Self-protection method and device for rail-mounted inspection robot in severe environment

Abstract

The invention provides a self-protection method and device for a rail-mounted inspection robot in a severe environment, and the method comprises the steps: shooting the high-definition image data of first target equipment in an inspection region through a high-definition camera, and extracting the operation state features of the first target equipment based on the high-definition image data; determining an abnormal condition of the first target equipment based on the operation state characteristics, and determining a danger level according to the abnormal condition; and when the danger level is higher than the target danger level, controlling the high-definition camera to enter a self-protection mode, and switching to use the radar detector to inspect the first target equipment. According to the invention, when the abnormal condition of the target equipment is analyzed to cause a damage risk to the high-definition camera of the rail-mounted inspection robot, the high-definition camera is controlled to perform a self-protection mode, so that the equipment is prevented from being damaged.

Description

Technical Field

[0001] The present invention relates to the technical field of inspection robots, and in particular to a self-protection method and device for a rail-hanging inspection robot against harsh environments. Background Art

[0002] Inspection robots are divided into track-type and non-track-type, and the track-type is further divided into ground track-type and suspended track-type. Due to the equipment placement in the inspection area, the reservation of the channel area, etc., the suspended track-type inspection robot (hanging track intelligent inspection robot) runs along the suspended track, does not rely on the ground environment, can climb slopes autonomously, and has its own liftable pan / tilt and high-definition camera, which is very suitable for inspections in narrow or complex areas.

[0003] However, the existing rail inspection robot lacks the ability to make self-protection decisions in harsh environments and cannot adjust its own self-protection strategy according to the actual harsh environment, resulting in an inability to balance the completion of the inspection task and the integrity of the inspection equipment. The present invention aims to improve this technical problem. Summary of the invention

[0004] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a self-protection method, device, electronic equipment, computer storage medium and computer program product of a rail-hanging inspection robot against harsh environments.

[0005] The present invention provides a self-protection method for a rail-hanging inspection robot against a harsh environment, the method comprising the following steps: Using a high-definition camera to capture high-definition image data of a first target device in the inspection area, and extracting operating status characteristics of the first target device based on the high-definition image data; Determine an abnormality of the first target device based on the operating status characteristics, and determine the danger level according to the abnormality; When the danger level is higher than the target danger level, the high-definition camera is controlled to enter a self-protection mode, and a radar detector is switched to inspect the first target device.

[0006] Optionally, the step of using a high-definition camera to capture high-definition image data of a first target device in the inspection area, and extracting operating status characteristics of the first target device based on the high-definition image data includes: Using a high-definition camera to capture first high-definition image data in the inspection area; Determine appearance feature data and inspection items of the first target device according to the current inspection task, and identify the first target device in the first high-definition image data based on the appearance feature data; A high-definition camera is used to capture second high-definition image data of the first target device, and the operating status characteristics of the first target device are extracted from the second high-definition image data based on the inspection items.

[0007] Optionally, the determining an abnormal situation of the first target device based on the operating status feature, and determining the danger level according to the abnormal situation includes: Using a first abnormality analysis model to analyze the operating state feature, obtain a number of abnormal conditions, and segment the operating state feature to obtain operating state sub-features corresponding to each of the abnormal conditions; Inputting each of the abnormal conditions and the corresponding operating status sub-feature into a second abnormality analysis model, the second abnormality analysis model outputting the predicted danger level; Among them, the first abnormality analysis model is trained based on the first set of training data, and each training data in the first set of training data includes the operating status sub-features of the first target device and the corresponding abnormal annotation data; the second abnormality analysis model is trained based on the second set of training data, and each training data in the second set of training data includes the operating status sub-features of the first target device, lens damage data and corresponding danger annotation data.

[0008] Optionally, when the acquisition distance of the operating status feature is higher than a preset distance, determining the abnormal condition of the first target device based on the operating status feature includes: Determine the abnormality type and the corresponding first abnormality degree of the first target device based on the operating status characteristics; Determine a second target device associated with the abnormal type of the first target device according to a preset association relationship, and predict the abnormality coefficient of the first target device according to the operating state characteristics of the second target device; wherein the second target device is closer to the rail inspection robot than the first target device; A second abnormality degree is calculated using the abnormality degree coefficient and the first abnormality degree, and the abnormal situation is determined according to the abnormality type and the second abnormality degree.

[0009] Optionally, the predicting of an abnormality coefficient of the first target device according to the operating state characteristics of the second target device includes: Inputting the operating state characteristics of the second target device and the preset association relationship into an abnormality coefficient prediction model to obtain the predicted abnormality coefficient; The abnormality coefficient prediction model is obtained by fine-tuning the pre-trained large model.

[0010] Optionally, controlling the high-definition camera to enter a self-protection mode includes: The high-definition camera is controlled to be turned off and a protective cover is used to protect the high-definition camera, and a radar detector is switched to inspect the first target device.

[0011] The present invention also provides a self-protection device for a rail-hanging inspection robot against a harsh environment, the device comprising a state extraction module, a hazard level analysis module, and a protection processing module; The state extraction module is used to use a high-definition camera to capture high-definition image data of a first target device in the inspection area, and extract the operating state characteristics of the first target device based on the high-definition image data; The danger level analysis module is used to determine the abnormal situation of the first target device based on the operating state characteristics, and determine the danger level according to the abnormal situation; The protection processing module is used to control the high-definition camera to enter a self-protection mode and switch to using a radar detector to inspect the first target device when the danger level is higher than the target danger level.

[0012] The present invention also provides an electronic device, comprising: a memory storing executable program code; a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute any of the methods described above.

[0013] The present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, any of the above methods is executed.

[0014] The present invention also provides a computer program product, which includes a computer program stored in a computer storage medium, and when the computer program is executed by a processor of an electronic device, it implements any of the methods described above.

[0015] The present invention normally performs image inspection on the first target device. When an abnormal situation is found in the first target device and the abnormal situation will cause a risk of damage to the high-definition camera of the rail-hanging inspection robot, the high-definition camera is controlled to perform self-protection, thereby preventing the high-definition camera from being damaged by fast floating objects, acidic gases, etc. in harsh environments; moreover, a radar detector with better self-protection capability is switched to be used to inspect the first target device, thereby ensuring that the inspection task can be completed smoothly and improving the inspection quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 It is a schematic diagram of a self-protection method of a rail-hanging inspection robot against a harsh environment disclosed in an embodiment of the present invention.

[0018] Figure 2 It is a schematic diagram of a self-protection device of a rail-hanging inspection robot against harsh environments disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0020] See also Figure 1 The embodiment of the present invention discloses a self-protection method of a rail-hanging inspection robot against a harsh environment, the method comprising the following steps: Using a high-definition camera to capture high-definition image data of a first target device in the inspection area, and extracting operating status characteristics of the first target device based on the high-definition image data; Determine an abnormality of the first target device based on the operating status characteristics, and determine the danger level according to the abnormality; When the danger level is higher than the target danger level, the high-definition camera is controlled to enter a self-protection mode, and a radar detector is switched to inspect the first target device.

[0021] The rail inspection robot of the present invention is equipped with at least two sets of inspection equipment, namely, a high-definition camera and a radar detector. The high-definition camera can provide high-resolution images, thereby realizing a higher-accuracy inspection of the first target device, while the detection resolution of the radar detector is not as good as that of the high-definition camera. However, the radar detector has a stronger adaptability to harsh environments, while the high-definition camera is easily affected by floating objects and acidic gases in harsh environments. For example, fast-moving floating objects in harsh environments will scratch the lens of the high-definition camera, and acidic gases will damage the coating on the lens of the high-definition camera. In view of the above technical problems, the present invention uses a high-definition camera to shoot high-definition image data of the first target device in the inspection area, and extracts the operating state characteristics of the first target device based on the high-definition image data obtained by shooting; then, the abnormal situation of the first target device is determined based on its operating state characteristics, and the danger level of the first target device for the high-definition camera of the rail inspection robot is determined based on the existing abnormal situation. When the danger level is higher than the target danger level, it indicates that the high-definition camera of the rail inspection robot is easily damaged by the abnormal situation. At this time, the high-definition camera is promptly controlled to enter the self-protection mode, and the radar detector is also switched to inspect the first target device.

[0022] Therefore, the present invention performs image inspection on the first target device under normal circumstances. When an abnormal situation is found in the first target device and the abnormal situation will cause a risk of damage to the high-definition camera of the rail-hanging inspection robot, the high-definition camera is controlled to perform self-protection, thereby preventing the high-definition camera from being damaged by fast floating objects, acidic gases, etc. in harsh environments; moreover, it also switches to use a radar detector with better self-protection capability to inspect the first target device, thereby ensuring that the inspection task can be completed smoothly and improving the inspection quality.

[0023] Optionally, the step of using a high-definition camera to capture high-definition image data of a first target device in the inspection area, and extracting operating status characteristics of the first target device based on the high-definition image data includes: Using a high-definition camera to capture first high-definition image data in the inspection area; Determine appearance feature data and inspection items of the first target device according to the current inspection task, and identify the first target device in the first high-definition image data based on the appearance feature data; A high-definition camera is used to capture second high-definition image data of the first target device, and the operating status characteristics of the first target device are extracted from the second high-definition image data based on the inspection items.

[0024] In this embodiment, the rail inspection robot receives an inspection task sent by the server, and the inspection task includes the appearance feature data and specific inspection items of the first target device to be inspected this time. Therefore, the rail inspection robot can identify the first target device in the first high-definition image data of the inspection area captured by the high-definition camera based on the appearance feature data. Then, in the process of getting closer to the first target device, the second high-definition image data of the first target device is captured, and the operating status characteristics of the first target device are extracted from the second high-definition image data based on the inspection items obtained by the analysis, and the operating status characteristics are, for example, gas leakage characteristics, exhaust emission characteristics, operating speed, etc.

[0025] Optionally, the determining an abnormal situation of the first target device based on the operating status feature, and determining the danger level according to the abnormal situation includes: Using a first abnormality analysis model to analyze the operating state feature, obtain a number of abnormal conditions, and segment the operating state feature to obtain operating state sub-features corresponding to each of the abnormal conditions; Inputting each of the abnormal conditions and the corresponding operating status sub-feature into a second abnormality analysis model, the second abnormality analysis model outputting the predicted danger level; Among them, the first abnormality analysis model is trained based on the first set of training data, and each training data in the first set of training data includes the operating status sub-features of the first target device and the corresponding abnormal annotation data; the second abnormality analysis model is trained based on the second set of training data, and each training data in the second set of training data includes the operating status sub-features of the first target device, lens damage data and corresponding danger annotation data.

[0026] In this embodiment, the present invention constructs two models, namely a first abnormality analysis model and a second abnormality analysis model. The first abnormality analysis model is used to perform abnormality analysis on the extracted operating status characteristics of the first target device to determine a number of abnormal conditions existing in the first target device, such as gas leakage, abnormal exhaust color, etc. The second abnormality analysis model is used to analyze the danger level caused by the abnormal condition to the lens of the high-definition camera based on the abnormal condition and the corresponding operating status sub-characteristics. Different danger levels represent, for example, whether the amount or concentration of the leaked acid gas will cause damage to the lens of the high-definition camera, and the probability of causing damage.

[0027] The operating status sub-feature is segmented from the operating status feature according to the corresponding abnormal situation.

[0028] Optionally, when the acquisition distance of the operating status feature is higher than a preset distance, determining the abnormal condition of the first target device based on the operating status feature includes: Determine the abnormality type and the corresponding first abnormality degree of the first target device based on the operating status characteristics; Determine a second target device associated with the abnormal type of the first target device according to a preset association relationship, and predict the abnormality coefficient of the first target device according to the operating state characteristics of the second target device; wherein the second target device is closer to the rail inspection robot than the first target device; A second abnormality degree is calculated using the abnormality degree coefficient and the first abnormality degree, and the abnormal situation is determined according to the abnormality type and the second abnormality degree.

[0029] In this embodiment, in order to further reduce the risk of the high-definition camera of the rail-hanging inspection robot being damaged by abnormal target devices, the rail-hanging inspection robot can acquire the operating status characteristics of the target device in advance. However, due to factors such as the shooting angle and occlusion between devices, the acquired operating status characteristics may be incomplete or lack confidence.

[0030] In response to this technical problem, the present invention is provided to first determine the abnormal type and the corresponding first abnormality degree of the first target device based on the extracted operating state characteristics of the first target device itself. Then, the operating state characteristics of the second target device that is closer to the rail inspection robot than the first target device are obtained, and the abnormality coefficient of the first target device is predicted based on the operating state characteristics of the second target device. The abnormality coefficient is the impact of the first target device obtained by indirect analysis under the abnormal condition of the second target device. For example, a certain abnormality of the second target device may also cause or aggravate gas leakage in the first target device. The influence relationship of such abnormal conditions between the devices can be determined in advance, and a corresponding preset association relationship can be established. Finally, the abnormality coefficient and the first abnormality are used to calculate the second abnormality, for example, the second abnormality = the first abnormality multiplied by the abnormality coefficient, and the abnormal condition of the target device is determined according to the abnormality type and the second abnormality.

[0031] Optionally, the predicting of an abnormality coefficient of the first target device according to the operating state characteristics of the second target device includes: Inputting the operating state characteristics of the second target device and the preset association relationship into an abnormality coefficient prediction model to obtain the predicted abnormality coefficient; The abnormality coefficient prediction model is obtained by fine-tuning the pre-trained large model.

[0032] In this embodiment, the present invention obtains an abnormality coefficient prediction model by fine-tuning the pre-trained large model, and then uses the abnormality coefficient prediction model to perform in-depth analysis on the operating state characteristics and preset association relationship of the aforementioned second target device, thereby obtaining the corresponding abnormality coefficient. Among them, the preset association relationship can be unstructured data, such as a description of the relationship between the second target device and the first target device in terms of the mutual influence of certain abnormal types.

[0033] Pre-trained large models are, for example, large models such as BERT and GPT series that are constructed based on the Transformer architecture. These large models are pre-trained on large-scale text data sets to learn the basic laws of language and rich semantic information, and then fine-tune training using the constructed training data samples to obtain the anomaly coefficient prediction model of the present invention.

[0034] Each piece of training data includes an operating status feature of the second target device, an operating status feature of the first target device, and a corresponding impact level label.

[0035] Optionally, controlling the high-definition camera to enter a self-protection mode includes: The high-definition camera is controlled to be turned off and a protective cover is used to protect the high-definition camera, and a radar detector is switched to inspect the first target device.

[0036] In this embodiment, the self-protection of the high-definition camera is achieved by shutting down and using a protective cover, wherein the protective cover can be a cover that protects the lens or even the body of the high-definition camera. In the self-protection mode, the protective cover is activated to cover the lens or even the body of the high-definition camera to isolate it from the harsh environment, thereby protecting the high-definition camera. In the conventional inspection mode, the protective cover is in a retracted state, which does not affect the inspection of the first target device in the inspection area based on image recognition technology using high-definition cameras. Of course, when the protective cover is in a retracted state, the radar detector is also in an off state or low-power mode.

[0037] See also Figure 2 As shown, an embodiment of the present invention further provides a self-protection device for a rail-hanging inspection robot against a harsh environment, the device comprising a state extraction module, a hazard level analysis module, and a protection processing module; the characteristics are: The state extraction module is used to use a high-definition camera to capture high-definition image data of a first target device in the inspection area, and extract the operating state characteristics of the first target device based on the high-definition image data; The danger level analysis module is used to determine the abnormal situation of the first target device based on the operating state characteristics, and determine the danger level according to the abnormal situation; The protection processing module is used to control the high-definition camera to enter a self-protection mode and switch to using a radar detector to inspect the first target device when the danger level is higher than the target danger level.

[0038] An embodiment of the present invention further provides an electronic device, comprising: a memory storing executable program code; a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the method described in any of the above embodiments.

[0039] An embodiment of the present invention further provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described in any of the above embodiments is executed.

[0040] An embodiment of the present invention further provides a computer program product, which includes a computer program stored in a computer storage medium, and when the computer program is executed by a processor of an electronic device, the method described in any of the above embodiments is implemented.

[0041] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (apparatus), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0042] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0043] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0044] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A self-protection method for a rail-hanging inspection robot against a harsh environment, characterized in that: The method comprises the following steps: Using a high-definition camera to capture high-definition image data of a first target device in the inspection area, and extracting operating status characteristics of the first target device based on the high-definition image data; Determine an abnormality of the first target device based on the operating status characteristics, and determine a danger level according to the abnormality; When the danger level is higher than the target danger level, the high-definition camera is controlled to enter a self-protection mode, and a radar detector is switched to inspect the first target device.

2. The self-protection method of a rail-hanging inspection robot against a harsh environment according to claim 1, characterized in that: Using a high-definition camera to capture high-definition image data of a first target device in the inspection area, and extracting the operating status characteristics of the first target device based on the high-definition image data, including: Using a high-definition camera to capture first high-definition image data in the inspection area; Determine appearance feature data and inspection items of the first target device according to the current inspection task, and identify the first target device in the first high-definition image data based on the appearance feature data; A high-definition camera is used to capture second high-definition image data of the first target device, and the operating status characteristics of the first target device are extracted from the second high-definition image data based on the inspection items.

3. The self-protection method of a rail-hanging inspection robot against a harsh environment according to claim 2, characterized in that: Determining an abnormal situation of the first target device based on the operating status characteristics, and determining the danger level according to the abnormal situation, includes: Using a first abnormality analysis model to analyze the operating state feature, obtain a number of abnormal conditions, and segment the operating state feature to obtain operating state sub-features corresponding to each of the abnormal conditions; Inputting each of the abnormal conditions and the corresponding operating status sub-feature into a second abnormality analysis model, the second abnormality analysis model outputting the predicted danger level; Among them, the first abnormality analysis model is trained based on the first set of training data, and each training data in the first set of training data includes the operating status sub-features of the first target device and the corresponding abnormal annotation data; the second abnormality analysis model is trained based on the second set of training data, and each training data in the second set of training data includes the operating status sub-features of the first target device, lens damage data and corresponding danger annotation data.

4. The self-protection method of a rail-hanging inspection robot against a harsh environment according to claim 3, characterized in that: When the acquisition distance of the operating status feature is higher than a preset distance, determining the abnormal condition of the first target device based on the operating status feature includes: Determine the abnormality type and the corresponding first abnormality degree of the first target device based on the operating status characteristics; Determine a second target device associated with the abnormal type of the first target device according to a preset association relationship, and predict the abnormality coefficient of the first target device according to the operating state characteristics of the second target device; wherein the second target device is closer to the rail inspection robot than the first target device; A second abnormality degree is calculated using the abnormality degree coefficient and the first abnormality degree, and the abnormal situation is determined according to the abnormality type and the second abnormality degree.

5. The self-protection method of a rail-hanging inspection robot against a harsh environment according to claim 4, characterized in that: Predicting and obtaining an abnormality coefficient of the first target device according to the operating state characteristics of the second target device includes: Inputting the operating state characteristics of the second target device and the preset association relationship into an abnormality coefficient prediction model to obtain the predicted abnormality coefficient; The abnormality coefficient prediction model is obtained by fine-tuning the pre-trained large model.

6. The self-protection method of a rail-hanging inspection robot against a harsh environment according to claim 5, characterized in that: Controlling the high-definition camera to enter a self-protection mode includes: The high-definition camera is controlled to be turned off and a protective cover is used to protect the high-definition camera, and a radar detector is switched to inspect the first target device.

7. A self-protection device for a rail inspection robot against a harsh environment, the device comprising a state extraction module, a hazard level analysis module, and a protection processing module; characterized in that: The state extraction module is used to use a high-definition camera to capture high-definition image data of a first target device in the inspection area, and extract the operating state characteristics of the first target device based on the high-definition image data; The danger level analysis module is used to determine the abnormal situation of the first target device based on the operating state characteristics, and determine the danger level according to the abnormal situation; The protection processing module is used to control the high-definition camera to enter a self-protection mode and switch to using a radar detector to inspect the first target device when the danger level is higher than the target danger level.

8. An electronic device comprising: A memory storing executable program code; A processor coupled to the memory; characterized in that: the processor calls the executable program code stored in the memory to execute the method according to any one of claims 1-6.

9. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is executed.

10. A computer program product, comprising a computer program stored in a computer storage medium, characterized in that: When the computer program is executed by a processor of an electronic device, the method according to any one of claims 1 to 6 is implemented.

Images