Intelligent and reliable collaborative method for multimodal transport one-box system based on artificial intelligence technology

Abstract

The present application relates to the technical field of multimodal transport, and particularly relates to a multimodal transport one-box intelligent trusted collaboration method based on artificial intelligence technology, which comprises the following steps: S1: standardized JSON data packets are obtained from transport data according to agreement rules, triple encryption is performed to obtain encrypted data packets, and the encrypted data packets are broadcast to a block chain; S2: a consensus verification is performed on Internet of Things equipment data and electronic waybill key fields by a verification node, and if the verification is passed, the data is written into the block chain; S3: a digital twin system fuses multi-source data, and extracts and calculates data, and an artificial intelligence calculates an optimal loading and unloading sequence and a transport path according to the calculated data; and S4: longitude and latitude data is uploaded to a block chain node through a Beidou positioning terminal and a 5G / LoRa gateway along a route, and a transport stage anomaly is automatically identified based on a geographic fence technology. The present application unifies data rules, realizes efficient connection of cross-transport modes, prevents manual input errors or irregular operations, reduces cargo loss risks and improves transfer efficiency through dynamic planning of a route, and reduces logistics time and cost consumption.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multimodal transport, in particular to a multimodal transport one-box intelligent trusted collaboration method, system, device and storage medium based on artificial intelligence technology. BACKGROUND

[0002] Multimodal transport is a comprehensive transportation strategy that integrates multiple transportation modes. By combining road, rail, air and sea transportation, it provides efficient and flexible door-to-door services, optimizes transportation efficiency, reduces transit time and cost, promotes green logistics, and reduces energy consumption and environmental pollution.

[0003] In the existing multimodal transport container transport process, the cargo information of the consignor, logistics enterprise and different transport subjects is not shared and connected, and the cargo needs to be reloaded multiple times between different transportation modes (such as sea transportation, railway and highway). The multiple loading and unloading of goods increases the risk of damage, labor cost and time cost; since the goods are not transported by using containers uniformly, when switching between different transportation modes, the goods often need to be re-packed or reinforced, making it difficult to realize the standardization and scaling of transportation, thereby affecting the transportation efficiency and service quality, which not only increases the complexity of operation, but also increases the cost; in addition, part of the logistics hub facilities are not matched, and the special multimodal transport station is insufficient, resulting in low container loading and unloading efficiency, frequent container unloading phenomenon, and lagging behind in the construction of cross-regional channels, especially the lack of connection between inland areas and coastal ports, which limits the coverage of "one container to the end"; in addition, the intelligent level of the logistics hub is not high, and the resource scheduling relies too much on manual experience, the allocation of cranes and trucks has strong subjectivity, and in the face of sudden weather and equipment failure, the dynamic adjustment capability is insufficient, it is difficult to quickly optimize the scheduling scheme, the fault tolerance is low, and there is a lack of global risk prediction mechanism, thereby causing the transportation link to be not smooth, the cost continues to rise, and the process of "multimodal transport one-box system" full efficient operation is restricted. SUMMARY

[0004] The purpose of the present application is to solve the problems in the prior art, and to provide a multimodal transport one-box intelligent trusted collaboration method based on artificial intelligence technology, comprising the following steps:

[0005] S1: converting the transportation data into a standardized JSON format according to the agreed rules to obtain a standardized JSON data package, and performing triple encryption on the standardized JSON data package to obtain an encrypted data package and broadcast it to the blockchain;

[0006] S2: Real-time collection of Internet of Things device data and electronic bill of lading key fields including bill of lading number, cargo category, and sender and receiver, consensus verification of the Internet of Things device data and the electronic bill of lading key fields by the verification node in the blockchain network, and writing into the blockchain if the verification is passed, and warning and data correction if there is data conflict;

[0007] S3: The digital twin system fuses multi-source data including physical device data, business system data and external environment data, extracts calculation data according to the multi-source data, and calculates the optimal loading and unloading sequence and transportation path according to the calculation data, generates operation instructions and issues them to the on-site automation equipment;

[0008] S4: Through the Beidou positioning terminal built in the container and the 5G / LoRa gateway along the way, the latitude and longitude data is uploaded to the blockchain node every minute, and the transportation stage anomaly is automatically identified based on the geographic fence technology. When the physical device data exceeds the smart contract threshold, an alarm is pushed to the consignor and the insurance company.

[0009] Preferably, in step S1, the transportation data is converted into a standardized JSON format according to the agreed rules to obtain a standardized JSON data package, which further includes:

[0010] The transportation data is input through a data conversion gateway, including structured data and unstructured data, wherein the structured data includes EDI messages, XML transportation documents, and CSV formats, and the unstructured data includes paper document scans, key fields extracted by ORC engines and NLP models;

[0011] The transportation data is output as a standardized JSON data package according to the agreed rules, wherein the agreed rules include defining a container structured data model and a space-time data rule, the container structured data model includes a unique identifier, physical parameters including size, weight, and box type, and safety inspection status including refrigerated box temperature control range, dangerous goods UN number, and inspection status, and the space-time data rule is that the timestamp is forced to use the ISO 8601 extended format for device end encryption, and the location data is fused with the WGS84 coordinate system and the UN / LOCODE port code.

[0012] Preferably, the standardized JSON data package is triple-encrypted to obtain an encrypted data package and broadcasted to the blockchain, further including:

[0013] S11: Symmetrically encrypting the original data of the standardized JSON data package using device end AES-256 to obtain symmetrically encrypted data;

[0014] S12: Dynamically generating a session key through the TLS protocol, and encrypting the symmetrically encrypted data according to the session key to obtain asymmetrically encrypted data;

[0015] S13: The asymmetric key is encrypted again by a cloud AES-256 key, and the encrypted data is subjected to SHA256 hash operation to generate an encrypted data packet and broadcast to the blockchain.

[0016] Preferably, in step S2, the verification nodes in the blockchain network verify the consensus of the Internet of Things device data and the electronic waybill key fields according to the encrypted data packet, further comprising:

[0017] The verification nodes in the blockchain network verify the consensus of the weight data obtained by the intelligent weighbridge, the declared weight in the electronic waybill, the container number and ISO check code extracted by the container number identification camera, the temperature and humidity, and the cold chain agreement, wherein the verification nodes include the nodes of the current port, the next segment carrier, and the customs supervision party;

[0018] If more than half of the verification nodes confirm that the data is true and valid, the Internet of Things device data and the electronic waybill key fields are written into a new block to form an unchangeable electronic file. If any verification node detects that the field is missing or the value range is out of bounds, it is rejected.

[0019] Preferably, in step S3, the digital twin system fuses multi-source data including physical device data, business system data, and external environment data, extracts calculation data according to the multi-source data, and the artificial intelligence calculates the optimal loading and unloading sequence and transportation path according to the calculation data, further comprising:

[0020] S31: The digital twin system fuses multi-source data including device sensor data, AI visual recognition of container position and barcode information, TOS port loading and unloading task list and transportation timetable, real-time weather and traffic conditions, extracts calculation data including device load rate, container urgency, and path congestion coefficient according to the multi-source data;

[0021] S32: Multi-agent deep deterministic policy gradient MADDPG assigns container hoisting priority to the rail-mounted crane and plans a dynamic path for the AGV according to the calculation data.

[0022] Preferably, in step S32, the multi-agent deep deterministic policy gradient MADDPG assigns container hoisting priority to the rail-mounted crane and plans a dynamic path for the AGV according to the calculation data, further comprising:

[0023] S321: discretize the digital twin yard into a two-dimensional grid map to obtain a digital twin yard topology map, and the grid attributes include static obstacles and dynamic forbidden areas, wherein the static obstacles include buildings, fixed equipment areas, and the dynamic forbidden areas include the radius range of the track crane operation and the temporary container stacking area;

[0024] S322: based on the digital twin yard topology map, an initial shortest path is generated using artificial intelligence technology, and whether there is a path conflict in the future time interval is predicted by the space-time grid method, if there is a path conflict, the AGV of the high emergency container is assigned a path priority, and other AGVs are triggered to locally detour to realize dynamic real-time adjustment;

[0025] Based on the digital twin yard topology map, an initial shortest path is generated using artificial intelligence technology, including:

[0026] The open list stores the starting point S (f=0), and the closed list is used to store the explored nodes and is initially empty, the cost function g(n) is defined as the actual movement cost from the starting point to the current node n, the adjacent grid distance is 1, and the diagonal grid distance is The cost function h(n) is defined as the Manhattan distance estimation cost of the current node n to the terminal G, and the cost function f(n) is defined as f(n)=g(n)+h(n).

[0027] The loop selects the node with the smallest cost function f(n) from the open list as the current node, checks the adjacent grids in the 8 directions of the current node, and filters invalid nodes, if the node is not in the open list, the open list is added and the parent node is recorded, if it is in the open list, the parent node and the cost are updated when the new cost function g(n) is smaller, if the terminal G is added to the closed list, the parent node is backtracked to generate the path, or the open list is empty and the path is returned. When the path passes through the dynamic forbidden area, the device operation plan in the future time interval of the area in the digital twin system is queried, if the device operation plan exists conflict, the dynamic forbidden area is converted into a static obstacle to re-plan the path.

[0028] Preferably, the edge computing gateway is used for fault prediction and processing of the device sensor, and further includes:

[0029] The vibration data of the crane vibration sensor, temperature sensor and truck load sensor data is preprocessed to obtain preprocessed data, and the preprocessed data is input into a random forest model.

[0030] The random forest model extracts low-frequency energy, medium-frequency energy and high-frequency energy of the vibration data, extracts sliding window mean and change rate of the temperature data, and extracts normalized current load and rated load ratio, each decision tree independently judges the category, counts the voting results of all trees, calculates the fault probability, if the fault probability is greater than or equal to 0.9, an emergency shutdown is triggered, if the fault probability is less than 0.9 and greater than or equal to 0.7, a high-risk reminder is triggered, and if the fault probability is less than 0.7, a low-risk is triggered.

[0031] Based on the same concept, the application also provides a multi-modal transport one-box intelligent trusted collaboration system based on artificial intelligence technology, comprising:

[0032] A data collaboration module, wherein the transportation data is converted into a standardized JSON format to obtain a standardized JSON data package, and the standardized JSON data package is triple-encrypted to obtain an encrypted data package and broadcast to a blockchain;

[0033] A verification module, wherein Internet of Things device data and electronic bill of lading key fields including bill of lading number, cargo category and consignee and consignor are collected in real time, the verification nodes in the blockchain network perform consensus verification on the Internet of Things device data and the electronic bill of lading key fields, and if the verification is passed, the data is written into the blockchain, if there is data conflict, a warning is given and the data is corrected;

[0034] An intelligent hub dynamic scheduling module, wherein a digital twin system fuses multi-source data including physical device data, business system data and external environment data, calculates data according to the multi-source data, and an artificial intelligence calculates an optimal loading and unloading sequence and a transportation path according to the calculated data, generates an operation instruction and issues the operation instruction to field automation equipment;

[0035] A tracking module, wherein a Beidou positioning terminal built in a container and a 5G / LoRa gateway along the way upload longitude and latitude data to a blockchain node every minute, and automatically identify transportation stage abnormalities based on geographic fence technology, and when physical device data exceeds an intelligent contract threshold, an alarm is pushed to a consignee and an insurance company.

[0036] Based on the same concept, the application also provides a computer device comprising a memory and one or more processors, the memory storing computer code, and the computer code being executed by the one or more processors to make the one or more processors execute the multi-modal transport one-box intelligent trusted collaboration method based on artificial intelligence technology as described in any one of the embodiments.

[0037] Based on the same concept, the application also provides a computer readable storage medium storing computer code, when the computer code is executed, the multi-modal intermodal one-box intelligent trusted collaboration method based on artificial intelligence technology as described in any of the embodiments is executed.

[0038] Compared with the prior art, the application has the following beneficial effects:

[0039] (1) The transport data of the application is converted into standardized JSON data packets according to the agreed rules, and the standardized JSON data packets are triple-encrypted to obtain encrypted data packets and broadcast to the blockchain, thereby breaking down data barriers, unifying operation rules, and realizing efficient connection across transportation modes.

[0040] (2) The application realizes consensus verification of the Internet of Things device data and the electronic bill of lading key fields including the bill of lading number, the cargo category, and the consignee and consignor by the verification nodes in the blockchain network, and if the verification is passed, it is written into the blockchain, if there is data conflict, it is warned and the data is corrected, realizing the consensus verification of the Internet of Things device data and the electronic bill of lading key fields by the verification nodes in the blockchain network, ensuring the strong consistency of business flow and logistics, and preventing manual input errors or irregular operations.

[0041] (3) The application realizes dynamic planning of railway branch lines and short transfer highway routes to expand the coverage of the "one-box-to-bottom" service, ensures the whole process of goods without unloading, thereby reducing the risk of cargo damage and improving the transfer efficiency, improving the overall efficiency of logistics operation, and reducing the time and cost consumption in the logistics link.

[0042] (4) The application realizes the improvement of the overall efficiency of multi-modal intermodal collaboration by comparing the planned path with the actual path through the Beidou positioning terminal built in the container and the 5G / LoRa gateway along the way, uploading the latitude and longitude data to the blockchain node every minute, and if the set rules are exceeded, triggering the transport path re-planning mechanism. BRIEF DESCRIPTION OF DRAWINGS

[0043] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are for purposes of illustration only and are not considered a limitation of the application.

[0044] Figure 1A flowchart of the multi-modal transport one-box intelligent trusted collaboration method based on artificial intelligence technology of the present application;

[0045] Figure 2 A verification flowchart of the multi-modal transport one-box intelligent trusted collaboration method based on artificial intelligence technology of the present application;

[0046] Figure 3 An architecture schematic diagram of the multi-modal transport one-box intelligent trusted collaboration system based on artificial intelligence technology of the present application. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solutions and advantages of the present application clearer and more apparent, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application. Obviously, the described examples are part of the examples of the present application, but not all the examples. Based on the examples in the present application, all other examples obtained by those skilled in the art without making creative efforts fall within the scope of protection of the present application.

[0048] Those skilled in the art can understand that, unless specifically stated, the singular form "a", "an", "the" used herein can also include the plural form. It should be further understood that the phrase "comprising" used in the specification of the present application means that the features, integers, steps, operations, elements and / or components exist, but does not exclude the existence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0049] The technical terms related to the present application are defined as follows:

[0050] Blockchain: a decentralized distributed database that records data in blocks arranged in chronological order through encryption algorithms and is maintained by multiple nodes in the network. It includes the core features of decentralization, data tamper resistance, high transparency, strong security, high availability and promotion of cross-border cooperation. These features enable blockchain to maintain data consistency in an environment without trust, reduce transaction costs, enhance system security and enable point-to-point value transmission.

[0051] Trusted system technology: a comprehensive technology system designed to ensure that computers and network systems can operate safely and reliably when facing various threats. It includes hardware and software security, encryption, security protocols, auditing, vulnerability management, intrusion detection and other aspects, aiming to achieve comprehensive security protection. The technology emphasizes the self-protection and recovery capabilities of the system, and through continuous monitoring and policy adjustment, it responds to changing threats to establish trust between users and the system and protect the confidentiality, integrity, availability and non-repudiation of information.

[0052] Internet of Things technology monitoring equipment: through sensor, network communication and data processing technology, realize the remote monitoring, control and management of physical equipment, to collect state data, analyze performance and predict maintenance needs.

[0053] Artificial intelligence decision-making: using core feature technology to simulate human decision-making process, through data collection, model establishment and training, feature selection and verification test, etc. Steps, realize the analysis of complex data and fast and accurate decision-making, improve the decision-making efficiency and quality.

[0054] First embodiment

[0055] A batch of automobile parts (high value, fragile) is shipped to a certain country;

[0056] Transportation path: container ship → a certain port → railway train → China-Europe train;

[0057] Emergency: the ship is delayed by 4 hours, and the transfer needs to be completed within 2 hours, otherwise the only railway train of the day will be missed.

[0058] Please refer to Figure 1 As shown, the multi-modal transport document data exchange credible system based on blockchain technology provided by the embodiment includes:

[0059] Join the blockchain network maintained by sea shipping companies, railway departments, highway transportation enterprises, etc. To establish a transportation data sharing chain, and each participant as a chain node.

[0060] S1: The transportation data is converted into a standardized JSON format according to the agreed rules to obtain a standardized JSON data packet, and the standardized JSON data packet is triple-encrypted to obtain an encrypted data packet and broadcast to the blockchain.

[0061] Preferably, in step S1, the transportation data is converted into a standardized JSON format according to the agreed rules to obtain a standardized JSON data packet, further comprising:

[0062] Through the data conversion gateway, input the transportation data including structured data and unstructured data, wherein the structured data includes EDI message, XML transportation document, CSV format, and the unstructured data includes paper document scan, ORC engine and NLP model extracted key field. Specifically, in this embodiment, the invalid fields (such as duplicate records, garbled code) in the transportation data are removed, the missing values are filled (the transportation time is completed by the historical data mean), and the data types are unified for data cleaning. The ORC / NLP extracted field is verified with the EDI message at the JSON level, and the original file such as bill of lading scan is hashed to prevent tampering;

[0063] The transport data outputs a standardized JSON data packet according to an agreement rule, wherein the agreement rule includes a container structured data model and a space-time data rule, the container structured data model includes a unique identifier ISO 6346 code, physical parameters including size, load, and box type, and safety inspection status including refrigerated box temperature control range, dangerous goods UN number, and inspection status, and the space-time data rule is that a time stamp is forced to use an ISO 8601 extended format (accurate to milliseconds + time zone) for device end encryption, and location data is fused with a WGS84 coordinate system and a UN / LOCODE port code. Specifically, in this embodiment, the UN number and the temperature control range are written into the blockchain in real time, the smart contract is triggered to alarm in an abnormal state, the port / railway / road carrier automatically identifies the transfer node through the UN / LOCODE, the ISO 8601 time stamp encryption ensures accurate and synchronous operation timing in each time zone, and AGV vehicle-mounted RFID reading of the container electronic tag is realized according to the permission mapping table in Table 1 to realize decryption permission hierarchical control.

[0064] Standardized JSON data packet instance:

[0065] {

[0066] "metadata": { / / Metadata layer: describes data source and processing information

[0067] "data_source": { / / Data source classification

[0068] "structured": [ / / Structured data source list

[0069] {

[0070] "type": "EDI", / / Data type: EDI electronic data interchange

[0071] "version": "EDIFACT-D96A", / / EDI standard version

[0072] "raw_data_hash": "sha256(...)" / / SHA-256 hash value of the original data (for data integrity protection, used for blockchain verification)

[0073] },

[0074] {

[0075] "type": "XML", / / Data type: XML extensible markup language

[0076] "schema": "CIM / SMDG", / / XML schema definition used

[0077] "raw_data_hash": "sha256(...)" / / Hash of the raw XML

[0078] }

[0079] ],

[0080] "unstructured": [ / / List of unstructured data sources

[0081] {

[0082] "type": "scanned_document", / / Data type: scanned document

[0083] "format": "PDF", / / File format

[0084] "ocr_engine": "Google Vision API", / / OCR engine used

[0085] "nlp_model": "BERT-Logistics", / / NLP model used

[0086] "extracted_fields": ["..."], / / List of extracted key fields

[0087] "raw_file_hash": "sha256(...)" / / SHA-256 hash of the raw file

[0088] }

[0089] ],

[0090] "gateway_id": "GATEWAY-EU-001", / / Unique identifier of the data transformation gateway

[0091] "processing_timestamp": "2025-01-01T15:30:45.123Z" / / Gateway processing time (ISO 8601 format, device-side encryption)

[0092] }

[0093] },

[0094] "container_model": { / / Container core data model

[0095] "identifier": { / / Container identification information

[0096] "container_number": "MSKU 601234", / / Container physical number (ISO 6346 standard)

[0097] "bic_code": "MSKU 001234" / / BIC International Container Bureau registration code

[0098] },

[0099] "physical_parameters": { / / Physical characteristics parameters, used to determine the physical characteristics of the transport tool selection

[0100] "size": "40 feet", / / Container size

[0101] "type": "High Cube", / / Container type

[0102] "tare_weight_kg": 3800, / / Container tare weight (kg)

[0103] "max_payload_kg": 28400, / / Maximum payload (kg)

[0104] "volume_cbm": 67.7 / / Internal volume (m 3 )

[0105] },

[0106] "safety_status": { / / Safety inspection status

[0107] "cargo_type": "Hazardous", / / Cargo type

[0108] "un_number": "UN3480", / / United Nations hazardous goods number

[0109] "inspection": { / / Inspection certificate information

[0110] "certificate_id": "SAFE-20240624-001", / / Inspection certificate ID

[0111] "expiry_date": "2025-12-31", / / Certificate validity period

[0112] "is_valid": true / / Current validity status

[0113] },

[0114] "reefer_settings": { / / Settings specific to refrigerated containers

[0115] "temperature_range": [-25, -18], / / Temperature control range [min, max] (in °C)

[0116] "humidity_percent": 65, / / Humidity setting (%)

[0117] "power_status": "ON" / / Power status

[0118] }

[0119] }

[0120] },

[0121] "spatiotemporal_data": { / / Spatiotemporal data layer

[0122] "timestamp": "2025-06-24T14:25:30.456Z", / / Event timestamp (ISO 8601 format, encrypted at device end)

[0123] "location": { / / Location information combining WGS84 coordinates and UN / LOCODE

[0124] "gps": { / / GPS coordinates (WGS84 coordinate system)

[0125] "latitude": 51.92442, / / Latitude

[0126] "longitude": 4.46773, / / Longitude

[0127] "coordinate_system": "WGS84" / / Coordinate system identifier

[0128] },

[0129] "unlocode": "NLRTM", / / UN / LOCODE port code (Rotterdam Port)

[0130] "facility": "ECT Delta Terminal" / / Specific facility name

[0131] },

[0132] "movement_status": "RAIL_LOADED" / / Standardized movement status enumeration value (Rail Loaded)

[0133] },

[0134] "compliance_checks": { / / Compliance check results

[0135] "iso_8601_verified": true, / / Timestamp format compliance

[0136] "unlocode_mapped": true / / Coordinate and UN / LOCODE consistency

[0137] },

[0138] "blockchain_link": { / / Blockchain association information

[0139] "previous_event_hash": "0x89a2...c7d1", / / Previous event blockchain hash value, forming an unalterable event chain

[0140] "data_package_hash": "sha256(...)", / / Current data package SHA-256 hash value, core data fingerprint

[0141] "smart_contract_address": "0x742d...b9e3" / / Associated smart contract address, triggering automated business logic

[0142] }

[0143] }

[0144] Table 1: Permission mapping table

[0145]

[0146] Preferably, the standardized JSON data package is triple-encrypted to obtain an encrypted data package and broadcast to the blockchain, further comprising:

[0147] S11: symmetrically encrypt the original data of the standardized JSON data packet by using the device end AES-256 to obtain symmetrically encrypted data, specifically, in the embodiment, a 256-bit key is generated by using a random number generator, and the original data such as temperature and humidity is symmetrically encrypted by AES-256 to obtain symmetrically encrypted data;

[0148] S12: dynamically generate a session key through a TLS1.3 protocol, and encrypt the symmetrically encrypted data according to the session key to obtain asymmetrically encrypted data, specifically, in the embodiment, a blockchain node generates a public key and a private key, the public key is public, and the private key is securely stored by the node, the symmetrically encrypted data is encrypted by using the public key of the node to prevent network monitoring and tampering;

[0149] S13: twice encrypt the asymmetrically encrypted data by using a cloud end AES-256 key, perform SHA256 hash operation on the twice encrypted data to generate an encrypted data packet, and broadcast the encrypted data packet to a blockchain, specifically, in the embodiment, the security and integrity of the static data are guaranteed, the final encrypted data and the hash value are stored in the blockchain, and the tamper-proofing and privacy protection are realized.

[0150] S2: real-time collection of Internet of Things device data and electronic bill of lading key fields including bill of lading number, cargo category, and consignee and consignor, consensus verification of the Internet of Things device data and the electronic bill of lading key fields by a verification node in a blockchain network, if the verification is passed, then writing into the blockchain, if there is data conflict, pre-warning and data correction, specifically, in the embodiment, the Internet of Things device actively reports data to an edge computing gateway at a preset frequency.

[0151] Please refer to Figure 2 In step S2, the verification node in the blockchain network performs consensus verification on the Internet of Things device data and the electronic bill of lading key fields according to the encrypted data packet, further comprising:

[0152] The verification node in the blockchain network performs consensus verification on the weight data obtained by the intelligent weighbridge, the declared weight in the electronic bill of lading, the container number extracted by the container number identification camera, the ISO check code (OCR reads ISO6346 code), the temperature and humidity, and the cold chain protocol, wherein the verification node includes nodes such as the current port, the next segment carrier, and the customs supervision party, specifically, in the embodiment, the verification node further includes two other transportation enterprise nodes.

[0153] The specific consensus verification process is as follows:

[0154] The current port calculates the difference rate of the weight data obtained by the intelligent weighbridge and the declared weight in the electronic waybill. If the difference rate exceeds 10%, a red early warning is sent and the data is rejected, triggering automatic calibration record retrieval of the weighbridge, requiring manual review of the container seal video, and verifying the validity of the container number again. If the difference rate does not exceed 10%, it is valid.

[0155] The next paragraph checks whether the real-time temperature and humidity data is within the cold chain temperature control range preset in the electronic waybill. If it exceeds the range, a hierarchical early warning is given according to the duration of the exceedance. If the duration does not exceed 30 minutes, a yellow early warning is issued, the standby refrigeration unit is started, and the shipper is notified to choose to continue transportation or change the container. If the duration exceeds 30 minutes, a red early warning is issued and the data is rejected. At the same time, the historical temperature and humidity data is compared to confirm that there is no sudden rise or fall (such as a fluctuation of not more than 2°C in the past 1 hour).

[0156] The customs supervision party checks whether the declared weight in the electronic waybill is consistent with the customs declaration. If it is not consistent, the data is rejected and the shipper is required to provide a written explanation. If the goods are hazardous, check whether the correct UN number is included in the container number safety inspection status. If there is no or incorrect number, the data is rejected. Check whether the container number is correctly bound to the container number field in the electronic waybill. If it is not bound or bound incorrectly, it is considered invalid data.

[0157] If more than half of the verification nodes confirm that the data is real and valid, the Internet of Things device data and the key fields of the electronic waybill are written into a new block to form an unalterable electronic file. If any verification node detects a missing field or value range overrun, it is rejected and a verification report is generated. Specifically, in this embodiment, data solidification is achieved through block structure. The block header records version number, timestamp, previous and next block correlation hash, data fingerprint tree root, and other blockchain basic information. The transaction body stores the key verification results, achieving the effect of data tamper-proofing and historical traceability.

[0158] Through this whole-process data mutual verification mechanism, the operation standards of different transportation segments are automatically aligned, fundamentally avoiding repeated loading and unloading or container replacement caused by information misplacement.

[0159] S3: The digital twin system fuses multi-source data including physical device data, business system data, and external environment data, extracts calculation data according to the multi-source data, and the artificial intelligence calculates the optimal loading and unloading sequence and transportation path according to the calculation data, generates operation instructions, and issues them to the on-site automation equipment.

[0160] Preferably, in step S3, the digital twin system fuses multi-source data including physical device data, business system data, and external environment data, extracts calculation data according to the multi-source data, and the artificial intelligence calculates the optimal loading and unloading sequence and transportation path further includes:

[0161] S31: The digital twin system fuses device sensor data, AI visual recognition container position and barcode information, TOS port loading and unloading task list and transportation timetable, real-time weather and traffic conditions, and extracts calculation data including device load rate, container urgency and path congestion coefficient from multi-source data. Specifically, in this embodiment, the device sensor real-time acquires sensor data including rail-mounted crane motor current, rail-mounted crane motor temperature (82°C), AGV battery power (38%), position encoder, AGV unmanned vehicle battery status, speed, etc., identifies the container position (ship shell position C7-32), barcode (MSKU 778901) and other information through AI vision, the business system data includes the loading and unloading task list from the TOS port operation system, the railway / road transportation timetable (remaining 110 minutes departure), and obtains external environment such as 7-level gust (affecting hoisting safety) and D area congestion in the yard. The multi-source data obtained is cleaned by removing sensor noise and repairing missing values, and the multi-source data is aligned with space-time labels, such as matching the container position identified by vision with the AGV unmanned vehicle coordinates to complete data fusion.

[0162] According to the multi-source data, the calculation data includes:

[0163] Device load rate = current task time consumption / device maximum processing capacity

[0164] Container urgency = inverse of distance to transportation tool departure time

[0165] Path congestion coefficient = real-time AGV unmanned vehicle flow / road segment capacity.

[0166] S32: Multi-agent deep deterministic policy gradient MADDPG assigns container hoisting priority to the rail-mounted crane and plans a dynamic path for the AGV unmanned vehicle according to the calculation data.

[0167] The rail-mounted crane is taken as the first agent and the AGV unmanned vehicle is taken as the second agent to construct the multi-agent deep deterministic policy gradient MADDPG. Specifically, in this embodiment, the rail-mounted crane decision: container MSKU 778901 hoisting priority: 1st (i.e. the first agent), operation parameters: maximum swing angle: ≤5° (anti-shock requirement), hoisting speed: 0.8 m / s (normal 1.2 m / s), target coordinates: [X:120, Y:350] (temporary storage area), AGV path planning: path A (shortest path): distance 350 m, but needs to pass through high wind speed area → high safety risk; path B (detour charging): distance 480 m, can supplement power → time-consuming increase; path C (tunnel passage): distance 420 m, wind avoidance but needs permission → optimal solution.

[0168] Define state space: including equipment load, container priority, path congestion, energy consumption, etc.

[0169] Define action space: assign container lifting sequence for rail-mounted cranes, AGV travel path speed and steering instructions.

[0170] Define reward function: including rail-mounted crane reward items and AGV reward items, where the positive reward of rail-mounted crane reward items is the number of completed lifting containers per unit time x emergency weight, the negative reward of rail-mounted crane reward items is the proportion of idle time x penalty coefficient, the positive reward of AGV reward items is the efficiency value of reaching the end point by the shortest path, and the negative reward of AGV reward items is the number of path conflicts x emergency loss weight.

[0171] Rail-mounted crane priority calculation: emergency = 0.9(high value) x e^(-0.1x110) x 1.3(fragile goods)=0.87;

[0172] AGV path parameters: congestion coefficient = 0.7(static obstacles) + 0.2(dynamic AGV) = 0.9; safety coefficient = 0.6(wind speed reduction 30%).

[0173] MADDPG technology realizes millisecond-level dynamic decision-making in complex environment, intelligent resolution of equipment coordination and resource conflict, and automatic binding of physical operation and blockchain contract for seamless conversion of multi-modal transport whole-process state.

[0174] Preferably, in step S32, the multi-agent deep deterministic policy gradient MADDPG assigns container lifting priority for rail-mounted cranes and plans dynamic path for AGV based on calculation data, further comprising:

[0175] S321: discretize digital twin station into a two-dimensional grid map to obtain a digital twin station topology map, the grid attributes include static obstacles and dynamic exclusion zones, where the static obstacles include obstacles such as buildings and fixed equipment areas, and the dynamic exclusion zones include rail-mounted crane operation radius range and temporary container stacking area, specifically, in this embodiment, the station ground is divided into 1m*1m grid for spatial discretization, and 1s interval is generated for future 30 time slices for time discretization, if two AGVs occupy the same grid in the same time slice, it is marked as a conflict;

[0176] S22: generating an initial shortest path based on the digital twin terminal topology map using artificial intelligence technology, and predicting whether there is a path conflict in the future for a plurality of time intervals through the space-time grid method, if there is a path conflict, assigning a path priority to the AGV of the high emergency container, triggering the local detour of other AGVs, to realize dynamic real-time adjustment, specifically, in this embodiment, the final decision is: [wharf] → [underground tunnel] → [charging pile] → [rail yard], and the estimated time consumption is: 8 minutes and 45 seconds (including 45 seconds of charging);

[0177] Generating an initial shortest path based on the digital twin terminal topology map using artificial intelligence technology, including:

[0178] The open list stores the starting point S (f=0), and the closed list is used to store the explored nodes and is initially empty, the cost function g(n) is defined as the actual movement cost from the starting point to the current node n, the adjacent grid distance is 1, and the diagonal grid distance is The cost function h(n) is defined as the Manhattan distance estimation cost of the current node n to the terminal G, and the cost function f(n) is defined as f(n)=g(n)+h(n);

[0179] The loop selects the node with the minimum cost function f(n) from the open list as the current node, checks the adjacent grids in the 8 directions of the current node, and filters invalid nodes, if the node is not in the open list, the node is added to the open list and the parent node is recorded, if it is in the open list, the parent node and the cost are updated when the new cost function g(n) is smaller, if the terminal G is added to the closed list, the parent node is backtracked to generate the path, or the open list is empty and the path is returned, the loop is stopped; when the path passes through the dynamic forbidden area, the equipment operation plan in the future for a plurality of time intervals in the area is queried in the digital twin system, if there is a conflict in the equipment operation plan, the dynamic forbidden area is converted into a static obstacle to re-plan the path, specifically, in this embodiment, the space-time cube overlapping area of different AGVs is detected, which is marked as a conflict point, when the conflict prediction occurs, the road right is re-allocated according to the container emergency degree (defined as the reciprocal of the distance from the transport tool departure time), the original path is reserved for the AGV of the high emergency container, and the local detour action is forced to be performed on other AGVs, that is, the conflict point is taken as the center, and a Dijkstra algorithm is used to search for a substitute sub-path within a radius of 5 m, the sub-path meets the distance deviation less than and equal to 15% of the initial path length, and no new space-time conflict, the adjusted path instruction is input into the action space of the MADDPG model for collaborative optimization.

[0180] When the rail crane and AGV need to cooperate, the waiting time of the container handover point is included in the state space. Specifically, in this embodiment, the digital identity of the container (matching MSKU 778901 on the blockchain) is verified before the AGV departs; the hoisting operation parameter hash value of the rail crane is recorded: HASH (speed 0.8 m / s, angle 5°); when the AGV arrives at the railway yard: automatically scan the container barcode, update the blockchain state to "port section completed, waiting for railway loading", and trigger the smart contract to release the sea freight;

[0181] If the AGV is expected to arrive later than the rail crane's idle time window, the MADDPG model is triggered to adjust synchronously, reducing the hoisting priority of the rail crane and increasing the path speed instruction for the associated AGV.

[0182] Based on the on-chain tracking data, a BI board is constructed to integrate multi-dimensional data such as location, temperature and humidity, and handover records. The full-process trajectory of the container is visualized on the map, and a transportation efficiency report is generated. Combined with historical data, a prediction model is trained to provide decision support for optimizing path planning and resource scheduling, and to improve the overall collaborative efficiency of multimodal transport.

[0183] Intelligent decision-making is based on artificial intelligence algorithms to analyze full-link data and dynamically optimize scheduling strategies. Combined with real-time weather, road conditions, and equipment status data, artificial intelligence automatically adjusts the loading and unloading priority or diversion path; predicts the flow direction of goods and the demand for empty containers, and plans the return of containers in advance to reduce empty running rate. Ultimately, it realizes "one container to the end" without disassembling or replacing containers, reducing labor intervention costs and the risk of damage.

[0184] Preferably, the edge computing gateway is used for fault prediction and processing of device sensors, further comprising:

[0185] The vibration data of the crane vibration sensor, the temperature sensor and the truck load sensor are preprocessed by the edge computing gateway to obtain preprocessed data, and the preprocessed data is input into a random forest model. Specifically, in this embodiment, a running monitoring module is installed on the crane, the truck and the like, vibration sensors are installed on bearings and gearboxes at key monitoring points of the crane to collect high-frequency vibration signals, temperature sensors are used to monitor motor winding and bearing temperature, stress sensors are used to monitor the load-bearing state and metal fatigue of the crane hook, load sensors are installed on the truck monitoring points to detect the pressure distribution on the truck bed, and GPS / IMU is installed to obtain real-time position, acceleration and turning inclination. The crane sensor data is transmitted to the edge gateway through 5G, the edge computing layer uses a moving average method to take the average of 10 consecutive data points, smooths random fluctuations for preprocessing and noise reduction (interference removal), sets vibration threshold values, temperature threshold values and the like, directly deletes or replaces the average of the previous and subsequent data for abnormal value filtering (error data exclusion), and only two core indicators obtained by the vibration sensor are retained: vibration intensity average for reflecting the overall vibration level of the device, and maximum vibration value for capturing sudden abnormal impact. After local preprocessing, the data is uploaded to a cloud analysis platform to realize real-time tracking and intelligent management of the device running state through multiple types of sensors and edge computing technology.

[0186] The random forest model extracts low-frequency energy, medium-frequency energy and high-frequency energy of the vibration data, extracts sliding window average and change rate of the temperature data, and extracts the ratio of the current load to the rated load after standardization. Each decision tree independently determines its class, and the voting results of all trees are counted to calculate the failure probability. If the failure probability is greater than or equal to 0.9, an emergency shutdown is triggered. If the failure probability is less than 0.9 and greater than or equal to 0.7, a high-risk reminder is triggered, and the truck is maintained within 6 hours. If the failure probability is less than 0.7, it is a low-risk, and it is included in the regular maintenance plan. Specifically, in this embodiment, the platform uses a machine learning model to fuse and analyze historical and real-time data, and generates a maintenance work order and pushes it to the operation and maintenance personnel 6-12 hours in advance to warn the crane bearing failure. The truck load distribution scheme is optimized in combination with the load and route data to reduce the empty running rate and the risk of uneven loading. When the predicted bearing failure probability is greater than 0.7, the influence of the bearing fracture on the overall structure is simulated by a twin body. If the failure probability is greater than 0.9 and there are spare parts in stock, an urgent replacement work order is pushed and the nearest maintenance team is dispatched. If the failure probability is greater than 0.7 and there are no spare parts in stock, it is recommended to run at a reduced load, and an automatic spare parts procurement process is triggered. When the container inclination sensor is greater than 15° for 5 seconds, the AGV emergency stop is triggered. When the temperature of the refrigerated container rises by 2℃ / min, the AGV automatically changes the route to the maintenance area.

[0187] By accessing the artificial intelligence scheduling system, integrating real-time data such as crane load, truck location, and yard capacity, and dynamically adjusting resource allocation using reinforcement learning algorithms, a risk prediction model trained with historical data and real-time sensor information can provide early warning of equipment failure and abnormal cargo, and real-time emergency instructions can be pushed to the operation terminal, reducing the possibility of low efficiency and unreasonable equipment allocation when manually operating yard equipment, ensuring the smoothness of container transportation, and significantly improving the efficiency of container transportation.

[0188] S4: Through the Beidou positioning terminal built-in the container and the 5G / LoRa gateway along the way, latitude and longitude data is uploaded to the blockchain node every minute, and the transportation stage is automatically identified based on the geographic fence technology. When the physical device data exceeds the threshold of the smart contract, an alarm is pushed to the consignor and the insurance company.

[0189] Smart contract automatic verification: AGV planning path generates a digital fingerprint HASH, and the actual driving path points are encrypted and chained (every 10 seconds), and the smart contract compares the deviation of the planned / actual path

[0190] The "transportation path as a contract" is realized, avoiding the risk of human tampering with the route, and the real-time effect is shown in Table 2:

[0191] Table 2: Implementation effect

[0192]

[0193] By deeply binding AGV path planning with blockchain container digital identity, cross-transportation mode timetable, and environmental risk data, the physical transportation flow and information control flow are trusted to cooperate, supporting "one-box" full-process responsibility tracing.

[0194] The Beidou positioning terminal built-in the container, combined with the 5G base station and LORA Internet of Things gateway along the way, uploads location data to the blockchain node every minute, automatically identifies the transportation stage such as sea transportation, railway station, and highway transportation through the geographic fence technology. After the location data is verified by consensus, it is updated to the chain. The consignor and the carrier can query the container location in real time through the system, compare the planned path with the actual situation, and send a detention warning if the sea transportation stays within the 200-nautical-mile polygon of the port for more than 72 hours, or stays within the 500-meter-radius circle of the railway station for more than the planned time + 4 hours, or deviates from the planned path buffer zone by more than 1 kilometer during highway transportation. If the cargo type is frozen seafood, check whether the temperature and humidity are within the smart contract range. If the threshold is exceeded, an alarm is pushed to the consignor and the insurance company. ±

[0195] ​The embodiment is based on a blockchain technology to construct a decentralized data sharing platform, encrypts records of container box specifications, loading and unloading records, temperature and humidity states and the like, solves data mutual trust and tampering risks across transportation modes; relies on artificial intelligence algorithms to integrate real-time transportation capacity data and equipment states of railways, ports and highways, such as crane load, yard capacity, vehicle position and the like, dynamically optimizes resource scheduling schemes, replaces traditional manual experience decision-making, and reduces equipment idle rate; realizes whole-process transparent monitoring through the Internet of Things and intelligent container tracking technology, combines an artificial intelligence risk prediction model to early warn abnormal events such as equipment failure and route congestion, and synchronously generates emergency bypassing or shunting schemes; based on unified document rules and operation standards of the multimodal transport data platform, connects information systems of shippers, logistics enterprises and different transportation subjects, eliminates rule conflicts such as box type adaptation and document circulation, ensures seamless connection of “one box to the end”, reduces unloading and opening operations, and comprehensively improves logistics efficiency and reduces damage rate.

[0196] Second embodiment

[0197] Referring to Figure 3 The embodiment provides a multimodal transport one-box intelligent trusted collaboration system based on artificial intelligence technology based on the same concept, which comprises:

[0198] The basic software and hardware include domestic information creation servers, Kunpeng servers, domestic information creation operating systems such as Kirin, application middleware, TongWeb, TongRDS, TongHttpServer, databases, Dameng databases, Beidou positioning terminals and artificial intelligence technology.

[0199] Based on secure and trusted hardware and trusted middleware, the system is composed of a full-link data collaboration module, intelligent hubs and automated interfaces, Internet of Things equipment monitoring and container whole-process tracking, and interfaces with data information of external agencies and related units.

[0200] The data collaboration module converts transportation data into standardized JSON format to obtain standardized JSON data packets, performs triple encryption on the standardized JSON data packets to obtain encrypted data packets, and broadcasts the encrypted data packets to the blockchain;

[0201] The verification module collects Internet of Things equipment data and electronic waybill key fields including bill of lading number, cargo category and consignee and consignor in real time, and the verification nodes in the blockchain network perform consensus verification on the Internet of Things equipment data and the electronic waybill key fields, and if the verification is passed, the data is written into the blockchain, and if there is data conflict, an early warning is given and the data is corrected;

[0202] An intelligent hub dynamic scheduling module, a digital twin system fuses multi-source data including physical device data, business system data and external environment data, extracts calculation data according to the multi-source data, an artificial intelligence calculates an optimal loading and unloading sequence and a transportation path according to the calculation data, generates an operation instruction and delivers the operation instruction to field automation equipment;

[0203] A tracking module uploads longitude and latitude data to a blockchain node every minute through a Beidou positioning terminal built in a container and a 5G / LoRa gateway along the way, and automatically identifies a transportation stage exception based on a geographic fence technology, and when physical device data exceeds an intelligent contract threshold, an alarm is pushed to a consignee and an insurance company.

[0204] Third embodiment

[0205] In this embodiment, a computer device is provided, including a memory and one or more processors, the memory storing computer code, and the computer code being executed by the one or more processors to cause the one or more processors to perform the steps of the multi-modal transport one-box intelligent trusted collaboration method based on artificial intelligence technology in the first embodiment.

[0206] In some embodiments of the present application, a computer readable storage medium is also provided, and the computer readable instructions are executed by one or more processors to cause the one or more processors to perform the steps of the multi-modal transport one-box intelligent trusted collaboration method based on artificial intelligence technology as described in any one of the first embodiment.

[0207] It can be understood that, for the aforementioned multi-modal transport one-box intelligent trusted collaboration method based on artificial intelligence technology, if each is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a plurality of instructions to make a computer device (which can be a personal computer server, or a network device, etc.) execute all or part of the steps of the embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (Read Only memory, ROM), a random access memory (Random Access memory, RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0208] The computer readable storage medium can include a tangible medium such as those previously described. Additionally, or alternatively, the computer readable storage medium can represent a transmission medium such as those described herein.

[0209] The above merely illustrates the preferred embodiments of the present application, and the protection scope of the present application is not limited to the above-mentioned embodiments. Any technical scheme falling within the concept of the present application shall fall within the protection scope of the present application. It should be noted that, for those of ordinary skill in the art, some improvements and refinements without departing from the principles of the present application shall also be considered as falling within the protection scope of the present application.

Claims

1. A multimodal transport single-container intelligent and reliable collaborative method based on artificial intelligence technology, characterized in that: Includes the following steps: S1: The transportation data is converted into a standardized JSON format according to the agreed rules to obtain a standardized JSON data packet. The standardized JSON data packet is triple-encrypted to obtain an encrypted data packet and broadcast to the blockchain. S2: Real-time collection of IoT device data and key fields of electronic waybill including bill of lading number, cargo category, consignor and consignee. Verification nodes in the blockchain network perform consensus verification of the IoT device data and the key fields of the electronic waybill according to the encrypted data packet. If the verification is successful, it is written to the blockchain. If there is a data conflict, an early warning is issued and the data is corrected. S3: The digital twin system integrates multi-source data, including physical equipment data, business system data, and external environment data. It extracts and calculates data based on the multi-source data. Artificial intelligence calculates the optimal loading and unloading sequence and transportation route based on the calculated data, generates operation instructions, and sends them to the on-site automated equipment. S4: Through the Beidou positioning terminal built into the container and the 5G / LoRa gateway along the route, latitude and longitude data are uploaded to the blockchain node every minute, and anomalies in the transportation stage are automatically identified based on geofencing technology. When the data of physical equipment exceeds the smart contract threshold, an alarm is pushed to the cargo owner and insurance company. In step S3, the digital twin system integrates multi-source data, including physical equipment data, business system data, and external environment data, and extracts computational data based on the multi-source data. The artificial intelligence then calculates the optimal loading / unloading sequence and transportation route based on the computational data, further including: S31: The digital twin system integrates multi-source data including equipment sensor data, container location and barcode information recognized by AI vision, TOS terminal loading and unloading task list and transportation timetable, real-time weather and traffic conditions, and extracts calculation data including equipment load rate, container urgency and path congestion coefficient based on the multi-source data. Among them, the edge computing gateway performs fault prediction and processing for equipment sensors. S32: The Multi-Agent Deep Deterministic Strategy Gradient (MADDPG) assigns container lifting priorities to the rail-mounted gantry crane and plans dynamic paths for the AGV (Automated Guided Vehicle) based on the calculated data. In step S32, the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) further includes allocating container lifting priorities to the rail-mounted gantry crane and planning dynamic paths for the AGV (Automated Guided Vehicle) based on the calculated data: S321: Discretize the digital twin site into a two-dimensional grid map to obtain the digital twin site topology map. The grid attributes include static obstacles and dynamic restricted areas. Static obstacles include obstacles such as buildings and fixed equipment areas, while dynamic restricted areas include restricted areas such as the operating radius of rail-mounted gantry cranes and temporary container storage areas. S322: Based on the digital twin site topology map, the initial shortest path is generated using artificial intelligence technology, and the path conflict is predicted in the future several time intervals using the spatiotemporal grid method. If the path conflict exists, the AGV of the high-urgency container is assigned path priority, triggering other AGVs to detour locally, so as to achieve dynamic real-time adjustment. Based on the digital twin site topology map, artificial intelligence technology is used to generate the initial shortest path, including: An open list stores the starting point S (f=0), and a closed list stores explored nodes, initially empty. A cost function g(n) is defined as the actual movement cost from the starting point to the current node n. The distance between adjacent grid cells is 1, and the distance between diagonal grid cells is... Let h(n) be the estimated cost of the Manhattan distance from the current node n to the destination G, and let f(n) be the cost function f(n) = g(n) + h(n). The loop selects the node with the smallest cost function f(n) from the open list as the current node. It checks the adjacent grids in the eight directions of the current node (up, down, left, right, and diagonal) and filters out invalid nodes. If the node is not in the open list, it is added to the open list and the parent node is recorded. If it is already in the open list, the parent node and cost are updated when the new cost function g(n) is smaller. The loop stops when the endpoint G is added to the closed list, backtracks the parent node to generate the path, or when the open list is empty and returns no path. When the path crosses a dynamic restricted area, the system queries the equipment operation plans for that area over several future time intervals. If there is a conflict in the equipment operation plans, the dynamic restricted area is converted into a static obstacle and the path is replanned.

2. The intelligent and reliable collaborative method for multimodal transport using a single container based on artificial intelligence technology according to claim 1, characterized in that, In step S1, the transport data is converted into a standardized JSON data packet according to agreed rules, further including: The transportation data, which includes structured and unstructured data, is input through a data conversion gateway. The structured data includes formats such as EDI messages, XML transport documents, and CSV, while the unstructured data includes scanned copies of paper documents and key fields extracted by the ORC engine and NLP model. The transportation data is output as a standardized JSON data packet according to the agreed rules. The agreed rules include defining a container structured data model and spatiotemporal data rules. The container structured data model includes a unique identifier, physical parameters including dimensions, load capacity, and container type, and safety inspection status including the temperature control range of refrigerated containers, UN numbers for dangerous goods, and inspection status. The spatiotemporal data rules require that timestamps be encrypted on the device side using the ISO 8601 extended format, and that location data be fused with the WGS84 coordinate system and the UN / LOCODE port code.

3. The intelligent and reliable collaborative method for multimodal transport using a single container based on artificial intelligence technology according to claim 1, characterized in that, The process of triple-encrypting the standardized JSON data packet to obtain an encrypted data packet and broadcasting it to the blockchain further includes: S11: The original data of the standardized JSON data packet is symmetrically encrypted using AES-256 on the device to obtain symmetrically encrypted data; S12: Dynamically generate a session key through the TLS protocol, and encrypt symmetric encrypted data based on the session key to obtain asymmetric encryption; S13: The asymmetric key is encrypted a second time using the cloud-based AES-256 key. The encrypted data is then hashed using SHA256 to generate an encrypted data packet, which is then broadcast to the blockchain.

4. The intelligent and reliable collaborative method for multimodal transport using a single container based on artificial intelligence technology according to claim 1, characterized in that, In step S2, the verification nodes in the blockchain network perform consensus verification on the IoT device data and the key fields of the electronic waybill based on the encrypted data packet, further including: The verification nodes in the blockchain network perform consensus verification on the weight data obtained through the smart weighbridge and the declared weight in the electronic waybill, the container number extracted by the container number recognition camera and the ISO verification code, and the temperature, humidity and cold chain protocol. The verification nodes include nodes such as the current port, the next carrier, and the customs supervisor. If more than half of the verification nodes confirm that the data is authentic and valid, the IoT device data and the key fields of the electronic waybill will be written into a new block to form an unalterable electronic file. If any verification node detects that a field is missing or the value range is out of bounds, it will refuse to write the data.

5. The intelligent and reliable collaborative method for multimodal transport using a single container based on artificial intelligence technology according to claim 1, characterized in that, Fault prediction and processing of device sensors through edge computing gateways further include: Preprocessed data is obtained by preprocessing data from crane vibration sensors, temperature sensors, truck load sensors, and vibration data through an edge computing gateway, and then the preprocessed data is input into a random forest model. The random forest model extracts the low-frequency, mid-frequency, and high-frequency energy from the vibration data, the sliding window mean and rate of change of the temperature data, and the standardized ratio of current load to rated load. Each decision tree independently determines its category, the voting results of all trees are statistically analyzed, and the failure probability is calculated. If the failure probability is greater than or equal to 0.9, an emergency shutdown is triggered; if the failure probability is less than 0.9 but greater than or equal to 0.7, a high-risk warning is issued; if it is less than 0.7, it is considered low-risk.

6. A multimodal transport single-container intelligent and reliable collaborative system based on artificial intelligence technology, characterized in that: include: The data collaboration module converts transportation data into a standardized JSON format according to agreed rules to obtain a standardized JSON data packet. The standardized JSON data packet is then triple-encrypted to obtain an encrypted data packet, which is then broadcast to the blockchain. The verification module collects data from IoT devices in real time and key fields of electronic waybills, including bill of lading number, cargo category, consignor and consignee. Verification nodes in the blockchain network perform consensus verification on the IoT device data and the key fields of the electronic waybill. If the verification is successful, the data is written to the blockchain. If there is a data conflict, an alert is issued and the data is corrected. The intelligent hub dynamic scheduling module integrates multi-source data, including physical equipment data, business system data, and external environment data, and extracts calculation data based on the multi-source data. Artificial intelligence calculates the optimal loading and unloading sequence and transportation route based on the calculation data, generates operation instructions, and sends them to the on-site automated equipment. The tracking module uploads latitude and longitude data to the blockchain node every minute through the Beidou positioning terminal built into the container and the 5G / LoRa gateway along the route. It also automatically identifies anomalies in the transportation stage based on geofencing technology. When the data of the physical equipment exceeds the threshold of the smart contract, it pushes an alarm to the cargo owner and insurance company. The digital twin system integrates multi-source data, including physical equipment data, business system data, and external environment data. It extracts computational data based on this multi-source data, and artificial intelligence calculates the optimal loading / unloading sequence and transportation route based on the computational data. This further includes: S31: The digital twin system integrates multi-source data including equipment sensor data, container location and barcode information recognized by AI vision, TOS terminal loading and unloading task list and transportation timetable, real-time weather and traffic conditions, and extracts calculation data including equipment load rate, container urgency and path congestion coefficient based on the multi-source data. Among them, the edge computing gateway performs fault prediction and processing for equipment sensors. S32: The Multi-Agent Deep Deterministic Strategy Gradient (MADDPG) assigns container lifting priorities to the rail-mounted gantry crane and plans dynamic paths for the AGV (Automated Guided Vehicle) based on the calculated data. In step S32, the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) further includes allocating container lifting priorities to the rail-mounted gantry crane and planning dynamic paths for the AGV (Automated Guided Vehicle) based on the calculated data: S321: Discretize the digital twin site into a two-dimensional grid map to obtain the digital twin site topology map. The grid attributes include static obstacles and dynamic restricted areas. Static obstacles include obstacles such as buildings and fixed equipment areas, while dynamic restricted areas include restricted areas such as the operating radius of rail-mounted gantry cranes and temporary container storage areas. S322: Based on the digital twin site topology map, the initial shortest path is generated using artificial intelligence technology, and the path conflict is predicted in the future several time intervals using the spatiotemporal grid method. If the path conflict exists, the AGV of the high-urgency container is assigned path priority, triggering other AGVs to detour locally, so as to achieve dynamic real-time adjustment. Based on the digital twin site topology map, artificial intelligence technology is used to generate the initial shortest path, including: An open list stores the starting point S (f=0), and a closed list stores explored nodes, initially empty. A cost function g(n) is defined as the actual movement cost from the starting point to the current node n. The distance between adjacent grid cells is 1, and the distance between diagonal grid cells is... Let h(n) be the estimated cost of the Manhattan distance from the current node n to the destination G, and let f(n) be the cost function f(n) = g(n) + h(n). The loop selects the node with the smallest cost function f(n) from the open list as the current node. It checks the adjacent grids in the eight directions of the current node (up, down, left, right, and diagonal) and filters out invalid nodes. If the node is not in the open list, it is added to the open list and the parent node is recorded. If it is already in the open list, the parent node and cost are updated when the new cost function g(n) is smaller. The loop stops when the endpoint G is added to the closed list, backtracks the parent node to generate the path, or when the open list is empty and returns no path. When the path crosses a dynamic restricted area, the system queries the equipment operation plans for that area over several future time intervals. If there is a conflict in the equipment operation plans, the dynamic restricted area is converted into a static obstacle and the path is replanned.

7. A computer device comprising a memory and one or more processors, the memory storing computer code that, when executed by the one or more processors, causes the one or more processors to perform the steps of the multimodal transport single-container intelligent trusted collaborative method based on artificial intelligence technology as described in any one of claims 1-5.

8. A computer-readable storage medium storing computer code, wherein when the computer code is executed, the steps of the multimodal transport single-container intelligent reliable collaborative method based on artificial intelligence technology as described in any one of claims 1-5 are performed.

Images