perform the same signal transformation and feature extraction process as the feature generation module on the image to be verified to generate the digital feature vector to be verified.

[0105] S503. Send the digital feature vector to be verified to the application verification module via the wireless network and initiate a verification request.

[0106] S504. Receive and parse the verification results returned by the application verification module, and intuitively display the verification conclusion to the user through the display screen or indicator lights.

[0107] In some implementations, after receiving the verification result, the terminal first verifies the signature of the result using the public key of the application verification module, and then parses the JSON format result data. If the verification is successful, the display shows "Verification Successful" and below it displays a summary of traceability information such as component model, batch, production time, and quality inspection results, while the green indicator light remains on. If the verification fails, the display shows "Verification Failed" and indicates the reason for the failure, while the red indicator light flashes. If the result parsing is abnormal, the display shows "Result parsing failed, please try again," and the yellow indicator light flashes.

[0108] Some of the data in the above formula are calculated by removing dimensions and taking their numerical values. The formula is the closest to the real situation obtained by software simulation of a large amount of collected data. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.

[0109] Working principle of the invention: First, by utilizing the non-replicable physical properties of the microstructures naturally formed on the surface of electronic components, a unique physical field signal is generated by the interaction of a high-energy beam with the microstructure. This signal serves as the original feature data, eliminating the defect of easily copied external identifiers from the physical source. Then, the original feature data is transformed from the spatial domain to the frequency domain through signal transformation, and the mid-to-high frequency features that can characterize the uniqueness of the microstructure are extracted. A fixed-length digital feature vector is generated by combining true random numbers and cryptographic hash operations, and the physical features of the component are transformed into a unique identity in the digital space. Next, relying on the distributed ledger and immutability of blockchain, digital feature vectors, hash values ​​of production process data, and key metadata are written into the blockchain. Through consensus algorithms, verification and storage are achieved across all network nodes, ensuring the trusted binding of production data and digital identity. Finally, the verification phase obtains the digital identifier of the component to be verified by reproducing the digital feature vector generation process, compares it with the on-chain evidence, and combines it with the complete production data stored off-chain for verification, so as to realize the authenticity of the component and traceability throughout its entire life cycle, thus providing dual protection for data credibility and traceability effectiveness from both physical and digital levels.

[0110] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A blockchain-based electronic component manufacturing production data traceability system, characterized by, include: The data acquisition module is used to collect raw characteristic data of electronic components; The original feature data represents the physical field signal generated by the interaction between the high-energy beam and the microstructure on the surface of the component. The feature generation module is used to transform the original feature data from the spatial domain to the transform domain through signal transformation, generating digital feature vectors of electronic components; specifically, it includes: S201. Perform a two-dimensional Fourier transform on the original feature data to convert the original feature data from a spatial domain image to a frequency domain spectrum; S202. Perform bandpass filtering on the frequency domain spectrum; S203. Calculate the energy intensity and distribution entropy of the filtered spectrum in multiple concentric rings and sector regions, and combine the calculated energy values, entropy values ​​and moment characteristics of the spectrum in a predetermined order to form a primary feature vector; S204. Obtain a true random number from a true random number generator, mix the primary feature vector with the true random number, perform a cryptographic hash operation on the mixed result, and use the resulting hash value as the digital feature vector of the electronic component. The blockchain evidence storage module is used to calculate the hash value of the digital feature vector and its associated production process data, and write the hash value into the blockchain distributed ledger for evidence storage, thereby obtaining the digital identity certificate of the electronic component. The application verification module is used to receive production data traceability verification requests for electronic components to be verified, compare the associated data on-chain and off-chain, and return the traceability information or verification results of the electronic components to be verified.

2. The blockchain-based electronic component manufacturing production data traceability system of claim 1, wherein, The process of obtaining the original feature data includes: A laser beam of a specific wavelength is used and focused by a lens group to irradiate the microscopic feature area on the surface of an electronic component; An image sensor is used to receive and record the laser scattering field pattern generated by the microstructure of the component surface to obtain the raw feature data.

3. The blockchain-based electronic component manufacturing data traceability system according to claim 1, characterized in that, The workflow of the blockchain evidence storage module includes: The system receives the digital feature vector from the feature generation module and the hash value of the associated production process data; the production process data represents various quality data generated by the electronic components during the production process, including one or more of process parameter data, equipment status data, and operation record data. A comprehensive hash value is generated by combining the hash value of the production process data with the digital feature vector. The combined hash value, the digital feature vector, the hash value of the production process data, the timestamp, and the off-chain storage address index are used to construct a transaction. The transaction is packaged into a new block using a consensus algorithm, and after being verified by all nodes in the blockchain network, the new block is appended to the end of the blockchain, completing distributed storage.

4. The blockchain-based electronic component manufacturing data traceability system according to claim 3, characterized in that, The process of packaging transactions into a new block means that consensus nodes in the blockchain network verify the validity of the stored transactions. After verification, a batch of transactions are packaged together with the hash value of the previous block in chronological order to form a new data block. The new data block is then broadcast to all nodes in the entire blockchain network for storage through the consensus mechanism, resulting in a chain-like storage structure.

5. The blockchain-based electronic component manufacturing data traceability system according to claim 1, characterized in that, The process of obtaining the traceability information includes: Receive a user's verification request, the request containing the digital feature vector of the electronic component to be verified; Based on the identification information or the digital identity credential retrieved through feature matching, the evidence storage record is queried from the blockchain distributed ledger; if the evidence storage record cannot be found, a verification failure message is returned. Based on the off-chain storage address index in the evidence storage record, retrieve the associated complete production process data from the off-chain storage system; The on-chain notarized transaction information is linked and integrated with the complete off-chain production process data to generate traceability information in chronological order.

6. The blockchain-based electronic component manufacturing data traceability system according to claim 5, characterized in that, The digital identity credential retrieved through feature matching includes: Calculate the similarity between the digital feature vector to be verified and each digital feature vector already stored in the blockchain; The system filters out the most similar evidence records with a similarity exceeding a preset threshold and extracts the digital identity credential corresponding to the evidence record; otherwise, it determines that the retrieval has failed and generates a verification failure message.

7. The blockchain-based electronic component manufacturing data traceability system according to claim 6, characterized in that, The preset threshold is an adaptive threshold that is dynamically adjusted based on historical verification data. The dynamic adjustment method includes: Periodically collect similarity scores from successfully matched historical verification records; Based on the distribution of the similarity values, a dynamic threshold that can maximize the distinction between genuine and counterfeit electronic components is calculated using a machine learning model. Update the dynamic threshold to the preset threshold used in the next verification.

8. The blockchain-based electronic component manufacturing data traceability system according to claim 1, characterized in that, The production data traceability system also includes a verification terminal module, the workflow of which includes: The built-in miniature laser emitter and image sensor scan a predetermined feature area of ​​the electronic component to be verified to obtain an image of the component to be verified. The same signal transformation and feature extraction process as the feature generation module is performed on the image to be verified to generate a digital feature vector to be verified. The digital feature vector to be verified is sent to the application verification module via a wireless network to initiate a verification request; Receive and parse the verification results returned by the application verification module, and intuitively display the verification conclusion to the user through a display screen or indicator light.

Images