CN113691579A - Robot AI service method and system based on cloud edge - Google Patents
Robot AI service method and system based on cloud edge Download PDFInfo
- Publication number
- CN113691579A CN113691579A CN202110736921.4A CN202110736921A CN113691579A CN 113691579 A CN113691579 A CN 113691579A CN 202110736921 A CN202110736921 A CN 202110736921A CN 113691579 A CN113691579 A CN 113691579A
- Authority
- CN
- China
- Prior art keywords
- model
- module
- edge
- robot
- cloud
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 19
- 230000006835 compression Effects 0.000 claims abstract description 23
- 238000007906 compression Methods 0.000 claims abstract description 23
- 238000012549 training Methods 0.000 claims description 14
- 238000005457 optimization Methods 0.000 claims description 10
- 238000013480 data collection Methods 0.000 claims description 6
- 238000013473 artificial intelligence Methods 0.000 description 84
- 238000011161 development Methods 0.000 description 4
- 230000001149 cognitive effect Effects 0.000 description 2
- 239000008186 active pharmaceutical agent Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
- H04L67/63—Routing a service request depending on the request content or context
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/04—Protocols for data compression, e.g. ROHC
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Theoretical Computer Science (AREA)
- Computing Systems (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computer Security & Cryptography (AREA)
- Manipulator (AREA)
Abstract
The invention discloses a robot AI service method and system based on cloud side end, belonging to the artificial intelligent robot technical field; the method comprises the following specific steps: the cloud end of the S1 automatically updates the AI model and sends the AI model to the edge end; s2, compressing and storing the AI model by the edge end according to different compression ratios; s3, operating the compressed AI model at the robot end; s4 robot selects AI model to reason according to user service state; the invention provides a robot AI service method and system based on a cloud side end, which integrates high computing power of cloud computing, timeliness of edge computing and AI computing capability of the robot end by utilizing a cloud end, the edge end and the robot end to provide two online and offline AI service modes for users so as to meet requirements of different users.
Description
Technical Field
The invention discloses a robot AI (Artificial intelligence) service method and system based on a cloud edge, and relates to the technical field of artificial intelligence robots.
Background
With the development of cloud computing and big data, various application systems and services are gradually clouded in various industry fields, deployed at a cloud end, collect more and more hardware resources at the cloud end, and the computing power of the cloud end is stronger and stronger. The rapid development of cloud computing and big data brings a new wave of artificial intelligence, the commercialization speed of the cloud computing and big data exceeds expectations, the cloud computing and big data is an important development strategy of all countries, the cloud computing and big data are also a novel field in which curve overtaking is expected in China, the optimization of an artificial intelligence model depends on training of a large number of data sets, and the model training needs super-strong computing power, so that the combination of the cloud computing and the artificial intelligence is inevitable.
At present, corresponding AI cloud services are introduced by various cloud service merchants, such as Microsoft Azure cognitive service, IBM Watson cognitive service, Tencent cloud, Ali cloud and the like, which are provided by the AI cloud services, users need to access the AI services of cloud platforms online through corresponding APIs, and do not support offline, and users need to upload own data sets if needing to use the AI services, so that personal data privacy is threatened easily.
The development of the embedded AI chip enables the robot computing unit to have certain AI computing capability, but due to the limitations of power consumption, volume, performance and the like, the computing capability is limited, and a large-scale AI computing model is difficult to operate. If the model is operated at the robot end, the model needs to be compressed;
therefore, the invention provides an online and offline in-the-lang AI service mode for a robot based on a cloud edge terminal AI service method and a system thereof, so as to solve the problems.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a robot AI service method and system based on a cloud edge terminal, and the adopted technical scheme is as follows: a robot AI service method based on cloud side end includes the following steps:
the cloud end of the S1 automatically updates the AI model and sends the AI model to the edge end;
s2, compressing and storing the AI model by the edge end according to different compression ratios;
s3, operating the compressed AI model at the robot end;
and S4, the robot selects an AI model through the edge terminal according to the service state of the user for reasoning.
The specific steps of automatically updating the AI model by the cloud end of the S1 and sending the AI model to the edge end include:
s101, training an AI model by using a public data set and issuing the AI model to an edge terminal;
s102, collecting a user feedback data set from an edge end periodically;
s103, optimizing and adjusting the AI model by utilizing the accumulated user feedback data set;
and S104, updating the AI model and sending the AI model to the edge terminal.
The specific steps of compressing and storing the AI model by the S2 edge terminal according to different compression rates include:
s201, setting a compression rate of an edge end to compress the AI model;
and S202, the edge terminal collects the user feedback data set and uploads the user feedback data set to the cloud for model optimization.
The specific steps of the robot S4 for reasoning through the AI model selected by the edge terminal according to the service state of the user include:
s401, selecting online or offline service by a user;
s402, when the online service is selected, the robot accesses the edge end and puts reasoning on the edge end for execution;
s412, when the offline service is selected, the robot downloads the compressed AI model from the edge end according to the self requirement and locally executes reasoning;
and S403, when the inference is wrong, the robot selects to upload wrong data to the edge terminal.
The utility model provides a robot AI service system based on cloud limit, the system specifically include high in the clouds module, edge end module, model operation module and robot end module:
cloud module: the cloud automatically updates the AI model and sends the AI model to the edge end;
an edge end module: the edge end compresses and stores the AI model according to different compression ratios;
a model operation module: running the compressed AI model at the robot end;
a robot end module: and the robot selects an AI model through the edge terminal according to the service state of the user for reasoning.
The cloud module specifically comprises a model training module, a feedback collecting module, a model adjusting module and a model updating module:
a model training module: training an AI model by using a public data set and issuing the AI model to an edge terminal;
a feedback collection module: collecting user feedback data sets from the edge terminal periodically;
a model adjustment module: optimizing and adjusting the AI model by utilizing the accumulated user feedback data set;
a model updating module: and updating the AI model and sending the AI model to the edge terminal.
The edge end module specifically comprises a compression setting module and a data uploading module:
a compression setting module: setting a compression rate of an edge end to compress the AI model;
the data uploading module: and the edge terminal collects the user feedback data set and uploads the user feedback data set to the cloud terminal for model optimization.
The robot end module specifically comprises a mode selection module, an online service module, an offline service module and a data collection module:
a mode selection module: a user selects an online or offline service;
an online service module: when the online service is selected, the robot accesses the edge terminal and puts reasoning on the edge terminal for execution;
an offline service module: when the offline service is selected, the robot downloads the compressed AI model from the edge end according to the self requirement and locally executes reasoning;
a data collection module: and when the inference is wrong, the robot selects to upload the wrong data to the edge terminal.
The invention has the beneficial effects that: the invention provides a robot AI service method and system based on a cloud side end, which integrates high computing power of cloud computing, timeliness of edge computing and AI computing capability of the robot end by utilizing a cloud end, the edge end and the robot end to provide two online and offline AI service modes for users so as to meet requirements of different users.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flow chart of the method of the present invention; fig. 2 is a schematic diagram of the system of the present invention.
Detailed Description
The present invention is further described below in conjunction with the following figures and specific examples so that those skilled in the art may better understand the present invention and practice it, but the examples are not intended to limit the present invention.
The first embodiment is as follows:
a robot AI service method based on cloud side end includes the following steps:
the cloud end of the S1 automatically updates the AI model and sends the AI model to the edge end;
s2, compressing and storing the AI model by the edge end according to different compression ratios;
s3, operating the compressed AI model at the robot end;
s4 robot selects AI model to reason according to user service state;
further, the specific steps of automatically updating the AI model by the cloud end of S1 and sending the AI model to the edge end include:
s101, training an AI model by using a public data set and issuing the AI model to an edge terminal;
s102, collecting a user feedback data set from an edge end periodically;
s103, optimizing and adjusting the AI model by utilizing the accumulated user feedback data set;
s104, updating the AI model and issuing the AI model to the edge terminal;
further, the specific steps of compressing and storing the AI model by the S2 edge according to different compression rates include:
s201, setting a compression rate of an edge end to compress the AI model;
and S202, the edge terminal collects the user feedback data set and uploads the user feedback data set to the cloud for model optimization.
Still further, the specific steps of the S4 robot performing inference through the edge-selected AI model according to the service status of the user include:
s401, selecting online or offline service by a user;
s402, when the online service is selected, the robot accesses the edge end and puts reasoning on the edge end for execution;
s412, when the offline service is selected, the robot downloads the compressed AI model from the edge end according to the self requirement and locally executes reasoning;
s403, when the inference is wrong, the robot selects to upload wrong data to an edge end;
the method is based on a cloud end, an edge end and a robot end; when the method is used, the cloud end trains the AI model by using a large number of public data sets and sends the model to the side end; collecting user feedback data sets from the edge terminal regularly, and after the accumulated data sets reach a certain number, the cloud terminal calls the parameter optimization model on the original model and sends the parameter optimization model to the edge terminal to update the model;
the edge terminal is responsible for receiving and storing the model issued by the cloud terminal, and meanwhile, the model is compressed and stored according to different compression ratios; providing AI service for the user, including using the original model to provide online AI service and providing the user with the downloading of the compressed model, and the user can use the offline AI service after downloading the compressed model; collecting a user feedback data set and uploading the user feedback data set to a cloud for model optimization;
the user robot end has certain intelligent computing power and computing power, a user can select online or offline service, the online service, namely accessing the edge end, puts inference on the edge end for execution, and the offline service, namely downloading a compressed model from the edge end according to self requirements and locally executing inference; if inference errors occur, error data can be selectively uploaded to the edge terminal,
example two:
the utility model provides a robot AI service system based on cloud limit, the system specifically include high in the clouds module, edge end module, model operation module and robot end module:
cloud module: the cloud automatically updates the AI model and sends the AI model to the edge end;
an edge end module: the edge end compresses and stores the AI model according to different compression ratios;
a model operation module: running the compressed AI model at the robot end;
a robot end module: and the robot selects an AI model through the edge terminal according to the service state of the user for reasoning.
The cloud module specifically comprises a model training module, a feedback collecting module, a model adjusting module and a model updating module:
a model training module: training an AI model by using a public data set and issuing the AI model to an edge terminal;
a feedback collection module: collecting user feedback data sets from the edge terminal periodically;
a model adjustment module: optimizing and adjusting the AI model by utilizing the accumulated user feedback data set;
a model updating module: and updating the AI model and sending the AI model to the edge terminal.
The edge end module specifically comprises a compression setting module and a data uploading module:
a compression setting module: setting a compression rate of an edge end to compress the AI model;
the data uploading module: and the edge terminal collects the user feedback data set and uploads the user feedback data set to the cloud terminal for model optimization.
The robot end module specifically comprises a mode selection module, an online service module, an offline service module and a data collection module:
a mode selection module: a user selects an online or offline service;
an online service module: when the online service is selected, the robot accesses the edge terminal and puts reasoning on the edge terminal for execution;
an offline service module: when the offline service is selected, the robot downloads the compressed AI model from the edge end according to the self requirement and locally executes reasoning;
a data collection module: and when the inference is wrong, the robot selects to upload the wrong data to the edge terminal.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (8)
1. A robot AI service method based on cloud side end is characterized in that the method comprises the following steps:
the cloud end of the S1 automatically updates the AI model and sends the AI model to the edge end;
s2, compressing and storing the AI model by the edge end according to different compression ratios;
s3, operating the compressed AI model at the robot end;
and S4, the robot selects an AI model through the edge terminal according to the service state of the user for reasoning.
2. The method of claim 1, wherein the step of automatically updating the AI model at the cloud end of the S1 and sending the AI model to the edge comprises:
s101, training an AI model by using a public data set and issuing the AI model to an edge terminal;
s102, collecting a user feedback data set from an edge end periodically;
s103, optimizing and adjusting the AI model by utilizing the accumulated user feedback data set;
and S104, updating the AI model and sending the AI model to the edge terminal.
3. The method as claimed in claim 1, wherein the step of compressing and storing the AI model by the S2 edge according to different compression rates comprises:
s201, setting a compression rate of an edge end to compress the AI model;
and S202, the edge terminal collects the user feedback data set and uploads the user feedback data set to the cloud for model optimization.
4. The method as claimed in claim 1, wherein the step of the S4 robot performing inference through the edge-selected AI model based on the service status of the user comprises:
s401, selecting online or offline service by a user;
s402, when the online service is selected, the robot accesses the edge end and puts reasoning on the edge end for execution;
s412, when the offline service is selected, the robot downloads the compressed AI model from the edge end according to the self requirement and locally executes reasoning;
and S403, when the inference is wrong, the robot selects to upload wrong data to the edge terminal.
5. The utility model provides a robot AI service system based on cloud limit, characterized by the system specifically include high in the clouds module, edge end module, model operation module and robot end module:
cloud module: the cloud automatically updates the AI model and sends the AI model to the edge end;
an edge end module: the edge end compresses and stores the AI model according to different compression ratios;
a model operation module: running the compressed AI model at the robot end;
a robot end module: and the robot selects an AI model through the edge terminal according to the service state of the user for reasoning.
6. The system of claim 5, wherein the cloud module specifically comprises a model training module, a feedback collection module, a model adjustment module, and a model update module:
a model training module: training an AI model by using a public data set and issuing the AI model to an edge terminal;
a feedback collection module: collecting user feedback data sets from the edge terminal periodically;
a model adjustment module: optimizing and adjusting the AI model by utilizing the accumulated user feedback data set;
a model updating module: and updating the AI model and sending the AI model to the edge terminal.
7. The system of claim 6, wherein the edge side module specifically comprises a compression setting module and a data uploading module:
a compression setting module: setting a compression rate of an edge end to compress the AI model;
the data uploading module: and the edge terminal collects the user feedback data set and uploads the user feedback data set to the cloud terminal for model optimization.
8. The system of claim 7, wherein the robot-side module specifically comprises a mode selection module, an online service module, an offline service module, and a data collection module:
a mode selection module: a user selects an online or offline service;
an online service module: when the online service is selected, the robot accesses the edge terminal and puts reasoning on the edge terminal for execution;
an offline service module: when the offline service is selected, the robot downloads the compressed AI model from the edge end according to the self requirement and locally executes reasoning;
a data collection module: and when the inference is wrong, the robot selects to upload the wrong data to the edge terminal.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110736921.4A CN113691579A (en) | 2021-06-30 | 2021-06-30 | Robot AI service method and system based on cloud edge |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110736921.4A CN113691579A (en) | 2021-06-30 | 2021-06-30 | Robot AI service method and system based on cloud edge |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113691579A true CN113691579A (en) | 2021-11-23 |
Family
ID=78576812
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110736921.4A Pending CN113691579A (en) | 2021-06-30 | 2021-06-30 | Robot AI service method and system based on cloud edge |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113691579A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107766889A (en) * | 2017-10-26 | 2018-03-06 | 济南浪潮高新科技投资发展有限公司 | A kind of the deep learning computing system and method for the fusion of high in the clouds edge calculations |
CN108093030A (en) * | 2017-11-29 | 2018-05-29 | 杭州古北电子科技有限公司 | A kind of artificial intelligence model dispositions method based on Cloud Server |
CN108667850A (en) * | 2018-05-21 | 2018-10-16 | 济南浪潮高新科技投资发展有限公司 | A kind of artificial intelligence service system and its method for realizing artificial intelligence service |
CN111199279A (en) * | 2019-10-30 | 2020-05-26 | 山东浪潮人工智能研究院有限公司 | Cloud edge calculation and artificial intelligence fusion method and device for police service industry |
-
2021
- 2021-06-30 CN CN202110736921.4A patent/CN113691579A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107766889A (en) * | 2017-10-26 | 2018-03-06 | 济南浪潮高新科技投资发展有限公司 | A kind of the deep learning computing system and method for the fusion of high in the clouds edge calculations |
CN108093030A (en) * | 2017-11-29 | 2018-05-29 | 杭州古北电子科技有限公司 | A kind of artificial intelligence model dispositions method based on Cloud Server |
CN108667850A (en) * | 2018-05-21 | 2018-10-16 | 济南浪潮高新科技投资发展有限公司 | A kind of artificial intelligence service system and its method for realizing artificial intelligence service |
CN111199279A (en) * | 2019-10-30 | 2020-05-26 | 山东浪潮人工智能研究院有限公司 | Cloud edge calculation and artificial intelligence fusion method and device for police service industry |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108550930B (en) | Bridging manager, bridging system and bridging method based on retired power battery pack | |
CN103761309A (en) | Operation data processing method and system | |
CN105045369A (en) | Dynamic CPU frequency adjustment method based on historical user-interaction information | |
CN102984037A (en) | Smart family system | |
CN102076064A (en) | Power control apparatus and method for cluster system | |
CN100382515C (en) | Power and environment concentrated monitoring management system | |
CN108770017B (en) | Dynamic equalization method and system for wireless resources | |
CN102143563A (en) | Control method, equipment and systems of short message center cluster | |
WO2023221846A1 (en) | Computing cluster and data acquisition method and device thereof, and storage medium | |
CN110650518A (en) | Energy-saving method and equipment for base station equipment | |
CN116383289A (en) | Digital twin model data processing method applied to energy storage Internet of things | |
CN105262890A (en) | Information update control system and method for applications on portable electronic device | |
CN113489802B (en) | Internet of things collection equipment | |
CN113691579A (en) | Robot AI service method and system based on cloud edge | |
CN110751382A (en) | Operating system of high-efficiency energy Internet | |
CN116526678B (en) | Intelligent computing center power supply elastic scheduling system and control method thereof | |
CN113806094A (en) | Cloud platform resource dynamic scheduling method based on deep learning | |
CN209605349U (en) | A kind of group control system | |
CN108196797B (en) | Data processing system based on cloud computing | |
CN114900558B (en) | Universal equipment management protocol control method and device | |
CN110689175A (en) | Energy consumption optimization method for distributed green cloud data center with chaotic multiple universes | |
CN112650338B (en) | Energy-saving and environment-friendly forestry seedling detection system and method based on Internet of things | |
CN115243349A (en) | Base station energy saving method and device, electronic equipment and storage medium | |
CN101572881B (en) | Equipment management method in TD-SCDMA | |
CN114301922A (en) | Reverse proxy method with delay perception load balancing and storage device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20211123 |
|
RJ01 | Rejection of invention patent application after publication |