CN-122020546-A - Port group multisource heterogeneous data fusion and transportation channel digitizing method and system

CN122020546ACN 122020546 ACN122020546 ACN 122020546ACN-122020546-A

Abstract

The invention belongs to the field of intelligent and integrated fusion of port planning data, and discloses a method and a system for fusing port group multisource heterogeneous data and digitizing a transportation channel, wherein the method comprises the steps of constructing a systematic data fusion frame based on port group multisource heterogeneous data, and establishing a unified space-time reference and data specification based on the systematic data fusion frame to form a fused data product; based on the fusion data product, a 'vehicle-ship aggregate' is constructed, a transportation group with operation characteristics meeting preset requirements is identified and classified based on the 'vehicle-ship aggregate', and a transportation channel meeting the preset requirements is extracted based on track data of the transportation group, so that digital and structural expression of the transportation channel is realized.

Inventors

SHEN CHEN
FANG ZHUO
HU YI
WANG DACHUAN
JI HAOJIE
TANG GUOLEI
WAN PENGCHENG
HAO JUN

Assignees

交通运输部规划研究院

Dates

Publication Date: 20260512
Application Date: 20260131

Claims (7)

1. A method for port group multisource heterogeneous data fusion and transportation channel digitization, the method comprising: Constructing a systematic data fusion framework based on port group multisource heterogeneous data, and constructing a unified space-time reference and data specification based on the systematic data fusion framework to form a fusion data product; Constructing a vehicle-ship aggregate based on the fusion data product, and identifying and classifying transportation groups with operation characteristics meeting preset requirements based on the vehicle-ship aggregate; Based on track data of transport groups, a transport channel meeting preset requirements is extracted, and digital and structural expression of the transport channel is realized.
2. The method of claim 1, wherein the method of constructing the systematic data fusion framework comprises: establishing a national 2000 geographic coordinate system and UTC time as unified space-time reference; based on the unified space-time reference, constructing a metadata model for port group service, and unifying semantic definition of key service fields; based on the semantic definition of the unified key service field, merging the multi-source repeated data by adopting a data source abundance-based deduplication algorithm; Based on a Lambda stream-batch integrated architecture, the collaborative processing of the multi-frequency data after the merging of the repeated points is realized, and the construction of the systematic data fusion framework is completed.
3. The method of claim 2, wherein the metadata model uses "entity-field-term-rule" as core logic to define entities such as ships, container wagons, railway trains, and port facilities, each entity is associated with "mandatory core field+optional extension field" to form a standardized data dictionary, and the expression is as follows Wherein, the method comprises the steps of, Is a collection of core entity ships, container trucks, railway trains and port facilities; A field set for each entity; A standard term set for an entity field; a rule set for mapping standard terms to standard terms, and a rule for iteratively updating U-based information table, As a field of the value of the number, Is that The fields of the entity are used to determine, Is that The mapping rules of the entity are set up, Is that Standard terminology for entities.
4. The method of claim 1, wherein the "car and boat aggregate" is a four-dimensional structuring element comprising "core members, association rules, business attributes, hierarchy identities" expressed mathematically as: ; wherein M is a core member set, all the vehicle examples contained are classified according to types, and the formula is expressed as follows: wherein Is a member of the ship and is provided with a plurality of control units, As a member of the i-th vessel, As a member of a container wagon, As a member of the j-th container wagon, Is a member of the class of railways, Is the first A railway train member; r is an association rule set, and defines constraint conditions of space-time cooperation and business association among members, wherein the constraint conditions are expressed as follows: wherein, the space-time collaboration rule The time difference between the arrival time of the ship and the connection time of the truck/train is less than or equal to delta t, the space distance between the connection point and the port berth/hub is less than or equal to delta s, and the service association rule is that Sharing the same container number, transportation mission number or service port cluster among members: , Wherein, the For the identification of port clusters, 、 Is the number of the container, 、 In order to serve a port cluster, 、 Is a transportation task; a is a core service attribute set, characterizes the overall functional characteristics of the aggregate, is generated by member characteristic aggregation, and is expressed as follows: Wherein RouteScope is the range of route radiation, cargoType is the transportation category, transMode is the intermodal mode, and capability is the total Capacity; L is a hierarchy identification, and is divided according to functional scale, wherein an ocean trunk group Cluster_A takes 10000-15000TEU large container ships as cores, 40 feet of heavy-load highway trunk line collection cards and 50-60 sections of grouping railway trunk line trains are matched to bear cross-intercontinental container transportation, a branch transportation group Cluster_B takes 500-3000TEU medium container ships as cores, 20-30 feet of highway branch line collection cards and 30-40 sections of grouping railway branch line trains are matched to bear container transportation of core ports and peripheral feeding ports, and a river-sea direct group Cluster_C takes 2000-5000DWT river-sea direct container ships as cores, and is matched with 15-20 feet of small highway collection cards, no railway cooperation and direct transportation of ports are borne.
5. The method according to claim 1, wherein the method of identifying and categorizing a transportation population with operational characteristics meeting preset requirements based on the "car and boat aggregate" comprises: extracting multidimensional features of transport entities from the fusion data, wherein the multidimensional features comprise space-time features, business attribute features and behavior pattern features; And performing unsupervised learning on the multidimensional features by adopting a DBSCAN density clustering algorithm to realize automatic identification and classification of transportation groups.
6. The method according to claim 1, wherein the method of extracting transport channels meeting preset requirements comprises: taking the history track of the vehicle and ship aggregate as input, and adopting a track density clustering algorithm to identify frequently used path corridor; the identified path corridors are abstracted into a digital network model, and each channel is given basic attributes, dynamic performance attributes and sustainability attributes.
7. A system for fusing multisource heterogeneous data of a port group and digitizing a transportation channel, which is used for realizing the method of any one of claims 1-6, and is characterized by comprising a fusing module, a classifying module and an extracting module; The fusion module is used for constructing a systematic data fusion frame based on port group multisource heterogeneous data, and establishing a unified space-time reference and data specification based on the systematic data fusion frame to form a fusion data product; the classifying module is used for constructing a vehicle-ship aggregate based on the fusion data product, and identifying and classifying transportation groups with operation characteristics meeting preset requirements based on the vehicle-ship aggregate; The extraction module is used for extracting the transportation channel meeting the preset requirements based on the track data of the transportation group, and realizing the digital and structural expression of the transportation channel.

Description

Port group multisource heterogeneous data fusion and transportation channel digitizing method and system Technical Field The invention belongs to the field of intelligent and integrated integration of port planning numbers, and particularly relates to a method and a system for fusing multisource heterogeneous data of port groups and digitizing a transportation channel. Background Currently, intelligent ports have become the core direction of global port transformation and upgrading, and intelligent technologies inside single-point ports, such as automatic shore bridge control, unmanned horizontal transportation and digital twin systems in harbors, have gradually entered into large-scale mature application stages. However, with the increasing demands of the global supply chain for "efficient collaboration" and "global toughness", port developments are facing strategic transitions from "single point wisdom" to "port group collaboration", from "port internal optimization" to "transportation channel global optimization". In this process, the prior art architecture exposes the following three significant limitations in terms of supporting cross-domain synergy and global decisions: (1) Data plane isomerism and information island problem The collaborative operation of the port group needs to rely on massive data of cross-main body, cross-dimension and cross-mode, including but not limited to high-frequency updated ship AIS dynamic data, structural data of a port operation scheduling system, port group space vector data and collection and dispatch data of railways, highways and the like of a port. These data differ significantly at the structural, timing, semantic level: Semantic heterogeneous is difficult to align, and different ports and different data source standards are different. Taking AIS data as an example, key fields (such as "ship arrival time") may have different definitions in different ports or service providers, and lack of unified mapping with port planning data systems, which results in a large amount of manual intervention during fusion, low efficiency and error liability. The multi-frequency data is difficult to cooperate, namely, the high-frequency data (such as AIS, updating second level) and the low-frequency data (such as space planning data, static state or annual older) are greatly different in processing paradigm, and the existing system is difficult to realize effective cooperation of real-time access and on-demand updating of the high-frequency data and the low-frequency data. (2) Model and algorithm layer localized trap and robustness loss Although data-driven optimization research is continuously emerging, such as fleet path planning by utilizing algorithms such as mixed integer programming and adaptive large neighborhood searching based on AIS data, or optimizing in-port collection card scheduling by adopting genetic algorithm, the optimization range is limited to a single link or a single transportation mode, and full-chain engagement points of 'sea transportation channel-port operation-railway transportation-highway connection' cannot be opened. In addition, existing models are mostly based on ideal assumptions, and lack robustness and adaptation capability to dynamic disturbances (e.g., extreme weather, equipment failure) that are common in transportation channels. (3) Fragmentation dilemma and evaluation blanking of application scenarios Current intelligent port technology applications present a "scene cut" feature. For example, digital twinning can better realize the joint scheduling of equipment at a container terminal, but the model granularity is insufficient at a bulk cargo terminal, and the application is more stayed at a macroscopic level. More importantly, the prior art system lacks unified definition and digital modeling standard for 'transportation channel', and cannot monitor the efficiency, toughness and sustainability (such as carbon emission) of the whole channel in real time, comprehensively evaluate and dynamically deduce. In summary, the prior art has limitations on three core layers of data fusion, algorithm optimization and scene application, and forms a technical bottleneck of mutual toggle, namely, the multi-source information cannot be effectively integrated due to data isomerization, so that the input quality of a full-chain optimization algorithm is limited, the localization of the algorithm and the fragmentation of the scene are limited, and the global management and control requirement of a transportation channel is difficult to fall to the ground. These drawbacks directly lead to poor port group synergic efficiency, insufficient overall efficiency of transportation channels and lack of toughness in practical business. Therefore, a systematic technical scheme is needed in the art, which can deeply integrate multisource heterogeneous data of a port group, and realize digital modeling and intelligent collaborative management and control of the whole process