Cloud-edge collaborative framework and system for home appliances

Abstract

This application provides a cloud-edge collaborative framework and system for home appliances, relating to the field of smart home technology. The cloud-edge collaborative framework includes: home appliances at the cloud and edge; the edge home appliances are communicatively connected to the cloud; the cloud is used for managing the installation and startup of AI applications on the edge home appliances, as well as managing and scheduling the computing power tasks corresponding to the AI ​​applications. This cloud-edge collaborative framework, by constructing a cloud-edge collaborative framework, manages and schedules applications for home appliances connected to the framework, thereby effectively improving scheduling efficiency and increasing the utilization rate of computing resources.

Description

Technical Field

[0001] This application relates to the field of smart home technology, and in particular to a cloud-edge collaborative framework and system for home appliances. Background Technology

[0002] Distributed computing technology mainly improves computing efficiency and reliability by allocating computing tasks across multiple nodes.

[0003] Existing kitchen appliances are typically equipped with at least one processor to meet the application needs in different scenarios and realize a variety of intelligent functions.

[0004] However, existing kitchen appliances typically use their own processors to perform computing tasks. This computing method results in uneven distribution of computing power and fails to achieve effective utilization of computing resources. Summary of the Invention

[0005] This application provides a cloud-edge collaborative framework and system for home appliances to achieve effective utilization of computing resources.

[0006] In a first aspect, this application provides a cloud-edge collaborative framework for home appliances, the cloud-edge collaborative framework including: home appliances at the cloud and the edge;

[0007] The home appliances at the edge are connected to the cloud for communication.

[0008] The cloud is used to manage the installation and startup of AI applications on the home appliances at the edge, as well as to manage and schedule the computing power tasks corresponding to the AI ​​applications.

[0009] The cloud-edge collaboration framework constructed in this embodiment utilizes the management and scheduling capabilities of the cloud to effectively manage home appliances at the edge, expands the effective utilization of distributed computing technology in AI application computing resources, and realizes remote management of edge AI applications, including various AI application-related tasks such as application installation, application update, application startup, and application shutdown, thereby improving the processing efficiency of AI applications, enhancing user satisfaction, and increasing the market competitiveness of home appliances.

[0010] In some embodiments, the cloud includes an edge cloud and an IoT platform, and the edge cloud and the IoT platform are communicatively connected; the edge cloud is specifically used for:

[0011] The AI ​​application computing power tasks corresponding to the home appliances at the edge are distributed from the IoT platform to the home appliances at the edge.

[0012] The cloud platform constructed in this embodiment can distribute AI application computing tasks, reducing cloud computing power, traffic, and costs. In a home, the network in the kitchen is generally poor, and network transmission is somewhat unstable. Data distribution can effectively avoid network fluctuations affecting the timeliness of computing. Furthermore, all tasks are managed uniformly in the cloud, increasing the possibility of personalized customization in the cloud, thereby achieving customized lightweight computing scheduling.

[0013] In some embodiments, the IoT platform is specifically used for:

[0014] Store tasks and data related to AI applications in home appliances at the edge.

[0015] In this embodiment, AI application-related tasks and data are stored on the IoT platform, which can reduce the storage pressure on edge cloud and edge home appliances, facilitate lightweight device design, and also make it easier to effectively manage and share data, as well as to uniformly update device algorithms and improve the flexibility of data application.

[0016] In some embodiments, the edge cloud is specifically used for:

[0017] Collect AI application launch requests uploaded by the connected edge devices;

[0018] Based on the AI ​​application launch request, the tasks and data related to application launch stored in the IoT platform are scheduled, and the scheduled tasks are forwarded to the home appliances at the edge.

[0019] In this embodiment, the device is located between the IoT platform and the edge home appliances, and mainly handles the collection and distribution of tasks for various AI applications without involving specific computational processing. Therefore, it can free up the computing power of the edge cloud and improve the data processing efficiency of the edge cloud.

[0020] In some embodiments, the edge-end home appliances include edge devices and end devices; the cloud is specifically used for:

[0021] The AI ​​application launch request uploaded by the edge device and the terminal device is obtained, as well as the computing power processing capability corresponding to the edge device and the terminal device. The AI ​​application launch request includes the AI ​​application computing power requirement.

[0022] Based on the computing power requirements of the AI ​​application and the computing power processing capabilities of the edge devices and end devices, the computing power tasks corresponding to the computing power requirements of the AI ​​application are distributed to the edge devices or end devices that meet the computing power processing capabilities; wherein, the computing power processing capability of the edge devices is greater than that of the end devices.

[0023] In this embodiment, the cloud can allocate the most suitable computing resources to different AI applications based on their actual needs, such as computing power, memory and video memory bandwidth, and energy consumption, thereby achieving load balancing in computing power management and scheduling, and improving scheduling efficiency and utilization of computing resources.

[0024] In some embodiments, the edge device includes an AI edge gateway, which is communicatively connected to the edge cloud; the AI ​​edge gateway is specifically used for:

[0025] Responding to computing power tasks issued by the edge cloud;

[0026] The computing power task is executed, and the intermediate processes and results of the execution of the computing power task are returned to the edge cloud.

[0027] In this embodiment, the AI ​​edge gateway is positioned between the edge cloud and the edge terminal. It can manage the edge terminal and forward the intermediate processes and results of computing tasks between the edge cloud and the edge terminal, thereby improving the multi-level management and enhancing the rationality of resource allocation.

[0028] In some embodiments, the edge device includes an AI edge terminal, which is communicatively connected to the edge cloud; the AI ​​edge terminal is specifically used for:

[0029] In response to the application installation and startup tasks and computing power tasks issued by the edge cloud;

[0030] Execute application installation tasks and computing power tasks, and return the intermediate processes and results of executing the installation tasks and computing power tasks to the edge cloud.

[0031] The AI ​​edge terminal provided in this embodiment can execute AI application computing power tasks issued by the system, and does not limit whether the issued tasks are its own tasks, so that computing power tasks can be performed locally and the processing time can be predicted.

[0032] In some embodiments, the cloud also includes a video cloud for collecting image data acquired by the AI ​​edge terminal. The video cloud is communicatively connected to the AI ​​edge terminal, and the edge device is specifically used for:

[0033] In response to the computing power task issued by the edge cloud, the image data stored in the video cloud is obtained;

[0034] Perform computational tasks related to the image data, and return the intermediate processes and results of performing the computational tasks to the edge cloud.

[0035] This embodiment utilizes video cloud to process image data-related tasks with large computational loads separately, thus not affecting the processing progress of other data in the cloud and improving overall data processing efficiency.

[0036] In some embodiments, the AI ​​edge terminal is further specifically used for:

[0037] Preprocess the acquired image data;

[0038] Perform model calculations on the pre-processed data;

[0039] Post-processing is performed on the data from the model calculations.

[0040] This embodiment further specifies that image data-related tasks are also processed at the AI ​​edge terminal, thereby reducing the processing load on the video cloud.

[0041] In some embodiments, the terminal device is communicatively connected to the IoT platform, and the terminal device is specifically used for:

[0042] Responding to the scripting language issued by the IoT platform;

[0043] Run the scripting language and return the results of the scripting language to the IoT platform.

[0044] This embodiment uses a scripting language to execute the script on the terminal device side and return the result to the scheduling side, realizing dynamic computing power scheduling. Furthermore, since the algorithm script can be written and sent to the backend, it also increases the possibility of customized requirements, enabling customized computing programs to be executed on low-resource hardware modules such as terminal devices, solving the problems of long data links and unstable network transmission.

[0045] Secondly, this application provides a cloud-edge collaboration system for home appliances, including home appliances and a cloud-edge collaboration framework for home appliances.

[0046] This application provides a cloud-edge collaborative framework for home appliances, involving a cloud, an edge device, and edge devices. The edge device is communicatively connected to both the cloud and the edge device. The cloud is used for remote management of edge AI applications, as well as computing power management and scheduling. Compared to the shortcomings of existing technologies, this application utilizes the cloud-edge collaborative framework to achieve remote management of edge AI applications, including application installation, updates, startup, and shutdown. It manages computing power and schedules computing power by managing all edge devices and end devices, scheduling multiple heterogeneous processors on kitchen appliances to implement various algorithms, improving scheduling efficiency and increasing the utilization of computing resources. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 A schematic diagram illustrating the application scenario of the cloud-edge collaborative framework for home appliances provided in this application embodiment;

[0049] Figure 2 This is a schematic diagram of the cloud-edge collaboration framework provided in the embodiments of this application;

[0050] Figure 3 A flowchart illustrating an application scheduling preparation method for a cloud-edge collaborative framework provided in this application embodiment;

[0051] Figure 4 A flowchart of an intelligent module application scheduling method for a cloud-edge collaborative framework provided in this application embodiment;

[0052] Figure 5 A flowchart of a cloud-edge collaborative algorithm application scheduling method for a cloud-edge collaborative framework provided in this application embodiment;

[0053] Figure 6 This is a schematic diagram of the cloud-edge collaboration framework provided in the embodiments of this application;

[0054] Figure 7 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application.

[0055] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0056] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0057] Existing kitchen appliances are typically equipped with at least one processor to activate the appliance's applications, process its functions, meet user needs in different scenarios, and enable various intelligent functions.

[0058] Based on the above-mentioned technical problems, the inventive concept of this application is to construct a cloud-edge collaborative framework to effectively schedule and manage computing resources at the edge to meet the computing needs of different applications. When there are many computing tasks, the scheduling system dynamically allocates tasks according to the overall load situation, realizing distributed and effective scheduling of edge devices and edge gateways, thereby improving scheduling efficiency and utilization of computing resources, aiming to solve the above-mentioned technical problems of the prior art.

[0059] The specific application scenarios for this application are as follows:

[0060] Figure 1 This is a schematic diagram illustrating an application scenario of the cloud-edge collaborative framework for home appliances provided in this application embodiment. For example... Figure 1 As shown, various home appliances 101 installed in the kitchen serve as edge devices, which are connected to the same cloud 102. The cloud includes an edge cloud 102 and an IoT platform 103. The edge devices are connected to the edge cloud 102, and then connected to the IoT platform 103 through the edge cloud 102, thereby achieving a reasonable allocation of computing resources for different home appliances 101 when the application starts.

[0061] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0062] Figure 2 This is a schematic diagram of the cloud-edge collaboration framework provided in an embodiment of this application. Figure 2 As shown, the cloud-edge collaboration framework includes:

[0063] IoT Platform 201, Edge Cloud 202, and Edge Terminal 203.

[0064] Among them, the IoT platform 201 includes programs for remote management of edge AI applications, such as product management, device management, storage services, rule services, and script services.

[0065] Edge Cloud 202 includes edge framework and scheduling services.

[0066] Optionally, the edge framework includes connectivity services, application management services, and resource management services.

[0067] The scheduling service includes task scheduling, data forwarding, and message bus.

[0068] For example, edge device 203 includes an AI edge gateway, an AI module terminal, and an AI edge terminal.

[0069] Optionally, the AI ​​edge gateway includes connectivity services, edge terminal management, gateway services, algorithm models, CPU, NPU, etc.

[0070] AI edge terminals include connectivity services, image capture, video streaming, algorithm models, CPU, NPU, etc.

[0071] For example, an AI edge terminal or AI edge gateway includes an edge all-in-one machine, whose hardware processor includes at least an NPU, a GPU, etc.

[0072] The AI ​​module terminal includes connectivity services and a lightweight application engine.

[0073] Alternatively, AI module terminals are typically terminal or module devices whose hardware processors only include simple end devices such as CPUs.

[0074] Figure 3 A flowchart illustrating an application scheduling preparation method for a cloud-edge collaborative framework provided in this application embodiment. (Reference) Figure 3 As shown, the method includes:

[0075] Users can manage scripts and applications through the IoT platform.

[0076] The AI ​​edge gateway, AI edge terminal, and AI module terminal are connected to the edge cloud via MQTT to transmit data from the AI ​​edge gateway, AI edge terminal, and AI module terminal to the edge cloud.

[0077] Figure 4 A flowchart illustrating an intelligent module application scheduling method within a cloud-edge collaborative framework, as provided in this application embodiment. (Reference) Figure 4 As shown, the method includes:

[0078] Users can edit and debug scripts through the IoT platform to adapt them to different models or functions of kitchen appliances, thereby improving the adaptability of control methods.

[0079] When the AI ​​module terminal has an application startup requirement, such as a formula application, the formula corresponds to a script stored on the IoT platform. Therefore, the IoT platform can send the edited or adjusted script to the AI ​​module terminal. After the AI ​​module terminal runs and processes the script, it sends the running feedback results back to the IoT platform.

[0080] For example, when a gas stove needs to activate an application that dynamically adjusts the stove's shut-off time based on humidity, the latest formula in the script can be sent to the gas stove so that it can dynamically adjust according to the latest formula to better meet the user's actual usage needs.

[0081] Since the algorithm's computation and protocol escaping are primarily performed in the cloud, or some simple computations or preprocessing are done on the MCU side, this raises the following issues:

[0082] The kitchen in a home typically has poor network connectivity, leading to some instability in network transmission.

[0083] The data link is long, and network fluctuations affect the timeliness of computation;

[0084] A large amount of device data is processed in the cloud, which increases the computing power, traffic, and cost of servers;

[0085] Since MCUs cannot be remotely pushed for upgrades, users of already sold products cannot easily enjoy the benefits of algorithm upgrades.

[0086] MCU resources are limited. As the electronic control unit, it should mainly handle I / O intensive tasks. Its computing power is limited and affects the timeliness of I / O processing, which may affect the user experience.

[0087] The development cycle is long, the flexibility is poor, and the algorithm is already determined after the firmware is burned.

[0088] The main reason for the above problem is that control commands cannot be transmitted in a timely manner, and this problem can be solved by using script functions.

[0089] For example, script functions may include customized computation scripts, such as one-dimensional function computations, unit conversions, and data conversions between different protocols.

[0090] Optionally, a resident executor runs on the device-side computing module and is called when the application starts. By dynamically distributing customized script functions with computing tasks on the IoT platform, the executor (script engine) completes the execution of the script on the main control computing power and returns the results to the scheduler, thereby realizing dynamic computing power scheduling.

[0091] In summary, by using scripting languages ​​to implement customized computing programs for small intelligent applications that would otherwise be implemented on the platform or MCU side, lightweight computing and protocol escaping can be executed on low-resource hardware modules. This solves the problems of long data links, unstable network transmission, limited MCU resources on the control side, complex OTA, and poor flexibility.

[0092] This embodiment implements the computing program using a scripting language (such as Lua, Python, etc.). By leveraging the fact that scripting languages ​​can run without compilation, customizable computing is offloaded to low-resource hardware modules or the system, thereby freeing up cloud computing power and reducing data traffic transmission, thus achieving customized lightweight computing scheduling.

[0093] For example, a low-resource hardware module with a main frequency of 100MHz, 1M storage, and 256KB memory.

[0094] Therefore, by using script functions, business processes can be performed locally, resulting in faster speeds and predictable processing times. Algorithm updates are simpler and more flexible, requiring only script distribution from the cloud, eliminating the need for module OTA updates. Since algorithm scripts can be written and distributed to the backend, the possibility of customizing personalized requirements is also increased. The MCU only needs to focus on device control and data acquisition, allowing for business focus, reduced development difficulty, shorter development cycles, and improved stability.

[0095] Figure 5 This document provides a flowchart of a cloud-edge collaborative algorithm application scheduling method for a cloud-edge collaborative framework, as illustrated in an embodiment of this application. (Reference) Figure 5 As shown, the method includes:

[0096] Users can package, upload, and deploy applications through the IoT platform.

[0097] The IoT platform installs and distributes algorithm applications to the edge cloud. The edge cloud then uses cloud-edge collaboration to distribute installation tasks to the AI ​​edge gateway, which in turn installs the applications based on these tasks.

[0098] After the application is installed, the IoT platform sends an application launch request to the edge cloud. Based on the application launch request, the edge cloud sends a computing power task distribution and launch request to the AI ​​edge gateway. The AI ​​edge gateway launches the application and returns the intermediate process and results of the task to the edge cloud.

[0099] Edge cloud sends installation and computing power task start requests to AI edge terminals through cloud-edge collaboration. The AI ​​edge terminals execute application installation and start, and return the intermediate process and results of the task to edge cloud.

[0100] Among them, the distributed heterogeneous computing power scheduling method based on cloud computing constructs a cloud-edge collaborative framework composed of multiple computing nodes, each with different computing capabilities, such as CPU, GPU, NPU, RISC-V, ARM, etc.

[0101] The cloud-edge collaboration framework monitors the load of each node and dynamically allocates computing tasks to achieve load balancing and improve scheduling efficiency.

[0102] The cloud-edge collaboration framework allocates the most suitable computing resources to different applications based on their needs, such as computing power, memory and GPU bandwidth, and energy consumption.

[0103] For example, for image recognition and generative applications that require a lot of computation, the cloud-edge collaborative framework will prioritize allocating high-performance computing resources such as GPUs and NPUs on the edge device; while for applications with smaller computational loads such as formula and function calculations, the cloud-edge collaborative framework will prioritize allocating CPU computing resources on the terminal or module device.

[0104] By building a cloud-edge collaboration framework, effective management of edge devices and edge gateways can be achieved.

[0105] For example, the management of edge devices and edge gateways includes operations such as application installation, application update, application startup, and application shutdown.

[0106] Meanwhile, the cloud-edge collaboration framework can also manage edge devices across the entire domain and the edge of local area networks, enabling computing power management and scheduling.

[0107] The cloud-edge collaboration framework employs an intelligent scheduling algorithm that automatically selects the optimal scheduling strategy based on the characteristics of the computing task and the available computing resources.

[0108] For example, for tasks with high real-time requirements, the cloud-edge collaborative framework will prioritize allocating resources with strong computing power; while for tasks with large computing loads, the cloud-edge collaborative framework will prioritize allocating resources with large storage space.

[0109] In summary, the cloud-edge collaboration framework utilizes a distributed heterogeneous computing power scheduling method to schedule multiple heterogeneous processors on kitchen appliances to achieve various algorithm applications and remote management of edge AI applications, including application installation, application updates, application startup, and application shutdown. It achieves computing power management and scheduling by managing edge devices and end devices across the entire domain.

[0110] In some embodiments, continue to refer to Figure 2 As shown, edge clouds can also include video clouds.

[0111] Figure 6 This is a schematic diagram of the cloud-edge collaboration framework provided in an embodiment of this application. (Reference) Figure 6 As shown, the edge video computing appliance includes an edge framework and a scheduling service.

[0112] The edge framework includes connectivity services, application management services, and resource management services. The scheduling service includes task scheduling and data forwarding. Both the edge framework and the scheduling service are connected to the message bus, which is also connected to an image processing terminal. The image processing terminal is equipped with a camera, and the camera realizes device discovery, image capture, and video stream storage services based on the launch of the acquisition application.

[0113] After the image is captured, it is first stored in the image cache. Then, it undergoes preprocessing, model calculation and postprocessing in sequence before being sent to the message bus. The message bus then feeds back to the edge cloud, and finally connects to the IoT platform through cloud-edge collaboration and data transmission.

[0114] In this embodiment, the edge framework mainly implements application management and provides basic component services, the scheduling application mainly implements algorithm task scheduling, the image acquisition application mainly completes image acquisition and image writing to disk, and the AI ​​algorithm mainly performs AI calculations on the images and classifies the images.

[0115] For example, it takes about 200ms to capture and write a 1920*1080 image, and it is estimated that it will only take 500ms to collect 30 channels of image data from 30 cameras at the same time.

[0116] Specifically, the aforementioned cloud-edge collaborative framework employs a distributed heterogeneous computing power scheduling method, including:

[0117] Obtain the computing power requirements of applications collected by the edge cloud;

[0118] The edge cloud is used to collect the computing power requirements of different applications after responding to the application launch request at the edge; the application launch request includes the application's computing power requirements and the edge identifier that sent the request.

[0119] Based on the application computing power requirements collected by the edge cloud, the computing power task corresponding to the application startup request is sent to the edge.

[0120] The edge device is equipped with both edge and end devices to handle computing tasks. The edge device has a greater computing task handling capacity than the end device. The edge device that initiates the application startup request may not be the same as the edge device that was issued the request.

[0121] For example, the method includes:

[0122] If the computing power requirements of the applications collected by the edge cloud meet the preset conditions, the computing power tasks corresponding to the computing power requirements will be sent to the edge devices for processing.

[0123] If the computing power requirements of the applications collected by the edge cloud do not meet the preset conditions, the computing power tasks corresponding to the computing power requirements will be sent to the end devices for processing.

[0124] Optionally, the preset condition includes a real-time requirement that is less than or equal to a preset time limit;

[0125] The method includes:

[0126] If the computing power requirement of the edge cloud is less than or equal to the preset time limit, the computing power task corresponding to the computing power requirement will be sent to the first edge device for processing; the computing power of the first edge device is greater than or equal to the preset computing power threshold.

[0127] Among them, the preset condition includes that the amount of computation is greater than or equal to the preset amount of computation;

[0128] The method includes:

[0129] If the computing power requirement of the edge cloud application is greater than or equal to the preset computing power, the computing power task corresponding to the computing power requirement will be sent to the second-side device for processing; the remaining storage space of the second-side device is greater than or equal to the preset space threshold.

[0130] In some embodiments, the method further includes:

[0131] Obtain the scripting language of the IoT platform; this IoT platform has a scripting service deployed to receive scripts edited by the user.

[0132] The scripting language of the IoT platform is distributed to the end devices to execute the scripting language processing tasks.

[0133] In other embodiments, the edge cloud also deploys a video cloud for storing image data; the method further includes:

[0134] Acquire the captured images and cache them.

[0135] The acquired images are forwarded to the side devices to perform image processing tasks.

[0136] This application also provides a cloud-edge collaboration system for home appliances, including home appliances and a cloud-edge collaboration framework for home appliances.

[0137] The cloud-edge collaboration system for home appliances provided in this embodiment can execute the cloud-edge collaboration framework of the above embodiment. Its implementation principle and technical effect are similar, and will not be described again in this embodiment.

[0138] In this embodiment of the invention, electronic devices or main control devices can be divided into functional modules according to the above method examples. For example, each function can be divided into its own functional modules, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional module. It should be noted that the module division in this embodiment of the invention is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0139] In the specific implementation of the aforementioned distributed heterogeneous computing power scheduling device, each module can be implemented as a processor. The processor can execute computer execution instructions stored in the memory, so that the processor executes the aforementioned distributed heterogeneous computing power scheduling method.

[0140] Figure 7 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of this application. Figure 7 As shown, the electronic device includes:

[0141] At least one processor 701 and memory 702.

[0142] The electronic device also includes a communication component 703.

[0143] The processor 701, memory 702, and communication component 703 are connected via a bus.

[0144] In the specific implementation process, at least one processor 701 executes the computer execution instructions stored in the memory 702, causing at least one processor 701 to execute the distributed heterogeneous computing power scheduling method as described above.

[0145] The specific implementation process of processor 701 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0146] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0147] The memory may include high-speed RAM, and may also include non-volatile storage (NVM), such as at least one disk storage.

[0148] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0149] The above describes the solutions provided by the embodiments of the present invention for the functions implemented by the electronic device and the main control device.

[0150] It is understandable that electronic devices or main control devices include hardware structures and / or software modules that perform the above functions in order to achieve the above functions.

[0151] By combining the units and algorithm steps of the various examples described in the embodiments of this invention, the embodiments of this invention can be implemented in hardware or a combination of hardware and computer software. Whether a certain function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the technical solutions of the embodiments of this invention.

[0152] This application also provides a computer program product, including a computer program that, when executed by a processor, implements a distributed heterogeneous computing power scheduling method.

[0153] The computer program product provided in this embodiment can execute the distributed heterogeneous computing power scheduling method of the above embodiment. Its implementation principle and technical effect are similar, and will not be described again in this embodiment.

[0154] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described distributed heterogeneous computing power scheduling method.

[0155] The computer-readable storage medium provided in this embodiment can execute the distributed heterogeneous computing power scheduling method of the above embodiment. Its implementation principle and technical effect are similar, and will not be described again in this embodiment.

[0156] The aforementioned computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0157] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in an electronic device or a host device.

[0158] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0159] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with relevant laws, regulations and standards, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0160] 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 foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A cloud-edge collaborative framework for home appliances, characterized in that, The cloud-edge collaboration framework includes: home appliances at the cloud and edge; The home appliances at the edge are connected to the cloud for communication. The cloud is used to manage the installation and startup of AI applications on the home appliances at the edge, as well as to manage and schedule the computing power tasks corresponding to the AI ​​applications.

2. The cloud-edge collaboration framework according to claim 1, characterized in that, The cloud includes an edge cloud and an IoT platform, and the edge cloud and the IoT platform are communicatively connected. The edge cloud is used to distribute AI application computing power tasks corresponding to the home appliances at the edge from the IoT platform to the home appliances at the edge. The IoT platform is used to store tasks and data related to AI applications in the home appliances at the edge.

3. The cloud-edge collaboration framework according to claim 2, characterized in that, The edge cloud is specifically used for: Collect AI application launch requests uploaded by the connected edge devices; Based on the AI ​​application launch request, the tasks and data related to application launch stored in the IoT platform are scheduled, and the scheduled tasks are forwarded to the edge home appliances so that the edge home appliances can execute them.

4. The cloud-edge collaboration framework according to claim 2, characterized in that, The edge-end home appliances include edge devices and end devices; the cloud is specifically used for: The AI ​​application launch request uploaded by the edge device and the terminal device is obtained, as well as the computing power processing capability corresponding to the edge device and the terminal device. The AI ​​application launch request includes the AI ​​application computing power requirement. Based on the computing power requirements of the AI ​​application and the computing power processing capabilities of the edge devices and end devices, the computing power tasks corresponding to the computing power requirements of the AI ​​application are distributed to the edge devices or end devices that meet the computing power processing capabilities; wherein, the computing power processing capability of the edge devices is greater than that of the end devices.

5. The cloud-edge collaboration framework according to claim 4, characterized in that, The edge device includes an AI edge gateway, which is communicatively connected to the edge cloud; the AI ​​edge gateway is specifically used for: Responding to computing power tasks issued by the edge cloud; The computing power task is executed, and the intermediate processes and results of the execution of the computing power task are returned to the edge cloud.

6. The cloud-edge collaboration framework according to claim 4, characterized in that, The edge device includes an AI edge terminal, which is communicatively connected to the edge cloud; the AI ​​edge terminal is specifically used for: In response to the application installation and startup tasks and computing power tasks issued by the edge cloud; Execute application installation tasks and computing power tasks, and return the intermediate processes and results of executing the installation tasks and computing power tasks to the edge cloud.

7. The cloud-edge collaboration framework according to claim 6, characterized in that, The cloud also includes a video cloud for collecting image data acquired by the AI ​​edge terminal. The video cloud is communicatively connected to the AI ​​edge terminal. The edge device is specifically used for: In response to the computing power task issued by the edge cloud, the image data stored in the video cloud is obtained; Perform computational tasks related to the image data, and return the intermediate processes and results of performing the computational tasks to the edge cloud.

8. The cloud-edge collaboration framework according to claim 7, characterized in that, The AI ​​edge terminal is also specifically used for: Preprocess the acquired image data; Perform model calculations on the pre-processed data; Post-processing is performed on the data from the model calculations.

9. The cloud-edge collaboration framework according to claim 4, characterized in that, The terminal device is communicatively connected to the IoT platform, and the terminal device is specifically used for: Responding to the scripting language issued by the IoT platform; Run the scripting language and return the results of the scripting language to the IoT platform.

10. A cloud-edge collaborative system for home appliances, characterized in that, This includes home appliances, and the cloud-edge collaboration framework for home appliances as described in any one of claims 1-9.

Images