A blockchain-based serialized traceability data verification method

Abstract

The present application relates to the technical field of data security, in particular to a serialization traceability data verification method based on block chain, comprising: collecting physical features of labeling and packaging equipment to generate a dynamic capture pulse entropy source, and constructing a label sequence identity coupling fingerprint containing instruction stream bias and physical state texture; combining traceability code and action time offset to perform XOR packaging to generate label sequence instantaneous credentials, and introducing historical feedback values to generate label sequence topology entropy chain through hash iteration; using distributed ledger consensus nodes to analyze the entropy chain, verifying the extracted logical feature vector and hardware base state characteristics, comparing the global state root closed loop to generate a logical anchor credential, and triggering asynchronous evidence storage accordingly. Through the deep coupling of physical entropy source and recursive topology chain, the present application realizes cross-dimensional verification of equipment identity, action timing, and physical terminal and block chain nodes, ensuring the authenticity and non-tamperability of traceability data in the generation and circulation process.

Description

Technical Field

[0001] This invention relates to the field of data security technology, specifically to a blockchain-based method for verifying serialized traceable data. Background Technology

[0002] In automated labeling and packaging scenarios, the serialized data generated in real time by terminal equipment integrated into the production line (such as frequently generated unique encrypted codes and serialized label IDs) is the core carrier for ensuring product data security and integrity verification. This data not only represents the product's digital identity but also serves as the underlying basis for subsequent logical verification by distributed systems.

[0003] Currently, for processing this type of serialized traceability data, existing methods mostly involve establishing a data association model in the cloud and using a hash function to hash user identities or operation records before storing them on the blockchain to generate blockchain credentials for subsequent verification. However, in high-frequency, continuous industrial production scenarios such as automatic labeling, the above technical approach still faces significant bottlenecks at the underlying data security level. First, at the data acquisition and transmission layer, the serialized identifier stream generated by the labeling machine has extremely strong temporal order. Existing hash verification mechanisms mostly focus on the integrity of static results, lacking real-time protection against identity forgery and replay attacks during high-frequency data stream transmission. This leads to a security disconnect between the original serial number generated by the physical device and the identity credentials recorded at the logical layer. Second, regarding the verification of logical consistency, existing distributed evidence storage methods often only verify single-dimensional data blocks, making it difficult to perform sequence consistency verification on serialized data with strict logical continuity. Once packet loss or illegal injection occurs in the intermediate transmission stage, the system struggles to achieve self-healing verification of the data sequence logical loop without relying on centralized authorization. Finally, due to the lack of deep protocol coupling between physical terminals and blockchain nodes, the root credentials of the traceability evidence chain are vulnerable to the threat of abuse of underlying permissions in the storage environment.

[0004] In response to the technical problems of insufficient identity authentication strength of serialized data streams, difficulty in verifying sequence consistency, and disconnection of coupling logic between physical terminals and blockchain nodes in the existing technologies, this invention proposes a blockchain-based serialized traceability data verification method. Summary of the Invention

[0005] The purpose of this invention is to provide a blockchain-based serialized traceability data verification method, which realizes deep coupling security verification between the automated production terminal of labeling and packaging equipment and blockchain nodes in terms of identity access, sequence logic consistency and physical execution state.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A blockchain-based method for verifying serialized traceability data includes: Acquire unwinding tension harmonic data, peeling blade data, and label suction negative pressure data of the labeling and packaging equipment to generate a label motor capture pulse entropy source; perform correlation transformation on the label motor capture pulse entropy source to extract logical feature vectors containing the underlying state bias of the instruction stream and the physical state texture of the execution end, and generate a label sequence identity coupled fingerprint; Obtain the serialized traceability identifier and the time coordinate offset of the labeling action at the automated execution end, and perform XOR encapsulation based on the label identity coupled with the fingerprint to generate the label instantaneous credential; obtain the labeling sequence historical feedback value, and perform chain aggregation and hash iteration processing on the label instantaneous credential and the label historical feedback value to generate the label topology entropy chain; Consensus nodes based on the distributed ledger perform reverse parsing on the sequence topology entropy chain to extract the logical feature vector from the sequence identity coupling fingerprint; perform identity verification by comparing the logical feature vector with the hardware base state features in the distributed ledger; compare the logical topology closed loop between the sequence topology entropy chain and the global state root of the distributed ledger to generate a logical anchor certificate; when the logical anchor certificate determines that the identity is consistent and the logical topology is closed, trigger asynchronous evidence storage update to complete data security verification.

[0007] Preferably, the process of generating the label-capturing pulse entropy source includes: acquiring the frequency domain characteristic distribution of the motor current at the AC servo drive end of the labeling and packaging equipment, and mapping it to the generated unwinding tension harmonic data; acquiring the pulse edge response time offset of the label peeling trigger signal and mapping it to the generated peeling blade data; acquiring the negative pressure pulsation transient data when the vacuum generator performs the label suction action and mapping it to the generated suction negative pressure data; and cascading and assembling the unwinding tension harmonic data, peeling blade data, and suction negative pressure data to generate the label-capturing pulse entropy source.

[0008] Preferably, the process of generating the serialized identity coupled fingerprint includes: inputting the serialized motion capture pulse entropy source into a nonlinear mapping algorithm to perform an association transformation and extract physical disturbance features; using the physical disturbance features to perform disturbance mapping on the discrete sampled data of the execution controller instruction execution cycle to generate the instruction stream bottom-level state bias; using the physical random disturbance features to perform orthogonal decomposition mapping on the waveform feature components of the action feedback signal of the physical execution mechanism at the execution end to generate the execution end physical state texture; concatenating the instruction stream bottom-level state bias with the execution end physical state texture execution features to construct a logical feature vector; and compressing and mapping the logical feature vector through a hash algorithm to generate the serialized identity coupled fingerprint.

[0009] Preferably, the process of generating the labeling instantaneous credential includes: obtaining a serialized traceability identifier code and using it as a globally unique sequence; obtaining the time coordinate offset of the labeling action of the automated execution terminal relative to the sampling zero point and converting it into a labeling machine phase vector; inputting the labeling identity coupled fingerprint, the globally unique sequence, and the labeling machine phase vector into an XOR encapsulation operator to perform bit-by-bit logical coupling to obtain a peeling-and-apply linkage instantaneous stream carrying hardware physical characteristics, data identity characteristics, and action spatiotemporal characteristics; inputting the peeling-and-apply linkage instantaneous stream into a discrete transformation operator to perform nonlinear obfuscation processing to generate the labeling instantaneous credential.

[0010] Preferably, the process of generating the labeling topology entropy chain includes: acquiring and identifying the production time sequence position of the current labeling action; if the production time sequence position is the first position, then defining the preset labeling seed initial vector as the labeling history feedback value; if the production time sequence position is an increasing position, then acquiring the labeling topology entropy chain generated by the previous production time sequence position as the labeling history feedback value; inputting the labeling instantaneous credentials and the labeling history feedback value into a cascade mapping algorithm to perform chain aggregation and construct sequence association features; inputting the sequence association features into a cyclic hash iteration operator to perform nonlinear logic compression and generate the labeling topology entropy chain of the current production time sequence position.

[0011] Preferably, the process of extracting the logical feature vector includes: retrieving the labeling topology entropy chain through the consensus node of the distributed ledger and performing labeling machine running state logic decomposition processing; separating the labeling identity coupling fingerprint of the current labeling position from the labeling topology entropy chain, inputting the labeling identity coupling fingerprint into a preset verification protocol to perform feature dimension restoration processing, and generating a peeled steady-state logical image; extracting discrete feature components containing instruction stream underlying state bias data and execution end physical state texture data from the peeled steady-state logical image, performing spatial reconstruction processing on the discrete feature components, and restoring to generate a logical feature vector.

[0012] Preferably, the process of executing the running state logic decomposition of the label includes: inputting the label topology entropy chain into a preset peeling and deconstruction model; using the pre-stored label history feedback value in the peeling and deconstruction model as the deconstruction benchmark, performing XOR residual extraction on the label topology entropy chain to offset the weight noise of the label history feedback value on the current position data, and separating the feature domain to be verified of the label instantaneous credential; inputting the feature domain to be verified into the inverse mapping logic of the cascaded mapping algorithm to perform bit logic decomposition, and restoring and generating the label identity coupled fingerprint.

[0013] Preferably, the process of generating logical anchor credentials and completing data security verification includes: performing identity verification on the logical feature vector and the hardware base state features of the distributed ledger, and outputting the identity matching weight; calling the pre-set peeling verification operator in the distributed ledger, comparing the logical topological loop of the index topological entropy chain with the global state root of the distributed ledger through the peeling verification operator, and outputting the sequence verification coefficient; inputting the identity matching weight and the sequence verification coefficient into the logical discriminant function for aggregation processing to generate logical anchor credentials; using the logical anchor credentials as a trigger signal for asynchronous blockchain notarization, when the logical anchor credentials determine that the identity is consistent and the logical topological loop is closed, triggering asynchronous notarization update to complete data security verification.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention constructs a label tracking pulse entropy source using unwinding tension harmonic data, peeling blade data, and label suction negative pressure data. It then maps and generates a coupled fingerprint containing the underlying bias of the instruction stream and the physical texture of the execution end, achieving underlying anchoring between the physical microscopic disturbances of the labeling and packaging equipment and the digital identity core. This solution utilizes the unclonable physical characteristics generated by automated equipment during high-speed operation as an identity source, preventing the security risks of illegal cloning or simulation of equipment identity from the physical source of data generation in automated labeling production scenarios.

[0015] 2. This invention generates instantaneous credentials by coupling fingerprints with tagged identity, serialized traceability identifiers, and time coordinate offsets through encapsulation. It also incorporates historical feedback values ​​from the previous tagged position to perform chained aggregation and hash iteration, achieving deep spatiotemporal coupling between production cycle time, temporal position, and logical credentials. This recursive locking mechanism based on physical execution order solves the problems of temporal sequence disorder, missing and duplicate tampering, and data tampering that easily occur in high-frequency serialized data streams in complex pipeline environments, ensuring the integrity and immutability of traceability data in the temporal dimension.

[0016] 3. This invention utilizes distributed ledger consensus nodes to perform reverse parsing on the sequence topology entropy chain, and coordinates a cross-verification method involving hardware base state feature verification and global state root logical topology closed-loop comparison to achieve cross-dimensional anchoring verification of the physical execution state, logical algorithm state, and ledger consensus state. This solution changes the current situation of data disconnect between physical terminals and blockchain nodes, and achieves deep security verification of serialized traceability identifiers by restoring the underlying physical feature vectors and performing logical closed-loop verification at the consensus layer. Attached Figure Description

[0017] Figure 1 This is a flowchart of a blockchain-based serialized traceability data verification method according to the present invention; Figure 2 This is a schematic diagram of the logical structure of the label topological entropy chain generation process according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating the distributed ledger verification and logical anchoring process according to an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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] Please see Figures 1 to 3 This invention provides a blockchain-based method for verifying serialized traceability data, the technical solution of which is as follows: Example 1

[0020] Reference Figure 1 This embodiment provides a specific application scenario for a blockchain-based serialized traceability data verification method. Specifically, on a production line integrated with high-speed automated labeling equipment, it performs source identity verification and tamper-proof evidence storage for the product's serialized identifier by collecting physical features and constructing a logical chain during the labeling process. In this application scenario, while the labeling and packaging equipment completes the physical unwinding, peeling, and affixing actions, it also performs deep coupling of underlying hardware features and digital identity at the logical layer. Through real-time interaction with distributed ledger nodes, it achieves secure verification and reliable evidence storage for massive amounts of high-frequency production data. The specific steps are as follows: Acquire unwinding tension harmonic data, peeling blade data, and label suction negative pressure data of the labeling and packaging equipment to generate a label motor capture pulse entropy source; perform correlation transformation on the label motor capture pulse entropy source to extract logical feature vectors containing the underlying state bias of the instruction stream and the physical state texture of the execution end, and generate a label sequence identity coupled fingerprint; Obtain the serialized traceability identifier and the time coordinate offset of the labeling action at the automated execution end, and perform XOR encapsulation based on the label identity coupled with the fingerprint to generate the label instantaneous credential; obtain the labeling sequence historical feedback value, and perform chain aggregation and hash iteration processing on the label instantaneous credential and the label historical feedback value to generate the label topology entropy chain; Consensus nodes based on the distributed ledger perform reverse parsing on the sequence topology entropy chain to extract the logical feature vector from the sequence identity coupling fingerprint; perform identity verification by comparing the logical feature vector with the hardware base state features in the distributed ledger; compare the logical topology closed loop between the sequence topology entropy chain and the global state root of the distributed ledger to generate a logical anchor certificate; when the logical anchor certificate determines that the identity is consistent and the logical topology is closed, trigger asynchronous evidence storage update to complete data security verification.

[0021] Furthermore, the process of generating the label-capturing pulse entropy source includes: acquiring the frequency domain characteristic distribution of the motor current at the AC servo drive end of the labeling and packaging equipment, and mapping it to the generated unwinding tension harmonic data; acquiring the pulse edge response time offset of the label peeling trigger signal and mapping it to the generated peeling blade data; acquiring the negative pressure pulsation transient data when the vacuum generator performs the label suction action and mapping it to the generated suction negative pressure data; and cascading and assembling the unwinding tension harmonic data, peeling blade data, and suction negative pressure data to generate the label-capturing pulse entropy source.

[0022] Specifically, the system acquires the operating data of the AC servo drive responsible for roll diameter control in the labeling and packaging equipment. The rising edge of the label peeling trigger signal is used as the zero point of the global sampling clock, synchronously triggering servo current sampling and transient acquisition by the pressure sensor to ensure that multi-source physical characteristics are within the same labeling action window on the time axis. High-frequency sampling extracts the 3rd and 5th harmonic components of the motor current in the 100Hz to 500Hz frequency band. Normalization is performed, and the maximum and minimum amplitudes of each harmonic component in the extracted current frequency domain feature distribution are calculated. Using a maximum-minimum algorithm, the difference between the minimum amplitude and the maximum amplitude is subtracted from the value of each sampling point in the sequence, and then divided by the discrete range between the maximum and minimum amplitudes. A linear scaling process maps the original amplitude values ​​proportionally to a standardized numerical range of 0 to 1, eliminating non-characteristic deviations caused by external voltage fluctuations or differences in equipment load. The scaled amplitude fluctuation curve is converted into a 128-bit fixed-point digital vector through linear quantization mapping, generating unwinding tension harmonic data.

[0023] The tag peeling trigger signal is captured synchronously by a high-speed photoelectric sensor. The response time between the time the signal command is issued and the time of physical sensing at the edge of the peeling plate is measured. The pulse edge response time offset in μs at the time of physical sensing at the edge is extracted. The data is in the range of 0 to 1023 μs. Linear discretization and quantization mapping is performed with a preset step size of 1 μs to obtain 10 initial feature values. These are then expanded into 64-bit depth feature value peeling blade data through hash mapping or logical padding.

[0024] Within 50ms of the suction action window, the pressure pulsation waveform of the pressure sensor at the vacuum generator end is acquired, and the pressure fluctuation component with a negative pressure peak between -0.05MPa and -0.08MPa is extracted. Due to the trend of industrial environmental interference on the pressure waveform, this embodiment pre-processes the pressure fluctuation component with a fuzzy extraction operator (low-pass filtering to remove high-frequency environmental noise), extracts a multi-dimensional feature vector (including the negative pressure change rate and pressure peak response time) that reflects the essential characteristics of the physical damping of the gas path, and then uses an adjoint error correction logic to quantize the multi-dimensional feature vector, maps the multi-dimensional feature vector to the nearest neighbor codeword point in the lattice space, calculates the distance offset between the feature vector and the standard codeword point, and generates a stable 64-bit transient bit code, i.e., suction negative pressure data, within the allowable physical offset threshold.

[0025] The aforementioned 128-bit unwinding tension harmonic data, 64-bit peeling blade data, and 64-bit suction negative pressure data are cascaded end-to-end at the bit level. Based on the physical execution logic chain of the labeling action, the unwinding tension harmonic data, representing the power source, is defined as the high-bit starting field; subsequently, the peeling blade data, representing the key timing of the action trigger, is used as the middle-bit transition field; finally, the suction negative pressure data, representing the end-of-line execution efficiency, is coupled as the low-bit ending field. The data fields are seamlessly connected through the logical shifts and splicing instructions of the bit vector splicing operator, without redundant check bits or padding bits, constructing a label motion capture pulse entropy source with a total length of 256 bits and complete physical characteristic dimensions.

[0026] This invention constructs a non-clonal entropy source that is deeply bound to the hardware operating state by extracting the discrete distribution of microscopic physical features during the labeling process. It establishes a logical association between traceability identifiers and physical device action features from the underlying data source, providing a physical foundation with device uniqueness for subsequent data verification.

[0027] Furthermore, the process of generating the serialized identity coupled fingerprint includes: inputting the serialized motion capture pulse entropy source into a nonlinear mapping algorithm to perform an association transformation and extract physical disturbance features; using the physical disturbance features to perform disturbance mapping on the discrete sampled data of the execution controller instruction execution cycle to generate the instruction stream bottom-level state bias; using the physical random disturbance features to perform orthogonal decomposition mapping on the waveform feature components of the action feedback signal of the physical execution mechanism at the execution end to generate the execution end physical state texture; concatenating the instruction stream bottom-level state bias with the execution end physical state texture execution features to construct a logical feature vector; and compressing and mapping the logical feature vector through a hash algorithm to generate the serialized identity coupled fingerprint.

[0028] Specifically, the correlation transformation includes the following steps: inputting a 256-bit standard motion capture pulse entropy source into a nonlinear mapping algorithm based on chaotic mapping logic, and generating a highly sensitive 128-bit initial value sequence in the continuous phase space by performing Logistic logic iterative operation with control parameter r of 3.99, which is defined as physical perturbation feature.

[0029] The micro-timing offset (i.e., clock jitter data) within the current instruction execution cycle of the execution controller, ranging from 10ns to 50ns, is acquired as discrete sampling data. A cyclic left shift XOR operation is performed between the 128-bit physical perturbation feature and the timing offset data. The 128-bit physical perturbation feature is defined as operator A, and the micro-timing offset data is zero-padded to 128 bits and defined as the block to be operated on, B. Based on the current phase period of the labeler, the number of cyclic left shift bits k (e.g., k=13) is determined, and a k-bit cyclic left shift operation is performed on operator A to redistribute the bit weights. Subsequently, the shifted operator A′ is XORed bit-by-bit with the block to be operated on, B. This produces a 128-bit binary block, generating the instruction stream underlying state bias containing the hardware micro-timing features. This avalanche effect mapping ensures that extremely small deviations in micro-timing (at the ns level) can be uniformly diffused throughout the entire field of the 128-bit instruction stream underlying state bias.

[0030] When the physical actuator performs label stripping or labeling actions, the amplitude components of the feedback signal waveform in the high-frequency band of 1kHz to 10kHz are extracted in real time and constructed as the original feature vector. Using orthogonal decomposition mapping logic (using the Schmidt orthogonalization process), the physical disturbance features are projected onto the vector space to which the original feature vector belongs, and feature projection stripping and signal normalization operations are performed. Specifically, the physical disturbance features are mapped to spatial basis vectors, and the original feature vector is defined as the vector to be processed. The projection components in the direction are calculated using the orthogonal projection operator, and the projection components are subtracted from the vector to be processed to achieve feature projection stripping, eliminate environmental common-mode interference, and extract the residual vector that only represents the physical properties unique to the actuator. Subsequently, the sum of squares of each dimension component in the residual vector is calculated, and the square root operation is performed on the sum of squares to obtain the magnitude of the residual vector. The reciprocal of the magnitude is used as a scaling factor and a weighted multiplication operation is performed with each dimension component of the residual vector to make the magnitude of the processed vector equal to 1. The amplitude values ​​of each dimension are mapped to the interval [0, 1] to generate a fixed-length 128-bit texture feature code that characterizes the micro-deformation and friction characteristics of the mechanical actuator, which serves as the physical state texture of the actuator.

[0031] The 128-bit instruction stream low-level state bias and the 128-bit execution-side physical state texture are concatenated in sequential order to construct a 256-bit logical feature vector. This logical feature vector is then input into a preset cryptographic hash algorithm (such as the SM3 algorithm) to perform data padding and a 512-bit block iterative compression operation. First, the 256-bit logical feature vector is padded by adding one "1" bit followed by n "0" bits, and a 64-bit length field is appended, bringing the total length to 512 bits, thus constructing a standard block message. Next, the standard block message undergoes a double sixteen-round nonlinear permutation logic process. In each iteration, the 512-bit data is first linearly expanded, generating multiple sets of word cyclic variables. Then, using the expanded word variables as control factors, the data undergoes sixty-four rounds of high-frequency obfuscation operations within a compression unit composed of nonlinear permutation logic and bit cyclic shift operators. In each round, the register state from the previous round is modulo-added with the expanded component of the current block, enabling deep nonlinear diffusion of physical features and instruction bias features at the bit level. Finally, the block results from the sixty-four iterations are concatenated and output to generate a fixed-length 256-bit digest, defined as a sequenced identity coupled fingerprint.

[0032] This invention cross-couples the logical timing deviation of the underlying instructions with the physical action signals of the actuator through nonlinear transformation, constructing a digital credential with dual locking effects of instruction state and physical state, ensuring the correspondence between the traceability data identity and the hardware device's operating logic.

[0033] Furthermore, the process of generating the labeling instantaneous credential includes: obtaining a serialized traceability identifier code and using it as a globally unique sequence; obtaining the time coordinate offset of the labeling action of the automated execution terminal relative to the sampling zero point and converting it into a labeling machine phase vector; inputting the labeling identity coupled fingerprint, the globally unique sequence, and the labeling machine phase vector into an XOR encapsulation operator to perform bit-by-bit logical coupling to obtain a peeling-and-apply linkage instantaneous stream carrying hardware physical characteristics, data identity characteristics, and action spatiotemporal characteristics; inputting the peeling-and-apply linkage instantaneous stream into a discrete transformation operator to perform nonlinear obfuscation processing to generate the labeling instantaneous credential.

[0034] Specifically, a 24-bit character serialized traceability identifier is extracted from the product surface using a visual scanning device on an automated production line. This identifier is then mapped to a 128-bit binary data block using a Base64 encoding algorithm and defined as a globally unique sequence. During the tag-peeling execution mechanism's operation cycle, a high-speed timer captures the time coordinate offset between the tag-peeling start signal and the tag-peeling completion signal relative to the sampling zero point. This value is mapped to a 64-bit binary sequence representing the phase information of the execution mechanism's current physical position, generating the tag-peeling phase vector.

[0035] A 256-bit identifier fingerprint, a 128-bit globally unique sequence, and a 64-bit machine phase vector are input into a preset XOR encapsulation operator. This XOR encapsulation operator is a logic processor configured with multi-input parallel XOR logic circuits. Its execution logic is as follows: first, the 128-bit globally unique sequence and the 64-bit machine phase vector are expanded to a 256-bit width through a cyclic copy operation. Then, phase alignment is performed on the data at each input terminal according to a preset cyclic left shift rule, and bit-by-bit parallel XOR logic calculations are performed. Bit-by-bit coupling generates a peeling-and-pasting instantaneous stream carrying hardware physical characteristics, data identity characteristics, and action spatiotemporal characteristics. This peeling-and-pasting instantaneous stream is a 448-bit composite bit vector. Its structure consists of a 256-bit physical feature area, a 128-bit identifier area, and a 64-bit machine phase identifier area rearranged. The bits of each functional area are interleaved according to a non-contiguous address space, forming a bit-level entangled logical whole.

[0036] The instantaneous stream of the peeling and pasting linkage is input into a discrete transformation operator consisting of multiple sets of 8x8 permutation box logic and a bit-level scrambling table. This operator performs nonlinear obfuscation processing based on a substitution permutation network. The process of generating a labeled instantaneous credential through nonlinear obfuscation processing includes: dividing the peeling and pasting linkage instantaneous stream into several data sub-blocks according to a preset step size, and inputting each data sub-block in parallel into a substitution network consisting of multiple sets of 8x8 nonlinear permutation boxes to perform byte substitution; performing cross-byte position rearrangement on the substituted data, and performing modulo-2 addition with the subkey of the current round to construct the iteration cycle of the substitution permutation network; after the preset iteration cycle is completed, calling the truncation operator to compress the dimension of the obfuscated bit vector, and extracting the bit component of the specified index bit to produce a fixed-length labeled instantaneous credential.

[0037] Specifically, this embodiment employs a discrete transform operator based on a substitution permutation network to perform obfuscation. First, the 448-bit peel-and-paste instantaneous stream is divided into seven 64-bit sub-blocks. Each sub-block is processed through eight sets of parallel 8x8 permutation boxes, using a finite field GF(2^3)... 8 The inverse operation and affine transformation on GF(2) are implemented to map the input 8-bit bytes to the finite field GF(2). 8 Find the elements in GF(2) and calculate the multiplicative inverse of each element in the field GF(2). 8 The irreducible polynomial of is taken as m(x) = x 8 +x 4 +x 3 +x 2+1, the inverse of the zero element is defined as itself; then, a modulo-2 addition multiplication operation is performed on the inverse vector using an 8×8 constant coefficient matrix, and a fixed 8-bit offset constant is superimposed. By using the nonlinearity provided by the multiplicative inverse operation and the algebraic confusion provided by the affine transformation, the input bytes are replaced with statistically irrelevant output bytes, thereby completely destroying the linear statistical properties of the original data.

[0038] After substitution, the data enters a bit-level scrambling table and undergoes global position rearrangement according to a preset diffusion matrix, ensuring that changes in each bit's input can rapidly spread across the entire bit width. The intermediate state generated in each round is XORed with the round key derived from the standard machine's phase vector. This process is repeated for 16 rounds to achieve deep obfuscation. Finally, a discrete transformation operator uses a truncation operator to extract a 256-bit fixed-length digest from the obfuscated 448-bit result using a hash index algorithm. The truncation operator uses the standard machine's phase vector as a seed value to generate a hash index table, dynamically extracting different 256-bit components at each sequence position, improving the credential's anti-predictability and generating instantaneous credentials. This technical solution, by constructing a nonlinear obfuscation mechanism based on finite field algebraic operations and a multi-round substitution network, deeply anchors the microscopic physical fluctuations of the dynamic execution mechanism to the product identity sequence, eliminating the linear statistical regularity of traceability data.

[0039] This invention constructs a digital credential deeply locked to a specific production moment and executed action by XOR coupling and discrete obfuscating the device hardware fingerprint, product identification code, and microsecond-level timing phase of mechanical actions, thereby enhancing the anti-counterfeiting capability of data during the flow process.

[0040] Further, the process of generating the labeling topology entropy chain includes: acquiring and identifying the production time sequence position of the current labeling action; if the production time sequence position is the first position, then defining the preset labeling seed initial vector as the labeling history feedback value; if the production time sequence position is an increasing position, then acquiring the labeling topology entropy chain generated by the previous production time sequence position as the labeling history feedback value; inputting the labeling instantaneous credential and the labeling history feedback value into a cascade mapping algorithm to perform chain aggregation, constructing a sequence association feature; inputting the sequence association feature into a cyclic hash iteration operator to perform nonlinear logic compression, generating the labeling topology entropy chain of the current production time sequence position, referring to... Figure 2 .

[0041] Specifically, the execution controller identifies the current production sequence position for the labeling action based on pulse signals fed back from the production line sensors. When the production sequence position is the first position, i.e., the first labeling action, a 256-bit binary random sequence pre-generated by the random number generator based on a hardware unique identifier is retrieved from the hardware secure memory. This sequence is defined as the initial vector of the labeling seed and used as the labeling history feedback value for the current position. When the production sequence position is the second or subsequent incrementing position, a 256-bit labeling topological entropy chain calculated from the previous production sequence position is acquired in real time and defined as the labeling history feedback value.

[0042] The cascaded mapping algorithm is a data fusion logic based on bit-level cross-arrangement. This algorithm concatenates and aggregates the 256-bit instantaneous index credentials generated by the current sequence position and the 256-bit historical index feedback value according to a preset odd-even bit cross-stepping rule. Specifically, the 256-bit instantaneous index credentials are defined as the current bit set A, and the 256-bit historical index feedback value is defined as the historical bit set B. Using cross-indexing logic, the nth bit of bit set A is mapped to the 2nth bit weight space (odd bit) of the 512-bit output sequence, and the nth bit of bit set B is mapped to the (2n+1)th bit weight space (even bit) of the output sequence. Through the discrete interleaving of the bit sequences, the instantaneous entropy and historical entropy are recombined at the bit level into a 512-bit sequence association feature with bidirectional logical dependence. The sequence association feature is a binary data block whose structure includes a double-helix association bit array formed by the bit-level spatial bit interleaving of the instantaneous feature bits of the current sequence position and the previous historical feedback feature bits.

[0043] The process of performing nonlinear logical compression on sequence-associated features to generate a sorted topological entropy chain includes: dividing the sequence-associated features into high-order byte blocks and low-order byte blocks, and inputting a preset cyclic hash iteration operator, wherein the operator adopts a custom cyclic hash operator based on the SP network structure; in each round, performing a preset number of cyclic left shift operations on the sequence-associated features, and using a permutation table to perform cross-byte bit value permutations on the shifted bit sequence; logically folding the iteratively confused bit sequence at the center position, so that the high-order byte blocks and low-order byte blocks perform bitwise XOR logical aggregation; and defining the compressed bit sequence as the sorted topological entropy chain of the current production time sequence position.

[0044] Specifically, the execution controller first loads the 512-bit sequence-associated features into a high-speed register. Based on the center bit (bit 256), it is divided into two equal-length logical blocks: the high-order byte block: carrying the first 256 bits of raw information (including the initial interleaving bits of instantaneous credentials and historical feedback); and the low-order byte block: carrying the last 256 bits of raw information.

[0045] The partitioned data is input into a cyclic hash iterative operator, which performs eight rounds of iterative operations. This cyclic hash iterative operator is a logic operation program that supports multi-round cyclic compression. Each round of iterative operation includes two actions: Cyclic left shift: The entire 512-bit sequence is cyclically shifted left by k bits (e.g., 13 or 17 bits). This action breaks the local parity correlation formed by cascaded mapping, allowing the previous historical feedback bits and the current instantaneous feature bits to be recombined within a larger bit width range. Cross-byte bit value permutation: Using a permutation table, a global mapping is performed on the shifted bit sequence. Any fluctuation in any 1 bit in the input sequence, after processing by the permutation table, will have its effect spread to multiple discontinuous byte spaces.

[0046] After eight rounds of deep obfuscation, the operator performs logical folding compression. This action no longer involves shifting, but instead "folds" the 512-bit sequence back to 256-bit space around the center bit. Each byte in the obfuscated high-order byte block is then XORed with its corresponding byte in the low-order byte block. Unlike simple truncation and discarding, folding XOR ensures that every feature point in the original 512-bit data logically contributes to the final output. Even with only minor differences in the high-order or low-order blocks, the folded result will exhibit a significant jump, defining the aggregated 256-bit fixed-length bit sequence as the index topological entropy chain of the current production time sequence. This index topological entropy chain not only contains the physical fingerprint information of the current tag but also, through XOR aggregation logic, locks down all historical changes in the entire production line since the "index seed initial vector," forming a digital certificate with strong time-series constraints. This scheme uses multi-round cyclic shifting and obfuscation and folding XOR mapping logic to reduce the feature bit width while deeply locking the correlation between instantaneous physical features and historical state features, ensuring that the generated topological entropy chain has extremely strong data integrity verification capabilities.

[0047] This invention establishes a recursive locking mechanism between data, ensuring that the entropy chain generated by each labeling sequence position contains the logical traces of all preceding actions. This effectively prevents code skipping, relabeling, or illegal data insertion of traceability identifiers during assembly line operations, thus ensuring the logical integrity of the entire production sequence.

[0048] Furthermore, the process of extracting the logical feature vector includes: retrieving the labeling topology entropy chain through the consensus node of the distributed ledger and performing label machine running state logic decomposition processing; separating the labeling identity coupling fingerprint of the current labeling position from the labeling topology entropy chain, inputting the labeling identity coupling fingerprint into a preset verification protocol to perform feature dimension restoration processing, and generating a peeled steady-state logical image; extracting discrete feature components containing instruction stream underlying state bias data and execution end physical state texture data from the peeled steady-state logical image, performing spatial reconstruction processing on the discrete feature components, and restoring to generate a logical feature vector.

[0049] Specifically, the consensus nodes of the distributed ledger obtain the 256-bit tag topology entropy chain corresponding to the current tagging sequence in real time, and use the decoding rules stored in the ledger state database to perform tagging machine running state logic decomposition processing.

[0050] By synchronously retrieving the stored historical feedback values, feature stripping based on reverse folding logic is performed on the indexed topological entropy chain. Specifically, the verification end first retrieves the 256-bit indexed topological entropy chain of the current position from the distributed ledger, and simultaneously retrieves the 256-bit historical feedback value of the indexed position referenced during its generation. Utilizing the reversible property of XOR operation, the historical feedback value of the indexed position is used as an operator to perform a bit-by-bit XOR logic reset operation with the indexed topological entropy chain, reconstructing a 256-bit intermediate mirror sequence containing complete feature information of the current position. This intermediate mirror sequence is then input into the reverse hash iteration operator. This operator performs a reverse cyclic shift (such as a right shift) that is completely symmetrical with the production end, as well as position restoration processing based on the reverse permutation table. Through 8 rounds of reverse diffusion mapping, the nonlinear confusion effect generated by the cyclic hash iteration of the production end is eliminated. After the reverse confusion elimination, the sequence's position order returns to the state before the cascade mapping. At this point, through a specific logical index, a bit stream representing the core physical and instruction characteristics of the current sequence position is extracted from the sequence, ultimately producing a fixed-length 256-bit sequenced identity coupled fingerprint.

[0051] The sequenced identity coupled fingerprint is input into a preset verification protocol. This protocol is a feature recovery logic based on an inverse permutation matrix and a linear feedback shift register. According to a pre-stored 256-order feature mapping table, bit-level inversion and permutation restoration are performed on the 256-bit sequenced identity coupled fingerprint, expanding the compressed data stream and mapping it back to the original feature space to generate a 256-bit peeled steady-state logic image. The peeled steady-state logic image consists of two independent feature components, separated by a preset mask extraction logic. Specifically, first, a first mask operator (high-order mask) with the same bit width as the peeled steady-state logic image is constructed. In this embodiment, this operator is a binary sequence where the first 128 bits are all "1"s and the last 128 bits are all "0"s (i.e., 0xFF...F00...0). Then, a parallel bitwise AND operation is performed between the 256-bit peeled steady-state logic image and the first mask operator. The low-order intervals of the mirror image are masked using logic gates, and a logical right shift operation is performed on the calculation result to align the high-order data to the standard starting address. Finally, 128-bit instruction stream low-level state bias data, representing the instruction timing characteristics, is extracted from the filtered bitstream. Similarly, the same operation is performed using the inverted second mask operator (low-order mask) to extract 128-bit execution-side physical state texture data.

[0052] The 128-bit instruction stream low-level bias data and the 128-bit execution-end physical state texture data are used as discrete feature components. The process of performing spatial reconstruction processing on the discrete feature components to restore and generate logical feature vectors includes: retrieving a preset tensor arrangement rule to establish a mapping topology relationship between the discrete feature components and the index bits of the target feature vector; according to the mapping topology relationship, the instruction stream low-level bias data is used as the first tensor component and concatenated and aligned with the execution-end physical state texture data as the second tensor component to construct an initial combined bit vector; according to the tensor arrangement rule, the initial 256-bit combined bit vector is reordered based on the position permutation matrix to correct the spatial bias of the feature data during the encoding process; after the bit reordering process, a fixed-length logical feature vector is restored.

[0053] The refactoring program retrieves a preset 256-bit tensor arrangement rule from the security configuration module. This rule is a spatial decoupling inverse transformation and includes a preset index mapping table. The index mapping table defines a bidirectional mapping function between each bit in the original logical feature vector and the bit order of the discrete feature components. Define the 128-bit instruction stream low-level bias data as a vector. Define the 128-bit execution-side physical state texture data as a vector. Perform a cascading alignment operation: Placed in the high range [255, 128], The bits are placed in the low-order interval [127, 0] and concatenated to form a temporary 256-bit initial combined bit vector. The position permutation matrix in the tensor arrangement rule is retrieved. This position permutation matrix is ​​a preset index transformation operator, specifically representing a set of nonlinear position mapping relationships. This matrix does not perform simple random permutation, but rather acts as a symmetric inverse transformation matrix generated based on the physical instruction space topology during feature extraction at the production end. By reprojecting the discrete bits in the initial combined bit vector back to their original logical coordinates, precise correction of the feature data space bias is achieved.

[0054] The program performs nonlinear repositioning of bits in the initial combined bit vector based on a position permutation matrix. Specifically, it iterates through each binary bit component of the initial vector, and based on the offset and spatial index provided by the permutation matrix, migrates the binary bit component to the specified coordinate position in the target vector. For example, if the matrix defines the original 10th bit as corresponding to the 3rd bit in the instruction stream, the program extracts the bit at offset address 3 in the initial combined bit vector and repositions it to the 10th index position in the target vector. During the repositioning process, the program automatically compensates for bit order offsets that may be introduced by pre-encoding (such as SM3 compression or SPN obfuscation). Integrity checks (such as parity checks or cyclic redundancy checks) are performed on the repositioned bit stream to ensure that all 256 bits are accurately reset. Finally, a fixed-length 256-bit binary sequence is restored, defined as the original logical feature vector. This vector not only restores the original data value but also reconstructs the spatial geometric distribution characteristics of the instruction features and physical features in the initial state. This scheme eliminates the "feature distortion" generated during multi-layer compression and obfuscation, achieving alignment of production-end features and verification-end features in the logical space.

[0055] This invention achieves logical restoration of highly compressed feature data through the reverse deconstruction logic of distributed ledger consensus nodes, ensuring that the underlying instruction logic and physical operation status information of the production end can be accurately extracted at the verification end, providing comparable data entities for subsequent cross-dimensional identity consistency retrieval.

[0056] Furthermore, the process of executing the runtime logic decomposition of the label includes: inputting the label topology entropy chain into a preset peeling and deconstruction model; using the pre-stored label history feedback value in the peeling and deconstruction model as the deconstruction benchmark, performing XOR residual extraction on the label topology entropy chain to offset the weight noise of the label history feedback value on the current position data, and separating the feature domain to be verified of the label instantaneous credential; inputting the feature domain to be verified into the inverse mapping logic of the cascaded mapping algorithm to perform bit logic decomposition, and restoring and generating the label identity coupled fingerprint.

[0057] Specifically, the distributed ledger consensus node imports the received 256-bit sequence topological entropy chain into the stripping and deconstruction model. This model, as a state-recovery-based logical operator, queries the ledger state database to anchor in real-time the 256-bit historical feedback value of the previous sequence bit associated with the current bit to be verified. The stripping and deconstruction model utilizes a reverse operator that is completely symmetrical to the "cyclic shift-permutation-folding" process on the production side to perform restoration: First, the model calls a pre-set state compensation operator, using the 256-bit sequence topological entropy chain as the input vector. By performing a reverse spatial mapping based on the previous historical feedback value, the bit weights generated by the "logical folding" are restored to a 512-bit intermediate state sequence. Subsequently, the peeling and deconstruction model executes the "XOR residual extraction" logic: a reverse cyclic shift, consistent with the production step size, is performed on the 512-bit intermediate sequence to eliminate global obfuscation; using the 256-bit historical feedback value of the previous sequence position as a timing reference mask, even-numbered bits in the 512-bit sequence are XORed bit-by-bit according to the parity cross-indexing rules of the encapsulation stage. This operation cancels out the coupling weight noise formed by historical data at the current timing position, thereby accurately extracting the 256-bit instantaneous credential component from the 512-bit sequence, achieving reconstruction from the compressed feature space to the original obfuscated data block. The folded dimensional features are remapped from 256 bits back to the original high-dimensional feature space, thus reconstructing a 448-bit original obfuscated data block containing all features of the current action, i.e., separating the instantaneous credential.

[0058] The stripping and deconstruction model invokes the reverse execution logic of the cascaded mapping algorithm to process the 448-bit instantaneous credential according to a preset bit reversal path. This process separates the intertwined instructions and physical features by performing a reverse mapping of the spatial coordinates of the bits: first, it removes the nonlinear bias of the 64-bit credential phase vector on the feature bits according to a preset index, and then separates a 128-bit globally unique sequence. Finally, from the remaining data components after stripping, the original 256-bit binary sequence is extracted through bit merging and tensor restoration, thus restoring and generating a credential identity coupled fingerprint that is logically symmetrical with the production end.

[0059] This invention achieves accurate decomposition of entropy chain data with strong temporal coupling by constructing a symmetric reverse deconstruction model and a weighted noise cancellation mechanism. This not only completely eliminates the logical interference of preceding data on the current identity features, but also ensures that lossless extraction and high-fidelity restoration of the underlying identity fingerprint can still be achieved in a high-dimensional data compression environment.

[0060] Furthermore, the process of generating logical anchored credentials and completing data security verification includes: performing identity verification by comparing the logical feature vector with the hardware base state features of the distributed ledger, and outputting the identity matching weight; calling the pre-set peeling verification operator in the distributed ledger, comparing the logical topological loop of the index topological entropy chain with the global state root of the distributed ledger through the peeling verification operator, and outputting the sequence verification coefficient; inputting the identity matching weight and the sequence verification coefficient into a logical discriminant function for aggregation processing to generate logical anchored credentials; using the logical anchored credentials as a trigger signal for asynchronous blockchain notarization, when the logical anchored credentials determine that the identity is consistent and the logical topological loop is closed, triggering asynchronous notarization update to complete data security verification. (Refer to...) Figure 3 This is a flowchart illustrating the distributed ledger verification and logical anchoring process according to an embodiment of the present invention.

[0061] Specifically, the distributed ledger consensus node aligns the parsed 256-bit logical feature vector with the 256-bit hardware base state features registered in the ledger. Before performing the alignment determination, the execution controller calls a preset fuzzy extraction operator, which is constructed based on a fault-tolerant reconstruction protocol for physically non-cloning functions. The operator retrieves auxiliary data associated with the current device in the distributed ledger in real time (i.e., redundant check bits generated using error correction coding during the device initialization registration phase) and uses this as a calibration benchmark. Subsequently, the operator uses the BCH decoding algorithm to perform bit-level error correction and restoration on the collected noisy logical feature vector: by calculating the feature offset between the noisy vector and the auxiliary data, it locates the out-of-position bits caused by physical environmental disturbances and performs logical inversion, thereby restoring a bit-level stable sequence identity coupled fingerprint in a production environment with physical background noise. Then, the stable fingerprint after error correction and restoration is used as input to perform a similarity evaluation based on Hamming distance. By performing a bitwise XOR operation, the total number of inconsistent bits between the two vectors is counted. When the number of out-of-position bits is within the preset dynamic fault tolerance threshold range, the physical features are determined to be consistent. The preset dynamic fault tolerance threshold ranges from 5 bits to 8 bits. Its determination is based on: the thermal noise distribution of analog-to-digital conversion for a specific actuator sensor, ensuring the threshold covers 99.9% of the steady-state fluctuation range; and the probability density of the 256-bit feature space, ensuring that the feature spaces of illegally tampered data and legitimate data are completely orthogonal under this threshold. The execution controller performs dynamic fine-tuning in 1-bit increments within this range based on the current labeling frequency and the signal-to-noise ratio of the sensor feedback signal, and outputs the identity matching weight based on the deviation. Through this cascaded verification mechanism of first correcting errors and then comparing, it is ensured that random noise generated at the physical execution end will not cause avalanche interference to subsequent hash iterations and XOR encapsulation.

[0062] The distributed ledger consensus nodes synchronously invoke and execute the stripping verification operator to perform cross-dimensional identity verification and temporal integrity logic determination. The stripping verification operator performs feature stripping logic based on reverse folding mapping and topological loop detection logic based on the global state root to perform multi-dimensional cross-alignment between the underlying physical features of the device to be verified and the global logical topology of the ledger, ultimately outputting a sequence verification coefficient to determine the authenticity of the data. The indexed topological entropy chain, as the payload of authorized transactions, is encapsulated in the Merkle leaf node of the block body. The stripping verification operator retrieves the currently verified 256-bit indexed topological entropy chain from the consensus node's local ledger database and extracts the corresponding 256-bit global state root (i.e., transaction tree root or state tree root) from the current block header. The stripping verification operator verifies whether the entropy chain satisfies the pointing continuity of the hash chain structure by performing Merkle proof path verification. The specific procedure is as follows: The peeling verification operator first obtains a set of proof path sequences composed of the hash values ​​of adjacent nodes from the consensus node based on the hash value of the entropy chain to be verified and its transaction index within the block. Then, the operator uses the entropy chain to be verified as the starting hash and performs a step-by-step cascading hash operation with the node hashes in the proof path sequence according to a preset field concatenation order, simulating the upward backtracking process of a Merkle tree. If the finally calculated root hash value is completely consistent with the global state root, it proves that the entropy chain satisfies the pointing continuity of the hash chain structure in the global logical topology of the ledger. Under the premise of confirming physical identity consistency and topological logical closure, the output sequence verification coefficient is 1.0, proving that the production time sequence logic has not been tampered with, ensuring the consistency of traceability data during generation and circulation.

[0063] The logical discriminant receives the identity matching weight and sequence verification coefficient, and performs aggregate discrimination. The specific process includes: the discriminant performs threshold logic judgment, i.e., the judgment result is true if and only if (identity matching weight ≥ 0.90) and (sequence verification coefficient = 1.0). When both physical matching degree and sequential continuity meet the criteria, i.e., after the judgment passes, the consensus node's private key is invoked to perform digital signature on the data packet containing the device ID, verification pass instruction, and current timestamp, generating a logical anchoring credential. This credential, as a state transition primitive, carries dual trust endorsement of physical and logical states. The logical anchoring credential, as an asynchronous evidence storage trigger signal, activates the consensus mechanism for state updates. The distributed ledger synchronously updates the serialized identifier and device characteristics for the current labeling action across all network nodes, completing the consensus-based solidification of data, thereby achieving a closed loop for data security verification.

[0064] This invention combines the fault tolerance of the underlying hardware fingerprint with the precise verification of the blockchain's global topology to construct an anchoring mechanism with self-healing and high reliability. It not only eliminates the interference of physical environmental noise on verification stability but also ensures strong coupling between the traceability chain and the physical space axis, achieving automated and secure management of production process data with "immediate verification upon production and authenticity upon storage."

[0065] This invention constructs a unique sequenced identity coupled fingerprint by deeply fusing the physical state texture (tension, negative pressure, etc.) of the execution end with the instruction flow bias, fundamentally solving the technical problem of easily forged device identities. Utilizing sequenced topological entropy chain technology, it aggregates instantaneous action features with historical feedback values ​​from the distributed ledger in a chain-like manner, achieving strong correlation and immutability of production data along the timeline. Combining reverse parsing of consensus nodes with topological closed-loop verification of the global state root, this solution effectively identifies sequence replay and device impersonation attacks, and achieves real-time credibility solidification and secure verification of traceability data in high-speed production environments through asynchronous evidence storage updates.

[0066] Example 2

[0067] This embodiment selects an automated high-speed packaging line of a large vaccine manufacturer as the application object. This line labels 600 bottles per minute, placing extremely high demands on the real-time performance and security of data processing. At the hardware deployment level, the intelligent labeling device serves as the core node, integrating a high-speed industrial control processor (such as an FPGA or high-performance MCU) and a distributed ledger secure communication module. A sensor matrix at the bottom of the device is connected to the processor via an isolated bus, ensuring the lossless capture of physical features within millisecond-level execution cycles. During the initialization phase, the device uploads its unique hardware state characteristics (including a unique factory serial number, actuator motor impedance fingerprint, and other 256-bit features) to the consensus node of the distributed ledger for registration, establishing a root of trust.

[0068] When the production line starts operating and the first vaccine vial enters the labeling station, the equipment's underlying sensing system acquires the current harmonic components of the unwinding servo drive in real time. This data, combined with the microsecond-level offset of the signal captured by the photoelectric sensor at the peeling plate and the negative pressure pulsation data of the vacuum suction cup at the moment of label pickup, is aggregated to generate the first label capture pulse entropy source. This entropy source performs decoupling and transformation of feature dimensions through a nonlinear mapping algorithm, producing a logical feature vector containing the underlying state bias of the instruction stream and the physical state texture of the execution end. Ultimately, it generates a coupled fingerprint with a physically unique label sequence identity.

[0069] Subsequently, the serialized traceability QR code on the vaccine vial is automatically read, and the millisecond-level time coordinate offset of the labeling action is captured. Through XOR encapsulation, the label identity is coupled with the fingerprint, the globally unique sequence corresponding to the traceability QR code, and the time coordinate offset, and bitwise logic is coupled to generate the label instantaneous credential for that position. Since it is currently the first labeling position, the preset label seed initial vector is called as the historical feedback value. Through the cascaded mapping algorithm and the cyclic hash iteration operator, the label instantaneous credential and the seed vector are subjected to chain-like aggregation calculation to generate the label topology entropy chain of the first position and uploaded to the distributed ledger consensus node.

[0070] As the production line continues to the next stage (such as the Nth labeling action), after generating the instantaneous labeling credential for the current stage, it automatically retrieves the labeling topology entropy chain produced in the (N-1)th stage as the labeling history feedback value. This is then processed through iterative cyclic hashing to generate the labeling topology entropy chain for the Nth stage. This recursive process ensures that the traceability data for each vaccine vial contains the logical imprint of all previous vials, forming a logically tight digital chain.

[0071] After receiving the entropy chain of the sequence topology for each position, the consensus node of the distributed ledger initiates the peeling and deconstruction model to perform logical decomposition of the machine's running state. It uses cyclic hashing reverse mapping logic to offset the weight noise of historical feedback values, separating and reconstructing the sequence identity coupled fingerprint from the entropy chain. The verification operator calculates the feature distance between the reconstructed fingerprint and the pre-stored hardware base state features in the ledger, and simultaneously compares the logical topological closure of the entropy chain with the global state root of the ledger.

[0072] When both the identity matching weight and the sequence verification coefficient meet the preset thresholds, a logical anchor certificate is generated and an asynchronous evidence update is triggered, permanently storing the vaccine vial's production time, device identity, and logical sequence number in the blockchain. If illegal label replacement or device identity anomalies occur during production, the deconstruction model will refuse to perform the evidence update and trigger an alarm because it cannot reconstruct a valid identity fingerprint or detects a break in the topological logic, thus ensuring the high security of the serialized traceability data.

[0073] This invention achieves a deep integration of physical execution characteristics and digital logic characteristics in the field of packaging labeling. Without increasing the computational burden on the production side, this solution leverages the decentralized nature of blockchain and the non-replicable nature of physical characteristics to provide robust underlying technical support for the full lifecycle traceability of high-security products.

[0074] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A blockchain-based method for verifying serialized traceability data, characterized in that, include: Acquire unwinding tension harmonic data, peeling blade data, and label suction negative pressure data of labeling and packaging equipment to generate a label motor capture pulse entropy source; Perform an association transformation on the flag capture pulse entropy source, extract the logical feature vector containing the underlying state bias of the instruction stream and the physical state texture of the execution end, and generate the flag sequence identity coupled fingerprint; Obtain the serialized traceability identifier and the time coordinate offset of the automated labeling action, and perform XOR encapsulation based on the label sequence identity coupled with the fingerprint to generate the label sequence instantaneous credential; Obtain the labeling sequence historical feedback value, perform chain aggregation and hash iteration processing on the labeling instant credential and labeling historical feedback value to generate a labeling topology entropy chain; Consensus nodes based on the distributed ledger perform reverse parsing on the sequence topology entropy chain to extract the logical feature vector from the sequence identity coupling fingerprint; and perform identity verification by comparing the logical feature vector with the hardware base state features in the distributed ledger. By comparing the logical topology loop of the indexed topology entropy chain with the global state root of the distributed ledger, a logical anchor certificate is generated. When the logical anchor certificate determines that the identity is consistent and the logical topology loop is closed, asynchronous evidence storage update is triggered to complete the data security verification.

2. The blockchain-based serialized traceability data verification method according to claim 1, characterized in that, The process of generating the label-capturing pulse entropy source includes: acquiring the frequency domain characteristic distribution of the motor current at the AC servo drive end of the labeling and packaging equipment, and mapping it to the generated unwinding tension harmonic data; acquiring the pulse edge response time offset of the label peeling trigger signal and mapping it to the generated peeling blade data; acquiring the negative pressure pulsation transient data when the vacuum generator performs the label suction action and mapping it to the generated suction negative pressure data; and cascading and assembling the unwinding tension harmonic data, peeling blade data, and suction negative pressure data to generate the label-capturing pulse entropy source.

3. The blockchain-based serialized traceability data verification method according to claim 1, characterized in that, The process of generating the serialized identity coupled fingerprint includes: inputting the serialized motion capture pulse entropy source into a nonlinear mapping algorithm to perform an association transformation and extract physical disturbance features; using the physical disturbance features to perform disturbance mapping on the discrete sampled data of the execution controller instruction execution cycle to generate the instruction stream bottom-level state bias; using the physical random disturbance features to perform orthogonal decomposition mapping on the waveform feature components of the physical actuator action feedback signal at the execution end to generate the execution end physical state texture; concatenating the instruction stream bottom-level state bias with the execution end physical state texture execution features to construct a logical feature vector; and compressing and mapping the logical feature vector through a hash algorithm to generate the serialized identity coupled fingerprint.

4. The blockchain-based serialized traceability data verification method according to claim 1, characterized in that, The process of generating the labeling instantaneous credential includes: obtaining a serialized traceability identifier code and using it as a globally unique sequence; obtaining the time coordinate offset of the labeling action of the automated execution terminal relative to the sampling zero point and converting it into a labeling machine phase vector; inputting the labeling identity coupled fingerprint, the globally unique sequence, and the labeling machine phase vector into an XOR encapsulation operator to perform bit-by-bit logical coupling to obtain a peeling-and-apply linkage instantaneous stream carrying hardware physical characteristics, data identity characteristics, and action spatiotemporal characteristics; inputting the peeling-and-apply linkage instantaneous stream into a discrete transformation operator to perform nonlinear obfuscation processing to generate the labeling instantaneous credential.

5. The blockchain-based serialized traceability data verification method according to claim 1, characterized in that, The process of generating the labeling topology entropy chain includes: acquiring and identifying the production time sequence position of the current labeling action; if the production time sequence position is the first position, then defining the preset labeling seed initial vector as the labeling history feedback value; if the production time sequence position is an increasing position, then acquiring the labeling topology entropy chain generated by the previous production time sequence position as the labeling history feedback value; inputting the labeling instantaneous credentials and the labeling history feedback value into a cascade mapping algorithm to perform chain aggregation and construct sequence association features; inputting the sequence association features into a cyclic hash iteration operator to perform nonlinear logic compression and generate the labeling topology entropy chain of the current production time sequence position.

6. The blockchain-based serialized traceability data verification method according to claim 1, characterized in that, The process of extracting the logical feature vector includes: retrieving the labeling topology entropy chain through the consensus node of the distributed ledger and performing label machine running state logic decomposition processing; separating the labeling identity coupling fingerprint of the current labeling position from the labeling topology entropy chain, inputting the labeling identity coupling fingerprint into a preset verification protocol to perform feature dimension restoration processing, and generating a peeled steady-state logical image; extracting discrete feature components containing instruction stream underlying state bias data and execution end physical state texture data from the peeled steady-state logical image, performing spatial reconstruction processing on the discrete feature components, and restoring to generate a logical feature vector.

7. A blockchain-based serialized traceability data verification method according to claim 6, characterized in that, The process of executing the running state logic decomposition of the label includes: inputting the label topology entropy chain into a preset peeling and deconstruction model; using the pre-stored label history feedback value in the peeling and deconstruction model as the deconstruction benchmark, performing XOR residual extraction on the label topology entropy chain to cancel the weight noise of the label history feedback value on the current position data, and separating the feature domain to be verified of the label instantaneous credential; inputting the feature domain to be verified into the inverse mapping logic of the cascaded mapping algorithm to perform bit logic decomposition, and restoring and generating the label identity coupled fingerprint.

8. A blockchain-based serialized traceability data verification method according to claim 1, characterized in that, The process of generating logical anchored credentials and completing data security verification includes: performing identity verification on the logical feature vector and the hardware base state features of the distributed ledger, and outputting the identity matching weight; calling the pre-set peeling verification operator in the distributed ledger, comparing the logical topological closed loop of the index topological entropy chain with the global state root of the distributed ledger through the peeling verification operator, and outputting the sequence verification coefficient; inputting the identity matching weight and the sequence verification coefficient into the logical discriminant function for aggregation processing to generate logical anchored credentials; using the logical anchored credentials as the trigger signal for asynchronous blockchain notarization, when the logical anchored credentials determine that the identity is consistent and the logical topological closed loop is achieved, triggering asynchronous notarization update to complete the data security verification.

Images