Product logistics management scheduling method based on block chain, medium and equipment

Abstract

The invention discloses a product logistics management scheduling method based on a block chain, a medium and equipment, and the method comprises the steps: generating an initial transportation path scheme through collecting logistics transportation demand information and supply chain node data; constructing a decentralized logistics resource pool, and collecting and verifying transportation fluctuation parameters to generate a credible state matrix; inputting the initial transportation path scheme and the logistics resource credible state matrix into a graph neural network scheduling model, and outputting a dynamic scheduling scheme; compliance verification of task allocation and scheduling instructions is automatically executed through the smart contract; and finally, updating the scheduling execution result and the logistics resource credible state matrix to a distributed account book, and outputting a logistics management report. Through deep cooperation of the intelligent contract and dynamic path optimization, real-time credible scheduling of the transportation process is realized, the problems of insufficient dynamic response and multi-party cooperation trust deficiency of an existing logistics system are solved, and the reliability and efficiency of cold chain and other sensitive commodity transportation are improved.

Description

Technical Field

[0001] The present invention relates to the field of blockchain intelligent logistics technology, and in particular to a blockchain-based product logistics management and scheduling method, medium, and equipment. Background Art

[0002] In modern logistics management, especially in scenarios involving the transportation of temperature-sensitive goods, achieving dynamic optimization of the transportation process and trusted collaboration among multiple parties has always been a focus of industry attention. Current logistics systems typically employ a centralized scheduling model, relying on preset rules or historical data for route planning, making it difficult to adapt to real-time changes in the transportation process. While the application of IoT technology enables the collection of basic data such as vehicle positioning and environmental monitoring, the data silos created by each link make it difficult to effectively integrate and utilize this information.

[0003] In recent years, blockchain technology has been introduced into the logistics sector to enhance data credibility. Some solutions attempt to automate the execution of transportation agreements through smart contracts, while other studies employ machine learning algorithms to optimize route planning. However, these approaches still face limitations in terms of dynamic response and system coordination. On the one hand, static data recording mechanisms struggle to support real-time decision-making; on the other hand, algorithm optimization and blockchain verification are often disconnected, failing to form a closed-loop management system. In high-value scenarios such as cross-border logistics and pharmaceutical cold chains, the conflict between the dynamic nature of the transportation environment and the static nature of scheduling strategies leads to delayed responses to exceptions. Furthermore, the lack of a reliable automated evaluation mechanism for multi-party collaboration still requires manual intervention and coordination. These issues limit the potential of logistics systems to improve efficiency and control costs, and also affect the overall reliability of complex supply chain networks. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to propose a blockchain-based product logistics management and scheduling method, medium and equipment, which realizes real-time and reliable adjustment of transportation routes through the deep coupling of smart contracts and dynamic routing optimization, and solves the problems of insufficient dynamic response capabilities of existing logistics systems and lack of trust in multi-party collaboration.

[0005] To achieve the above technical objectives, in a first aspect, the present application provides a blockchain-based product logistics management and scheduling method, comprising: Collect product logistics and transportation demand information and supply chain node data. Logistics and transportation demand information includes product type, transportation priority, temperature control requirements, and target distribution area. Supply chain node data includes storage location, transportation vehicle status, and carrier qualifications. Dynamically decompose logistics and transportation demand information, use a multi-objective optimization algorithm to generate an initial transportation route plan, and calculate the path feasibility score of the initial transportation route plan based on historical transportation data; A decentralized logistics resource pool is built based on supply chain node data. Transport fluctuation parameters, including vehicle GPS trajectory, temperature and humidity sensor data, and electronic lock status, are collected in real time through edge computing nodes. A spatiotemporal consistency verification algorithm is used to filter abnormal data. The verified transport fluctuation parameters are then stored on-chain to generate a trusted status matrix for logistics resources. The initial transport path plan and the logistics resource trust state matrix are input into the graph neural network scheduling model. The spatiotemporal dependencies between transport nodes are modeled through the graph attention mechanism, and the path weights are dynamically adjusted to output a dynamic scheduling plan that includes real-time path optimization suggestions, emergency scheduling instructions, and resource allocation strategies. The smart contract is triggered based on the dynamic scheduling plan. The smart contract automatically allocates transportation tasks according to preset rules, freezes the transaction rights of abnormal batches, and calculates the carrier's incentive token. At the same time, the compliance of the scheduling instructions is verified through zero-knowledge proof to obtain the scheduling execution results. Based on the blockchain consensus mechanism, the scheduling execution results and the logistics resource trust status matrix are updated to the distributed ledger, and a logistics management report including the completion rate of transportation tasks, temperature control compliance rate and violation event statistics is generated and output.

[0006] In some embodiments, dynamic task decomposition is performed on logistics transportation demand information, an initial transportation route plan is generated using a multi-objective optimization algorithm, and a path feasibility score of the initial transportation route plan is calculated in combination with historical transportation data, including: Based on the product type, transportation priority, and temperature control requirements in the logistics transportation demand information, a transportation constraint weight matrix is ​​constructed through the analytic hierarchy process. The transportation constraint weight matrix is ​​used to quantify the differentiated requirements of different product categories for timeliness, temperature stability, and safety. The transportation constraint weight matrix is ​​aligned with the historical transportation data in time and space, and an improved non-dominated sorting genetic algorithm is used to perform multi-objective optimization and generate a set of candidate routing solutions containing the Pareto optimal solution set. Perform a dynamic risk assessment on each candidate route in the candidate route set. Long-short-term memory networks are used to predict the potential traffic congestion probability and temperature control deviation risk of each candidate route in the target delivery area. Combined with real-time vehicle status data from the logistics resource trust state matrix, a comprehensive feasibility score is calculated for each candidate route. Based on the comprehensive feasibility score, the candidate path plans in the candidate path plan set are sorted and screened, and the candidate path plans with a comprehensive feasibility score lower than the preset feasibility score threshold are eliminated. The remaining candidate path plans are matched with the carrier qualifications in the supply chain node data for matching analysis to generate the initial transportation path plan and the corresponding path feasibility score.

[0007] In some embodiments, building a decentralized logistics resource pool based on supply chain node data includes: The edge computing nodes in the blockchain network collect dynamic resource data from each supply chain node in real time. Dynamic resource data includes the real-time inventory status of storage nodes, the load capacity and available refrigeration unit information of transportation vehicles, and the current task load status of carriers. A fast verification algorithm based on Bloom filter is used to preliminarily screen dynamic resource data, eliminate invalid or duplicate data records, and obtain verified dynamic resource data; The verified dynamic resource data is associated and mapped with the static qualification information in the supply chain node data to construct a multi-dimensional resource feature vector, which includes space availability, equipment compatibility and serviceability; Multi-dimensional resource feature vectors are classified and integrated according to preset rules to generate a decentralized logistics resource pool with a timestamp. Each resource unit in the decentralized logistics resource pool is accompanied by a verifiable credit score and historical service record.

[0008] In some embodiments, real-time collection of transport fluctuation parameters by edge computing nodes includes: Real-time collection of raw sensor data, including vehicle GPS positioning coordinates, multi-zone temperature and humidity sensor readings, and electronic lock switch status signals; Adopting an adaptive sampling strategy to dynamically collect raw sensor data, the sampling frequency is automatically increased when the temperature and humidity change rate exceeds the preset threshold or the GPS signal is lost; The collected data is denoised in real time using a lightweight filtering algorithm, and the edge node digital signature and timestamp are added to generate transport fluctuation parameters with identity authentication.

[0009] In some embodiments, a spatiotemporal consistency verification algorithm is used to filter abnormal data, and the verified transportation fluctuation parameters are stored on the chain to generate a logistics resource trust status matrix, including: Receive the transport fluctuation parameters, parse and obtain the vehicle GPS trajectory sequence, temperature and humidity time series data, and electronic lock status change records, and record them as the first processed data; Construct a spatiotemporal correlation analysis model, input the first processed data into the spatiotemporal correlation analysis model, identify the matching degree between the GPS trajectory and the temperature and humidity changes through a multi-dimensional clustering algorithm, detect whether there is human tampering or equipment failure, and obtain the second processed data; The second processed data is anonymized using a differential privacy protection algorithm to generate a standard data unit that meets the storage requirements of the blockchain; The processed standard data units are written into the blockchain through smart contracts to generate a trusted status matrix of logistics resources and update the real-time status records in the decentralized logistics resource pool.

[0010] In some embodiments, the initial transportation path plan and the logistics resource trust state matrix are input into the graph neural network scheduling model. The spatiotemporal dependencies between transportation nodes are modeled through the graph attention mechanism, and the path weights are dynamically adjusted to output a dynamic scheduling plan that includes real-time path optimization suggestions, emergency scheduling instructions, and resource allocation strategies, including: Construct a transportation network topology graph, converting warehousing nodes, transportation vehicles, and transit stations into graph nodes. Establish node feature vectors corresponding to graph nodes based on the spatiotemporal characteristics and resource availability indicators in the logistics resource trust state matrix, and construct edge features based on the actual transportation route connection relationships. A multi-layer graph attention network is used to learn the features of the transportation network topology. The spatial dependency weights between graph nodes are calculated through a multi-head attention mechanism. Combined with a temporal convolutional network to capture the dynamic change pattern of transportation resources, the optimized path weights of the initial transportation path are obtained. The message passing mechanism of the graph neural network is used to iteratively update the priority scores of each initial transportation path and generate a set of candidate optimized paths; Based on a set of candidate optimized paths, a reinforcement learning strategy is used to dynamically generate real-time path optimization suggestions. When abnormal fluctuations in the logistics resource trust status matrix are detected, emergency dispatch instructions including backup routes and emergency resource deployment are automatically triggered. The optimized path weights are integrated with the initial transport path plan to output a dynamic scheduling plan that includes optimal path recommendations, resource reallocation plans, and risk warning information.

[0011] In some embodiments, the smart contract execution is triggered based on the dynamic scheduling scheme. The smart contract automatically allocates transportation tasks according to preset rules, freezes abnormal batch transaction permissions, and calculates carrier incentive tokens. At the same time, the compliance of the scheduling instructions is verified through zero-knowledge proof to obtain the scheduling execution results, including: Analyze the path optimization suggestions and resource allocation strategies in the dynamic scheduling plan, assign transportation tasks to the optimal carrier node through smart contracts, and generate task allocation results. The allocation process is based on the real-time available resources in the logistics resource trust status matrix and the carrier's credit score to make a multi-dimensional selection; When it is detected that the transportation fluctuation parameters exceed the preset safety threshold, the exception handling smart contract is triggered, automatically freezing the trading rights of the relevant batch of products and generating an unalterable violation record on the blockchain. At the same time, an encrypted alert is sent to the supervision node, and an exception handling record is generated. The number of incentive tokens is dynamically calculated based on the carrier's task completion quality, timeliness, and temperature control compliance rate. The token value anchor coefficient is adjusted through decentralized oracle data to obtain token incentive details. Build a zero-knowledge proof verification module to generate a validity certificate that the dispatch instruction complies with the product sales rules, so that regulators can conduct compliance spot checks and obtain compliance certificates; The task assignment results, exception handling records, token incentive details and compliance certificates are packaged to generate the scheduling execution results.

[0012] In some embodiments, the scheduling execution results and the logistics resource trust status matrix are updated to the distributed ledger according to the blockchain consensus mechanism, and a logistics management report containing transportation task completion, temperature control compliance rate, and violation event statistics is generated and output, including: Based on the scheduling execution results, the smart contract automatically calculates the task completion indicators of each transportation batch. Task completion indicators include timeliness achievement rate, path deviation and resource utilization efficiency; Extract historical data from temperature control sensors in the trusted state matrix of logistics resources, use a sliding window algorithm to calculate the temperature control compliance rate at each transportation stage, and identify the frequency and duration of abnormal temperature fluctuation events; Through the violation record smart contract on the blockchain, the distribution of violation types and spatial distribution characteristics within the preset period are counted to generate analysis results of violation hotspots; The results of task completion indicators, temperature control compliance rates, and violation hotspot analysis are weighted and integrated to generate carrier credit score update recommendations and supply chain optimization plans. The carrier's credit score update suggestions and supply chain optimization plans are integrated with the task completion indicators, temperature control compliance rates, and violation hotspot analysis results into a logistics management report and output. The logistics management report includes a transportation performance trend chart, a temperature control quality heat map, and a violation risk warning prompt.

[0013] In a second aspect, the present invention further provides a computer-readable storage medium storing computer program instructions, which implement the method described in the first aspect when executed by a processor.

[0014] In a third aspect, the present invention further provides an electronic device comprising a memory and a processor, wherein the memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method described in the first aspect.

[0015] By adopting the above-mentioned technical solution, the present invention has the following beneficial effects compared with the prior art: by collecting logistics and transportation demand information and supply chain node data, an initial transportation route plan is generated; a decentralized logistics resource pool is constructed, and transportation fluctuation parameters are collected and verified to generate a trusted state matrix; the initial transportation route plan and the logistics resource trusted state matrix are input into the graph neural network scheduling model to output a dynamic scheduling plan; compliance verification of task allocation and scheduling instructions is automatically executed through smart contracts; and finally, the scheduling execution results and the logistics resource trusted state matrix are updated to the distributed ledger and a logistics management report is output. The present invention realizes real-time trusted scheduling of the transportation process through the deep collaboration of smart contracts and dynamic path optimization, solves the problems of insufficient dynamic response of existing logistics systems and lack of trust in multi-party collaboration, and improves the reliability and efficiency of transportation of sensitive commodities such as cold chain. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0017] Figure 1 is a method step diagram of steps S101 to S106 of the method described in the specific embodiment; Figure 2 is a method step diagram of steps S201 to S204 of the method described in the specific embodiment; Figure 3 Schematic diagram of the structure of the electronic device described in the specific implementation.

[0018] The reference numerals in the above drawings are described as follows: 1. Electronic equipment; 11. Memory; 12. Processor. DETAILED DESCRIPTION

[0019] The present invention will be described in further detail below with reference to the accompanying drawings and examples. It is particularly noted that the following examples are intended only to illustrate the present invention and are not intended to limit the scope of the present invention. Similarly, the following examples are only some embodiments of the present invention and are not intended to be exhaustive. All other embodiments obtained by those of ordinary skill in the art without creative effort are intended to fall within the scope of protection of the present invention.

[0020] See also Figure 1 In a first aspect, this embodiment provides a product logistics management and scheduling method based on blockchain, including: S101. Collect product logistics and transportation demand information and supply chain node data. Logistics and transportation demand information includes product type, transportation priority, temperature control requirements, and target distribution area. Supply chain node data includes storage location, transportation vehicle status, and carrier qualifications. S102. Dynamically decompose the logistics transportation demand information, generate an initial transportation route plan using a multi-objective optimization algorithm, and calculate the path feasibility score of the initial transportation route plan in combination with historical transportation data; S103. Build a decentralized logistics resource pool based on supply chain node data. Use edge computing nodes to collect transport fluctuation parameters in real time. These parameters include vehicle GPS trajectories, temperature and humidity sensor data, and electronic lock status. Use a spatiotemporal consistency verification algorithm to filter abnormal data. Verified transport fluctuation parameters are stored on-chain to generate a trusted state matrix for logistics resources. S104. Input the initial transportation route plan and the logistics resource trust state matrix into the graph neural network scheduling model, model the spatiotemporal dependencies between transportation nodes through the graph attention mechanism, dynamically adjust the path weights, and output a dynamic scheduling plan that includes real-time route optimization suggestions, emergency scheduling instructions, and resource allocation strategies; S105. Trigger the execution of the smart contract based on the dynamic scheduling plan. The smart contract automatically allocates transportation tasks according to preset rules, freezes the transaction permissions of abnormal batches, and calculates the carrier incentive token. At the same time, it verifies the compliance of the scheduling instructions through zero-knowledge proof and obtains the scheduling execution result. S106. Update the scheduling execution results and the logistics resource trust status matrix to the distributed ledger according to the blockchain consensus mechanism, generate and output a logistics management report including the completion rate of transportation tasks, temperature control compliance rate and violation event statistics.

[0021] In step S101, logistics and transportation demand information is used as the original input of the scheduling system. Its product type field is used to distinguish between regular goods and special categories such as pharmaceutical cold chain. The transportation priority can adopt the industry-standard T1-T4 grading standard. Temperature control requires recording the allowable temperature fluctuation range of the product. The target distribution area is defined by geographic information system coordinates. Preferably, supply chain node data is collected through IoT terminals deployed in warehouses and transportation vehicles, wherein the storage location information is associated with the inventory data of the enterprise resource planning system, the transportation vehicle status includes real-time monitoring parameters such as engine operating conditions and fuel remaining, and the carrier qualification information needs to be verified through the blockchain node to verify the validity of its digital certificate. Furthermore, the data collection process uses an encrypted transmission protocol to ensure information security, thereby establishing a reliable data foundation for subsequent intelligent scheduling.

[0022] The dynamic task decomposition technology employed in step S102 breaks down complex logistics orders into subtasks with temporal and spatial constraints. A multi-objective optimization algorithm simultaneously considers transportation costs, timeliness, and carbon emissions during route generation, generating an initial transportation route plan with a detailed sequence of transit nodes and time window arrangements. Preferably, a route feasibility score is calculated by analyzing the frequency and resolution efficiency of abnormal events in historical transport records. This scoring mechanism effectively identifies potentially high-risk route segments and provides a basis for subsequent dynamic adjustments.

[0023] The decentralized logistics resource pool constructed in step S103 realizes the digital twin of physical transportation resources. Preferably, edge computing nodes are deployed at key logistics facilities, which can reduce the impact of network transmission delays through local preprocessing. The vehicle GPS trajectory data collected by it needs to be cross-verified by multi-base station positioning, the temperature and humidity sensor readings meet the requirements of the national metrology verification regulations, and the electronic lock status information contains the encrypted operating user identity. The spatiotemporal consistency verification algorithm identifies anomalies by analyzing the time series characteristics and spatial correlation of device data, such as vehicle signals appearing in different locations at the same time or abnormal temperature change records. The generated logistics resource trust state matrix adopts a multi-dimensional data structure, which includes key dimensions such as resource availability, environmental compliance and operational credibility. Furthermore, the logistics resource trust state matrix realizes distributed storage and real-time synchronization through the blockchain network, providing the system with the latest logistics status snapshot.

[0024] The graph neural network scheduling model in step S104 uses a graph attention mechanism to capture the implicit spatiotemporal correlations within the transportation network. When the logistics resource trust status matrix indicates an anomaly on a particular route, the model adaptively adjusts the weight coefficients of the relevant routes and comprehensively considers the distribution of remaining transport capacity to generate an optimization plan. The three types of instructions output by the dynamic scheduling plan are synergistic: real-time route optimization recommendations focus on geographical route adjustments, emergency dispatch instructions address sudden equipment failures, and resource allocation strategies optimize the matching of vehicles and cargo. This multi-layered response mechanism enables the system to effectively address various uncertainties in the logistics process.

[0025] During the execution of the smart contract in step S105, the transportation task allocation mechanism is automatically triggered through the preset contract logic based on the verified path optimization suggestions and resource allocation strategies in the dynamic scheduling plan. The smart contract first verifies the real-time vehicle status and carrier qualifications recorded in the logistics resource trust status matrix. When it detects that the transportation fluctuation parameters exceed the preset safety threshold, it immediately freezes the on-chain transaction permissions of the relevant batches of products and generates an unalterable violation record. The task allocation process strictly follows the resource priority ranking in the dynamic scheduling plan, and at the same time, the best match is made based on the real-time updated carrier credit score. For each successfully assigned transportation task, the smart contract automatically calculates the corresponding incentive token base, which is bound to the core parameters required by the task, such as timeliness and temperature control standards.

[0026] All dispatch instructions undergo compliance verification via zero-knowledge proof before taking effect, ensuring they adhere to basic product transportation regulations. Data generated during the execution process, including task assignment records, exception handling results, and token calculation details, collectively constitute the dispatch execution results and generate corresponding blockchain transaction vouchers. This step, through the automated execution of smart contracts, transforms dynamic dispatch plans into binding on-chain operations, ensuring the enforceability of dispatch decisions while providing structured input data for subsequent blockchain consensus verification.

[0027] Step S106 uses a blockchain consensus mechanism to ensure that all participants reach agreement on the dispatch results. The dispatch execution results recorded in the distributed ledger contain a complete chain of evidence for the operation. The generated logistics management report uses structured data to display key indicators of the transportation process. The completion rate of transportation tasks reflects the degree of compliance with the plan, the temperature control compliance rate reflects the stability of environmental control, and the violation statistics reveal anomalies in system operation. This data provides an objective basis for continuous improvement for logistics companies.

[0028] This embodiment uses the logistics resource trust state matrix to map the transportation resource status in real time, and combines the graph neural network scheduling model to dynamically model the spatiotemporal correlation characteristics in the transportation network, forming a closed-loop feedback mechanism from data collection to solution optimization. The multi-objective optimization algorithm comprehensively considers transportation costs, timeliness and environmental indicators in the path planning stage. The edge computing node ensures the real-time nature of the data through localized processing, and the spatiotemporal consistency verification algorithm effectively eliminates abnormal data interference. The smart contract automatically executes task allocation and exception handling based on preset business rules. The zero-knowledge proof mechanism ensures the compliance verification of scheduling instructions, and the blockchain consensus ensures the credible evidence of the execution results. This method realizes the organic unity of resource status perception, dynamic decision optimization and automated execution in the logistics scheduling process, ensuring temperature control compliance and operation traceability while improving transportation efficiency, and provides reliable technical support for intelligent scheduling in complex supply chain environments.

[0029] See also Figure 2 In some embodiments, dynamic task decomposition is performed on logistics transportation demand information, an initial transportation route plan is generated using a multi-objective optimization algorithm, and a path feasibility score of the initial transportation route plan is calculated in combination with historical transportation data, including: S201. Based on the product type, transportation priority, and temperature control requirements in the logistics transportation demand information, construct a transportation constraint weight matrix using the analytic hierarchy process. The transportation constraint weight matrix is ​​used to quantify the differentiated requirements of different product categories for timeliness, temperature stability, and safety. S202, aligning the transportation constraint weight matrix with the historical transportation data in time and space, and performing multi-objective optimization using an improved non-dominated sorting genetic algorithm to generate a set of candidate route solutions including a Pareto optimal solution set; S203. Dynamically assess the risk of each candidate route in the candidate route set. Long-short-term memory (LSTM) networks are used to predict the potential traffic congestion probability and temperature control deviation risk of each candidate route in the target delivery area. Combined with the real-time vehicle status data in the logistics resource trustworthy status matrix, a comprehensive feasibility score for each candidate route is calculated. S204. Sort and screen the candidate path plans in the candidate path plan set based on the comprehensive feasibility score, eliminate the candidate path plans with a comprehensive feasibility score lower than the preset feasibility score threshold, and perform a matching analysis on the remaining candidate path plans with the carrier qualifications in the supply chain node data to generate an initial transportation path plan and a corresponding path feasibility score.

[0030] In step S201, the transportation constraint weight matrix converts the product transportation demand into quantifiable decision parameters through the hierarchical analysis method, where the product type field is mapped to the temperature stability coefficient, the transportation priority corresponds to the timeliness weight, and the temperature control requirement is converted into a safety tolerance value within the allowable fluctuation range.

[0031] The spatiotemporal alignment operation in step S202 maps spatiotemporal features from historical transportation data, such as route duration and temperature control records, to current transportation constraints. The improved non-dominated sorting genetic algorithm, based on the traditional NSGA-II, introduces a transportation resource occupancy penalty factor, resulting in a Pareto-optimal solution set that simultaneously meets the triple objectives of cost minimization, time efficiency optimization, and carbon emission control. Each solution in the candidate routing set includes a complete node sequence, time window arrangement, and detailed expected resource consumption.

[0032] In the dynamic risk assessment model used in step S203, a long-short-term memory network analyzes historical traffic flow data and meteorological records to establish a mapping relationship between spatiotemporal characteristics and congestion probability. Preferably, the network structure incorporates an attention mechanism layer to capture the influence of special periods such as holidays. The temperature control deviation risk prediction module integrates product characteristic curves and route environmental parameters to establish a risk prediction model by analyzing temperature fluctuation patterns in historical abnormal events. The comprehensive feasibility score calculation process can incorporate a real-time vehicle status correction factor. When the logistics resource trust status matrix indicates a warning signal for a vehicle's refrigeration system, the temperature control risk coefficient of the corresponding routing plan will be dynamically adjusted upward.

[0033] In step S204, the preset feasibility score threshold is dynamically adjusted using a sliding window mechanism. The window size is determined by the moving average of historical scoring errors, ensuring that the screening criteria are consistent with the current state of the logistics network. The final initial transportation route plan generation process considers the matching of carrier qualifications. Their credit score data is derived from the immutable historical performance records on the blockchain, ensuring a precise match between transportation capacity and task requirements.

[0034] This embodiment combines static path planning with real-time risk prediction by constructing a dynamically evolving evaluation system, adopts the analytic hierarchy process to transform subjective business rules into a computable transportation constraint weight matrix, introduces a long-short-term memory network time series model to enhance the predictability of candidate path solution evaluation, and establishes a preset feasibility score threshold adaptive adjustment mechanism to improve system robustness, so that the path feasibility score can dynamically respond to changes in the logistics network state, while ensuring the feasibility of the solution and taking into account multi-objective optimization requirements, providing high-quality initial input for the subsequent graph neural network scheduling model. The decision-making evidence chain formed during the entire evaluation process will be automatically recorded on the chain through smart contracts, providing data support for the transportation strategy optimization of logistics companies.

[0035] In some embodiments, building a decentralized logistics resource pool based on supply chain node data includes: The edge computing nodes in the blockchain network collect dynamic resource data from each supply chain node in real time. Dynamic resource data includes the real-time inventory status of storage nodes, the load capacity and available refrigeration unit information of transportation vehicles, and the current task load status of carriers. A fast verification algorithm based on Bloom filter is used to preliminarily screen dynamic resource data, eliminate invalid or duplicate data records, and obtain verified dynamic resource data; The verified dynamic resource data is associated and mapped with the static qualification information in the supply chain node data to construct a multi-dimensional resource feature vector, which includes space availability, equipment compatibility and serviceability; Multi-dimensional resource feature vectors are classified and integrated according to preset rules to generate a decentralized logistics resource pool with a timestamp. Each resource unit in the decentralized logistics resource pool is accompanied by a verifiable credit score and historical service record.

[0036] In this embodiment, a decentralized logistics resource pool built based on supply chain node data implements dynamic aggregation management of logistics resources through edge computing nodes within the blockchain network. Edge computing nodes collect dynamic resource data from each supply chain node in real time. This data is automatically accessed through IoT devices and enterprise information system interfaces, ensuring the real-time and integrity of the data source.

[0037] The initial screening of dynamic resource data uses a fast verification algorithm based on a Bloom filter. Through a preset set of hash functions, the key fields of the data are mapped to a bit array structure to achieve efficient duplicate data detection. Preferably, the storage space configuration of the Bloom filter needs to balance the error rate and computational overhead. In engineering practice, the size of the bit array is usually dynamically adjusted according to the resource update frequency. Dynamic resource data has significant spatiotemporal locality characteristics, which is manifested in the fact that most resource status updates are concentrated within the jurisdiction of adjacent edge nodes. Therefore, a two-layer mechanism combining local verification and global verification is adopted to ensure data accuracy while optimizing network transmission efficiency.

[0038] Verified dynamic resource data is mapped and correlated with static qualification information from supply chain node data to construct a multidimensional resource feature vector encompassing three dimensions: spatial availability, equipment compatibility, and serviceability. Spatial availability is calculated based on the topological relationship between geographic coordinates and the mission area. Equipment compatibility is determined by the matching of transport vehicle and cargo characteristics. Serviceability is derived from a combination of credit scores and real-time load status assessments. These feature vectors are categorized and integrated using a timestamp-based sliding window mechanism to ensure the timeliness of information in the resource pool. In the resulting decentralized logistics resource pool, each resource unit is accompanied by a credit score automatically calculated by the smart contract and a traceable historical service record.

[0039] This embodiment achieves real-time resource data collection through edge computing nodes, employs Bloom filters for efficient data verification, and integrates blockchain technology to ensure data immutability. A multidimensional feature vector construction method transforms heterogeneous logistics resources into standardized, computable units, providing a reliable data foundation for path feasibility scoring and graph neural network scheduling. The utilization of spatiotemporal locality further optimizes computing resource allocation at the edge layer, improving the overall system's responsiveness.

[0040] In some embodiments, real-time collection of transport fluctuation parameters by edge computing nodes includes: Real-time collection of raw sensor data, including vehicle GPS positioning coordinates, multi-zone temperature and humidity sensor readings, and electronic lock switch status signals; Adopting an adaptive sampling strategy to dynamically collect raw sensor data, the sampling frequency is automatically increased when the temperature and humidity change rate exceeds the preset threshold or the GPS signal is lost; The collected data is denoised in real time using a lightweight filtering algorithm, and the edge node digital signature and timestamp are added to generate transport fluctuation parameters with identity authentication.

[0041] In this embodiment, transport fluctuation parameters are a set of key indicators reflecting the dynamic changes in the logistics transport process. They are composed of raw sensor data collected in real time by edge computing nodes. This raw sensor data includes GPS coordinates for tracking the continuity of the transport route, temperature and humidity sensor readings for monitoring the stability of the cargo storage environment, and electronic lock switch status signals for indicating the security of cargo packaging during transport.

[0042] An adaptive sampling strategy balances monitoring accuracy and resource consumption by dynamically adjusting data collection frequency. When the rate of change in temperature and humidity exceeds a preset threshold or when the GPS signal is lost, the system automatically increases the sampling frequency. This threshold is determined based on cargo characteristics (such as the thermal sensitivity of cold chain pharmaceuticals) and historical transportation environment data. A lightweight filtering algorithm uses a modified moving average method to perform real-time denoising on collected data. Different filter window sizes are set to account for GPS coordinate fluctuations and instantaneous drift in the temperature and humidity sensors, ensuring that the denoised data retains valid fluctuation characteristics.

[0043] Preferably, the additional process of digital signature and timestamp of edge nodes is implemented through a hardware security module, and an asymmetric encryption algorithm is used to sign the filtered data to ensure the non-tamperability of the transport fluctuation parameters. The generated transport fluctuation parameters with identity authentication will be used for subsequent anomaly detection, in which GPS trajectory jumps, temperature and humidity anomalies, and electronic lock status anomalies constitute the basis for multi-dimensional correlation detection. For the case of multi-sensor data conflicts, a cross-modal arbitration strategy based on Kalman filtering is adopted. By establishing a state space model of each sensor data, the anomaly probabilities of different modalities are dynamically weighted and fused, and finally a comprehensive anomaly score is output. For example, when the temperature and humidity sensor readings are abnormal but the electronic lock status is normal, the system will combine the GPS trajectory continuity to determine whether the environmental parameter fluctuations are caused by the temporary stop of the vehicle.

[0044] This embodiment uses edge computing nodes to achieve multi-dimensional dynamic monitoring of the transportation process. Adaptive sampling and lightweight filtering ensure the real-time and reliability of data collection. The identity authentication mechanism based on the hardware security module provides a trusted data source for subsequent anomaly detection, and the cross-modal arbitration strategy supported by the Kalman filter effectively solves the problem of multi-sensor data conflicts and overcomes the shortcomings of single temperature detection. This embodiment achieves efficient and reliable monitoring of the transportation process through multi-dimensional sensor data fusion and edge intelligent processing, significantly improving the accuracy of anomaly detection and system reliability.

[0045] In some embodiments, a spatiotemporal consistency verification algorithm is used to filter abnormal data, and the verified transportation fluctuation parameters are stored on the chain to generate a logistics resource trust status matrix, including: Receive the transport fluctuation parameters, parse and obtain the vehicle GPS trajectory sequence, temperature and humidity time series data, and electronic lock status change records, and record them as the first processed data; Construct a spatiotemporal correlation analysis model, input the first processed data into the spatiotemporal correlation analysis model, identify the matching degree between the GPS trajectory and the temperature and humidity changes through a multi-dimensional clustering algorithm, detect whether there is human tampering or equipment failure, and obtain the second processed data; The second processed data is anonymized using a differential privacy protection algorithm to generate a standard data unit that meets the storage requirements of the blockchain; The processed standard data units are written into the blockchain through smart contracts to generate a trusted status matrix of logistics resources and update the real-time status records in the decentralized logistics resource pool.

[0046] In this embodiment, the spatiotemporal consistency verification algorithm identifies anomalies by analyzing the temporal and spatial correlations of transportation data. This algorithm is used to verify the physical plausibility of GPS trajectory sequences and temperature and humidity time series data. Specifically, the GPS trajectory sequence in the first processed data is spatiotemporally aligned with the electronic lock status change records. For example, an anomaly flag is triggered when the electronic lock is opened outside a pre-set geofenced area. Meanwhile, a sudden change in temperature and humidity without a vehicle stop (GPS stationary) is considered a possible device failure. Preferably, a multidimensional clustering algorithm employs the density-based DBSCAN method to cluster GPS coordinates and temperature and humidity readings according to spatiotemporal windows. Outliers (e.g., sudden temperature and humidity readings that do not match a stable driving trajectory) are identified using intra-cluster distance thresholds to detect tampering or sensor failure.

[0047] For differential privacy protection algorithms, it is preferred to use - Differential privacy parameters control the strength of data anonymization, where the privacy budget allocation strategy is based on data sensitivity classification: GPS trajectory privacy budget is higher ( ), to prevent location re-identification; the privacy budget of temperature and humidity data is low ( ), due to its relatively weak sensitivity. Differential privacy algorithms are implemented by adding Laplace noise, with the noise scale inversely proportional to the privacy budget, ensuring a balance between data availability (such as preserving temperature and humidity trends) and privacy protection. The standard data unit after differential privacy processing consists of anonymized trajectory segments, aggregated temperature and humidity intervals, and encrypted electronic lock events. Its data structure meets the key-value storage requirements of blockchain.

[0048] Smart contracts perform double verification when writing to the blockchain: first, they check whether the hash value of the data unit is consistent with the spatiotemporal verification result, and secondly, they verify whether the differential privacy parameters meet the preset standards. The passed logistics resource trust state matrix will record the data trust score of each edge node and update the resource pool in conjunction with the transportation fluctuation parameters of the aforementioned embodiment. Preferably, the data trust score of each edge node is dynamically adjusted based on the anomaly detection results. For example, when a node frequently triggers temperature and humidity anomalies but the GPS trajectory is normal, its trust score decreases, and the resource pool will reduce the task allocation to the node.

[0049] This embodiment achieves trusted processing and privacy-secure storage of logistics and transportation data through the synergy of the spatiotemporal consistency verification algorithm and differential privacy protection. The spatiotemporal correlation analysis model uses a multi-dimensional clustering algorithm to identify physical contradictions between GPS trajectories and temperature and humidity data, effectively detecting human tampering and equipment failures; the differential privacy protection algorithm uses sensitivity-graded noise injection to prevent location re-identification while retaining data trends. The smart contract's dual verification mechanism of hash consistency verification and privacy parameter review ensures the integrity and compliance of the on-chain data. The generated logistics resource trust status matrix dynamically updates the edge node trust score, providing an accurate status basis for the decentralized resource pool. This embodiment achieves privacy security while ensuring data authenticity, providing a reliable basis for the trusted scheduling of logistics resources.

[0050] In some embodiments, the initial transportation path plan and the logistics resource trust state matrix are input into the graph neural network scheduling model. The spatiotemporal dependencies between transportation nodes are modeled through the graph attention mechanism, and the path weights are dynamically adjusted to output a dynamic scheduling plan that includes real-time path optimization suggestions, emergency scheduling instructions, and resource allocation strategies, including: Construct a transportation network topology graph, converting warehousing nodes, transportation vehicles, and transit stations into graph nodes. Establish node feature vectors corresponding to graph nodes based on the spatiotemporal characteristics and resource availability indicators in the logistics resource trust state matrix, and construct edge features based on the actual transportation route connection relationships. A multi-layer graph attention network is used to learn the features of the transportation network topology. The spatial dependency weights between graph nodes are calculated through a multi-head attention mechanism. Combined with a temporal convolutional network to capture the dynamic change pattern of transportation resources, the optimized path weights of the initial transportation path are obtained. The message passing mechanism of the graph neural network is used to iteratively update the priority scores of each initial transportation path and generate a set of candidate optimized paths; Based on a set of candidate optimized paths, a reinforcement learning strategy is used to dynamically generate real-time path optimization suggestions. When abnormal fluctuations in the logistics resource trust status matrix are detected, emergency dispatch instructions including backup routes and emergency resource deployment are automatically triggered. The optimized path weights are integrated with the initial transport path plan to output a dynamic scheduling plan that includes optimal path recommendations, resource reallocation plans, and risk warning information.

[0051] In this embodiment, the transportation network topology graph is a network representation method that abstracts the warehousing nodes, transport vehicles, and transfer stations in the logistics system into graph nodes and constructs edge features based on the actual transport route connection relationships. Node feature vectors are generated by combining spatiotemporal features and resource availability indicators in the logistics resource trust status matrix. Spatiotemporal features include GPS trajectory verification results and temperature and humidity data credibility scores. Resource availability indicators include parameters such as vehicle cargo capacity and temperature and humidity compliance status. Edge features are quantitatively constructed using data such as travel time and road grade of historical transport routes.

[0052] The graph attention mechanism uses a multi-head attention structure to calculate the spatial dependency weights between nodes, while simultaneously integrating it with a temporal convolutional network to analyze the dynamic changes in transportation resources. The spatial dependency weight calculation takes into account the physical distance between nodes and the matching degree of transportation capacity, while the temporal convolutional network is used to capture the temporal patterns of vehicle speeds and cargo status changes. This joint spatiotemporal modeling approach can more accurately reflect the actual degree of connectivity between nodes in the transportation network.

[0053] The message-passing mechanism updates path priority scores through multiple iterations. The first layer aggregates real-time resource status information from adjacent nodes. The second layer adjusts the connection strength between nodes based on spatiotemporal dependency weights. The third layer introduces a time-sensitive factor to prioritize transport tasks approaching deadlines. This mechanism ensures that the path optimization process accounts for both static network topology and dynamic resource changes.

[0054] A reinforcement learning strategy generates scheduling decisions based on real-time transportation status. Its state space includes candidate route priority scores and resource availability data, while its action space defines operations such as route adjustments and resource reallocation. When an anomaly is detected in the logistics resource trust matrix, such as abnormal temperature and humidity data at a node but normal GPS trajectory, an alternate route bypassing that node is automatically generated, and emergency resources are deployed through related storage nodes. The resulting dynamic scheduling plan includes optimal route recommendations, resource allocation adjustment suggestions, and anomaly risk warnings.

[0055] This embodiment combines the temporal characteristics of GPS trajectories with the spatial correlation analysis of the graph attention mechanism to construct a spatiotemporal joint representation that can accurately reflect the actual state of the transportation network. This not only ensures the spatial rationality of the path optimization plan, but also ensures that the scheduling decision can respond to dynamic changes in the transportation process in a timely manner. Through real-time interaction with the trusted state matrix of logistics resources, it achieves efficient scheduling of transportation resources and rapid handling of abnormal situations.

[0056] In some embodiments, the smart contract execution is triggered based on the dynamic scheduling scheme. The smart contract automatically allocates transportation tasks according to preset rules, freezes abnormal batch transaction permissions, and calculates carrier incentive tokens. At the same time, the compliance of the scheduling instructions is verified through zero-knowledge proof to obtain the scheduling execution results, including: Analyze the path optimization suggestions and resource allocation strategies in the dynamic scheduling plan, assign transportation tasks to the optimal carrier node through smart contracts, and generate task allocation results. The allocation process is based on the real-time available resources in the logistics resource trust status matrix and the carrier's credit score to make a multi-dimensional selection; When it is detected that the transportation fluctuation parameters exceed the preset safety threshold, the exception handling smart contract is triggered, automatically freezing the trading rights of the relevant batch of products and generating an unalterable violation record on the blockchain. At the same time, an encrypted alert is sent to the supervision node, and an exception handling record is generated. The number of incentive tokens is dynamically calculated based on the carrier's task completion quality, timeliness, and temperature control compliance rate. The token value anchor coefficient is adjusted through decentralized oracle data to obtain token incentive details. Build a zero-knowledge proof verification module to generate a validity certificate that the dispatch instruction complies with the product sales rules, so that regulators can conduct compliance spot checks and obtain compliance certificates; The task assignment results, exception handling records, token incentive details and compliance certificates are packaged to generate the scheduling execution results.

[0057] In this embodiment, the optimal carrier node selection process is implemented through a smart contract. This contract makes a multi-dimensional selection based on the real-time available resources and carrier credit scores in the logistics resource trust status matrix. Real-time available resources include the current location of the transport vehicle, load capacity, and refrigeration unit status, which are continuously updated by edge computing nodes. The carrier credit score is derived from historical performance records stored on the blockchain, including task completion rate, temperature control compliance rate, and violation statistics, ensuring the immutability and objectivity of the scoring data. A weighted scoring mechanism is used in the allocation process, prioritizing carriers with high resource matching and excellent credit scores, thereby ensuring transportation efficiency while reducing operational risks.

[0058] When transport fluctuation parameters exceed preset safety thresholds, the system automatically triggers an exception handling smart contract. Preferably, these thresholds are dynamically adjusted based on product characteristics (e.g., drug temperature control range) and historical transport data (e.g., equipment failure rates) to ensure accurate anomaly detection. The smart contract first freezes on-chain transaction permissions for the relevant batch of products, preventing the anomalous batch from entering subsequent circulation. It also generates a violation record on the blockchain, including a timestamp, the type of anomaly, and the appropriate action. This violation record is stored in a distributed manner to ensure immutability, and encrypted alerts are sent to supervisory nodes for timely intervention. The generation of the exception handling record strictly adheres to pre-set rules, ensuring standardized and traceable action measures.

[0059] The calculation of incentive tokens comprehensively considers the carrier's task completion quality, timeliness, and temperature control compliance rate. The token value anchor coefficient is dynamically adjusted through decentralized oracle data obtained from external market conditions, aligning incentives with actual market conditions. Token allocation preferably uses a contribution algorithm based on Shapley values ​​to quantify the actual contributions of multiple carriers to collaborative tasks, avoiding the unfairness caused by simple equal distribution. A deflationary mechanism is also introduced: when a carrier's credit score falls below a set threshold, a portion of its tokens will be destroyed, maintaining the long-term stability of the token economy and the effectiveness of incentives.

[0060] The zero-knowledge proof verification module is preferably implemented using zk-SNARKs technology. The smart contract compiles scheduling instructions and pre-set rules into a verifiable circuit, generating a compliance proof without disclosing the original data. This ensures the compliance of scheduling instructions without revealing the specific instruction content. This module verifies that scheduling instructions comply with pre-set product transportation rules, allowing regulators to conduct spot checks and verification. The resulting scheduling execution results include task assignment results, exception handling records, token incentive details, and compliance proofs. All data is stored on the blockchain to ensure integrity and auditability.

[0061] This embodiment achieves full automation and trustworthiness in dynamic logistics scheduling through the collaborative mechanism of smart contracts and zero-knowledge proofs. Rules such as route optimization, resource allocation, exception handling, and incentive calculation are encoded into executable logic, leveraging the blockchain's immutable nature to ensure transparent and auditable data at all stages. During the resource allocation phase, smart contracts dynamically match optimal carrier nodes using a multi-dimensional selection process based on a real-time, updated logistics resource trust matrix and carriers' historical credit scores. The exception handling phase precisely triggers the freezing of transaction permissions and the posting of violation records to the blockchain through a dynamic threshold adjustment mechanism. Incentive tokens are anchored to market data through a decentralized oracle and introduce a contribution algorithm, achieving a strong correlation between value distribution and credit scores. A zero-knowledge proof module verifies scheduling compliance without exposing raw data, creating a closed-loop supervisory system. This embodiment significantly improves resource allocation efficiency in cold chain logistics, reduces the risk of human intervention through the automated execution of on-chain rules, and leverages cryptographic technology to ensure a balance between commercial privacy and regulatory compliance, providing a reliable and efficient scheduling paradigm for the transportation of high-value products.

[0062] In some embodiments, the scheduling execution results and the logistics resource trust status matrix are updated to the distributed ledger according to the blockchain consensus mechanism, and a logistics management report containing transportation task completion, temperature control compliance rate, and violation event statistics is generated and output, including: Based on the scheduling execution results, the smart contract automatically calculates the task completion indicators of each transportation batch. Task completion indicators include timeliness achievement rate, path deviation and resource utilization efficiency; Extract historical data from temperature control sensors in the trusted state matrix of logistics resources, use a sliding window algorithm to calculate the temperature control compliance rate at each transportation stage, and identify the frequency and duration of abnormal temperature fluctuation events; Through the violation record smart contract on the blockchain, the distribution of violation types and spatial distribution characteristics within the preset period are counted to generate analysis results of violation hotspots; The results of task completion indicators, temperature control compliance rates, and violation hotspot analysis are weighted and integrated to generate carrier credit score update recommendations and supply chain optimization plans. The carrier's credit score update suggestions and supply chain optimization plans are integrated with the task completion indicators, temperature control compliance rates, and violation hotspot analysis results into a logistics management report and output. The logistics management report includes a transportation performance trend chart, a temperature control quality heat map, and a violation risk warning prompt.

[0063] In this example, the blockchain's immutability ensures the authenticity and trustworthiness of the transport data, providing a reliable data foundation for subsequent analysis. The distributed ledger allows all participants to access the latest data in real time, facilitating collaborative management.

[0064] In calculating task completion metrics, the timeliness achievement rate is determined by comparing the actual completion time of the transport task with the time threshold preset by the smart contract. If the actual time is earlier than or equal to the threshold, it is recorded as fully achieved. Path deviation is calculated based on the standard deviation of the spatial distance between the vehicle's GPS trajectory and the planned path, reflecting the degree of route adherence during the transport process. Resource efficiency is comprehensively evaluated based on the ratio of load capacity to actual load volume, combined with the operating power consumption of the refrigeration unit. These three metrics are converted to a unified dimension using the smart contract's built-in normalization algorithm, and then the entropy weight method is used to determine the weight of each metric. The final weighted summation results in an overall task completion score. This scoring mechanism prevents the excessive impact of a single metric deviation on the overall assessment, providing an objective basis for subsequent credit score updates.

[0065] The temperature control compliance rate is calculated using a sliding window algorithm that processes historical sensor data. The window size is set based on the transportation time span and data sampling frequency, ensuring that the statistical results reflect both overall trends and local fluctuations. Temperature anomalies are identified based on preset threshold ranges, recording the duration and frequency of threshold violations, providing a quantitative basis for cold chain quality control. Preset threshold ranges can be dynamically set based on industry standards and product characteristics, and the corresponding threshold parameters can be automatically loaded by smart contracts.

[0066] Violation analysis relies on immutable records stored on the blockchain. Smart contracts categorize and compile statistics on violations within a pre-set period, analyzing the temporal distribution of different types of violations. Spatial distribution patterns are visualized using geographic information system technology, and high-risk areas are identified based on the topology of the transportation network.

[0067] Credit score update recommendations are generated using a multi-metric fusion approach. The weight of each metric is dynamically adjusted based on its impact on transportation quality, ensuring fairness and guidance in the scoring results. Supply chain optimization solutions are based on historical data and current problem analysis to propose targeted improvement measures.

[0068] The resulting logistics management report integrates various analytical results. Transport performance trend charts visually demonstrate changes in carrier performance, temperature control quality heat maps highlight critical control points for cold chain transportation, and non-compliance risk warnings help managers proactively mitigate potential risks. The report's output format supports viewing on a variety of devices, making it easier for users of different roles to access the information they need.

[0069] This implementation uses a blockchain consensus mechanism to synchronize dispatch execution results and the logistics resource trust status matrix to a distributed ledger. Smart contracts are then used to automatically calculate transport task completion metrics. A sliding window algorithm is then used to analyze temperature control compliance rates and abnormal events. Furthermore, the spatial distribution of violations is analyzed based on the blockchain's immutable nature. Using an entropy-weighted method to dynamically weight and integrate multiple metrics, recommendations for carrier credit score updates and supply chain optimization solutions are generated. Ultimately, these recommendations are integrated into a logistics management report that includes transport performance trend charts, temperature control quality heat maps, and violation risk warnings.

[0070] This embodiment realizes the full-chain trusted traceability and multi-dimensional quantitative evaluation of the transportation process, ensures data authenticity and calculation objectivity through the automated execution of smart contracts, improves risk identification accuracy with the help of spatiotemporal feature fusion analysis, and provides closed-loop support for cold chain logistics management from real-time monitoring to decision optimization, significantly enhancing the systematicness and collaborative efficiency of transportation quality control.

[0071] In a second aspect, this embodiment further provides a computer-readable storage medium storing computer program instructions, which implement the method described in the first aspect when executed by a processor.

[0072] The computer program involved in this embodiment can be stored in a computer device readable storage medium, which includes but is not limited to a disk, a magnetic tape, a magnetic card, a floppy disk, a flash memory, an optical disc, an optical card, a read-only memory (ROM), a random access memory (RAM), an erasable programmable ROM (EPROM) and an electrically erasable programmable ROM (EEPROM), etc., and also includes other biological, physical or chemical structures that can achieve similar or equivalent functions to the storage media listed above, such as DNA, RNA, proteins and other units with information storage capabilities. In a specific embodiment, the storage medium involved can be one of the above-mentioned media types or a combination of the above-mentioned media types. In different embodiments, the computer program involved in the embodiment can be stored in a single medium in a centralized manner or in a distributed manner in multiple media. The memory containing the computer device readable storage medium can be a non-volatile memory or a random access memory. These computer device readable storage media can be built into the device or can be connected to the device involved in the embodiment as an external device or part of an external device. In some embodiments, a memory having a computer-readable storage medium is deployed locally; in other embodiments, a solution of deploying the memory away from the processor may also be adopted, such as a network-attached memory accessed via an RF circuit or an external port and a communication network, wherein the communication network may be the Internet, one or more intranets, a local area network (LAN), a wide area network (WLAN), a storage area network (SAN), etc., or a suitable combination thereof, as long as the computer device can access the memory. In addition, the computer program involved in the embodiment can be stored in plaintext / ciphertext form, or can be designed as training data, which can be integrated and reorganized through model training and implicitly stored in the parameter state of a deep neural network or other machine learning model.

[0073] See also Figure 3 In a third aspect, this embodiment further provides an electronic device 1, comprising a memory 11 and a processor 12, wherein the memory 11 is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor 12 to implement the method described in the first aspect.

[0074] The processor described in this embodiment can be implemented by hardware, firmware, software or a combination thereof, and can use a circuit, a single or multiple application-specific integrated circuits (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller, or at least one of a microprocessor. It also includes other physical, biological or chemical structures that can achieve similar or equivalent functions to the processors listed above, such as biological neurons, quantum computing units, DNA computing units, etc., so that the processor can execute some or all of the steps in the computer program or method involved in each embodiment of the present application, or any combination of the steps mentioned therein.

[0075] Different from the existing technology, the above technical solution has the following beneficial effects: by constructing a trusted state matrix of logistics resources, dynamic digital mapping of transportation resources is realized, and the transportation network is modeled in real time in combination with the spatiotemporal attention mechanism of the graph neural network scheduling model, forming a closed-loop management from demand collection, path optimization to dynamic scheduling; the blockchain consensus mechanism is used to ensure the reliable synchronization of scheduling execution results and resource status, and the task completion index, temperature control compliance rate and violation event statistics are automatically calculated through smart contracts, and a logistics management report containing credit score update suggestions and optimization plans is generated.

[0076] The above technical solution combines real-time data collection from edge computing nodes, a spatiotemporal consistency verification algorithm, and differential privacy protection to address the issues of insufficient data authenticity and delayed response in traditional logistics management. Through the collaborative optimization of graph attention networks and reinforcement learning, dynamic weight adjustment of transportation routes and rapid response to anomalies are achieved. Zero-knowledge proofs are used to verify scheduling compliance, ensuring commercial privacy while meeting regulatory requirements. Ultimately, this achieves trusted traceability and optimized resource allocation throughout the entire product logistics process, significantly improving transportation efficiency, temperature control compliance, and anomaly handling capabilities, providing a safe and reliable intelligent scheduling solution for high-value product supply chains, especially cold chain logistics.

[0077] The above descriptions are only some embodiments of the present invention and do not limit the scope of protection of the present invention. Any equivalent device or equivalent process transformation made by using the contents of the description and drawings of the present invention, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A product logistics management and scheduling method based on blockchain, characterized in that: include: Collect product logistics and transportation demand information and supply chain node data, including product type, transportation priority, temperature control requirements, and target delivery area; and supply chain node data including storage location, transportation vehicle status, and carrier qualifications; Dynamically decomposing the logistics transportation demand information, generating an initial transportation route plan using a multi-objective optimization algorithm, and calculating a path feasibility score of the initial transportation route plan in combination with historical transportation data; A decentralized logistics resource pool is built based on the supply chain node data. Transport fluctuation parameters, including vehicle GPS trajectory, temperature and humidity sensor data, and electronic lock status, are collected in real time through edge computing nodes. A spatiotemporal consistency verification algorithm is used to filter abnormal data. The verified transport fluctuation parameters are then stored on-chain to generate a trusted status matrix for logistics resources. The initial transport path plan and the logistics resource trust state matrix are input into the graph neural network scheduling model, the spatiotemporal dependency relationship between transport nodes is modeled through the graph attention mechanism, the path weights are dynamically adjusted, and a dynamic scheduling plan including real-time path optimization suggestions, emergency scheduling instructions and resource allocation strategies is output; The dynamic scheduling scheme triggers the execution of a smart contract, which automatically allocates transportation tasks, freezes abnormal batch transaction permissions, and calculates carrier incentive tokens according to preset rules. At the same time, it verifies the compliance of the scheduling instructions through zero-knowledge proof and obtains the scheduling execution result; Based on the blockchain consensus mechanism, the scheduling execution results and the logistics resource trust status matrix are updated to the distributed ledger, and a logistics management report including the completion rate of transportation tasks, temperature control compliance rate and violation event statistics is generated and output.

2. The product logistics management and scheduling method based on blockchain according to claim 1 is characterized in that: Dynamically decompose the logistics transportation demand information, use a multi-objective optimization algorithm to generate an initial transportation route plan, and calculate the path feasibility score of the initial transportation route plan in combination with historical transportation data, including: Based on the product type, transportation priority, and temperature control requirements in the logistics transportation demand information, a transportation constraint weight matrix is ​​constructed using the analytic hierarchy process. The transportation constraint weight matrix is ​​used to quantify the differentiated requirements of different product categories for timeliness, temperature stability, and safety; The transportation constraint weight matrix is ​​aligned with the historical transportation data in time and space, and an improved non-dominated sorting genetic algorithm is used to perform multi-objective optimization and solve the problem to generate a set of candidate path solutions including a Pareto optimal solution set; Performing a dynamic risk assessment on each candidate route in the candidate route set, using a long short-term memory network to predict the potential traffic congestion probability and temperature control deviation risk of each candidate route in the target delivery area, and combining the real-time vehicle status data in the logistics resource trust state matrix to calculate a comprehensive feasibility score for each candidate route; Based on the comprehensive feasibility score, the candidate path plans in the candidate path plan set are sorted and screened, and the candidate path plans with a comprehensive feasibility score lower than the preset feasibility score threshold are eliminated. The remaining candidate path plans are matched with the carrier qualifications in the supply chain node data and the initial transportation path plan and the corresponding path feasibility score are generated.

3. The product logistics management and scheduling method based on blockchain according to claim 1 is characterized in that: Constructing a decentralized logistics resource pool based on the supply chain node data, including: The edge computing nodes in the blockchain network collect dynamic resource data from each supply chain node in real time. The dynamic resource data includes the real-time inventory status of the storage node, the load capacity and available refrigeration unit information of the transport vehicle, and the current task load status of the carrier; Using a fast verification algorithm based on a Bloom filter to preliminarily screen the dynamic resource data, eliminating invalid or duplicate data records, and obtaining verified dynamic resource data; Associating and mapping the verified dynamic resource data with the static qualification information in the supply chain node data to construct a multi-dimensional resource feature vector, the multi-dimensional resource feature vector including space availability, equipment compatibility, and serviceability; The multi-dimensional resource feature vectors are classified and integrated according to preset rules to generate a decentralized logistics resource pool with a timestamp. Each resource unit in the decentralized logistics resource pool is accompanied by a verifiable credit score and historical service record.

4. The product logistics management and scheduling method based on blockchain according to claim 1 is characterized in that: Real-time collection of transport fluctuation parameters through edge computing nodes, including: Real-time collection of raw sensor data, including vehicle GPS positioning coordinates, multi-zone temperature and humidity sensor readings, and electronic lock switch status signals; Adopting an adaptive sampling strategy to dynamically collect the raw sensor data, the sampling frequency is automatically increased when the temperature and humidity change rate exceeds a preset threshold or the GPS signal is lost; The collected data is denoised in real time using a lightweight filtering algorithm, and the edge node digital signature and timestamp are added to generate transport fluctuation parameters with identity authentication.

5. The product logistics management and scheduling method based on blockchain according to claim 4 is characterized in that: A spatiotemporal consistency verification algorithm is used to filter abnormal data, and the verified transportation fluctuation parameters are stored on the chain to generate a trusted state matrix of logistics resources, including: Receive the transport fluctuation parameters, parse and obtain the vehicle GPS trajectory sequence, temperature and humidity time series data, and electronic lock status change records, and record them as the first processed data; Constructing a spatiotemporal correlation analysis model, inputting the first processed data into the spatiotemporal correlation analysis model, identifying the matching degree between the GPS trajectory and the temperature and humidity changes through a multidimensional clustering algorithm, detecting whether there is human tampering or equipment failure, and obtaining second processed data; The second processed data is anonymized using a differential privacy protection algorithm to generate a standard data unit that meets the storage requirements of the blockchain; The processed standard data units are written into the blockchain through smart contracts to generate a trusted status matrix of logistics resources and update the real-time status records in the decentralized logistics resource pool.

6. The product logistics management and scheduling method based on blockchain according to claim 1 is characterized in that: The initial transport path plan and the logistics resource trust state matrix are input into the graph neural network scheduling model. The spatiotemporal dependencies between transport nodes are modeled through the graph attention mechanism, and the path weights are dynamically adjusted to output a dynamic scheduling plan that includes real-time path optimization suggestions, emergency scheduling instructions, and resource allocation strategies, including: Construct a transportation network topology graph, converting warehousing nodes, transportation vehicles, and transfer stations into graph nodes, establishing node feature vectors corresponding to the graph nodes based on the spatiotemporal characteristics and resource availability indicators in the logistics resource trust state matrix, and constructing edge features based on the actual transportation route connection relationship; A multi-layer graph attention network is used to learn the features of the transportation network topology graph. The spatial dependency weights between graph nodes are calculated through a multi-head attention mechanism. Combined with a temporal convolutional network to capture the dynamic change pattern of transportation resources, the optimized path weights of the initial transportation path are obtained. The message passing mechanism of the graph neural network is used to iteratively update the priority scores of each initial transportation path and generate a set of candidate optimized paths; Based on the candidate optimization path set, a real-time path optimization suggestion is dynamically generated through a reinforcement learning strategy. When an abnormal fluctuation in the logistics resource trust state matrix is ​​detected, an emergency dispatch instruction including backup routes and emergency resource deployment is automatically triggered; The optimized path weights are integrated with the initial transport path plan to output a dynamic scheduling plan including optimal path recommendations, resource reallocation plans, and risk warning information.

7. The product logistics management and scheduling method based on blockchain according to claim 1 is characterized in that: The dynamic scheduling scheme triggers the execution of a smart contract, which automatically allocates transportation tasks, freezes abnormal batch transaction permissions, and calculates carrier incentive tokens according to preset rules. At the same time, the compliance of the scheduling instructions is verified through zero-knowledge proof, and the scheduling execution results are obtained, including: Analyze the path optimization suggestions and resource allocation strategies in the dynamic scheduling solution, assign the transportation tasks to the optimal carrier node through smart contracts, and generate task allocation results. The allocation process is based on the real-time available resources and carrier credit scores in the logistics resource trust status matrix to make multi-dimensional optimal selections; When it is detected that the transportation fluctuation parameters exceed the preset safety threshold, the exception handling smart contract is triggered, automatically freezing the trading rights of the relevant batch of products and generating an unalterable violation record on the blockchain. At the same time, an encrypted alert is sent to the supervision node, and an exception handling record is generated. The number of incentive tokens is dynamically calculated based on the carrier's task completion quality, timeliness, and temperature control compliance rate. The token value anchor coefficient is adjusted through decentralized oracle data to obtain token incentive details. Build a zero-knowledge proof verification module to generate a validity certificate that the dispatch instruction complies with the product sales rules, so that regulators can conduct compliance spot checks and obtain compliance certificates; The task assignment results, exception handling records, token incentive details and compliance certificates are packaged to generate the scheduling execution results.

8. The product logistics management and scheduling method based on blockchain according to claim 1 is characterized in that: Based on the blockchain consensus mechanism, the dispatch execution results and the logistics resource trust status matrix are updated to the distributed ledger, and a logistics management report containing transportation task completion, temperature control compliance rate, and violation statistics is generated and output, including: Based on the scheduling execution results, the task completion indicators of each transport batch are automatically calculated through smart contracts. The task completion indicators include timeliness achievement rate, path deviation and resource utilization efficiency; Extracting historical data of temperature control sensors from the logistics resource trustworthy state matrix, using a sliding window algorithm to calculate the temperature control compliance rate for each transportation stage, and identifying the frequency and duration of abnormal temperature fluctuation events; Through the violation record smart contract on the blockchain, the distribution of violation types and spatial distribution characteristics within the preset period are counted to generate analysis results of violation hotspots; The task completion index, temperature control compliance rate and violation hotspot analysis results are weighted and integrated to generate carrier credit score update suggestions and supply chain optimization plans; The carrier credit score update suggestions and supply chain optimization plans are integrated with the task completion indicators, temperature control compliance rates, and violation hotspot analysis results into a logistics management report and output. The logistics management report includes a transportation performance trend chart, a temperature control quality heat map, and a violation risk warning prompt.

9. A computer-readable storage medium storing computer program instructions, characterized in that: The computer program instructions implement the method according to any one of claims 1 to 8 when executed by a processor.

10. An electronic device comprising a memory and a processor, characterized in that: The memory is configured to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method according to any one of claims 1 to 8.

Images