A method and system for describing and scheduling heterogeneous resource pooling on the edge for embodied tasks.
By pooling and standardizing the description of heterogeneous hardware resources of embodied intelligent devices, the problem of inconsistent resource descriptions is solved, enabling precise matching and efficient scheduling of resource capabilities, and improving task execution efficiency and resource utilization.
Patent Information
- Application Number
- CN202610244143.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-02
- Publication Date
- 2026-06-02
AI Technical Summary
In existing technologies, the heterogeneous hardware resources of embodied intelligent devices lack a unified capability description method, making it difficult for the cloud to accurately know the capabilities and status of the resources on the edge, resulting in low task scheduling and matching efficiency and low resource utilization.
By identifying heterogeneous hardware resources on the edge, pooling modeling is performed to generate standardized resource description files, which are then uploaded to the cloud to achieve precise matching between resource capabilities and task requirements.
It achieves a unified abstraction and standardized expression of resource capabilities, improves task execution efficiency and resource utilization, and enhances system stability and scalability.
Smart Images

Figure CN122132177A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of embodied intelligent system architecture technology, and in particular to a method and system for describing and scheduling heterogeneous resource pooling on the edge for embodied tasks. Background Technology
[0002] Embodied intelligent devices typically include a variety of heterogeneous hardware resources, such as sensors, motor actuators, graphics processing units, or neural network processing units. In the current technology, there is a lack of a unified way to describe the capabilities of different hardware resources, making it difficult for the cloud to accurately know the capabilities and status of the resources on the edge, resulting in low efficiency in task scheduling and matching and low resource utilization.
[0003] Therefore, a unified method for modeling and describing resource pooling is needed. Summary of the Invention
[0004] Purpose of the invention The purpose of this invention is to provide a method and system for describing and scheduling heterogeneous resources on the edge for specific tasks, in order to solve the problems of low task execution efficiency and low resource utilization caused by the lack of a unified capability description method for various heterogeneous hardware resources on the edge and the inability of cloud scheduling to accurately match resource capabilities.
[0005] Technical solution To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a method for describing and scheduling heterogeneous resource pooling on the edge for embodied tasks, including: The edge device identifies its own heterogeneous hardware resources, which include sensor modules, actuator modules, and computing processing modules. Pooling modeling is performed on the heterogeneous hardware resources, and the capability parameters, real-time working status parameters and interface protocol type of each resource are collected; A standardized resource description file is generated based on the aforementioned capability parameters, real-time working status parameters, and interface protocol types. Upload the standardized resource description file to the cloud server; Based on the resource requirements of the specific task, the cloud server matches the heterogeneous resources on the edge side that meet the requirements from the resource description file, and sends scheduling instructions to the edge device.
[0006] Preferably, the capability parameters include computational capability parameters, accuracy parameters, response time parameters, or driving capability parameters.
[0007] Preferably, the real-time operating status parameters include resource utilization rate, operating status identifier, and fault status identifier.
[0008] Preferably, the standardized resource description file is stored in a structured data format and includes a resource identifier field, a capability field, a status field, and a protocol field.
[0009] This invention also provides a task-oriented edge-side heterogeneous resource pooling description and scheduling adaptation system, comprising: The resource identification module is used to identify heterogeneous hardware resources on the terminal side; The resource pooling modeling module is used to abstract and encapsulate resources and build a resource pool structure; The resource description generation module is used to generate standardized resource description files; The communication module is used to upload resource description files to the cloud; A cloud-based matching and scheduling module is used to match suitable resources and issue scheduling instructions based on specific task requirements. This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method steps.
[0010] Beneficial effects Compared with the prior art, the present invention has the following beneficial effects: By pooling and modeling heterogeneous resources on the edge side, a unified abstraction and standardized expression of resource capabilities can be achieved. By generating standardized resource description files, the cloud can accurately obtain the resource capabilities and real-time status information of the edge side; Achieve precise matching between resource capabilities and specific task requirements to improve task execution efficiency; Improve the utilization rate of edge hardware resources and reduce resource waste; Improve the stability and scalability of the overall scheduling of embodied intelligence systems. Attached Figure Description
[0011] Figure 1 This is a schematic diagram of the end-side heterogeneous resource pooling system structure; Figure 2 Flowchart of the method for describing and adapting resource pooling on the edge side; Figure 3 This is a schematic diagram of the structure of a standardized resource description file.
[0012] Figure 1 Structural diagram include: End-side equipment 100 Resource identification module 110 Resource pooling module 120 Description file generation module 130 Communication Module 140 Cloud server 200 Cloud scheduling module 210 Data flow: Resource identification → Pooling → Description generation → Upload → Cloud matching → Command issuance.
[0013] Figure 2 Method Flowchart S201 Identify heterogeneous resources; S202 Establish a resource pool model; S203 Generate standardized resource description files; S204 Upload to the cloud; S205 Match resources in the cloud; S206 Issue scheduling instructions.
[0014] Figure 3 Description file structure diagram include: Resource Identification Area 301 Capability Parameters Section 302 Real-time status area 303 Protocol Fields 304 Resources are stored in a structured entry format.
Claims
1. A method for describing and scheduling heterogeneous resource pooling on the edge for embodied tasks, characterized in that, include: S1. The end-side device identifies its own heterogeneous hardware resources, which include a sensor module, an actuator module, and a computing processing module. S2. Perform pooling modeling on the heterogeneous hardware resources, and collect the capability parameters, real-time working status parameters and interface protocol type of each resource. S3. Generate a standardized resource description file based on the aforementioned capability parameters, real-time working status parameters, and interface protocol type; S4. Upload the standardized resource description file to the cloud server; S5. The cloud server matches the heterogeneous resources on the end side that meet the requirements from the resource description file based on the resource requirements of the specific task, and sends scheduling instructions to the end-side device.
2. The method according to claim 1, characterized in that: The capability parameters include computational capability parameters, accuracy parameters, or driving capability parameters.
3. The method according to claim 1, characterized in that: The real-time operating status parameters include resource utilization rate, operating status indicator, or fault status indicator.
4. The method according to claim 1, characterized in that: The standardized resource description file is stored in a structured data format.
5. A task-oriented edge-side heterogeneous resource pooling description and scheduling adaptation system, characterized in that, include: Resource identification module; Resource pooling modeling module; Resource description generation module; Communication upload module; Cloud-based matching and scheduling module.
6. A computer-readable storage medium having a program stored thereon that, when executed by a processor, implements the method of any one of claims 1-4.
