CN114338281A - Model distribution application design method and system based on edge computing gateway - Google Patents

Abstract

The invention provides a model distribution application design method and system based on edge computing gateway. The scheme includes selecting a preset template according to the device model and application scenario through the human-computer interaction interface, and entering the device configuration parameters according to the preset template; using an encryption algorithm to process information of a general model; using an encryption algorithm to process information of a specific model ; Determine whether the edge computing gateway service device has an active acquisition model request, and deploy and update the device computing model through the edge device; obtain model configuration parameters, determine the application scenario of the device, and adjust the parameters of the target model; Determine the edge computing gateway service device Whether there is an abnormal state, and feedback the obtained abnormal state to the cloud management service device. This solution obtains the model of the server by means of instructions, and physically connects with the edge device through the host computer terminal device, and the host computer terminal device deploys the model to the edge device.

Description

A model distribution application design method and system based on edge computing gateway

technical field

The present invention relates to the technical field of edge computing, and more particularly, to a model distribution application design method and system based on an edge computing gateway.

Background technique

With the rapid development of the Internet of Things, the Internet of Everything has become the trend of future development, and embedded node devices will show a substantial growth. The data center on the traditional cloud side interacts with the tens of thousands of IoT nodes, which not only costs a lot of computing resources, but also has network congestion, which is more obvious during peak hours. Although 5G wireless mobile communication with high bandwidth is gradually popularized, some emerging application scenarios, such as: unmanned driving, safety monitoring, roadside perception data, etc., require low latency, high reliability, traditional nodes to the cloud, The needs of these application services cannot be met. Therefore, various industries are now exploring new edge AI computing, which has a crucial impact on the security and performance improvement of IoT systems. Since emerging application services require customization of specific scene models, large-scale deployment of general scene models, and configuration such as parameter quantization, a specific method is urgently needed to solve this difficulty in frequent replacement and deployment. question.

SUMMARY OF THE INVENTION

In view of the above problems, the present invention proposes a model distribution application design method and system based on an edge computing gateway. The model of the server is obtained by means of instructions, and is physically connected to the edge device through the host computer terminal device. Model deployment to edge devices.

According to the first aspect of the embodiments of the present invention, a method for designing a model distribution application based on an edge computing gateway is provided.

In one or more embodiments, preferably, the method for designing a model distribution application based on an edge computing gateway includes:

In the model distribution system of the edge computing gateway, through the human-computer interaction interface, a preset template is selected according to the device model and application scenario, and the device configuration parameters are entered according to the preset template;

Obtain the device configuration parameters, determine whether there is a general model, and after obtaining the general model, use an encryption algorithm to process the information of the general model;

Obtaining the device configuration parameters, judging whether there is a specific model, and after obtaining the specific model, using the encryption algorithm to process the information of the specific model;

judging whether the edge computing gateway service device has a device active acquisition model request, and deploying and updating the device computing model through the edge device;

Obtaining model configuration parameters, judging the application scenario of the device, and adjusting the parameters of the target model;

Judging whether the edge computing gateway service device has an abnormal state, and feeding back the obtained abnormal state to the cloud management service device;

Wherein, the model distribution system of the edge computing gateway includes a cloud management service device and an edge computing gateway service device, wherein the cloud management service device includes an interaction management service device and a processing management service device;

The encryption algorithm specifically adopts a symmetric encryption method, and the specific process is as follows:

Delivering a model distribution request to the edge computing gateway service device at the cloud management service device;

The edge computing gateway service device randomly selects a compression method and an encryption method, and generates a symmetric encryption key;

sending the symmetric encryption key to the processing management service device, parsing the message, and encrypting the model to generate a target model;

delivering the target model to the edge computing gateway service device.

In one or more embodiments, preferably, in the model distribution system of the edge computing gateway, through a human-computer interaction interface, a preset template is selected according to the device model and application scenario, and the device is entered according to the preset template. Configuration parameters, including:

Connect the IoT device node to the edge computing gateway through one or more of R485, network, and IO bus methods;

Set the device model and application scenario through the human-computer interface;

Automatically obtain the corresponding device configuration parameters in the preset calculation model according to the device model and the application scenario;

The corresponding device configuration parameters are issued through one or more of the R485, network, and IO bus modes.

In one or more embodiments, preferably, the device configuration parameters are obtained to determine whether there is a general model, and after the general model is obtained, an encryption algorithm is used to perform information processing of the general model:

Utilize the edge computing gateway service device to issue a general scenario model request;

According to the general scenario model request, the general scenario model information is encrypted and reported to the processing management service device using the encryption algorithm, wherein the general scenario model information includes the device model, performance, compilation chain, application corresponding to the general scenario Scenario type, application scenario adaptability;

Using the processing management service device to report the general scene model information to the interaction management service device;

Retrieve all generic target models from the model database according to the device model, application scenario and compilation chain corresponding to the generic scenario;

The general target model is sent to the edge computing gateway management service device through the processing management service device.

In one or more embodiments, preferably, the device configuration parameters are obtained to determine whether a specific model exists, and after the specific model is obtained, the encryption algorithm is used to perform information processing of the specific model, which specifically includes:

Obtaining a specific scenario model request, the edge computing gateway service device uses the encryption algorithm to report the specific scenario model information, where the specific scenario model information includes the device model, performance, compilation chain, application scenario type corresponding to the specific scenario, Adaptability to application scenarios;

Replying the first reply request to the edge computing gateway service device through the processing management service device;

After the edge computing gateway service device receives the first reply request, the specific scenario model information is sent back to the processing management service device, and then to the interaction management service device;

Screening, marking, and selecting a specific target model for data according to a specific scenario through the interaction management service device;

The interaction management service device feeds back a second reply request to the edge computing gateway service device through the processing management service device.

In one or more embodiments, preferably, the judging whether the edge computing gateway service device has a device active acquisition model request, and deploying and updating the device computing model through the edge device, specifically includes:

Judging whether the edge computing gateway service device has a device active acquisition model request;

Obtain the device's active acquisition model request, and upload the corresponding device parameter value to the cloud management service device;

Obtain the corresponding model according to the preset model library and the corresponding device parameter value, and then package it to generate an active acquisition model through an encryption algorithm;

Obtain the active acquisition model and deliver it to the edge computing gateway service device.

In one or more embodiments, preferably, the acquiring model configuration parameters, judging the application scenario of the device, and adjusting the parameters of the target model specifically include:

obtaining the model configuration parameters in the interaction management service device;

Delivering model parameters to the corresponding edge computing gateway service device according to the model configuration parameters by the processing management service device;

The target model is modified according to the model configuration parameters.

In one or more embodiments, preferably, the judging whether the edge computing gateway service device has an abnormal state, and feeding back the obtained abnormal state to the cloud management service device, specifically includes:

Upload device parameters and abnormal data to the processing management service device according to the edge computing gateway service device;

The processing management service device feeds back the abnormal situation to the interaction management service device;

In the interaction management service device, the staff analyzes according to the situation feedback, and adjusts the target model;

Re-delivery the target model to the edge computing gateway service device.

According to the second aspect of the embodiments of the present invention, a model distribution application design system based on an edge computing gateway is provided.

In one or more embodiments, preferably, the model distribution application design system based on edge computing gateway includes:

The device configuration module is used to select a preset template according to the device model and application scenario through the human-computer interaction interface in the model distribution system of the edge computing gateway, and enter the device configuration parameters according to the preset template;

The general model information module is used to obtain the device configuration parameters, determine whether there is a general model, and after obtaining the general model, use an encryption algorithm to process the information of the general model;

A specific model information module, used to obtain the device configuration parameters, determine whether there is a specific model, and after obtaining the specific model, use the encryption algorithm to process the information of the specific model;

an active model update module, configured to determine whether the edge computing gateway service device has a device actively acquiring a model request, and deploy and update the device computing model through the edge device;

A passive model update module, used for acquiring model configuration parameters, judging the application scenario of the device, and adjusting the parameters of the target model;

The abnormal state analysis module is used for judging whether the edge computing gateway service device has an abnormal state, and feeding back the obtained abnormal state to the cloud management service device.

According to a third aspect of the embodiments of the present invention, there is provided a computer-readable storage medium on which computer program instructions are stored, and when executed by a processor, the computer program instructions implement any one of the first aspect of the embodiments of the present invention. method described.

According to a fourth aspect of an embodiment of the present invention, an electronic device is provided, including a memory and a processor, the memory being used to store one or more computer program instructions, wherein the one or more computer program instructions are processed by the processing The controller executes the steps described in any one of the first aspects of the embodiments of the present invention.

The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

1) The scheme of the present invention has the characteristics of a high degree of automation. Specifically, it realizes the unified management of a unified cloud center, reduces manual intervention, realizes automatic data monitoring, and eliminates abnormal faults;

2) The solution of the present invention has the characteristics of high equipment flexibility and strong expansibility. Specifically, the remote deployment, distribution and upgrade models can be targeted according to the needs of different actual scenarios, which improves the flexibility and application of end-side edge computing. Through this function, it can adapt to different application scenarios and realize different functions according to different hardware peripherals;

3) The solution of the present invention has the characteristics of convenient operation and maintenance and low cost. Specifically, it facilitates software upgrade and maintenance, does not need to go to the site to deploy the model of the edge computing gateway, and reduces manual maintenance, especially when the edge computing gateway is used. When the number of data reaches a certain order of magnitude, large-scale general model deployment can be realized through cloud services, with less cost and expense;

4) The solution of the present invention has the characteristics of high security. Specifically, compared with the on-site deployment of the host computer, the cloud-to-edge gateway method is directly adopted, which greatly prevents the leakage of data packets and solves the problem of information security.

Other features and advantages of the present invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description, claims, and drawings.

The technical solutions of the present invention will be further described in detail below through the accompanying drawings and embodiments.

Description of drawings

In order to illustrate the technical solutions in the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings used in the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained from these drawings without creative effort.

FIG. 1 is a flowchart of a method for designing a model distribution application based on an edge computing gateway according to an embodiment of the present invention.

Fig. 2 is a model distribution application design method based on an edge computing gateway in an embodiment of the present invention, in the model distribution system of the edge computing gateway, through the human-computer interaction interface, according to the device model and application scenario selection preset template, A flowchart of inputting device configuration parameters according to the preset template.

3 is an embodiment of the present invention in a model distribution application design method based on an edge computing gateway to obtain the device configuration parameters, determine whether there is a general model, after obtaining the general model, use an encryption algorithm to perform information on the general model Process flow chart.

FIG. 4 is a method for obtaining the device configuration parameters in a model distribution application design method based on an edge computing gateway according to an embodiment of the present invention, judging whether there is a specific model, and after obtaining the specific model, using the encryption algorithm to perform the specific model The flow chart of information processing.

Fig. 5 is a method for designing a model distribution application based on an edge computing gateway according to an embodiment of the present invention, which is a process of judging whether the edge computing gateway service device has a device active acquisition model request, and deploying and updating the device computing model through the edge device. flow chart.

6 is a flowchart of obtaining model configuration parameters, judging the application scenario of the device, and adjusting parameters of the target model in an edge computing gateway-based model distribution application design method according to an embodiment of the present invention.

FIG. 7 is a method for judging whether the edge computing gateway service device has an abnormal state in an edge computing gateway-based model distribution application design method according to an embodiment of the present invention, and feeding back the obtained abnormal state to the cloud management service device flow chart.

FIG. 8 is a structural diagram of a model distribution application design system based on an edge computing gateway according to an embodiment of the present invention.

FIG. 9 is a structural diagram of an electronic device in an embodiment of the present invention.

Detailed ways

In some of the processes described in the description and claims of the present invention and the above-mentioned drawings, various operations are included in a specific order, but it should be clearly understood that these operations may not be in accordance with the order in which they appear herein. For execution or parallel execution, the sequence numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the sequence numbers themselves do not represent any execution order. Additionally, these flows may include more or fewer operations, and these operations may be performed sequentially or in parallel. It should be noted that the descriptions such as "first" and "second" in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit "first" and "second" are different types.

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present invention.

With the rapid development of the Internet of Things, the Internet of Everything has become the trend of future development, and embedded node devices will show a substantial growth. The data center on the traditional cloud side interacts with the tens of thousands of IoT nodes, which not only costs a lot of computing resources, but also has network congestion, which is more obvious during peak hours. Although 5G wireless mobile communication with high bandwidth is gradually popularized, some emerging application scenarios, such as: unmanned driving, safety monitoring, roadside perception data, etc., require low latency, high reliability, traditional nodes to the cloud, The needs of these application services cannot be met. Therefore, various industries are now exploring new edge AI computing, which has a crucial impact on the security and performance improvement of IoT systems. Since emerging application services require customization of specific scene models, large-scale deployment of general scene models, and configuration such as parameter quantization, a specific method is urgently needed to solve this difficulty in frequent replacement and deployment. question.

Before the technology of the present invention, in the prior art, for the model deployment of the edge computing gateway device, the manual deployment scheme is still adopted, which has the following problems:

1) High operational requirements for staff. Due to the deployment of equipment on site and the deployment of models through physical connection, sometimes the installation environment of the edge computing gateway is too harsh and inconvenient to operate, which provides extremely strict usage requirements for on-site support.

2) The operation and maintenance cost is high and the efficiency is low. The host computer terminal equipment needs to be geographically close to the edge terminal equipment for physical connection, which provides extremely strict usage requirements for the on-site support of the host computer terminal equipment, so specific training for the staff is required. Moreover, the staff can only operate and maintain one edge device, which requires a lot of labor cost and time cost, and the operation and maintenance efficiency is extremely low.

3) Central data is difficult to aggregate statistics. Due to the complex deployment process of edge computing gateways, configuration such as parameter quantification is required, and different application scenarios are required for distribution, and different models are required for distribution. Manual deployment is difficult to collect detailed information and feed it back to the server in real time.

4) Fallibility. Due to the complexity of the deployment environment and a large number of devices, it is easy to cause deployment errors, resulting in abnormal failures and difficulties in troubleshooting.

5) Security is not high. Due to the manual deployment on site and the deployment of the host computer, the computing model data packets of the edge computing gateway are required, and the lack of security certification is likely to cause data packet leakage.

In the embodiments of the present invention, a method and system for designing a model distribution application based on an edge computing gateway are provided. This solution obtains the model of the server by means of instructions, and physically connects with the edge device through the host computer terminal device, and the host computer terminal device deploys the model to the edge device.

According to the first aspect of the embodiments of the present invention, a method for designing a model distribution application based on an edge computing gateway is provided.

FIG. 1 is a flowchart of a method for designing a model distribution application based on an edge computing gateway according to an embodiment of the present invention.

As shown in FIG. 1, in one or more embodiments, preferably, the method for designing a model distribution application based on an edge computing gateway includes:

S101. In a model distribution system of an edge computing gateway, through a human-computer interaction interface, select a preset template according to the device model and application scenario, and enter device configuration parameters according to the preset template;

S102, obtaining the device configuration parameters, judging whether there is a general model, and after obtaining the general model, use an encryption algorithm to process the information of the general model;

S103, obtaining the device configuration parameters, judging whether there is a specific model, and after obtaining the specific model, use the encryption algorithm to process the information of the specific model;

S104. Determine whether the edge computing gateway service device has a device active acquisition model request, and deploy and update the device computing model through the edge device;

S105, obtaining model configuration parameters, judging the application scenario of the device, and adjusting the parameters of the target model;

S106. Determine whether the edge computing gateway service device has an abnormal state, and feed back the obtained abnormal state to the cloud management service device;

Wherein, the model distribution system of the edge computing gateway includes a cloud management service device and an edge computing gateway service device, wherein the cloud management service device includes an interaction management service device and a processing management service device;

The encryption algorithm specifically adopts a symmetric encryption method, and the specific process is as follows:

Delivering a model distribution request to the edge computing gateway service device at the cloud management service device;

The edge computing gateway service device randomly selects a compression method and an encryption method, and generates a symmetric encryption key;

sending the symmetric encryption key to the processing management service device, parsing the message, and encrypting the model to generate a target model;

delivering the target model to the edge computing gateway service device.

In the embodiment of the present invention, the edge computing gateway collects and summarizes the information of the nodes of the Internet of Things, such as point cloud data of lidar, roadside images of cameras, weather temperature and humidity, etc., as the original data of the training model. The staff filters and filters the data in the cloud through the human-computer interaction interface. After training, the corresponding model is compiled according to the edge computing gateway model and application scenario, and the corresponding parameters are specified and distributed to the corresponding edge computing gateway. Scale Distribute general models to edge computing gateways, or specify difficult deployment problems for specific edge computing gateways. At the same time, when the model of the edge computing gateway is abnormal, the data can be sent back to the cloud, which is convenient for recording and statistics, and solves the problem of difficult troubleshooting during model distribution. By deploying the management system for model distribution in the cloud, the model of the remote edge computing gateway can be managed, deployed and updated frequently, which solves the complex problem of edge computing gateway operation and maintenance, thereby reducing operating costs.

Specifically, the encryption adopts a symmetric encryption method. The message protocol in the transmission process includes an encryption key, a compression method, and an encryption type. The encryption type adopts the commonly used AES, DES, etc., and the compression methods zip, rar, etc. The packet data is parsed, the corresponding value is obtained, the model firmware is encrypted and compressed, and then sent to the client. Finally, the client decompresses, decrypts and deploys the model. Finally, the distribution process of the specific model of the general model machine is realized.

Fig. 2 is a model distribution application design method based on an edge computing gateway in an embodiment of the present invention, in the model distribution system of the edge computing gateway, through the human-computer interaction interface, according to the device model and application scenario selection preset template, A flowchart of inputting device configuration parameters according to the preset template.

As shown in FIG. 2 , in one or more embodiments, preferably, in the model distribution system of the edge computing gateway, a preset template is selected according to the device model and application scenario through a human-computer interaction interface, and according to the described Preset templates enter device configuration parameters, including:

S201. Connect the IoT device node to the edge computing gateway through one or more of R485, network, and IO bus;

S202, setting the device model and application scenario through the human-computer interaction interface;

S203, automatically obtaining the corresponding device configuration parameters in a preset computing model according to the device model and the application scenario;

S204 , delivering the corresponding device configuration parameters through one or more of R485, network, and IO bus modes.

In the embodiment of the present invention, based on the cloud deployment model distribution technology, the cost of deep learning model deployment, operation and maintenance, and technology iteration is reduced.

The cloud includes the following services: human-computer interaction management, model distribution, data processing, data interaction services, and OTA. The cloud is mainly based on the model distribution service, and other services are used as auxiliary modules for realizing model distribution.

Specifically, OTA realizes the technology of remote management of mobile terminal equipment and SIM card data through the air interface of mobile communication.

Specifically, an edge computing gateway is a gateway deployed on the edge side of the network. It connects the physical and digital worlds through functions such as network connection and protocol conversion, and provides lightweight connection management, real-time data analysis, and application management functions.

Specifically, different application scenarios have different computing models. Specific instructions: traffic flow analysis using cameras, object recognition and tracking using lidar, regional weather forecasting using temperature and humidity sensor nodes, intelligent dynamic control of traffic lights, pedestrian recognition using cameras, red light running recognition and other algorithms.

3 is an embodiment of the present invention in a model distribution application design method based on an edge computing gateway to obtain the device configuration parameters, determine whether there is a general model, after obtaining the general model, use an encryption algorithm to perform information on the general model Process flow chart.

As shown in FIG. 3 , in one or more embodiments, preferably, the device configuration parameters are obtained to determine whether there is a general model, and after obtaining the general model, an encryption algorithm is used to perform information processing of the general model:

S301, using the edge computing gateway service device to issue a general scenario model request;

S302. According to the general scenario model request, encrypt the general scenario model information using the encryption algorithm and report it to the processing management service device, where the general scenario model information includes the device model, performance, compilation chain corresponding to the general scenario , Application scenario type, application scenario adaptability;

S303. Use the processing management service device to report the general scene model information to the interaction management service device;

S304. Retrieve all general target models from the model database according to the device model, application scenario and compilation chain corresponding to the general scenario;

S305. Send the general target model to the edge computing gateway management service device through the processing management service device.

In the embodiment of the present invention, after the device is successfully initialized, the general scene model information (basic information such as device model, performance, compilation chain, application scene type, application scene applicability, etc.) is registered and reported, and the management service will configure the model parameters and types. Perform protocol analysis to identify a certain application scenario of this device as a general type. A large number of models are pre-stored in the cloud server. Through database matching, an appropriate lightweight library is selected mainly according to the performance of the device, and the application scenario is generated. Arguments to compile and then selectively compile through the compile chain. If there is a compilation error, the staff will deal with the warning after the warning. The solution is based on the automatic distribution method, which stores information in the cloud to solve the problem of frequently deployed software update records. At the same time, the abnormal information reporting method is convenient for fault analysis, and realizes the flexibility and high scalability of the application of the pull solution.

FIG. 4 is a method for obtaining the device configuration parameters in a model distribution application design method based on an edge computing gateway according to an embodiment of the present invention, judging whether there is a specific model, and after obtaining the specific model, using the encryption algorithm to perform the specific model The flow chart of information processing.

As shown in FIG. 4 , in one or more embodiments, preferably, the device configuration parameters are obtained to determine whether there is a specific model, and after the specific model is obtained, the encryption algorithm is used to process the information of the specific model. , including:

S401. Obtain a request for a specific scenario model, and the edge computing gateway service device uses the encryption algorithm to report the specific scenario model information, where the specific scenario model information includes the device model, performance, compilation chain, and application scenario corresponding to the specific scenario. Type, application scenario adaptability;

S402, replying a first reply request to the edge computing gateway service device through the processing management service device;

S403. After the edge computing gateway service device receives the first reply request, return the specific scenario model information to the processing management service device, and then transmit it to the interaction management service device;

S404. Screen, mark, and select a specific target model for data according to a specific scenario through the interaction management service device;

S405. The interaction management service device feeds back a second reply request to the edge computing gateway service device through the processing management service device.

In the embodiment of the present invention, the process of registering on the device side is the same as that of the general model. The information collected by itself, such as images and point clouds, is uploaded to the server, and the staff screen, mark and select the model training according to the specific scene to form a specific model. A model of the scene. The distribution process of the model involves the method of secure encryption. The encryption adopts the symmetric encryption method. The message protocol in the transmission process includes the encryption key, the compression method, and the encryption type. The encryption type adopts the commonly used AES, DES, etc., and the compression method zip, rar, etc., the server parses the message data, obtains the corresponding value, encrypts and compresses the model firmware, and then sends it to the client. Finally, the client decompresses, decrypts and deploys the model. The solution has the function of security authentication, which can better protect intellectual property rights and prevent information leakage.

Fig. 5 is a method for judging whether the edge computing gateway service device has a request to actively acquire a model in an edge computing gateway-based model distribution application design method according to an embodiment of the present invention, and deploying and updating the device computing model through the edge device flow chart.

As shown in FIG. 5 , in one or more embodiments, preferably, the judging whether the edge computing gateway service device has a device active acquisition model request, and the edge device is used to deploy and update the device computing model, specifically including:

S501. Determine whether the edge computing gateway service device has a device active acquisition model request;

S502, obtaining the device's active acquisition model request, and uploading the corresponding device parameter value to the cloud management service device;

S503, obtaining a corresponding model according to a preset model library and the corresponding device parameter value, and then packaging, and generating an active acquisition model through an encryption algorithm;

S504. Obtain the active acquisition model and deliver it to the edge computing gateway service device.

In the embodiment of the present invention, a visual operation interface is provided, which is easy to operate and maintain.

6 is a flowchart of obtaining model configuration parameters, judging the application scenario of the device, and adjusting parameters of the target model in an edge computing gateway-based model distribution application design method according to an embodiment of the present invention.

As shown in FIG. 6 , in one or more embodiments, preferably, the acquiring model configuration parameters, judging the application scenario of the device, and adjusting the parameters of the target model specifically include:

S601. Obtain the model configuration parameter in the interaction management service device;

S602, delivering model parameters to the corresponding edge computing gateway service device according to the model configuration parameters by the processing management service device;

S603. Modify the target model according to the model configuration parameter.

In the embodiment of the present invention, the staff configures the specified device parameters through the human-computer interaction interface, the management service model generates configuration information according to the parameters, and delivers it to the specified device through the network, and the edge computing gateway modifies according to the delivery model parameters, Used for identification of corresponding application scenarios. Finally, based on the cloud deployment model and the information interaction method through the edge gateway, the generalized large-scale deployment and specific deployment functions can be performed.

FIG. 7 is a method for judging whether the edge computing gateway service device has an abnormal state in an edge computing gateway-based model distribution application design method according to an embodiment of the present invention, and feeding back the obtained abnormal state to the cloud management service device flow chart.

As shown in FIG. 7 , in one or more embodiments, preferably, the judging whether the edge computing gateway service device has an abnormal state, and feeding back the obtained abnormal state to the cloud management service device, specifically includes: :

S701. Upload device parameters and abnormal data to the processing management service device according to the edge computing gateway service device;

S702, the processing management service device feeds back the abnormal situation to the interaction management service device;

S703, in the interaction management service device, the staff analyzes according to the situation feedback, and adjusts the target model;

S704. Re-deliver the target model to the edge computing gateway service device.

In the embodiment of the present invention, the staff analyzes the abnormal situation through the human-computer interaction interface, reconfigures and specifies the equipment parameters, and regenerates the model according to the parameters of the management service model, and sends it to the specified equipment through the network. The gateway is deployed according to the delivery model, so as to solve the problem of abnormality of different applications.

According to the second aspect of the embodiments of the present invention, a model distribution application design system based on an edge computing gateway is provided.

FIG. 8 is a structural diagram of a model distribution application design system based on an edge computing gateway according to an embodiment of the present invention.

As shown in FIG. 8 , in one or more embodiments, preferably, the model distribution application design system based on an edge computing gateway includes:

The device configuration module 801 is used to select a preset template according to the device model and application scenario through the human-computer interaction interface in the model distribution system of the edge computing gateway, and enter the device configuration parameters according to the preset template;

The general model information module 802 is used to obtain the device configuration parameters, determine whether there is a general model, and after obtaining the general model, use an encryption algorithm to process the information of the general model;

The specific model information module 803 is used to obtain the device configuration parameters, determine whether there is a specific model, and after obtaining the specific model, use the encryption algorithm to process the information of the specific model;

An active model update module 804, configured to determine whether the edge computing gateway service device has a device active acquisition model request, and deploy and update the device computing model through the edge device;

A passive model update module 805, configured to obtain model configuration parameters, determine the application scenario of the device, and adjust the parameters of the target model;

The abnormal state analysis module 806 is configured to determine whether the edge computing gateway service device has an abnormal state, and feed back the obtained abnormal state to the cloud management service device.

In the embodiment of the present invention, general and specific models are deployed, models are actively acquired by devices, and direct parameter delivery and modification models, exception handling, and the like are configured. Realize the unified management of edge gateways in the unified cloud center, reduce manual intervention, realize automatic data monitoring, and eliminate abnormal faults. According to the needs of different actual scenarios, targeted remote deployment, distribution, and upgrade models have improved the flexibility and application scope of edge computing on the device side, adapted to different application scenarios, and realized different computing model functions; there is no need to go to the site for edge computing. The gateway deploys the model, which reduces manual maintenance and greatly reduces costs and expenses. On the other hand, the cloud directly to the device is adopted, and the security encryption authentication method is adopted, which greatly avoids the problem of data packet security leakage.

According to a third aspect of the embodiments of the present invention, there is provided a computer-readable storage medium on which computer program instructions are stored, and when executed by a processor, the computer program instructions implement any one of the first aspect of the embodiments of the present invention. method described.

According to a fourth aspect of the embodiments of the present invention, an electronic device is provided. FIG. 9 is a structural diagram of an electronic device in an embodiment of the present invention. The electronic device shown in FIG. 9 is a general edge computing gateway model distribution apparatus, which includes a general computer hardware structure, which at least includes a processor 901 and a memory 902 . The processor 901 and the memory 902 are connected by a bus 903 . Memory 902 is adapted to store instructions or programs executable by processor 901 . The processor 901 may be an independent microprocessor, or may be a set of one or more microprocessors. Thus, the processor 901 executes the instructions stored in the memory 902 to execute the above-described method flow of the embodiments of the present invention to process data and control other devices. The bus 903 connects the above-mentioned various components together, while connecting the above-mentioned components to the display controller 904 and the display device and the input/output (I/O) device 905 . The input/output (I/O) device 905 may be a mouse, keyboard, modem, network interface, touch input device, somatosensory input device, printer, and other devices known in the art. Typically, input/output devices 905 are connected to the system through input/output (I/O) controllers 906 .

The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

1) The scheme of the present invention has the characteristics of a high degree of automation. Specifically, it realizes the unified management of a unified cloud center, reduces manual intervention, realizes automatic data monitoring, and eliminates abnormal faults;

2) The solution of the present invention has the characteristics of high equipment flexibility and strong expansibility. Specifically, the remote deployment, distribution and upgrade models can be targeted according to the needs of different actual scenarios, which improves the flexibility and application of end-side edge computing. Through this function, it can adapt to different application scenarios and realize different functions according to different hardware peripherals;

3) The solution of the present invention has the characteristics of convenient operation and maintenance and low cost. Specifically, it facilitates software upgrade and maintenance, does not need to go to the site to deploy the model of the edge computing gateway, and reduces manual maintenance, especially when the edge computing gateway is used. When the number of data reaches a certain order of magnitude, large-scale general model deployment can be realized through cloud services, with less cost and expense;

4) The solution of the present invention has the characteristics of high security. Specifically, compared with the on-site deployment of the host computer, the cloud-to-edge gateway method is directly adopted, which greatly prevents the leakage of data packets and solves the problem of information security.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media having computer-usable program code embodied therein, including but not limited to disk storage, optical storage, and the like.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each process and/or block in the flowchart illustrations and/or block diagrams, and combinations of processes and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device produce Means for implementing the functions specified in a flow or flow of a flowchart and/or a block or blocks of a block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions The apparatus implements the functions specified in the flow or flow of the flowcharts and/or the block or blocks of the block diagrams.

These computer program instructions can also be loaded on a computer or other programmable data processing device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process such that The instructions provide steps for implementing the functions specified in the flow or blocks of the flowcharts and/or the block or blocks of the block diagrams.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit and scope of the invention. Thus, provided that these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims (10)

1. A model distribution application design method based on an edge computing gateway is characterized by specifically comprising the following steps:

selecting a preset template in a model distribution system of an edge computing gateway through a human-computer interaction interface according to the equipment model and an application scene, and inputting equipment configuration parameters according to the preset template;

acquiring the equipment configuration parameters, judging whether a universal model exists or not, and processing the information of the universal model by using an encryption algorithm after the universal model is acquired;

acquiring the equipment configuration parameters, judging whether a specific model exists or not, and processing information of the specific model by using the encryption algorithm after the specific model is acquired;

judging whether the edge computing gateway service equipment has an equipment active model acquisition request or not, and carrying out deployment updating on an equipment computing model through the edge equipment;

obtaining model configuration parameters, judging an application scene of equipment, and performing parameter adjustment on the target model;

judging whether the edge computing gateway service equipment has an abnormal state or not, and feeding back the obtained abnormal state to cloud management service equipment;

the model distribution system of the edge computing gateway comprises cloud management service equipment and edge computing gateway service equipment, wherein the cloud management service equipment comprises interaction management service equipment and processing management service equipment;

the encryption algorithm specifically adopts a symmetric encryption mode, and the specific flow is as follows:

issuing a model distribution request to the edge computing gateway service equipment at the cloud management service equipment;

the edge computing gateway service equipment randomly selects a compression mode and an encryption mode to generate a symmetric encryption key;

sending the symmetric encryption key to the processing management service equipment, analyzing a message, and encrypting a model to generate a target model;

and issuing the target model to the edge computing gateway service equipment.

2. The method according to claim 1, wherein the model distribution application design method based on the edge computing gateway selects a preset template according to the device model and the application scene through a human-computer interface in the model distribution system of the edge computing gateway, and inputs device configuration parameters according to the preset template, and specifically comprises:

connecting the Internet of things equipment node with the edge computing gateway through one or more of R485, network and IO bus modes;

setting equipment models and application scenes through a human-computer interaction interface;

automatically obtaining corresponding equipment configuration parameters in a preset calculation model according to the equipment model and the application scene;

and issuing the corresponding equipment configuration parameters through one or more of R485, network and IO bus modes.

3. The method according to claim 1, wherein the obtaining of the device configuration parameters, the determining of whether the generic model exists, and after obtaining the generic model, performing information processing of the generic model using an encryption algorithm:

sending a universal scene model request by utilizing the edge computing gateway service equipment;

encrypting and reporting general scene model information to the processing management service equipment by using the encryption algorithm according to the general scene model request, wherein the general scene model information comprises equipment model, performance, a compiling chain, an application scene type and application scene adaptability corresponding to a general scene;

reporting the general scene model information to the interactive management service equipment by using the processing management service equipment;

retrieving all the general target models from a model database according to the equipment models, the application scenes and the compiling chains corresponding to the general scenes;

and sending the general target model to edge computing gateway management service equipment through the processing management service equipment.

4. The method according to claim 1, wherein the obtaining of the device configuration parameters, the determining of whether a specific model exists, and after obtaining the specific model, performing information processing on the specific model by using the encryption algorithm specifically includes:

acquiring a specific scene model request, and reporting specific scene model information by the edge computing gateway service equipment by using the encryption algorithm, wherein the specific scene model information comprises equipment model, performance, compiling chain, application scene type and application scene adaptability corresponding to a specific scene;

replying a first reply request to the edge computing gateway service equipment through the processing management service equipment;

after the edge computing gateway service equipment receives the first reply request, the specific scene model information is transmitted back to the processing management service equipment and then transmitted to the interaction management service equipment;

screening, marking and selecting a specific target model for data according to a specific scene through the interactive management service equipment;

and the interactive management service equipment feeds back a second reply request to the edge computing gateway service equipment through the processing management service equipment.

5. The method according to claim 1, wherein the determining whether the edge computing gateway service device has a device active model acquisition request includes:

judging whether the edge computing gateway service equipment has an equipment active model acquisition request or not;

acquiring a model request actively acquired by the equipment, and uploading a corresponding equipment parameter value to the cloud management service equipment;

obtaining a corresponding model according to a preset model library and the corresponding equipment parameter values, further packaging, and generating an active acquisition model through an encryption algorithm;

and obtaining the active acquisition model and issuing the active acquisition model to the edge computing gateway service equipment.

6. The method according to claim 1, wherein the obtaining of the model configuration parameters, the determining of the application scenario of the device, and the parameter adjustment of the target model specifically comprise:

obtaining the model configuration parameters in the interaction management service equipment;

issuing model parameters to the corresponding edge computing gateway service equipment through the processing management service equipment according to the model configuration parameters;

and modifying the target model according to the model configuration parameters.

7. The method according to claim 1, wherein the determining whether the edge computing gateway service device has an abnormal state and feeding back the obtained abnormal state to a cloud management service device specifically includes:

uploading the parameters and abnormal data of the equipment to the processing management service equipment according to the edge computing gateway service equipment;

the processing management service equipment feeds back the abnormal condition to the interactive management service equipment;

in the interactive management service equipment, a worker analyzes according to condition feedback and adjusts the target model;

and re-issuing the target model to the edge computing gateway service equipment.

8. A model distribution application design system based on an edge computing gateway, the system comprising:

the device configuration module is used for selecting a preset template according to the device model and the application scene through a human-computer interaction interface in a model distribution system of the edge computing gateway, and inputting device configuration parameters according to the preset template;

the universal model information module is used for acquiring the equipment configuration parameters, judging whether a universal model exists or not, and processing the information of the universal model by using an encryption algorithm after the universal model is acquired;

the specific model information module is used for acquiring the equipment configuration parameters, judging whether a specific model exists or not, and processing the information of the specific model by using the encryption algorithm after the specific model is acquired;

the active model updating module is used for judging whether the edge computing gateway service equipment has an equipment active model acquisition request or not and carrying out deployment updating on the equipment computing model through the edge equipment;

the passive model updating module is used for acquiring model configuration parameters, judging an application scene of equipment and adjusting parameters of the target model;

and the abnormal state analysis module is used for judging whether the edge computing gateway service equipment has an abnormal state or not and feeding back the obtained abnormal state to the cloud management service equipment.

9. A computer-readable storage medium on which computer program instructions are stored, which, when executed by a processor, implement the method of any one of claims 1-8.

10. An electronic device comprising a memory and a processor, wherein the memory is configured to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the steps of any of claims 1-8.

Images