River-sea combined transportation transit hub intelligent control method and system

Abstract

The invention relates to the technical field of intelligent scheduling, and particularly discloses an intelligent control method and system for a river-sea combined transportation transit hub. According to the method, a data processing algorithm based on deep learning is adopted to carry out deep conjoint analysis on transfer operation request information of each to-be-operated inland ship of a transfer hub and operation resource occupation state information of a port; the transfer operation priority of each inland ship is intelligently determined by comprehensively considering the transfer date requirements, cargo types and loading and unloading requirements of the inland ships and the occupation states of operation resources such as berths, loading and unloading equipment and storage space of a port, and the operation resources of the port are dynamically scheduled according to the transfer operation priority, so that the transfer operation efficiency is maximized. In this way, efficient configuration and utilization of port operation resources can be achieved, the waiting time of transfer operation is shortened, and the overall logistics efficiency is improved.

Description

Technical Field

[0001] This application relates to the technical field of greenhouse gas detection, and more specifically, to an intelligent control method and system for a river-sea intermodal transfer hub. Background Art

[0002] As an efficient logistics transportation mode, river-sea intermodal transportation plays a crucial role in the global logistics system. Among them, the river-sea intermodal transfer hub, as a key node connecting inland waterway transportation and sea transportation, its operation efficiency directly affects the smoothness and economy of the entire intermodal transportation chain. With the booming development of global trade and the continuous increase in the volume of inland waterway and sea transportation business, the transfer hub is facing increasing operating pressure, which poses higher requirements for its operating efficiency and intelligent control level. Under the traditional operation mode of the river-sea intermodal transfer hub, the priority arrangement of the transfer operations of inland vessels often generates a scheduling plan based on preset rules (such as earliest due date first, cargo value weighting method). However, the port operation resources (berths, loading and unloading equipment, storage yards, etc.) are in a state of staggered occupation in time and space, making it difficult for the traditional scheduling method based on preset rules to comprehensively consider the complex relationship between operation requests and resource characteristics, which may lead to problems such as unreasonable resource allocation and low operation efficiency.

[0003] Therefore, an optimized intelligent control method and system for the river-sea intermodal transfer hub are expected. Summary of the Invention

[0004] To solve the above technical problems, this application is proposed. The embodiments of this application provide an intelligent control method and system for a river-sea intermodal transfer hub, which uses a data processing algorithm based on deep learning to deeply and jointly analyze the transfer operation request information of each inland vessel waiting for operation in the transfer hub and the occupation status information of the port operation resources, so as to comprehensively consider the transfer date requirements, cargo types, loading and unloading requirements of inland vessels, and the occupation status of port operation resources such as berths, loading and unloading equipment, and storage space, and intelligently determine the transfer operation priority of each inland vessel, and accordingly dynamically schedule the port operation resources to maximize the transfer operation efficiency. In this way, the efficient allocation and utilization of port operation resources can be realized, the waiting time for transfer operations can be reduced, and the overall logistics efficiency can be improved.

[0005] Correspondingly, according to one aspect of this application, an intelligent control method for a river-sea intermodal transfer hub is provided, which includes: Obtain the transfer operation request information of all inland vessels waiting for operation to obtain a set of transfer operation request information, where the transfer operation request information includes vessel identification, vessel basic attributes, arrival time, cargo type, cargo volume, loading and unloading requirements, and transfer date requirements; Retrieve the port operation resource occupancy status information, where the port operation resource occupancy status information includes berth occupancy status information, handling equipment occupancy status information, yard occupancy status information, warehouse storage status information, and human resource allocation status information; Based on the set of the transshipment operation request information and the port operation resource occupancy status information, calculate the execution priority scores for each transshipment operation request; According to the execution priority scores of each transshipment operation request, allocate port operation resources in sequence to execute the corresponding transshipment operation requests, where the port operation resources include berth resources, handling equipment resources, yard resources, warehouse storage resources, and human resources.

[0006] According to another aspect of the present application, there is provided an intelligent control system for a river-sea intermodal transfer hub, which includes: A transshipment operation request acquisition module, configured to acquire the transshipment operation request information of all inland river vessels waiting for operations to obtain a set of transshipment operation request information, where the transshipment operation request information includes vessel identification, vessel basic attributes, arrival time, cargo type, cargo volume, handling requirements, and transfer date requirements; An operation resource information acquisition module, configured to retrieve the port operation resource occupancy status information, where the port operation resource occupancy status information includes berth occupancy status information, handling equipment occupancy status information, yard occupancy status information, warehouse storage status information, and human resource allocation status information; A priority calculation module, configured to calculate the execution priority scores for each transshipment operation request based on the set of the transshipment operation request information and the port operation resource occupancy status information; An operation resource allocation module, configured to allocate port operation resources in sequence according to the execution priority scores of each transshipment operation request to execute the corresponding transshipment operation requests, where the port operation resources include berth resources, handling equipment resources, yard resources, warehouse storage resources, and human resources.

[0007] Compared with the prior art, the intelligent control method and system for a river-sea intermodal transfer hub provided by the present application adopt a data processing algorithm based on deep learning to perform in-depth joint analysis on the transshipment operation request information of each inland river vessel waiting for operations in the transfer hub and the port operation resource occupancy status information, so as to comprehensively consider the transfer date requirements, cargo types, handling requirements of inland river vessels, and the occupancy status of port operation resources such as berths, handling equipment, and storage spaces, intelligently determine the transshipment operation priorities of each inland river vessel, and dynamically schedule the port operation resources accordingly to maximize the transshipment operation efficiency. In this way, the efficient configuration and utilization of port operation resources can be achieved, the waiting time for transshipment operations can be reduced, and the overall logistics efficiency can be improved. Brief Description of the Drawings

[0008] The above and other objects, features, and advantages of the present application will become more apparent by describing the embodiments of the present application in more detail with reference to the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. They are used together with the embodiments of the present application to explain the present application and do not constitute a limitation to the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0009] Figure 1 It is a flowchart of an intelligent control method for a river-sea intermodal transfer hub according to an embodiment of the present application.

[0010] Figure 2 It is a flowchart of step S3 in the intelligent control method for a river-sea intermodal transfer hub according to an embodiment of the present application.

[0011] Figure 3 It is a schematic diagram of data flow in step S3 of the intelligent control method for a river-sea intermodal transfer hub according to an embodiment of the present application.

[0012] Figure 4 It is a flowchart of step S31 in the intelligent control method for a river-sea intermodal transfer hub according to an embodiment of the present application.

[0013] Figure 5 It is a flowchart of step S32 in the intelligent control method for a river-sea intermodal transfer hub according to an embodiment of the present application.

[0014] Figure 6 It is a block diagram of an intelligent control system for a river-sea intermodal transfer hub according to an embodiment of the present application. Detailed Embodiments

[0015] Next, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited by the exemplary embodiments described herein.

[0016] Figure 1 It is a flowchart of an intelligent control method for a river-sea intermodal transfer hub according to an embodiment of the present application. As Figure 1As shown in the figure, the intelligent control method for the river-sea combined transport transfer hub according to the embodiments of the present application includes the steps of: S1, obtaining the transfer operation request information of all inland river vessels waiting for operation to obtain a set of transfer operation request information, where the transfer operation request information includes vessel identification, basic vessel attributes, arrival time, cargo type, cargo volume, loading and unloading requirements, and transfer date requirements; S2, retrieving the port operation resource occupancy status information, where the port operation resource occupancy status information includes berth occupancy status information, loading and unloading equipment occupancy status information, yard occupancy status information, warehouse storage status information, and human resource allocation status information; S3, calculating the execution priority scores of each transfer operation request based on the set of transfer operation request information and the port operation resource occupancy status information; S4, sequentially allocating port operation resources according to the execution priority scores of each transfer operation request to execute the corresponding transfer operation request, where the port operation resources include berth resources, loading and unloading equipment resources, yard resources, warehouse storage resources, and human resources.

[0017] In the above intelligent control method for the river-sea combined transport transfer hub, in step S1, the transfer operation request information of all inland river vessels waiting for operation is obtained to obtain a set of transfer operation request information, where the transfer operation request information includes vessel identification, basic vessel attributes, arrival time, cargo type, cargo volume, loading and unloading requirements, and transfer date requirements. It should be understood that the river-sea combined transport hub needs to process a large number of transfer requests of inland river vessels every day, and there are problems of data fragmentation in the traditional manual input method. Through standardized data collection, the present application constructs a multi-dimensional feature space covering information such as the basic physical attributes of vessels (draft depth, vessel type dimensions), business attributes (cargo type, loading and unloading technology), and time limit constraints (transfer deadline), so as to form transfer operation request information data in a unified format, which helps to provide a standardized and high-quality information basis for subsequent transfer operation scheduling.

[0018] In the above intelligent control method for the river-sea intermodal transfer hub, in step S2, the port operation resource occupancy status information is retrieved. The port operation resource occupancy status information includes berth occupancy status information, handling equipment occupancy status information, yard occupancy status information, warehouse storage status information, and human resource allocation status information. It should be understood that the berth occupancy status information includes the current usage of the berth, the estimated idle time, and the physical attributes of the berth (such as length, depth, etc.), which is particularly important for arranging large or special types of ships. The handling equipment occupancy status information details the current occupancy status and the estimated available time of various types of handling equipment, helping to intelligently allocate equipment according to handling requirements. The yard occupancy status information covers the cargo storage situation, the remaining space, and the cargo in-and-out plan of the yard, ensuring that the cargo can be stacked orderly and transferred in a timely manner. The warehouse storage status information provides key data such as the storage capacity of the warehouse, the distribution of cargo types, and the inventory turnover rate, providing a decision-making basis for cargo storage and outbound. The human resource allocation status information reflects in real time the working status, skill level, and deployability of port staff, ensuring the efficient utilization of human resources and meeting various requirements of the transfer operation. By comprehensively obtaining the above information, it helps to perceive the spatio-temporal occupancy status of port resources in real time, realize the comprehensive monitoring and intelligent scheduling of port operation resources, and provide dynamic environment perception input for subsequent priority scoring.

[0019] In the above intelligent control method for the river-sea intermodal transfer hub, in step S3, based on the set of the transfer operation request information and the port operation resource occupancy status information, the execution priority score of each transfer operation request is calculated. It should be understood that traditional river-sea intermodal transportation scheduling overly relies on simple rules and cannot quantify the non-linear trade-offs between multiple objectives (such as accelerating the turnover of high-value goods vs. avoiding the cost of quay crane idling), making it difficult to meet the diverse ship operation requirements. For example, only considering the cargo type (such as refrigerated goods and dangerous goods having priority), however, in some cases, the remaining operation resources at the port (such as specific types of handling equipment or berths) may be insufficient, resulting in these ships being unable to complete the operation in a timely manner even though they have a high priority. At the same time, other ships with a smaller cargo volume and simple handling requirements may be delayed due to waiting for high-priority ships to complete the operation, and the remaining operation resources at the port are not fully utilized, leading to a waste of resources and a decline in the overall efficiency of the transfer operation. Therefore, this application further introduces a data processing algorithm based on deep learning to utilize the powerful learning ability of the neural network to capture the complex associations and potential laws between the transfer operation request information and the port operation resource occupancy status information, so as to comprehensively consider various factors (such as cargo value, transfer urgency, port resource matching degree, etc.), achieve a more reasonable priority ranking, and ensure the efficient execution of the transfer operation and the maximization of resource utilization.

[0020] Figure 2It is a flowchart of step S3 in the intelligent control method for the river-sea combined transport transfer hub according to the embodiment of the present application. Figure 3 It is a schematic diagram of data flow in step S3 in the intelligent control method for the river-sea combined transport transfer hub according to the embodiment of the present application. As Figure 2 and Figure 3 shown, the step S3 includes: S31, performing embedding encoding on each transfer operation request information in the set of transfer operation request information and the port operation resource occupancy status information to obtain a set of transfer operation request feature embedding encoding vector sets and port operation resource feature embedding encoding vectors; S32, respectively performing feature-driven operation request-resource global association interaction on each transfer operation request feature embedding encoding vector in the set of transfer operation request feature embedding encoding vectors and the port operation resource feature embedding encoding vectors to obtain a set of transfer operation request-port operation resource feature global significant interaction association matrices; S33, respectively performing feature decoding on each transfer operation request-port operation resource feature global significant interaction association matrix in the set of transfer operation request-port operation resource feature global significant interaction association matrices to obtain the execution priority scores of the respective transfer operation requests.

[0021] Specifically, in the step S31, embedding encoding is performed on each transfer operation request information in the set of transfer operation request information and the port operation resource occupancy status information to obtain a set of transfer operation request feature embedding encoding vector sets and port operation resource feature embedding encoding vectors. Figure 4 It is a flowchart of step S31 in the intelligent control method for the river-sea combined transport transfer hub according to the embodiment of the present application. As Figure 4 shown, the step S31 includes: S311, performing embedding encoding processing based on the Bert model on each transfer operation request information in the set of transfer operation request information to obtain the set of transfer operation request feature embedding encoding vectors; S312, performing embedding encoding processing based on the Bert model on the port operation resource occupancy status information to obtain the port operation resource feature embedding encoding vectors.

[0022] Specifically, in step S311, each transfer operation request information in the set of transfer operation request information is subjected to embedding encoding processing based on the Bert model to obtain a set of transfer operation request feature embedding encoding vectors. It should be understood that the transfer operation request information covers multi-dimensional data such as ship identification and cargo type. Traditional feature engineering (such as One-hot encoding and manually defined weights) is difficult to capture the semantic associations of multi-modal data in ship operation requests. For example, there is an implicit dependency between the cargo type (such as "cold chain medicine") and the loading and unloading requirements ("constant temperature power supply duration"). Manually preset rules cannot quantify such non-linear interactions, resulting in insufficient information utilization. In response to this, the present application introduces the Bert model to perform deep semantic encoding on each transfer operation request information in the set of transfer operation request information. Specifically, through pre-training on a large-scale corpus, the Bert model has powerful semantic understanding and representation capabilities. In the present application, the Bert model learns the context-aware representation of the transfer operation request text through the bidirectional self-attention mechanism (Self-Attention), converts the unstructured / semi-structured ship operation request information into high-dimensional dense vectors, and at the same time retains the semantic hierarchical features and cross-field correlations to obtain a set of transfer operation request feature embedding encoding vectors.

[0023] Specifically, in step S312, the port operation resource occupancy status information is subjected to embedding encoding processing based on the Bert model to obtain the port operation resource feature embedding encoding vector. Similarly, the port resource status has strong spatio-temporal dependence, and traditional numerical encoding (such as binary occupancy flags) loses the topological associations between resources. Here, the present application also adopts an embedding encoding method based on the Bert model to process the port operation resource occupancy status information, so as to utilize the deep semantic understanding ability of the Bert model to capture the collaborative associations between port operation resources, and encode discrete resource status events (such as "berth A: occupied (remaining 15 minutes), quay crane 3: idle, yard B2: occupancy rate 82%") into continuous vectors, generating port operation resource feature embedding encoding vectors for characterizing the collaborative constraints and state evolution trends between resources.

[0024] Specifically, in step S32, each of the transfer operation request feature embedding encoding vectors in the set of transfer operation request feature embedding encoding vectors is subjected to a feature-driven operation request-resource global association interaction with the port operation resource feature embedding encoding vector to obtain a set of transfer operation request-port operation resource feature global significant interaction association matrices. It should be understood that due to the multi-granularity characteristics of the association between ship requests and resource statuses (such as the need to consider global resource competition and local equipment adaptation simultaneously), traditional dot product similarity calculations cannot distinguish the importance of different interaction patterns. Based on this, the present application proposes a feature-driven operation request-resource global association interaction method. By performing principal component analysis (PCA) dimensionality reduction on the transfer operation request feature embedding encoding vector and the port operation resource feature embedding encoding vector respectively, key feature components are extracted to achieve directional coupling analysis of demand-resource features. By mining the non-linear relationships and potential correlations between interaction features, a global association map between operation requests and resource statuses is constructed, revealing the deep dependencies and constraints between each attribute in the ship request (such as cargo type) and each dimension of the resource status (such as the remaining capacity of the yard), so as to accurately depict the complex association relationship between the transfer operation request and the port operation resource, and obtain a set of transfer operation request-port operation resource feature global significant interaction association matrices.

[0025] Figure 5 It is a flowchart of step S32 in the intelligent control method for the river-sea intermodal transfer hub according to an embodiment of the present application. As Figure 5 shown, step S32 includes: S321, performing kernel association encoding based on feature principal components on the transfer operation request feature embedding encoding vector and the port operation resource feature embedding encoding vector to obtain a set of kernel association encoding vectors between the transfer operation request-port operation resource feature principal components; S322, performing global significant interaction aggregation based on the correlation topological structure on the set of kernel association encoding vectors between the transfer operation request-port operation resource feature principal components to obtain the transfer operation request-port operation resource feature global significant interaction association matrix.

[0026] More specifically, step S321 includes: First, performing feature principal component analysis on the transfer operation request feature embedding encoding vector and the port operation resource feature embedding encoding vector to obtain a set of transfer operation request feature principal component encoding vectors and a set of port operation resource feature principal component encoding vectors, which is expressed by the formula: where represents the transfer operation request feature embedding encoding vector, denotes the principal component analysis network, is the covariance matrix of the transfer operation request feature embedding coding vectors, denotes the matrix composed of the set of the transfer operation request feature principal component coding vectors obtained by performing eigenvalue decomposition on , denotes transpose, denotes the diagonal matrix composed of the set of the transfer operation request principal component eigenvalues obtained by performing eigenvalue decomposition on , and respectively denote the first and the th eigenvalues in the set of the transfer operation request principal component eigenvalues, , and respectively denote the first, the second and the th transfer operation request feature principal component coding vectors in , is the number of the transfer operation request feature principal component coding vectors, denotes the port operation resource feature embedding coding vectors, is the covariance matrix of the port operation resource feature embedding coding vectors, denotes the matrix composed of the set of the port operation resource feature principal component coding vectors obtained by performing eigenvalue decomposition on , denotes transpose, denotes the diagonal matrix composed of the set of the port operation resource principal component eigenvalues obtained by performing eigenvalue decomposition on , and respectively denote the first and the th eigenvalues in the set of the port operation resource principal component eigenvalues, , and respectively denote the first, the second and the th port operation resource feature principal component coding vectors in .

[0027] Here, considering that the characteristic dimensions of the ship operation request and the port resource status are relatively high, directly using the original features will lead to an explosion in computational complexity and there are problems of multicollinearity (for example, there are strong correlations between features such as the draft of the ship and the water depth of the berth, and the type of goods and the model of the loading and unloading equipment). Traditional direct linear correlation analysis is vulnerable to interference from redundant information. Therefore, in this application, orthogonal transformation is performed on the embedded coding vectors of the transfer operation request features and the embedded coding vectors of the port operation resource features to extract the principal components, realizing feature space decoupling and noise reduction, and obtaining a set of principal component coding vectors of the transfer operation request features and a set of principal component coding vectors of the port operation resource features, thereby providing a redundant-free base space for subsequent correlation interaction analysis.

[0028] Next, each corresponding principal component coding vector of the transfer operation request features and the principal component coding vector of the port operation resource features in the set of principal component coding vectors of the transfer operation request features and the set of principal component coding vectors of the port operation resource features are respectively input into the principal component kernel correlation coding network to obtain a set of kernel correlation coding vectors between the principal components of the transfer operation request-port operation resource features, which is expressed by the formula: Among them, represents the th principal component coding vector of the transfer operation request features in represents the th principal component coding vector of the port operation resource features in ， represents the norm of the vector, and are different weighting parameters, represents and the kernel correlation coding vector between the principal components of the transfer operation request-port operation resource features.

[0029] That is, because there is a complex non-linear relationship between the principal components of the ship request and the port resources (such as the exponential temperature correlation of cold chain goods that need to match the refrigerated yard), in order to accurately capture the complex non-linear relationship between the two, this application designs a principal component kernel correlation coding network. By using the kernel method (such as polynomial kernel, etc.), the high-dimensional principal component features are mapped to a higher-dimensional feature space, making the originally linearly inseparable relationship linearly separable, thereby effectively capturing the non-linear interaction between the principal components of the transfer operation request features and the port operation resource features, and obtaining a set of kernel correlation coding vectors between the principal components of the transfer operation request-port operation resource features.

[0030] More specifically, step S322 includes: First, calculate the performance operator between any two kernel correlation coding vectors of the transfer operation request-port operation resource feature principal components in the set of kernel correlation coding vectors of the transfer operation request-port operation resource feature principal components to obtain the transfer operation request-port operation resource feature performance operator correlation topology matrix. In a specific example of the present application, calculate the position-difference vector between any two kernel correlation coding vectors of the transfer operation request-port operation resource feature principal components in the set of kernel correlation coding vectors of the transfer operation request-port operation resource feature principal components, and take the square root of the sum of the squares of each eigenvalue in the position-difference vector as the performance operator between the two kernel correlation coding vectors of the transfer operation request-port operation resource feature principal components. It is expressed by the formula: Where , , and respectively represent the 1st, th, th, and th kernel correlation coding vectors between the transfer operation request-port operation resource feature principal components in the set of kernel correlation coding vectors of the transfer operation request-port operation resource feature principal components, represents the eigenvalue at the th position in , represents the eigenvalue at the th position in

[0031]

[0032] is the characteristic scale value of the kernel correlation coding vector of the transfer operation request-port operation resource feature principal component, represents the performance operator calculation function, represents the transfer operation request-port operation resource feature performance operator correlation topology matrix.

[0031] Here, in order to further quantify the synergistic effect of ship-resource associations in different dimensions on the global scheduling goal, the present application defines a performance operator to evaluate the potential correlation between different operation requests and resource combinations, and constructs a global association map between operation requests and resource states, that is, the transfer operation request-port operation resource feature performance operator correlation topology matrix. Among them, each element value in the transfer operation request-port operation resource feature performance operator correlation topology matrix reflects the association strength and collaborative efficiency between the corresponding two operation requests and resource combinations, providing a key basis for subsequent intelligent scheduling.

[0032] Next, based on the transfer operation request-port operation resource characteristic performance operator association topology matrix, the set of kernel association coding vectors between the principal components of the transfer operation request-port operation resource characteristic is subjected to global significant interaction aggregation coding to obtain the transfer operation request-port operation resource characteristic global significant interaction correlation matrix. In a specific example of the present application, the set of kernel association coding vectors between the principal components of the transfer operation request-port operation resource characteristic and the transfer operation request-port operation resource characteristic performance operator association topology matrix are input into the graph convolutional neural network model to obtain the transfer operation request-port operation resource characteristic global significant interaction correlation matrix, which is expressed as: in, is a graph convolutional neural network model, It represents the global significant interaction correlation matrix of the transit operation request-port operation resource characteristics.

[0033] It should be understood that the graph convolutional neural network model has significant advantages in processing graph structure data, and can capture the complex relationships between nodes and mine potential interaction patterns. In this application, the transfer operation request-port operation resource feature performance operator association topology matrix is ​​used as the input graph, and each transfer operation request-port operation resource feature principal component kernel association coding vector is regarded as a node in the graph, and the edge weights between nodes are determined by the element values ​​in the transfer operation request-port operation resource feature performance operator association topology matrix. Through the iterative propagation mechanism of the graph convolutional neural network model, the node features can aggregate information from neighboring nodes, so that the global state information can be aggregated through multi-layer graph convolution, and the transfer operation request-port operation resource feature global significant interaction association matrix considering the multi-body high-order interaction of ships and resources is generated, which provides an effective basis for subsequent intelligent scheduling.

[0034] Specifically, in a preferred example of the present application, based on the transfer operation request-port operation resource characteristic performance operator association topology matrix, a global significant interaction aggregation encoding is performed on the set of kernel association encoding vectors between the transfer operation request-port operation resource characteristic principal components to obtain the transfer operation request-port operation resource characteristic global significant interaction association matrix, including: based on the transfer operation request-port operation resource characteristic performance operator association topology matrix, performing a spatial distribution topology calibration on each kernel association encoding vector between the transfer operation request-port operation resource characteristic principal components in the set of kernel association encoding vectors between the transfer operation request-port operation resource characteristic principal components to obtain an optimized set of kernel association encoding vectors between the transfer operation request-port operation resource characteristic principal components; inputting the optimized set of kernel association encoding vectors between the transfer operation request-port operation resource characteristic principal components and the transfer operation request-port operation resource characteristic performance operator association topology matrix into a graph convolutional neural network model to obtain the transfer operation request-port operation resource characteristic global significant interaction association matrix.

[0035] Here, due to the complex dynamic coupling relationship between the ship request characteristics and the port resource characteristics, the original kernel association encoding (i.e., the kernel association encoding vector between the transfer operation request-port operation resource characteristic principal components) is prone to spatial distribution fragmentation (such as the discrete distribution of the embedding vectors of similar operation requests) under the influence of the random potential field, resulting in the graph convolutional neural network being difficult to effectively capture the true association topology. Based on this, the present application further reconstructs the characteristic manifold of the kernel association encoding vector between the transfer operation request-port operation resource characteristic principal components through differential geometry constraints.

[0036] Specifically, first, construct a spatial transition calibration matrix to serve as the spatial standard transition gauge field. In order to correct the spatial distribution disorder problem caused by the random potential field in the spatial transition calibration matrix , first, multiply each kernel association encoding vector between the transfer operation request-port operation resource characteristic principal components with the corresponding initialized spatial transition calibration matrix to obtain a compactified vector , and then two-dimensionally splice the compactified vectors corresponding to each kernel association encoding vector between the transfer operation request-port operation resource characteristic principal components to obtain a compactified matrix : : where represents the matrix multiplication operation, and represents Initial space transition calibration matrix denote Corresponding kernel correlation coding vector between the compactified transfer operation request - port operation resource feature principal components denote the kernel correlation coding matrix between the compactified transfer operation request - port operation resource feature principal components

[0037] In this way, by calculating the matrix and the transfer operation request - port operation resource feature performance operator associated topological matrix The Gaussian correlation coefficient between them is calculated, and the space transition calibration matrix is iterated by making the Gaussian correlation coefficient tend to zero wherein denote position difference is the natural exponential function denote the F - norm of the matrix denote and the variance of the set composed of all matrix values of

[0038] Thus, the kernel correlation coding vector between the compactified transfer operation request - port operation resource feature principal components is optimized by the iterated space transition calibration matrix wherein is the iterated space transition calibration matrix corresponding to denote the optimized kernel correlation coding vector between the compactified transfer operation request - port operation resource feature principal components corresponding to

[0039] In this way, during the calculation process of the graph convolutional neural network model, the problem of spatial distribution disorder brought by the kernel correlation coding vectors between the compactified transfer operation request - port operation resource feature principal components to the topological form space of the topological matrix can be solved, thus improving the calculation results of the graph convolutional neural network model

[0040] Specifically, in step S33, feature decoding is performed on each transfer operation request - port operation resource feature global significant interaction correlation matrix in the set of transfer operation request - port operation resource feature global significant interaction correlation matrices to obtain the execution priority scores of the respective transfer operation requests. Specifically, in order to convert the high - order interaction information in the transfer operation request - port operation resource feature global significant interaction correlation matrix into an executable scheduling instruction, the present application uses a feature decoding technology based on deep learning (such as a fully - connected neural network) to decode the latent information in the matrix. During the decoding process, each transfer operation request - port operation resource feature global significant interaction correlation matrix is regarded as an independent input sample, processed through the forward propagation path of the network, and the corresponding execution priority score of the transfer operation request is output. The level of the execution priority score reflects the urgency of the transfer operation request and the quality of the match with port operation resources, providing an intuitive decision - making basis for the intelligent scheduling system. After obtaining the execution priority scores of all transfer operation requests, the transfer operation requests can be sorted based on the execution priority scores, and port operation resources can be arranged in sequence for scheduling and execution.

[0041] In the above - mentioned intelligent control method for the river - sea intermodal transfer hub, in step S4, port operation resources are sequentially allocated according to the execution priority scores of the respective transfer operation requests to execute the corresponding transfer operation requests. The port operation resources include berth resources, loading and unloading equipment resources, yard resources, warehouse storage resources, and human resources. That is, for the transfer operation request with the highest score, the required port operation resources are preferentially allocated to ensure that the request can be processed in a timely manner. The port operation resources include, but are not limited to, berth resources for docking ships; loading and unloading equipment resources, such as cranes, conveyor belts, etc., for cargo loading and unloading operations; yard resources for temporarily storing goods; warehouse storage resources for long - term storage of goods; and human resources, such as dock workers, management personnel, etc., responsible for specific operation and management tasks.

[0042] In summary, the intelligent control method for the river - sea intermodal transfer hub according to the embodiments of the present application is elucidated. It uses a data - processing algorithm based on deep learning to deeply and jointly analyze the transfer operation request information of inland river vessels waiting for operations at the transfer hub and the occupation status information of port operation resources, comprehensively considering the transfer date requirements, cargo types, loading and unloading requirements of inland river vessels, and the occupation status of port operation resources such as berths, loading and unloading equipment, and storage space, intelligently determining the transfer operation priorities of each inland river vessel, and dynamically scheduling port operation resources accordingly to maximize the transfer operation efficiency. In this way, the efficient allocation and utilization of port operation resources can be achieved, the waiting time for transfer operations can be reduced, and the overall logistics efficiency can be improved.

[0043] Furthermore, the present application also provides an intelligent control system for a river-sea combined transport transfer hub.

[0044] Figure 6 It is a block diagram of the intelligent control system for a river-sea combined transport transfer hub according to an embodiment of the present application. As Figure 6 shown, the intelligent control system 100 for a river-sea combined transport transfer hub according to an embodiment of the present application includes: a transfer operation request acquisition module 110, configured to acquire transfer operation request information of all inland vessels waiting for operations to obtain a set of transfer operation request information, where the transfer operation request information includes vessel identification, vessel basic attributes, arrival time, cargo type, cargo volume, loading and unloading requirements, and transfer date requirements; an operation resource information acquisition module 120, configured to retrieve port operation resource occupancy status information, where the port operation resource occupancy status information includes berth occupancy status information, loading and unloading equipment occupancy status information, yard occupancy status information, warehouse storage status information, and human resource allocation status information; a priority calculation module 130, configured to calculate an execution priority score for each transfer operation request based on the set of transfer operation request information and the port operation resource occupancy status information; and an operation resource allocation module 140, configured to sequentially allocate port operation resources to execute corresponding transfer operation requests according to the execution priority scores of the respective transfer operation requests, where the port operation resources include berth resources, loading and unloading equipment resources, yard resources, warehouse storage resources, and human resources.

[0045] Here, those skilled in the art can understand that the specific operations of the various modules in the above intelligent control system for a river-sea combined transport transfer hub have been described in detail in the description of the above intelligent control method for a river-sea combined transport transfer hub, and therefore, the repeated description thereof will be omitted. Figures 1 to 5 Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

[0046] ​

Claims

1. An intelligent control method for a river-sea transport transit hub, characterized in that: include: Acquire the transshipment operation request information of all inland vessels waiting for operation to obtain a collection of transshipment operation request information, wherein the transshipment operation request information includes the ship identification, basic attributes of the ship, arrival time, cargo type, cargo capacity, loading and unloading requirements, and transshipment date requirements; Retrieving port operation resource occupation status information, wherein the port operation resource occupation status information includes berth occupation status information, loading and unloading equipment occupation status information, yard occupation status information, warehouse storage status information and human resource allocation status information; Calculating the execution priority score of each transit operation request based on the set of transit operation request information and the port operation resource occupation status information; According to the execution priority scores of the various transit operation requests, port operation resources are allocated in sequence to execute the corresponding transit operation requests, and the port operation resources include berth resources, loading and unloading equipment resources, yard resources, warehouse storage resources and human resources.

2. The intelligent control method for river-sea transport transfer hub according to claim 1 is characterized in that: Based on the set of the transfer operation request information and the port operation resource occupation status information, the execution priority score of each transfer operation request is calculated, including: Embedding and coding each transfer operation request information in the set of transfer operation request information and the port operation resource occupancy status information to obtain a set of transfer operation request feature embedded coding vectors and a port operation resource feature embedded coding vector; Each of the transfer operation request feature embedding coding vectors in the set of the transfer operation request feature embedding coding vectors is respectively subjected to feature-driven operation request-resource global correlation interaction with the port operation resource feature embedding coding vector to obtain a set of transfer operation request-port operation resource feature global significant interaction correlation matrices; Feature decoding is performed on each transfer operation request-port operation resource feature global significant interaction correlation matrix in the set of the transfer operation request-port operation resource feature global significant interaction correlation matrix to obtain the execution priority score of each transfer operation request.

3. The intelligent control method for river-sea transport transfer hub according to claim 2 is characterized in that: Each transfer operation request information in the set of transfer operation request information and the port operation resource occupancy status information are embedded and encoded to obtain a set of transfer operation request feature embedded coding vectors and a port operation resource feature embedded coding vector, including: Performing an embedded coding process based on a Bert model on each transfer operation request information in the set of transfer operation request information to obtain a set of transfer operation request feature embedded coding vectors; The port operation resource occupancy status information is subjected to embedded coding processing based on the Bert model to obtain the port operation resource feature embedded coding vector.

4. The intelligent control method for river-sea transport transfer hub according to claim 1 is characterized in that: Each of the transfer operation request feature embedding coding vectors in the set of the transfer operation request feature embedding coding vectors is respectively subjected to feature-driven operation request-resource global association interaction with the port operation resource feature embedding coding vector to obtain a set of transfer operation request-port operation resource feature global significant interaction association matrices, including: Performing kernel association coding based on feature principal components on the transit operation request feature embedding coding vector and the port operation resource feature embedding coding vector to obtain a set of kernel association coding vectors between transit operation request and port operation resource feature principal components; The set of kernel association coding vectors between principal components of the transit operation request-port operation resource characteristics is subjected to global significant interaction aggregation based on the correlation topological structure to obtain the global significant interaction association matrix of the transit operation request-port operation resource characteristics.

5. The intelligent control method for river-sea transport transfer hub according to claim 4 is characterized in that: The feature embedding coding vector of the transit operation request and the feature embedding coding vector of the port operation resource are subjected to kernel association coding based on feature principal components to obtain a set of kernel association coding vectors between the transit operation request and the port operation resource feature principal components, including: Performing feature principal component analysis on the transit operation request feature embedding coding vector and the port operation resource feature embedding coding vector to obtain a set of transit operation request feature principal component coding vectors and a set of port operation resource feature principal component coding vectors; Each corresponding group of transit operation request feature principal component coding vectors and port operation resource feature principal component coding vectors in the set of transit operation request feature principal component coding vectors and the set of port operation resource feature principal component coding vectors are respectively input into the principal component kernel association coding network to obtain the set of kernel association coding vectors between the transit operation request and port operation resource feature principal components.

6. The intelligent control method for river-sea transport transfer hub according to claim 5 is characterized in that: The set of kernel association coding vectors between principal components of the transit operation request-port operation resource features is subjected to global significant interaction aggregation based on the correlation topological structure to obtain the global significant interaction association matrix of the transit operation request-port operation resource features, including: Calculating the performance operator between any two kernel correlation coding vectors between principal components of the transit operation request-port operation resource feature in the set of kernel correlation coding vectors between principal components of the transit operation request-port operation resource feature to obtain a transit operation request-port operation resource feature performance operator correlation topology matrix; Based on the transfer operation request-port operation resource feature performance operator association topology matrix, the set of kernel association coding vectors between the principal components of the transfer operation request-port operation resource features is subjected to global significant interaction aggregation coding to obtain the transfer operation request-port operation resource feature global significant interaction association matrix.

7. The intelligent control method for river-sea transport transfer hub according to claim 6 is characterized in that: Calculating the performance operator between any two of the transfer operation request-port operation resource feature principal component kernel association coding vectors in the set of the transfer operation request-port operation resource feature principal component kernel association coding vectors to obtain a transfer operation request-port operation resource feature performance operator association topology matrix, including: Calculate the position difference vector between any two transit operation request-port operation resource feature principal component kernel association coding vectors in the set of transit operation request-port operation resource feature principal component kernel association coding vectors, and use the square root of the sum of squares of each eigenvalue in the position difference vector as the performance operator between the two transit operation request-port operation resource feature principal component kernel association coding vectors.

8. The intelligent control method for river-sea transport transfer hub according to claim 7 is characterized in that: Based on the transfer operation request-port operation resource feature performance operator association topology matrix, a set of kernel association coding vectors between principal components of the transfer operation request-port operation resource feature is subjected to global significant interaction aggregation coding to obtain the transfer operation request-port operation resource feature global significant interaction association matrix, including: The set of kernel association coding vectors between the principal components of the transit operation request-port operation resource characteristics and the performance operator association topology matrix of the transit operation request-port operation resource characteristics are input into the graph convolutional neural network model to obtain the global significant interaction association matrix of the transit operation request-port operation resource characteristics.

9. The intelligent control method for river-sea transport transfer hub according to claim 7, characterized in that: Based on the transfer operation request-port operation resource feature performance operator association topology matrix, a set of kernel association coding vectors between principal components of the transfer operation request-port operation resource feature is subjected to global significant interaction aggregation coding to obtain the transfer operation request-port operation resource feature global significant interaction association matrix, including: Based on the transfer operation request-port operation resource characteristic performance operator association topology matrix, each transfer operation request-port operation resource characteristic principal component kernel association coding vector in the set of transfer operation request-port operation resource characteristic principal component kernel association coding vectors is spatially distributed topologically calibrated to obtain a set of optimized transfer operation request-port operation resource characteristic principal component kernel association coding vectors; The set of kernel association coding vectors between the principal components of the optimized transit operation request-port operation resource characteristics and the transit operation request-port operation resource characteristic performance operator association topology matrix are input into the graph convolutional neural network model to obtain the global significant interaction association matrix of the transit operation request-port operation resource characteristics.

10. An intelligent control system for a river-sea transport transfer hub, characterized in that: include: A transfer operation request acquisition module is used to acquire the transfer operation request information of all inland vessels waiting for operation to obtain a collection of transfer operation request information, wherein the transfer operation request information includes ship identification, basic ship attributes, arrival time, cargo type, cargo capacity, loading and unloading requirements, and transshipment date requirements; An operation resource information acquisition module is used to retrieve the port operation resource occupation status information, wherein the port operation resource occupation status information includes berth occupation status information, loading and unloading equipment occupation status information, yard occupation status information, warehouse storage status information and human resource allocation status information; A priority calculation module, used to calculate the execution priority score of each transit operation request based on the set of transit operation request information and the port operation resource occupation status information; The operation resource allocation module is used to allocate port operation resources in sequence to execute the corresponding transit operation requests according to the execution priority scores of the various transit operation requests. The port operation resources include berth resources, loading and unloading equipment resources, yard resources, warehouse storage resources and human resources.

Images