Search

CN-122027649-A - Edge cloud cooperative data acquisition method and system based on intelligent body time sensitive network

CN122027649ACN 122027649 ACN122027649 ACN 122027649ACN-122027649-A

Abstract

The invention discloses a method and a system for acquiring side cloud collaborative data based on a body intelligent time sensitive network, and relates to the technical field of data acquisition, wherein the method comprises the steps of 1 integrating multiple types of sensors including a visual sensor, a vibration sensor, a temperature sensor, an inertial navigation module and a dust concentration sensor, dynamically adjusting acquisition frequency to be 10Hz-2000Hz by utilizing a decision algorithm and combining historical acquisition data and a service demand threshold value, acquiring scene environment data, equipment running state, position information and moving speed in real time, 2 constructing a special data transmission network based on TSN, optimizing transmission requirements of different scenes through protocols and hardware deployment, 3 performing side cloud collaborative intelligent processing and splitting, optimizing deployment edge nodes, configuring splitting strategies, performing cloud global optimization, and 4 adopting a hierarchical encryption strategy and dynamic adaptation mechanism to balance data security and transmission instantaneity and cope with scene change and link fluctuation.

Inventors

  • WANG AILONG
  • WU SHUAI
  • YANG PENG

Assignees

  • 浪潮通信技术有限公司

Dates

Publication Date
20260512
Application Date
20260203

Claims (8)

  1. 1. A cloud-edge collaborative data acquisition method based on a self-contained intelligent time-sensitive network is characterized by comprising the following steps: the method comprises the steps of 1 integrating multiple types of sensors, including a visual sensor, a vibration sensor, a temperature sensor, an inertial navigation module and a dust concentration sensor, dynamically adjusting the acquisition frequency to 10Hz-2000Hz by combining historical acquisition data and a business demand threshold value through a decision algorithm, and acquiring scene environment data, equipment running state, position information and moving speed in real time; Step 2, constructing a special data transmission network based on TSN, and adapting transmission requirements of different scenes with hardware deployment through protocol optimization: Aiming at a general scene, a PTP time synchronization protocol is adopted to realize microsecond time synchronization of an edge node and a cloud, a transmission bandwidth is allocated through a time slot scheduling algorithm based on data priority, the time slot of high priority data accounts for 30% -50%, Aiming at the track traffic inspection scene, 1 TSN repeater is deployed along each section of the tunnel according to the signal shielding problem of the tunnel, a wired and wireless dual-mode backup transmission mode is adopted, when the signal to noise ratio of a wireless link is lower than 15dB, the wired transmission is automatically switched, The method comprises the steps of 3, performing edge cloud collaborative intelligent processing and distribution, namely optimally deploying edge nodes, configuring a distribution strategy, namely, controlling real-time control class data to be processed locally, controlling response time of a general scene to be less than or equal to 50ms, enabling the response time of the general scene to be less than or equal to 100ms, and enabling the general scene to be free from uploading cloud, performing asynchronous cloud uploading through a TSN low-priority channel after non-real-time analysis class data are compressed, performing cloud global optimization, namely, receiving compressed data uploaded by edges, performing big data analysis and AI model training, wherein the general scene transmits an optimization instruction to the edge nodes every 24 hours, and transmitting an optimization instruction to the edge nodes every 12 hours; And 4, balancing data security and transmission instantaneity by adopting a hierarchical encryption strategy and a dynamic adaptation mechanism, and coping with scene change and link fluctuation.
  2. 2. The method for acquiring the edge cloud collaborative data based on the intelligent body time-sensitive network according to claim 1, wherein in step 1, the method comprises the following steps: setting the equipment vibration overrun and temperature anomaly as high-priority acquisition triggering conditions aiming at the general scene, Aiming at a track traffic inspection scene, the method comprises the steps of obtaining the moving speed of an inspection robot, dynamically adjusting the image acquisition frequency of a steel rail to 1000Hz-2000Hz, automatically improving the exposure and contrast of images when the dust concentration is more than 50mg/m < 3 >, avoiding image blurring caused by dust, and setting the conditions that the steel rail crack is more than or equal to 1mm, the abrasion of a contact net is more than or equal to 0.5mm, and the ultrasonic detection sedimentation is more than or equal to 0.1mm as high-priority acquisition triggering conditions.
  3. 3. The method for acquiring the edge cloud collaborative data based on the intelligent body time-sensitive network according to claim 1, wherein the optimizing deployment of the edge node in the step 3 comprises the following steps: the method comprises the steps of deploying DRM (digital rights management) of a distributed resource manager at an edge node, monitoring the power load in real time, integrating a cross-protocol conversion engine, supporting more than 23 industrial protocols and analysis of a special track traffic protocol, configuring a data compression unit, adopting a conventional compression algorithm for a general scene, enabling the compression ratio to be more than or equal to 10:1, adopting an H.266 video compression standard for a track traffic inspection scene, enabling the compression ratio to be 40:1, and greatly reducing cloud transmission pressure.
  4. 4. The method for acquiring the edge cloud collaborative data based on the intelligent body time-sensitive network according to claim 1, wherein in step 4, the method comprises the following steps: Hierarchical encryption is carried out, namely, in a general scene, a national encryption algorithm is adopted to encrypt control instructions, lightweight verification is adopted to encrypt non-real-time data, in a track traffic inspection scene, a national encryption SM9 identifier is adopted to encrypt core data, When dynamic adaptation adjustment is carried out, the calculation load of the edge node is monitored in real time, when the load rate of a general scene exceeds 85 percent and the track traffic inspection scene exceeds 80 percent, the low-priority data acquisition frequency is automatically reduced, the quality of a transmission link is monitored in real time, when the signal to noise ratio of the link is reduced and the transmission delay is increased, the redistribution of TSN time slots is triggered, the high-priority data transmission is preferentially ensured, and when the scene environment is changed, the acquisition parameters are dynamically adjusted, and the data quality is ensured.
  5. 5. An edge cloud cooperative data acquisition system based on a body intelligent time sensitive network is characterized by comprising a body intelligent sensing module, a TSN communication module, an edge processing module, a cloud service module and a security encryption module, The intelligent body sensing module integrates multiple types of sensors and comprises a visual sensor, a vibration sensor, a temperature sensor, an inertial navigation module and a dust concentration sensor, and dynamically adjusts the acquisition frequency to be 10Hz-2000Hz by combining historical acquisition data and a business demand threshold value through a decision algorithm to acquire scene environment data, equipment running state, position information and moving speed in real time; The TSN communication module constructs a special data transmission network based on TSN, and adapts transmission requirements of different scenes through protocol optimization and hardware deployment: Aiming at a general scene, a PTP time synchronization protocol is adopted to realize microsecond time synchronization of an edge node and a cloud, a transmission bandwidth is allocated through a time slot scheduling algorithm based on data priority, the time slot of high priority data accounts for 30% -50%, Aiming at the track traffic inspection scene, 1 TSN repeater is deployed along each section of the tunnel according to the signal shielding problem of the tunnel, a wired and wireless dual-mode backup transmission mode is adopted, when the signal to noise ratio of a wireless link is lower than 15dB, the wired transmission is automatically switched, The edge processing module and the cloud service module perform edge cloud collaborative intelligent processing and distribution, wherein the edge processing module firstly optimizes and deploys edge nodes; the cloud service module performs cloud global optimization by receiving compressed data uploaded by edges, performing big data analysis and AI model training, wherein the general scene transmits an optimization instruction to edge nodes every 24 hours, and the track traffic inspection scene transmits an optimization instruction to the edge nodes every 12 hours; The security encryption module adopts a hierarchical encryption strategy and a dynamic adaptation mechanism to balance data security and transmission instantaneity and cope with scene change and link fluctuation.
  6. 6. The system for acquiring the edge cloud collaborative data based on the intelligent time-sensitive network of claim 5, wherein the intelligent sensing module sets the equipment vibration overrun and the temperature anomaly as high-priority acquisition triggering conditions aiming at the general scene, Aiming at a track traffic inspection scene, the method comprises the steps of obtaining the moving speed of an inspection robot, dynamically adjusting the image acquisition frequency of a steel rail to 1000Hz-2000Hz, automatically improving the exposure and contrast of images when the dust concentration is more than 50mg/m < 3 >, avoiding image blurring caused by dust, and setting the conditions that the steel rail crack is more than or equal to 1mm, the abrasion of a contact net is more than or equal to 0.5mm, and the ultrasonic detection sedimentation is more than or equal to 0.1mm as high-priority acquisition triggering conditions.
  7. 7. The intelligent self-contained time-sensitive network-based edge cloud collaborative data acquisition system according to claim 5, wherein the edge processing module optimizes deployment of edge nodes comprising: the method comprises the steps of deploying DRM (digital rights management) of a distributed resource manager at an edge node, monitoring the power load in real time, integrating a cross-protocol conversion engine, supporting more than 23 industrial protocols and analysis of a special track traffic protocol, configuring a data compression unit, adopting a conventional compression algorithm for a general scene, enabling the compression ratio to be more than or equal to 10:1, adopting an H.266 video compression standard for a track traffic inspection scene, enabling the compression ratio to be 40:1, and greatly reducing cloud transmission pressure.
  8. 8. The cloud-based collaborative data acquisition system based on the intelligent body-equipped time-sensitive network according to claim 5, wherein the security encryption module performs hierarchical encryption, wherein in a general scene, a national encryption algorithm is adopted to encrypt control instructions, a lightweight check is adopted to encrypt non-real-time data, in a track traffic inspection scene, a national encryption SM9 identifier is adopted to encrypt core data, When dynamic adaptation adjustment is carried out, the calculation load of the edge node is monitored in real time, when the load rate of a general scene exceeds 85 percent and the track traffic inspection scene exceeds 80 percent, the low-priority data acquisition frequency is automatically reduced, the quality of a transmission link is monitored in real time, when the signal to noise ratio of the link is reduced and the transmission delay is increased, the redistribution of TSN time slots is triggered, the high-priority data transmission is preferentially ensured, and when the scene environment is changed, the acquisition parameters are dynamically adjusted, and the data quality is ensured.

Description

Edge cloud cooperative data acquisition method and system based on intelligent body time sensitive network Technical Field The invention discloses a cloud-edge collaborative data acquisition method and system based on a body-equipped intelligent time-sensitive network, and relates to the technical field of data acquisition. Background In the fields of industrial manufacture, intelligent inspection, smart cities and the like, data acquisition is a core foundation for realizing equipment monitoring, quality control, intelligent decision making and safe operation and maintenance. Along with the deep digital transformation of industry, the real-time performance, reliability, intellectualization and multi-equipment synergy requirements of a time-sensitive scene on data acquisition are increasingly severe, but the existing cloud-edge collaborative data acquisition architecture still has some defects that the actual application requirements are difficult to meet: Bandwidth resources are severely squeezed, a single device in an industrial scene can generate GB-level vibration and visual data every second, the high-definition image/video data in the intelligent inspection scene is huge in quantity, the network is congested due to the fact that the cloud is uploaded in full quantity, the enterprise needs to continuously add bandwidth investment, and transmission efficiency is low. The communication delay is unstable, and even if a 5G MEC technology is adopted, the round trip delay of a cloud instruction of the existing architecture still is difficult to meet the millisecond control requirement. The problem of data island is remarkable in that communication protocols adopted by different devices are large in difference, such as protocols of quality inspection devices, AGV (automatic guided vehicle) and mechanical arms in industrial scenes are incompatible, so that data cannot be effectively fused. The calculation power distribution is unreasonable, namely the calculation power scheduling of the edge nodes and the cloud is lack of an intelligent adaptation mechanism, the situation that the edge terminals are overloaded and dead or the cloud resources are wasted is easy to occur, and the calculation power utilization rate is generally lower than 65%. The special scene adaptation is insufficient, such as the problems of signal shielding, dust interference, high mobility of inspection equipment and the like in a track traffic tunnel, the conventional TSN network lacks a relay backup mechanism, and the acquisition parameters are dynamically adjusted by a self-intelligent unbonded scene, so that the data transmission loss rate is high, and the defect identification accuracy is low. The security and the instantaneity conflict that the physical exposure risk of the edge terminal is high, the real-time property of data transmission is affected by the full-quantity encryption processing, the single encryption strategy can not balance the requirements of the edge terminal and the data transmission, and particularly the requirements of security sensitive scenes such as rail transit and the like on data encryption are higher. In the prior art, bian Yun collaborative architecture focuses on simple task splitting and data transmission, does not deeply integrate the scene self-adaptation capability with body intelligence and the deterministic transmission characteristics of TSN, lacks customized design for special scenes such as rail transit, and cannot fundamentally solve the pain points. Disclosure of Invention Aiming at the problems of the prior art, the invention provides a method and a system for acquiring edge-cloud cooperation data based on an intelligent body time-sensitive network, which solve the technical problems of high bandwidth occupation, unstable time delay, data island, unreasonable calculation power distribution, insufficient special scene adaptation and conflict between safety and real-time performance in the existing data acquisition process, and particularly meet the core requirements of high real-time performance, high reliability, multi-equipment cooperation and interference resistance of scenes such as track traffic inspection. The specific scheme provided by the invention is as follows: the invention provides a cloud-edge collaborative data acquisition method based on an intelligent time-sensitive network, which comprises the following steps: the method comprises the steps of 1 integrating multiple types of sensors, including a visual sensor, a vibration sensor, a temperature sensor, an inertial navigation module and a dust concentration sensor, dynamically adjusting the acquisition frequency to 10Hz-2000Hz by combining historical acquisition data and a business demand threshold value through a decision algorithm, and acquiring scene environment data, equipment running state, position information and moving speed in real time; Step 2, constructing a special data transmission network based on TSN, and adaptin