Intelligent inspection method and system for wind driven generator

Abstract

The invention discloses an intelligent inspection method for a wind driven generator, and the method comprises the following steps: S1, equipment installation: arranging an AI edge server in each fan, and carrying out the real-time algorithm analysis of all collected data in the fan; s2, fan cabin inspection: a first track robot and a first track are arranged in the fan cabin, the first track is provided with a U-shaped bend at the upper part of the cabin, so that the first track robot moves on the first track to inspect the cabin, and an inspection result is output; s3, inspecting a vertical tower drum; s4, wheel hub inspection; and S5, performing man-machine interaction. The invention further discloses an intelligent inspection system for the wind driven generator. According to the inspection mode adopted by the invention, remote, intelligent and regular inspection can be realized, the safe operation of machinery equipment can be ensured, the personnel can be kept away from danger, the inspection efficiency is improved, the personnel cost can be reduced, and the inspection system has the advantages of safety, reliability and low cost.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind turbine inspection, and in particular relates to an intelligent inspection method and system for wind turbines. Background Art

[0002] At present, onshore wind power is continuously developing towards low wind speed areas. In order to better develop wind resources in low wind speed areas, wind turbine manufacturers have developed ultra-high tower units. As high towers, especially flexible towers, have become more and more popular in the industry, especially in the past two years, there have been many tower collapse accidents of high tower units. The safety of high towers has also become the focus of attention of owners and wind turbine manufacturers. For this reason, it is necessary to conduct inspections on wind turbines.

[0003] There are still some problems with the existing intelligent inspection methods and systems for wind turbines: most wind turbines are installed in remote, sparsely populated and difficult to reach places, such as Gobi, mountain tops, and shallow sea areas. Due to the remote location, the workers work and live in a difficult environment. At the same time, some operations require climbing, which is dangerous to personnel; wind turbines are generally used on a large scale, with many equipment and many contents to be inspected (motors, inverters, sensors, cables, etc.). The inspectors need to master many skills and tools. Due to the lack of inspection personnel and the untimely inspection work, the wind turbines are often shut down. For this reason, we propose an intelligent inspection method and system for wind turbines. Summary of the invention

[0004] The purpose of the present invention is to provide a wind turbine intelligent inspection method and system to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a wind turbine intelligent inspection method, comprising the following steps:

[0006] S1. Equipment installation: An AI edge server is installed in each wind turbine to perform real-time algorithm analysis on all collected data in the wind turbine.

[0007] S2. Inspection of the wind turbine cabin: a first track robot and a first track are arranged in the wind turbine cabin, a U-shaped bend is arranged on the upper part of the cabin of the first track, so that the first track robot moves on the first track to inspect the cabin and output the inspection results;

[0008] S3. Vertical tower inspection: Install a second track on the existing lifting system in the tower, use wire traction as the power mode, and then set up a second track robot that can move vertically up and down. The second track is arranged vertically in the tower, and the driving traction system is arranged at the bottom of the tower silo, so that the second track robot moves on the second track to realize the inspection of the tower and output the inspection results;

[0009] S4. Wheel hub inspection: A dual-spectrum gimbal is installed inside the wheel hub to fully cover the inside of the wheel hub and obtain high-definition video. An industrial listening sensor is installed inside the wheel hub to collect data on sound, temperature, and vibration. The AI ​​edge server then performs AI algorithm analysis to determine whether there are any abnormalities and outputs the inspection results.

[0010] S5. Human-computer interaction: The inspection results are transmitted back to the control center through the 5G network, and the inspection results are displayed and remotely controlled until the inspection work is completed.

[0011] Preferably, the U-shaped bend is set as one horizontal U-shaped bend, which is mainly used to inspect the equipment in the upper and middle areas of the cabin, or three U-shaped bends, one of which is a horizontal U-shaped bend and two are vertical U-shaped bends, which are used to inspect different areas of the cabin.

[0012] Preferably, the S2 wind turbine cabin inspection also includes: installing a high-definition camera and an infrared thermal imager on the first track robot, obtaining high-definition pictures of the inspection equipment and the area through the camera, obtaining temperature information of the inspection equipment through the thermal imager, performing AI algorithm analysis through the AI ​​edge server, and outputting the results to determine whether there are any abnormalities.

[0013] Preferably, the vertical tower inspection in S3 also includes: carrying three dual-light collectors on the second track robot to collect infrared imaging and visible light video data of the tower, and performing AI algorithm analysis through the AI ​​edge server to determine whether there are any abnormalities.

[0014] Preferably, the AI ​​algorithm adopts a neural network-based deep learning framework to perform analysis by extracting data features. The entire process includes steps such as data collection, feature labeling, data preprocessing, AI model training, and model optimization and adjustment.

[0015] Preferably, the data preprocessing includes detecting and removing outliers through a machine learning model, processing missing values ​​using mean filling, removing high-frequency noise using a Butterworth filter, and normalization and feature extraction processing.

[0016] The intelligent inspection system for wind turbines includes an interaction layer, a platform layer, a network layer and a perception layer. The interaction layer includes a web background and an IOC large screen. The platform layer includes a system management module, a business management module, an AI platform module and an API gateway. The network layer transmits data through a 5G network. The perception layer is used for data collection and is transmitted to the network layer through Ethernet and WIFI.

[0017] The system management module is used for organization management, user management, area management, system configuration and authority management;

[0018] The business management module is used for equipment management, alarm management, alarm push, inspection management, report management, temperature and vibration monitoring, team management, video monitoring and data analysis;

[0019] The AI ​​platform module is used for model management, training, annotation and algorithm push, and provides an AI platform;

[0020] The perception layer includes an acoustic, temperature and vibration system, a wind turbine cabin inspection module, a vertical tower inspection module, a hub inspection module, a fixed-point monitoring module and an edge server.

[0021] Preferably, the acoustic temperature vibration system is used to perform sound spectrum analysis, temperature field analysis and equipment vibration analysis;

[0022] The wind turbine nacelle inspection module is used to inspect the nacelle by means of a first track robot and a first track arranged in the wind turbine nacelle, and by means of the first track robot moving on the first track, and output the inspection result;

[0023] The vertical tower inspection module is used to inspect the tower by means of a second track and a second track robot installed in the tower, and the second track robot moves on the second track to output the inspection result;

[0024] The wheel hub inspection module is used to obtain high-definition video through a dual-spectrum pan / tilt installed inside the wheel hub, and is also used to collect data on sound, temperature, and vibration through an industrial listening sensor installed inside the wheel hub. In addition, an AI algorithm analysis is performed through an AI edge server to determine whether there is an abnormality and output the inspection result;

[0025] The fixed-point monitoring module is used to perform fixed-point monitoring on the fan.

[0026] Preferably, the service management module includes an inspection management module, the inspection management module is used to implement the inspection management function, and the inspection management module includes a device management unit, a line management unit, a plan management unit, a remote monitoring unit, an inspection record unit and a user management unit;

[0027] The equipment management unit is used to manage the inspection robot and various data acquisition and control equipment fixed to the robot;

[0028] The line management unit is used to reasonably set the inspection content and line of the inspection robot;

[0029] The plan management unit is used to formulate inspection plans and tasks and monitor execution results;

[0030] The remote monitoring unit is used to control the robot to monitor the area that needs to be viewed using a variety of technical means;

[0031] The inspection recording unit is used to record various execution data of the inspection for statistical analysis and subsequent query optimization;

[0032] The user management unit is used to perform hierarchical management of different permissions for users accessing the system.

[0033] Preferably, the AI ​​platform module includes a basic layer, a technical layer and an application layer, the basic layer is used for data collection, data processing and data understanding, the technical layer is used for feature extraction, model building, model evaluation, hyper-parameter debugging and model training, and the application layer is used for predictive reasoning, threshold strategy and model application.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] (1) The inspection method adopted by the present invention can realize remote, intelligent and regular inspection, which can not only ensure the safe operation of machinery and equipment, but also keep personnel away from danger. While improving the inspection efficiency, it can reduce personnel costs and has the advantages of safety, reliability and low cost.

[0036] (2) The inspection system of the present invention utilizes automation, intelligence and artificial intelligence technologies to realize automatic monitoring and detection of equipment, facilities, environment, etc. in the production process, meeting or exceeding the requirements of manual inspection, thereby improving inspection efficiency, discovering problems in a timely manner, ensuring personnel safety, and reducing labor intensity. At the same time, the inspection data is fully recorded for easy traceability, remote monitoring, and manual rapid investigation and confirmation. Data sharing is possible to achieve emergency linkage. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a flow chart of the present invention;

[0038] Figure 2 It is a structural framework diagram of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] See also Figure 1-Figure 2 The present invention provides a technical solution: a wind turbine intelligent inspection method, comprising the following steps:

[0041] S1. Equipment installation: An AI edge server is installed in each wind turbine to perform real-time algorithm analysis on all collected data in the wind turbine.

[0042] S2. Inspection of the wind turbine cabin: a first track robot and a first track are arranged in the wind turbine cabin, a U-shaped bend is arranged on the upper part of the cabin of the first track, so that the first track robot moves on the first track to inspect the cabin and output the inspection results;

[0043] S3. Vertical tower inspection: Install a second track on the existing lifting system in the tower, use wire traction as the power mode, and then set up a second track robot that can move vertically up and down. The second track is arranged vertically in the tower, and the driving traction system is arranged at the bottom of the tower silo, so that the second track robot moves on the second track to realize the inspection of the tower and output the inspection results;

[0044] S4. Wheel hub inspection: A dual-spectrum gimbal is installed inside the wheel hub to fully cover the inside of the wheel hub and obtain high-definition video. An industrial listening sensor is installed inside the wheel hub to collect data on sound, temperature, and vibration. The AI ​​edge server then performs AI algorithm analysis to determine whether there are any abnormalities and outputs the inspection results.

[0045] S5. Human-computer interaction: The output inspection results are transmitted back to the control center through the 5G network, and the inspection results are displayed and remotely controlled until the inspection work is completed.

[0046] In this embodiment, preferably, the U-shaped bend is set as one horizontal U-shaped bend, which is mainly used to inspect the equipment in the upper and middle areas of the cabin, or three U-shaped bends, one of which is a horizontal U-shaped bend and two are vertical U-shaped bends, which are used to inspect different areas of the cabin.

[0047] In this embodiment, preferably, the S2 wind turbine cabin inspection also includes: mounting a high-definition camera and an infrared thermal imager on the first track robot, obtaining high-definition pictures of the inspection equipment and the area through the camera, obtaining temperature information of the inspection equipment through the thermal imager, performing AI algorithm analysis through the AI ​​edge server, and outputting the results to determine whether there are any abnormalities.

[0048] In this embodiment, preferably, the vertical tower inspection in S3 also includes: carrying three dual-light collectors on the second track robot to collect infrared imaging and visible light video data of the tower, and performing AI algorithm analysis through the AI ​​edge server to determine whether there are any abnormalities.

[0049] In this embodiment, preferably, the AI ​​algorithm adopts a neural network deep learning framework to perform analysis by extracting data features. The whole process includes steps such as data collection, feature labeling, data preprocessing, AI model training, and model optimization and adjustment.

[0050] In this embodiment, preferably, the data preprocessing includes detecting and eliminating outliers through a machine learning model, processing missing values ​​with mean filling, removing high-frequency noise with a Butterworth filter, and normalization and feature extraction processing.

[0051] The intelligent inspection system for wind turbines includes an interaction layer, a platform layer, a network layer and a perception layer. The interaction layer includes a web background and an IOC large screen. The platform layer includes a system management module, a business management module, an AI platform module and an API gateway. The network layer transmits data through a 5G network. The perception layer is used for data collection and is transmitted to the network layer through Ethernet and WIFI.

[0052] The system management module is used for organization management, user management, area management, system configuration and authority management;

[0053] The business management module is used for equipment management, alarm management, alarm push, inspection management, report management, temperature and vibration monitoring, team management, video monitoring and data analysis;

[0054] The AI ​​platform module is used for model management, training, annotation and algorithm push, and provides an AI platform;

[0055] The perception layer includes an acoustic, temperature and vibration system, a wind turbine cabin inspection module, a vertical tower inspection module, a hub inspection module, a fixed-point monitoring module and an edge server.

[0056] In this embodiment, preferably, the acoustic temperature vibration system is used to perform sound spectrum analysis, temperature field analysis and equipment vibration analysis;

[0057] The wind turbine nacelle inspection module is used to inspect the nacelle by means of a first track robot and a first track arranged in the wind turbine nacelle, and by means of the first track robot moving on the first track, and output the inspection result;

[0058] The vertical tower inspection module is used to inspect the tower by means of a second track and a second track robot installed in the tower, and the second track robot moves on the second track to output the inspection result;

[0059] The wheel hub inspection module is used to obtain high-definition video through a dual-spectrum pan / tilt installed inside the wheel hub, and is also used to collect data on sound, temperature, and vibration through an industrial listening sensor installed inside the wheel hub. In addition, an AI algorithm analysis is performed through an AI edge server to determine whether there is an abnormality and output the inspection result;

[0060] The fixed-point monitoring module is used to perform fixed-point monitoring on the fan.

[0061] In this embodiment, preferably, the service management module includes an inspection management module, the inspection management module is used to implement the inspection management function, and the inspection management module includes a device management unit, a line management unit, a plan management unit, a remote monitoring unit, an inspection record unit and a user management unit;

[0062] The equipment management unit is used to manage the inspection robot and various data acquisition and control equipment fixed to the robot;

[0063] The line management unit is used to reasonably set the inspection content and line of the inspection robot;

[0064] The plan management unit is used to formulate inspection plans and tasks and monitor execution results;

[0065] The remote monitoring unit is used to control the robot to monitor the area that needs to be viewed using a variety of technical means;

[0066] The inspection recording unit is used to record various execution data of the inspection for statistical analysis and subsequent query optimization;

[0067] The user management unit is used to perform hierarchical management of different permissions for users accessing the system.

[0068] In this embodiment, preferably, the AI ​​platform module includes a basic layer, a technical layer and an application layer, the basic layer is used for data collection, data processing and data understanding, the technical layer is used for feature extraction, model construction, model evaluation, hyper-parameter debugging and model training, and the application layer is used for predictive reasoning, threshold strategy and model application.

[0069] Principles and advantages of the present invention:

[0070] The inspection method adopted by the present invention can realize remote, intelligent and regular inspection, which can not only ensure the safe operation of machinery and equipment, but also keep personnel away from danger. While improving the inspection efficiency, it can reduce personnel costs and has the advantages of safety, reliability and low cost.

[0071] The inspection system of the present invention utilizes automation, intelligence and artificial intelligence technologies to realize automatic monitoring and detection of equipment, facilities, environment, etc. in the production process, meeting or exceeding the requirements of manual inspection, thereby improving inspection efficiency, discovering problems in a timely manner, ensuring personnel safety, and reducing labor intensity. At the same time, the inspection data is fully recorded for easy traceability, remote monitoring, and manual rapid investigation and confirmation. Data sharing is possible to achieve emergency linkage.

[0072] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent inspection method for wind turbines, characterized in that: The following steps are involved: S1. Equipment installation: An AI edge server is installed in each wind turbine to perform real-time algorithm analysis on all collected data in the wind turbine. S2. Inspection of the wind turbine cabin: a first track robot and a first track are arranged in the wind turbine cabin, a U-shaped bend is arranged on the upper part of the cabin of the first track, so that the first track robot moves on the first track to inspect the cabin and output the inspection results; S3. Vertical tower inspection: Install a second track on the existing lifting system in the tower, use wire traction as the power mode, and then set up a second track robot that can move vertically up and down. The second track is arranged vertically in the tower, and the driving traction system is arranged at the bottom of the tower silo, so that the second track robot moves on the second track to realize the inspection of the tower and output the inspection results; S4. Wheel hub inspection: A dual-spectrum gimbal is installed inside the wheel hub to fully cover the inside of the wheel hub and obtain high-definition video. An industrial listening sensor is installed inside the wheel hub to collect data on sound, temperature, and vibration. The AI ​​edge server then performs AI algorithm analysis to determine whether there are any abnormalities and outputs the inspection results. S5. Human-computer interaction: The output inspection results are transmitted back to the control center through the 5G network, and the inspection results are displayed and remotely controlled until the inspection work is completed.

2. The intelligent inspection method for wind turbines according to claim 1 is characterized in that: The U-shaped bend is set as a horizontal U-shaped bend, which is mainly used to inspect the equipment in the upper and middle areas of the cabin, or three U-shaped bends, one of which is a horizontal U-shaped bend and two are vertical U-shaped bends, which are used to inspect different areas of the cabin.

3. The intelligent inspection method for wind turbines according to claim 1 is characterized in that: The S2 wind turbine cabin inspection also includes: installing a high-definition camera and an infrared thermal imager on the first track robot, obtaining high-definition pictures of the inspection equipment and area through the camera, obtaining temperature information of the inspection equipment through the thermal imager, performing AI algorithm analysis through the AI ​​edge server, and outputting the results to determine whether there are any abnormalities.

4. The intelligent inspection method for wind turbines according to claim 3 is characterized in that: The vertical tower inspection in S3 also includes: carrying three dual-light collectors on the second track robot to collect infrared imaging and visible light video data of the tower, and performing AI algorithm analysis through the AI ​​edge server to determine whether there are any abnormalities.

5. The intelligent inspection method for wind turbines according to claim 4 is characterized in that: The AI ​​algorithm adopts a neural network deep learning framework to perform analysis by extracting data features. The entire process includes steps such as data collection, feature labeling, data preprocessing, AI model training, and model optimization and adjustment.

6. The intelligent inspection method for wind turbines according to claim 5 is characterized in that: The data preprocessing includes detecting and removing outliers through a machine learning model, processing missing values ​​with mean filling, removing high-frequency noise with a Butterworth filter, and normalization and feature extraction.

7. The intelligent inspection system for wind turbines is characterized by: It includes an interaction layer, a platform layer, a network layer and a perception layer. The interaction layer includes a web background and an IOC large screen. The platform layer includes a system management module, a business management module, an AI platform module and an API gateway. The network layer transmits data through a 5G network. The perception layer is used for data collection and is transmitted to the network layer through Ethernet and WIFI. The system management module is used for organization management, user management, area management, system configuration and authority management; The business management module is used for equipment management, alarm management, alarm push, inspection management, report management, temperature and vibration monitoring, team management, video monitoring and data analysis; The AI ​​platform module is used for model management, training, annotation and algorithm push, and provides an AI platform; The perception layer includes an acoustic, temperature and vibration system, a wind turbine cabin inspection module, a vertical tower inspection module, a hub inspection module, a fixed-point monitoring module and an edge server.

8. The intelligent inspection system for wind turbines according to claim 7 is characterized in that: The acoustic temperature vibration system is used to perform sound spectrum analysis, temperature field analysis and equipment vibration analysis; The wind turbine nacelle inspection module is used to inspect the nacelle by means of a first track robot and a first track arranged in the wind turbine nacelle, and by means of the first track robot moving on the first track, and output the inspection result; The vertical tower inspection module is used to inspect the tower by means of a second track and a second track robot installed in the tower, and the second track robot moves on the second track to output the inspection result; The wheel hub inspection module is used to obtain high-definition video through a dual-spectrum pan / tilt installed inside the wheel hub, and is also used to collect data on sound, temperature, and vibration through an industrial listening sensor installed inside the wheel hub. In addition, an AI algorithm analysis is performed through an AI edge server to determine whether there is an abnormality and output the inspection result; The fixed-point monitoring module is used to perform fixed-point monitoring on the fan.

9. The intelligent inspection system for wind turbines according to claim 7 is characterized in that: The service management module includes an inspection management module, which is used to implement the inspection management function. The inspection management module includes a device management unit, a line management unit, a plan management unit, a remote monitoring unit, an inspection record unit and a user management unit; The equipment management unit is used to manage the inspection robot and various data acquisition and control equipment fixed to the robot; The line management unit is used to reasonably set the inspection content and line of the inspection robot; The plan management unit is used to formulate inspection plans and tasks and monitor execution results; The remote monitoring unit is used to control the robot to monitor the area that needs to be viewed using a variety of technical means; The inspection recording unit is used to record various execution data of the inspection for statistical analysis and subsequent query optimization; The user management unit is used to perform hierarchical management of different permissions for users accessing the system.

10. The intelligent inspection system for wind turbines according to claim 7, characterized in that: The AI ​​platform module includes a basic layer, a technical layer and an application layer. The basic layer is used for data collection, data processing and data understanding, the technical layer is used for feature extraction, model construction, model evaluation, hyperparameter debugging and model training, and the application layer is used for predictive reasoning, threshold strategy and model application.

Images