CN-121982735-A - Method, device, equipment and medium for acquiring scheduling scheme based on digital asset

CN121982735ACN 121982735 ACN121982735 ACN 121982735ACN-121982735-A

Abstract

The application relates to the technical field of digital assets and the technical field of data validation, and discloses a scheduling scheme acquisition method, a device, equipment and a medium based on the digital assets, wherein the method comprises the steps of acquiring text information, image information and structured field information of the digital assets; the method comprises the steps of carrying out fusion operation on semantic vectors of text information, semantic vectors of image information and semantic vectors of structured field information to generate multi-mode fusion vectors of digital assets, integrating the multi-mode fusion vectors of the digital assets, metadata sets of the digital assets, fingerprint sets of the digital assets, dependence graphs of the digital assets and strategy parameters of the digital assets to generate uplink demands of the digital assets, processing the uplink demands of the digital assets through a dynamic scheduling strategy to obtain a plurality of candidate scheduling schemes, and determining an optimal scheduling scheme of the digital assets based on a fitness function and the plurality of candidate scheduling schemes. The application can improve the acquisition efficiency of the optimal scheduling scheme of the digital asset.

Inventors

ZHANG XINYU
CHEN HANBING
CAO WENZHI
YI GUODONG
ZHANG WEIWEI

Assignees

湖南红普创新科技发展有限公司

Dates

Publication Date: 20260505
Application Date: 20251230

Claims (10)

1. A scheduling scheme acquisition method based on digital assets, which is applied to electronic equipment, the scheduling scheme acquisition method comprising: acquiring text information, image information and structured field information of a digital asset; Performing fusion operation on the semantic vector of the text information, the semantic vector of the image information and the semantic vector of the structured field information to generate a multi-mode fusion vector of the digital asset; Integrating the multimodal fusion vector of the digital asset, the metadata set of the digital asset, the fingerprint set of the digital asset, the dependency graph of the digital asset and the policy parameters of the digital asset to generate the uplink demand of the digital asset; Processing the uplink demand of the digital asset through a dynamic scheduling strategy to obtain a plurality of candidate scheduling schemes; an optimal scheduling scheme for the digital asset is determined based on the predefined fitness function and the plurality of candidate scheduling schemes.
2. The scheduling method of claim 1, wherein performing weighted aggregation on the semantic vector of the text information, the semantic vector of the image information, and the semantic vector of the structured field information to generate the multimodal fusion vector of the digital asset comprises: generating semantic vectors of text information through a pre-training language model, generating semantic vectors of image information through a visual encoder, and generating semantic vectors of structured field information through a semantic model; And performing fusion operation on the semantic vector of the text information, the semantic vector of the image information and the semantic vector of the structured field information by adopting a weighted fusion mode to generate a multi-mode fusion vector of the digital asset.
3. The scheduling method according to claim 1, wherein integrating the multimodal fusion vector of digital assets, the metadata set of digital assets, the fingerprint set of digital assets, the dependency graph of digital assets, the policy parameters of digital assets, and generating the uplink requirements of digital assets comprises: forming metadata of text information, metadata of image information and metadata of structured field information into metadata sets of digital assets; Forming a fingerprint set of the digital asset from the data fingerprint of the text information, the data fingerprint of the image information and the data fingerprint of the structured field information; Integrating the multimodal fusion vector of the digital asset, the metadata set of the digital asset, the fingerprint set of the digital asset, the dependency graph of the digital asset and the policy parameters of the digital asset to generate the uplink demand of the digital asset.
4. A scheduling method according to claim 3, wherein the composing the metadata of the text information, the metadata of the image information, the metadata of the structured field information into the metadata set of the digital asset comprises: the method comprises the steps of forming metadata of text information by source identification of the text information, acquisition time of the text information and acquisition environment information of the text information; the source identification of the image information, the acquisition time of the image information and the acquisition environment information of the image information are combined into metadata of the image information; The source identification of the structured field information, the acquisition time of the structured field information and the acquisition environment information of the structured field information are combined into metadata of the structured field information; metadata of the text information, metadata of the image information, and metadata of the structured field information are organized into a metadata collection of the digital asset.
5. A scheduling method according to claim 3, wherein the composing the data fingerprint of the text information, the data fingerprint of the image information, the data fingerprint of the structured field information into a fingerprint set of the digital asset comprises: Performing a stitching operation on the text information and the metadata of the text information to generate first stitching data, performing a stitching operation on the image information and the metadata of the image information to generate second stitching data, and performing a stitching operation on the structured field information and the metadata of the structured field information to generate third stitching data; Processing the first spliced data through a hash function to obtain a first hash value, selecting the first hash value as a data fingerprint of text information, processing the second spliced data through the hash function to obtain a second hash value, selecting the second hash value as a data fingerprint of image information, processing the third spliced data through the hash function to obtain a third hash value, selecting the third hash value as a data fingerprint of structured field information, and forming a fingerprint set of the digital asset from the data fingerprint of the text information, the data fingerprint of the image information and the data fingerprint of the structured field information.
6. A scheduling method according to claim 3, wherein integrating the multimodal fusion vector of digital assets, the metadata set of digital assets, the fingerprint set of digital assets, the dependency graph of digital assets, and the policy parameters of digital assets, generating the uplink requirements of digital assets comprises: acquiring a priority index of the digital asset, a privacy class of the digital asset, a cost range and a maximum time delay of the digital asset; the method comprises the steps of forming a policy parameter of a digital asset by a priority index of the digital asset, a privacy class, a cost range and a maximum time delay of the digital asset, integrating a multi-modal fusion vector of the digital asset, a metadata set of the digital asset, a fingerprint set of the digital asset, a dependency graph of the digital asset and the policy parameter of the digital asset, and generating a uplink demand of the digital asset.
7. The scheduling method of claim 1, wherein determining an optimal scheduling scheme for the digital asset based on the predefined fitness function and the plurality of candidate scheduling schemes comprises: Generating the adaptability of each candidate scheduling scheme through a predefined adaptability function, and selecting the candidate scheduling scheme with the minimum adaptability as the optimal scheduling scheme of the digital asset; the fitness function is defined as follows: ; represent the first The adaptability of the candidate scheduling schemes, the first The greater the fitness of each candidate scheduling scheme, the description of the first The weaker the overall performance of each candidate scheduling scheme in three aspects of total completion time, execution cost and task failure rate is The smaller the fitness of each candidate scheduling scheme, the description of the first The weaker the overall performance of the candidate scheduling schemes in three aspects of total completion time, execution cost and task failure rate; represent the first Total completion time of the candidate scheduling schemes; represent the first The execution cost of the candidate scheduling schemes; represent the first Task failure rates of the candidate scheduling schemes; a first weight coefficient is represented and a second weight coefficient is represented, A second weight coefficient is represented and is used to represent, Representing a third weight coefficient.
8. A digital asset based scheduling scheme acquisition apparatus, for use with an electronic device, comprising: The acquisition module is used for acquiring text information, image information and structured field information of the digital asset; The fusion module is used for executing fusion operation on the semantic vector of the text information, the semantic vector of the image information and the semantic vector of the structured field information to generate a multi-mode fusion vector of the digital asset; The generation module is used for integrating the multimodal fusion vector of the digital asset, the metadata set of the digital asset, the fingerprint set of the digital asset, the dependency graph of the digital asset and the strategy parameters of the digital asset to generate the uplink demand of the digital asset; The processing module is used for processing the uplink demands of the digital assets through a dynamic scheduling strategy to obtain a plurality of candidate scheduling schemes; a scheduling module for determining an optimal scheduling scheme for the digital asset based on the predefined fitness function and the plurality of candidate scheduling schemes.
9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the scheduling scheme acquisition method of any one of claims 1 to 7 when executing the computer program.
10. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the scheduling scheme acquisition method of any one of claims 1 to 7.

Description

Method, device, equipment and medium for acquiring scheduling scheme based on digital asset Technical Field The present application relates to the field of digital asset technologies and the field of data right verification technologies, and in particular, to a method, an apparatus, a device, and a medium for acquiring a scheduling scheme based on a digital asset. Background Digital assets are rights and interests sets which exist in digital form, have quantifiable value and are protected by law, and the core logic of the digital assets is to realize the rights, circulation and value change of the assets by means of digital technology. Digital asset chaining, also known as asset digitizing, is the recording of physical or digital asset information on a chain by blockchain technology to form a tamper-proof digital voucher. Digital asset uplink is not a simple data storage operation and therefore requires advanced scheduling schemes for digital assets. However, the acquisition process of the optimal scheduling scheme of the existing digital asset is complicated, which is not beneficial to improving the acquisition efficiency of the optimal scheduling scheme. The method is characterized in that the prior art mainly adopts a manual acquisition mode to acquire the optimal scheduling scheme of the digital asset, and the manual acquisition mode consumes a large amount of human resources and time resources, so that the acquisition time of the optimal scheduling scheme of the digital asset is increased, and the optimal scheduling scheme is easily influenced by manual intervention, thus being unfavorable for improving the acquisition efficiency of the optimal scheduling scheme. Disclosure of Invention The embodiment of the application provides a scheduling scheme acquisition method, device, equipment and medium based on digital assets, which are used for solving the technical problems that the acquisition process of the optimal scheduling scheme of the existing digital assets is complicated and the acquisition efficiency of the optimal scheduling scheme is not improved. In a first aspect, an embodiment of the present application provides a method for acquiring a scheduling scheme based on a digital asset, which is applied to an electronic device, where the method for acquiring the scheduling scheme includes: acquiring text information, image information and structured field information of a digital asset; Performing fusion operation on the semantic vector of the text information, the semantic vector of the image information and the semantic vector of the structured field information to generate a multi-mode fusion vector of the digital asset; Integrating the multimodal fusion vector of the digital asset, the metadata set of the digital asset, the fingerprint set of the digital asset, the dependency graph of the digital asset and the policy parameters of the digital asset to generate the uplink demand of the digital asset; Processing the uplink demand of the digital asset through a dynamic scheduling strategy to obtain a plurality of candidate scheduling schemes; an optimal scheduling scheme for the digital asset is determined based on the predefined fitness function and the plurality of candidate scheduling schemes. In a possible implementation manner of the first aspect, the performing weighted aggregation on the semantic vector of the text information, the semantic vector of the image information, and the semantic vector of the structured field information to generate a multimodal fusion vector of the digital asset includes: generating semantic vectors of text information through a pre-training language model, generating semantic vectors of image information through a visual encoder, and generating semantic vectors of structured field information through a semantic model; And performing fusion operation on the semantic vector of the text information, the semantic vector of the image information and the semantic vector of the structured field information by adopting a weighted fusion mode to generate a multi-mode fusion vector of the digital asset. In a possible implementation manner of the first aspect, the integrating the multimodal fusion vector of the digital asset, the metadata set of the digital asset, the fingerprint set of the digital asset, the dependency graph of the digital asset, and the policy parameter of the digital asset, to generate the uplink requirement of the digital asset includes: forming metadata of text information, metadata of image information and metadata of structured field information into metadata sets of digital assets; Forming a fingerprint set of the digital asset from the data fingerprint of the text information, the data fingerprint of the image information and the data fingerprint of the structured field information; Integrating the multimodal fusion vector of the digital asset, the metadata set of the digital asset, the fingerprint set of the digital asset, the dependency graph o