WO2020143094A1 - 一种基于边缘计算的智能管理方法和系统 - Google Patents

一种基于边缘计算的智能管理方法和系统 Download PDF

Info

Publication number
WO2020143094A1
WO2020143094A1 PCT/CN2019/074462 CN2019074462W WO2020143094A1 WO 2020143094 A1 WO2020143094 A1 WO 2020143094A1 CN 2019074462 W CN2019074462 W CN 2019074462W WO 2020143094 A1 WO2020143094 A1 WO 2020143094A1
Authority
WO
WIPO (PCT)
Prior art keywords
edge computing
data
software
original data
computing software
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.)
Ceased
Application number
PCT/CN2019/074462
Other languages
English (en)
French (fr)
Inventor
刘永超
邹焕英
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wangsu Science and Technology Co Ltd
Original Assignee
Wangsu Science and Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Wangsu Science and Technology Co Ltd filed Critical Wangsu Science and Technology Co Ltd
Priority to US17/421,391 priority Critical patent/US20220086052A1/en
Priority to EP19908290.0A priority patent/EP3907641A4/en
Publication of WO2020143094A1 publication Critical patent/WO2020143094A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/04Network management architectures or arrangements
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00Computing arrangements based on specific mathematical models
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14Network analysis or design
    • H04L41/142Network analysis or design using statistical or mathematical methods
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/16Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/12Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/2866Architectures; Arrangements
    • H04L67/289Intermediate processing functionally located close to the data consumer application, e.g. in same machine, in same home or in same sub-network
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L69/00Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
    • H04L69/08Protocols for interworking; Protocol conversion
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00Machine learning
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/01Protocols
    • H04L67/10Protocols in which an application is distributed across nodes in the network

Definitions

  • the invention relates to the technical field of edge computing, in particular to an intelligent management method and system based on edge computing.
  • Enterprises can connect machine equipment to the network, and use the cloud computing platform to centrally calculate the data generated by the machine equipment, so as to realize more efficient decision management of the enterprise through the calculation results.
  • the machine equipment of an enterprise when the machine equipment of an enterprise is running, it can periodically upload the locally generated raw data to the cloud computing platform through the Internet of Things technology. After that, the cloud computing platform can aggregate and store the received raw data, and perform centralized analysis and calculation. Then, the cloud computing platform can feed back the obtained calculation results to the machine equipment and mobile management terminal of the enterprise, so that the machine equipment automatically runs according to the calculation results, or manually manages the inside of the enterprise according to the calculation results through the mobile management terminal.
  • the scale of the enterprise has expanded, and the number of IoT machines and devices has increased rapidly. Correspondingly, the amount of data that needs to be transferred to the cloud on the machines and devices of the enterprise has also increased dramatically.
  • the original cloud computing platform for computing management will consume a lot of Bandwidth resources are used to transmit data generated by machine equipment, and the feedback of calculation results takes a long time, the pressure of data calculation on the cloud computing platform is greater, the load and resource overhead are higher, and network interruption and cloud downtime will affect the machine
  • the operation of the equipment causes losses to the enterprise; in addition, the data generated by the machine equipment is completely clouded, and the data security caused by the malicious attack of the cloud server is greater.
  • embodiments of the present invention provide an intelligent management method and system based on edge computing.
  • the technical solution is as follows:
  • an intelligent management method based on edge computing is provided.
  • the method is applied to an intranet management system.
  • the intranet management system includes at least a machine device and an edge computing software or device that communicates with the machine device. among them:
  • the edge computing software or equipment obtains the original data of the machine equipment
  • the edge computing software or device analyzes the original data based on locally stored mathematical models and strategies to generate data analysis results
  • the edge computing software or device feeds back the data analysis result to the preset result receiving device, so that the result receiving device implements intelligent management based on the data analysis result.
  • the method further includes:
  • the edge computing software or device performs data cleaning processing on the original data, and performs data aggregation and compression on the original data after data cleaning processing;
  • the edge computing software or device uploads the compressed raw data to the edge computing platform, so that the edge computing platform creates and trains mathematical models and strategies based on the operating data;
  • the edge computing software or device receives and stores the mathematical model and strategy fed back by the edge computing platform.
  • the internal network management system further includes an internal network edge node deployed in the internal computer room;
  • the method also includes:
  • the intranet edge node obtains the original data uploaded by the edge computing software or device;
  • the internal network edge node performs internal processing on the original data to generate the internal processing result
  • the internal network edge node feeds back the internal processing result to a preset result receiving device.
  • the preset result receiving device is a machine device, a data center device, or an operation management device.
  • the edge computing software or device feeds back the data analysis result to the preset result receiving device, including:
  • the edge computing software or device feeds back the data analysis result to the machine device through the OT system in the form of control instructions.
  • the edge computing software or device obtains the original data of the machine device, it further includes:
  • the edge computing software or device converts the original data under different protocols into the original data under the specified protocol based on preset protocol conversion rules.
  • the method further includes:
  • the edge computing software or device obtains the system operating data of the OT system and uploads the system operating data to the edge computing platform, so that the edge computing platform creates and trains the OT system decision model based on the system operating data;
  • the edge computing software or device receives the OT system decision model fed back by the edge computing platform, and intelligently manages the OT system based on the OT system decision model.
  • an intelligent management system based on edge computing includes at least machine equipment and edge computing software or equipment in communication with the machine equipment, wherein the edge computing software or equipment is used to in:
  • the data analysis result is fed back to the preset result receiving device, so that the result receiving device implements intelligent management based on the data analysis result.
  • the edge computing software or device is also used for:
  • the intelligent management system further includes an internal network edge node deployed in the internal computer room;
  • the internal network edge node is used to:
  • the preset result receiving device is a machine device, a data center device, or an operation management device.
  • the edge computing software or device feeds back the data analysis result to the preset result receiving device, including:
  • the edge computing software or device feeds back the data analysis result to the machine device through the OT system in the form of control instructions.
  • the edge computing software or device is also used for:
  • the original data under different protocols is converted into the original data under the specified protocol.
  • the edge computing software or device is also used for:
  • the edge computing software or device obtains the system operating data of the OT system and uploads the system operating data to the edge computing platform, so that the edge computing platform creates and trains the OT system decision model based on the system operating data;
  • the edge computing software or device receives the OT system decision model fed back by the edge computing platform, and intelligently manages the OT system based on the OT system decision model.
  • edge computing software or equipment near the bottom of the machine equipment can make data analysis processing closer to the data source, data can be processed faster, and the delay time of data analysis processing is shortened.
  • the amount of data that needs to be uploaded to the edge computing platform after being processed by the edge computing software or equipment is greatly reduced, saving bandwidth resources for transmitting data.
  • the use of edge computing software or equipment plus edge computing platform to jointly implement data processing can make full use of edge storage resources and computing resources, reducing the pressure of data computing and data storage on the edge computing platform.
  • the data that has security requirements is handled by the internal network, and the data does not need to be transmitted to the external network, which can improve the security of the data.
  • FIG. 1 is a schematic diagram of an intelligent management system based on edge computing provided by an embodiment of the present invention
  • FIG. 2 is a flowchart of an intelligent management method based on edge computing provided by an embodiment of the present invention
  • FIG. 3 is a schematic diagram of an intelligent management system based on edge computing provided by an embodiment of the present invention.
  • An embodiment of the present invention provides an intelligent management method based on edge computing.
  • the method is applied to an intelligent management system.
  • the intelligent management system may use edge computing technology to perform equipment, such as enterprises, factories, buildings, transportation, and medical treatment, that require networking.
  • the intelligent management system includes at least machine equipment and edge computing software or equipment that communicates with the machine equipment.
  • the machine equipment can be any machine equipment that needs to be connected to the cloud, such as industrial production machines, robots, smart equipment, motors, smart homes, meters, cameras, and sensors.
  • the machine equipment can generate or collect raw materials in the working state. Data, and send raw data to edge computing software or devices.
  • the edge computing software or device may be a software or device pre-deployed by the edge computing provider on the internal network and physically close to the machine device.
  • the edge computing software or device may generate or collect the machine device through the edge computing service on it Data collection, calculation, storage, transmission and application services.
  • edge computing software or devices can have AI vision, AR/VR simulation, deep learning, voiceprint recognition and other capabilities, supporting ultra-clear images and ultra-clear The local processing capability of the video.
  • a large number of edge computing software or devices can be deployed in the internal network, and each edge computing software or device can be connected to one or more machine devices.
  • Edge computing services can be mathematical models, strategies, data analysis results, etc. trained by edge computing platforms, or data preprocessing, calculation, storage, communication protocols, applications, and complete functional services of edge computing software or devices themselves.
  • Step 201 the edge computing software or device obtains the original data of the machine device.
  • the original data can be sensor instrument data such as temperature, pressure, vibration, humidity, electrical parameters, camera pictures, camera video and sound, production data, quality data, machine equipment operation data, process parameters, mac and ID addresses, etc.
  • the edge computing provider may deploy edge computing software or equipment on the machine equipment side in advance, and establish communication between the edge computing software or equipment and the underlying machine equipment.
  • the edge computing software or equipment and machine equipment can be connected by wired communication technology, such as Ethernet, Modbus, TCP/IP, RS232/RS485/RS482 serial port technology, etc., wireless communication technologies iBeacon, Bluetooth, ZigBee, etc. Connect.
  • wired communication technology such as Ethernet, Modbus, TCP/IP, RS232/RS485/RS482 serial port technology, etc.
  • wireless communication technologies iBeacon, Bluetooth, ZigBee, etc. Connect Taking the target enterprise as an example, when the machine and equipment of the target enterprise are running, pre-deployed edge computing software or equipment can continuously monitor the machine and equipment, and obtain the original data and information of the machine and equipment.
  • Step 202 the edge computing software or device analyzes the original data based on locally stored mathematical models and strategies, and generates data analysis results.
  • the edge computing software or device can call up locally stored mathematical models and strategies, analyze the original data, and generate data analysis results.
  • the edge computing software or device may run an edge computing framework and container for computing local data.
  • the computing function of the edge computing software or device may be provided by the edge computing framework and container.
  • the container can realize the specific processing of analyzing the original data. It is worth mentioning that a mathematical model and strategy can be used simultaneously on multiple edge computing software or devices, and multiple mathematical models can be stored on each edge computing software or device.
  • Step 203 The edge computing software or device feeds back the data analysis result to the preset result receiving device, so that the result receiving device implements internal intelligent management based on the data analysis result.
  • the edge computing software or device after the edge computing software or device generates the data analysis result, the data analysis result may be fed back to the preset result receiving device.
  • the result receiving device after the result receiving device obtains the data analysis result, it can implement intelligent management based on the data analysis result.
  • intelligent management can also be performed in scenarios such as management, development, engineering analysis, equipment research and development, equipment maintenance, and on-site diagnostic analysis, which can improve management efficiency, increase the speed of management decisions, and production capacity, improve Service quality reduces operating costs.
  • strategies such as intelligent decision support and intelligent scheduling and scheduling can improve the enterprise's energy utilization rate, reduce energy consumption, and use resources reasonably.
  • the edge computing software or device can also provide the original data to the edge computing platform, so that the edge computing platform can further optimize the mathematical model and strategy.
  • the corresponding processing can be as follows: the edge computing software or device performs data cleaning processing on the original data , And perform data aggregation and compression on the original data after data cleaning; the edge computing software or device uploads the aggregated and compressed original data to the edge computing platform, so that the edge computing platform creates and trains mathematics based on the original data of the machine equipment Models and strategies; edge computing software or devices receive and store mathematical models and strategies fed back by edge computing platforms.
  • the edge computing platform can be deployed on the external network, built by the edge computing server according to the region, and used to serve all enterprises, factories, buildings, transportation, and medical computing platforms in each region. It can aggregate a large number of edge computing software or equipment to upload The original data is processed in depth, such as modeling, storage, and calculation.
  • the edge computing platform can be an edge node in the CDN cluster.
  • edge computing software or device after the edge computing software or device obtains the original data of the machine equipment, it can first perform data cleaning processing on the original data, and the redundant and redundant original data (such as the timestamp repeated data within the period and the owner's repeated information) , MAC address duplication information, etc.) Screening and clearing, and complete the missing original data, at the same time, you can also correct or delete the wrong original data. Later, edge computing software or equipment can also aggregate the raw data after data cleaning through edge computing frameworks, containers, applications, etc., and interact with local data centers to support machines and devices to go to the cloud with only necessary data and information. Data can be used internally to protect data security.
  • the edge computing software or device can compress the aggregated original data through the edge computing framework, container, and application, and store the compressed original data locally. In this way, the edge computing software or device can upload the processed raw data to the edge computing platform without reporting all the original raw data to the edge computing platform, thereby reducing traffic consumption during data transmission.
  • the edge computing platform may first store the original data in local memory or a preset database. After that, the edge computing platform performs big data analysis on the stored raw data, so that machine-learned mathematical models can be created through machine learning, and the created mathematical models and strategies can be continuously trained based on the original data. Furthermore, the edge computing platform can feed back the trained mathematical model and strategy to the edge computing software or device. After receiving the mathematical model and strategy, the edge computing software or device can store the mathematical model and strategy to update the locally running mathematical model , Strategies, so as to achieve continuous optimization of mathematical models and strategies.
  • the edge computing software or device and the edge computing platform can communicate through cloud communication technology, and the cloud communication technology can communicate based on a LAN line or an Internet line.
  • edge computing software or equipment can be mainly responsible for edge computing that is close to the data source of the machine equipment, responding to data analysis and processing with high real-time requirements, and analyzing and processing some complex calculations that require large amounts of data (such as The above model training), the edge computing software or device can upload the necessary data for processing to the edge computing platform, which is implemented by the edge computing platform.
  • the intelligent management system further includes an internal network edge node deployed in the internal computer room, and the edge computing software or device may send the acquired raw data to the internal network edge node to pass the internal network edge node Run edge computing frameworks, containers and applications, or private clouds to implement the calculation, storage, network, and applications of data reported by all internally connected machine devices to achieve high-performance, high-concurrency, and large-flow data and information operations.
  • step 201 The following processes may exist: the internal network edge node obtains the original data uploaded by the edge computing software or device; the internal network edge node performs internal processing on the original data to generate an internal processing result; the internal network edge node feeds back the internal processing result to the preset The receiving device.
  • the edge computing service provider may pre-deploy an internal network edge node in the internal computer room.
  • the internal network edge node may be dedicated to providing internal data processing services.
  • an internal local autonomous computing network may be formed.
  • the edge node of the internal network is regarded as the core of the autonomous computing network, and the edge node of the internal network is responsible for the interconnection and interworking between the devices of the internal network. Therefore, after the edge computing software or device obtains the original data, it can upload the original data to the internal network edge node.
  • the internal network edge node can obtain the original data, and process the original data according to preset data processing rules (may be referred to as internal processing) to generate an internal processing result.
  • the internal network edge node can feed back the internal processing result to the preset result receiving device.
  • the internal device intelligent management can be realized only through the intranet.
  • data processing can be implemented on the internal network without transmission through the external network, which effectively avoids data leakage during the transmission of the external network and improves data security. It is worth mentioning that after the internal network edge node internally processes the original data, it can continue to upload the original data to the edge computing platform, so that the edge computing platform can perform subsequent processing based on the original data.
  • the edge computing software or device and the internal network edge node can communicate through the Internet of Things, which can include 3G/4G/5G network, WiFi, LoRa network or NB-IoT network, etc.; the internal network edge node and edge computing platform Communication can be done through HTTP/HTTPS networks, which can include wired and wireless networks.
  • the data analysis results can be provided to different internal devices for multi-directional intelligent management.
  • the preset result receiving devices can be machine devices, data center devices, and operation management devices.
  • different result receiving devices may be preset for different data analysis results. After receiving the corresponding data analysis results, the result receiving device may implement intelligent management in different scenarios.
  • the result receiving device may be a machine device. After receiving the data analysis result, the machine device may be automatically arranged as an operation instruction according to the data analysis result to realize the independent intelligent optimization of the machine device; of course, the result receiving device may also be a data center device After receiving the data analysis results, the data center equipment can store, manage, and update the local data based on the data analysis results, and realize the independent intelligent optimization of the data center equipment.
  • the result receiving equipment can also be an operation management equipment (such as Mobile phone terminal), after receiving the data analysis result, the operation management device can perform intelligent analysis and decision support on the internal operation status based on the data analysis result, and realize the independent intelligent optimization of the operation.
  • the edge computing software or device can support data conversion of multiple protocols.
  • the following processing may exist after step 201: the edge computing software or device converts the original data under different protocols based on preset protocol conversion rules The original data under the specified protocol.
  • the edge computing software or device can be pre-configured with data analysis functions under multiple protocols and protocol conversion rules for different data. In this way, after the edge computing software or device obtains the original data of the machine, it can be based on the corresponding The data parsing function under the protocol parses the above original data, and then converts the original data under different protocols into the original data under the specified protocol through protocol conversion rules. In this way, through the processing of edge computing software or equipment, data communication can be achieved between new and old machinery and equipment, so that traditional industrial equipment can be seamlessly and efficiently connected to modern edge computing platforms. It can be understood that if the data processing result needs to be fed back to the machine and device subsequently, the edge computing software or device may also perform protocol conversion on the data processing result and convert the data processing result into content under the protocol supported by the machine device.
  • the edge computing platform can feed back the data processing results to the machine equipment through the OT (Operation Technology) layer.
  • the processing in step 203 can be as follows: the edge computing software or device can pass the OT system in the form of control instructions Feedback the deep processing results to the machine equipment.
  • the edge computing platform may send the deep processing result to the edge computing software or device.
  • the edge computing software or device can feed back the deep processing result to the underlying machine equipment through the OT system in the form of control instructions, so that the machine equipment can dynamically adjust the operation logic based on the deep processing result and generate optimization After the original data.
  • the data of the OT system can be processed by the edge computing platform to achieve intelligent management of the OT system.
  • the corresponding processing can be as follows: the edge computing software or device obtains the system operating data of the OT system and uploads the system operating data To the edge computing platform, so that the edge computing platform creates and trains the OT system decision model based on the system operating data; the edge computing software or device receives the OT system decision model fed back by the edge computing platform, and intelligently manages the OT system based on the OT system decision model
  • the edge computing software or device can obtain the system operation data of the OT system while acquiring the original data of the machine equipment, and then upload the system operation data to the edge computing platform.
  • the edge computing platform can store and organize the system operation data after receiving the system operation data.
  • the edge computing platform can build the OT system decision model based on the system operation data, and continuously train the OT system decision model through the system operation data.
  • the edge computing platform can feed back the trained OT system decision model to the edge computing software or device, so that the edge computing software or device can apply the OT system decision model to the OT system to achieve intelligent management of the OT system.
  • the edge computing local computing software obtains the original data of the machine equipment; the edge computing local computing software is based on the edge computing container, framework, and model running on the side of the local machine equipment, and the industry application analysis, learning, and use of the machine given by it
  • the original data generated by the equipment generates mathematical models and data analysis results; the edge computing local computing software runs the preset mathematical models and strategies on the one hand to send and receive, feed back the raw data analysis results generated by the machine, on the other hand, the mathematical models and strategies
  • the related data information or the data information required for the newly established mathematical model interacts with the edge computing platform; and the result receiving device can realize intelligent management based on the data analysis results.
  • Deploying edge computing software or equipment near the bottom of the machine equipment can make the data analysis processing closer to the data source, the data can be processed faster, and the delay time for data analysis processing is shortened, and there is no need to access the external network.
  • the internal network can build a complete intelligent management system.
  • edge computing software or equipment after being processed by edge computing software or equipment, the amount of data that needs to be uploaded to the edge computing platform is greatly reduced, saving bandwidth resources for transmitting data.
  • the use of edge computing software or equipment plus edge computing platform to jointly implement data processing can make full use of edge storage resources and computing resources, reducing the pressure of data computing and data storage on the edge computing platform.
  • the data that has security requirements is handled by the internal network, and the data does not need to be transmitted to the external network, which can improve the security of the data.
  • an embodiment of the present invention also provides an intelligent management system based on edge computing.
  • the intelligent management system includes at least a machine device and an edge computing software or device connected to the machine device.
  • Edge computing software or equipment for:
  • the data result is fed back to the preset result receiving device, so that the result receiving device implements intelligent management based on the data result.
  • the edge computing software or device is also used for:
  • the intelligent management system further includes an internal network edge node deployed in the internal computer room;
  • the internal network edge node is used to:
  • the preset result receiving device is a machine device, a data center device, or an operation management device.
  • the edge computing software or device feeds back the data analysis result to the preset result receiving device, including:
  • the edge computing software or device feeds back the data analysis result to the machine device through the OT system in the form of control instructions.
  • the edge computing software or device is also used for:
  • the original data under different protocols is converted into the original data under the specified protocol.
  • the edge computing software or device is also used for:
  • the edge computing software or device obtains the system operating data of the OT system, and uploads the system operating data to the edge computing platform, so that the edge computing platform creates and trains the OT system decision model based on the system operating data;
  • the edge computing software or device receives the OT system decision model fed back by the edge computing platform, and intelligently manages the OT system based on the OT system decision model.
  • edge computing software or equipment near the bottom of the machine equipment can make the data analysis processing closer to the data source, the data can be processed faster, and the delay time of the data analysis processing is shortened.
  • the amount of data that needs to be uploaded to the edge computing platform after being processed by the edge computing software or equipment is greatly reduced, saving bandwidth resources for transmitting data.
  • the use of edge computing software or equipment plus an edge computing platform to jointly implement data processing can make full use of edge storage resources and computing resources, reducing the pressure of data computing and data storage on the edge computing platform.
  • handing over data that has security requirements to internal processing, data does not need to be transmitted to the external network can improve data security.
  • the program may be stored in a computer-readable storage medium.
  • the mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Medical Informatics (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Theoretical Computer Science (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Computing Systems (AREA)
  • Algebra (AREA)
  • Mathematical Analysis (AREA)
  • Computer Security & Cryptography (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Databases & Information Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Probability & Statistics with Applications (AREA)
  • Computer And Data Communications (AREA)
  • Stored Programmes (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种基于边缘计算的智能管理方法和系统,属于边缘计算技术领域。该方法可以缩短数据分析处理的时间而更加低延时,并在网络中断、云宕机时保证边缘计算在内部网络既可离线运行本地计算软件、实现本地自治,又能通过离线运行的本地边缘计算组实现机器设备间的互联互通而搭建完整的智能管理网络和系统,同时可以在联网时节省传输数据的带宽资源,有效缓解底层的机器设备上云时的数据计算和存储压力、流量过高以及应用开发难且智能不足的问题。

Description

一种基于边缘计算的智能管理方法和系统 技术领域
本发明涉及边缘计算技术领域,特别涉及一种基于边缘计算的智能管理方法和系统。
背景技术
随着云计算技术的不断发展,越来越多的企业开始将云计算技术应用到实际的生产经营和管理过程中。企业可以将机器设备接入网络,并通过云计算平台对机器设备产生的数据进行中心计算,从而通过计算结果实现企业更高效的决策管理。
具体来说,企业的机器设备在运行时,可以周期性地将本地生成的原始数据,通过物联网技术上传至云计算平台。之后,云计算平台可以对接收到的原始数据进行汇总存储,并进行集中的分析、计算。继而,云计算平台可以将得到的计算结果反馈给企业的机器设备和移动管理终端,使得机器设备按照计算结果自动运行,或者人工通过移动管理终端按照计算结果对企业内部进行管理。
在实现本发明的过程中,发明人发现现有技术至少存在以下问题:
企业的规模扩大,物联网机器设备的数量快速增加,相应的,企业的机器设备上云需要传输的数据量也剧增,而继续采用原有的云计算平台进行计算管理,将会消耗大量的带宽资源以用来传输机器设备产生的数据,且计算结果的反馈耗时较长、云计算平台上数据计算的压力较大、负载和资源开销较高,同时网络中断和云宕机将影响机器设备的运行而对企业造成损失;另外机器设备产生的数据完全上云,当云服务器被恶意攻击时造成的数据安全危害较大。
发明内容
为了解决现有技术的问题,本发明实施例提供了一种基于边缘计算的智能管理方法和系统。所述技术方案如下:
第一方面,提供了一种基于边缘计算的智能管理方法,所述方法应用于内 网管理系统,所述内网管理系统至少包括机器设备和与所述机器设备通讯的边缘计算软件或设备,其中:
所述边缘计算软件或设备获取机器设备的原始数据;
所述边缘计算软件或设备基于本地存储的数理模型、策略,分析所述原始数据,生成数据分析结果;
所述边缘计算软件或设备向预设的结果接收设备反馈所述数据分析结果,以使所述结果接收设备基于所述数据分析结果实现智能管理。
可选的,所述方法还包括:
所述边缘计算软件或设备对所述原始数据进行数据清洗处理,并对数据清洗处理后的原始数据进行数据聚合和压缩;
所述边缘计算软件或设备将压缩后的原始数据上传至边缘计算平台,以使所述边缘计算平台基于所述运行数据创建并训练数理模型、策略;
所述边缘计算软件或设备接收并存储所述边缘计算平台反馈的数理模型、策略。
可选的,所述内网管理系统还包括部署在内部机房的内网边缘节点;
所述方法还包括:
所述内网边缘节点获取所述边缘计算软件或设备上传的所述原始数据;
所述内网边缘节点对所述原始数据进行内部处理,生成所述内部处理结果;
所述内网边缘节点将所述内部处理结果反馈至预设的结果接收设备。
可选的,所述预设的结果接收设备为机器设备、数据中心设备或者运营管理设备。
可选的,所述边缘计算软件或设备向预设的结果接收设备反馈所述数据分析结果,包括:
所述边缘计算软件或设备以控制指令的方式,通过OT系统将所述数据分析结果反馈至所述机器设备。
可选的,所述边缘计算软件或设备获取机器设备的原始数据之后,还包括:
所述边缘计算软件或设备基于预设的协议转化规则,将不同协议下的原始数据转化为指定协议下的原始数据。
可选的,所述方法还包括:
所述边缘计算软件或设备获取OT系统的系统运行数据,将所述系统运行数 据上传至边缘计算平台,以使所述边缘计算平台基于所述系统运行数据创建并训练OT系统决策模型;
所述边缘计算软件或设备接收所述边缘计算平台反馈的OT系统决策模型,并基于所述OT系统决策模型对所述OT系统进行智能管理。
第二方面,提供了一种基于边缘计算的智能管理系统,所述智能管理系统至少包括机器设备和与所述机器设备通讯的边缘计算软件或设备,其中,所述边缘计算软件或设备,用于:
获取所述机器设备的原始数据;
基于本地存储的数理模型、策略,分析所述原始数据,生成数据分析结果;
向预设的结果接收设备反馈所述数据分析结果,以使所述结果接收设备基于所述数据分析结果实现智能管理。
可选的,所述边缘计算软件或设备,还用于:
对所述原始数据进行数据清洗处理,并对数据清洗处理后的原始数据进行数据聚合和压缩;
将压缩后的原始数据上传至边缘计算平台,以使所述边缘计算平台基于所述运行数据创建并训练数理模型、策略;
接收并存储所述边缘计算平台反馈的数理模型、策略。
可选的,所述智能管理系统还包括部署在内部机房的内网边缘节点;
所述内网边缘节点,用于:
获取所述边缘计算软件或设备上传的所述原始数据;
对所述原始数据进行内部处理,生成所述内部处理结果;
将所述内部处理结果反馈至预设的结果接收设备。
可选的,所述预设的结果接收设备为机器设备、数据中心设备或者运营管理设备。
可选的,所述边缘计算软件或设备向预设的结果接收设备反馈所述数据分析结果,包括:
所述边缘计算软件或设备以控制指令的方式,通过OT系统将所述数据分析结果反馈至所述机器设备。
可选的,所述边缘计算软件或设备,还用于:
基于预设的协议转化规则,将不同协议下的原始数据转化为指定协议下的原始数据。
可选的,所述边缘计算软件或设备,还用于:
所述边缘计算软件或设备获取OT系统的系统运行数据,将所述系统运行数据上传至边缘计算平台,以使所述边缘计算平台基于所述系统运行数据创建并训练OT系统决策模型;
所述边缘计算软件或设备接收所述边缘计算平台反馈的OT系统决策模型,并基于所述OT系统决策模型对所述OT系统进行智能管理。
本发明实施例提供的技术方案带来的有益效果是:
本发明实施例中,在底层的机器设备处就近部署边缘计算软件或设备,一方面可以使得数据分析处理更接近数据来源,数据能够得到更快的处理,缩短了数据分析处理的延迟时间,并无需接入外网在内部网络既可搭建完整的智能管理系统,另一方面经过边缘计算软件或设备处理,需要上传至边缘计算平台的数据量大幅降低,节省了用于传输数据的带宽资源。同时,采用边缘计算软件或设备加边缘计算平台共同实现数据处理,可以充分利用边缘的存储资源和计算资源,降低了边缘计算平台上的数据计算压力和数据存储压力。此外,将存在安全需求的数据交由内部网络处理,数据无需传输到外网,可以提高数据的安全性。
附图说明
为了更清楚地说明本发明实施例中的技术方案,下面将对实施例描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的一种基于边缘计算的智能管理系统的框架示意图;
图2是本发明实施例提供的一种基于边缘计算的智能管理方法流程图;
图3是本发明实施例提供的一种基于边缘计算的智能管理系统的框架示意图。
具体实施方式
为使本发明的目的、技术方案和优点更加清楚,下面将结合附图对本发明实施方式作进一步地详细描述。
本发明实施例提供了一种基于边缘计算的智能管理方法,该方法应用于智能管理系统,智能管理系统可以是利用边缘计算技术对企业、工厂、楼宇、交通、医疗等需要联网的机器设备进行智能管理的系统。如图1所示,智能管理系统至少包括机器设备和与机器设备通讯的边缘计算软件或设备。其中,机器设备可以是任何需要联网上云的机器设备,如可以是工业生产机器、机器人、智能装备、电机、智能家居、仪表、摄像头和传感器等,机器设备在工作状态下可以产生或采集原始数据,并将原始数据发送给边缘计算软件或设备。边缘计算软件或设备可以是由边缘计算提供方预先部署在内部网络,物理位置上接近机器设备的软件或设备,边缘计算软件或设备可以通过其上的边缘计算服务,实现对机器设备产生或采集的数据进行汇总、计算、存储、传输和应用等服务,具体的,边缘计算软件或设备可以具备AI视觉、AR/VR仿真、深度学习、声纹识别等能力,支持对超清图像和超清视频的本地处理能力。内部网络中可以部署有大量边缘计算软件或设备,每个边缘计算软件或设备可以与一台或多台机器设备相连接。边缘计算服务可以是边缘计算平台训练的数理模型、策略、数据分析结果等,还可以是边缘计算软件或设备自身的数据预处理、计算、存储、通讯协议、应用、完全等功能服务。
下面将结合具体实施方式,对图2所示的处理流程进行详细的说明,内容可以如下:
步骤201,边缘计算软件或设备获取机器设备的原始数据。
其中,原始数据可以是温度、压力、振动、湿度、电参数等传感器仪表数据,照相机图片、摄像机视频和声音、生产数据、品质数据、机器设备运行数据、工艺参数、mac和ID地址等。
在实施中,为了实现对底层的机器设备进行联网并智能化,边缘计算提供方可以预先在机器设备侧部署边缘计算软件或设备,并将边缘计算软件或设备与底层的机器设备建立通讯。具体的,边缘计算软件或设备与机器设备可以通过有线通信技术实现连接,如可以通过Ethernet、Modbus、TCP/IP、 RS232/RS485/RS482串口技术等实现连接,无线通讯技术iBeacon、蓝牙、ZigBee等实现连接。以目标企业为例,在目标企业的机器设备运行时,预先部署的边缘计算软件或设备可以对机器设备进行持续监测,并获取机器设备的原始数据和信息。
步骤202,边缘计算软件或设备基于本地存储的数理模型、策略,分析原始数据,生成数据分析结果。
在实施中,边缘计算软件或设备在获取到机器设备的原始数据之后,可以调取本地存储的数理模型、策略,对原始数据进行分析,生成数据分析结果。具体的,边缘计算软件或设备上可以运行有用于对实现本地数据计算的边缘计算框架和容器,边缘计算软件或设备的计算功能均可以由边缘计算框架和容器提供,这样,通过边缘计算框架和容器即可以实现对原始数据进行分析的具体处理。值得一提的是,一个数理模型、策略可以在多个边缘计算软件或设备上同时使用,每个边缘计算软件或设备上可以存储有多个数理模型。
步骤203,边缘计算软件或设备向预设的结果接收设备反馈数据分析结果,以使结果接收设备基于数据分析结果实现内部智能管理。
在实施中,边缘计算软件或设备生成数据分析结果之后,可以将数据分析结果反馈给预设的结果接收设备。这样,结果接收设备在获取到数据分析结果之后,可以基于数据分析结果实现对智能管理。当然,基于不同的数据分析结果,还可以在管理、开发、工程分析、设备研发、设备维护、现场诊断分析等场景进行智能管理,从而可以提升管理效率,提高管理决策的速度以及生产产能,改善服务品质,降低经营成本。此外,对于企业来说,基于边缘计算软件和设备生成的数据分析结果,通过智能决策支持、智能调度排产等策略可以提升企业的能源利用率,降低能源消耗,合理利用资源。
可选的,边缘计算软件或设备还可以将原始数据提供给边缘计算平台,以使边缘计算平台进一步优化数理模型、策略,相应的处理可以如下:边缘计算软件或设备对原始数据进行数据清洗处理,并对数据清洗处理后的原始数据进行数据聚合和压缩;边缘计算软件或设备将聚合和压缩后的原始数据上传至边缘计算平台,以使边缘计算平台基于机器设备的原始数据创建并训练数理模型、策略;边缘计算软件或设备接收并存储边缘计算平台反馈的数理模型、策略。
其中,边缘计算平台可以是部署在外网,由边缘计算服务器按照区域搭建 的,用于服务各区域内所有企业、工厂、楼宇、交通、医疗的计算平台,其可以汇总大量边缘计算软件或设备上传的原始数据,并进行建模、存储、计算等深度处理,具体来说,边缘计算平台可以是CDN集群中的边缘节点。
在实施中,边缘计算软件或设备在获取到机器设备的原始数据后,可以先对原始数据进行数据清洗处理,将重复、多余的原始数据(如周期内的时间戳重复数据、归属者重复信息、MAC地址重复信息等)筛选清除,并将缺失的原始数据补充完整,同时还可以将错误的原始数据纠正或者删除。之后,边缘计算软件或设备还可以通过边缘计算框架、容器、应用等对数据清洗处理后的原始数据进行聚合,并与本地数据中心进行交互,支持机器设备只以必要数据和信息上云,大量数据得以在本地内部使用,以保护数据安全。进一步的,边缘计算软件或设备可以通过边缘计算框架、容器和应用对聚合后的原始数据进行压缩,并将压缩后的原始数据存储在本地。这样,边缘计算软件或设备可以经上述处理后的原始数据上传到边缘计算平台,而无需将全部原始的原始数据上报至边缘计算平台,从而减少了数据传输时的流量消耗。
边缘计算平台在获取到所有边缘计算软件或设备上传的原始数据之后,可以先将原始数据存储在本地内存或预设数据库中。之后,边缘计算平台对已存储的原始数据进行大数据分析,从而可以通过机器学习的方式创建设备智能的数理模型,并进一步基于原始数据对创建的数理模型、策略进行持续训练。进而,边缘计算平台可以将训练得到的数理模型、策略反馈给边缘计算软件或设备,边缘计算软件或设备在接收数理模型、策略后,可以存储该数理模型、策略,以更新本地运行的数理模型、策略,从而实现数理模型、策略的不断优化。
此处,边缘计算软件或设备与边缘计算平台可以通过云通讯技术进行通信,云通讯技术可以是基于局域网线路或者互联网线路进行通信。
值得一提的是,边缘计算软件或设备可以主要负责贴近机器设备数据源的边缘计算,响应实时性需求较高的数据分析处理,而对于一些较为运算复杂的、需要大量数据的分析处理(如上述模型训练),边缘计算软件或设备则可以将处理所需的必要数据上传给边缘计算平台,由边缘计算平台来实现。
可选的,如图3所示,智能管理系统还包括部署在内部机房的内网边缘节点,边缘计算软件或设备可以将获取到的原始数据发送给内网边缘节点,以通过内网边缘节点运行边缘计算框架、容器和应用或私有云来实现内部全部联网 机器设备上报数据的计算、存储、网络和应用,来实现高性能、高并发、大流量的数据和信息运算,相应的,步骤201后可以存在如下处理:内网边缘节点获取边缘计算软件或设备上传的原始数据;内网边缘节点对原始数据进行内部处理,生成内部处理结果;内网边缘节点将内部处理结果反馈至预设的结果接收设备。
在实施中,边缘计算服务方可以预先在内部机房中部署内网边缘节点,该内网边缘节点可以专用于提供内部的数据处理服务,通过该内网边缘节点可以组建内部的本地自治计算网络,将内网边缘节点作为自治计算网络的核心,并由内网边缘节点负责内网各设备间的互联互通。故而,边缘计算软件或设备获取到原始数据后,可以将原始数据上传至内网边缘节点。内网边缘节点可以获取该原始数据,并按照预设的数据处理规则对原始数据进行处理(可称为内部处理),生成内部处理结果。之后,内网边缘节点可以将内部处理结果反馈至预设的结果接收设备。这样,仅通过内网也可实现内部的设备智能管理。同时,对于一些安全需求较高的数据,可以在内网即实现数据处理,无需经过外网传输,有效避免了外网传输过程中的数据泄露,提高了数据的安全性。值得一提的是,内网边缘节点在对原始数据进行内部处理后,可以继续将原始数据上传至边缘计算平台,以使边缘计算平台基于原始数据进行后续处理。此处,边缘计算软件或设备与内网边缘节点可以通过物联网进行通信,物联网可以包括3G/4G/5G网络、WiFi、LoRa网络或NB-IoT网络等;内网边缘节点与边缘计算平台可以通过HTTP/HTTPS网络进行通信,HTTP/HTTPS网络可以包括有线网络和无线网络。
可选的,数据分析结果可以提供给内部的不同设备来多方位的实现智能管理,相应的,上述预设的结果接收设备可以是机器设备、数据中心设备、运营管理设备。
在实施中,可以针对不同的数据分析结果预先设定不同的结果接收设备,结果接收设备接收到相应的数据分析结果后,可以实现不同场景下的智能管理。具体的,结果接收设备可以是机器设备,机器设备接收到数据分析结果之后,可以按照数据分析结果自动编排为操作指令,实现机器设备的自主智能优化;当然,结果接收设备也可以是数据中心设备,数据中心设备接收到数据分析结果之后,可以基于数据分析结果对本地数据进行存储和数据治理、更新,实现 了数据中心设备的自主智能优化;此外,结果接收设备还可以是运营管理设备(如手机移动终端),运营管理设备接收到数据分析结果之后,可以基于数据分析结果对内部运营状况进行智能的分析决策支持,实现了运营的自主智能优化。
可选的,边缘计算软件或设备可以支持多种协议的数据转换,相应的,步骤201之后可以存在以下处理:边缘计算软件或设备基于预设的协议转化规则,将不同协议下的原始数据转化为指定协议下的原始数据。
在实施中,边缘计算软件或设备中可以预先配置有多种协议下的数据解析功能和不同数据的协议转换规则,这样,边缘计算软件或设备在获取到机器设备的原始数据之后,可以基于相应协议下的数据解析功能解析上述原始数据,然后通过协议转换规则将不同协议下的原始数据转化为指定协议下的原始数据。这样,通过边缘计算软件或设备的处理,可以让新型机器设备和旧式机器设备之间实现数据互通,从而使得传统工业设备可以无缝且高效地连接到现代的边缘计算平台。可以理解,如果后续需要向机器设备反馈数据处理结果,边缘计算软件或设备也可以对数据处理结果进行协议转化,将数据处理结果转换为机器设备支持的协议下的内容。
可选的,边缘计算平台可以通过OT(Operation Technology,操作技术)层向机器设备反馈数据处理结果,相应的,步骤203的处理可以如下:边缘计算软件或设备以控制指令的方式,通过OT系统将深度处理结果反馈至机器设备。
在实施中,边缘计算平台在基于原始数据生成深度处理结果后,可以向边缘计算软件或设备发送该深度处理结果。边缘计算软件或设备接收到该深度处理结果之后,可以以控制指令的方式,通过OT系统将深度处理结果反馈至底层的机器设备,从而机器设备可以基于深度处理结果动态调整运行逻辑,并生成优化后的原始数据。
可选的,OT系统的数据可以交由边缘计算平台进行处理,以实现对OT系统的智能管理,相应的处理可以如下:边缘计算软件或设备获取OT系统的系统运行数据,将系统运行数据上传至边缘计算平台,以使边缘计算平台基于系统运行数据创建并训练OT系统决策模型;边缘计算软件或设备接收边缘计算平台反馈的OT系统决策模型,并基于OT系统决策模型对OT系统进行智能管理
在实施中,边缘计算软件或设备在获取机器设备的原始数据的同时,还可以获取OT系统的系统运行数据,然后将该系统运行数据上传至边缘计算平台。 这样,边缘计算平台可以在接收到系统运行数据之后,可以对系统运行数据进行存储和整理。之后,边缘计算平台可以基于系统运行数据搭建OT系统决策模型,并通过系统运行数据对OT系统决策模型进行不断训练。进而,边缘计算平台可以向边缘计算软件或设备反馈训练得到的OT系统决策模型,以使边缘计算软件或设备将OT系统决策模型应用到OT系统内,以实现OT系统的智能管理。
本发明实施例中,边缘计算本地计算软件获取机器设备的原始数据;边缘计算本地计算软件基于本地机器设备侧运行的边缘计算容器、框架以及模型,通过其赋予的行业应用分析、学习、使用机器设备产生的原始数据,生成数理模型和数据分析结果;边缘计算本地计算软件通过运行预设的数理模型和策略一方面收发、反馈机器设备产生的原始数据分析结果,另一方面将数理模型和策略相关的数据信息或新建数理模型所需的数据信息与边缘计算平台进行交互;并以使结果接收设备基于数据分析结果实现智能管理。在底层的机器设备处就近部署边缘计算软件或设备,一方面可以使得数据分析处理更接近数据来源,数据能够得到更快的处理,缩短了数据分析处理的延迟时间,并无需接入外网在内部网络既可搭建完整的智能管理系统,另一方面经过边缘计算软件或设备处理,需要上传至边缘计算平台的数据量大幅降低,节省了用于传输数据的带宽资源。同时,采用边缘计算软件或设备加边缘计算平台共同实现数据处理,可以充分利用边缘的存储资源和计算资源,降低了边缘计算平台上的数据计算压力和数据存储压力。此外,将存在安全需求的数据交由内部网络处理,数据无需传输到外网,可以提高数据的安全性。
基于相同的技术构思,本发明实施例还提供了一种基于边缘计算的智能管理系统,所述智能管理系统至少包括机器设备和与所述机器设备相连的边缘计算软件或设备,其中,所述边缘计算软件或设备,用于:
获取所述机器设备的原始数据;
基于本地运行的数理模型、策略和应用,分析所述原始数据,生成数据结果;
向预设的结果接收设备反馈所述数据结果,以使所述结果接收设备基于所述数据结果实现智能管理。
可选的,所述边缘计算软件或设备,还用于:
对所述原始数据进行数据清洗处理,并对数据清洗处理后的原始数据进行数据聚合和压缩;
将压缩后的原始数据上传至边缘计算平台,以使所述边缘计算平台基于所述运行数据创建并训练数理模型、策略和应用;
接收并存储所述边缘计算平台反馈的数理模型、策略。
可选的,所述智能管理系统还包括部署在内部机房的内网边缘节点;
所述内网边缘节点,用于:
获取所述边缘计算软件或设备上传的所述原始数据;
对所述原始数据进行内部处理,生成所述内部处理结果;
将所述内部处理结果反馈至预设的结果接收设备。
可选的,所述预设的结果接收设备为机器设备、数据中心设备或者运营管理设备。
可选的,所述边缘计算软件或设备向预设的结果接收设备反馈所述数据分析结果,包括:
所述边缘计算软件或设备以控制指令的方式,通过OT系统将所述数据分析结果反馈至所述机器设备。
可选的,所述边缘计算软件或设备,还用于:
基于预设的协议转化规则,将不同协议下的原始数据转化为指定协议下的原始数据。
可选的,所述边缘计算软件或设备,还用于:
所述边缘计算软件或设备获取OT系统的系统运行数据,将所述系统运行数据上传至边缘计算平台,以使所述边缘计算平台基于所述系统运行数据创建并训练OT系统决策模型;
所述边缘计算软件或设备接收所述边缘计算平台反馈的OT系统决策模型,并基于所述OT系统决策模型对所述OT系统进行智能管理。
本发明实施例中,在底层的机器设备处就近部署边缘计算软件或设备,一方面可以使得数据分析处理更接近数据来源,数据能够得到更快的处理,缩短了数据分析处理的延迟时间,并无需接入外网在内部网络既可搭建完整的智能管理系统,另一方面经过边缘计算软件或设备处理,需要上传至边缘计算平台的数据量大幅降低,节省了用于传输数据的带宽资源。同时,采用边缘计算软 件或设备加边缘计算平台共同实现数据处理,可以充分利用边缘的存储资源和计算资源,降低了边缘计算平台上的数据计算压力和数据存储压力。此外,将存在安全需求的数据交由内部处理,数据无需传输到外网,可以提高数据的安全性。
本领域普通技术人员可以理解实现上述实施例的全部或部分步骤可以通过硬件来完成,也可以通过程序来指令相关的硬件完成,所述的程序可以存储于一种计算机可读存储介质中,上述提到的存储介质可以是只读存储器,磁盘或光盘等。
以上所述仅为本发明的较佳实施例,并不用以限制本发明,凡在本发明的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本发明的保护范围之内。

Claims (14)

  1. 一种基于边缘计算的智能管理方法,其特征在于,所述方法应用于内网管理系统,所述内网管理系统至少包括机器设备和与所述机器设备通讯的边缘计算软件或设备,其中:
    所述边缘计算软件或设备获取机器设备的原始数据;
    所述边缘计算软件或设备基于本地存储的数理模型、策略,分析所述原始数据,生成数据分析结果;
    所述边缘计算软件或设备向预设的结果接收设备反馈所述数据分析结果,以使所述结果接收设备基于所述数据分析结果实现智能管理。
  2. 根据权利要求1所述的方法,其特征在于,所述方法还包括:
    所述边缘计算软件或设备对所述原始数据进行数据清洗处理,并对数据清洗处理后的原始数据进行数据聚合和压缩;
    所述边缘计算软件或设备将压缩后的原始数据上传至边缘计算平台,以使所述边缘计算平台基于所述运行数据创建并训练数理模型、策略;
    所述边缘计算软件或设备接收并存储所述边缘计算平台反馈的数理模型、策略。
  3. 根据权利要求1所述的方法,其特征在于,所述内网管理系统还包括部署在内部机房的内网边缘节点;
    所述方法还包括:
    所述内网边缘节点获取所述边缘计算软件或设备上传的所述原始数据;
    所述内网边缘节点对所述原始数据进行内部处理,生成所述内部处理结果;
    所述内网边缘节点将所述内部处理结果反馈至预设的结果接收设备。
  4. 根据权利要求1-3任一项所述的方法,其特征在于,所述预设的结果接收设备为机器设备、数据中心设备或者运营管理设备。
  5. 根据权利要求4所述的方法,其特征在于,所述边缘计算软件或设备向 预设的结果接收设备反馈所述数据分析结果,包括:
    所述边缘计算软件或设备以控制指令的方式,通过OT系统将所述数据分析结果反馈至所述机器设备。
  6. 根据权利要求1所述的方法,其特征在于,所述边缘计算软件或设备获取机器设备的原始数据之后,还包括:
    所述边缘计算软件或设备基于预设的协议转化规则,将不同协议下的原始数据转化为指定协议下的原始数据。
  7. 根据权利要求1所述的方法,其特征在于,所述方法还包括:
    所述边缘计算软件或设备获取OT系统的系统运行数据,将所述系统运行数据上传至边缘计算平台,以使所述边缘计算平台基于所述系统运行数据创建并训练OT系统决策模型;
    所述边缘计算软件或设备接收所述边缘计算平台反馈的OT系统决策模型,并基于所述OT系统决策模型对所述OT系统进行智能管理。
  8. 一种基于边缘计算的智能管理系统,其特征在于,所述智能管理系统至少包括机器设备和与所述机器设备通讯的边缘计算软件或设备,其中,所述边缘计算软件或设备,用于:
    获取所述机器设备的原始数据;
    基于本地存储的数理模型、策略,分析所述原始数据,生成数据分析结果;
    向预设的结果接收设备反馈所述数据分析结果,以使所述结果接收设备基于所述数据分析结果实现智能管理。
  9. 根据权利要求8所述的系统,其特征在于,所述边缘计算软件或设备,还用于:
    对所述原始数据进行数据清洗处理,并对数据清洗处理后的原始数据进行数据聚合和压缩;
    将压缩后的原始数据上传至边缘计算平台,以使所述边缘计算平台基于所述运行数据创建并训练数理模型、策略;
    接收并存储所述边缘计算平台反馈的数理模型、策略。
  10. 根据权利要求8所述的系统,其特征在于,所述智能管理系统还包括部署在内部机房的内网边缘节点;
    所述内网边缘节点,用于:
    获取所述边缘计算软件或设备上传的所述原始数据;
    对所述原始数据进行内部处理,生成所述内部处理结果;
    将所述内部处理结果反馈至预设的结果接收设备。
  11. 根据权利要求8-10任一项所述的系统,其特征在于,所述预设的结果接收设备为机器设备、数据中心设备或者运营管理设备。
  12. 根据权利要求11所述的系统,其特征在于,所述边缘计算软件或设备向预设的结果接收设备反馈所述数据分析结果,包括:
    所述边缘计算软件或设备以控制指令的方式,通过OT系统将所述数据分析结果反馈至所述机器设备。
  13. 根据权利要求8所述的系统,其特征在于,所述边缘计算软件或设备,还用于:
    基于预设的协议转化规则,将不同协议下的原始数据转化为指定协议下的原始数据。
  14. 根据权利要求8所述的系统,其特征在于,所述边缘计算软件或设备,还用于:
    所述边缘计算软件或设备获取OT系统的系统运行数据,将所述系统运行数据上传至边缘计算平台,以使所述边缘计算平台基于所述系统运行数据创建并训练OT系统决策模型;
    所述边缘计算软件或设备接收所述边缘计算平台反馈的OT系统决策模型,并基于所述OT系统决策模型对所述OT系统进行智能管理。
PCT/CN2019/074462 2019-01-09 2019-02-01 一种基于边缘计算的智能管理方法和系统 Ceased WO2020143094A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US17/421,391 US20220086052A1 (en) 2019-01-09 2019-02-01 Intelligent management method and system based on edge computing
EP19908290.0A EP3907641A4 (en) 2019-01-09 2019-02-01 INTELLIGENT MANAGEMENT PROCESS AND SYSTEM BASED ON EDGE COMPUTER

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910021104.3 2019-01-09
CN201910021104.3A CN109815733A (zh) 2019-01-09 2019-01-09 一种基于边缘计算的智能管理方法和系统

Publications (1)

Publication Number Publication Date
WO2020143094A1 true WO2020143094A1 (zh) 2020-07-16

Family

ID=66604089

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/074462 Ceased WO2020143094A1 (zh) 2019-01-09 2019-02-01 一种基于边缘计算的智能管理方法和系统

Country Status (4)

Country Link
US (1) US20220086052A1 (zh)
EP (1) EP3907641A4 (zh)
CN (1) CN109815733A (zh)
WO (1) WO2020143094A1 (zh)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112165511A (zh) * 2020-09-07 2021-01-01 天地(常州)自动化股份有限公司 煤矿监控系统、方法及计算机存储介质
CN112181668A (zh) * 2020-11-04 2021-01-05 深圳市蓝波湾通信技术有限公司 一种基于边缘计算端的智能管理方法、装置以及系统
CN112486677A (zh) * 2020-11-25 2021-03-12 深圳市中博科创信息技术有限公司 一种数据链与图传产品的实现方法及装置
CN113110360A (zh) * 2021-05-19 2021-07-13 中铁工程装备集团有限公司 隧道施工智能装备集群协同管控系统、方法及边缘计算站
CN113329358A (zh) * 2021-04-29 2021-08-31 广州中国科学院软件应用技术研究所 一种基于risc-v指令集的aiot多制式边缘网关通信系统及设备
CN114001440A (zh) * 2021-10-29 2022-02-01 珠海格力电器股份有限公司 一种空调组网方法
CN114257417A (zh) * 2021-11-26 2022-03-29 中国南方电网有限责任公司 网络安全数据处理方法、装置、计算机设备和存储介质
CN114301809A (zh) * 2021-12-31 2022-04-08 郑州云海信息技术有限公司 一种边缘计算平台架构
CN115314850A (zh) * 2022-07-01 2022-11-08 山东新一代信息产业技术研究院有限公司 一种基于云边协同控制的智能运动系统
WO2023147649A1 (en) * 2022-02-04 2023-08-10 Begin Ai Inc. Systems and methods for generating processable data for machine learning applications
CN116611536A (zh) * 2023-07-19 2023-08-18 支付宝(杭州)信息技术有限公司 模型训练方法及装置、电子设备及存储介质
CN117834454A (zh) * 2024-03-04 2024-04-05 中铁四局集团有限公司 物联网设备协议适配的快速接入方法
CN119017374A (zh) * 2024-07-22 2024-11-26 清华大学 基于移动巡检机器人的公共空间安全风险识别系统

Families Citing this family (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110493304B (zh) * 2019-07-04 2022-11-29 上海数据交易中心有限公司 边缘计算系统以及交易系统
CN110602176A (zh) * 2019-08-23 2019-12-20 江西憶源多媒体科技有限公司 一种利用边缘计算控制智慧路灯的方法
CN110602178B (zh) * 2019-08-26 2021-11-26 杭州电子科技大学 一种基于边缘压缩计算处理温度传感器数据的方法
US12052260B2 (en) * 2019-09-30 2024-07-30 International Business Machines Corporation Scalable and dynamic transfer learning mechanism
CN110867105B (zh) * 2019-10-23 2020-12-04 重庆特斯联智慧科技股份有限公司 一种基于边缘计算的家庭学习监督方法和系统
CN110543146A (zh) * 2019-10-30 2019-12-06 苏州市伯太数字科技有限公司 一种基于边缘计算的循圆加工方法及数控机床加工系统
WO2021087778A1 (zh) * 2019-11-05 2021-05-14 北京小米移动软件有限公司 数据处理系统、方法、装置、设备及可读存储介质
CN111277557B (zh) * 2019-12-04 2023-01-03 珠海派诺科技股份有限公司 一种实时通信方法、设备、存储介质
CN110850789A (zh) * 2019-12-09 2020-02-28 常州市小先信息技术有限公司 一种基于边缘计算的机器人能效优化方法
CN111182083B (zh) * 2020-02-11 2021-06-11 腾讯科技(深圳)有限公司 设备控制方法、装置和边缘设备
CN111371874B (zh) * 2020-02-28 2021-10-08 腾讯科技(深圳)有限公司 一种工业数据传输方法、装置、存储介质和电子设备
CN111901381A (zh) * 2020-06-29 2020-11-06 北京智芯微电子科技有限公司 基于边缘计算的物联代理装置及数据决策方法
CN112165479A (zh) * 2020-09-22 2021-01-01 北京智芯微电子科技有限公司 配电网的云边协同管控系统
CN112148568B (zh) * 2020-09-25 2023-09-22 昆仑数智科技有限责任公司 一种设备监控系统及方法
CN112291200B (zh) * 2020-09-30 2023-04-14 国电南瑞科技股份有限公司 一种边缘计算app可信接入物联边缘代理的方法及系统
CN112349404A (zh) * 2020-11-03 2021-02-09 中国人民解放军总医院 基于云-边-端架构的多中心医疗设备大数据云平台
CN112181382B (zh) * 2020-11-27 2021-08-27 北京和利时系统工程有限公司 一种基于软件定义的工业智能控制系统
CN112653703A (zh) * 2020-12-25 2021-04-13 杭州中科先进技术研究院有限公司 一种基于边缘计算的多医疗协议转换解析方法和系统
CN112650183B (zh) * 2020-12-29 2022-04-12 煤炭科学研究总院 工业互联网平台
CN112835318A (zh) * 2021-01-22 2021-05-25 茂名市智慧城市研究中心 一种工农业物联网数据采集方法及系统
CN113395323A (zh) * 2021-05-12 2021-09-14 罗贤德 基于边缘计算的可穿透内网的云测试平台及其测试方法
CN113357083A (zh) * 2021-08-09 2021-09-07 东方电气风电有限公司 一种风力发电机组智能控制系统及方法
CN113673856B (zh) * 2021-08-13 2024-04-26 南京邮电大学 基于3d的医院感控模拟培训考核系统
CN114697199B (zh) * 2022-05-31 2022-08-23 苏州映赛智能科技有限公司 一种基于边缘计算的高鲁棒性数据采集方法和系统
CN115499259A (zh) * 2022-08-26 2022-12-20 青岛海尔空调器有限总公司 用于控制家电设备的方法及装置、网关空调、存储介质
CN115551017A (zh) * 2022-09-28 2022-12-30 西安电子科技大学 一种基于微服务的云-网-边智能协同与管理编排方法
CN115766709A (zh) * 2022-11-22 2023-03-07 中国长江三峡集团有限公司 一种基于5g的声像电子文件自动归档方法
CN115801529B (zh) * 2023-01-18 2023-05-16 中冶南方(武汉)自动化有限公司 物联网边缘侧应用交互的方法和系统
CN116599787B (zh) * 2023-07-17 2023-10-03 中通服建设有限公司 一种智能家居环境监控系统
CN119814775B (zh) * 2024-11-18 2025-11-07 广西电网有限责任公司 一种电网边缘计算平台及部署方法
CN119644779B (zh) * 2024-12-11 2026-02-10 珠海格力电器股份有限公司 智能家居的管理方法、装置、电子设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104539732A (zh) * 2015-01-19 2015-04-22 成都创信云通科技有限公司 面向大量独立设备流数据的多级系统架构
US20170048308A1 (en) * 2015-08-13 2017-02-16 Saad Bin Qaisar System and Apparatus for Network Conscious Edge to Cloud Sensing, Analytics, Actuation and Virtualization
CN108667924A (zh) * 2018-04-18 2018-10-16 大连理工大学 一种提供边缘计算服务的网关设备建立方法
CN109151072A (zh) * 2018-10-26 2019-01-04 上海方融科技有限责任公司 一种基于雾节点的边缘计算系统

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111367187B (zh) * 2015-08-27 2023-10-24 江森自控泰科知识产权控股有限责任公司 用于改进对分布式网络中的传感器流数据的处理的方法
CN107766889B (zh) * 2017-10-26 2021-06-04 浪潮集团有限公司 一种云端边缘计算融合的深度学习计算系统和方法
CN107992949A (zh) * 2017-11-09 2018-05-04 北京许继电气有限公司 工业数据分析方法和系统
CN108805195A (zh) * 2018-04-03 2018-11-13 北京理工大学 一种基于二值深层神经网络的电机群故障诊断方法
CN109617796A (zh) * 2018-11-15 2019-04-12 江苏东洲物联科技有限公司 一种基于规则引擎的边缘计算网关

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104539732A (zh) * 2015-01-19 2015-04-22 成都创信云通科技有限公司 面向大量独立设备流数据的多级系统架构
US20170048308A1 (en) * 2015-08-13 2017-02-16 Saad Bin Qaisar System and Apparatus for Network Conscious Edge to Cloud Sensing, Analytics, Actuation and Virtualization
CN108667924A (zh) * 2018-04-18 2018-10-16 大连理工大学 一种提供边缘计算服务的网关设备建立方法
CN109151072A (zh) * 2018-10-26 2019-01-04 上海方融科技有限责任公司 一种基于雾节点的边缘计算系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3907641A4 *

Cited By (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112165511A (zh) * 2020-09-07 2021-01-01 天地(常州)自动化股份有限公司 煤矿监控系统、方法及计算机存储介质
CN112181668A (zh) * 2020-11-04 2021-01-05 深圳市蓝波湾通信技术有限公司 一种基于边缘计算端的智能管理方法、装置以及系统
CN112181668B (zh) * 2020-11-04 2024-05-14 深圳市蓝波湾通信技术有限公司 一种基于边缘计算端的智能管理方法、装置以及系统
CN112486677B (zh) * 2020-11-25 2024-01-12 深圳市中博科创信息技术有限公司 一种数据的图传方法及装置
CN112486677A (zh) * 2020-11-25 2021-03-12 深圳市中博科创信息技术有限公司 一种数据链与图传产品的实现方法及装置
CN113329358A (zh) * 2021-04-29 2021-08-31 广州中国科学院软件应用技术研究所 一种基于risc-v指令集的aiot多制式边缘网关通信系统及设备
CN113110360A (zh) * 2021-05-19 2021-07-13 中铁工程装备集团有限公司 隧道施工智能装备集群协同管控系统、方法及边缘计算站
CN114001440A (zh) * 2021-10-29 2022-02-01 珠海格力电器股份有限公司 一种空调组网方法
CN114257417A (zh) * 2021-11-26 2022-03-29 中国南方电网有限责任公司 网络安全数据处理方法、装置、计算机设备和存储介质
CN114301809A (zh) * 2021-12-31 2022-04-08 郑州云海信息技术有限公司 一种边缘计算平台架构
CN114301809B (zh) * 2021-12-31 2024-02-09 郑州云海信息技术有限公司 一种边缘计算平台架构
WO2023147649A1 (en) * 2022-02-04 2023-08-10 Begin Ai Inc. Systems and methods for generating processable data for machine learning applications
GB2622545A (en) * 2022-02-04 2024-03-20 Begin Ai Inc Systems and methods for generating processable data for machine learning applications
CN115314850A (zh) * 2022-07-01 2022-11-08 山东新一代信息产业技术研究院有限公司 一种基于云边协同控制的智能运动系统
CN116611536B (zh) * 2023-07-19 2023-09-29 支付宝(杭州)信息技术有限公司 模型训练方法及装置、电子设备及存储介质
CN116611536A (zh) * 2023-07-19 2023-08-18 支付宝(杭州)信息技术有限公司 模型训练方法及装置、电子设备及存储介质
CN117834454A (zh) * 2024-03-04 2024-04-05 中铁四局集团有限公司 物联网设备协议适配的快速接入方法
CN117834454B (zh) * 2024-03-04 2024-06-07 中铁四局集团有限公司 物联网设备协议适配的快速接入方法
CN119017374A (zh) * 2024-07-22 2024-11-26 清华大学 基于移动巡检机器人的公共空间安全风险识别系统

Also Published As

Publication number Publication date
EP3907641A4 (en) 2022-03-02
EP3907641A1 (en) 2021-11-10
CN109815733A (zh) 2019-05-28
US20220086052A1 (en) 2022-03-17

Similar Documents

Publication Publication Date Title
WO2020143094A1 (zh) 一种基于边缘计算的智能管理方法和系统
CN116962409B (zh) 基于通感算与虚拟控制器智能融合的工业互联网系统
CN113810953B (zh) 一种基于数字孪生的无线传感网络资源调度方法及系统
Chen et al. Edge computing in IoT-based manufacturing
CN110807905B (zh) 一种基于端-边-云架构的社区火灾监测系统
CN112804280A (zh) 电力物联网系统及其数据处理方法
CN112866096A (zh) 一种面向工业物联网的智慧物联网关
CN113542371B (zh) 基于边缘网关的资源调度方法及系统
CN112650581A (zh) 一种面向智能楼宇的云边协同任务调度方法
CN109450790B (zh) 一种基于流量感知的网络服务功能链智慧供应方法
CN106378436A (zh) 一种基于移动互联网的压铸生产信息监控系统
CN108924228B (zh) 基于边缘计算的工业互联网优化系统
CN117527846A (zh) 轮胎工业互联网边缘控制设备及方法
CN112162789A (zh) 一种基于软件定义的边缘计算随机卸载决策方法及系统
CN114866593A (zh) 一种端边协同的工业网络控制架构及资源联合分配方法
CN118739398A (zh) 基于边缘计算的多台区分布式光伏协同优化方法及系统
CN115293378A (zh) 一种基于机器人的海上风电运维方法及系统
CN116755404A (zh) 一种基于边缘计算的离散制造车间管理系统及方法
CN117750408A (zh) 一种基于物联网的通信故障嗅探方法
CN116684472B (zh) 一种面向端侧算力网络的服务部署系统及其服务部署方法
CN116224791A (zh) 智能制造协作机器人边缘系统的协作训练控制方法
CN115550112A (zh) 电力边缘智能网关及电力边缘数据处理方法
Li et al. Sdn based computation offloading for industrial internet of things
CN114640568B (zh) 基于深度强化学习的网络智能管控架构系统和运行方法
Xu et al. Application and experiments of 5g technology powered industrial internet

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19908290

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2019908290

Country of ref document: EP

Effective date: 20210809

ENP Entry into the national phase

Ref document number: 2019908290

Country of ref document: EP

Effective date: 20210806