A port tally vehicle monitoring management method and system

Abstract

The application provides a port tally vehicle monitoring management method and system, and relates to the technical field of port automation management. The method comprises the following steps: collecting multi-dimensional tally characteristics data of port tally vehicles through an Internet of Things sensing network and preprocessing; extracting multi-dimensional tally characteristic vectors, combining port traffic situations, and constructing a dynamic weighted directed graph model; planning an optimal tally driving path for the vehicle based on the dynamic weighted directed graph model, generating an optimal tally operation plan, and pushing the plan to a vehicle-mounted intelligent terminal; when the vehicle executes the plan, monitoring changes in the vehicle position and attitude and changes in the cargo state through multi-sensor fusion technology, triggering a hierarchical early warning mechanism and adjusting the plan when an abnormal event is monitored; and after the vehicle completes the plan, verifying operation process data based on a blockchain distributed ledger, generating a digital twin report, and synchronizing the report to the monitoring end, so that efficient, safe and intelligent monitoring and management of port tally vehicles can be achieved, and the operation efficiency and safety of the port can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of port automation management, and particularly relates to a port tally vehicle monitoring management method and system. BACKGROUND

[0002] With the rapid development of technologies such as the Internet of Things, big data, and artificial intelligence, port automation and intelligence have become the key to improving port operation efficiency and competitiveness.

[0003] In the traditional port tally process, the monitoring and management of tally vehicles mainly rely on manual operation and simple informationization means, which has problems such as low tally efficiency, low data accuracy, insufficient real-time monitoring capability, and low transparency of the tally process, and cannot meet the requirements of modern ports for efficient and transparent tally operations.

[0004] Therefore, it is necessary to provide a port tally vehicle monitoring management method and system to solve the above technical problems. SUMMARY

[0005] To solve the above technical problems, the present application provides a port tally vehicle monitoring management method and system to solve the problem that the traditional port tally vehicle monitoring management method mainly relies on manual operation and simple informationization means, and has problems such as low tally efficiency, low data accuracy, insufficient real-time monitoring capability, and low transparency of the tally process.

[0006] The present application provides a port tally vehicle monitoring management method, which comprises:

[0007] Real-time collection of multi-dimensional tally feature data of port tally vehicles through a port Internet of Things sensing network, and synchronous transmission to an edge computing node for preprocessing, wherein the multi-dimensional tally feature data includes electronic tag information, vehicle three-dimensional point cloud data, and cargo spectral feature data of the port tally vehicles;

[0008] Extraction of multi-dimensional tally feature vectors corresponding to the preprocessed multi-dimensional tally feature data, combined with real-time traffic conditions of the port, to construct a dynamic weighted directed graph model, wherein the multi-dimensional tally feature vectors include vehicle identity features, cargo state features, and tally operation priorities of the port tally vehicles;

[0009] Based on the dynamic weighted directed graph model, an optimal tally driving path containing space-time constraints is planned for the port tally vehicles, an optimal tally operation plan is generated, and the plan is pushed to a vehicle-mounted intelligent terminal through 5G vehicle networking technology;

[0010] When the port cargo handling vehicle executes the optimal cargo handling operation plan, the vehicle position and posture change data and the cargo state change data are monitored in real time through multi-sensor fusion technology, and when an abnormal event is detected, a hierarchical early warning mechanism is triggered and the optimal cargo handling operation plan is automatically adjusted.

[0011] After the port cargo handling vehicle completes the optimal cargo handling operation plan, the integrity and consistency of the cargo handling operation process data are verified based on a blockchain distributed ledger, a digital twin cargo handling report is generated, and the report is synchronized to the monitoring end through a quantum secure channel.

[0012] Preferably, the multi-dimensional cargo handling feature data of the port cargo handling vehicle is collected in real time through a port Internet of Things sensing network, and specifically includes:

[0013] The attention weight coefficient of the sensor data in the port Internet of Things sensing network is dynamically calculated based on a multi-head attention mechanism, and a time decay factor is introduced to adjust the attention weight coefficient.

[0014] Based on the adjusted attention weight coefficient, the sensor data is weighted and summed to obtain fused port Internet of Things sensing data, i.e., the multi-dimensional cargo handling feature data:

[0015] In the formula, indicates the fused port Internet of Things sensing data, i.e., the multi-dimensional cargo handling feature data; B indicates the total amount of sensor data participating in fusion in the port Internet of Things sensing network; indicates the attention weight coefficient of the bth sensor data in the port Internet of Things sensing network; indicates the bth sensor data in the port Internet of Things sensing network; indicates the bth sensor data after being processed by the multi-head attention mechanism; indicates a time decay term, which is used to adjust the attention weight coefficient of the sensor data according to the difference between the collection time of the sensor data and the current time; indicates a time decay factor, ; indicates the collection time of the bth sensor data; t indicates the current time; indicates the query matrix, key matrix, and value matrix of the bth sensor data, respectively; indicates the transpose of the key matrix ; indicates the dimension of the key matrix ; and Softmax() indicates an activation function.

[0016] Preferably, the dynamic weighted directed graph model includes port traffic nodes and port traffic connection road segments.

[0017] Based on the dynamic weighted directed graph model, a path planning objective function and constraints are set, and the path planning objective function is solved to determine the optimal tally driving path. The expressions of the path planning objective function and constraints are as follows:

[0018] Where i and j represent the port traffic nodes in the dynamic weighted directed graph model; ij represents the port traffic connection segment from port traffic node i to port traffic node j in the dynamic weighted directed graph model; A represents the set of all port traffic connection segments in the dynamic weighted directed graph model; represents the generalized cost function about the current time t; represents the decision variable for road segment selection, Indicates the selection of the port transportation connection section from port transportation node i to port transportation node j, Indicates that the port transportation connection section from port transportation node i to port transportation node j is not selected. Indicates the selection of the port transportation connection section from port transportation node j to port transportation node i, Indicates that the port transportation connection section from port transportation node j to port transportation node i is not selected; represents the path planning balance coefficient; L represents the set of all port traffic nodes in the dynamic weighted directed graph model; represents the weight coefficient of port transportation node i; represents the deviation value associated with port traffic node i; They represent the starting port transportation node and the ending port transportation node of the optimal tally driving route respectively; They represent the time when the port tally vehicle arrives at the port traffic node i and j respectively; represents the communication time of the port tally vehicle from port transportation node i to port transportation node j, which depends on the decision variable .

[0019] Preferably, for the port traffic connection section ij, the traffic congestion index predicted by integrating the spatiotemporal graph convolutional network is and dynamic reward values ​​generated by reinforcement learning , we get the generalized cost function for time t :

[0020] Where, Represents the fusion weight coefficient, which is used to adjust the traffic congestion index and dynamic reward value In the generalized cost function The proportion of .

[0021] Preferably, the triggering step of the hierarchical early warning mechanism is as follows:

[0022] D-S evidence theory is adopted to fuse multi-source abnormality evidence and calculate abnormality index , i.e. to aggregate the trust degree distribution of each evidence source on the abnormal event and deduct the cross-conflict trust degree, the calculation formula of the abnormality index is as follows:

[0023] In the formula, r represents the total number of multi-source abnormality evidence participating in fusion, i.e. the total number of evidence sources; A represents the abnormal event; represents the complement of the abnormal event A, i.e. the non-abnormal event; k and l represent the index of the evidence source, , , ; represents the trust degree distribution of the kth evidence source on the abnormal event A, i.e. the trust degree of the kth evidence source on the occurrence of the abnormal event A; represents the trust degree distribution of the lth evidence source on the non-abnormal event , i.e. the trust degree of the lth evidence source on the occurrence of the non-abnormal event ;

[0024] Two-level early warning thresholds are set, including a first-level early warning threshold and a second-level early warning threshold ;

[0025] According to the comparison result of the abnormality index and the two-level early warning thresholds, the corresponding early warning level level is triggered:

[0026] .

[0027] Preferably, the automatic adjustment process of the optimal cargo handling operation plan is as follows:

[0028] A plan adjustment objective function and constraint conditions are set, the plan adjustment objective function is solved, and the adjusted optimal cargo handling operation plan is obtained, and the expression of the plan adjustment objective function and constraint conditions is as follows:

[0029] In the formula, M represents the total number of cargo handling operations; N represents the total number of resources for executing cargo handling operations; represents the cost of the mth cargo handling operation executed by the nth resource; represents a cargo handling operation distribution decision vector, represents that the mth cargo handling operation is distributed to the nth resource, represents that the mth cargo handling operation is not distributed to the nth resource; a weight coefficient representing an adjustment deviation of the optimal cargo handling operation plan, used to balance the deviation degree of actual cargo handling operation allocation cost and target cargo handling operation allocation cost; F represents the total number of adjustment stages; an adjustment weight coefficient of the mthcargo handling operation, used to distinguish the importance of different cargo handling operations in the adjustment process of the optimal cargo handling operation plan; representing the actual completion time of the mthcargo handling operation; representing the target completion time of the mthcargo handling operation; representing the processing time of the mthcargo handling operation; representing the capacity limit of the nthresource, i.e. the longest cargo handling operation processing time borne by the nthresource; respectively representing the time nodes of the f+1thand fthadjustment stages; representing an indication vector, when the mthcargo handling operation is processed in the fthadjustment stage, when the mthcargo handling operation is not processed in the fthadjustment stage, .

[0030] Preferably, after the port cargo handling vehicle completes the optimal cargo handling operation plan, the integrity and consistency of the cargo handling operation process data are verified based on the blockchain distributed ledger, a digital twin cargo handling report is generated and synchronized to the monitoring end through a quantum secure channel, specifically including:

[0031] After the port cargo handling vehicle completes the optimal cargo handling operation plan, the cargo handling operation process data of the port cargo handling vehicle are automatically collected, including cargo loading and unloading records, cargo handling operation time series, vehicle position coordinate sets and equipment interaction logs, and after standardized processing, an original cargo handling operation data set is generated;

[0032] An SM3 hash algorithm is used to calculate a unique hash value of the original cargo handling operation data set, and the original cargo handling operation data set is bound to form a to-be-verified cargo handling operation data set;

[0033] The to-be-verified cargo handling operation data set is pushed to a blockchain distributed node, the blockchain distributed node calls an improved practical Byzantine fault tolerance consensus algorithm to perform multi-node cross-verification on the to-be-verified cargo handling operation data set, verifies the integrity and consistency of the cargo handling operation process data by comparing the chain-like continuity of the cargo handling operation time series and the consistency of the unique hash value, and generates a blockchain notarization voucher containing a digital signature of the blockchain distributed node and binds it to the cargo handling operation process data after verification;

[0034] constructing the digital twin cargo report including a vehicle entity three-dimensional model, a cargo handling operation process time sequence chain, a cargo attribute parameter set, and a physical-virtual mapping rule based on the cargo handling operation process data including the blockchain storage certificate;

[0035] calling a quantum key distribution protocol to generate a one-time session key, encrypting the digital twin cargo report, and pushing it to the monitoring end through a quantum secure channel;

[0036] After the monitoring end decrypts the encrypted digital twin cargo report using the pairing key, it verifies the authenticity of the unique hash value and the blockchain storage certificate, and if the verification is passed, it triggers a report reception confirmation signal and updates the local database.

[0037] A port cargo handling vehicle monitoring and management system, the monitoring and management system comprises:

[0038] A cargo handling data acquisition module for real-time acquisition of multi-dimensional cargo handling feature data of port cargo handling vehicles through a port Internet of Things sensing network and synchronous transmission to an edge computing node for preprocessing, wherein the multi-dimensional cargo handling feature data includes electronic tag information, vehicle three-dimensional point cloud data, and cargo spectral feature data of the port cargo handling vehicles.

[0039] A directed graph model construction module for extracting multi-dimensional cargo handling feature vectors corresponding to the preprocessed multi-dimensional cargo handling feature data and constructing a dynamic weighted directed graph model in combination with real-time traffic situation of the port, wherein the multi-dimensional cargo handling feature vectors include vehicle identity features, cargo state features, and cargo handling operation priority of the port cargo handling vehicles.

[0040] A cargo handling path planning module for planning an optimal cargo handling driving path including space-time constraints for the port cargo handling vehicles based on the dynamic weighted directed graph model, generating an optimal cargo handling operation plan, and pushing it to a vehicle-mounted intelligent terminal through 5G vehicle networking technology.

[0041] An early warning mechanism triggering module for real-time monitoring of vehicle position and attitude change data and cargo state change data through multi-sensor fusion technology when the port cargo handling vehicles execute the optimal cargo handling operation plan, triggering a hierarchical early warning mechanism and automatically adjusting the optimal cargo handling operation plan when an abnormal event is detected.

[0042] A cargo handling report generation module for verifying the integrity and consistency of cargo handling operation process data based on a blockchain distributed ledger after the port cargo handling vehicles complete the optimal cargo handling operation plan, generating a digital twin cargo report, and synchronizing it to the monitoring end through a quantum secure channel.

[0043] An electronic device comprises a memory and a processor, the memory stores a computer program, when the processor runs the computer program stored in the memory, the processor executes the steps of the port tally vehicle monitoring management method of any one of the above.

[0044] A readable storage medium, the readable storage medium stores a computer program, the computer program is executed by a processor to implement the steps of the port tally vehicle monitoring management method of any one of the above.

[0045] Compared with the related art, the port tally vehicle monitoring management method and system provided by the application has the following beneficial effects:

[0046] The application collects multi-dimensional tallying feature data of the port tally vehicle in real time through a port Internet of Things sensing network, and synchronously transmits the multi-dimensional tallying feature data to an edge computing node for preprocessing, wherein the multi-dimensional tallying feature data comprises electronic tag information, three-dimensional point cloud data and cargo spectrum feature data of the port tally vehicle; a multi-dimensional tallying feature vector corresponding to the preprocessed multi-dimensional tallying feature data is extracted, and a dynamic weighted directed graph model is constructed in combination with a real-time traffic situation of the port, wherein the multi-dimensional tallying feature vector comprises vehicle identity features, cargo state features and tallying operation priorities of the port tally vehicle; based on the dynamic weighted directed graph model, an optimal tallying driving path containing a time-space constraint is planned for the port tally vehicle, an optimal tallying operation plan is generated and pushed to a vehicle-mounted intelligent terminal through 5G vehicle networking technology; when the port tally vehicle executes the optimal tallying operation plan, vehicle position and posture change data and cargo state change data are monitored in real time through multi-sensor fusion technology, when an abnormal event is monitored, a hierarchical early warning mechanism is triggered and the optimal tallying operation plan is automatically adjusted; after the port tally vehicle completes the optimal tallying operation plan, the completeness and consistency of tallying operation process data are verified based on a blockchain distributed ledger, a digital twin tallying report is generated and synchronously transmitted to a monitoring end through a quantum secure channel, so that efficient, safe and intelligent monitoring and management of the port tally vehicle can be realized, and the operation efficiency and safety of the port are significantly improved.

[0047] The application collects multi-dimensional cargo handling feature data in real time through a port Internet of Things sensing network, including electronic tag information, vehicle three-dimensional point cloud data and cargo spectrum feature data, and pre-processes by using an edge computing node, greatly improving the efficiency and accuracy of data processing; then, a dynamic weighted directed graph model can be constructed in combination with a real-time traffic situation of the port, to plan an optimal cargo handling driving path of the port cargo handling vehicle containing time and space constraints, the model ensures that the cargo handling vehicle completes the work in the shortest time by solving a path planning objective function, and improves the port cargo handling efficiency; when the port cargo handling vehicle executes the optimal cargo handling operation plan, the application monitors vehicle position and attitude change data and cargo state change data in real time through multi-sensor fusion technology, and once an abnormal event is monitored, a hierarchical early warning mechanism is triggered immediately, and the optimal cargo handling operation plan is automatically adjusted, effectively avoiding potential safety risks; in addition, after the port cargo handling vehicle executes the optimal cargo handling operation plan, the integrity and consistency of the cargo handling operation process data can be verified based on a blockchain distributed ledger, to generate a digital twin cargo handling report, so that the data is ensured to be tamper-proof and traceable, providing reliable data support for port management; finally, the application can synchronize the digital twin cargo handling report to the monitoring end through a quantum secure channel, and generate a one-time session key for encryption by using a quantum key distribution protocol, so as to ensure the security of the data transmission process and prevent data leakage and illegal access. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 A flow chart of a port cargo handling vehicle monitoring management method provided by an embodiment of the application is shown in

[0049] Figure 2 A system block diagram of a port cargo handling vehicle monitoring management system provided by an embodiment of the application is shown in

[0050] Figure 3 A hardware structure schematic diagram of an electronic device provided by an embodiment of the application is shown in DETAILED DESCRIPTION

[0051] To make the objectives, technical solutions and advantages of the embodiments of the application clearer, the technical solutions in the embodiments of the application will be described below in conjunction with the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0052] As Figure 1 shown is a flow chart of a port cargo handling vehicle monitoring management method provided by an embodiment of the application, Figure 1The execution subject of the method shown can be a software and / or hardware device. The execution subject of the present application can include, but is not limited to, at least one of the following: user equipment, network equipment, etc. Among them, the user equipment can include, but is not limited to, a computer, a smart phone, a personal digital assistant (PDA), and the above-mentioned electronic equipment, etc. The network equipment can include, but is not limited to, a single network server, a server group composed of multiple network servers, or a cloud composed of a large number of computers or network servers based on cloud computing, wherein cloud computing is a kind of distributed computing, which is a super virtual computer composed of a group of loosely coupled computers. The present embodiment does not make any limitation. It includes steps S1 to S5, which are as follows:

[0053] S1, real-time collection of multi-dimensional cargo handling feature data of port cargo handling vehicles through a port Internet of Things sensing network, and synchronous transmission to an edge computing node for preprocessing, wherein the multi-dimensional cargo handling feature data includes electronic tag information, vehicle three-dimensional point cloud data and cargo spectral feature data of the port cargo handling vehicles;

[0054] Among them, the port Internet of Things sensing network refers to a distributed sensing system deployed in the whole port, which is composed of multiple types of sensors and is used for real-time collection of multi-dimensional feature data such as electronic tag information, three-dimensional point cloud data and cargo spectral feature data of cargo handling vehicles, providing data support for cargo handling vehicle monitoring and management. The multi-dimensional cargo handling feature data refers to a multi-dimensional data set reflecting the port cargo handling vehicles and the working state. The edge computing node refers to a computing unit deployed near the port Internet of Things sensing network, which has the ability of local data processing and can perform preprocessing operations such as cleaning, noise reduction and format conversion on the collected multi-dimensional cargo handling feature data, reducing data transmission volume and cloud computing pressure, and improving system response speed.

[0055] The electronic tag information refers to the identity identification, ownership information and other structured data stored in the memory of the radio frequency tag carried by the vehicle or the cargo, which is used for quickly identifying the identity of the subject. The vehicle three-dimensional point cloud data refers to the three-dimensional spatial coordinate set of the vehicle generated by laser radar scanning, which can accurately represent the shape, size and spatial position of the vehicle. The cargo spectral feature data refers to the wavelength distribution information of the reflected light on the surface of the cargo collected by the spectrometer, which is used to analyze the physical properties such as material and composition of the cargo.

[0056] In practical applications, multi-dimensional feature information of the cargo handling vehicle can be captured in real time through an Internet of Things sensing network deployed in the entire port. These information specifically include identity and ownership data carried by the vehicle electronic tag, three-dimensional point cloud spatial coordinates generated by laser radar scanning, and spectral characteristics of the cargo surface reflected by the spectrometer. Then, the collected multi-dimensional cargo handling feature data can be transmitted to an edge computing node for data cleaning, noise reduction, format standardization and other preprocessing operations. At the same time, the localized processing capability of edge computing can be used to reduce data transmission pressure and improve system response efficiency.

[0057] S2, extracting a multi-dimensional cargo handling feature vector corresponding to the pre-processed multi-dimensional cargo handling feature data, and constructing a dynamic weighted directed graph model in combination with a real-time traffic situation of the port, wherein the multi-dimensional cargo handling feature vector includes vehicle identity features, cargo state features and cargo handling operation priority of the cargo handling vehicle in the port;

[0058] It can be understood that the multi-dimensional cargo handling feature vector refers to a high-dimensional numerical vector extracted from the pre-processed multi-dimensional cargo handling feature data. The real-time traffic situation of the port includes dynamic traffic information such as road traffic conditions, vehicle density, and congestion degree of the operation area in the port.

[0059] The vehicle identity features refer to a feature set for uniquely identifying the vehicle, such as license plate number, device number, etc. The cargo state features refer to feature parameters for reflecting the loading state, integrity, quantity, etc. of the cargo. The cargo handling operation priority refers to an operation sequencing index determined based on the urgency of the cargo, the port scheduling rules, etc.

[0060] Further, a multi-dimensional cargo handling feature vector representing core information can be extracted from the pre-processed multi-dimensional cargo handling feature data. The vector includes identity features for uniquely identifying the vehicle, state features reflecting the loading state and integrity of the cargo, and priority features determined based on the urgency of the operation. Then, a dynamic weighted directed graph model can be constructed in combination with the real-time traffic situation of the port. The model takes the port traffic nodes as vertices, which can be cargo handling operation areas, intersections, etc. The connection sections between the port traffic nodes are directed edges, and the weights of the edges will be dynamically adjusted according to the real-time traffic conditions and cargo handling operation demands, providing a quantitative analysis framework for subsequent path planning.

[0061] S3, planning an optimal cargo handling driving path containing time and space constraints for the cargo handling vehicle in the port based on the dynamic weighted directed graph model, generating an optimal cargo handling operation plan and pushing it to a vehicle-mounted intelligent terminal through 5G vehicle networking technology;

[0062] The 5G vehicle networking technology refers to wireless communication technology between vehicles and infrastructure based on the fifth generation mobile communication technology, supports low latency and high reliability data transmission, and is used to realize real-time information interaction between vehicles and dispatch centers and other vehicles. The vehicle-mounted intelligent terminal refers to an embedded intelligent device installed on a handling vehicle, which has data receiving, display, positioning and short-distance communication functions, and is used to receive operation plans and feedback vehicle status. The optimal handling driving path refers to a driving path planned for a port handling vehicle based on a dynamic weighted directed graph model, which meets the space-time constraints and has the minimum generalized cost in the port handling scene. The optimal handling operation plan refers to an efficient and feasible operation execution scheme formulated for a port handling vehicle based on the optimal handling driving path, which is a comprehensive planning for the whole handling process.

[0063] In practical applications, based on the constructed dynamic weighted directed graph model, the optimal driving path of the handling vehicle that meets the space-time constraints can be planned by solving the path planning objective function and minimizing the generalized cost that integrates the traffic congestion index and the dynamic reward value. The space-time constraints here include time dimension restrictions such as operation time window and road passage time limit, and also include space dimension restrictions such as road width limit and no-entry area. Then, the generated optimal handling operation plan can be pushed to the vehicle-mounted intelligent terminal in real time through the 5G vehicle networking technology to ensure that the vehicle obtains accurate operation execution scheme in time.

[0064] S4, when the port handling vehicle executes the optimal handling operation plan, the vehicle position and attitude change data and the cargo state change data are monitored in real time through multi-sensor fusion technology, and when an abnormal event is detected, a hierarchical early warning mechanism is triggered and the optimal handling operation plan is automatically adjusted;

[0065] It can be understood that the multi-sensor fusion technology refers to a technology of fusing multi-source sensor data carried by the port tally vehicle through Kalman filtering, D-S evidence theory and other algorithms, which can integrate the advantages of different sensors, offset the errors of a single sensor, improve the accuracy and robustness of vehicle position, attitude and cargo state monitoring, and ensure reliable monitoring results in complex port environment. The vehicle position and attitude change data refers to a dynamic data set for reflecting the real-time motion state of the port tally vehicle, including the spatial position coordinates of the vehicle, such as latitude and longitude, three-dimensional coordinates, etc., and attitude parameters, such as heading angle, pitch angle, roll angle, etc., for judging whether the vehicle is driving according to the planned path, whether there is deviation or abnormal motion. The cargo state change data refers to a dynamic data set for representing the real-time state change of the cargo during the tallying process, including the loading position deviation of the cargo, integrity, such as whether damaged, quantity change, surface features, etc., for judging whether the cargo is in a normal state during the transfer and loading process. The abnormal event refers to a sudden situation deviating from the normal operation state during the execution of the operation plan of the port tally vehicle, including but not limited to deviation of the vehicle from the planned path, damage or abnormal position of the cargo, vehicle failure, traffic congestion, etc. In addition, these abnormal events will seriously affect the efficiency or safety of the tallying operation. The hierarchical early warning mechanism refers to a multi-level response warning system set based on the severity of the abnormal event.

[0066] During the execution of the optimal tallying operation plan of the port tally vehicle, the dynamic information of the vehicle and the cargo can be monitored in real time by means of multi-sensor fusion technology. This technology integrates the data collected by the multi-source sensors on the vehicle, including but not limited to visual information captured by the camera, attitude parameters recorded by the gyroscope, position coordinates obtained by the GPS, three-dimensional point cloud generated by the laser radar, and spectral features of the cargo detected by the spectrometer, etc. Through Kalman filtering, D-S evidence theory and other algorithms, the data fusion processing is carried out, so as to effectively offset the measurement errors of a single sensor and improve the robustness and accuracy of the monitoring.

[0067] Among them, the real-time monitored vehicle position and attitude change data covers the real-time spatial coordinates of the vehicle, dynamic parameters such as heading angle, pitch angle, roll angle, etc., for accurately judging whether the vehicle is driving according to the planned path, whether there is deviation or abnormal motion; the cargo state change data includes the loading position deviation of the cargo, integrity, quantity change and surface spectral feature variation, etc., for real-time mastering the state of the cargo during the transfer and loading process.

[0068] When an abnormal event is monitored, such as deviation of the vehicle from the preset path, deviation or damage of the goods, equipment failure, etc., a hierarchical early warning mechanism is triggered immediately. The mechanism fuses multi-source abnormal evidence based on D-S evidence theory, calculates the abnormality index quantifying the severity of the abnormality, distributes the trust degree of each evidence source to the abnormal event and deducts the cross conflict trust degree, and then compares the abnormality index with the preset two-level warning threshold to determine the corresponding warning level.

[0069] At the same time of triggering the early warning, the adjustment process of the optimal cargo handling operation plan can be automatically started. This process is realized by setting the plan adjustment objective function and constraint conditions. The plan adjustment objective function takes minimizing the operation allocation cost and adjustment deviation as the core, and takes into account the adjustment priority of different operations; the constraint conditions ensure that each operation is only assigned to one resource, the resource load does not exceed the capacity limit, and the time nodes of the adjustment stage are consecutive and orderly. By solving the plan adjustment objective function that meets the constraint conditions, the adjusted operation plan can be generated, ensuring that the cargo handling task can be completed efficiently and safely under abnormal conditions, and realizing the dynamic adaptation and closed-loop control of the operation process.

[0070] S5, after the port cargo handling vehicle completes the optimal cargo handling operation plan, the integrity and consistency of the cargo handling operation process data are verified based on the blockchain distributed ledger, a digital twin cargo handling report is generated, and the report is synchronized to the monitoring end through a quantum secure channel.

[0071] It should be noted that the blockchain distributed ledger refers to a decentralized data storage system deployed in the port tallying scene, which maintains the chain record of the tallying operation process data through distributed nodes, and its core characteristics are non-tamperability and traceability, to ensure the integrity and consistency of the tallying process data, and provide a reliable basis for subsequent data auditing and tracing. The digital twin tallying report refers to a virtual digital mirror report constructed based on the tallying operation process data stored by the blockchain, which is a precise mapping of the physical tallying process in the virtual space, including the vehicle entity three-dimensional model, the tallying operation process time sequence chain, the cargo attribute parameter set, and the physical-virtual mapping rule. Through the spatiotemporal consistency verification, the time synchronization and spatial matching degree of the virtual model and the physical operation process are ensured, which can be used for operation review, visual management and multi-party collaborative verification, and intuitively presents the details of the whole tallying process. The quantum secure channel refers to an encrypted communication link constructed based on the quantum key distribution protocol, which is used to ensure the security of the digital twin tallying report transmission. It encrypts the report by generating a one-time session key, and uses the quantum characteristics of single-photon polarization state to transmit the key, which can theoretically resist quantum computing attacks and ensure that the report is not stolen or tampered with during transmission, providing unconditional security communication guarantee for cross-subject data synchronization. The monitoring end refers to a terminal system that receives and processes the digital twin tallying report, which is usually deployed in the port dispatch center or relevant management departments, and is the core node for realizing the whole process visual monitoring and data archiving of the tallying operation.

[0072] After the tallying vehicle completes the operation, it can automatically collect the whole process operation data, including cargo loading and unloading records, time sequence, position coordinates, and equipment interaction logs, etc., which are verified by the blockchain distributed ledger. The blockchain, with the distributed node consensus mechanism, cross- verifies the chain continuity and hash value consistency of the tallying operation process data through multiple nodes, to ensure the integrity and consistency of the tallying operation process data, and generates a blockchain storage certificate with node signature after verification. Based on these reliable data, a digital twin tallying report is constructed, which includes the vehicle three-dimensional model, the operation time sequence chain, the cargo parameter set, and the physical-virtual mapping rule. The report is a precise virtual mirror of the physical operation process. Finally, the report is encrypted and transmitted to the monitoring end through the quantum secure channel, and the monitoring end verifies the authenticity of the hash value and the blockchain storage certificate after decryption, completes data synchronization and updates the local database, forming a complete tallying management closed loop.

[0073] The multi-dimensional tallying feature data of the port tallying vehicle is collected in real time through the port Internet of Things sensing network, specifically including:

[0074] The attention weight coefficient of the sensor data in the port Internet of Things sensing network is dynamically calculated based on the multi-head attention mechanism, and a time decay factor is introduced to adjust the attention weight coefficient;

[0075] Based on the adjusted attention weight coefficient, the sensor data is weighted and summed to obtain the fused port Internet of Things perception data, i.e. the multi-dimensional cargo handling feature data:

[0076] In the formula, indicates the fused port Internet of Things perception data, i.e. the multi-dimensional cargo handling feature data; B indicates the total amount of sensor data participating in fusion in the port Internet of Things perception network; indicates the attention weight coefficient of the bth sensor data in the port Internet of Things perception network; indicates the bth sensor data in the port Internet of Things perception network; indicates the bth sensor data processed by the multi-head attention mechanism; indicates a time decay term, which is used to adjust the attention weight coefficient of the sensor data according to the difference between the collection time of the sensor data and the current time; indicates a time decay factor, ; indicates the collection time of the bth sensor data; t indicates the current time; indicates the query matrix, key matrix and value matrix of the bth sensor data, respectively; indicates the transpose of the key matrix ; indicates the dimension of the key matrix ; and Softmax() indicates an activation function.

[0077] In actual application, first, the attention weight coefficients of the sensor data in the port Internet of Things perception network can be dynamically calculated based on the multi-head attention mechanism. In this process, the query matrix, key matrix and value matrix corresponding to each sensor data need to be constructed, and the sensor data processed by the attention mechanism is generated through matrix operation and combined with the Softmax activation function, so as to quantify the importance of different sensor data in the current scene and form the initial attention weight coefficient.

[0078] Secondly, in order to optimize the timeliness of the weight coefficient, a time decay factor can be introduced to adjust the above weight coefficient. The value range of the time decay factor is 0.1 to 0.5, and the adjustment is realized by constructing a time decay term, which is calculated according to the difference between the collection time of each sensor data and the current time. The greater the difference, the more obvious the weight decay of the corresponding sensor data, so as to reduce the influence of outdated data on the fusion result.

[0079] After the dynamic calculation of the weight coefficient and the time decay adjustment are completed, the sensor data can be weighted and summed based on the adjusted attention weight coefficient. This process integrates the raw data collected by different sensors, and finally obtains the fused port Internet of Things sensing data, i.e. the multi-dimensional cargo handling feature data required. The fused data comprehensively considers the importance difference and data timeliness of different sensors, and can more accurately and comprehensively reflect the actual state of the port cargo handling vehicle and the cargo, providing a high-quality data basis for subsequent feature extraction and model construction.

[0080] The dynamic weighted directed graph model comprises a port traffic node and a port traffic connecting road section;

[0081] Based on the dynamic weighted directed graph model, a path planning objective function and a constraint condition are set, and the path planning objective function is solved to determine the optimal cargo handling driving path. The expression of the path planning objective function and the constraint condition is as follows:

[0082] In the formula, i and j represent the port traffic nodes in the dynamic weighted directed graph model; ij represents the port traffic connecting road section from the port traffic node i to the port traffic node j in the dynamic weighted directed graph model; A represents the set of all port traffic connecting road sections in the dynamic weighted directed graph model; represents the generalized cost function about the current time t; represents the road section selection decision variable, represents the selection of the port traffic connecting road section from the port traffic node i to the port traffic node j, represents the non-selection of the port traffic connecting road section from the port traffic node i to the port traffic node j, represents the selection of the port traffic connecting road section from the port traffic node j to the port traffic node i, represents the non-selection of the port traffic connecting road section from the port traffic node j to the port traffic node i; represents the path planning balance coefficient; L represents the set of all port traffic nodes in the dynamic weighted directed graph model; represents the weight coefficient of the port traffic node i; represents the deviation value related to the port traffic node i; respectively represent the starting port traffic node and the terminal port traffic node of the optimal cargo handling driving path; respectively represent the time when the port cargo handling vehicle arrives at the port traffic nodes i and j; represents the communication time of the port cargo handling vehicle from the port traffic node i to the port traffic node j, which depends on the decision variable .

[0083] It should be noted that the core component of the dynamic weighted directed graph model is the port traffic node and the port traffic connection section, wherein the port traffic node represents a key position such as a work area and an intersection in the port, and the port traffic connection section represents a passable path between the port traffic nodes, and the port traffic connection section has directionality to embody the one-way constraint of vehicle passing.

[0084] Then, based on the dynamic weighted directed graph model, a path planning objective function and a constraint condition need to be set, and by solving the path planning objective function, an optimal rational driving path is determined.

[0085] The path planning objective function takes minimizing the total cost as the core, and the total cost includes two parts: one is the sum of the generalized cost of all selected sections, and the generalized cost function is a function of the current time, reflecting the passing cost of the section at a specific time; the other is the product of the path planning balance coefficient and the weighted sum of all node deviation values, wherein the node deviation value is weighted by the node weight coefficient, and the path planning balance coefficient is used to adjust the proportion of the section cost and the node deviation in the total cost. The selection of the port traffic connection section is realized through the section selection decision variable, if the section selection decision variable takes the value of 1, it means selecting the corresponding port traffic connection section; if the section selection decision variable takes the value of 0, it means not selecting the port traffic connection section, so as to clearly define the composition of the optimal rational driving path.

[0086] The constraint condition includes two aspects: on the one hand, the flow balance constraint, for the starting node of the optimal rational driving path, the difference between the sum of the section selection decision variables flowing out of the node and the sum of the section selection decision variables flowing into the node is 1; for the terminal node, the difference is -1; for other nodes, the difference is 0, to ensure that the optimal rational driving path starts from the starting point and finally reaches the ending point without redundancy cycle. On the other hand, the timing constraint, the time of the vehicle reaching the subsequent node should not be less than the sum of the time of reaching the previous node and the section passing time between the two nodes, and the section passing time depends on the section selection decision variable, to ensure the coherence and feasibility of the optimal rational driving path in the time dimension.

[0087] By solving the path planning objective function that satisfies the above constraint condition, the optimal rational driving path with the minimum generalized cost and conforming to the space-time logic can be determined, providing efficient passing guidance for the tally vehicle.

[0088] For the port traffic connection section ij, the traffic congestion index predicted by fusing the space-time graph convolution network and the dynamic reward value generated by reinforcement learning , the generalized cost function about time t is obtained :

[0089] In the formula, denotes a fusion weight coefficient, used to adjust the traffic congestion index and the dynamic reward value In the generalized cost function , the proportion of .

[0090] , the generalized cost function of the port traffic connection road segment ij with respect to time t is constructed by fusing two types of core indicators. One type of indicator is the traffic congestion index predicted by the spatio-temporal graph convolution network, which reflects the degree of traffic congestion of the port traffic connection road segment ij at time t; the other type of indicator is the dynamic reward value generated by reinforcement learning, which is used to quantify the adaptability and efficiency benefit of the port traffic connection road segment ij in the current operation scenario.

[0091] In order to balance the contribution proportion of the two types of indicators in the generalized cost function, a fusion weight coefficient can be introduced. This coefficient dynamically adjusts the proportion of the traffic congestion index and the dynamic reward value, so that the generalized cost function can reflect both the real-time traffic pressure of the road segment and the dynamic demand of the port tally operation. The value range of the fusion weight coefficient ensures the reasonable combination of the two types of indicators, neither over-relying on the congestion condition and ignoring the operation efficiency, nor ignoring the actual traffic conditions due to the pursuit of rewards. The generalized cost function formed finally can accurately quantify the comprehensive traffic cost of the road segment ij at time t, providing a scientific cost evaluation basis for the planning of the optimal tally driving path.

[0092] The triggering step of the hierarchical early warning mechanism is as follows:

[0093] D-S evidence theory is used to fuse multi-source abnormal evidence and calculate the abnormality index , that is, the confidence degree distribution of each evidence source to the abnormal event is summarized, and the cross-conflict confidence degree is deducted, and the calculation formula of the abnormality index is as follows:

[0094] In the formula, r represents the total number of multi-source abnormal evidence participating in fusion, that is, the total number of evidence sources; A represents an abnormal event; represents the complement of the abnormal event A, that is, a non-abnormal event; k and l represent the index of the evidence source, , , ; represents the confidence degree distribution of the kth evidence source to the abnormal event A, that is, the confidence degree of the kth evidence source to the occurrence of the abnormal event A; represents the confidence degree distribution of the lth evidence source to the non-abnormal event , that is, the confidence degree of the lth evidence source to the occurrence of the non-abnormal event ;

[0095] Set two levels of warning thresholds, including the first level warning threshold and the second level warning threshold ;

[0096] According to the abnormality index The comparison result with the two-level warning threshold triggers the corresponding warning level:

[0097] .

[0098] It is understandable that we can first fuse the multi-source abnormal evidence based on the DS evidence theory and calculate the abnormality index. Specifically, we need to summarize the trust distribution of all evidence sources for abnormal events, and at the same time deduct the cross-conflict trust between different evidence sources, that is, the product of the trust of a certain evidence source for abnormal events and the trust of another evidence source for non-abnormal events, so as to obtain the abnormality index used to quantify the severity of abnormal events. Among them, the total number of evidence sources involved in the fusion is r, k and l represent the indexes of different evidence sources and are different from each other; abnormal events are recorded as A, and its complement is non-abnormal events. The k-th evidence source's trust distribution for abnormal event A reflects the degree of trust that the evidence source believes that abnormal event A has occurred, while the l-th evidence source's trust distribution for non-abnormal events reflects the degree of trust that the non-abnormal event has occurred.

[0099] Then, two levels of warning thresholds can be preset, namely the first level warning threshold and the second level warning threshold, to define abnormal events of different degrees.

[0100] Finally, the corresponding warning level can be determined based on the comparison results of the calculated abnormality index and the two-level warning threshold. When the abnormality index is less than the first-level warning threshold, the triggered warning level is 1, which is a mild warning; when the abnormality index is greater than or equal to the first-level warning threshold and less than the second-level warning threshold, the triggered warning level is 2, which is a moderate warning; when the abnormality index is greater than or equal to the second-level warning threshold, the triggered warning level is 3, which is a severe warning. Through this grading method, the corresponding warning level can be triggered according to the severity of the abnormal event, and different warning levels correspond to different disposal measures, thereby achieving accurate and graded response to abnormal events and reducing the risk impact of abnormal events.

[0101] The automatic adjustment process of the optimal tallying operation plan is as follows:

[0102] Set a plan adjustment objective function and constraints, solve the plan adjustment objective function, and obtain the adjusted optimal tallying operation plan. The expressions of the plan adjustment objective function and constraints are as follows:

[0103] wherein M represents the total number of tallying operations; N represents the total number of resources for performing the tallying operations; represents the cost of the mth tallying operation performed by the nth resource; represents a tallying operation allocation decision vector, represents that the mth tallying operation is allocated to the nth resource, represents that the mth tallying operation is not allocated to the nth resource; represents a weight coefficient of adjustment deviation of the optimal tallying operation plan, used to balance the deviation degree of actual tallying operation allocation cost and target tallying operation allocation cost; F represents the total number of adjustment stages; represents an adjustment weight coefficient of the mth tallying operation, used to distinguish the importance of different tallying operations in the adjustment process of the optimal tallying operation plan; represents the actual completion time of the mth tallying operation; represents the target completion time of the mth tallying operation; represents the processing time of the mth tallying operation; represents the capacity limit of the nth resource, i.e. the longest tallying operation processing time borne by the nth resource; respectively represent the time nodes of the f+1th and fth adjustment stages; represents an indication vector, when the mth tallying operation is processed in the fth adjustment stage, when the mth tallying operation is not processed in the fth adjustment stage, .

[0104] In actual application, the plan adjustment target function contains two core contents. One is the total cost of tallying operation allocation, i.e. the summation of all combinations of tallying operations and execution resources, the cost of each combination being the product of the cost of the mth tallying operation performed by the nth resource and the corresponding allocation decision vector, so as to quantify the direct cost of operation allocation. The other is the weighted summation of adjustment deviation, the deviation degree of actual operation allocation cost and target allocation cost being balanced through the adjustment deviation weight coefficient, while the adjustment weight coefficient of the mth tallying operation is introduced to distinguish the importance of different operations in adjustment, and then the absolute value of the difference between the actual completion time and the target completion time of each operation is combined to form a comprehensive consideration of adjustment deviation.

[0105] The constraint conditions ensure the feasibility of the plan adjustment from three dimensions. First, each tallying operation must be and only be assigned to one execution resource, that is, for each operation, the sum of all resource combination assignment decision vectors is 1, ensuring that all operations have a clear execution subject. Second, the total processing time of each operation carried by each resource must not exceed its capacity limit, that is, the sum of the product of the processing time of all operations assigned to the nth resource and the corresponding assignment decision vector does not exceed the maximum carrying time of the resource, ensuring that the resource load is within a reasonable range. Third, the time node of each adjustment stage must meet the timing constraint, and the time node of the f+1 adjustment stage cannot be earlier than the sum of the time node of the f stage and the total time of all operations processed in the stage, where whether the operation belongs to the f stage is identified by the indicator vector, thereby ensuring the continuity of the adjustment process in the time dimension.

[0106] By solving the plan adjustment objective function that satisfies the above constraint conditions, the optimal tallying operation plan can be adjusted to take into account cost control, resource adaptability and time rationality, and to achieve efficient adaptation of the operation plan in a dynamic scenario.

[0107] After the port tallying vehicle completes the optimal tallying operation plan, the integrity and consistency of the tallying operation process data are verified based on a blockchain distributed ledger, a digital twin tallying report is generated and synchronized to the monitoring end through a quantum secure channel, specifically including:

[0108] After the port tallying vehicle completes the optimal tallying operation plan, the port tallying vehicle automatically collects the tallying operation process data, including cargo loading and unloading records, tallying operation time series, vehicle position coordinate set and equipment interaction log, and generates an original tallying operation data set after standardization processing;

[0109] An SM3 hash algorithm is used to calculate a unique hash value of the original tallying operation data set, and the unique hash value is bound with the original tallying operation data set to form a to-be-verified tallying operation data set;

[0110] The to-be-verified tallying operation data set is pushed to a blockchain distributed node, and the blockchain distributed node calls an improved practical Byzantine fault tolerance consensus algorithm to perform multi-node cross-verification on the to-be-verified tallying operation data set. By comparing the chain continuity of the tallying operation time series and the consistency of the unique hash value, the integrity and consistency of the tallying operation process data are verified, and after verification, a blockchain evidence credential containing a digital signature of the blockchain distributed node is generated and bound with the tallying operation process data;

[0111] Based on the cargo handling operation data containing the blockchain storage certificate, the digital twin cargo handling report containing a vehicle entity three-dimensional model, a cargo handling operation process time sequence, a cargo attribute parameter set, and a physical-virtual mapping rule is constructed;

[0112] A quantum key distribution protocol is called to generate a one-time session key, encrypt the digital twin cargo handling report, and push it to the monitoring end through a quantum secure channel.

[0113] After the monitoring end decrypts the encrypted digital twin cargo handling report using the pairing key, it verifies the authenticity of the unique hash value and the blockchain storage certificate. If the verification is passed, a report receiving confirmation signal is triggered and the local database is updated.

[0114] After the port cargo handling vehicle completes the optimal cargo handling operation plan, it can automatically collect the cargo handling operation process data of the port cargo handling vehicle, including cargo loading and unloading records, cargo handling operation time sequence, vehicle position coordinate set, and equipment interaction log, etc. After standardization processing, the original cargo handling operation data set is formed to ensure the uniformity of data format and the completeness of content.

[0115] Then, the SM3 hash algorithm can be used to calculate the unique hash value of the original cargo handling operation data set, and the original cargo handling operation data set is bound to form a to-be-verified cargo handling operation data set. The uniqueness of the hash value can realize the preliminary check of the subsequent data integrity.

[0116] Then, the to-be-verified cargo handling operation data set can be pushed to the blockchain distributed node, and the blockchain distributed node can call the improved practical Byzantine fault tolerance consensus algorithm to carry out multi-node cross verification. In this process, each blockchain distributed node can carry out collaborative verification on the received to-be-verified cargo handling operation data set based on the preset consensus rules, focusing on comparing the chain continuity of the cargo handling operation time sequence, i.e. ensuring that the operation data forms a coherent chain in chronological order without interruption or tampering traces; at the same time, the consistency of the unique hash value is verified to ensure that the operation data has not been tampered with.

[0117] The improved practical Byzantine fault tolerance consensus algorithm can ensure that the majority of nodes reach a consensus and ensure the reliability of the verification result even in the presence of some abnormal nodes by optimizing the communication efficiency and fault tolerance mechanism of the nodes. After multi-node cross verification, if it is confirmed that the cargo handling operation process data meets the requirements of integrity and consistency, the blockchain distributed node will generate a blockchain storage certificate containing its own digital signature and bind it with the cargo handling operation process data, so as to realize the non-tamperability and traceability of the data, and provide a reliable basis for subsequent generation of digital twin cargo handling report and data synchronization.

[0118] Based on the cargo handling operation process data containing the blockchain storage certificate, a digital twin cargo handling report can be constructed, which contains a vehicle entity three-dimensional model, a cargo handling operation process time sequence chain, a cargo attribute parameter set and a physical-virtual mapping rule, so that accurate mapping of the physical operation process and the virtual model can be realized.

[0119] In order to guarantee the transmission security of the digital twin cargo handling report, a quantum key distribution protocol can be called to generate a one-time session key, encrypt the digital twin cargo handling report, and push it to the monitoring end through a quantum secure channel. After receiving the encrypted report, the monitoring end uses the paired key to decrypt it and verifies the authenticity of the unique hash value and the blockchain storage certificate. After verification, a report receiving confirmation signal is triggered, and the local database is updated, completing the data closed-loop management of the entire cargo handling operation.

[0120] As shown in Figure 2 , it is a system block diagram of a port cargo handling vehicle monitoring management system provided by an embodiment of the application. The monitoring management system comprises:

[0121] A cargo handling data acquisition module is configured to acquire multi-dimensional cargo handling feature data of a port cargo handling vehicle in real time through a port Internet of Things sensing network, and synchronously transmit the multi-dimensional cargo handling feature data to an edge computing node for preprocessing. The multi-dimensional cargo handling feature data includes electronic tag information, vehicle three-dimensional point cloud data and cargo spectral feature data of the port cargo handling vehicle.

[0122] A directed graph model construction module is configured to extract a multi-dimensional cargo handling feature vector corresponding to the preprocessed multi-dimensional cargo handling feature data, and construct a dynamic weighted directed graph model in combination with a real-time traffic situation of the port. The multi-dimensional cargo handling feature vector includes vehicle identity features, cargo state features and cargo handling operation priority of the port cargo handling vehicle.

[0123] A cargo handling path planning module is configured to plan an optimal cargo handling driving path containing time and space constraints for the port cargo handling vehicle based on the dynamic weighted directed graph model, generate an optimal cargo handling operation plan and push the optimal cargo handling operation plan to a vehicle-mounted intelligent terminal through 5G vehicle networking technology.

[0124] An early warning mechanism triggering module is configured to monitor vehicle position and attitude change data and cargo state change data in real time through multi-sensor fusion technology when the port cargo handling vehicle executes the optimal cargo handling operation plan, and trigger a hierarchical early warning mechanism and automatically adjust the optimal cargo handling operation plan when an abnormal event is detected.

[0125] A cargo handling report generation module is configured to verify the integrity and consistency of cargo handling operation process data based on a blockchain distributed ledger after the port cargo handling vehicle completes the optimal cargo handling operation plan, generate a digital twin cargo handling report and synchronously push the digital twin cargo handling report to a monitoring end through a quantum secure channel.

[0126] Figure 2 The device of the embodiment can be used to perform the steps of the method embodiment correspondingly. Figure 1 The steps in the method embodiment are similar in implementation principle and technical effects, and thus will not be described here again.

[0127] An electronic device comprises a memory and a processor, the memory stores a computer program, when the processor runs the computer program stored in the memory, the processor performs the steps of the port tally vehicle monitoring management method according to any one of the above.

[0128] As Figure 3 shown, it is a hardware structure schematic diagram of an electronic device provided by the embodiment of the present application, the electronic device 30 comprises: a processor 31, a memory 32 and a computer program; wherein

[0129] The memory 32 is used for storing the computer program, and the memory can also be a flash memory. The computer program is, for example, an application program, a functional module and the like for implementing the above method.

[0130] The processor 31 is used for executing the computer program stored in the memory to implement each step performed by the device in the above method. For details, please refer to the related description in the above method embodiment.

[0131] Optionally, the memory 32 can be independent or integrated with the processor 31.

[0132] When the memory 32 is a device independent of the processor 31, the device can further comprise:

[0133] A bus 33 is used for connecting the memory 32 and the processor 31.

[0134] A readable storage medium, the readable storage medium stores a computer program, the computer program is executed by a processor to implement the steps of the port tally vehicle monitoring management method according to any one of the above.

[0135] Wherein, the readable storage medium can be a computer storage medium or a communication medium. The communication medium includes any medium that facilitates transfer of a computer program from one place to another. The computer storage medium can be any available medium that can be accessed by a general purpose or special purpose computer. For example, the readable storage medium is coupled to the processor, so that the processor can read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (ASIC). In addition, the ASIC can be located in the user equipment. Of course, the processor and the readable storage medium can also exist as discrete components in the communication device. The readable storage medium can be read-only memory (ROM), random access memory (RAM), CD-ROM, magnetic tape, floppy diskette, and optical data storage device, etc.

[0136] The present application also provides a program product, which includes execution instructions stored in a readable storage medium. At least one processor of the device can read the execution instructions from the readable storage medium, and the at least one processor executes the execution instructions so that the device implements the method provided by the various embodiments described above.

[0137] In the embodiments of the above device, it should be understood that the processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the present application can be directly embodied as execution completed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

[0138] Through the introduction of the above embodiments, the multi-dimensional tallying feature data of the port tallying vehicle is collected in real time through the port Internet of Things sensing network, and is transmitted to an edge computing node for preprocessing in synchronization, wherein the multi-dimensional tallying feature data includes electronic tag information, three-dimensional point cloud data and cargo spectrum feature data of the port tallying vehicle; a multi-dimensional tallying feature vector corresponding to the preprocessed multi-dimensional tallying feature data is extracted, a dynamic weighted directed graph model is constructed in combination with a real-time traffic situation of the port, wherein the multi-dimensional tallying feature vector includes a vehicle identity feature, a cargo state feature and a tallying operation priority of the port tallying vehicle; based on the dynamic weighted directed graph model, an optimal tallying driving path containing a time and space constraint is planned for the port tallying vehicle, an optimal tallying operation plan is generated and pushed to a vehicle-mounted intelligent terminal through 5G vehicle networking technology; when the port tallying vehicle executes the optimal tallying operation plan, vehicle position and attitude change data and cargo state change data are monitored in real time through multi-sensor fusion technology, and when an abnormal event is monitored, a hierarchical early warning mechanism is triggered and the optimal tallying operation plan is automatically adjusted; after the port tallying vehicle completes the optimal tallying operation plan, the integrity and consistency of the tallying operation process data are verified based on a blockchain distributed ledger, a digital twin tallying report is generated and synchronized to the monitoring end through a quantum secure channel, so that efficient, safe and intelligent monitoring and management of the port tallying vehicle can be realized, and the operation efficiency and safety of the port are significantly improved.

[0139] The multi-dimensional tallying feature data is collected in real time through the port Internet of Things sensing network, including electronic tag information, three-dimensional point cloud data and cargo spectrum feature data, and preprocessing is performed by using an edge computing node, so that the efficiency and accuracy of data processing are greatly improved; then, a dynamic weighted directed graph model can be constructed in combination with a real-time traffic situation of the port, an optimal tallying driving path containing a time and space constraint is planned for the port tallying vehicle, and the model ensures that the tallying vehicle completes the operation in the shortest time by solving a path planning objective function, thereby improving the port tallying efficiency; when the port tallying vehicle executes the optimal tallying operation plan, the vehicle position and attitude change data and the cargo state change data are monitored in real time through multi-sensor fusion technology, once an abnormal event is monitored, a hierarchical early warning mechanism is triggered immediately, and the optimal tallying operation plan is automatically adjusted, so that potential safety risks are effectively avoided; in addition, after the port tallying vehicle executes the optimal tallying operation plan, the integrity and consistency of the tallying operation process data can be verified based on a blockchain distributed ledger, a digital twin tallying report is generated, so that the data is ensured to be tamper-proof and traceable, and reliable data support is provided for port management; finally, the digital twin tallying report can be synchronized to the monitoring end through a quantum secure channel, and a one-time session key is generated by using a quantum key distribution protocol for encryption, so that the safety during data transmission is ensured, and data leakage and illegal access are prevented.

[0140] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit the present application; although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions recorded in the above embodiments can be modified, or some or all of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for monitoring and managing port tally vehicles, characterized in that: The monitoring and management method includes: The multi-dimensional tally feature data of port tally vehicles is collected in real time through the port Internet of Things sensing network and synchronously transmitted to the edge computing node for preprocessing, wherein the multi-dimensional tally feature data includes the electronic tag information of the port tally vehicles, the vehicle's three-dimensional point cloud data, and the cargo spectral feature data; Extracting a multidimensional tally feature vector corresponding to the pre-processed multidimensional tally feature data, and constructing a dynamic weighted directed graph model in combination with the real-time traffic situation of the port, wherein the multidimensional tally feature vector includes the vehicle identity characteristics, cargo status characteristics, and tallying operation priority of the port tally vehicle; Based on the dynamic weighted directed graph model, an optimal tallying driving path with time and space constraints is planned for the port tallying vehicle, and an optimal tallying operation plan is generated and pushed to the on-board intelligent terminal via 5G vehicle networking technology; When the port tallying vehicle executes the optimal tallying operation plan, it monitors the vehicle position and posture change data and cargo status change data in real time through multi-sensor fusion technology. When an abnormal event is detected, a hierarchical early warning mechanism is triggered and the optimal tallying operation plan is automatically adjusted; After the port tallying vehicle completes the optimal tallying operation plan, the integrity and consistency of the tallying operation process data are verified based on the blockchain distributed ledger, and a digital twin tallying report is generated and synchronized to the monitoring end through the quantum secure channel; The automatic adjustment process of the optimal tallying operation plan is as follows: Set a plan adjustment objective function and constraints, solve the plan adjustment objective function, and obtain the adjusted optimal tallying operation plan. The expressions of the plan adjustment objective function and constraints are as follows: Where M represents the total number of tally operations; N represents the total number of resources required to perform tally operations; U mn represents the cost of the mth tallying operation performed by the nth resource; q mn represents the tally operation allocation decision vector, q mn =1 means the mth tallying operation is assigned to the nth resource, q mn =0 means that the mth tallying operation is not allocated to the nth resource; δ represents the weight coefficient of the adjustment deviation of the optimal tallying operation plan, which is used to balance the deviation between the actual tallying operation allocation cost and the target tallying operation allocation cost; F represents the total number of adjustment stages; τ m represents the adjustment weight coefficient of the mth tallying operation, which is used to distinguish the importance of different tallying operations in the adjustment process of the optimal tallying operation plan; t m represents the actual completion time of the mth tallying operation; represents the target completion time of the mth tallying operation; t' m represents the processing time of the mth tallying operation; Z n represents the capacity limit of the nth resource, that is, the maximum tally operation processing time carried by the nth resource; H f+1 、H f They represent the time nodes of the f+1th and fth adjustment stages respectively; Represents the indicator vector. When the mth tally operation is processed in the fth adjustment stage, When the mth tallying operation is not processed in the fth adjustment stage, 2. A method for monitoring and managing port tally vehicles according to claim 1, characterized in that: The real-time collection of multi-dimensional tally feature data of port tally vehicles through the port Internet of Things sensing network specifically includes: Dynamically calculating the attention weight coefficient of the sensor data in the port IoT perception network based on a multi-head attention mechanism, and introducing a time decay factor to adjust the attention weight coefficient; Based on the adjusted attention weight coefficient, the sensor data is weighted and summed to obtain the fused port IoT perception data, i.e., the multi-dimensional tally feature data: Where D fusion represents the fused port IoT perception data, i.e., multi-dimensional tally feature data; B represents the total amount of sensor data involved in the fusion in the port IoT perception network; α b represents the attention weight coefficient of the bth sensor data in the port IoT perception network; D b represents the bth sensor data in the port IoT perception network; Att(D b ) represents the b-th sensor data after being processed by the multi-head attention mechanism; represents the time decay term, which is used to adjust the attention weight coefficient of the sensor data according to the difference between the acquisition time of the sensor data and the current time; λ represents the time decay factor, 0.1<λ<0.5; t b represents the acquisition time of the bth sensor data; t represents the current time; Q b , K b 、V b Represent the query matrix, key matrix, and value matrix of the b-th sensor data respectively; K b T Denotes the bond matrix K b The transpose of Denotes the bond matrix K b Dimension; Softmax() represents the activation function.

3. A method for monitoring and managing port tally vehicles according to claim 1, characterized in that: The dynamic weighted directed graph model includes port traffic nodes and port traffic connecting sections; Based on the dynamic weighted directed graph model, a path planning objective function and constraints are set, and the path planning objective function is solved to determine the optimal tally driving path. The expressions of the path planning objective function and constraints are as follows: min∑ ij∈A c ij (t)*x ij +β*∑ i∈L oh i *p i Where i and j represent the port traffic nodes in the dynamic weighted directed graph model; ij represents the port traffic connection section from port traffic node i to port traffic node j in the dynamic weighted directed graph model; A represents the set of all port traffic connection sections in the dynamic weighted directed graph model; c ij (t) represents the generalized cost function with respect to the current time t; x ij 、x ji represents the road section selection decision variable, x ij =1 means selecting the port transportation connection section from port transportation node i to port transportation node j, x ij =0 means that the port transportation connection section from port transportation node i to port transportation node j is not selected, x ji =1 means selecting the port transportation connection section from port transportation node j to port transportation node i, x ji =0 means that the port transportation connection section from port transportation node j to port transportation node i is not selected; β represents the path planning balance coefficient; L represents the set of all port transportation nodes in the dynamic weighted directed graph model; ω i represents the weight coefficient of port traffic node i; p i represents the deviation value related to the port transportation node i; s1 and s2 represent the starting port transportation node and the ending port transportation node of the optimal tally driving path respectively; t i , t j They represent the time when the port tally vehicle arrives at the port traffic node i and j respectively; t ij (x ij ) represents the communication time of the port tally vehicle from port transportation node i to port transportation node j, which depends on the decision variable x ij .

4. A method for monitoring and managing port tally vehicles according to claim 3, characterized in that: For the port traffic connection section ij, the traffic congestion index C is predicted by integrating the spatiotemporal graph convolutional network congestion and the dynamic reward value Y generated by reinforcement learning reward , we get the generalized cost function c about time t ij (t): c ij (t)=γ*C congestion +(1-γ)*Y reward In the formula, γ represents the fusion weight coefficient, which is used to adjust the traffic congestion index C congestion and dynamic reward value Y reward In the generalized cost function c ij The proportion in (t), 0.3<γ<0.

7.

5. A method for monitoring and managing port tally vehicles according to claim 1, characterized in that: The triggering steps of the hierarchical early warning mechanism are as follows: The DS evidence theory is used to fuse multi-source abnormal evidence and calculate the abnormality index θ. That is, the trust distribution of each evidence source on the abnormal event is summarized and the cross-conflicting trust is deducted. The calculation formula of the abnormality index θ is as follows: Where r represents the total number of multi-source anomaly evidence involved in the fusion, that is, the total number of evidence sources; A represents the abnormal event; represents the complement of abnormal event A, i.e., non-abnormal events; k and l represent the index of the evidence source, 1≤k≤r, 1≤l≤r, l≠k; R k (A) represents the trust distribution of the k-th evidence source to the abnormal event A, that is, the degree of trust that the k-th evidence source believes that the abnormal event A has occurred; Indicates the lth source of evidence for non-abnormal events The trust distribution of the lth evidence source is that it is not an abnormal event the level of trust that occurs; Set two levels of warning thresholds, including the first level warning threshold θ1 and the second level warning threshold θ2; According to the comparison result of the abnormality index θ and the two-level warning threshold, the corresponding warning level is triggered:

6. A method for monitoring and managing port tally vehicles according to claim 1, characterized in that: After the port tallying vehicle completes the optimal tallying operation plan, the integrity and consistency of the tallying operation process data are verified based on the blockchain distributed ledger, and a digital twin tallying report is generated and synchronized to the monitoring end through the quantum secure channel, specifically including: After the port tallying vehicle completes the optimal tallying operation plan, the tallying operation process data of the port tallying vehicle is automatically collected, including cargo loading and unloading records, tallying operation time series, vehicle position coordinate sets, and equipment interaction logs, and an original tallying operation data set is generated after standardization processing; Calculate the unique hash value of the original tally operation dataset using the SM3 hash algorithm, and bind it with the original tally operation dataset to form the tally operation dataset to be verified; Pushing the to-be-verified tallying operation dataset to a blockchain distributed node, the blockchain distributed node calls an improved practical Byzantine fault-tolerant consensus algorithm to perform multi-node cross-validation on the to-be-verified tallying operation dataset, and verifying the integrity and consistency of the tallying operation process data by comparing the chain continuity of the tallying operation time series and the consistency of the unique hash value. After the verification is passed, a blockchain evidence certificate containing the digital signature of the blockchain distributed node is generated and bound to the tallying operation process data; Based on the tallying operation process data including the blockchain evidence, construct the digital twin tallying report including the vehicle entity three-dimensional model, the tallying operation process time sequence chain, the cargo attribute parameter set and the physical-virtual mapping rules; Calling the quantum key distribution protocol to generate a one-time session key, encrypting the digital twin tally report, and pushing it to the monitoring end through the quantum secure channel; After the monitoring end uses the paired key to decrypt the encrypted digital twin tally report, it verifies the authenticity of the unique hash value and the blockchain evidence certificate. After the verification is passed, it triggers the report reception confirmation signal and updates the local database.

7. A port tally vehicle monitoring and management system, applied to a port tally vehicle monitoring and management method according to any one of claims 1 to 6, characterized in that: The monitoring and management system includes: A tally data acquisition module is used to collect multi-dimensional tally feature data of port tally vehicles in real time through the port Internet of Things perception network, and synchronously transmit it to the edge computing node for preprocessing, wherein the multi-dimensional tally feature data includes the electronic tag information of the port tally vehicles, the vehicle's three-dimensional point cloud data, and the cargo spectral feature data; a directed graph model construction module, configured to extract a multidimensional tally feature vector corresponding to the preprocessed multidimensional tally feature data, and construct a dynamic weighted directed graph model based on the real-time port traffic situation, wherein the multidimensional tally feature vector includes the vehicle identity characteristics, cargo status characteristics, and tallying operation priority of the port tally vehicle; A tallying route planning module is used to plan an optimal tallying route with time and space constraints for the port tallying vehicle based on the dynamic weighted directed graph model, generate an optimal tallying operation plan, and push it to the on-board intelligent terminal via 5G vehicle networking technology; An early warning mechanism triggering module is used to monitor vehicle position and posture change data and cargo status change data in real time through multi-sensor fusion technology when the port tallying vehicle executes the optimal tallying operation plan. When an abnormal event is detected, a hierarchical early warning mechanism is triggered and the optimal tallying operation plan is automatically adjusted; The tallying report generation module is used to verify the integrity and consistency of the tallying operation process data based on the blockchain distributed ledger after the port tallying vehicle completes the optimal tallying operation plan, generate a digital twin tallying report, and synchronize it to the monitoring end through the quantum secure channel.

8. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor runs the computer program stored in the memory, the processor executes the steps of the port tally vehicle monitoring and management method according to any one of claims 1 to 6.

9. A readable storage medium storing a computer program, wherein: When the computer program is executed by a processor, it is used to implement the steps of a port tally vehicle monitoring and management method according to any one of claims 1 to 6.

Images