A Method and System for Real-time Synchronization of Receiving Information and Abnormal Early Warning

Abstract

The present invention discloses a method and system for real-time synchronization of recipient information and anomaly warning, which relates to the technical fields of intelligent logistics information processing and data security. It includes collecting logistics trajectory data and constructing a spatio-temporal graph model, and extracting trajectory features through a spatio-temporal graph convolutional network; using a Transformer model for trajectory time series prediction and judging anomaly events based on an anomaly detection threshold; combining zero-knowledge proof and blockchain technology to encrypt and verify logistics data and store evidence, and constructing a logistics scheduling optimization and dynamic anomaly recovery strategy. The method of the present invention constructs a complete system for real-time synchronization of logistics information and anomaly warning through spatio-temporal graph modeling, ST-GCN trajectory feature extraction, Transformer prediction analysis, blockchain evidence storage, and reinforcement learning scheduling optimization; by introducing blockchain and ZKP technology, it ensures the integrity and privacy security of data and reduces the risks of logistics fraud and information tampering.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent logistics information processing and data security, and specifically provides a method and system for real-time synchronization of recipient information and anomaly warning. Background Art

[0002] With the rapid development of e-commerce and intelligent logistics, the real-time synchronization of logistics information and anomaly warning have become one of the key technologies to improve logistics efficiency and user experience. At present, the logistics industry widely applies technologies such as GPS positioning, RFID identification, wireless sensor network (WSN), and big data analysis to monitor the transportation trajectory of express parcels, and uses machine learning or rule-based algorithms for anomaly detection. In recent years, the application of deep learning in the logistics field has gradually increased. For example, anomaly detection methods based on long short-term memory network (LSTM) and recurrent neural network (RNN) can perform time series modeling on logistics trajectories and provide certain prediction capabilities. In addition, blockchain technology has also been applied in logistics tracking to improve the transparency and security of logistics information. However, although these technologies have improved the intelligent level of logistics trajectory monitoring to a certain extent, there are still problems such as data synchronization delay, insufficient anomaly detection accuracy, and lack of data security guarantee, which restrict the further development of intelligent logistics systems.

[0003] The existing logistics information synchronization and anomaly detection methods mainly rely on traditional time series modeling methods and rule-based anomaly detection, but there are many limitations in practical applications. The time series prediction models based on RNN and LSTM are difficult to handle long time series dependencies. Especially in logistics trajectory data, due to the large time span of parcel transportation, the existing methods are difficult to accurately capture anomaly patterns across regions and time periods, resulting in a low anomaly recognition rate. The existing processing methods for trajectory data are relatively simple, lacking the modeling of the spatial relationship of the logistics path. Most systems only rely on the trajectory changes of GPS data and ignore the topological relationship between logistics nodes, resulting in low accuracy of path anomaly detection. The data security problem is prominent. The traditional logistics data storage method relies on a centralized server, which is easily subject to data tampering and forgery. Moreover, most current blockchain evidence storage schemes can only provide static data traceability and cannot effectively combine privacy computing technology for security verification. For anomaly warning, the existing systems usually adopt fixed thresholds or detection methods based on statistical features, failing to implement a dynamic and adaptive anomaly warning strategy and unable to provide accurate scheduling optimization in the case of complex changes in the logistics environment. Based on the above problems, traditional methods are difficult to meet the requirements of efficient, accurate, and traceable logistics information synchronization and anomaly detection, and there is an urgent need for an innovative solution that combines spatio-temporal modeling, deep learning, and blockchain security evidence storage to solve the limitations of existing technologies and improve the intelligent level of logistics systems. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed.

[0005] Therefore, the technical problems solved by the present invention are as follows: the logistics trajectory data modeling method lacks the deep learning ability for spatio-temporal features, has a low accuracy of anomaly detection, lacks the prediction ability for long-time series trajectories, has insufficient logistics data security, cannot prevent data tampering, and how to achieve accurate anomaly detection based on spatio-temporal graph neural networks and Transformer, and combine zero-knowledge proof and blockchain technology to ensure the security and traceability of logistics data.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a method for real-time synchronization of recipient information and anomaly warning, including collecting logistics trajectory data and constructing a spatio-temporal graph model, and extracting trajectory features through a spatio-temporal graph convolutional network; using a Transformer model for trajectory time series prediction, and judging anomaly events based on an anomaly detection threshold; combining zero-knowledge proof and blockchain technology to perform encrypted verification and storage of logistics data, and constructing a logistics scheduling optimization and dynamic anomaly recovery strategy; the trajectory feature extraction includes using a three-layer spatio-temporal graph convolutional network, the first layer performs local spatio-temporal feature extraction to process short-time logistics path information; the second layer performs global information aggregation to learn long-time series cross-regional logistics features; the third layer performs anomaly pattern recognition to detect abnormal trajectories and provide interpretable anomaly scores.

[0007] As a preferred embodiment of the method for real-time synchronization of recipient information and anomaly warning according to the present invention, wherein: constructing the spatio-temporal graph model includes constructing an express delivery transportation network using a graph data structure, where nodes represent logistics locations, edges represent the transportation paths of packages, and the weights of the edges are dynamically updated according to transportation time and package status.

[0008] As a preferred embodiment of the method for real-time synchronization of recipient information and anomaly warning according to the present invention, wherein: the trajectory feature extraction includes extracting the sequential transportation mode and spatial path information of the package;

[0009] Set a historical trajectory window, embed time series information into the graph structure, and capture abnormal patterns of different express delivery paths.

[0010] As a preferred embodiment of the method for real-time synchronization of recipient information and anomaly warning according to the present invention, wherein: performing trajectory time series prediction includes inputting the logistics trajectory features extracted by the spatio-temporal graph convolutional network into the Transformer model, and using the multi-head self-attention mechanism to analyze the historical trajectory patterns of the package; through the time series prediction ability of the Transformer, calculate the normal trajectory distribution of the express delivery package at future time points.

[0011] As a preferred solution of the real-time synchronization and exception warning method for recipient information described in the present invention, wherein: the determination of exception events includes setting an exception detection threshold, and using a calculation method based on Euclidean distance or Mahalanobis distance to measure the deviation between the actual trajectory and the predicted trajectory; when it is detected that the trajectory deviation exceeds the set exception detection threshold, an exception event is triggered, and a warning signal is sent; when an exception occurs, warning information is sent to the delivery personnel and the management platform; different coping strategies are triggered according to the type of exception; the express packages marked as exceptions are rechecked to confirm the authenticity of the exceptions.

[0012] As a preferred solution of the real-time synchronization and exception warning method for recipient information described in the present invention, wherein: the encryption verification and storage of logistics data includes using zk-SNARKs to encrypt and prove the key logistics data; before storing the logistics data, a zero-knowledge proof is generated, and the authenticity of the data can be verified without accessing the original logistics information; the blockchain storage technology is used to store the hash value of the logistics event in the distributed ledger; the Merkle tree structure is used to organize the logistics data, and the authenticity of the record is verified through the root hash; when a logistics exception event is detected, based on the historical data stored in the blockchain, the abnormal trajectory is traced back, and it is verified whether the data has been tampered with during the storage process; the intelligent contract is used to automatically execute the abnormal backtracking logic, and when data inconsistency is found, a warning is sent to the relevant responsible parties and the violation behavior is recorded.

[0013] As a preferred solution of the real-time synchronization and exception warning method for recipient information described in the present invention, wherein: the construction of the logistics scheduling optimization and dynamic exception recovery strategy includes using reinforcement learning to optimize the reallocation path of abnormal packages and minimizing delays when the courier re-plans the route; training the RL model in combination with historical exception data to enable the platform to autonomously learn the optimal scheduling strategy; on the basis of blockchain storage, a verifiable exception recovery process is constructed, and exception handling is carried out based on logistics specifications.

[0014] Another object of the present invention is to provide a real-time synchronization and exception warning system for recipient information, which can predict the trajectory time series by using the Transformer model and judge the exception event based on the exception detection threshold, and solves the problem that the current logistics trajectory data modeling technology lacks the prediction ability for long-time series trajectories.

[0015] As a preferred solution of the real-time synchronization and abnormal warning system for received information described in the present invention, it includes a data processing module, an abnormal warning module, and an optimization and recovery module; the data processing module is used to collect logistics track data and construct a spatio-temporal graph model, and extract track features through a spatio-temporal graph convolutional network; the abnormal warning module is used to predict the track time series by using a Transformer model and judge abnormal events based on an abnormal detection threshold; the optimization and recovery module is used to combine zero-knowledge proof and blockchain technology to encrypt and verify logistics data and store evidence, and construct a logistics scheduling optimization and dynamic abnormal recovery strategy.

[0016] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the method for real-time synchronization and abnormal warning of received information are realized.

[0017] A computer-readable storage medium stores a computer program thereon. When the computer program is executed by a processor, the steps of the method for real-time synchronization and abnormal warning of received information are realized.

[0018] The beneficial effects of the present invention: The method for real-time synchronization and abnormal warning of received information provided by the present invention constructs a complete real-time synchronization and abnormal warning system for logistics information through spatio-temporal graph modeling, ST-GCN track feature extraction, Transformer prediction analysis, blockchain evidence storage, and reinforcement learning scheduling optimization; compared with traditional logistics information processing methods, the present invention not only improves the accuracy of track data, enhances the abnormal detection ability, but also optimizes logistics scheduling and improves the response speed of the system; by introducing blockchain and ZKP technology, the integrity and privacy security of data are guaranteed, and the risks of logistics fraud and information tampering are reduced. The present invention has better effects in terms of efficiency, reliability, and accuracy. Description of the Drawings

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0020] Figure 1 It is the overall flowchart of a method for real-time synchronization and abnormal warning of received information provided in Embodiment 1 of the present invention.

[0021] Figure 2 It is the overall flowchart of a real-time synchronization and abnormal warning system for received information provided in Embodiment 3 of the present invention. Detailed Embodiments

[0022] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0023] Example 1, reference Figure 1 , is an embodiment of the present invention, and provides a method for real-time synchronization of received information and abnormal warning, comprising:

[0024] S1: Collect logistics trajectory data and build a spatiotemporal graph model, and extract trajectory features through the spatiotemporal graph convolutional network.

[0025] Furthermore, constructing a spatiotemporal graph model includes constructing an express delivery network using a graph data structure, where nodes represent logistics locations, edges represent the delivery routes of packages, and edge weights are dynamically updated based on the delivery time and package status.

[0026] It should be noted that trajectory feature extraction includes extracting the temporal transportation mode and spatial path information of the package; setting the historical trajectory window, embedding the time series information into the graph structure, and capturing abnormal patterns of different express delivery paths.

[0027] It should also be noted that a preferred solution for constructing a spatiotemporal graph model specifically includes extracting GPS coordinates, timestamps, package status, distribution station numbers and other information from the express delivery system; filtering missing data. If the GPS coordinates at a certain moment are missing, linear interpolation of the previous and next data is used to supplement them; defining logistics nodes: each warehouse, distribution station, and recipient address is regarded as a node; calculating the two logistics nodes and The straight-line distance between , in meters; the exponential decay method is used to calculate the spatial proximity weight: if If the distance is less than 500 meters, a higher weight (close to 1) is set; if If the distance is greater than 5000 meters, a lower weight (close to 0) is set; between 500 meters and 5000 meters, the weight value is gradually reduced in an exponential manner; the time relationship weight is calculated to extract the time interval between adjacent nodes If the logistics flow time between two nodes is lower than the historical average, a higher time weight is assigned; if the time interval is more than twice the historical average, a lower time weight is assigned; construct a space-time graph using an adjacency matrix with time weights , where each element It is calculated by weighting the spatial relationship weight and the temporal relationship weight. = 0.7 × spatial weight + 0.3 × temporal weight; Set a dynamic update mechanism to recalculate the adjacency matrix every 10 minutes based on new express delivery data to adapt to the latest changes in the logistics path.

[0028] It should also be noted that a preferred solution for extracting trajectory features through a spatio-temporal graph convolutional network specifically includes setting the number of layers and optimization parameters of the spatio-temporal graph convolutional network (ST-GCN). The input data includes logistics nodes, and each node has the following features: location coordinates (latitude and longitude); package status (in transit, arrived at the distribution station, signed for); transportation time information (historical average, current transportation time). The data dimension is in the form of (batch size, time step, number of nodes, feature dimension). The number of layers and functions of ST-GCN: The first layer (local trajectory feature extraction layer): Calculate the path change of the current express delivery within 5 minutes; Identify short-range delivery patterns within the same city and extract trajectory features from the warehouse to the terminal distribution station; The second layer (global trajectory pattern learning layer): Calculate the express delivery transfer situation within 24 hours; Handle cross-regional deliveries, extract cross-province / cross-city transportation patterns, and calculate typical patterns of cargo transfer in combination with historical data; The third layer (anomaly detection layer): Compare the deviation between the actual trajectory and the ST-GCN predicted trajectory; If the package path deviation exceeds 2 standard deviations, mark this trajectory as abnormal; The optimization parameter setting includes adopting a three-layer GCN structure and normalizing the input data to improve calculation stability; When training, use the Adam optimizer, set the initial learning rate to 0.001, and decay by 10% every 10 rounds; During the training process, use L2 regularization to avoid overfitting, so that the model can better generalize to different express delivery path patterns.

[0029] It should also be noted that by constructing a spatio-temporal graph model, the dynamic characteristics of the logistics trajectory can be accurately described; The spatial and temporal weights between logistics nodes are updated in real time using a dynamic adjacency matrix, enabling the model to adapt to changes in the logistics network; Compared with traditional methods that only rely on GPS trajectory points, the present invention can capture the actual situation of the express delivery transportation process more comprehensively, thereby effectively reducing the false alarm rate caused by factors such as path changes and couriers taking detours, improving the credibility of the trajectory data, and providing a more accurate data basis for subsequent anomaly detection. Through a three-layer ST-GCN structure, the present invention extracts trajectory features at three levels: short time, long time, and global mode, and can identify abnormal patterns at different time scales; Compared with traditional statistical analysis or rule-based methods, ST-GCN can effectively model the temporal and spatial relationships of logistics trajectories and improve the robustness of anomaly detection; For example, it can accurately identify problems such as couriers taking detours, staying, and abnormally deviating from the expected trajectory, and reduce false alarms in cases where the path similarity is high but there is an actual anomaly, improving the anomaly detection ability.

[0030] S2: Use the Transformer model for trajectory time series prediction and judge abnormal events based on the anomaly detection threshold.

[0031] Furthermore, the trajectory time series prediction includes inputting the logistics trajectory features extracted by the spatio-temporal graph convolutional network into the Transformer model, and using the multi-head self-attention mechanism to analyze the historical trajectory patterns of the packages; through the time series prediction ability of the Transformer, calculate the normal trajectory distribution of the express packages at future time points.

[0032] It should be noted that judging abnormal events includes setting the anomaly detection threshold and using a calculation method based on Euclidean distance or Mahalanobis distance to measure the deviation between the actual trajectory and the predicted trajectory; when it is detected that the trajectory deviation exceeds the set anomaly detection threshold, trigger an abnormal event and send a warning signal to the logistics management system; send warning information to the delivery personnel and the management platform when an anomaly occurs; trigger different coping strategies according to the type of anomaly; conduct secondary detection on the express packages marked as abnormal to confirm the authenticity of the anomaly.

[0033] It should also be noted that the express trajectory data processed by ST-GCN contains the trajectory feature vectors at each time step. Since the Transformer requires sequential input, the format of the ST-GCN output data is converted to ensure that the data is in the form of (batch size, number of time steps, trajectory feature dimension). Since the Transformer itself does not have time information, a timestamp encoding is added to each express trajectory data point. Using the Positional Encoding method, the timestamp is converted into a set of time features, enabling the Transformer to distinguish the trajectory data at different time points. Use the multi-head self-attention mechanism of the Transformer to calculate the trajectory pattern: calculate the importance weights of each time step in the historical trajectory, and focus on those historical trajectory points that lead to abnormal situations. Calculate the similarity between the current express trajectory and the historical trajectory, and score the trajectory deviation. The express trajectory features are calculated by ST-GCN, and the time pattern analysis results are calculated by the Transformer. The results of the two are weighted and fused: if both consider a certain section of the trajectory to deviate from the normal path, the anomaly score is increased. If only ST-GCN finds an anomaly while the Transformer considers the trajectory pattern normal, the anomaly score is reduced. Once the comprehensive score exceeds the preset anomaly detection threshold, the system sends an anomaly alert to the logistics management platform: generate an abnormal trajectory report, mark the possible reasons for the anomaly (such as the courier taking a detour or being delayed). Trigger a warning message to notify the delivery personnel to check the current status of the express. If the abnormal situation persists, automatically assign a new courier for remediation.

[0034] It should also be noted that using the Transformer model for trajectory time series prediction can effectively address the deficiencies of traditional LSTM in dealing with long-term dependency relationships. The multi-head self-attention mechanism of Transformer can identify key points in the logistics historical trajectory and predict future trajectory deviations, enabling intervention before anomalies occur. For example, when an express delivery fails to reach a certain distribution node as expected, early warnings can be issued and the delivery plan can be adjusted, rather than waiting passively after the anomaly occurs. This mechanism helps reduce logistics delivery delays and improve logistics timeliness and operational efficiency.

[0035] S3: Combine zero-knowledge proof and blockchain technology to encrypt and verify logistics data and store evidence, and construct logistics scheduling optimization and dynamic anomaly recovery strategies.

[0036] Furthermore, encrypting and verifying logistics data and storing evidence includes using zk-SNARKs to encrypt and prove key logistics data; generating zero-knowledge proofs before storing logistics data, and verifying the authenticity of data without accessing the original logistics information; using blockchain evidence storage technology to store the hash values of logistics events in a distributed ledger; using the Merkle tree structure to organize logistics data and verifying the authenticity of records through the root hash; when detecting logistics anomaly events, based on the historical data stored in the blockchain, trace back the abnormal trajectory and verify whether the data has been tampered with during storage; using smart contracts to automatically execute the abnormal traceback logic, and when data inconsistencies are found, send warnings to relevant responsible parties and record violations.

[0037] It should also be noted that a preferred implementation of zero-knowledge proof (ZKP) includes: using zk-SNARKs (Zero-Knowledge Succinct Non-Interactive Argument of Knowledge) to generate proofs for key logistics data. The generating party (logistics company) performs a hash operation on the timestamp, transportation route, and status information of the express delivery and converts it into a mathematical proof. The verifying party (supervision system) can verify whether the proof is correct without accessing the original data to ensure that the data has not been tampered with. Using the Groth16 proof system to calculate the proof off-chain and reduce the on-chain computing burden, the specific steps include: Parameter setting: In the initial stage, a trusted third party generates a set of security parameters and stores them in the database of the logistics system; Proof generation: When a logistics event occurs (such as an express delivery arriving at the distribution station), the hash value of the event is input into the zk-SNARKs proof circuit to calculate a verifiable proof; Proof verification: The regulatory party or a third party can use the pre-set verification key to quickly verify the generated proof without accessing the original information of the express delivery; Storage and update: The result of the zero-knowledge proof is uploaded to the blockchain and compared with the historical records to ensure the integrity of the data chain. The optimization scheme for the computational complexity of zero-knowledge proof (ZKP) includes: Adopting a hierarchical proof structure to reduce the computational burden: In the first layer, local pre-computation is performed. When the courier scans or signs for a package, some proofs are pre-computed in advance to reduce the computing pressure on the server; In the second layer, batch proof calculation is performed. The logistics server calculates the ZKP proofs for multiple express deliveries in batches every 5 minutes instead of calculating the proof for each express delivery separately to reduce the computing time; In the third layer, quick verification is performed. The verifying party can compare multiple proofs through the Merkle tree root hash to verify multiple logistics records at one time. Optimization measures to reduce the proof generation time: Using an optimized version of the Groth16 proof system, using the Curve25519 elliptic curve to accelerate the calculation and reduce the multiplication operation overhead; Using parallel computing, using multi-threading to generate proofs in parallel, splitting the calculation into multiple independent tasks to improve the ZKP computing efficiency; Using an offline computing mode, performing partial proof calculation on logistics events in advance and directly calling when a data verification request occurs to avoid server overload during peak hours. It should also be noted that zk-SNARKs is an abbreviation for "Zero-Knowledge Succinct Non-Interactive Argument of Knowledge", which can be translated into Chinese as "Zero-Knowledge Succinct Non-Interactive Proof of Knowledge". It is a cryptographic technology that allows one party (the prover) to prove to another party (the verifier) that a certain statement is true without revealing any other information.

[0038] It should also be noted that for the evidence storage structure based on the Merkle tree, the hash value of each express delivery data is used as a leaf node to construct the Merkle tree; the hash value of each layer is calculated and stored on the blockchain, enabling the verifier to verify data integrity by only storing the root hash. If the express delivery data is tampered with, its hash value will change, resulting in a mismatch of the Merkle tree root hash, thereby detecting anomalies. Smart contracts are used for data evidence storage and verification. The smart contract defines the storage rules for express delivery status. Whenever the logistics status is updated, the system automatically generates a transaction and records it on the blockchain. The smart contract includes a data consistency check mechanism. When new express delivery information is detected, it is compared with the historical evidence storage data. If data inconsistency is found, an alarm is triggered. Logistics companies can call the verification interface of the smart contract to query the historical records of specific express deliveries and verify whether they are consistent with the data on the blockchain. Distributed storage is adopted. The blockchain only stores the hash values of logistics data, while the actual express delivery information is stored in a distributed storage system (such as IPFS) to ensure data integrity and privacy.

[0039] It should be noted that constructing the logistics scheduling optimization and dynamic anomaly recovery strategy includes using reinforcement learning to optimize the reallocation path of abnormal packages and minimizing delays when couriers re-plan routes; training the RL model by combining historical anomaly data to enable the platform to autonomously learn the optimal scheduling strategy; on the basis of blockchain evidence storage, constructing a verifiable anomaly recovery process and handling anomalies based on logistics specifications.

[0040] It should also be noted that by combining zero-knowledge proofs (ZKP) and blockchain evidence storage, the authenticity and immutability of logistics data are ensured; zk-SNARKs are used for data verification, enabling third parties to verify the authenticity of logistics status without accessing the original data, thus ensuring data credibility while protecting privacy. In addition, the Merkle tree evidence storage structure based on the blockchain provides an efficient way to verify data integrity, ensuring that logistics companies, users, and regulatory agencies can all safely conduct data interactions in an environment of distrust, fundamentally preventing data fraud or malicious tampering. The present invention uses reinforcement learning (RL) to optimize express delivery scheduling and combines blockchain evidence storage to improve the transparency of anomaly handling. After an anomaly occurs, the express delivery allocation plan can be automatically adjusted based on historical anomaly data to implement an intelligent and dynamic scheduling strategy, avoiding relying on manual intervention. Blockchain evidence storage ensures the traceability of the anomaly handling process, prevents human tampering of anomaly recovery records, and improves the fairness of express delivery anomaly handling. By automatically optimizing the delivery path and dynamically adjusting the task allocation of couriers, our invention improves the efficiency of anomaly handling, reduces customer complaints caused by delivery anomalies, and enhances the user experience.

[0041] Example 2, an embodiment of the present invention, provides a method for real-time synchronization of recipient information and anomaly warning. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0042] First, the experiment selects the express delivery data of a certain express company in the past 30 days, covering distribution points in multiple cities. The experimental data includes information such as GPS trajectories, express status, logistics timestamps, express delivery duration, anomaly detection results, etc.; the data sources include 10,000 actual delivery data, of which 8,000 are normal data and 2,000 are abnormal data, including path deviation, delay, package retention, etc., and the integrity of the test data is ensured by simulating actions such as couriers taking detours, express packages being lost, and system data tampering; the experiment uses two schemes for comparison. Scheme 1 adopts an anomaly detection and evidence storage scheme of ST-GCN + Transformer + ZKP + blockchain, and Scheme 2 adopts a traditional scheme based on rule thresholds and LSTM trajectory prediction and stores logistics data through a centralized database; the experiment first parses the GPS trajectory data and constructs a dynamic adjacency matrix to adapt to the changes in the express delivery network, and then extracts logistics trajectory features through ST-GCN and sets a historical trajectory window to capture short-term and long-term trajectory patterns; in the anomaly detection stage, Transformer is used for time series prediction, calculates the future trajectory and compares the current express trajectory deviation. If the anomaly score exceeds 0.75, it is determined as an anomaly; in the data evidence storage stage, zk-SNARKs is used to calculate the zero-knowledge proof of logistics data, and the blockchain is used to record the hash value to ensure data authenticity; referring to Table 1, the experiment compares the anomaly detection accuracy, false alarm rate, logistics data tampering detection ability, calculation overhead, storage consumption, and warning response time of the two schemes, so as to evaluate the advantages of the present invention in logistics management.

[0043] Table 1 Logistics experiment result table

[0044]

[0045] The comparison of experimental data shows that the logistics anomaly detection and data archiving method of the present invention is significantly superior to the traditional solution in multiple key performance indicators; the anomaly detection accuracy rate of the present invention reaches 96.8%, an increase of 14.3% compared to 82.5% of the traditional method. This is mainly because ST-GCN combined with Transformer can accurately identify various anomaly types such as route detours and delivery anomalies, while the traditional rule-based threshold method is limited by the set fixed patterns and is difficult to adapt to complex logistics environments; in addition, the false alarm rate of the present invention is only 2.3%, far lower than 12.5% of the traditional solution. This benefits from the ability of Transformer to predict trajectory patterns, which can effectively reduce false alarms caused by route changes and traffic fluctuations; in terms of data security, the present invention combines ZKP and blockchain archiving to achieve a 100% logistics data tampering detection rate, while the traditional centralized storage solution cannot detect data tampering and has data security risks; in terms of computing performance, the computing overhead of the present invention is 68 ms / time, a reduction of 23.6% compared to 89 ms / time of the traditional method. Due to the use of ST-GCN trajectory feature extraction and Transformer parallel computing, the detection efficiency is greatly improved; in addition, the storage consumption of the present invention is 0.12 MB / time, a reduction of 62.5% compared to 0.32 MB / time of the traditional method. This mainly benefits from the fact that the blockchain archiving uses the Merkle tree structure and only needs to store hash values instead of complete logistics data; finally, in terms of early warning response speed, the early warning response time of the present invention is 1.2 seconds, while that of the traditional solution is 5.6 seconds, an increase of 366%. Due to the time series modeling and pattern recognition capabilities of Transformer, the system can analyze possible anomaly patterns based on historical trajectory data, improve the accuracy of anomaly detection, and provide early warnings before the anomaly trend forms; overall, the present invention is superior to the prior art in terms of detection accuracy, false alarm rate, data security, computing overhead, storage optimization, etc., especially outstanding in terms of data security, computing efficiency, and anomaly detection accuracy, providing an efficient, accurate, and verifiable logistics information management solution for modern logistics systems.

[0046] Example 3, referring to Figure 2 , which is an embodiment of the present invention, provides a real-time synchronization and anomaly early warning system for recipient information, including a data processing module, an anomaly early warning module, and an optimization and recovery module.

[0047] Among them, the data processing module is used to collect logistics trajectory data and construct a spatio-temporal graph model, and extract trajectory features through a spatio-temporal graph convolutional network; the anomaly early warning module is used to perform trajectory time series prediction using a Transformer model and judge anomaly events based on an anomaly detection threshold; the optimization and recovery module is used to combine zero-knowledge proof and blockchain technology to encrypt and verify logistics data and archive it, and construct a logistics scheduling optimization and dynamic anomaly recovery strategy.

[0048] If a function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs, etc., various media that can store program codes.

[0049] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a defined sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in combination with an instruction execution system, apparatus, or device.

[0050] More specific examples (nonexhaustive list) of computer-readable media include the following: electrical connection parts with one or more wirings (electronic devices), portable computer disk cartridges (magnetic devices), random access memories (RAMs), read-only memories (ROMs), erasable programmable read-only memories (EPROMs or flash memories), optical fiber devices, and portable compact disc read-only memories (CDROMs). Additionally, a computer-readable medium can even be paper or other suitable media on which a program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, then editing, interpreting, or otherwise processing it as necessary, and then storing it in a computer memory.

[0051] It should be understood that each part of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following technologies well known in the art can be used: discrete logic circuits with logic gate circuits for implementing logic functions on data signals, application specific integrated circuits with appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc. It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

[0052] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for real-time synchronization of received information and abnormal warning, characterized in that: include: Collect logistics trajectory data and build a spatiotemporal graph model, and extract trajectory features through the spatiotemporal graph convolutional network; Logistics trajectory data includes GPS coordinates, timestamps, package status, and delivery station numbers extracted from express delivery systems; Building a spatiotemporal graph model involves building an express delivery network using a graph data structure, where nodes represent logistics locations, edges represent the delivery paths of packages, and edge weights are dynamically updated based on the delivery time and package status; Trajectory feature extraction includes extracting the temporal transportation mode and spatial path information of the package; Set the historical trajectory window, embed the time series information into the graph structure, and capture the abnormal patterns of different express routes; The Transformer model is used to predict trajectory time series and abnormal events are judged based on the anomaly detection threshold; Combining zero-knowledge proof and blockchain technology, the logistics data is encrypted, verified and stored, and logistics scheduling optimization and dynamic abnormal recovery strategies are built; Trajectory feature extraction includes the use of a three-layer spatiotemporal graph convolutional network. The first layer extracts local spatiotemporal features and processes short-term logistics path information; The second layer aggregates global information and learns long-term cross-regional logistics characteristics; The third layer performs abnormal pattern recognition, detects abnormal trajectories and provides interpretable anomaly scores; The first local trajectory feature extraction layer includes calculating the path changes of the current express within 5 minutes; identifying the short-distance delivery mode in the same city, and extracting the trajectory features from the warehouse to the terminal delivery station; The second global trajectory pattern learning layer includes calculating the express delivery flow within 24 hours; processing cross-regional distribution, extracting cross-provincial / cross-city transportation patterns, and calculating the typical pattern of cargo flow based on historical data.

2. The method for real-time synchronization of received information and abnormal warning according to claim 1, characterized in that: The trajectory time series prediction includes inputting the logistics trajectory features extracted by the spatiotemporal graph convolutional network into the Transformer model, and using the multi-head self-attention mechanism to analyze the historical trajectory pattern of the package; Through the time series prediction capability of Transformer, the normal trajectory distribution of express packages at future time points is calculated.

3. The method for real-time synchronization of received information and abnormal warning according to claim 2, characterized in that: The determining of abnormal events includes setting an abnormality detection threshold and using a calculation method based on Euclidean distance or Mahalanobis distance to measure the deviation between the actual trajectory and the predicted trajectory; When the detected trajectory deviation exceeds the set abnormal detection threshold, an abnormal event is triggered and an early warning signal is sent; Send early warning information to delivery personnel and management platform when an abnormality occurs; Trigger different response strategies based on the type of exception; Carry out secondary inspection on express parcels that have been marked as abnormal to confirm the authenticity of the abnormality.

4. The method for real-time synchronization of received information and abnormal warning according to claim 3, characterized in that: The encryption verification and storage of logistics data includes using zk-SNARKs to encrypt and prove key logistics data; Before the logistics data is stored, a zero-knowledge proof is generated to verify the authenticity of the data without accessing the original logistics information; Using blockchain evidence storage technology, the hash value of logistics events is stored in a distributed ledger; The Merkle tree structure is used to organize logistics data, and the authenticity of the records is verified through the root hash; When an abnormal logistics event is detected, the abnormal track is traced back based on the historical data stored in the blockchain, and it is verified whether the data has been tampered with during the storage process; Smart contracts are used to automatically execute abnormal backtracking logic. When data inconsistencies are found, warnings are sent to the relevant responsible parties and violations are recorded.

5. The method for real-time synchronization of received information and abnormal warning according to claim 4, characterized in that: The said construction of logistics scheduling optimization and dynamic abnormal recovery strategy includes using reinforcement learning to optimize the redistribution path of abnormal packages and minimize delays when couriers re-plan routes; Combine historical abnormal data to train the RL model, so that the platform can autonomously learn the optimal scheduling strategy; Based on blockchain evidence storage, a verifiable exception recovery process is built to handle exceptions based on logistics specifications.

6. A system using the method for real-time synchronization of received information and abnormal warning as claimed in any one of claims 1 to 5, characterized in that: Including data processing module, abnormal warning module, optimization and recovery module; The data processing module is used to collect logistics trajectory data and construct a spatiotemporal graph model, and extract trajectory features through a spatiotemporal graph convolutional network; The abnormal warning module is used to use the Transformer model to predict trajectory time series and judge abnormal events based on the abnormal detection threshold; The optimization and recovery module is used to combine zero-knowledge proof and blockchain technology to encrypt, verify and store logistics data, and build logistics scheduling optimization and dynamic abnormality recovery strategies.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method for real-time synchronization of received information and abnormal warning according to any one of claims 1 to 6 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for real-time synchronization of received information and abnormal warning according to any one of claims 1 to 6 are implemented.

Images