A Zero-Trust-Based Dynamic Authorization Method for Cross-Domain Data Exchange among Multi-Agents

By adopting a zero-trust-based dynamic authorization method, which utilizes unique identification codes and dual verification mechanisms to dynamically adjust permissions, the system solves the security and privacy protection issues in cross-domain data exchange among multiple agents. It achieves accurate authorization authentication of agent identities and real-time identification of abnormal behaviors, thereby improving the system's security and robustness, while also possessing green and energy-saving advantages.

CN123074757APending Publication Date: 2026-11-10GUANGZHOU ZHIKA LOGISTICS TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611083748.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-21
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing multi-agent cross-domain data exchange suffers from insufficient security, static authorization mechanisms, lack of data privacy protection, and weak ability to detect abnormal behavior.

Method used

A zero-trust-based dynamic authorization method is adopted, which uses unique identification code configuration, dual verification mechanism and hierarchical temporary connection to achieve accurate authorization and continuous authentication of intelligent agent identity, dynamically adjust permissions, and combine carbon emission calculation for data access control.

Benefits of technology

It effectively prevents abuse of permissions, achieves fine-grained data privacy protection, identifies and handles abnormal behavior in real time, improves the security and robustness of multi-agent systems in cross-domain data exchange, and has green and energy-saving advantages.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123074757A_ABST
    Figure CN123074757A_ABST
Patent Text Reader

Abstract

This invention discloses a dynamic authorization method for cross-domain data exchange among multiple agents based on zero trust. The method includes configuring unique identification codes for each agent using a data interaction platform; acquiring original task information; the data interaction platform sending request data to each agent; and when an agent receives the request data or needs to actively send a data retrieval request to the data interaction platform, the agent sends first verification data to the platform, which then verifies the first verification data. Through unique identification code configuration and a dual verification mechanism, precise authorization and continuous authentication of agent identities are achieved. A hierarchical temporary connection mechanism is introduced, dynamically establishing or dissolving connections based on task assignment, breaking the limitations of traditional static authorization and effectively preventing abuse of permissions. Simultaneously, a data access control strategy based on trust zones achieves fine-grained data privacy protection.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of smart logistics technology, and in particular to a dynamic authorization method for cross-domain data exchange between multiple agents based on zero trust. Background Technology

[0002] With the deep integration of artificial intelligence and the Internet of Things (IoT) technologies, multi-agent systems have been widely applied in complex scenarios such as smart logistics, intelligent transportation, and industrial automation. In these scenarios, multiple heterogeneous agents need to frequently exchange data and collaborate across different trust domain boundaries to complete complex business processes. For example, in the field of smart logistics, different types of agents such as pickup robots, transport vehicles, and delivery drones need to share task data, location information, and environmental perception data in real time to achieve efficient cargo flow.

[0003] Traditional multi-agent data exchange typically employs a boundary-based security model, establishing trusted zones within the system and granting default trust to agents within those zones. This model usually uses a static authorization mechanism, where agents are granted fixed data access permissions upon registration, which remain unchanged throughout their lifecycle. However, this static authorization model has significant security vulnerabilities: once an agent's identity credentials are stolen or forged, an attacker can gain full access to that agent's data, leading to large-scale leaks of sensitive data. Furthermore, traditional models struggle to dynamically adjust permissions based on the agent's real-time state, failing to meet the principle of least privilege.

[0004] In terms of data privacy protection, existing technologies often employ an all-or-nothing data distribution strategy, meaning that agents either receive the complete raw data or are unable to access any data at all. This coarse-grained access control cannot adapt to the differentiated data needs of agents with different functions. Taking logistics scenarios as an example, transportation agents only need to obtain the destination information of goods, without needing to know the user's contact information. Existing technologies lack refined data anonymization and distribution mechanisms based on agent type and trust domain, which can easily lead to the excessive exposure of user privacy information. When agents exhibit suspicious behaviors such as deviating from the planned route, frequently requesting data outside of working hours, or abnormal network environments, the system often fails to identify and take timely blocking measures, significantly increasing the risk of data misuse or leakage. Summary of the Invention

[0005] The technical problem solved by this invention is that existing intelligent agents have problems such as insufficient security, static authorization mechanisms, lack of data privacy protection, and weak ability to detect abnormal behavior when exchanging data across domains.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a dynamic authorization method for cross-domain data exchange between multiple agents based on zero trust, comprising: Step S1: Configure a unique identification code for each intelligent agent using the data interaction platform; Step S2: Obtain the original task information; the data interaction platform sends request data to each intelligent agent. Step S3: When the intelligent agent receives the requested data or needs to actively send a data retrieval request to the data interaction platform, the intelligent agent sends the first verification data to the data interaction platform. The data interaction platform verifies the first verification data. After the first verification data passes the verification, the data interaction platform establishes a first temporary connection with the intelligent agent and adjusts the original task information into the first task information according to the type of the intelligent agent, and sends it to the intelligent agent that established the first temporary connection. Step S4: The intelligent agent receives the first task information and sends the second verification data and task claim status to the data interaction platform. Step S5: The data interaction platform generates a task list based on the task claiming status and the second verification data, terminates the first temporary connection with the unclaimed task agent, sends the second task information to the agent that has claimed the task, and upgrades the first temporary connection to the second temporary connection. The agent receives the second task information. Step S6: The data interaction platform performs dynamic management and control of the first temporary connection and the second temporary connection based on the first verification data, request data and the second verification data sent by the intelligent agent.

[0007] Preferably, the unique identification code is bound to the agent's type, manufacturer ID, product ID, serial number, and trust domain, and the unique identification codes of all agents are organized to generate a unique identification code library.

[0008] Preferably, the original task information includes waybill, party information, cargo category, cargo weight, cargo volume, cargo status, service content, cargo value, risk consequences, mode of transport, cost, contact number, arrival time, origin warehousing node and destination warehousing node; The intelligent agent includes a shipping intelligent agent, a transportation intelligent agent, and a receiving intelligent agent; The shipping agent establishes a first temporary contact with the data interaction platform, and the shipping agent sends the original task information to the data interaction platform. The intelligent transportation agent is used to transport goods from the origin storage node to the destination storage node, and utilizes real-time road condition and traffic information collected by cameras. After the warehousing node receives the goods, the receiving end intelligent agent establishes a first temporary contact with the data interaction platform. The receiving end intelligent agent sends the waybill, the actual type of goods received, the actual weight of the goods received, the actual volume of the goods received, and the actual status of the goods received to the data interaction platform.

[0009] Preferably, the request data includes shipping agent request data, transportation agent request data, and receiving agent request data; The data requested by the shipping agent includes the original task information update request and hidden encoding; The data requested by the transport agent includes a request to update transportable cargo data and a hidden code. The transportable cargo data includes the transport agent's current address information, current status, current time, remaining transport weight, remaining transport space, and hidden code. The receiving end intelligent agent requests data including receiving data update requests and hidden codes. The receiving data includes the warehouse node receiving the waybill, the current time, the actual type of goods received, the actual weight of the goods received, the actual volume of the goods received, and the actual status of the goods received.

[0010] Preferably, the first verification data includes the agent's type, manufacturer ID, product ID, serial number, acceptable tasks, current geographical location, historical carbon emission data, network environment, current time, and hidden code; The historical carbon emission data includes urban carbon emission data, suburban carbon emission data, and highway carbon emission data; After receiving the first verification data, the data interaction platform verifies the first verification data by including: Based on the ID of the agent that sent the first verification data, the manufacturer ID, product ID, and serial number in the first verification data are obtained and matched in the unique identification code database. If the manufacturer ID, product ID, and serial number match, the verification is successful. If one or more of the matched manufacturer ID, product ID, and serial number do not match, the verification fails. The relevant manufacturer ID, product ID, and serial number are retrieved from the unique identification code library, marked as risky, and verification information is sent. After the data interaction platform sends the verification information, it will not send any more information proactively until the verification is successful. If the hidden code is not identified, mark it as the first trusted zone; If the hidden code is identified, the key is retrieved to decrypt the hidden code, and the decrypted code is obtained. The first verification data and the decrypted code are sent to the data interaction platform, and the data interaction platform marks the second trust zone for it. The data interaction platform establishes a first temporary connection with the verified agent.

[0011] Preferably, the data interaction platform extracts the origin and destination warehouse nodes from the original task information for navigation path planning, calculates the length of urban roads, suburban roads and highways respectively, and calculates the total carbon emissions of each transportation agent according to the urban carbon emission data, suburban carbon emission data and highway carbon emission data, and sorts the total carbon emissions in descending order.

[0012] Preferably, the original task information is adjusted to the first task information, including: The data interaction platform adjusts the original task information to first task information based on the agent's trust domain, including: The data interaction platform extracts the cargo category, cargo weight, cargo volume, cargo status, service content, arrival time, cargo value, and risk consequences from the original task information as the first task information. The data interaction platform then sends the first task information to the transportation intelligent agent that has established the first temporary connection. The transportation agent receives the first task information and sends the task acceptance status and the second verification data to the data interaction platform. The task acceptance status includes accepting the task and not accepting the task. If the transportation agent does not send the task acceptance status and the second verification data within a unit of time, the data interaction platform regards the transportation agent as not accepting the task and terminates the first temporary connection with the transportation agent that has not accepted the task. The second verification data includes the current geographical location, current network environment, current time, navigation route planning and actual carbon emissions, as well as real-time road condition information and traffic information; The data interaction platform receives task claiming status and second verification data, selects the current geographical location from the transportation agents claiming tasks, calculates the distance between the current geographical location of each transportation agent claiming a task and the address of the origin warehouse node, selects the transportation agents whose distance is less than a predetermined distance threshold, and publishes the second task information. The transportation agents that receive the second task information upgrade from the first temporary connection to the second temporary connection. The second task information includes waybill, cargo category, cargo weight, cargo volume, cargo status, service content, cargo value, risk consequences, mode of transport, origin warehouse node address, destination warehouse node address, and task confirmation notification.

[0013] Preferably, the second temporary connection includes: The transport agent sends second verification data to the data interaction platform at unit time intervals, and the data interaction platform shares with the transport agent the real-time locations of the nearest X transport agents, as well as the real-time collected road condition information and traffic information; The transportation agent adjusts its navigation route planning in real time based on the latest road condition and traffic information, and sends the adjusted navigation route planning and actual carbon emissions to the data interaction platform. The data interaction platform records the actual carbon emissions of the transportation agent in real time and calculates the estimated carbon emissions of the adjusted navigation route planning.

[0014] Preferably, after the data interaction platform establishes a first temporary connection with the intelligent agent, the intelligent agent sends a data retrieval request to the data interaction platform; The agent has no second trust zone marker; If the agent is a shipping agent, it can retrieve the historical original task information recorded by the shipping agent. If the agent is a transportation agent, it can retrieve the historical second task information of the transportation agent; If the intelligent agent is a receiving end intelligent agent, it can retrieve the historical receiving data of the receiving end intelligent agent; If the intelligent agent has a second trust zone marker, it can, in addition to having the authority to access data information without a second trust zone marker, also access all original task information related to the second task information in a single instance, except for the party's information, fees, and contact numbers. If the intelligent agent is a transportation intelligent agent, it can also access the remaining transportation weight and remaining transportation space of the nearest X transportation intelligent agents.

[0015] Preferably, step S6 specifically includes: If a single agent sends M data retrieval requests to the data interaction platform during non-working hours, or sends N data retrieval requests to the data interaction platform within a short period of time, and if the agent and the data interaction platform are in a first temporary connection, the data interaction platform will disconnect the first temporary connection with the agent, mark the agent as having a risk, and send verification information. If the intelligent agent and the data interaction platform are connected in a second temporary manner, the data interaction platform will disconnect the second temporary connection with the intelligent agent and send verification information. After receiving the verification information, the intelligent agent will send verification materials, including the manufacturer's license, product ID image and product photos. Before the verification is passed, the data interaction platform will no longer send any information to the intelligent agent. After the verification is passed, the data interaction platform and the intelligent agent will send information normally. If the agent fails to send the second verification data to the data interaction platform at regular intervals under good network conditions, or if the agent's current geographical location is more than a predetermined distance threshold from the predetermined route, the data interaction platform will disconnect the second temporary connection with the agent, mark the agent as a risk, and send verification information.

[0016] The beneficial effects of this invention are as follows: Through unique identification code configuration and a dual verification mechanism, precise authorization and continuous authentication of intelligent agent identities are achieved. A hierarchical temporary connection mechanism is introduced, dynamically establishing or dissolving connections based on task assignment, breaking the limitations of traditional static authorization and effectively preventing abuse of permissions. Simultaneously, a data access control strategy based on trust zones achieves fine-grained data privacy protection. Dynamic management enables real-time identification and automatic handling of abnormal agent behavior, improving the security and robustness of multi-agent systems during cross-domain data exchange. The scheduling logic combined with carbon emission calculation further demonstrates the technological advantages of green energy conservation. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the basic process of a dynamic authorization method for cross-domain data exchange based on zero trust provided in an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of the data interaction between the data interaction platform and the agents in a dynamic authorization method for cross-domain data exchange based on zero trust provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] In multi-agent collaborative operation scenarios, different types of agents need to exchange data across trust domain boundaries. However, traditional security models, which employ static authorization mechanisms, struggle to address security threats such as identity forgery, data leakage, and abnormal behavior. In this invention, zero trust means that the data interaction platform does not grant default trust to any agent. All data access and task allocation require dynamic identity verification and continuous behavioral evaluation. Zero trust in this invention includes identity zero trust, behavior zero trust, and data zero trust. Identity zero trust means that each agent does not obtain any data access permissions before completing identity verification and establishing a connection with the data interaction platform. Behavior zero trust means that during task execution, the data interaction platform continuously monitors the agent's behavioral characteristics (request frequency, location trajectory, carbon emissions, connection status), automatically adjusting or severing permissions upon detecting abnormal behavior. Data zero trust means that the data interaction platform strictly limits the range of data that an agent can access based on the agent's trust domain definition, ensuring that each agent only obtains the minimum dataset necessary for its business. Through unique identification code configuration, dual verification mechanisms, and dynamic management of temporary connections, the security goal of never trusting and always verifying is achieved.

[0021] This embodiment uses a smart logistics scenario as an example. In this scenario, the data interaction platform is deployed on a cloud server and is responsible for the overall management of identity authentication, permission allocation, and data distribution for all intelligent agents. The intelligent agents include three categories: shipping agents, transportation agents, and receiving agents, which respectively perform tasks such as cargo pickup, trunk transportation, and last-mile delivery. Each intelligent agent communicates with the data interaction platform via a wireless network and is equipped with data acquisition devices such as positioning modules, cameras, and sensors.

[0022] Example 1, referring to Figure 1 and Figure 2 As an embodiment of the present invention, a dynamic authorization method for cross-domain data exchange based on zero trust multi-agents is provided, comprising: Step S1: Configure a unique identification code for each intelligent agent using the data interaction platform; In this embodiment, the unique identifier is a set of multi-dimensional attribute codes used to uniquely identify the identity of an agent in the system, enabling accurate identification and classification management of agents. The unique identifier is also known as the AgentID (Globally Unique Identifier), a globally unique identifier assigned to each agent. The unique identifier is generated and assigned by the data interaction platform during agent registration, stored in a unique identifier database, and serves as the foundational data source for all subsequent authentication and permission checks. It also works in conjunction with the agent's trust domain definition to achieve differentiated authorization.

[0023] Step S2: Obtain the original task information; the data interaction platform sends request data to each intelligent agent. The unique identifier is bound to the agent's type, manufacturer ID, product ID, serial number, and trust domain. The unique identifiers of all agents are then organized to generate a unique identifier library.

[0024] In this embodiment, the unique identifier is bound to the agent's type, manufacturer ID, product ID, serial number, and trust domain to facilitate the identification of the agent. Specifically, the agent's type is a classification identifier for the business functions performed by the agent, used to determine the agent's data access permission scope; the manufacturer ID is a unique code for the agent's manufacturer, used to trace the origin of the equipment; the product ID is a unique code for the agent's product model, used to distinguish different specifications and models of equipment; and the serial number is the serial number of a single agent, used to achieve unique identification at the device level.

[0025] The original task information includes waybill, party information, cargo category, cargo weight, cargo volume, cargo status, service content, cargo value, risk consequences, mode of transport, cost, contact number, arrival time, origin warehouse node and destination warehouse node; the original task information is the information entered when receiving logistics tasks, and is the basic data source for subsequent task distribution and data processing.

[0026] In this embodiment, the waybill is the unique order number for this logistics task; the party information refers to the basic information of the originating and destination warehousing nodes; the cargo category is the classification of the items within the cargo (such as documents, electronic products, frozen products, bulk goods, hazardous chemicals, or fragile items); the cargo weight and volume are the physical attribute parameters of the cargo; the service content is the service type selected by the user (preconditions such as hazardous chemical transportation / cold chain transportation / carbon emission quotas, etc.); the fee is the amount receivable for this service; the contact information includes the mobile phone number of the consignor or recipient; the transportation method specifies whether the transportation requires special vehicles such as hazardous chemical transportation, cold chain transportation, or fresh food transportation; and the originating and destination warehousing nodes are the originating and ending geographical locations of the cargo.

[0027] Intelligent agents include shipping intelligent agents, transportation intelligent agents, and receiving intelligent agents; The shipping-side intelligent agent establishes a first temporary connection with the data interaction platform, sending raw task information to the platform. The shipping-side intelligent agent refers to an intelligent terminal device deployed at the logistics start and end nodes (such as the origin and destination warehouse nodes) to digitally create and initialize logistics tasks. In this invention, the shipping-side intelligent agent typically exists in the form of a handheld PDA (Personal Digital Assistant), an industrial tablet computer, or a PC client. Its core function is to collect basic attribute data of the goods and encapsulate it into standardized raw task information. When sending the raw task information, the shipping-side intelligent agent simultaneously sends its own device fingerprint (such as MAC address and device serial number) as auxiliary information for identity verification. The platform receives and stores this raw task information as baseline data for subsequent scheduling and calculation.

[0028] The transportation agent is used to transport goods from the origin storage node to the destination storage node, and uses real-time road condition and traffic information collected by cameras; In this embodiment, the transportation intelligent agent is used to transport goods from the originating storage node to the destination storage node, utilizing real-time road and traffic information collected by cameras. The transportation intelligent agent refers to an onboard intelligent terminal installed on a transport vehicle (such as a truck or drone) to perform the spatial displacement task of the goods and perceive the transportation environment in real time. The transportation intelligent agent is the node with the most frequent data interaction, and its hardware typically includes a main control unit, a positioning module (GPS / BeiDou), a network communication module (4G / 5G), onboard cameras, and various sensors. The transportation intelligent agent receives the second task information from the data interaction platform, parses the destination coordinates, calls the navigation software to plan the route, and controls the vehicle's movement. The transportation intelligent agent uses onboard camera equipment to collect real-time road and traffic information. The transportation intelligent agent analyzes the camera footage through image recognition algorithms to identify road smoothness (e.g., bumps, potholes), road surface type (urban roads, highways, rural roads), and weather conditions (rain, snow, fog). This data is used to assist the platform in assessing transportation risks and the safety status of the goods. The intelligent transportation agent uses cameras to identify traffic signs (speed limits, no-entry zones), traffic density, and congestion conditions. This data is used to adjust navigation route planning in real time and is transmitted back to the platform for global scheduling. The intelligent transportation agent packages and sends the collected road condition information, traffic information, and its own location, speed, fuel consumption, and other status data to the data interaction platform at preset time intervals (e.g., every 30 seconds) or event-triggered mechanisms (e.g., entering an electronic fence). In one embodiment, the intelligent transportation agent has edge computing capabilities, enabling it to perform preliminary analysis of the video streams captured by the cameras locally, uploading only key feature data (e.g., "congestion 500 meters ahead," "pothole warning") to the platform to save network bandwidth and improve response speed.

[0029] After the warehouse node receives the goods, the receiving intelligent agent establishes a first temporary contact with the data interaction platform. The receiving intelligent agent sends the waybill, the actual type of goods received, the actual weight of the goods received, the actual volume of the goods received, and the actual status of the goods received to the data interaction platform.

[0030] In this embodiment, the intelligent terminal devices at the logistics start and end nodes (such as the origin and destination warehouse nodes), and the receiving intelligent agent, typically integrate intelligent gate systems, intelligent weighing platforms, high-precision barcode scanners, and surveillance cameras. Upon detecting the arrival signal of the transport vehicle (e.g., triggered by a Bluetooth beacon or license plate recognition), the receiving intelligent agent automatically establishes a connection with the data interaction platform. This connection is used to request task verification permissions and upload acceptance data. The receiving agent scans the barcode or RFID tag on the outer packaging of the goods to obtain the actual category of goods received; automatically weighs the goods using an intelligent weighbridge to obtain the actual weight of the goods received; obtains the actual volume of the goods received using a laser rangefinder or volume measurement light curtain; and confirms whether the outer packaging of the goods is damaged using a surveillance camera or with manual assistance, recording the actual received status data (e.g., "intact," "damaged," "damp"). The receiving intelligent agent locally compares the collected actual data with the waybill information issued by the platform. If the data deviation is within the allowable range (e.g., weight error <1%), an acceptance report is automatically generated; if the deviation exceeds the limit, an anomaly alarm is triggered. The receiving agent sends the waybill number, actual received goods category, weight, volume, and status data (such as "intact," "damaged," or "damp") to the data exchange platform. This data is used by the platform to update the task status to "completed" and to settle transportation costs and calculate carbon emissions. The data uploaded by the receiving agent serves as the basis for verifying the performance quality of the transportation agent. If the actual received goods status differs from the status in the original task information, the data exchange platform will automatically trigger a claims process or credit deduction mechanism according to preset smart contract rules, achieving a reliable closed loop across the entire logistics chain.

[0031] The requested data includes request data from the shipping agent, request data from the transportation agent, and request data from the receiving agent. The data requested by the shipping agent includes the original task information update request and hidden encoding; The receiving agent requests data including receiving data update requests and hidden codes. The receiving data includes the warehouse node receiving the waybill, the current time, the actual type of goods received, the actual weight of the goods received, the actual volume of the goods received, and the actual status of the goods received.

[0032] The data requested by the receiving agent includes cargo information, services, fees, destination warehouse nodes, and a hidden code. The hidden code is generated by the data interaction platform and embedded in the request data. If the agent can correctly decrypt the hidden code after receiving the request data, it indicates that the agent belongs to the company or a deeply cooperating company, is considered an insider, has a higher level of trust, and will be marked as a second trust zone agent.

[0033] In this embodiment, the hidden code is an identity identifier generated by the data interaction platform using a preset encryption algorithm, used to identify the trust level of the agent during the verification process. The core design concept of the hidden code is that only agents with the correct key pre-configured can successfully decrypt, thereby achieving automatic identification and hierarchical authorization of trust levels. The generation of the hidden code adopts a composite encryption architecture of "multi-layer encryption + dynamic obfuscation", specifically including a basic information encoding layer, a symmetric encryption layer, an asymmetric encapsulation layer, a dynamic obfuscation layer, and an encoding output layer. The basic information coding layer includes: Table 1, Basic Information Table.

[0034] The fields in Table 1 are concatenated in a preset order to form the original byte sequence.

[0035] Symmetric encryption uses a preset master key to encrypt basic information, thereby ensuring data confidentiality.

[0036] The encryption algorithm uses AES-256-GCM (Advanced Encryption Standard - 256-bit - Galois / Counter Mode), which has the following technical features: a key length of 256 bits (32 bytes), providing strong data protection; GCM mode provides both encryption and integrity verification functions, and supports Additional Authentication Data (AAD) to enhance security. The encryption process specifically includes: Obtain the current version's master key MK from the key management module, generate a 12-byte initialization vector IV, use the original byte sequence as plaintext P, set the additional authentication data AAD to the agent number, perform AES-GCM encryption, and output ciphertext C and authentication tag T. The asymmetric encapsulation layer uses an asymmetric encryption algorithm to protect the symmetric key, thereby enabling secure key distribution.

[0037] The encryption algorithm uses RSA-OAEP (RSA-Optimal Asymmetric Encryption Padding), which has the following technical characteristics: the key length is 2048 bits, providing sufficient computational security; the OAEP padding scheme provides semantic security, preventing chosen-plaintext attacks; and the hash algorithm uses SHA-256 as the mask generation function. The packaging process specifically includes: Obtain the public key PK (Public Key) of the current version from the key management module, use the master key MK as the data to be encapsulated, perform RSA-OAEP encryption, and output the encapsulated key EK (Encapsulated Key). Dynamic obfuscation layers rearrange and replace bytes in encrypted data to increase the difficulty of reverse analysis.

[0038] The obfuscation algorithm employs time-seed-based byte rearrangement and replacement, specifically including: using the lower 32 bits of the current timestamp as the obfuscation seed, using the Fisher-Yates shuffle algorithm to generate a byte rearrangement table, performing byte rearrangement on the ciphertext C, and using a preset replacement table S-Box to perform byte replacement, thereby generating an obfuscated data sequence; The encoding output layer converts the processed binary data into a transmittable string format.

[0039] The encoding method uses Base64URL encoding, which has the following characteristics: Using the URL-safe character set (A-Za-z0-9_-), without the padding character "=", it facilitates embedding URL or JSON data, and the encoded data length is approximately 1.33 times that of the original data. The final output hidden encoding format is: HE<Version Number>_<Base64 URL of Encapsulation Key EK>_<Base64 URL of Ciphertext C>_<Base64 URL of Authentication Tag T>; where the version number is used to identify the current key version and algorithm version, and is in the format of "V" followed by two digits, such as "V01" or "V02".

[0040] In this embodiment, the specific content of the requested data is customized according to the type of intelligent agent, following the principle of least privilege, and only includes the minimum dataset required for the intelligent agent of that type to perform the task; the hidden code is an encrypted identifier generated by the data interaction platform, used to identify whether the intelligent agent belongs to the company or a deeply cooperating company, in order to confirm the trust level of the intelligent agent and avoid the risk of information overload and privacy leakage caused by the intelligent agent indiscriminately obtaining all data.

[0041] In step S3, when the intelligent agent receives requested data or needs to actively send a data retrieval request to the data interaction platform, the intelligent agent sends first verification data to the data interaction platform. The data interaction platform verifies the first verification data. After the first verification data passes verification, the data interaction platform establishes a first temporary connection with the intelligent agent and adjusts the original task information into first task information according to the type of the intelligent agent, and sends it to the intelligent agent that established the first temporary connection. The first verification data is the identity authentication request information sent by the intelligent agent to the data interaction platform, which is used to prove the legality of the intelligent agent's identity and apply to establish a communication connection.

[0042] In this embodiment, the unique identification code database is a database that stores the unique identification codes of all registered smart agents, used to support fast matching queries during identity verification. The unique identification code database is maintained by the data interaction platform. When a new smart agent registers, a new record is written; when a smart agent's status changes or it is deregistered, the corresponding record is updated or deleted.

[0043] The first verification data includes the agent's type, vendor ID, product ID, serial number, acceptable tasks, current geographic location, historical carbon emission data, network environment, current time, and hidden code; acceptable tasks are the types of tasks that the agent is currently capable of performing and related parameters; geographic location is the agent's current GPS coordinates; network environment is the agent's current network connection status (such as WiFi, 4G, 5G, etc.) and signal strength; current time is the timestamp of the agent sending the verification request.

[0044] Historical carbon emission data includes urban carbon emission data, suburban carbon emission data, and highway carbon emission data; After receiving the first verification data, the data interaction platform verifies the first verification data, including: After receiving the request data, the data interaction platform verifies the first verification data, including: In this embodiment, the first verification data is verified through a data interaction platform to determine the legitimacy of the intelligent agent's identity and its ability to accept tasks, and to determine its trust level.

[0045] Based on the ID of the agent that sent the first verification data, the manufacturer ID, product ID, and serial number in the first verification data are obtained and matched in the unique identification code database. If the manufacturer ID, product ID, and serial number match, the verification is successful. If one or more of the matched manufacturer ID, product ID, and serial number do not match, the verification fails. The relevant manufacturer ID, product ID, and serial number are retrieved from the unique identification code library, marked as risky, and verification information is sent. After the data interaction platform sends the verification information, it will not send any more information proactively until the verification is successful. If the hidden code is not identified, it is marked as the first trust zone; the smart agents in the first trust zone are usually ordinary registered devices with low permissions (generally social vehicles or temporary transportation capacity).

[0046] If the hidden code is identified, the key is retrieved to decrypt the hidden code, and the decrypted code is obtained. The first verification data and the decrypted code are sent to the data interaction platform, which marks it as a second trust zone. The intelligent agent in the second trust zone is usually a deeply authenticated device or a partner device, which enjoys higher privileges.

[0047] The data interaction platform establishes an initial temporary connection with the verified agent.

[0048] In this embodiment, the first temporary connection is the initial communication session established between the data interaction platform and the verified agent, used to support basic data exchange functions. The first temporary connection is established immediately after successful verification and is time-sensitive, remaining valid only until the agent completes task claiming. The first temporary connection can only access data contained in the first task information. The first task information is task data generated by the data interaction platform after de-identifying and adjusting the original task information according to the agent's trust domain, used to achieve role-based minimum data distribution.

[0049] Adjust the original task information to the first task information, including: The data interaction platform adjusts the original task into primary task information based on the agent's trust domain, including: The data interaction platform extracts the origin and destination warehouse nodes from the original task information to plan navigation routes, calculates the length of urban roads, suburban roads and highways respectively, and calculates the total carbon emissions of each transportation agent according to the urban carbon emission data, suburban carbon emission data and highway carbon emission data, and sorts the total carbon emissions in descending order.

[0050] In this embodiment, after receiving the first verification data sent by the intelligent agent, the data interaction platform first performs identity matching in the unique identification code library. After successful verification, the trust zone is marked according to the hidden code identification result. Then, a first temporary connection is established, and desensitized first task information is generated according to the intelligent agent type and trust domain and sent to the intelligent agent. This realizes an integrated verification-authorization-distribution process, ensuring that each intelligent agent can only obtain the minimum dataset necessary for its business, effectively preventing the excessive exposure of sensitive information. Total carbon emissions = urban carbon emission data x urban road length + suburban carbon emission data x suburban road length + highway carbon emission data x highway road length.

[0051] In step S4, the agent receives the first task information and sends the second verification data and task claiming status to the data interaction platform. The second verification data is the enhanced identity and status information submitted by the agent when claiming the task, which is used to support data interaction during task execution.

[0052] The data interaction platform adjusts the original task information into primary task information based on the agent's trust domain, including: The data interaction platform extracts the cargo category, cargo weight, cargo volume, cargo status, service content, arrival time, cargo value, and risk consequences from the original task information as the first task information. The data interaction platform then sends the first task information to the transportation intelligent agent that has established the first temporary connection. The transportation agent receives the first task information and sends the task acceptance status and second verification data to the data interaction platform. The task acceptance status includes accepting the task and not accepting the task. If the transportation agent does not send the task acceptance status and second verification data within a unit of time (e.g., 30 seconds), the data interaction platform regards the transportation agent as not accepting the task and terminates the first temporary connection with the transportation agent that has not accepted the task. In this embodiment, after receiving the first task information, the transportation agent evaluates the task based on its remaining transportation weight, remaining transportation space, and task route. If it decides to claim the task, it generates a message containing second verification data and a claim declaration and sends it to the data interaction platform. The proactive claiming and status synchronization of tasks by the transportation agent provides data support for subsequent task allocation optimization. If no transportation agent proactively claims the task, a new transportation agent is scheduled to complete it, or the task is reassigned during the next task allocation, or the remaining tasks from the previous day are prioritized for allocation on the next day, ensuring timely task allocation.

[0053] The second task information includes waybill, cargo category, cargo weight, cargo volume, cargo status, service content, cargo value, risk consequences, mode of transport, origin warehouse node address, destination warehouse node address, and task confirmation notification; the second task information is a complete task instruction sent by the data interaction platform to the transportation agent that is finally determined to execute the task, which is used to guide the transportation agent to execute the specific task.

[0054] The second set of verification data includes the current geographical location, current network environment, current time, navigation route planning, actual carbon emissions, and real-time road condition and traffic information.

[0055] In this embodiment, the current geographical location is the real-time GPS coordinates of the transportation agent; the current network environment is the network connection status of the transportation agent; the navigation path planning is the navigation path planning for the transportation agent's planned task execution; the current time is the timestamp of data transmission; road condition information is road congestion, construction, and other conditions collected by cameras; traffic information is traffic control information such as traffic lights and traffic restrictions. Accepting a task means the transportation agent explicitly expresses its desire to execute the task; not accepting a task means the transportation agent indicates that it is unable or unwilling to execute the task.

[0056] The data interaction platform receives task claiming status and second verification data. It selects the current geographical location from the transportation agents that have claimed the task, calculates the distance between the current geographical location of each transportation agent that has claimed the task and the address of the origin warehouse node, selects the transportation agents whose distance is less than a predetermined distance threshold, and publishes the second task information. The transportation agents that receive the second task information upgrade from the first temporary connection to the second temporary connection. In this embodiment, the data interaction platform extracts the origin and destination warehouse nodes from the original task information, calls the map service interface to plan the navigation route, and calculates the lengths of urban roads, suburban roads, and highways respectively. Based on the historical carbon emission data reported by each transportation agent (including unit emissions under urban, suburban, and highway conditions), the platform calculates the estimated total carbon emissions for the transportation agent performing this task. The data interaction platform selects transportation agents whose current geographical location is less than a predetermined distance threshold (e.g., 5 kilometers) from the origin warehouse node as a candidate set. Within the candidate set, the transportation agent with the lowest estimated total carbon emissions is prioritized as the target transportation agent. The data interaction platform sends the second task information to the target agent. Simultaneously, the platform upgrades the agent's connection status from "first temporary connection" to "second temporary connection." The second temporary connection has a longer lifespan and higher data access permissions, marking the transportation agent's formal entry into the task execution phase.

[0057] The second temporary connection is as follows: The transport agent sends second verification data to the data interaction platform at regular intervals. The data interaction platform also shares the real-time locations of the nearest X transport agents, as well as real-time road condition and traffic information. The second temporary connection is a high-level communication session established between the data interaction platform and the agent that formally claimed the task, used to support richer data exchange and collaborative functions.

[0058] In this embodiment, the transport agent sends second verification data to the data interaction platform at intervals to achieve continuous identity verification and status synchronization. The data interaction platform shares the real-time locations of the nearest X transport agents, as well as real-time road condition and traffic information, with the transport agent to achieve collaborative operation and route optimization. The unit time is a preset sending interval, such as 30 seconds or 1 minute; X is a preset threshold for the number of nearby transport agents, such as 5 or 10. Nearby transport agents can provide assistance when the current transport agent encounters a fault. After receiving the task acceptance status from the transport agent, the data interaction platform first terminates the first temporary connection with unacknowledged transport agents, releasing communication resources. Then, it performs distance calculations and prioritizes the transport agents that have accepted tasks based on total carbon emissions to determine the final task executor. Finally, it sends the second task information to the selected transport agent and upgrades its connection from the first temporary connection to the second temporary connection, ensuring efficient task allocation and real-time monitoring of the transport agent during subsequent task execution. The second temporary connection has a longer lifespan and higher data access permissions, marking the agent's formal entry into the task execution phase.

[0059] The transportation agent adjusts its navigation route planning in real time based on the latest road condition and traffic information, and sends the adjusted navigation route planning and actual carbon emissions to the data exchange platform. The data exchange platform records the actual carbon emissions of the transportation agent in real time and calculates the estimated carbon emissions of the adjusted navigation route planning.

[0060] In this embodiment, under the second temporary connection state, the data interaction platform pushes real-time status data of the X (e.g., 5) other transportation agents closest to the current transportation agent to the current agent according to a preset coordination strategy. This data specifically includes: The location of neighboring agents is used to determine the vehicle density of the road segment ahead. Road condition information collected by neighboring agents includes road surface damage, road construction, and accident hotspots. Traffic information collected by neighboring agents includes average vehicle speed, congestion length, and traffic light waiting times. The transportation agent utilizes an edge computing module, combining its original navigation path with received collaborative data, to reconstruct the path cost function, the mathematical expression of which is: in, Represents the path cost function. This represents the time cost (the travel time predicted based on real-time adjustments to the navigation route planning). This indicates the cost of fuel or electricity (derived from historical average electricity or fuel consumption per kilometer). The risk costs (vehicle wear and cargo safety risks assessed based on road condition information), represented by α, β, and γ, are weighted coefficients. When a neighboring agent reports severe congestion ahead (leading to a surge in time costs) or poor road conditions (increasing risk costs), the transportation agent automatically triggers a replanning algorithm (such as an improved Dijkstra algorithm) to calculate the cost of an alternative path. If the alternative path has a lower cost, an adjusted navigation path plan is generated. The transportation agent reads the engine's instantaneous fuel consumption rate, engine speed, vehicle speed, and air intake volume in real time via the OBD (On-Board Diagnostics) interface or CAN bus. The data exchange platform compares the calculated projected carbon emissions with the actual carbon emission trends reported by the transportation agent. If the deviation is within the allowable range (e.g., ±10%), the path adjustment is deemed reasonable, and the projected total emissions for the task are updated. If the deviation is too large (such as actual emissions far exceeding expected emissions), the platform will trigger an anomaly warning to determine whether there is vehicle malfunction, overloading, or data falsification, and notify the dynamic control module to intervene (specifically through manual telephone or video follow-up). Through the second temporary connection, real-time monitoring of the transportation intelligent agent during transportation and the ability to respond in real time according to actual road conditions are ensured, while also ensuring the data interaction platform's accurate control of global carbon emission data, thus realizing closed-loop management of green logistics.

[0061] In step S5, the data interaction platform generates a task list based on the task claiming status and the second verification data. It then terminates the first temporary connection with unclaimed task agents, sends second task information to the agents that have claimed tasks, and upgrades the first temporary connection to a second temporary connection. The agent receives the second task information. The task list is a task allocation scheme generated by the data interaction platform based on the task claiming status of the agents, used to specify which agent will execute each task. In one embodiment, the task list is generated by the data interaction platform after optimization decisions based on the second verification data of the agents that have claimed tasks, comprehensively considering factors such as distance, carrying capacity, and route matching, to select the optimal agent to execute the task. The data interaction platform extracts the current geographical location from the agents that have claimed tasks, calculates the distance between the current geographical location of each agent and the originating warehouse node address, and prioritizes the agent with the shortest distance as the task executor. The distance calculation uses GPS coordinate straight-line distance or path planning distance algorithms.

[0062] After the data interaction platform establishes the first temporary connection with the intelligent agent, the intelligent agent sends a data retrieval request to the data interaction platform; The agent lacks a second trust zone marker; If the agent is a shipping agent, it can retrieve the historical original task information recorded by the shipping agent; If the agent is a transportation agent, it can retrieve the second task information from the transportation agent's history; If the agent is a receiving agent, it can retrieve the historical receiving data of that receiving agent; If an agent has a second trust zone marker, it not only has the authority to access data without a second trust zone marker, but can also retrieve all original task information related to the second task information in a single instance, excluding the party's information, fees, and contact numbers. If the agent is a transportation agent, it can also retrieve the remaining transportation weight and remaining transportation space of the nearest X transportation agents. By obtaining the remaining capacity (remaining load capacity, remaining space) of the transportation agents, the current transportation agent can make intelligent carpooling decisions or coordinate transportation capacity. For example, if it finds that a nearby transportation agent has spare space and is on the same route, it can apply to the data exchange platform to transfer some goods, achieving dynamic capacity optimization. In this embodiment, the distinction between the first trust zone and the second trust zone is the core defense line for ensuring privacy, security, and collaborative efficiency in cross-domain data exchange. If an agent does not have a second trust zone identifier (i.e., the agent has only passed basic identity verification, is marked as belonging to the first trust zone, is not an ally, and has low trustworthiness), the agent is restricted to accessing only its own historical data, and cross-entity access is prohibited. An agent possessing a second trust zone identifier (i.e., the agent has passed hidden code decryption verification, is an ally, and has high trustworthiness)...

[0063] Step S6: The data interaction platform dynamically manages the first temporary connection and the second temporary connection based on the first verification data, request data, and second verification data sent by the intelligent agent, including: If a single agent sends M data retrieval requests to the data interaction platform during non-working hours, or sends N data retrieval requests to the data interaction platform within a short period of time, and if the agent and the data interaction platform are in the first temporary connection, the data interaction platform will disconnect the first temporary connection with the agent, mark the agent as having a risk, and send verification information. If the intelligent agent and the data interaction platform are connected via a second temporary connection, the data interaction platform will disconnect the second temporary connection with the intelligent agent and send verification information. After receiving the verification information, the intelligent agent will send verification materials, including the manufacturer's license, product ID image, and product photos. Before the verification is passed, the data interaction platform will not send any information to the intelligent agent. After the verification is passed, the data interaction platform and the intelligent agent will send information normally. If the agent fails to send the second verification data to the data interaction platform at regular intervals under good network conditions, or if the agent's current geographical location exceeds a predetermined distance threshold from the predetermined route, the data interaction platform will disconnect the second temporary connection with the agent, mark the agent as risky, and send verification information. In this embodiment, dynamic control runs through the interaction between the intelligent agent and the data interaction platform. By analyzing the intelligent agent's request frequency, location trajectory, connection status and other behavioral characteristics, security risks are assessed in real time, and response actions such as connection disconnection, risk marking and manual verification are automatically triggered.

[0064] Dynamic management and control specifically includes the following anomaly detection and handling mechanisms: First type of exception: Frequent request exception; Frequent request anomalies are suspicious behaviors in which an agent frequently sends data retrieval requests to the data interaction platform within a short period of time or outside of working hours. This may indicate that the agent has been hijacked or has the intention to steal data.

[0065] The detection rule is that a single agent sends M data retrieval requests to the data interaction platform during non-working hours, or a single agent sends N data retrieval requests to the data interaction platform within a short period of time. Here, "non-working hours" refers to the time range outside the agent's normal working hours, such as nighttime or rest days; M is the threshold for the number of requests during non-working hours, such as 5 or 10 times; "short period of time" refers to a preset time window, such as 1 minute or 5 minutes; and N is the threshold for the number of requests within a short period of time, such as 10 or 20 times.

[0066] The handling measures vary depending on the connection type: If the agent and the data interaction platform have only a first temporary connection, the data interaction platform will directly sever the first temporary connection with the agent, mark the agent as risky, and send verification information. This approach is relatively strict because the first temporary connection corresponds to a basic trust level, and the risk tolerance for frequent requests is low.

[0067] If the agent and the data interaction platform are connected via a second temporary connection, the data interaction platform will sever the connection and send verification information. Upon receiving the verification information, the agent must submit verification materials for manual review. These materials include the manufacturer's license, product ID image, and product photos. Before successful verification, the data interaction platform will not send any further information to the agent, effectively blocking communication; after successful verification, normal communication between the data interaction platform and the agent will resume. This approach provides agents in the second trust zone with greater fault tolerance, but still requires manual verification to confirm their identity.

[0068] Second exception type: Connection interruption exception; A connection interruption exception is the behavior of an agent failing to send the second verification data on time in the second temporary connection state, which may indicate that the agent has malfunctioned, the network has been interrupted, or it has been maliciously controlled.

[0069] The detection rule is that the agent does not send second verification data to the data interaction platform at unit intervals when the current network environment is good. A good network environment means that the agent's network signal strength is higher than a preset threshold and there is no network failure; unit interval is a preset heartbeat interval.

[0070] The data interaction platform disconnects the second temporary connection with the agent, marks the agent as a risk, and sends verification information.

[0071] Third anomaly type: Route deviation anomaly; Route deviation anomaly is a behavior in which the actual position of the agent deviates significantly from the planned route, which may indicate that the agent has been hijacked, navigation failure, or abnormal stop.

[0072] The detection rule is that the agent's current geographical location is more than a predetermined distance threshold from the planned route. The distance threshold is the maximum allowable deviation distance, such as 500 meters or 1 kilometer.

[0073] The solution is for the data interaction platform to sever the second temporary connection with the agent, mark the agent as a risk, and send verification information.

[0074] In practice, the data interaction platform continuously monitors the behavioral data of all intelligent agents. When any of the above-mentioned abnormal behaviors are detected, the corresponding processing flow is immediately triggered, including disconnecting the connection, marking the risk, and sending verification, thus achieving rapid response and isolation of risks. This mechanism constructs a comprehensive security protection network that can effectively identify and block various abnormal behaviors, ensuring the security of data exchange.

[0075] Fourth type of anomaly: Carbon emission exceeding standards; The detection rule is that the real-time carbon emissions of the transportation agent during transportation exceed the carbon emission target value preset in the waybill. The handling measures are that the data interaction platform pushes carbon emission reduction scheduling suggestions to the agent. If the emissions continue to exceed the target, the trust level of the agent is downgraded, and its priority in subsequent task allocation is reduced. If the emissions exceed the target severely, the second temporary connection is directly cut off, and the agent is marked as a high-risk agent.

[0076] Fifth type of anomaly: Cargo status exceeding limits; The detection rule is that the cargo compartment temperature, humidity, vibration, and pressure data reported by the onboard sensors of the transport agent exceed the preset safety thresholds for the corresponding cargo category (bulk goods, cold chain, hazardous chemicals). The handling measure is that the data interaction platform simultaneously pushes a cargo safety warning to both the shipping agent and the transport agent, immediately disconnects the current second temporary connection, and marks the agent as having a high cargo risk.

[0077] Example 2 is another embodiment of the present invention. This embodiment is based on the previous embodiment. The difference between this embodiment and the previous embodiment is that the following uses a complete logistics and distribution process as an example to illustrate the specific implementation process of the present invention.

[0078] User A places an order through a logistics company to transport a batch of goods from Haidian District, Beijing to Pudong New Area, Shanghai. The logistics company uses the multi-agent cross-domain data exchange dynamic authorization method provided in this invention for task management.

[0079] Based on the origin and destination warehouse nodes, the data interaction platform determines the transportation route as follows: from the Haidian warehouse node in Beijing to the Pudong warehouse node in Shanghai. The data interaction platform generates transportation agent request data, including cargo information, the addresses of the Haidian and Pudong warehouse nodes in Beijing and a hidden code, and sends it to all online transportation agents.

[0080] After receiving the request data, the transportation agent T001 sends the first verification data to the data interaction platform: type is transportation agent, vendor ID is V001, product ID is P001, number is T001, and acceptable tasks include transportation tasks, from Beijing to Shanghai, remaining transportation weight is 500kg, and remaining transportation space is 10m. 3The current location is near the Chaoyang warehouse node in Beijing, the network environment is 4G with good performance, the current time is 2024-01-15 10:00:00, and the hidden code is an encrypted string.

[0081] After receiving the initial verification data, the data exchange platform matches and verifies it in the unique identifier database: if the manufacturer ID, product ID, and serial number match, the verification is successful. The data exchange platform identifies the hidden code, retrieves the key, and successfully decrypts it, marking T001 as the second trusted zone. The data exchange platform establishes the first temporary connection with T001 and sends the first task information: cargo information, Beijing Haidian warehouse node address, and Shanghai Pudong warehouse node address.

[0082] After assessing its own status, the transportation agent T001 decides to claim the task and sends the second verification data and task claim status to the data interaction platform: current GPS location is (39.92, 116.46), network environment is good 4G, the planned route is Beijing-Shanghai Expressway, the current time is 2024-01-15 10:05:00, road condition information is Beijing-Shanghai Expressway is smooth, and the claimed task is BJ20240115001.

[0083] Meanwhile, the transport agent T002 also sent a claim request. Its current GPS location is (40.01, 116.40), which is far from the Beijing Haidian warehouse node.

[0084] The data exchange platform calculates the distances: T001 is approximately 8 kilometers from the Beijing Haidian warehouse node, and T002 is approximately 15 kilometers from the Beijing Haidian warehouse node. The data exchange platform prioritizes executing the task for T001, which has the shorter distance, generates a task list, and sends the second task information to T001: cargo information, the address of the Beijing Haidian warehouse node, the address of the Shanghai Pudong warehouse node, and a task confirmation notification. The data exchange platform upgrades the connection between T001 from the first temporary connection to the second temporary connection and terminates the first temporary connection with T002.

[0085] During transportation, the transportation agent T001 sends second verification data to the data interaction platform every minute. The data interaction platform shares the real-time locations of the five nearest transportation agents and road condition information along the route with T001, supporting T001 in route optimization.

[0086] If T001 deviates from the planned route by more than 1 kilometer during transportation, the data exchange platform will detect the abnormal route deviation, automatically disconnect the second temporary connection, mark T001 as having a risk, and send verification information to T001 requesting verification materials.

[0087] If T001 sends 8 data retrieval requests consecutively during non-working hours at night, the data interaction platform will detect the frequent request anomaly. Since T001 is currently in the second temporary connection, the data interaction platform will disconnect the connection and send verification information, requiring T001 to provide the manufacturer's license, product ID image, and product photo for manual verification.

[0088] After the goods arrive at the Shanghai Pudong warehouse node, the receiving end intelligent agent sends the waybill, the actual type of goods received, the actual weight of the goods received, the actual volume of the goods received, and the actual status of the goods received to the data interaction platform.

[0089] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0090] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A dynamic authorization method for cross-domain data exchange based on zero-trust multi-agent systems, characterized in that, include: Step S1: Configure a unique identification code for each intelligent agent using the data interaction platform; Step S2: Obtain the original task information; the data interaction platform sends request data to each intelligent agent. Step S3: When the intelligent agent receives the requested data or needs to actively send a data retrieval request to the data interaction platform, the intelligent agent sends the first verification data to the data interaction platform. The data interaction platform verifies the first verification data. After the first verification data passes the verification, the data interaction platform establishes a first temporary connection with the intelligent agent and adjusts the original task information into the first task information according to the type of the intelligent agent, and sends it to the intelligent agent that established the first temporary connection. Step S4: The intelligent agent receives the first task information and sends the second verification data and task claim status to the data interaction platform. Step S5: The data interaction platform generates a task list based on the task claiming status and the second verification data, terminates the first temporary connection with the unclaimed task agent, sends the second task information to the agent that has claimed the task, and upgrades the first temporary connection to the second temporary connection. The agent receives the second task information. Step S6: The data interaction platform performs dynamic management and control of the first temporary connection and the second temporary connection based on the first verification data, request data and the second verification data sent by the intelligent agent.

2. The dynamic authorization method for cross-domain data exchange based on zero trust of multiple agents as described in claim 1, characterized in that: The unique identification code is bound to the agent's type, manufacturer ID, product ID, serial number, and trust domain. All the unique identification codes of the agents are then organized to generate a unique identification code library.

3. The dynamic authorization method for cross-domain data exchange based on zero trust of multiple agents as described in claim 2, characterized in that: The original task information includes waybill, party information, cargo category, cargo weight, cargo volume, cargo status, service content, cargo value, risk consequences, mode of transport, cost, contact number, arrival time, origin warehouse node and destination warehouse node; The intelligent agent includes a shipping intelligent agent, a transportation intelligent agent, and a receiving intelligent agent; The shipping agent establishes a first temporary contact with the data interaction platform, and the shipping agent sends the original task information to the data interaction platform. The intelligent transportation agent is used to transport goods from the origin storage node to the destination storage node, and utilizes real-time road condition and traffic information collected by cameras. After the warehousing node receives the goods, the receiving end intelligent agent establishes a first temporary contact with the data interaction platform. The receiving end intelligent agent sends the waybill, the actual type of goods received, the actual weight of the goods received, the actual volume of the goods received, and the actual status of the goods received to the data interaction platform.

4. The dynamic authorization method for cross-domain data exchange based on zero trust of multiple agents as described in claim 3, characterized in that: The request data includes request data from the shipping agent, request data from the transportation agent, and request data from the receiving agent. The data requested by the shipping agent includes the original task information update request and hidden encoding; The data requested by the transport agent includes a request to update transportable cargo data and a hidden code. The transportable cargo data includes the transport agent's current address information, current status, current time, remaining transport weight, remaining transport space, and hidden code. The receiving end intelligent agent requests data including receiving data update requests and hidden codes. The receiving data includes the warehouse node receiving the waybill, the current time, the actual type of goods received, the actual weight of the goods received, the actual volume of the goods received, and the actual status of the goods received.

5. The dynamic authorization method for cross-domain data exchange based on zero trust of multiple agents as described in claim 4, characterized in that: The first verification data includes the agent's type, manufacturer ID, product ID, serial number, acceptable tasks, current geographical location, historical carbon emission data, network environment, current time, and hidden code; The historical carbon emission data includes urban carbon emission data, suburban carbon emission data, and highway carbon emission data; After receiving the first verification data, the data interaction platform verifies the first verification data by including: Based on the ID of the agent that sent the first verification data, the manufacturer ID, product ID, and serial number in the first verification data are obtained and matched in the unique identification code database. If the manufacturer ID, product ID, and serial number match, the verification is successful. If one or more of the matched manufacturer ID, product ID, and serial number do not match, the verification fails. The relevant manufacturer ID, product ID, and serial number are retrieved from the unique identification code library, marked as risky, and verification information is sent. After the data interaction platform sends the verification information, it will not send any more information proactively until the verification is successful. If the hidden code is not identified, mark it as the first trusted zone; If the hidden code is identified, the key is retrieved to decrypt the hidden code, and the decrypted code is obtained. The first verification data and the decrypted code are sent to the data interaction platform, and the data interaction platform marks the second trust zone for it. The data interaction platform establishes a first temporary connection with the verified agent.

6. The dynamic authorization method for cross-domain data exchange based on zero trust of multiple agents as described in claim 5, characterized in that: The data interaction platform extracts the origin and destination warehouse nodes from the original task information to plan navigation routes, calculates the length of urban roads, suburban roads and highways respectively, and calculates the total carbon emissions of each transportation agent according to the urban carbon emission data, suburban carbon emission data and highway carbon emission data, and sorts the total carbon emissions in descending order.

7. The dynamic authorization method for cross-domain data exchange based on zero trust of multiple agents as described in claim 6, characterized in that: Adjust the original task information to the first task information, including: The data interaction platform adjusts the original task information to first task information based on the agent's trust domain, including: The data interaction platform extracts the cargo category, cargo weight, cargo volume, cargo status, service content, arrival time, cargo value, and risk consequences from the original task information as the first task information. The data interaction platform then sends the first task information to the transportation intelligent agent that has established the first temporary connection. The transportation agent receives the first task information and sends the task acceptance status and the second verification data to the data interaction platform. The task acceptance status includes accepting the task and not accepting the task. If the transportation agent does not send the task acceptance status and the second verification data within a unit of time, the data interaction platform regards the transportation agent as not accepting the task and terminates the first temporary connection with the transportation agent that has not accepted the task. The second verification data includes the current geographical location, current network environment, current time, navigation route planning and actual carbon emissions, as well as real-time road condition information and traffic information; The data interaction platform receives task claiming status and second verification data, selects the current geographical location from the transportation agents claiming tasks, calculates the distance between the current geographical location of each transportation agent claiming a task and the address of the origin warehouse node, selects the transportation agents whose distance is less than a predetermined distance threshold, and publishes the second task information. The transportation agents that receive the second task information upgrade from the first temporary connection to the second temporary connection. The second task information includes waybill, cargo category, cargo weight, cargo volume, cargo status, service content, cargo value, risk consequences, mode of transport, origin warehouse node address, destination warehouse node address, and task confirmation notification.

8. The dynamic authorization method for cross-domain data exchange based on zero trust of multiple agents as described in claim 7, characterized in that: The second temporary connection includes: The transport agent sends second verification data to the data interaction platform at unit time intervals, and the data interaction platform shares with the transport agent the real-time locations of the nearest X transport agents, as well as the real-time collected road condition information and traffic information; The transportation agent adjusts its navigation route planning in real time based on the latest road condition and traffic information, and sends the adjusted navigation route planning and actual carbon emissions to the data interaction platform. The data interaction platform records the actual carbon emissions of the transportation agent in real time and calculates the estimated carbon emissions of the adjusted navigation route planning.

9. The dynamic authorization method for cross-domain data exchange based on zero trust of multiple agents as described in claim 8, characterized in that: After the data interaction platform establishes a first temporary connection with the intelligent agent, the intelligent agent sends a data retrieval request to the data interaction platform. The agent has no second trust zone marker; If the agent is a shipping agent, it can retrieve the historical original task information recorded by the shipping agent. If the agent is a transportation agent, it can retrieve the historical second task information of the transportation agent; If the intelligent agent is a receiving end intelligent agent, it can retrieve the historical receiving data of the receiving end intelligent agent; If the intelligent agent has a second trust zone marker, it can, in addition to having the authority to access data information without a second trust zone marker, also access all original task information related to the second task information in a single instance, except for the party's information, fees, and contact numbers. If the intelligent agent is a transportation intelligent agent, it can also access the remaining transportation weight and remaining transportation space of the nearest X transportation intelligent agents.

10. The dynamic authorization method for cross-domain data exchange based on zero trust of multiple agents as described in claim 9, characterized in that: Step S6 specifically includes: If a single agent sends M data retrieval requests to the data interaction platform during non-working hours, or sends N data retrieval requests to the data interaction platform within a short period of time, and if the agent and the data interaction platform are in a first temporary connection, the data interaction platform will disconnect the first temporary connection with the agent, mark the agent as having a risk, and send verification information. If the intelligent agent and the data interaction platform are connected in a second temporary manner, the data interaction platform will disconnect the second temporary connection with the intelligent agent and send verification information. After receiving the verification information, the intelligent agent will send verification materials, including the manufacturer's license, product ID image and product photos. Before the verification is passed, the data interaction platform will no longer send any information to the intelligent agent. After the verification is passed, the data interaction platform and the intelligent agent will send information normally. If the agent fails to send the second verification data to the data interaction platform at regular intervals under good network conditions, or if the agent's current geographical location is more than a predetermined distance threshold from the predetermined route, the data interaction platform will disconnect the second temporary connection with the agent, mark the agent as a risk, and send verification information.