Robot navigation method, device, electronic device and storage medium for generator rotor

Abstract

The present application belongs to the field of robot navigation technology, and relates to a robot navigation method, device, electronic device and storage medium for generator rotors. The present application obtains the current posture, environmental image and point cloud data of the robot; determines the next target posture based on the current posture of the robot and a preset discrete posture sequence; determines the environmental QR code based on the environmental image; determines the point cloud position by matching the point cloud data with the preset point cloud map; determines the robot positioning information based on the environmental QR code, point cloud position, laser odometer and point cloud fusion QR code map; and navigates the robot based on the robot positioning information and the next target posture. The present application combines the environmental QR code and the point cloud fusion QR code map, fuses the QR code with the point cloud, and accurately locates the robot, which can realize the navigation of the robot in the generator rotor plant, and use the robot to maintain the rotor to reduce the physical damage caused by manual maintenance.

Description

Technical Field

[0001] The present application belongs to the field of robot navigation technology, and in particular, relates to a robot navigation method, device, electronic device and storage medium for a generator rotor. Background Art

[0002] The rotor of a hydro-turbine generator is a rotating part of a hydro-turbine generator, which is used to generate a magnetic field, convert energy, and transmit torque. When the rotor is working, it rotates at a high speed, and the surface is prone to sludge formed by the mixture of grease and dust in the air that escapes from the generator. This will increase the thermal resistance of the main insulation, causing overheating and discharge, affecting the main insulation, so regular cleaning and maintenance are required. At present, rotor cleaning is mainly done manually, which generally faces the dilemma of high maintenance workload, low maintenance efficiency, and short maintenance time window. In addition, the reagents used for rotor maintenance will cause health hazards.

[0003] The generator plant is an open and long corridor environment, and the generator rotors are distributed in the corridor with geometric symmetry. This environmental geometric feature will cause perception degradation, thereby reducing the reliability of 2D laser point cloud perception and making it difficult for the robot to accurately locate. In addition, since the rotor is suspended at a height higher than the horizontal height of the mobile robot, the 2D laser beam cannot capture the rotor contour shape truthfully, and it is impossible to realize the operation path planning for maintenance tasks.

[0004] Therefore, how to achieve precise robot positioning in the generator rotor plant and perform maintenance on the rotor through the robot to reduce the physical damage caused by manual maintenance has become an urgent problem that needs to be solved. Summary of the invention

[0005] The present application provides a robot navigation method, device, electronic device and storage medium for a generator rotor, which realizes precise robot positioning for a generator rotor plant. Through precise positioning, the robot is used to maintain the rotor to reduce physical damage caused by manual maintenance.

[0006] In a first aspect, an embodiment of the present application provides a robot navigation method for a generator rotor, the robot comprising a camera and a laser odometer, the camera being used to capture an environment image, the laser odometer being used to obtain a point cloud, the method comprising: obtaining a current posture, a current environment image and current point cloud data of the robot; determining a next target posture based on the current posture of the robot and a preset discrete posture sequence; determining a current environment QR code based on the current environment image; determining a current point cloud position by matching the current point cloud data with a preset point cloud map; determining robot positioning information based on the current environment QR code, the current point cloud position, the laser odometer and a point cloud fusion QR code map; and navigating the robot based on the robot positioning information and the next target posture.

[0007] In a possible implementation, before obtaining the robot's current posture, current environment image and current point cloud data, the method includes: laying a QR code on a generator set building, wherein the QR code includes location information, and the generator set building includes multiple generator rotors; obtaining a rotor image, IMU sensor data and a single-frame laser point cloud through the robot; constructing the point cloud fused QR code map based on the QR code, the IMU sensor data and the single-frame laser point cloud; and determining the path planning information for the rotor maintenance operation based on the point cloud fused QR code map.

[0008] In a possible implementation, the construction of a point cloud fused two-dimensional code map based on the two-dimensional code, the IMU sensor data and the single-frame laser point cloud includes: determining a pre-integration value through pre-integration calculation based on the IMU sensor data; performing feature matching based on the single-frame laser point cloud to determine a feature matching result; determining a point cloud registration pose through a preset algorithm based on the pre-integration value and the feature matching result; constructing a multiple pose constraint factor graph based on the point cloud registration pose and the pre-integration value; and embedding the two-dimensional code into the multiple pose constraint factor graph to determine the point cloud fused two-dimensional code map.

[0009] In a possible implementation, the path planning information for the rotor maintenance operation is determined based on the point cloud fused with the two-dimensional code map, including: intercepting a single-frame laser point cloud at a fixed height in the rotor area; determining a cross-sectional projection image of the single-frame laser point cloud at a fixed height in the rotor area; fitting the outer contour of the rotor through a random sampling consistency algorithm based on the cross-sectional projection image; determining a target pose point that surrounds the rotor and maintains a fixed distance based on the rotor outer contour; and determining the path planning information based on the target pose point.

[0010] In a possible implementation, the current point cloud position is determined by matching the current point cloud data with a preset point cloud map, including: performing grid division based on the current point cloud data to determine a point cloud subset; performing Gaussian distribution fitting and intensity feature extraction on the point cloud subset to determine a feature matrix; and determining the current point cloud position based on matching the feature matrix with a preset point cloud map.

[0011] In a possible implementation, the determining of the robot positioning information based on the current environment QR code, the current point cloud position, the laser odometer and the point cloud fusion QR code map includes: determining the camera pose based on the current environment QR code; determining the laser odometer pose based on the current point cloud position and the laser odometer; determining the robot positioning information based on the camera pose, the laser odometer pose and the point cloud fusion QR code map.

[0012] In a possible implementation, the preset algorithm is an iterative closest point algorithm or a normal distribution transformation algorithm; accordingly, based on the pre-integrated value and the feature matching result, the point cloud registration pose is determined by a preset algorithm, including: based on the pre-integrated value and the feature matching result, the point cloud registration of frame-to-frame and frame-to-map is realized by an iterative closest point algorithm or a normal distribution transformation algorithm.

[0013] In a second aspect, an embodiment of the present application provides a robot navigation device for a generator rotor, the device comprising: an acquisition module, used to acquire the robot's current posture, current environment image and current point cloud data; a determination module, used to determine the next target posture based on the robot's current posture and a preset discrete posture sequence; based on the current environment image, determine the current environment QR code; determine the current point cloud position by matching the current point cloud data with a preset point cloud map; the determination module is also used to determine the robot positioning information based on the current environment QR code, the current point cloud position, the laser odometer and the point cloud fusion QR code map; a processing module, used to navigate the robot based on the robot positioning information and the next target posture.

[0014] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method described in the first aspect or any one of the implementation methods thereof is implemented.

[0015] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect or any one of the implementation methods thereof is implemented.

[0016] In a fifth aspect, an embodiment of the present application provides a computer program product. When the computer program product runs on an electronic device, the electronic device executes the method described in the first aspect or any one of the implementation methods thereof.

[0017] The solution provided by this application is: obtaining the current posture of the robot, the current environment image and the current point cloud data; determining the next target posture based on the current posture of the robot and the preset discrete posture sequence; determining the current environment QR code based on the current environment image; determining the current point cloud position by matching the current point cloud data with the preset point cloud map; determining the robot positioning information based on the current environment QR code, the current point cloud position, the laser odometer and the point cloud fusion QR code map; navigating the robot based on the robot positioning information and the next target posture. With respect to the prior art, due to the geometric symmetry of the generator rotor plant, the laser point cloud perception is degraded, making it difficult for the robot to accurately locate. This application combines the environment QR code and the point cloud fusion QR code map, fuses the QR code with the point cloud, and accurately locates the robot, which can realize the navigation of the robot in the generator rotor plant, and use the robot to maintain the rotor to reduce the physical damage caused by manual maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 Schematic diagram of the working environment of the hydro-generator rotor maintenance robot provided in this application;

[0020] Figure 2 A schematic flow chart of a robot navigation method for a generator rotor provided in one embodiment of the present application;

[0021] Figure 3 A schematic diagram of rotor outer contour recognition and safe operation path generation provided in this application;

[0022] Figure 4 A structural block diagram of a robot navigation device for a generator rotor provided in one embodiment of the present application;

[0023] Figure 5 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0024] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0025] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0026] It should also be understood that the term “and / or” used in the specification and appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0027] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0028] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0029] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0030] The rotor of a hydro-turbine generator is a rotating part of a hydro-turbine generator, used to generate a magnetic field, convert energy, and transmit torque. When the rotor is working, it is in a high-speed rotating state, and the surface is prone to sludge formed by the mixture of grease and dust in the air that escape from the generator. This will increase the thermal resistance of the main insulation, resulting in overheating and discharge, affecting the main insulation, so regular cleaning and maintenance are required. At present, rotor cleaning is mainly done manually, and generally faces the dilemma of high maintenance workload, low maintenance efficiency, and short maintenance time window. In addition, the reagents used for rotor maintenance will cause health hazards, so the development and application of intelligent maintenance robots for generator rotors are of great significance.

[0031] Autonomous mobile navigation is the core technology of the rotor maintenance robot of the hydro-turbine generator set. It supports the robot's autonomous movement around the rotor to achieve the transfer of onboard maintenance equipment and traversal of work points. Autonomous mobile navigation technology includes three aspects: map construction, work path planning, and positioning and mobile control. At present, there will be different technical developments in indoor cleaning robots / service robots / industrial AGVs according to scene characteristics. Existing mobile navigation technology solutions for indoor robots generally unify map construction and mobile positioning under the framework of two-dimensional laser SLAM, and identify obstacles by constructing a 2D occupancy grid map, thereby realizing global path planning and autonomous movement for the target point. The generator set plant is an open and long corridor environment (such as Figure 1 The generator rotors are distributed in the corridors. When the generator is working normally, the rotors are hidden under the ground. The rotors that need to be maintained and cleaned are lifted to the ground through the gantry and placed in the air (as shown in Fig. Figure 1 b). In this scenario, the overall environment is like an open corridor, while the rotor has geometric symmetry. This environmental geometric feature may cause perception degradation, thereby reducing the reliability of 2D laser point cloud perception; secondly, irregular objects such as cargo and tool piles are randomly placed in the surrounding environment, and workers will also randomly walk within the laser perception range. This messy and dynamic scene feature will reduce the accuracy of laser point cloud perception; in addition, since the rotor is suspended at a height higher than the horizontal height of the mobile robot, the 2D laser beam cannot capture the rotor contour shape truthfully, and the operation path planning for maintenance tasks cannot be achieved. Finally, during the operation, the robot's traversable area has environmental challenges such as narrow space and uncertain road conditions. In order to ensure the safe and collision-free navigation of high-load mobile robots, the real-time and accuracy of navigation positioning must have good performance. In summary, the existing 2D laser SLAM solution cannot meet the above requirements.

[0032] Figure 2 A schematic flow chart of a robot navigation method for a generator rotor provided in an embodiment of the present application, the method comprising:

[0033] S210, obtaining the robot's current posture, current environment image and current point cloud data.

[0034] Exemplarily, when the robot is working, it obtains the current posture of the robot, the current environment image taken by the camera, the environment image includes a QR code of the current area, and the current point cloud data.

[0035] In one possible implementation, before obtaining the robot's current posture, current environment image and current point cloud data, the method includes: laying a QR code on a generator set building, the QR code including location information, and the generator set building including multiple generator rotors; obtaining a rotor image, IMU sensor data and a single-frame laser point cloud through the robot; constructing a point cloud fused QR code map based on the QR code, inertial measurement unit (IMU) sensor data and a single-frame laser point cloud; and determining the path planning information for the rotor maintenance operation based on the point cloud fused QR code map.

[0036] For example, the staff will pre-arrange a series of QR codes in the generator plant. These QR codes serve as artificial landmarks of known locations and can provide accurate location references for the robot. When the robot scans these QR codes through the environment image, it can obtain its own relative position information through decoding, thereby assisting positioning and navigation.

[0037] Next, the staff controls the robot (such as through a remote control or a preset program) to move in the environment, especially around areas of special concern (such as rotors). During the movement, the robot collects a variety of data, including images (for visual recognition and environmental understanding), single-frame laser point clouds (a collection of points in three-dimensional space, used to build three-dimensional maps), and IMU (inertial measurement unit) sensor data (such as acceleration, angular velocity, etc., used to assist positioning and attitude estimation).

[0038] Furthermore, the collected single-frame laser point cloud and IMU sensor data are used to construct a 3D map, and the location information of the QR code is added to the map as a known point. By fusing these two types of data, a composite map containing both precise location information (from the QR code) and detailed environmental structure (from the point cloud) can be generated, namely, a point cloud fused QR code map.

[0039] Furthermore, the constructed point cloud map is processed to extract rotor information, path planning information, and navigation positioning information from the three-dimensional disordered, discrete point cloud map. After obtaining the point cloud fusion QR code map, further data processing is required. This includes extracting key information from the three-dimensional point cloud map, such as the specific position, shape, size, etc. of the rotor, as well as identifying obstacles and walkable areas in the environment. This information includes path planning information (determining the best path from the starting point to the end point) and navigation positioning information (determining its own position in real time and adjusting the direction of movement).

[0040] In a possible implementation, a point cloud fused QR code map is constructed based on the QR code, IMU sensor data and single-frame laser point cloud, including: determining a pre-integration value through pre-integration calculation based on the IMU sensor data; performing feature matching based on the single-frame laser point cloud to determine the feature matching result; determining the point cloud registration pose through a preset algorithm based on the pre-integration value and the feature matching result; constructing a multiple pose constraint factor graph based on the point cloud registration pose and the pre-integration value; and embedding the QR code into the multiple pose constraint factor graph to determine the point cloud fused QR code map.

[0041] For example, in the construction of point cloud fusion QR code map, the fusion of laser point cloud and artificial landmarks is mainly realized by relying on the framework of laser SLAM. Specifically, IMU (inertial measurement unit) data can provide information such as acceleration and angular velocity of the robot in real time. The pre-integration calculation is to process these raw data and determine the pre-integration value to estimate the relative posture change of the robot between two consecutive time points. This helps to fill the gap in posture estimation through IMU data when the laser point cloud data is sparse or missing.

[0042] Extract geometric features of single-frame laser point cloud and extract global invariant descriptors, match the extracted features, and determine the feature matching results. Laser sensors can generate point cloud data of the environment, which contains rich three-dimensional geometric information. By extracting geometric features of point clouds (such as edges, planes, etc.) and global invariant descriptors (such as FPFH, SHOT, etc.), the robot can identify specific structures or objects in the environment and use them for subsequent map construction and pose estimation.

[0043] In the pose estimation stage, the robot uses the pre-integrated values ​​(from IMU) and the feature matching results between point cloud frames to construct a multi-pose constraint factor graph through a preset algorithm.

[0044] The registered pose and pre-integration values ​​are constructed into a factor graph with multiple pose constraints. A factor graph is a graph structure used to represent and optimize complex constraint relationships. In this factor graph, each node represents a pose (or state) and each edge represents a constraint (such as point cloud registration results, IMU pre-integration results, etc.). Factor graph optimization is essentially a nonlinear least squares problem, whose goal is to find a set of poses (or states) that satisfy all constraints so that the overall error is minimized. The optimized pose is the estimated value of the laser odometer that satisfies the geometric consistency constraint.

[0045] The robot uses the images collected by the camera to identify the QR code in the environment and embed it into the multiple pose constraint factor graph to determine the point cloud fusion QR code map.

[0046] Exemplarily, the preset algorithm is an iterative closest point algorithm or a normal distribution transformation algorithm. Based on the pre-integration value and feature matching results, the iterative closest point algorithm or the normal distribution transformation algorithm is used to achieve frame-to-frame and frame-to-map point cloud registration to estimate the current position of the robot.

[0047] In one possible implementation, based on the point cloud fused with the QR code map, the path planning information for the rotor maintenance operation is determined, including: intercepting a single-frame laser point cloud at a fixed height in the rotor area; determining a cross-sectional projection image of the single-frame laser point cloud at a fixed height in the rotor area; based on the cross-sectional projection image, fitting the outer contour of the rotor through a random sampling consistency algorithm; based on the outer contour of the rotor, determining the target posture points that surround the rotor and maintain a fixed distance; based on the target posture points, determining the path planning information.

[0048] Specifically, first, in order to reduce unnecessary data interference and improve processing efficiency, the laser point cloud at a fixed height (such as Figure 3 This is because in real-world scenarios, the laser scanner will obtain point cloud data within the entire scanning range, but only part of the data is related to the rotor. By intercepting the point cloud at a specific height, interference from other objects such as the ground and walls can be eliminated.

[0049] Next, the intercepted 3D laser point cloud is projected onto a horizontal plane. This step is to simplify the point cloud data in 3D space into data on a 2D plane for subsequent processing. Since the rotor is usually placed horizontally, projecting it onto a horizontal plane can retain the main shape information of the rotor while reducing processing complexity.

[0050] In the projected two-dimensional point cloud, the Random Sample Consensus (RANSAC) algorithm is used to perform model-based rotor contour fitting. RANSAC is a robust parameter estimation method that can handle data sets containing a lot of noise and outliers. Through the RANSAC algorithm, the outer contour of the rotor can be fitted and its geometric information (such as radius) and position information (coordinates on the two-dimensional plane) can be determined.

[0051] Furthermore, with the geometric information and position information of the rotor's outer contour, a series of target pose points around the rotor and maintaining a fixed distance can be further generated. This usually involves expanding the rotor's outer contour (i.e., expanding outward a certain distance based on the original contour), and then discretely sampling on this expanded contour to obtain a series of points as target pose points. These points will serve as the basis for path planning to guide robots or automated equipment on how to perform maintenance operations around the rotor.

[0052] S220, based on the current posture of the robot and the preset discrete posture sequence, determine the next target posture; based on the current environment image, determine the current environment QR code; by matching the current point cloud data with the preset point cloud map, determine the current point cloud position.

[0053] Exemplarily, based on the current posture, the robot will access the discrete posture sequence (i.e., the result of path planning) stored or received through the network. This sequence contains a series of positions and postures (postures) that the robot needs to reach. The robot will parse this sequence to determine the next target posture to be reached. According to the current environment image, the current environment QR code is determined to obtain the rough position information of the current position relative to the QR code map. At the same time, the robot will extract global feature descriptors (such as FPFH, SHOT, etc.) from the point cloud data, and match these descriptors with the pre-built point cloud map to obtain more accurate global position information. Because it is not associated with the QR code information, the position information determined at this time may be multiple.

[0054] In one possible implementation, the current point cloud position is determined by matching the current point cloud data with a preset point cloud map, including: performing grid division based on the current point cloud data to determine a point cloud subset; performing Gaussian distribution fitting and intensity feature extraction on the point cloud subset to determine a feature matrix; and determining the current point cloud position based on matching the feature matrix with a preset point cloud map.

[0055] Specifically, grid division is performed based on the current point cloud data to determine the point cloud subset, and then Gaussian distribution fitting and intensity feature extraction are used to construct a feature matrix with perspective, rotation invariance and feature uniqueness. Using these feature matrices and the geometric information of the point cloud map, the current point cloud frame can be aligned to determine the current point cloud position.

[0056] S230, determining the robot positioning information based on the current environment QR code, the current point cloud position, the laser odometer and the point cloud fusion QR code map.

[0057] In one possible implementation, the robot positioning information is determined based on the current environment QR code, the current point cloud position, the laser odometry and the point cloud fusion QR code map, including: determining the camera pose based on the current environment QR code; determining the laser odometry pose based on the current point cloud position and the laser odometry; determining the robot positioning information based on the camera pose, the laser odometry pose and the point cloud fusion QR code map.

[0058] Exemplarily, the location information of the QR code is used to visually identify the QR code and implement camera extrinsic parameter inference based on the camera and multi-view constraints to obtain the estimated camera posture.

[0059] By using the current point cloud position to realize the current point cloud frame registration, the current laser radar (laser odometer) sensing posture can be inferred.

[0060] Furthermore, the robot positioning information is determined based on the camera pose, laser odometer pose and point cloud fusion QR code map.

[0061] S240, navigating the robot based on the robot positioning information and the next target position.

[0062] The technical solution provided by the present application obtains the current posture of the robot, the current environment image and the current point cloud data; determines the next target posture based on the current posture of the robot and the preset discrete posture sequence; determines the current environment QR code based on the current environment image; determines the current point cloud position by matching the current point cloud data with the preset point cloud map; determines the robot positioning information based on the current environment QR code, the current point cloud position, the laser odometer and the point cloud fusion QR code map; and navigates the robot based on the robot positioning information and the next target posture. With respect to the prior art, due to the geometric symmetry of the generator rotor plant, the laser point cloud perception is degraded, making it difficult for the robot to accurately locate. The present application combines the environment QR code and the point cloud fusion QR code map, fuses the QR code with the point cloud, and accurately locates the robot, which can realize the navigation of the robot in the generator rotor plant and use the robot to maintain the rotor to reduce the physical damage caused by manual maintenance.

[0063] Figure 4 This is a structural block diagram of a robot navigation device for a generator rotor provided in one embodiment of the present application. For ease of explanation, only the parts related to the embodiment of the present application are shown. Figure 4The robot navigation device 400 for a generator rotor includes an acquisition module 401 , a determination module 402 , and a processing module 403 .

[0064] In one implementation, the apparatus 400 may be used to implement the above Figure 2 For example, the acquisition module 401 is used to implement S210, the determination module 402 is used to implement S220 and S230, and the processing module 403 is used to implement S240.

[0065] The technical solution provided by the present application obtains the current posture of the robot, the current environment image and the current point cloud data; determines the next target posture based on the current posture of the robot and the preset discrete posture sequence; determines the current environment QR code based on the current environment image; determines the current point cloud position by matching the current point cloud data with the preset point cloud map; determines the robot positioning information based on the current environment QR code, the current point cloud position, the laser odometer and the point cloud fusion QR code map; and navigates the robot based on the robot positioning information and the next target posture. With respect to the prior art, due to the geometric symmetry of the generator rotor plant, the laser point cloud perception is degraded, making it difficult for the robot to accurately locate. The present application combines the environment QR code and the point cloud fusion QR code map, fuses the QR code with the point cloud, and accurately locates the robot, which can realize the navigation of the robot in the generator rotor plant and use the robot to maintain the rotor to reduce the physical damage caused by manual maintenance.

[0066] Figure 5 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 5 As shown, the electronic device 5 of this embodiment includes: at least one processor 50 ( Figure 5 Only one is shown in the figure) a processor, a memory 51, and a computer program 52 stored in the memory 51 and executable on the at least one processor 50, and when the processor 50 executes the computer program 52, the steps in any of the above method embodiments are implemented.

[0067] The electronic device 5 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The electronic device may include, but is not limited to, a processor 50 and a memory 51. Those skilled in the art will appreciate that Figure 5 It is only an example of the electronic device 5 and does not constitute a limitation on the electronic device 5. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include input and output devices, network access devices, etc.

[0068] The processor 50 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0069] In some embodiments, the memory 51 may be an internal storage unit of the electronic device 5, such as a hard disk or memory of the electronic device 5. In other embodiments, the memory 51 may also be an external storage device of the electronic device 5, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 5. Further, the memory 51 may also include both an internal storage unit and an external storage device of the electronic device 5. The memory 51 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program. The memory 51 may also be used to temporarily store data that has been output or is to be output.

[0070] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0071] An embodiment of the present application also provides an electronic device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor implements the steps of any of the above-mentioned method embodiments when executing the computer program.

[0072] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0073] An embodiment of the present application provides a computer program product. When the computer program product runs on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0074] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the camera / electronic device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electric carrier signal, a telecommunication signal, and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0075] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0076] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0077] In the embodiments provided in the present application, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0078] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0079] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A robot navigation method for a generator rotor, characterized in that: The robot includes a camera and a laser odometer, the camera is used to capture an environment image, and the laser odometer is used to obtain a point cloud. The method includes: Obtaining the current posture of the robot, the current environment image and the current point cloud data; Determine the next target posture based on the current posture of the robot and the preset discrete posture sequence; determine the current environment QR code based on the current environment image; determine the current point cloud position by matching the current point cloud data with the preset point cloud map; Determine the robot positioning information based on the current environment QR code, the current point cloud position, the laser odometer and the point cloud fusion QR code map; The robot is navigated based on the robot positioning information and the next target position.

2. The method according to claim 1, characterized in that Before obtaining the current posture of the robot, the current environment image and the current point cloud data, the method includes: Paving a generator plant with a QR code, the QR code including location information, the generator plant including a plurality of generator rotors; Obtain rotor images, IMU sensor data and single-frame laser point clouds through the robot; Based on the two-dimensional code, the IMU sensor data and the single-frame laser point cloud, construct the point cloud fusion two-dimensional code map; Based on the point cloud fused QR code map, path planning information for the rotor maintenance operation is determined.

3. The method according to claim 2, characterized in that The step of constructing a point cloud fused two-dimensional code map based on the two-dimensional code, the IMU sensing data and the single-frame laser point cloud comprises: Based on the IMU sensing data, a pre-integration value is determined through pre-integration calculation; based on the single-frame laser point cloud, feature matching is performed to determine a feature matching result; Based on the pre-integrated value and the feature matching result, a point cloud registration pose is determined by a preset algorithm; Constructing a multiple pose constraint factor graph based on the point cloud registration pose and the pre-integrated value; The two-dimensional code is embedded in the multiple posture constraint factor graph to determine the point cloud fusion two-dimensional code map.

4. The method according to claim 2, characterized in that: Determining the path planning information of the rotor maintenance operation based on the point cloud fusion two-dimensional code map includes: Capture a single-frame laser point cloud at a fixed height in the rotor area; Determine a cross-sectional projection image of a single-frame laser point cloud at a fixed height of the rotor region; Based on the cross-sectional projection image, fitting the outer contour of the rotor by a random sampling consistency algorithm; Based on the outer contour of the rotor, determining a target posture point surrounding the rotor and maintaining a fixed distance; Based on the target pose point, the path planning information is determined.

5. The method according to claim 1, characterized in that The determining the current point cloud position by matching the current point cloud data with a preset point cloud map includes: Performing grid division based on the current point cloud data to determine a point cloud subset; Performing Gaussian distribution fitting and intensity feature extraction on the point cloud subset to determine a feature matrix; Based on matching the feature matrix with a preset point cloud map, the current point cloud position is determined.

6. The method according to claim 5, characterized in that The determining of the robot positioning information based on the current environment two-dimensional code, the current point cloud position, the laser odometer and the point cloud fusion two-dimensional code map includes: Determine the camera position based on the current environment QR code; Determine the laser odometry pose based on the current point cloud position and the laser odometry; The robot positioning information is determined based on the camera posture, the laser odometer posture and the point cloud fusion QR code map.

7. The method according to claim 3, characterized in that The preset algorithm is an iterative closest point algorithm or a normal distribution transformation algorithm; Accordingly, the step of determining the point cloud registration pose based on the pre-integrated value and the feature matching result by a preset algorithm includes: Based on the pre-integration value and the feature matching result, frame-to-frame and frame-to-map point cloud registration is achieved through an iterative nearest point algorithm or a normal distribution transformation algorithm.

8. A robot navigation device for a generator rotor, characterized in that: The device comprises: The acquisition module is used to obtain the robot's current posture, current environment image and current point cloud data; A determination module is used to determine the next target posture based on the current posture of the robot and a preset discrete posture sequence; determine the current environment QR code based on the current environment image; and determine the current point cloud position by matching the current point cloud data with a preset point cloud map; The determination module is further used to determine the robot positioning information based on the current environment QR code, the current point cloud position, the laser odometer and the point cloud fusion QR code map; A processing module is used to navigate the robot based on the robot positioning information and the next target posture.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images