Search

CN-121979638-A - Digital twinning-based intelligent park data processing method, device and system

CN121979638ACN 121979638 ACN121979638 ACN 121979638ACN-121979638-A

Abstract

The invention provides a digital twinning-based intelligent park data processing method, a device and a system, which are used for generating a cross-park twinning index library by receiving capability self-declaration data and resource reporting data transmitted by each park twinning node, screening cooperative participation nodes to issue encryption initial model content, obtaining a global optimization model according to feedback fusion, combing a plurality of subtask execution requirements and dependency relations, dynamically binding each subtask with the twin state of the cooperative participation node to generate a subtask distribution list and a data interaction specification, driving each cooperative participation node to execute corresponding subtasks after symbiotic calibration and evolution, acquiring real-time operation data and result data, carrying out integration analysis to generate an optimization effect data set, and driving a global optimization model and the bidirectional evolution of the twin state of each cooperative participation node to generate a global optimization model and a task distribution strategy after encryption processing. The intelligent park cross-node collaborative operation method and system can improve stability and suitability of intelligent park cross-node collaborative operation.

Inventors

  • Ye Woxing

Assignees

  • 广州市立腾智能科技有限公司

Dates

Publication Date
20260505
Application Date
20260123

Claims (10)

  1. 1. The intelligent park data processing method based on digital twinning is characterized by comprising the steps of receiving capability self-declaration data and resource reporting data transmitted by each park twinning node, calling a preset decentralizing trans-park twinning peer-to-peer interconnection architecture and an encryption interconnection gateway, and generating a trans-park twinning index library containing park twinning capability images and real-time resource load information; combining the cross-park twin index library, pre-adapting the twin state of each park twin node, screening out park twin nodes meeting the collaborative optimization target as collaborative participation nodes, issuing encryption initial model content to each collaborative participation node, receiving encryption model content fragments fed back by each collaborative participation node and generated based on local twin data training, carrying out fusion processing on the encryption model content fragments to obtain a global optimization model, analyzing park twin capacity images and real-time resource load information in the cross-park twin index library, disassembling the collaborative optimization target to obtain a plurality of subtasks, carding execution requirements and dependency relations of each subtask, dynamically binding each subtask with the twin state of the collaborative participation node, distributing subtasks based on binding results and subtask dependency relations, generating a subtask distribution list and a data interaction specification, calling the global optimization model, the subtask distribution list and the data interaction specification, carrying out symbiotic calibration and evolution on local twin decision logic corresponding to each collaborative twin nodes to drive each collaborative participation node to execute corresponding subtasks, acquiring execution requirements and real-time data of the collaborative participation nodes, analyzing the execution results of the subtask and real-time data, and after the optimization effect data set is encrypted, driving the bidirectional evolution of the global optimization model and the twin states of all the cooperative participation nodes based on the optimization effect data set, and adjusting a subtask distribution list by combining the real-time resource load change of all the cooperative participation nodes to generate an optimized global optimization model and a task distribution strategy and synchronizing the optimized global optimization model and the task distribution strategy to a local twin system corresponding to each cooperative participation node.
  2. 2. The method of claim 1, wherein the invoking the preset decentralised cross-campus twin peer-to-peer interconnect architecture and the encrypted interconnect gateway to generate a cross-campus twin index library comprising a campus twin capability image and real-time resource load information comprises mapping the capability self-declaration data and the time sequence change of the resource source reporting data to different dimensions of a state evolution track of a corresponding campus twin node respectively, and generating a mapping relation set comprising the capability self-declaration data time sequence change, the resource source reporting data time sequence change and a state evolution track multi-dimensional mapping relation; based on the mapping relation set, triggering a twin capacity portrait state unit corresponding to a park twin node to adjust, wherein the adjustment direction is the same as the trend of the time sequence change of the capacity self-declaration data, embedding the resource into the adjusted state unit from the trend of the time sequence change of the reporting data, and generating a twin capacity portrait primary version adjusted by the state unit; the method comprises the steps of correspondingly binding the current content of the resource source reporting data to a state unit of an adjusted twin capacity image primary version, enabling the execution logic of the resource source reporting data content and the state unit to form an association map, generating a bound twin capacity image association version, calling the decentralization cross-park twin peer-to-peer interconnection architecture, carrying out cross-node state synchronization on the bound twin capacity image association version of each park twin node, keeping the consistency of the execution logic of the state unit, the mapping relation of the resource source reporting data and the capacity self-declaration data in the synchronization process, generating a cross-node synchronized image set, based on the cross-node synchronized image set, combining the twinning capability images of each park twinning node and the resource report data binding content into unified index entries, wherein each entry comprises a state evolution track, state unit content, a capability self-declaration data mapping relation and resource report data binding information, generating an entry set of a cross-park twinning index library, dynamically sequencing the entry set according to the state life cycle of the corresponding park twinning node, and generating the cross-park twinning index library comprising the park twinning capability images and real-time resource load information according to the matching degree of the current stage of the state evolution track and the capability self-declaration data time sequence change.
  3. 3. The method of claim 2, wherein mapping the time sequence changes of the capability self-declaration data and the resource source reporting data to different dimensions of a state evolution track of a corresponding park twinning node respectively, generating a mapping relation set containing the time sequence changes of the capability self-declaration data and the time sequence changes of the resource source reporting data and the multidimensional mapping relation of the state evolution track, comprises extracting the time sequence change content of each field of the capability self-declaration data, mapping the change track of each field to a capability dimension change path of the state evolution track of the corresponding park twinning node, generating a corresponding relation of the capability self-declaration data field-capability dimension, extracting the time sequence change content of each field of the resource source reporting data, mapping the change track of each field to a resource dimension change path of the state evolution track of the corresponding park twinning node, generating a corresponding relation of the resource source reporting data field-resource dimension, integrating the same field from the state evolution data to the contour track based on the corresponding relation of the capability self-declaration the corresponding park twinning data field, integrating the same-capability profile track of the corresponding park twinning node, and integrating all the capability data fields to the initial contour track, the method comprises the steps of generating a resource dimension track contour set, associating the capability dimension track contour set with the resource dimension track contour set to enable capability dimension contours and resource dimension contours of twin nodes in the same park to form parallel evolution state evolution tracks, generating an associated state evolution track set, marking and associating the capability self-declaration data time sequence change and the resource self-declaration data time sequence change with corresponding dimensions of the associated state evolution tracks respectively, and generating a mapping relation set containing multi-dimensional mapping relation.
  4. 4. The method of claim 1, wherein the combining the cross-park twin index library performs pre-adaptation processing on twin states of all park twin nodes, screens out park twin nodes meeting a collaborative optimization target as collaborative participation nodes, issues encrypted initial model content to all collaborative participation nodes, receives encrypted model content fragments fed back by all collaborative participation nodes and generated based on local twin data training, performs fusion processing on the encrypted model content fragments to obtain a global optimization model, and comprises the steps of calling a park twin capacity image and real-time resource load information in the cross-park twin index library, pre-aligning a state evolution track of the park twin capacity image and a change trend of the real-time resource load information with an expected state track and an expected resource trend of the collaborative optimization target respectively, and generating a track-trend pre-alignment result set; based on the track-trend prealignment result set, screening out park twin nodes with the coincidence degree of the state evolution track and the expected state track and the coincidence degree of the resource change trend and the expected resource trend, determining the nodes as preadapted nodes, generating a preadapted node list, issuing an encryption initial model content to each preadapted node in the preadapted node list, wherein the encryption initial model content comprises capability dimension logic corresponding to the expected state track and resource dimension logic corresponding to the expected resource trend, generating a model content issuing result, receiving encryption model content fragments fed back by each preadapted node and generated based on local twin data training, and each fragment comprises park twin capability image adjustment logic corresponding to the park twin node, the method comprises the steps of reporting data adaptive logic, generating a training fragment set associated with nodes, carrying out dimension level fusion on the training fragment set associated with the nodes, carrying out cross-node aggregation on corresponding capability dimension logic in each fragment according to a preset model aggregation rule, generating an aggregated capability dimension logic set, carrying out cross-node aggregation on corresponding resource dimension logic in each fragment, generating an aggregated resource dimension logic set, binding the integrated capability dimension logic set and the resource dimension logic set according to a time sequence corresponding relation between an expected state track and an expected resource trend, and generating a global optimization model.
  5. 5. The method of claim 4, wherein the pre-aligning the state evolution trace of the park twin capability image and the change trend of the real-time resource load information with the expected state trace of the collaborative optimization target respectively to generate a trace-trend pre-alignment result set, comprises extracting a core capability stage sequence of the state evolution trace of each park twin node in the cross-park twin index library, key capability executing nodes of each stage corresponding to the trace to generate a node core capability stage sequence set, extracting a core capability stage sequence of the expected state trace of the collaborative optimization target, wherein the sequence structure is the same as the node core capability stage sequence to generate an expected core capability stage sequence, extracting a core resource stage sequence of the change trend of the real-time resource load information of each park twin node in the cross-park twin index library, generating a node core resource stage sequence set, extracting a core resource stage sequence of the expected resource trend of the collaborative optimization target, wherein the sequence structure is the same as the node core resource stage sequence, comparing the two-dimensional position of each park twin node in the cross-park twin index library with the expected resource stage sequence, comparing the two-dimensional position of each park twin node in the expected resource twin index library with the expected core resource stage sequence, comparing the two-dimensional position and the expected core resource stage sequence to the expected core resource load information respectively, comparing the two-dimensional position and the expected core resource stage sequence to the expected core capability sequence, a set of track-trend pre-alignment results is generated.
  6. 6. The method of claim 1, wherein the parsing the campus twin capability image and the real-time resource load information in the cross-campus twin index library, disassembling a collaborative optimization target to obtain a plurality of subtasks, carding execution requirements and dependency relationships of the subtasks, dynamically binding the subtasks with twin states of collaborative participation nodes, distributing the subtasks based on binding results and the subtask dependency relationships, generating a subtask distribution list and a data interaction specification, and comprising the steps of calling the campus twin capability image and the real-time resource load information in the cross-campus twin index library, disassembling the collaborative optimization target into a plurality of subtasks, wherein each subtask corresponds to one capability stage of an expected state track and one resource stage of an expected resource trend, and generating a corresponding relation set of the subtasks and the double stages; based on the corresponding relation set, combing the execution requirement of each subtask, the execution requirement is correspondingly matched with the capacity execution logic of the capacity stage and the resource consumption logic of the resource stage to generate a subtask set containing the execution requirement, combing the dependency relation among the subtasks, which is the sequence of the capacity stage corresponding to the subtask and the resource stage to generate a subtask dependency relation set, corresponding the execution requirement of each subtask to the capacity dimension logic of the park twin capacity image of the collaborative participation node and the resource dimension logic of the resource source reporting data, enabling the subtask execution logic and the node double dimension logic to form binding, generating a binding relation set of the subtask and the node, carrying out node allocation on each subtask based on the binding relation set and the subtask dependency relation set, following the double-stage sequence corresponding to the subtask in the allocation process, and adding data interaction requirements in the execution process of the subtasks into the preliminary list, wherein the requirements are the same as the information flow rule of the bound node bi-dimensional logic, and the subtask distribution list and the data interaction specification are generated.
  7. 7. The method of claim 6, wherein the method comprises the steps of associating the execution requirement of each subtask with the capability dimension logic of the park twin capability image of the cooperative participation node, decomposing the execution requirement into a plurality of capability execution action units and resource consumption action units by the execution requirement of each subtask in the subtask set, generating a set of binding relation between the subtask and the node by the resource dimension logic of the resource source reporting data, extracting the capability dimension logic of the park twin capability image of the cooperative participation node, decomposing the capability dimension logic into a plurality of capability action units, generating a set of resource consumption action units by the resource dimension logic of the resource source of the cooperative participation node, matching the resource dimension logic of the resource participation node with the resource consumption action units, and matching the resource dimension logic of the resource consumption action units in the cooperative participation node according to the matching sequence of the binding relation between the resource dimension logic units of the resource consumption action units of the cooperative participation node, wherein the capability dimension logic of the park twin capability image of the cooperative participation node is extracted to generate a plurality of capability action units, the unit structure is the same as the capability action units of the resource capability execution action units of the cooperative participation node, the node capability action unit is generated, the node capability action unit is set, the resource dimension logic of the resource unit of the resource source reporting data is generated, the resource dimension logic of the sub-role resource units of the resource source data is bound, the resource units of the cooperative participation node is generated, the resource units of the resource units are the resource units of the resource consumption action units are bound, the resource units of the sub-unit units are generated, the resource units are the resource units of the resource units are bound by the resource units, and the resource units are bound by the resource units, and are the resource units, and the resource units are bound by the resource units, and resource units are respectively corresponding units and resource units are matched, and resource units and source units and are matched, the method comprises the steps of generating a primary binding relation set, carrying out logic consistency check on the primary binding relation set, ensuring that a continuous execution chain is formed after sub-task execution logic and node two-dimensional logic are bound, generating a binding check result set, merging binding content of each sub-task in the binding check result set and a corresponding matching result into a unified item, and generating a binding relation set of the sub-task and the node.
  8. 8. The method of claim 1, wherein invoking the global optimization model, the subtask allocation list and the data interaction specification, performing symbiotic calibration and evolution on local twinning decision logic corresponding to each cooperative participation node, driving each cooperative participation node to execute corresponding subtasks, acquiring real-time running data and result data in the execution process of the subtasks, performing integrated analysis to generate an optimized effect data set comprising execution efficiency and cooperative adaptation degree, comprising invoking capacity dimension logic and resource dimension logic of the global optimization model, respectively injecting the capacity dimension logic and resource dimension logic into corresponding dimension positions of the local twinning decision logic of the corresponding cooperative participation node, enabling the global twinning decision logic and the local twinning decision logic to form symbiotic logic unit sets, calibrating each symbiotic logic unit in the symbiotic logic unit sets based on the subtask allocation list and the data interaction specification, generating a calibrated symbiotic logic unit set based on a double-dimensional information flow rule of double-action unit logic and data interaction required by the execution of the subtasks, driving each cooperative participation node to execute corresponding subtask units, generating a state-dependent data-request data-state of the corresponding to the execution of the cooperative participation nodes, and real-time data-consumption-state-dependent on the data of the cooperative participation nodes, and real-time resource-time state-required by the data-area on the execution requests of the cooperative participation nodes, the method comprises the steps of generating a result data and an execution requirement association set, extracting execution efficiency related content according to a preset efficiency calculation rule based on the real-time dual-state data set and the result data association set, extracting collaborative fitness related content according to a preset collaborative degree evaluation rule, classifying the content according to collaborative participation nodes, classifying dimensions into capability dimension efficiency and resource dimension fitness to generate an efficiency fitness related content set, carrying out item arrangement on the efficiency fitness related content set, merging two dimension related content of each node into a unified item, and generating an optimization effect data set comprising execution efficiency and collaborative fitness.
  9. 9. The data processing device is characterized by comprising a data receiving module, a data processing module and a data processing module, wherein the data receiving module is used for receiving capability self-declaration data and resource source reporting data transmitted by each park twin node, calling a preset decentralizing cross-park twin peer-to-peer interconnection architecture and an encryption interconnection gateway, and generating a cross-park twin index library containing park twin capability images and real-time resource load information; the system comprises a model construction module, a state carding module, an analysis integration module, a data interaction module, a data analysis module and a data analysis module, wherein the model construction module is used for carrying out pre-adaptation processing on the twinning state of each park twinning node in combination with the cross park twinning index library, screening out park twinning nodes which meet the cooperative optimization target as cooperative participation nodes, issuing encryption initial model content to each cooperative participation node, receiving encryption model content fragments fed back by each cooperative participation node and generated based on local twinning data training, carrying out fusion processing on the encryption model content fragments to obtain a global optimization model, the state carding module is used for analyzing park twinning capacity images and real-time resource load information in the cross park twinning index library, disassembling the cooperative optimization target to obtain a plurality of subtasks, carding execution requirements and dependency relations of each subtask, dynamically binding each subtask with the twinning state of the cooperative participation node, generating a subtask distribution list and a data interaction specification based on the binding results and the subtask dependency relations, carrying out calibration and evolution of local twinning decision logic corresponding to each cooperative participation node, carrying out the analysis and the real-time data interaction specification in the execution process of the corresponding subtask driving the cooperative participation node, the system comprises an optimization effect data set, an encryption optimization module, a sub-task allocation list, a local twinning system and a local network node, wherein the optimization effect data set comprises execution efficiency and cooperative adaptation degree, the encryption optimization module is used for driving the bidirectional evolution of a global optimization model and a twinning state of each cooperative participation node based on the optimization effect data set after encryption processing is carried out on the optimization effect data set, and the sub-task allocation list is adjusted by combining the real-time resource load change of each cooperative participation node to generate an optimized global optimization model and task allocation strategy and synchronizing the optimized global optimization model and task allocation strategy to the local twinning system corresponding to each cooperative participation node.
  10. 10. A computer system comprising a memory for storing computer executable instructions or computer programs and a processor for implementing the digital twinning-based intelligent campus data processing method of any one of claims 1 to 9 when executing the computer executable instructions or computer programs stored in the memory.

Description

Digital twinning-based intelligent park data processing method, device and system Technical Field The invention relates to the field of data processing, in particular to a digital twinning-based intelligent park data processing method, device and system. Background Along with the deep fusion of the digital twin technology and intelligent park management, collaborative optimization can be realized by constructing digital twin mapping nodes of a park physical scene and depending on cross-node data interaction. In the prior art, a centralized architecture is adopted to collect the capacity and resource data of each twin node, a unified index library is constructed, collaborative nodes are screened based on static evaluation standards, a global collaborative model is generated by aggregating basic model parameters, tasks are distributed according to subtask complexity, and finally the global model is updated unidirectionally. However, the centralized index library is difficult to adapt to real-time dynamic changes of each node, static screening standards lead to insufficient adaptability of the nodes and cooperative demands, a global model is difficult to attach to local operation logic of each node, subtask allocation cannot match with real-time states of the nodes, a unidirectional update mode cannot realize synchronous adaptation of global and local states, stability and resource utilization efficiency of cross-node cooperative operation of a park are always difficult to achieve expectations, and how to improve the adaptability and stability of cross-node cooperative operation of an intelligent park is a problem to be solved in the field. Disclosure of Invention The invention provides a digital twinning-based intelligent park data processing method, device and system. In a first aspect, the embodiment of the invention provides a digital twinning-based intelligent park data processing method, which comprises the steps of receiving capability self-declaration data and resource source reporting data transmitted by each park twinning node, calling a preset decentralizing trans-park twinning peer-to-peer interconnection architecture and an encryption interconnection gateway, and generating a trans-park twinning index library containing park twinning capability images and real-time resource load information; the method comprises the steps of combining a cross-park twin index library, carrying out pre-adaptation processing on twin states of all park twin nodes, screening out park twin nodes meeting a collaborative optimization target as collaborative participation nodes, issuing encryption initial model content to all collaborative participation nodes, receiving encryption model content fragments fed back by all collaborative participation nodes and generated based on local twin data training, carrying out fusion processing on the encryption model content fragments to obtain a global optimization model, analyzing park twin capacity images and real-time resource load information in the cross-park twin index library, disassembling a collaborative optimization target to obtain a plurality of subtasks, carding execution requirements and dependency relations of all subtasks, dynamically binding all subtasks with twin states of the collaborative participation nodes, distributing subtasks based on binding results and the subtask dependency relations, generating a subtask distribution list and a data interaction specification, calling the global optimization model, the subtask distribution list and the data interaction specification, carrying out symbiotic calibration and evolution on local twin decision logic corresponding to all collaborative participation nodes to drive all collaborative participation nodes to execute corresponding subtasks, obtaining real-time running data and integrated data in a sub task execution process, carrying out integrated data and integrated data optimization effect, carrying out encryption optimization processing effect set after the result is obtained, and driving the bidirectional evolution of the global optimization model and the twin state of each cooperative participation node based on the optimization effect data set, and adjusting a subtask distribution list by combining the real-time resource load change of each cooperative participation node to generate an optimized global optimization model and a task distribution strategy, and synchronizing the optimized global optimization model and the task distribution strategy to a local twin system corresponding to each cooperative participation node. In a second aspect, the embodiment of the invention provides a data processing device, which comprises a data receiving module, a data processing module and a data processing module, wherein the data receiving module is used for receiving capability self-declaration data and resource reporting data transmitted by each park twin node, calling a preset decentralizing trans-park twin peer-to-peer i