In-situ inspection robot, in-situ inspection method, and storage medium

Abstract

The application discloses an in-situ detection robot, an in-situ detection method and a storage medium, and belongs to the technical field of in-situ overhauling of water turbines. The in-situ detection robot comprises a movable carrier provided on a bearing surface below a runner of a water turbine to be detected; a multi-stage spatial positioning mechanism provided on the movable carrier; and a composite detection terminal provided at the tail end of the multi-stage spatial positioning mechanism and comprising a three-dimensional topography scanner; and a hierarchical positioning controller in communication connection with the movable carrier and the multi-stage spatial positioning mechanism, respectively. The in-situ detection robot eliminates the risk of secondary damage to the runner caused by the robot falling. The hierarchical positioning controller implements a two-stage coupling strategy combining global coarse positioning and end precise positioning, realizes full-automatic positioning and detection, and the whole process does not need manual intervention, so that the standardization and automation of the detection process are realized, the detection result is not affected by human factors, the reliability is high, and the detection positioning precision is improved.

Description

Technical Field

[0001] This application relates to the field of water turbine maintenance technology, specifically to an in-situ inspection robot, an in-situ inspection method, and a storage medium. Background Technology

[0002] Impulse turbines are the core energy conversion equipment in high-head hydropower stations. Their runners and buckets operate under the impact of high-speed water flow, enduring alternating stress, silt erosion, and cavitation over long periods. Under high-head, high-silt conditions, the bucket surfaces are highly susceptible to defects such as abrasion, cavitation, and cracks. If these defects are not detected and addressed promptly, they will directly affect the unit's power generation efficiency and, in severe cases, may even lead to bucket failure, threatening the safe operation of the hydropower station. Therefore, defect detection of the runner during unit shutdown and maintenance is a crucial step in ensuring the safe and stable operation of the hydropower station. Current technologies involve in-situ testing using handheld instruments or offline testing after runner disassembly, requiring manual high-altitude work after unit shutdown, which carries risks. Alternatively, using a wall-climbing robot attached to the runner for testing carries the risk of the robot falling and damaging the runner. Summary of the Invention

[0003] This application provides an in-situ inspection robot, an in-situ inspection method, and a storage medium, aiming to solve the technical problem that the turbine runner maintenance process requires manual high-altitude operation after the unit is shut down, or that there is a risk of falling when the inspection is carried out by a wall-climbing robot adsorbed on the runner.

[0004] On one hand, this application provides an in-situ inspection robot, which includes: a movable platform disposed on a bearing surface below the turbine runner to be tested; a multi-level spatial positioning mechanism disposed on the movable platform; a composite inspection terminal disposed at the end of the multi-level spatial positioning mechanism, including a three-dimensional topography scanner; and a hierarchical positioning controller, which is communicatively connected to the movable platform and the multi-level spatial positioning mechanism respectively; wherein, the hierarchical positioning controller is used to: control the movable platform to perform global coarse positioning based on a preset path and environmental visual features to determine the current position of the movable platform on the bearing surface; and based on the current position... The first pose deviation between the composite detection terminal and the preset detection window is used to control the movable platform to move to the preset detection window. After the movable platform reaches the preset detection window, the multi-level spatial positioning mechanism is controlled to perform end-effector precision positioning based on the point cloud data of the turbine runner surface to be tested collected by the three-dimensional topography scanner, and to determine the second pose deviation between the current pose of the composite detection terminal and the turbine runner surface to be tested. Based on the second pose deviation, the multi-level spatial positioning mechanism is controlled to move so that the composite detection terminal is aligned with the turbine runner surface to be tested. The composite detection terminal is then controlled to perform in-situ detection of the turbine runner to be tested.

[0005] In some embodiments, the multi-level spatial positioning mechanism includes: a multi-level lifting platform, vertically disposed on the movable platform; a multi-degree-of-freedom robotic arm, disposed at the top of the multi-level lifting platform; and a composite detection terminal disposed at the end of the multi-degree-of-freedom robotic arm. The hierarchical positioning controller is further configured to: control the multi-level lifting platform to vertically lift the composite detection terminal, so that the composite detection terminal moves to the height of the turbine runner under test; and control the multi-degree-of-freedom robotic arm to adjust the current pose of the composite detection terminal, so that the composite detection terminal is aligned with the surface of the turbine runner under test.

[0006] In some embodiments, the multi-degree-of-freedom robotic arm includes multiple joints; wherein, the hierarchical positioning controller is further configured to: drive each joint in the position control group to perform position compensation one by one according to a preset sequence, and lock the joint in the position control group that has completed the position compensation; after all joints in the position control group are locked, drive each joint in the attitude control group to adjust the attitude of the composite detection terminal one by one, so that the composite detection terminal is aligned with the surface of the turbine runner to be tested.

[0007] In some embodiments, the multi-stage lifting platform is equipped with an electronic soft limit device, which includes an absolute encoder and a laser ranging sensor. The hierarchical positioning controller is further configured to: control the absolute encoder to provide feedback on the lifting position of the multi-stage lifting platform to determine whether the multi-stage lifting platform has reached a preset travel boundary; control the laser ranging sensor to detect the distance between the composite detection terminal and the surface of the turbine runner under test to determine whether the distance is less than a preset safety threshold; and restrict the operation of the multi-stage lifting platform's drive when the lifting position reaches the preset travel boundary or the distance is less than the preset safety threshold.

[0008] In some embodiments, the composite detection terminal further includes: a visual imaging module; both the three-dimensional topography scanner and the visual imaging module are disposed at the end of the multi-level spatial positioning mechanism; wherein, the hierarchical positioning controller is further configured to: monitor the measurement confidence level of the point cloud data acquired by the three-dimensional topography scanner; when the measurement confidence level is lower than a preset threshold, switch to a backup positioning mode guided by the visual imaging module; in the backup positioning mode, acquire the deviation between the current image features acquired by the visual imaging module and the target image features; convert the deviation into joint speed control commands for the multi-level spatial positioning mechanism; and based on the joint speed control commands, drive the movement of each joint in the multi-level spatial positioning mechanism so that the composite detection terminal is aligned with the surface of the turbine runner to be tested.

[0009] In some embodiments, the robot further includes a dual-redundant energy supply module, which includes: a primary energy source configured to be connected to an external power source via a drag cable; and a backup energy source configured as an onboard energy storage battery. The hierarchical positioning controller is further configured to: determine the minimum backup power required for the backup energy source based on the motor output torque and angular velocity of each joint in the multi-level spatial positioning mechanism, a preset safe return time, the power consumption of the robot's control system, and a preset safety margin coefficient.

[0010] On the other hand, this application embodiment also provides an in-situ detection method, applicable to any of the in-situ detection robots provided in this application embodiment; the in-situ detection method includes: controlling the movable platform to perform global coarse localization based on a preset path and environmental visual features to determine the current position of the movable platform on the bearing surface; controlling the movable platform to move to the preset detection window based on a first pose deviation between the current position and a preset detection window; after the movable platform reaches the preset detection window, controlling the multi-level spatial positioning mechanism to perform end-effector fine localization based on the point cloud data of the turbine runner surface to be tested collected by the three-dimensional topography scanner to determine a second pose deviation between the current pose of the composite detection terminal and the turbine runner surface to be tested; controlling the multi-level spatial positioning mechanism to move based on the second pose deviation so that the composite detection terminal is aligned with the turbine runner surface to be tested; and controlling the composite detection terminal to perform in-situ detection on the turbine runner to be tested.

[0011] In some embodiments, controlling the multi-level spatial positioning mechanism to perform end-effector precision positioning based on the point cloud data of the turbine runner surface to be tested acquired by the three-dimensional topography scanner, and determining the second pose deviation between the current pose of the composite detection terminal and the turbine runner surface to be tested, includes: determining the current pose of the composite detection terminal in the global coordinate system by positive kinematic mapping according to the current pose state of each moving component in the multi-level spatial positioning mechanism; wherein the current pose state is determined by the target extension of the multi-level lifting platform, the circumferential position of the movable platform, and the current joint angle of the multi-degree-of-freedom robotic arm; the target extension is determined according to at least one of the following: the platform height of the movable platform, the vertical projection height of the multi-degree-of-freedom robotic arm under the current joint angle, and the distance between the composite detection terminal and the turbine runner surface to be tested; and calculating the second pose deviation between the current pose and the desired pose of the turbine runner surface to be tested based on the current pose and the point cloud data.

[0012] In some embodiments, controlling the movement of the multi-level spatial positioning mechanism based on the second pose deviation to align the composite detection terminal with the surface of the turbine runner under test includes: determining a position compensation amount based on a second position deviation in the second pose deviation; determining an attitude compensation amount based on a second attitude deviation in the second pose deviation; solving for the compensation amounts of each joint in the position control group and the attitude control group of the multi-level spatial positioning mechanism respectively using inverse kinematics based on the position compensation amount and the attitude compensation amount; generating target position commands for each joint in the multi-level spatial positioning mechanism respectively based on the compensation amounts of each joint in the position control group and the attitude control group; and driving each joint in the multi-level spatial positioning mechanism to move according to the target position commands to align the composite detection terminal with the surface of the turbine runner under test.

[0013] In some embodiments, driving each joint in the multi-level spatial positioning mechanism to move according to the target position command so that the composite detection terminal is aligned with the surface of the turbine runner under test includes: driving each joint in the position control group to perform position compensation one by one according to a preset sequence, and locking the joint in the position control group that has completed the position compensation; after all joints in the position control group are locked, driving each joint in the attitude control group to adjust the attitude of the composite detection terminal one by one so that the composite detection terminal is aligned with the surface of the turbine runner under test.

[0014] In some embodiments, the method further includes: monitoring the measurement confidence level of the point cloud data acquired by the three-dimensional topography scanner; switching to a backup positioning mode guided by the visual imaging module when the measurement confidence level is lower than a preset threshold; in the backup positioning mode, acquiring the deviation between the current image features acquired by the visual imaging module and the target image features; converting the deviation into joint speed control commands for the multi-level spatial positioning mechanism; and driving the movement of each joint in the multi-level spatial positioning mechanism based on the joint speed control commands, so that the composite detection terminal is aligned with the surface of the turbine runner to be tested.

[0015] On the other hand, embodiments of this application also provide a computer program product, including a computer program or instructions, which, when executed by a processor, implement the steps in any of the in-situ detection methods provided in embodiments of this application.

[0016] On the other hand, embodiments of this application also provide a computer-readable storage medium storing a computer program or instructions thereon, including a computer program or instructions that, when executed by a processor, implement the steps in any of the in-situ detection methods provided in embodiments of this application.

[0017] The in-situ inspection robot provided in this application embodiment has a movable platform that is always positioned on the bearing surface below the turbine runner under test (such as the bottom floor of the turbine chamber). The composite inspection terminal is lifted from the bearing surface upwards to the runner surface for inspection. Throughout the inspection process, the movable platform maintains stable contact with the bearing surface, eliminating the risk of falling and causing secondary damage to the runner. This also shortens the inspection cycle and reduces maintenance costs. A two-level coupling strategy combining global coarse positioning and end-effector fine positioning is implemented through a hierarchical positioning controller, improving the inspection and positioning accuracy and ensuring the consistency and comparability of the inspection data. A highly reliable global positioning is achieved through a global coarse positioning strategy that coordinates a preset path with environmental visual features. This application embodiment achieves fully automatic positioning and inspection through a hierarchical positioning mechanism, requiring no manual intervention throughout the process. This standardizes and automates the inspection process, and the inspection results are unaffected by human factors, resulting in high reliability. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the structure of an in-situ inspection robot provided in an embodiment of this application; Figure 2 A flowchart illustrating an in-situ detection method for a water turbine provided in an embodiment of this application; Figure 3 This is a schematic diagram of another in-situ inspection robot provided in an embodiment of this application; Figure 4 This is a schematic diagram of another in-situ inspection robot provided in an embodiment of this application; Figure 5 A flowchart illustrating a precise end-effector positioning method provided in this application embodiment; Figure 6 This is a flowchart illustrating another in-situ detection method provided in an embodiment of this application.

[0020] Explanation of reference numerals in the attached figures: 10. In-situ inspection robot; 110. Movable platform; 120. Multi-level spatial positioning mechanism; 121. Multi-level lifting platform; 122. Multi-degree-of-freedom robotic arm; 130. Composite inspection terminal; 140. Hierarchical positioning controller; 150. Dual redundant energy supply module; 151. Primary energy source; 152. Backup energy source; 20. Turbine runner under test; 30. Bearing surface. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] In the description of this application, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0023] The use of "applies to" or "configured to" in this application implies open and inclusive language, which does not exclude the applicability to or configuration to devices performing additional tasks or steps. Additionally, the use of "based on" implies openness and inclusivity, because processes, steps, calculations, or other actions "based on" one or more of the stated conditions or values ​​may in practice be based on additional conditions or values ​​beyond those stated.

[0024] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0025] In related technologies, the following methods are mainly used for the inspection of impulse turbine runners: I. Manual Scaffolding Inspection. This method involves erecting scaffolding around the turbine runner, where inspectors, armed with testing instruments (such as ultrasonic flaw detectors, hardness testers, and visual inspection mirrors), climb to various parts of the runner to conduct point-by-point inspections. This method has the following drawbacks: the erection and dismantling of scaffolding is labor-intensive and time-consuming, often requiring several days to weeks for a single unit's inspection, severely impacting the unit's availability; inspectors must work in high-altitude, confined, and damp environments, posing serious safety risks such as falls and electric shocks; and the inspection results are highly dependent on the operator's experience and skill level, resulting in poor consistency between results from different personnel and batches, making it difficult to establish standardized inspection data.

[0026] II. Offline Testing After Disassembly. This method involves removing the runner from the main shaft and hoisting it to the maintenance station for detailed testing. While this method can obtain relatively comprehensive test data, it has the following drawbacks: Disassembly and reinstallation of the runner involve precision processes such as shaft alignment and bolt tightening; improper operation may cause secondary damage to the runner or main shaft. Large impact runners can weigh tens of tons, requiring large lifting equipment and transport vehicles for disassembly, hoisting, and transfer, resulting in high logistics costs and long cycles. The unit cannot operate at all during disassembly and testing; for hydropower stations with high grid peak-shaving pressure, the downtime losses are enormous.

[0027] III. Wall-Climbing Robot Inspection. This method utilizes the attraction force provided by permanent magnets or electromagnets to enable a wall-climbing robot to crawl on the surface of a rotating wheel, carrying detection sensors for defect detection. However, this method faces many challenges in practical applications: the rotating wheel's water tank is a complex three-dimensional curved surface, and the magnetic attraction stability of the wall-climbing robot on the curved surface is poor, making it difficult to achieve full-coverage inspection; the rotating wheel surface, after long-term operation, is covered with mud, sand, and rust products, which seriously affects the reliability of the magnetic attraction force, and the robot is at risk of falling from a height, which would cause serious secondary damage to the rotating wheel; the wall-climbing robot is limited by the magnetic attraction force, making it difficult to carry large or heavy inspection equipment, thus limiting the diversity of inspection methods.

[0028] Based on this, the embodiments of this application provide a complete technical solution that enables the detection terminal to achieve high-precision automatic alignment and in-situ detection of the turbine runner's water bucket surface in a narrow and complex water turbine chamber environment without disassembling the turbine runner, without manual high-altitude operations, and without relying on magnetic adsorption.

[0029] Figure 1 This is a schematic diagram of an in-situ inspection robot provided in an embodiment of this application. Please refer to [link / reference]. Figure 1 The in-situ inspection robot 10 includes: a movable platform 110, a multi-level spatial positioning mechanism 120, a composite inspection terminal 130, and a hierarchical positioning controller 140.

[0030] A movable platform 110 is mounted on a bearing surface 30 below the turbine runner 20 under test, and is used to carry the multi-level spatial positioning mechanism 120 and the composite testing terminal 130 to move autonomously on the bearing surface 30. The bearing surface 30 refers to any surface located below the turbine runner 20 under test that can support the movable platform 110 and allow it to move, including but not limited to the floor of the turbine room, maintenance platforms, and test benches. It should be understood that the movable platform 110 is essentially the chassis and walking assembly of the entire in-situ testing robot 10, including but not limited to various forms such as multi-legged robots, tracked vehicles, humanoid robots, wheeled robots, and on-orbit vehicles. In some embodiments, the movable platform 110 may include a drive module and a walking mechanism. The drive module provides the walking power for the movable platform 110, and typically includes a drive motor and its associated reducer and transmission mechanism. The walking mechanism is in direct contact with the bearing surface 30 and, driven by the drive module, propels the entire robot to move on the bearing surface 30. In some embodiments, the movable platform 110 can be connected to other functional modules above it via detachable interfaces such as anti-loosening bolts, so that the appropriate functional modules can be replaced according to the environmental conditions of different bearing surfaces 30.

[0031] A multi-level spatial positioning mechanism 120 is mounted on a movable platform 110 to move the composite detection terminal 130 to the height of the turbine runner 20 under test and to adjust the spatial orientation of the composite detection terminal 130. The composite detection terminal 130, located at the end of the multi-level spatial positioning mechanism 120, includes a three-dimensional topography scanner (not shown) for in-situ detection of the surface of the turbine runner 20 under test. Exemplarily, the three-dimensional topography scanner can be a laser three-dimensional scanner or a structured light three-dimensional scanner. A hierarchical positioning controller 140 is communicatively connected to both the movable platform 110 and the multi-level spatial positioning mechanism 120, and is used to control the coordinated movement of the movable platform 110 and the multi-level spatial positioning mechanism 120.

[0032] Figure 2 This is a flowchart illustrating an in-situ monitoring method for a water turbine provided in an embodiment of this application. Please refer to... Figure 1 and Figure 2In actual operation, the hierarchical positioning controller 140 uses a two-level coupling positioning strategy that combines global coarse positioning and end fine positioning to guide the composite detection terminal 130 from a large area of ​​space on the bearing surface 30 to the millimeter-level detection position on the surface of the turbine runner 20 to be tested. The hierarchical positioning controller 140 is specifically used for: controlling the movable platform 110 to perform global coarse positioning based on a preset path and environmental visual features to determine the current position of the movable platform 110 on the bearing surface 30; controlling the movable platform 110 to move to the preset detection window based on the first pose deviation between the current position and the preset detection window; after the movable platform 110 reaches the preset detection window, controlling the multi-level spatial positioning mechanism 120 to perform end-point fine positioning based on the point cloud data of the surface of the turbine runner 20 to be tested collected by the 3D topography scanner to determine the second pose deviation between the current pose of the composite detection terminal 130 and the surface of the turbine runner 20 to be tested; controlling the multi-level spatial positioning mechanism 120 to move based on the second pose deviation, so that the composite detection terminal 130 is aligned with the surface of the turbine runner 20 to be tested; and controlling the composite detection terminal 130 to perform in-situ detection of the turbine runner 20 to be tested.

[0033] In this embodiment, the preset detection window is a target area range for the hierarchical positioning controller 140 when performing global coarse positioning. It includes the circumferential position range on the bearing surface 30 and the vertical height range of the composite detection terminal 130. Global coarse positioning is completed when the movable platform 110 is located within the circumferential position range of the preset detection window and the composite detection terminal 130 is located within the vertical height range of the preset detection window.

[0034] The global coarse positioning stage includes circumferential coarse positioning and vertical coarse positioning. During the circumferential coarse positioning process, the hierarchical positioning controller 140 controls the movable platform 110 to perform global coarse positioning based on a preset path and environmental visual features to determine the current position of the movable platform 110 on the bearing surface 30. The preset path can be a path information pre-planned based on the waterwheel chamber CAD model, or a path planned in real time using the waterwheel chamber CAD model during the movement of the movable platform 110; this application does not limit this. The hierarchical positioning controller 140 extracts the two-dimensional planar layout of the bearing surface 30 from the waterwheel chamber CAD model. For example, this layout may include the projection position of the turbine main shaft on the bearing surface 30, the position and outline of each fixed obstacle on the bearing surface 30, and the projection position and circumferential distribution of the turbine runner 20 to be tested on the bearing surface 30. Based on the above information, the hierarchical positioning controller 140 obtains a collision-free path on the two-dimensional plane of the bearing surface 30 from the current starting position of the movable platform 110 to the axial position range of the preset detection window.

[0035] As the movable platform 110 moves along a preset path, a vision camera (not shown in the figure) mounted on the movable platform 110 acquires real-time environmental images of the surrounding environment of the bearing surface 30. The hierarchical positioning controller 140 extracts environmental visual features from these images. These environmental visual features include, but are not limited to, feature points with known coordinates in the waterwheel chamber CAD model, such as fixed structures, bolt holes, scale marks, and columns on the inner wall of the waterwheel chamber. The hierarchical positioning controller 140 matches the acquired environmental visual features with the feature points in the waterwheel chamber CAD model to determine the absolute position of the movable platform 110 on the bearing surface 30, thus obtaining the current position of the movable platform 110.

[0036] The preset path provides macroscopic path information along which the movable platform 110 should travel, i.e., the overall route the movable platform 110 travels from its starting position towards the preset detection window. Environmental visual features provide an absolute position reference for the movable platform 110 during its travel. By matching the real-time acquired environmental visual features with feature points in the CAD model, the cumulative error caused by relying solely on the odometer during long-distance travel of the movable platform 110 can be eliminated. The preset path and environmental visual features work together to achieve circumferential coarse positioning of the movable platform 110. During vertical coarse positioning, the hierarchical positioning controller 140 controls the multi-level spatial positioning mechanism 120 to vertically lift the composite detection terminal 130, moving the composite detection terminal 130 to a height near the turbine runner 20 to be tested.

[0037] After determining the current position of the movable platform 110, the hierarchical positioning controller 140 calculates the first pose deviation between the current position and the preset detection window. The first pose deviation includes a first position deviation and a first attitude deviation. Based on the first position deviation and the first attitude deviation, the hierarchical positioning controller 140 generates a drive command to control the movable platform 110 and the multi-level spatial positioning mechanism 120 to adjust and move to the preset detection window respectively.

[0038] For example, the positioning accuracy of the global coarse positioning stage can be at the centimeter level (e.g., ±3cm to ±5cm), which can send the base of the multi-level spatial positioning mechanism 120 into the working range where end-effector fine positioning can be performed.

[0039] After the movable platform 110 and the multi-level spatial positioning mechanism 120 reach the preset detection window, the hierarchical positioning controller 140 starts the end-point precision positioning process.

[0040] First, the hierarchical positioning controller 140 controls a 3D topography scanner to acquire point cloud data of the surface of the turbine runner 20 under test, and performs end-effector fine positioning based on the point cloud data acquired by the 3D topography scanner to determine the second pose deviation between the current pose of the composite detection terminal 130 and the surface of the turbine runner 20 under test. For example, the hierarchical positioning controller 140 determines the current pose of the composite detection terminal 130 in the global coordinate system through positive kinematic mapping based on the current pose states of each moving component in the multi-level spatial positioning mechanism 120 (including the lifting height, joint angles, etc. of the multi-level spatial positioning mechanism 120); simultaneously, based on the point cloud data acquired by the 3D topography scanner, it identifies the target detection point on the surface of the turbine runner 20 under test, determines the desired pose (including desired position and desired attitude) at the target detection point; then, it calculates the deviation between the current pose and the desired pose to obtain the second pose deviation.

[0041] Then, based on the aforementioned second pose deviation, the hierarchical positioning controller 140 controls the movement of the multi-stage spatial positioning mechanism 120 to align the composite detection terminal 130 with the surface of the turbine runner 20 under test. This process is a pose closed-loop control process: the hierarchical positioning controller 140 continuously acquires the real-time pose of the composite detection terminal 130 through the encoders of each joint in the multi-stage spatial positioning mechanism 120, calculates the real-time deviation between it and the desired pose, and generates drive commands based on the real-time deviation to control the movement of the multi-stage spatial positioning mechanism 120 until the real-time deviation approaches zero, that is, the composite detection terminal 130 is precisely aligned with the surface of the turbine runner 20 under test.

[0042] End-efficiency fine positioning is a second-level positioning that differs from global coarse positioning. Its purpose is to further converge the pose of the composite detection terminal 130 from centimeter-level accuracy to millimeter-level accuracy, ensuring that the 3D topography scanner can acquire point cloud data of the wheel surface under the optimal measurement position and attitude.

[0043] After the composite inspection terminal 130 is aligned with the surface of the turbine runner 20 under test, the graded positioning controller 140 controls the composite inspection terminal 130 to perform in-situ inspection of the surface of the turbine runner 20. The graded positioning controller 140 controls the three-dimensional topography scanner to perform three-dimensional topography scanning on the surface of the turbine runner 20 under test, and obtains three-dimensional point cloud data of the runner surface for subsequent defect identification and analysis (such as detection and evaluation of defects such as abrasion, cavitation, and cracks).

[0044] After the detection of a detection position is completed, the graded positioning controller 140 can control the movable platform 110 to move circumferentially along the bearing surface 30 to the next preset detection window, repeating the above-mentioned global coarse positioning, end fine positioning and in-situ detection steps until the detection of all areas to be inspected of the turbine runner 20 to be tested is completed.

[0045] The in-situ inspection robot 10 provided in this embodiment has a movable platform 110 that is always positioned on the bearing surface 30 below the turbine runner 20 to be tested (such as the bottom floor of the turbine chamber). The composite inspection terminal 130 is lifted from the bearing surface 30 upwards to the runner surface for inspection. Throughout the inspection process, the movable platform 110 maintains stable contact with the bearing surface 30, eliminating the risk of falling and causing secondary damage to the runner. This also shortens the inspection cycle and reduces maintenance costs. A two-level coupling strategy combining global coarse positioning and end-point fine positioning is implemented through a hierarchical positioning controller 140, improving the inspection positioning accuracy and ensuring the consistency and comparability of the inspection data. A highly reliable global positioning is achieved through a global coarse positioning strategy that coordinates a preset path with environmental visual features. This embodiment achieves fully automatic positioning and inspection through a hierarchical positioning controller, requiring no manual intervention throughout the process. This standardizes and automates the inspection process, and the inspection results are unaffected by human factors, resulting in high reliability.

[0046] Figure 3 This is a schematic diagram of another in-situ inspection robot provided in an embodiment of this application. Figure 4 This is a schematic diagram of another in-situ inspection robot provided in an embodiment of this application. Please refer to... Figure 3 and Figure 4 In some embodiments, the multi-level spatial positioning mechanism 120 includes a multi-level lifting platform 121 and a multi-degree-of-freedom robotic arm 122.

[0047] The multi-stage lifting platform 121 is vertically mounted on the movable platform 110. For example... Figure 4 As shown, the multi-stage lifting platform 121 can employ multiple sets of vertical synchronous lifting mechanisms, such as multi-stage hydraulic cylinders, multi-stage electric cylinders, or screw lifting mechanisms, etc., which are not limited in this application. The multi-stage lifting platform 121 is used to lift the composite testing terminal 130 in the vertical direction, so that the composite testing terminal 130 is moved from the height of the bearing surface 30 to the vicinity of the height of the turbine runner 20 under test. The multi-stage lifting platform 121 achieves a large range of vertical displacement through its own multi-stage telescopic structure, thereby compensating for the height difference between the bearing surface 30 of the movable platform 110 and the turbine runner 20 under test.

[0048] A multi-degree-of-freedom robotic arm 122 is mounted on top of a multi-stage lifting platform 121. A composite detection terminal 130 is mounted at the end of the multi-degree-of-freedom robotic arm 122. The multi-degree-of-freedom robotic arm 122 is used to adjust the spatial orientation of the composite detection terminal 130 so that the composite detection terminal 130 can be aligned with the surface of the turbine runner 20 under test in a suitable orientation.

[0049] The hierarchical positioning controller 140 is communicatively connected to the multi-stage lifting platform 121 and the multi-degree-of-freedom robotic arm 122. In actual operation, the hierarchical positioning controller 140 controls the multi-stage lifting platform 121 and the multi-degree-of-freedom robotic arm 122 to move in coordination. Specifically, the hierarchical positioning controller 140 is also used to: control the multi-stage lifting platform 121 to lift the composite detection terminal 130 in the vertical direction, so that the composite detection terminal 130 moves to the height of the turbine runner 20 to be tested; and control the multi-degree-of-freedom robotic arm 122 to adjust the current posture of the composite detection terminal 130 so that the composite detection terminal 130 is aligned with the surface of the turbine runner 20 to be tested.

[0050] In the vertical coarse positioning stage of global coarse positioning, the hierarchical positioning controller 140 controls the multi-stage lifting platform 121 to lift the composite detection terminal 130 in the vertical direction, so that the composite detection terminal 130 moves to a height near the turbine runner 20 under test. In some embodiments, the hierarchical positioning controller 140 determines the target extension amount of the multi-stage lifting platform 121 based on the platform height of the movable platform 110, the vertical projection height of the multi-degree-of-freedom robotic arm 122 at the current joint angle, and the distance between the composite detection terminal 130 and the surface of the turbine runner 20 under test, and controls the multi-stage lifting platform 121 to extend to the target extension amount, so that the composite detection terminal 130 is located within the vertical height range of the preset detection window.

[0051] During the final positioning stage, the hierarchical positioning controller 140 controls the multi-degree-of-freedom robotic arm 122 to adjust the current pose of the composite inspection terminal 130 so that the composite inspection terminal 130 is aligned with the surface of the turbine runner 20 under test. Based on the point cloud data acquired by the 3D topography scanner, the hierarchical positioning controller 140 determines the pose deviation between the current pose of the composite inspection terminal 130 and the desired pose of the surface of the turbine runner 20 under test. Then, it drives the joints of the multi-degree-of-freedom robotic arm 122 to move to eliminate this pose deviation, ensuring that the composite inspection terminal 130 is precisely aligned with the surface of the turbine runner 20 under test.

[0052] Understandably, the multi-stage lifting platform 121 performs coarse adjustment of vertical displacement over a large range, lifting the composite detection terminal 130 from the bearing surface 30 to near the height of the rotating wheel; while the multi-degree-of-freedom robotic arm 122 performs fine adjustment of spatial pose over a small range, precisely aligning the composite detection terminal 130 from near the height of the rotating wheel to the detection point on the rotating wheel surface. Under the control of the hierarchical positioning controller 140, the two work together to achieve high-precision positioning of the composite detection terminal 130 from the bearing surface 30 to the rotating wheel surface.

[0053] During the lifting and lowering motion of the multi-stage lifting platform 121, the multi-degree-of-freedom robotic arm 122 can maintain a safe retracted posture to avoid interference between the robotic arm and the surrounding environment. After the multi-stage lifting platform 121 reaches the target height, the multi-degree-of-freedom robotic arm 122 extends and adjusts its posture.

[0054] The multi-level spatial positioning mechanism 120 provided in this embodiment achieves decoupled control of large-range vertical displacement and small-range pose fine adjustment through the coordinated cooperation of the multi-level lifting platform 121 and the multi-degree-of-freedom robotic arm 122, thereby improving the overall positioning accuracy and system reliability.

[0055] In some embodiments, the multi-degree-of-freedom robotic arm 122 includes multiple joints. Exemplarily, the multi-degree-of-freedom robotic arm 122 is a six-degree-of-freedom serial robotic arm, including six rotary joints, such as... The multi-degree-of-freedom robotic arm 122 achieves flexible positioning and attitude adjustment in three-dimensional space through the coordinated rotation of its joints. The hierarchical positioning controller 140 divides the multiple joints of the multi-degree-of-freedom robotic arm 122 into a position control group and an attitude control group. The position control group controls the spatial position of the composite detection terminal 130 in the global coordinate system, i.e., the X, Y, and Z coordinates. The attitude control group controls the pointing attitude of the composite detection terminal 130, i.e., the pitch angle, yaw angle, and roll angle. For example, the position control group includes the first three of the six joints (…). The attitude control group includes the last three joints ( It should be understood that the above grouping method is only an example, and those skilled in the art can make adaptive adjustments according to the specific configuration and joint arrangement of the multi-degree-of-freedom robotic arm 122.

[0056] The hierarchical positioning controller 140 is also specifically used for: driving each joint in the position control group to perform position compensation one by one according to a preset sequence, and locking the joints in the position control group that have completed position compensation; after all joints in the position control group are locked, driving each joint in the attitude control group to adjust the attitude of the composite detection terminal 130 one by one, so that the composite detection terminal 130 is aligned with the surface of the turbine runner 20 to be tested.

[0057] Joint compensation employs a single-axis sequential compensation control logic. For example, the hierarchical positioning controller 140 drives the first joint in the position control group to move according to its corresponding joint compensation amount, completing the position compensation for that joint. After the joint reaches the target position, the hierarchical positioning controller 140 controls the joint's brake or servo driver to lock the joint in its current position, preventing passive displacement under external forces. Subsequently, based on the state of the locked joints, the hierarchical positioning controller 140 recalculates the compensation amount for the remaining joints, then drives the second joint in the position control group to move according to the updated joint compensation amount, completing the position compensation for that joint and locking it. This process continues, with the hierarchical positioning controller 140 sequentially completing the position compensation and locking of each joint in the position control group according to a preset sequence, until all joints in the position control group have been compensated and locked.

[0058] After all joints in the position control group are locked, the hierarchical positioning controller 140 drives each joint in the attitude control group to move one by one according to a preset sequence, adjusting the pointing attitude of the composite detection terminal 130, including pitch angle, yaw angle, and roll angle, until the attitude of the composite detection terminal 130 is consistent with the normal direction of the target detection point on the surface of the turbine runner 20 under test. Each joint in the attitude control group can also adopt a strategy of compensating and locking one by one to ensure that no additional position disturbance is introduced during the attitude adjustment process.

[0059] In this embodiment, the hierarchical positioning controller 140 uses single-axis sequential compensation control logic to drive each joint of the multi-degree-of-freedom robotic arm 122. By locking the joint after each compensation is completed, the inverse kinematics solution of subsequent joints can be performed based on the locked, determined state, eliminating coupling errors in multi-axis linkage. Since each joint completes compensation and locking independently, positioning accuracy is improved. Through the above-mentioned grouped single-axis compensation control, the hierarchical positioning controller 140 can accurately align the composite detection terminal 130 with the surface of the turbine runner 20 under test during the end-effector fine positioning stage, ensuring that the 3D topography scanner can collect point cloud data of the runner surface under the optimal measurement posture.

[0060] like Figure 3 and Figure 4 As shown, in some embodiments, the multi-stage lifting platform 121 is equipped with a limiting device (not shown) to constrain the movement range of the multi-stage lifting platform 121 during lifting motion, preventing the multi-stage lifting platform 121 from exceeding its mechanical travel boundary and preventing the composite detection terminal 130 from colliding with the turbine runner 20 under test or other surrounding environments. The limiting device includes at least one of the following: mechanical hard limiting and electronic soft limiting.

[0061] In some embodiments, the mechanical hard limit is a limit block disposed at the end of the stroke of the multi-stage lifting platform 121. The limit block is disposed at the end position of the movement trajectory of the lifting column or lifting end of the multi-stage lifting platform 121, for example, at the top and bottom stroke ends of the multi-stage lifting platform 121. When the lifting module of the multi-stage lifting platform 121 moves to the end of its stroke and contacts the limit block during the lifting process, the limit block forcibly blocks the movement of the lifting module through physical contact. The limit block can be a rigid structural component, such as a metal block or a shock-absorbing buffer block. When the lifting module contacts the limit block, the limit block physically prevents the lifting module from continuing to move in a direction beyond the stroke boundary, thereby preventing the multi-stage lifting platform 121 from exceeding its designed maximum extension or minimum retraction. For example, when the lifting module reaches a position 50mm away from the limit block, it is controlled to move slowly. The mechanical hard limit is a purely physical structure that does not rely on electrical signals or software control. Even if the electronic soft limit fails or the control system malfunctions, the mechanical hard limit can still function independently to rigidly constrain the movement range of the multi-stage lifting platform 121.

[0062] In some embodiments, the multi-stage lifting platform 121 is equipped with an electronic soft limit device (not shown in the figure), which includes an absolute encoder and a laser rangefinder. The absolute encoder is mounted on the drive motor of the multi-stage lifting platform 121 or on the linkage shaft of the lifting column, and is mechanically coupled to the lifting module of the multi-stage lifting platform 121.

[0063] The hierarchical positioning controller 140 is also specifically used to: control the absolute encoder to feed back the lifting position of the multi-stage lifting platform 121 to determine whether the multi-stage lifting platform 121 has reached the preset travel boundary; control the laser ranging sensor to detect the distance between the composite detection terminal 130 and the surface of the turbine runner 20 to be tested to determine whether the distance is less than the preset safety threshold; and restrict the operation of the drive of the multi-stage lifting platform 121 when the lifting position reaches the preset travel boundary or the distance is less than the preset safety threshold.

[0064] An absolute encoder is used to provide real-time feedback on the lifting position of the multi-stage lifting platform 121, i.e., the extension amount or absolute height position of the lifting module at the current moment. A hierarchical positioning controller 140 communicates with the absolute encoder to acquire the lifting position of the multi-stage lifting platform 121 in real time at a set frequency (e.g., 100Hz or higher). The hierarchical positioning controller 140 has preset travel boundaries for the lifting position. These preset travel boundaries include an upper travel boundary (the maximum height the lifting module is allowed to extend) and a lower travel boundary (the minimum height the lifting module is allowed to retract), corresponding to the highest and lowest permissible positions of the multi-stage lifting platform 121 within a safe range, respectively. During the lifting motion, the hierarchical positioning controller 140 continuously compares the lifting position fed back by the absolute encoder with the preset travel boundaries to determine whether the multi-stage lifting platform 121 has reached the preset travel boundaries. The hierarchical positioning controller 140 determines whether the lifting position has approached or reached the upper or lower travel boundary. For example, when the distance between the lifting position and the upper or lower travel boundary is less than a preset threshold (e.g., 10mm), the graded positioning controller 140 controls the multi-stage lifting platform 121 to decelerate; when the lifting position reaches the preset travel boundary, the graded positioning controller 140 immediately restricts the driver of the multi-stage lifting platform 121 from continuing to run in a direction beyond the boundary and triggers a safety warning.

[0065] A laser rangefinder is mounted at the end of the multi-level spatial positioning mechanism 120 or on the composite detection terminal 130, for example, on the housing of the composite detection terminal 130, adjacent to the 3D topography scanner. The laser rangefinder's detection direction is towards the surface of the turbine runner 20 under test, used to detect the real-time distance between the composite detection terminal 130 and the surface of the turbine runner 20. The hierarchical positioning controller 140 is communicatively connected to the laser rangefinder and acquires this distance data in real time at a set frequency. The hierarchical positioning controller 140 has a preset safety threshold. The preset safety threshold is used to evaluate whether the composite detection terminal 130 is too close to the turbine runner 20 under test, posing a collision risk. The specific value of the preset safety threshold can be determined comprehensively based on the working distance range of the 3D topography scanner, the positioning accuracy of the multi-degree-of-freedom robotic arm 122, and the topographic features of the runner surface. For example, the preset safety threshold can be 30mm to 50mm. During vertical coarse positioning, end-effector fine positioning, or in-place detection, the hierarchical positioning controller 140 continuously compares the distance detected by the laser rangefinder with a preset safety threshold to determine whether the distance is less than the preset safety threshold. When the distance is less than the preset safety threshold, the hierarchical positioning controller 140 immediately restricts the drive of the multi-stage lifting platform 121 from continuing to move towards the wheel and sends a collision warning signal to the system.

[0066] In some embodiments, mechanical hard limits and electronic soft limits together constitute a multi-level safety protection system for the multi-stage lifting platform 121. The electronic soft limit, as the first layer of protection, monitors the lifting position and end distance in real time through an absolute encoder and a laser rangefinder, limiting the drive operation in advance when approaching the limit position, thus achieving active safety protection. The mechanical hard limit, as the second layer of protection, forcibly blocks movement through the physical blocking action of the limit block when the electronic soft limit fails or the lifting module crosses the electronic soft limit boundary due to abnormal circumstances, thus achieving passive safety protection. The limiting device provided in this embodiment, through the combination of active monitoring by electronic soft limits and physical blocking by mechanical hard limits, avoids damage to the multi-stage lifting platform 121 due to exceeding its travel range or collision between the composite detection terminal 130 and the rotating wheel, improving the safety of the in-situ detection robot 10 operation.

[0067] In some embodiments, the composite inspection terminal 130 includes a 3D topography scanner and a visual imaging module (not shown in the figure). Both the 3D topography scanner and the visual imaging module are located at the end of the multi-level spatial positioning mechanism 120. Specifically, the hierarchical positioning controller 140 is used to: monitor the measurement confidence level of the point cloud data acquired by the 3D topography scanner; switch to a backup positioning mode guided by the visual imaging module when the measurement confidence level is lower than a preset threshold; in the backup positioning mode, acquire the deviation between the current image features acquired by the visual imaging module and the target image features; convert the deviation into joint speed control commands for the multi-level spatial positioning mechanism 120; and, based on the joint speed control commands, drive the movement of each joint in the multi-level spatial positioning mechanism 120 so that the composite inspection terminal 130 is aligned with the surface of the turbine runner 20 to be tested.

[0068] The visual imaging module is used to acquire environmental or visual images of the surface of the turbine runner 20 under test. The visual imaging module may include industrial cameras, binocular cameras, or depth cameras, etc., and this application does not limit this to any particular type. During the global coarse positioning stage, the visual imaging module is used to acquire environmental images around the bearing surface 30, so that the hierarchical positioning controller 140 can extract environmental visual features and match them with the turbine chamber CAD model, thereby assisting in the circumferential coarse positioning of the movable platform 110. Furthermore, during the environmental adaptive switching process, the visual imaging module is also used to provide image feature guidance in the backup positioning mode.

[0069] Both the 3D topography scanner and the visual imaging module are located at the end of the multi-level spatial positioning mechanism 120, and move together with the mechanism. In some embodiments, the 3D topography scanner and the visual imaging module are installed adjacent to each other, and their detection directions are both directed towards the surface of the turbine runner 20 under test, so that their detection areas are consistent or substantially overlapped, facilitating subsequent fusion processing of multi-sensor data. The 3D topography scanner can provide high-precision 3D spatial information, but may fail under harsh environments such as abnormal lighting or water mist interference; the visual imaging module can provide rich texture and color information and has strong adaptability to changes in ambient lighting, but cannot directly provide 3D spatial coordinates. By jointly located at the end of the multi-level spatial positioning mechanism 120, and through multi-sensor data fusion, both can leverage their respective advantages, improving the system's adaptability and reliability under different environmental conditions.

[0070] like Figure 1 As shown, in some embodiments, the robot also includes a dual-redundant energy supply module 150. The dual-redundant energy supply module 150 provides power to various electrical components such as the movable platform 110, the multi-level spatial positioning mechanism 120, the composite detection terminal 130, and the hierarchical positioning controller 140. The dual-redundant energy supply module 150 includes a primary energy source 151 and a backup energy source 152. The primary energy source 151 is configured to connect to an external power source via a drag cable. One end of the drag cable is connected to an external power source (such as a 220V AC power supply from a hydroelectric power plant or a DC power supply from a power plant panel), and the other end is connected to the power supply interface of the in-situ inspection robot 10. When the in-situ inspection robot 10 moves autonomously along the bearing surface 30, the drag cable moves with the robot, continuously providing power to the robot. The primary energy source 151 provides continuous power for the normal inspection operations of the in-situ inspection robot 10 and can simultaneously charge the backup energy source 152 when the external power source is normal. The backup energy source 152 is configured as an onboard energy storage battery. The onboard energy storage battery can be mounted on the movable platform 110 and move with the platform 110. When the tow cable is normally connected to the external power source, the onboard energy storage battery is in a float charging state; when the tow cable is disconnected or the external power source fails, the hierarchical positioning controller 140 automatically switches to the backup energy source 152 for power supply, and the onboard energy storage battery provides emergency power to the in-situ inspection robot 10.

[0071] The hierarchical positioning controller 140 is also specifically used to: determine the minimum reserve power required for the backup energy source 152 based on the motor output torque and angular velocity of each joint in the multi-level spatial positioning mechanism 120, the preset safe return time, the power consumption of the robot's control system, and the preset safety margin coefficient. This minimum reserve power refers to the minimum amount of electricity that the backup energy source 152 needs to store in an emergency to ensure that the in-situ detection robot 10 can safely return to a safe area from any working position.

[0072] In some embodiments, the hierarchical positioning controller 140 incorporates an emergency energy budget model for calculating minimum reserve power. The calculation relationship satisfies the following formula (1): Formula (1): .

[0073] in, The preset safe return time refers to the maximum time required for the in-situ inspection robot 10 to return to the safe point from its current inspection position along a preset safe path. For the ( i The motor output torque of each joint is expressed in Newton-meters (N·m). For the ( i The angular velocity of each joint, expressed in radians per second (rad / s). Indicates the ( i The instantaneous mechanical power of a joint, expressed in watts (W). The power consumption of the control system of the in-situ inspection robot 10 is expressed in joules (J). The power consumption of the control system includes the power consumption of the hierarchical positioning controller 140 itself, the power consumption of the 3D topography scanner, the power consumption of the vision imaging module, the power consumption of each sensor, and the power consumption of the communication module, etc., representing the total energy consumption within the safe homing time. The preset safety margin coefficient is dimensionless. The preset safety margin coefficient is used to reserve energy margin for uncertainties such as battery aging, load fluctuations, changes in ambient temperature, and changes in friction, to ensure that the backup energy source 152 can still meet the requirements for safe return to its original position under various non-ideal operating conditions.

[0074] When the tow cable is disconnected or the external power supply fails (e.g., the cable is accidentally crushed and broken or the power station suddenly loses power), the hierarchical positioning controller 140 automatically switches to the backup power source 152 and triggers the one-key reset function. After the one-key reset function is triggered, the hierarchical positioning controller 140 controls the movable platform 110 to return to the safe area along the preset safe path according to the preset safe return strategy, controls the multi-stage lifting platform 121 to retract to the lowest position, and controls the multi-degree-of-freedom robotic arm 122 to retract to the safe posture, ensuring that the in-situ inspection robot 10 can safely evacuate the inspection area of ​​the turbine runner 20 under any unexpected circumstances.

[0075] The dual-redundant energy supply module 150 provided in this application embodiment improves the power supply reliability and emergency survivability of the in-situ inspection robot 10 in the complex operating environment of a hydropower station by combining the main power supply from the drag cable with the backup power supply from the onboard energy storage battery. This is achieved through a quantitative calculation of backup power using an emergency energy budget model and a one-click reset function. Even in extreme cases such as accidental disconnection of the drag cable or external power failure, the in-situ inspection robot 10 still has sufficient backup power to complete a safe return to its original position, avoiding safety risks caused by the robot being stranded in the rotary inspection area due to power outages.

[0076] Compared to methods using magnetic crawling robots for inspection, which risk damage to the turbine runner from falling, the in-situ inspection robot 10 provided in this application can perform runner inspection without compromising runner safety and is unaffected by runner shape, thus having wide applicability. Compared to track-based inspection robots, the in-situ inspection robot 10 provided in this application can achieve automatic path planning and inspection on load-bearing surfaces 30 such as those inside un-tracked waterwheel chambers. It ensures end-point accuracy through graded positioning, runner safety through mechanical and electronic dual-limiting, and provides one-click recovery capability in emergencies through dual-redundant power supply. Compared to manual inspection methods, the in-situ inspection robot 10 provided in this application can achieve fully automated inspection, avoiding the dangers of manual high-altitude operations, and the inspection process is standardized with consistent results.

[0077] On the other hand, this application also provides an in-situ detection method, applicable to any of the in-situ detection robots 10 provided in this application. Please refer to... Figure 2 The method includes the following steps S110 to S150.

[0078] S110: The movable platform is controlled to perform global coarse positioning based on a preset path and environmental visual features to determine its current position on the bearing surface. The hierarchical positioning controller 140 controls the movable platform 110 to perform global coarse positioning based on a preset path and environmental visual features to determine its current position on the bearing surface 30. The specific implementation of this step is consistent with the description of circumferential coarse positioning in the aforementioned global coarse positioning stage. The movable platform 110 travels along a preset path pre-planned according to the waterwheel chamber CAD model. The visual camera set on the movable platform 110 collects environmental images around the bearing surface 30 in real time. The hierarchical positioning controller 140 extracts environmental visual features from the environmental images and matches the collected environmental visual features with feature points in the waterwheel chamber CAD model to determine the absolute position of the movable platform 110 on the bearing surface 30.

[0079] S120: Based on the first pose deviation between the current position and the preset detection window, control the movable stage to move to the preset detection window. After determining the current position of the movable stage 110, the hierarchical positioning controller 140 calculates the first pose deviation between the current position and the preset detection window, and generates a drive command based on the first pose deviation to control the movable stage 110 and the multi-level spatial positioning mechanism 120 to move to the preset detection window. The preset detection window is the target area range for global coarse positioning, including the circumferential position range on the bearing surface 30 and the vertical height range of the composite detection terminal 130.

[0080] It should be noted that, in this embodiment, the movable platform 110 moving to the preset detection window means that the movable platform 110 enters the circumferential position range of the preset detection window. At this time, the hierarchical positioning controller 140 determines that the movable platform 110 has reached the preset detection window. It should also be noted that, during the process of the movable platform 110 moving along the preset path towards the preset detection window, and when the movable platform 110 reaches the preset detection window, the hierarchical positioning controller 140 also controls the multi-level spatial positioning mechanism 120 to perform vertical coarse positioning, vertically lifting the composite detection terminal 130 to within the vertical height range of the preset detection window. That is, in step S120, the movable platform 110 moving to the preset detection window and the vertical coarse positioning can be performed in parallel or sequentially, and both together complete the global coarse positioning.

[0081] S130: After the movable platform reaches the preset detection window, the multi-level spatial positioning mechanism is controlled to perform end-effector fine positioning based on the point cloud data of the turbine runner surface to be tested collected by the 3D topography scanner, and to determine the second pose deviation between the current pose of the composite detection terminal and the turbine runner surface to be tested. After the movable platform 110 reaches the circumferential position range of the preset detection window and the composite detection terminal 130 is located within the vertical height range of the preset detection window, the hierarchical positioning controller 140 starts the end-effector fine positioning process. The hierarchical positioning controller 140 controls the 3D topography scanner to acquire point cloud data of the surface of the turbine runner 20 under test, and determines the current pose of the composite detection terminal 130 in the global coordinate system through positive kinematic mapping based on the current pose state of each moving part in the multi-level spatial positioning mechanism 120 (including the lifting height, joint angle, etc. of the multi-level spatial positioning mechanism 120). At the same time, based on the point cloud data acquired by the 3D topography scanner, the target detection point on the surface of the turbine runner 20 under test is identified, and the desired pose at the target detection point is determined. Then, the deviation between the current pose and the desired pose is calculated to obtain the second pose deviation between the current pose of the composite detection terminal 130 and the surface of the turbine runner 20 under test.

[0082] S140: Based on the second pose deviation, control the movement of the multi-level spatial positioning mechanism to align the composite detection terminal with the surface of the turbine runner under test. The hierarchical positioning controller 140 controls the movement of the multi-level spatial positioning mechanism 120 based on the second pose deviation determined in step S130 to align the composite detection terminal 130 with the surface of the turbine runner 20 under test.

[0083] S150: Control the composite inspection terminal to perform in-situ inspection of the turbine runner under test. After the composite inspection terminal 130 is aligned with the surface of the turbine runner 20 under test, the graded positioning controller 140 controls the composite inspection terminal 130 to perform in-situ inspection of the surface of the turbine runner 20 under test. The graded positioning controller 140 controls the three-dimensional topography scanner to perform three-dimensional topography scanning on the surface of the turbine runner 20 under test, and obtains three-dimensional point cloud data of the runner surface for subsequent defect identification and analysis (such as detection and evaluation of defects such as abrasion, cavitation, and cracks).

[0084] After the detection of a detection position is completed, the graded positioning controller 140 can control the movable platform 110 to move circumferentially along the bearing surface 30 to the next preset detection window, repeating the above steps S110 to S150 until the detection of all areas to be inspected of the turbine runner 20 to be tested is completed.

[0085] The in-situ detection method provided in this application implements a two-level coupled positioning strategy of global coarse positioning and end-point fine positioning through a hierarchical positioning controller 140, achieving stepwise precise positioning from a large area on the bearing surface 30 to a millimeter-level detection position on the surface of the rotating wheel. This method completes in-situ detection without disassembling the rotating wheel, and the detection process is fully automatic, standardized, and the detection results are objective and reliable.

[0086] In some embodiments, step S130 above includes the following sub-steps S131 and S132.

[0087] S131: Based on the current pose of each moving component in the multi-level spatial positioning mechanism, the current pose of the composite detection terminal in the global coordinate system is determined by positive kinematic mapping; wherein, the current pose is determined by the target extension of the multi-level lifting platform, the circumferential position of the movable platform, and the current joint angle of the multi-degree-of-freedom robotic arm; the target extension is determined based on at least one of the following: the platform height of the movable platform, the vertical projection height of the multi-degree-of-freedom robotic arm under the current joint angle, and the distance between the composite detection terminal and the surface of the turbine runner to be tested.

[0088] After the movable platform 110 reaches the circumferential position range of the preset detection window and the multi-stage lifting platform 121 has reached the target extension amount, the hierarchical positioning controller 140 starts the end fine positioning process, which first needs to determine the current pose of the composite detection terminal 130 in the global coordinate system.

[0089] The hierarchical positioning controller 140 acquires the current pose state of each moving component in the multi-level spatial positioning mechanism 120, including the target extension of the multi-level lifting platform 121. (i.e., current extension amount), the circumferential position of the movable platform 110 on the bearing surface 30 and the current joint angles of each joint of the multi-degree-of-freedom robotic arm 122. .

[0090] Based on the aforementioned current pose state, the hierarchical positioning controller 140 determines the current pose of the composite detection terminal 130 in the global coordinate system through forward kinematic mapping. For example, this forward kinematic mapping relationship can be established using a kinematic model based on the improved DH parameter method, and its forward kinematic equation is expressed as formula (2): Formula (2): .

[0091] in: The coordinates of the composite detection terminal 130 in the global coordinate system are the position components of the current pose. The motion mapping function of the multi-degree-of-freedom robotic arm 122 relative to its base coordinate system is used to calculate the position of the composite detection terminal 130 relative to the base of the multi-degree-of-freedom robotic arm 122 by means of the improved DH parameter method based on the current joint angle θ of each joint of the multi-degree-of-freedom robotic arm 122. The transformation matrix of the robot arm base in the global coordinate system is given by the circumferential position of the movable platform 110. and target stretch Determined in real time.

[0092] During the vertical coarse positioning process in the global coarse positioning stage, the hierarchical positioning controller 140 needs to determine the target extension amount of the multi-stage lifting platform 121 and control the multi-stage lifting platform 121 to extend to the target extension amount so as to lift the composite detection terminal 130 from the bearing surface 30 to the vertical height range of the preset detection window.

[0093] In this embodiment, the preset detection window can be a three-dimensional spatial range, including the circumferential position range on the bearing surface 30 and the vertical height range of the composite detection terminal 130. The goal of vertical coarse positioning is to lift the composite detection terminal 130 to this vertical height range. In order to accurately control the extension amount of the multi-stage lifting platform 121, the hierarchical positioning controller 140 determines the target extension amount of the multi-stage lifting platform 121 according to the following geometric constraints.

[0094] The dynamic geometric constraint equation satisfied by the height parameters of the multi-stage lifting platform 121 can be expressed by the following formula (3): Formula (3): .

[0095] Wherein, H is the vertical distance from the bearing surface 30 (such as the bottom of the waterwheel chamber) to the surface of the turbine runner 20 to be measured. The value of H is determined by the turbine installation dimensions and is a known constant that can be pre-stored in the graded positioning controller 140. The platform height of the movable platform 110 is the vertical distance between the bottom of the traveling mechanism of the movable platform 110 (the part in contact with the bearing surface 30) and its top surface (the part connected to the multi-stage lifting platform 121). The value is a known fixed value, determined by the mechanical structure of the movable platform 110. This refers to the real-time extension of the multi-stage lifting platform 121, specifically the vertical extension of the multi-stage lifting platform 121 relative to its retracted state. This value is controlled in real-time by the graded positioning controller 140 and is also the target extension amount that needs to be determined in this step. This is the vertical projection height of the multi-degree-of-freedom robotic arm 122 at the current joint angle θ, i.e., the vertical projection distance of the end of the multi-degree-of-freedom robotic arm 122 (i.e., the installation position of the composite detection terminal 130) relative to the top of the multi-stage lifting platform 121. This value changes with the configuration of the multi-degree-of-freedom robotic arm 122 and can be calculated by forward kinematics based on the current angles of each joint of the multi-degree-of-freedom robotic arm 122. This is the real-time distance between the composite detection terminal 130 and the surface of the turbine runner 20 under test. This value is obtained in real time by a laser rangefinder sensor installed on the composite detection terminal 130.

[0096] In practical applications, the hierarchical positioning controller 140 determines the target extension amount based on the aforementioned geometric constraints. That is, given H, , and In this case, the hierarchical positioning controller 140 can determine the target extension amount by the following formula (4): Formula (4): .

[0097] It is understandable that the total vertical distance H from the bearing surface 30 of the turbine chamber to the impeller may differ in different hydropower stations, and the distance from the impeller surface to the bearing surface 30 at different heights within the same hydropower station may also vary depending on the detection position. In this embodiment, the hierarchical positioning controller 140 calculates the target extension in real time based on the aforementioned geometric constraints, rather than using fixed lifting height commands. This allows the multi-stage lifting platform 121 to adaptively adjust to a suitable height, thereby adapting to the working conditions of different hydropower stations, different impellers, and different detection positions on the impeller.

[0098] In this embodiment, by introducing the aforementioned dynamic geometric constraint equations, the hierarchical positioning controller 140 can accurately determine the target extension of the multi-stage lifting platform 121 during the vertical coarse positioning stage. This ensures that the composite detection terminal 130, after considering the platform height of the movable platform 110, the vertical projection height of the current posture of the multi-degree-of-freedom robotic arm 122, and the safe distance between the composite detection terminal 130 and the surface of the rotating wheel, is precisely located within the vertical height range of the preset detection window. Compared to fixed lifting commands or manual estimation of lifting height, this application calculates the target extension through dynamic geometric constraint equations. This allows for adaptation to variations in the vertical distance from the bearing surface 30 of different hydropower stations to the rotating wheel, different detection heights, and the coupling effect of different robotic arm configurations on the lifting height. This ensures the accuracy and consistency of the vertical coarse positioning, providing a reliable spatial prerequisite for subsequent end-effector fine positioning. Simultaneously, by... By controlling the lifting within a preset safety threshold range, the excessive lifting of the multi-stage lifting platform 121 is avoided, which could cause the composite detection terminal 130 to collide with the rotating wheel, thus ensuring safety during the lifting process.

[0099] It should be understood that, unlike traditional fixed-base robotic arms, the base of the multi-degree-of-freedom robotic arm 122 provided in this embodiment is not fixed, but moves circumferentially with the movable platform 110. (Changes) and the lifting movement of the multi-stage lifting platform 121 ( (Changes) and changes dynamically. Therefore, the base transformation matrix It changes in real time and needs to be dynamically substituted into the positive kinematics mapping.

[0100] S132: Based on the current pose and point cloud data, calculate the second pose deviation between the current pose and the expected pose of the turbine runner surface under test.

[0101] The hierarchical positioning controller 140 controls a 3D topography scanner to acquire point cloud data of the surface of the turbine runner 20 under test. After acquiring the point cloud data, the hierarchical positioning controller 140 identifies target detection points on the surface of the turbine runner 20 under test (e.g., the center point of the water-dividing blade on the water bucket surface or a pre-set detection reference point) based on the point cloud data, and determines the desired pose of the target detection point, including the desired position. (3D spatial coordinates of the target detection point) and desired pose (The direction of the normal to the surface of the wheel at the target detection point).

[0102] Then, the hierarchical positioning controller 140 calculates the second pose deviation between the current pose and the desired pose determined in step S131.

[0103] In some embodiments, step S140 above includes the following sub-steps S141 to S145.

[0104] S141: Determine the position compensation amount based on the second position deviation in the second pose deviation; S142: Determine the attitude compensation amount based on the second attitude deviation in the second pose deviation.

[0105] Position compensation amount This represents the three-dimensional spatial position deviation between the current position and the desired position of the composite detection terminal 130; the attitude compensation amount represents the current pointing direction of the composite detection terminal 130 and the desired attitude (i.e., the direction of the normal to the surface of the turbine runner). The attitude deviation between ( ).

[0106] S143: Based on the position compensation and attitude compensation, solve for the joint compensation values ​​of each joint in the position control group and the attitude control group of the multi-level spatial positioning mechanism using inverse kinematics. The hierarchical positioning controller 140 calculates the position compensation values ​​determined in steps S141 and S142 respectively. And attitude compensation amount, combined with the current base transformation matrix The joint compensation required for each joint of the multi-degree-of-freedom robotic arm 122 is solved by inverse kinematics. .

[0107] S144: Based on the joint compensation amounts in the position control group and the attitude control group, the target position command for each joint in the multi-level spatial positioning mechanism is generated. The hierarchical positioning controller 140 generates the target position command (current position plus compensation amount) for each joint in the position control group based on the joint compensation amounts of each joint in the position control group; simultaneously, it generates the target position command for each joint in the attitude control group based on the joint compensation amounts of each joint in the attitude control group. The target position command is the target angle value that each joint needs to reach.

[0108] S145: According to the target position command, drive each joint in the multi-level spatial positioning mechanism to move separately, so that the composite detection terminal is aligned with the surface of the turbine runner under test. The hierarchical positioning controller 140 sends the target position command generated in step S144 to the servo driver of each joint in the multi-level spatial positioning mechanism 120, driving each joint to move to the target position, so that the composite detection terminal 130 is accurately aligned with the surface of the turbine runner 20 under test.

[0109] The embodiments of this application introduce a dynamic base transformation matrix into the positive kinematics mapping. The circumferential position of the movable platform 110 and the target extension of the multi-stage lifting platform 121 are used as necessary parameters to determine the current pose of the composite detection terminal 130. The inverse kinematics solution is based on the position compensation and attitude compensation between the current pose and the desired pose, and solves the joint compensation of the position control group and the attitude control group respectively. The forward kinematics mapping provides accurate feedback on where we are now, and the inverse kinematics solution provides the driving instructions on how to reach the target. Together, they form a complete closed loop for fine end-effector localization, enabling the composite detection terminal 130 to further converge from centimeter-level accuracy after global coarse localization to millimeter-level accuracy.

[0110] In some embodiments, step S145 above includes the following sub-steps S1451 and S1452.

[0111] S1451: Drive each joint in the position control group to perform position compensation one by one according to the preset sequence, and lock the joints in the position control group that have completed position compensation; S1452: After all joints in the position control group are locked, the attitude control group adjusts the attitude of the composite detection terminal one by one so that the composite detection terminal is aligned with the surface of the turbine runner under test.

[0112] After obtaining the target position commands for each joint through inverse kinematics calculation, the hierarchical positioning controller 140 does not simultaneously drive all joints to move. Instead, it employs a control strategy of sequential compensation across single axes in groups. Taking a six-DOF serial robotic arm as an example, it includes six rotary joints. .

[0113] The position control group includes the first three of the six joints ( The hierarchical positioning controller 140 is used to control the spatial position of the composite detection terminal 130 in the global coordinate system, i.e., the X, Y, and Z coordinates. The hierarchical positioning controller 140 operates according to a preset joint sequence (e.g., ...). The system sequentially drives each joint to move to the angle indicated by its corresponding target position command. After all joints in the position control group are locked, the spatial position of the composite detection terminal 130 has accurately reached the target position.

[0114] The attitude control group includes the last three joints ( This is used to control the pointing attitude of the composite detection terminal 130, namely pitch angle, yaw angle, and roll angle. The hierarchical positioning controller 140 follows a preset joint sequence (e.g., The system sequentially drives each joint to move to the angle indicated by its corresponding target position command. Each joint in the attitude control group completes attitude adjustment and locks one by one until the pointing attitude of the composite detection terminal 130 is consistent with the normal direction of the target detection point on the surface of the turbine runner 20 under test, that is, the composite detection terminal 130 is precisely aligned with the surface of the turbine runner 20 under test.

[0115] In this embodiment, the hierarchical positioning controller 140 employs grouped single-axis sequential compensation control logic to drive each joint of the multi-degree-of-freedom robotic arm 122. Compared to traditional multi-axis simultaneous linkage schemes, single-axis sequential compensation uses a decoupling strategy of compensating and locking one joint at a time. Once the compensation of each joint is completed, the joint is locked, allowing the inverse kinematics solution of subsequent joints to be performed based on the locked, determined state. This eliminates coupling errors in multi-axis linkage, improves positioning accuracy, and simplifies the control algorithm.

[0116] As a concrete example, Figure 5 This is a schematic flowchart illustrating a precise end-effector positioning method provided in an embodiment of this application. Please refer to [link / reference]. Figure 5 The precise positioning process for the end point includes the following steps S501 to S507.

[0117] S501: End-efficiency fine positioning begins. The hierarchical positioning controller 140 determines that the global coarse positioning has been completed (the movable platform 110 is located within the circumferential position range of the preset detection window, and the composite detection terminal 130 is located within the vertical height range of the preset detection window), and starts the end-efficiency fine positioning process.

[0118] S502: 3D key point detection. The hierarchical positioning controller 140 controls the 3D topography scanner to acquire point cloud data of the surface of the turbine runner 20 under test.

[0119] S503: Position compensation calculation. The hierarchical positioning controller 140 calculates the position deviation between the current position of the composite detection terminal 130 and the desired position of the surface of the turbine runner 20 under test, including position compensation and attitude compensation.

[0120] S504: Joint compensation calculation. Based on the position compensation and attitude compensation determined in step S503, and combined with the current base transformation matrix, the hierarchical positioning controller 140 solves for the joint compensation in the position control group and the joint compensation in the attitude control group of the multi-level spatial positioning mechanism 120 through inverse kinematics.

[0121] S505: Single joint compensation locking. The hierarchical positioning controller 140, based on the joint compensation amount obtained in step S504, uses single-axis sequential compensation control logic to drive each joint in the multi-level spatial positioning mechanism 120 to move separately.

[0122] S506: Determine if joint compensation is complete. If not, repeat step S505; if yes, proceed to step S507. The hierarchical positioning controller 140 determines whether all joints in the multi-level spatial positioning mechanism 120 have been compensated and locked. If there are still joints that have not been compensated (including joints in the position control group that have not been compensated or joints in the attitude control group that have not been attitude adjusted), return to step S505 and continue to perform the single-joint compensation and locking operation for the next joint. If all joints have been compensated and locked, proceed to step S507.

[0123] S507: Fine positioning completed, begin water bucket inspection. The graded positioning controller 140 confirms that all joints have been compensated and locked, and the deviation between the current pose of the composite inspection terminal 130 and the desired pose of the surface of the turbine runner 20 under test is close to zero, thus completing the end-effector fine positioning. Subsequently, the graded positioning controller 140 controls the composite inspection terminal 130 to perform in-situ inspection of the surface of the turbine runner 20 under test, that is, controls the three-dimensional topography scanner to collect three-dimensional point cloud data of the runner surface for subsequent defect identification and analysis.

[0124] In actual waterwheel operating environments, adverse factors such as abnormal lighting (e.g., localized strong light causing overexposure or missing point cloud data) and excessively high water mist concentration (e.g., residual water film or water mist on the rotor surface causing laser scattering) may exist, leading to a significant decrease in the quality of point cloud data acquired by the 3D topography scanner. To address this, the in-situ inspection robot 10 provided in this embodiment also possesses environmental adaptability, automatically switching to a backup positioning mode when the 3D topography scanner is subjected to environmental interference, ensuring the continuity and reliability of end-effector precision positioning.

[0125] In some embodiments, the method further includes the following steps S160 to S200.

[0126] S160: Measurement confidence level of point cloud data acquired by the 3D topography scanner. Measurement confidence level is used to evaluate the quality and reliability of the currently acquired point cloud data. It can be generated by the 3D topography scanner based on indicators such as point cloud density, signal-to-noise ratio, and registration residual, or it can be obtained by the hierarchical positioning controller 140 through quality analysis of the point cloud data.

[0127] S170: If the measurement confidence level is lower than a preset threshold, switch to the backup positioning mode guided by the vision imaging module. When the measurement confidence level is lower than the preset threshold, it indicates that the quality of the point cloud data currently acquired by the 3D topography scanner is insufficient to support reliable end-effector precision positioning. The hierarchical positioning controller 140 automatically switches the guidance source from the 3D topography scanner to the vision imaging module, that is, the composite detection terminal 130 switches to the backup positioning mode guided by the vision imaging module.

[0128] S180: In standby positioning mode, acquire the deviation between the current image features acquired by the visual imaging module and the target image features. Target image features These are pre-stored template image features of the surface of the turbine runner 20 to be tested, such as a standard view template generated by rendering a CAD model of the turbine chamber, or reference image features of the same detection location acquired in advance under good lighting conditions. Current image features and / or target image features It can include salient visual feature points such as corners, edges, and contours.

[0129] S190: Converts deviations into joint speed control commands for a multi-level spatial positioning mechanism. This conversion is based on the principle of visual servo control and is achieved through a composite mapping between the image Jacobian matrix and the robotic arm Jacobian matrix.

[0130] The hierarchical positioning controller 140 can be based on the image Jacobian matrix. Visual servo control is executed, converting image feature deviations into joint speed control commands. The principle of this visual servo control can be expressed by the following formula (5): Formula (5): .

[0131] in, The joint velocity control command vector of the multi-level spatial positioning mechanism 120 represents the motion velocity that each joint needs to execute, in radians per second (rad / s). It is the pseudo-inverse of the Jacobian matrix of the multi-degree-of-freedom robotic arm 122. It is the pseudo-inverse of the image Jacobian matrix. This refers to the current image feature vector acquired in real time by the visual imaging module, such as the set of corner pixel coordinates within the target detection area in the current image. The target image feature vector is the set of pixel coordinates of corresponding feature points in the pre-stored template image. The gain coefficient is used to adjust the convergence speed of visual servo control.

[0132] S200: Based on joint speed control commands, it drives the movement of each joint in the multi-level spatial positioning mechanism so that the composite detection terminal is aligned with the surface of the turbine runner under test.

[0133] The graded positioning controller 140 controls the joint speed according to the joint speed control command generated in step S190. The multi-level spatial positioning mechanism 120 (specifically, the multi-degree-of-freedom robotic arm 122) is driven to move its joints at a specified speed to adjust the pose of the composite detection terminal 130, so that the current image features gradually approach the target image features, that is, the image feature deviation (s_target - s) gradually decreases to near zero, thereby maintaining the alignment state between the composite detection terminal 130 and the surface of the turbine runner 20 under test.

[0134] Once environmental conditions are restored (e.g., water mist dissipates or supplemental lighting takes effect), if the measurement confidence level recovers to a level higher than or equal to the preset threshold, the hierarchical positioning controller 140 automatically switches back to the main positioning mode guided by the 3D topography scanner, that is, the point cloud data collected by the 3D topography scanner is used again as the main data source for fine positioning of the end point.

[0135] The environmentally adaptive backup positioning mode provided in this embodiment automatically switches to visual servo control guided by the vision imaging module when the point cloud data quality deteriorates by real-time monitoring of the measurement confidence level of the 3D topography scanner. It utilizes image feature deviations to drive the movement of the multi-level spatial positioning mechanism 120 to maintain alignment, enabling the in-situ inspection robot 10 to autonomously adapt to harsh environments. Compared to solutions relying on a single sensor, this embodiment improves the system's robustness and reliability in complex waterwheel chamber environments through the coordinated operation of the 3D topography scanner and the vision imaging module.

[0136] As a concrete example, such as Figure 6 As shown in the embodiment of this application, an in-situ detection method includes the following steps S601 to S609.

[0137] Before starting in-situ detection, the in-situ detection robot 10 is initialized. The hierarchical positioning controller 140 is powered on and starts up, completes the self-test of each module and the establishment of communication links, and loads the preset path information and preset detection window parameters.

[0138] S601: The movable platform is in operation. The hierarchical positioning controller 140 controls the movable platform 110 to perform global coarse positioning based on a preset path and environmental visual features to determine the current position of the movable platform 110 on the bearing surface 30.

[0139] S602: Has the designated position been reached? The hierarchical positioning controller 140 determines whether the movable stage 110 has entered the circumferential position range of the preset detection window based on the first pose deviation between the current position of the movable stage 110 and the preset detection window. If the determination result is negative, the system returns to step S601 and continues to control the movement of the movable stage 110; if the determination result is positive, the system executes step S603.

[0140] S603: Multi-stage lifting platform operation. The graded positioning controller 140 controls the multi-stage lifting platform 121 to lift the composite detection terminal 130 in the vertical direction, so that the composite detection terminal 130 moves to the vicinity of the height of the turbine runner 20 to be tested.

[0141] S604: Has the specified height limit been reached? This determines whether the composite detection terminal 130 is within the vertical height range of the preset detection window. The hierarchical positioning controller 140 determines whether the multi-stage lifting platform 121 has reached the target extension amount and whether the distance between the composite detection terminal 130 and the surface of the turbine runner 20 under test meets the preset safety threshold, based on the lifting position feedback from the absolute encoder and the distance detected by the laser rangefinder. If the determination result is negative, the process returns to step S603 and continues to control the lifting of the multi-stage lifting platform 121; if the determination result is positive, step S605 is executed.

[0142] S605: Multi-degree-of-freedom robotic arm operation. The hierarchical positioning controller 140 controls the multi-degree-of-freedom robotic arm 122 to perform end-effector precision positioning based on the point cloud data of the surface of the turbine runner 20 under test collected by the 3D topography scanner, determines the second pose deviation between the current pose of the composite detection terminal 130 and the surface of the turbine runner 20 under test, and solves the joint compensation amount of each joint through inverse kinematics.

[0143] S606: Determine the initial position. The hierarchical positioning controller 140 determines whether the second pose deviation between the current pose and the desired pose of the composite detection terminal 130 has converged to the preset accuracy range. If the determination result is negative, return to step S605 and continue to control the multi-degree-of-freedom robotic arm 122 to perform end-effector precision positioning; if the determination result is positive, execute step S607.

[0144] S607: The composite inspection terminal is in operation. The hierarchical positioning controller 140 controls the three-dimensional topography scanner in the composite inspection terminal 130 to perform three-dimensional topography scanning on the surface of the turbine runner 20 under test, and acquire three-dimensional point cloud data of the runner surface.

[0145] S608: Is the detection complete? The hierarchical positioning controller 140 determines whether the current detection position is the preset last detection position. If the determination result is no, it controls the movable stage 110 to move to the next preset detection window and returns to step S607; if the determination result is yes, it executes step S609.

[0146] S609: In-situ inspection robot initialization. After the inspection process ends, the hierarchical positioning controller 140 controls the movable platform 110 to return to the safe area, controls the multi-stage lifting platform 121 to retract to the lowest position, and controls the multi-degree-of-freedom robotic arm 122 to retract to the safe posture.

[0147] Through the above steps, a two-level coupled positioning system combining global coarse positioning and end-effector fine positioning was achieved for the in-situ inspection robot 10, ensuring the accuracy of the inspection results data.

[0148] The specific implementation of each of the above steps can be found in the description of the relevant embodiments above, and will not be repeated here.

[0149] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0150] Therefore, embodiments of this application provide a computer-readable storage medium storing a computer program thereon, the computer program being loaded by a processor to execute the steps described in the above-described method embodiments of this application. For example, the computer program being loaded by a processor can execute the steps of any of the above-described methods.

[0151] For details on the implementation of each of the above operations / steps, please refer to the previous examples, which will not be repeated here.

[0152] The computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0153] Since the computer program stored in the computer-readable storage medium can execute the steps in any of the above method embodiments provided in the embodiments of this application, the beneficial effects that the methods described in any of the above method embodiments can achieve can be realized, as detailed in the preceding embodiments, and will not be repeated here.

[0154] This application also provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the methods provided in the various optional implementations of the above embodiments.

[0155] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0156] The foregoing has provided a detailed description of an in-situ detection robot, an in-situ detection method, and a storage medium provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. An in-situ inspection robot, characterized in that, The in-situ detection robot includes: A movable platform is set on the bearing surface below the turbine runner to be tested; A multi-level spatial positioning mechanism is mounted on the movable platform; A composite detection terminal, located at the end of the multi-level spatial positioning mechanism, includes a three-dimensional topography scanner. A hierarchical positioning controller is communicatively connected to the movable platform and the multi-level spatial positioning mechanism, respectively. The hierarchical positioning controller is used for: The movable platform is controlled to perform global coarse positioning based on a preset path and environmental visual features to determine the current position of the movable platform on the bearing surface; Based on the first pose deviation between the current position and the preset detection window, the movable platform is controlled to move to the preset detection window; After the movable platform reaches the preset detection window, the multi-level spatial positioning mechanism is controlled to perform end-point fine positioning based on the point cloud data of the turbine runner surface to be tested collected by the three-dimensional topography scanner, and to determine the second pose deviation between the current pose of the composite detection terminal and the turbine runner surface to be tested. Based on the second pose deviation, the movement of the multi-level spatial positioning mechanism is controlled so that the composite detection terminal is aligned with the surface of the turbine runner to be tested. The composite detection terminal is controlled to perform in-situ detection on the turbine runner under test.

2. The in-situ inspection robot according to claim 1, characterized in that, The multi-level spatial positioning mechanism includes: A multi-stage lifting platform is vertically mounted on the movable platform; A multi-degree-of-freedom robotic arm is positioned at the top of the multi-stage lifting platform; the composite detection terminal is positioned at the end of the multi-degree-of-freedom robotic arm. The hierarchical positioning controller is further used for: The multi-stage lifting platform is controlled to lift the composite detection terminal in the vertical direction, so that the composite detection terminal is moved to the height of the turbine runner under test; The multi-degree-of-freedom robotic arm is controlled to adjust the current pose of the composite detection terminal so that the composite detection terminal is aligned with the surface of the turbine runner to be tested.

3. The in-situ inspection robot according to claim 2, characterized in that, The multi-degree-of-freedom robotic arm includes multiple joints; The hierarchical positioning controller is further used for: Each joint in the drive position control group is compensated for position one by one according to a preset sequence, and the joint in the position control group that has completed the position compensation is locked. After all joints in the position control group are locked, each joint in the drive attitude control group is adjusted one by one to adjust the attitude of the composite detection terminal so that the composite detection terminal is aligned with the surface of the turbine runner to be tested.

4. The in-situ inspection robot according to claim 2, characterized in that, The multi-stage lifting platform is equipped with an electronic soft limit device, which includes an absolute encoder and a laser rangefinder. The hierarchical positioning controller is further used for: The absolute encoder is controlled to provide feedback on the lifting position of the multi-stage lifting platform in order to determine whether the multi-stage lifting platform has reached the preset travel boundary. The laser ranging sensor is controlled to detect the distance between the composite detection terminal and the surface of the turbine runner under test, so as to determine whether the distance is less than a preset safety threshold. When the lifting position reaches the preset travel boundary, or when the distance is less than the preset safety threshold, the operation of the multi-stage lifting platform's drive is restricted.

5. The in-situ inspection robot according to claim 1, characterized in that, The composite detection terminal further includes a visual imaging module; both the three-dimensional topography scanner and the visual imaging module are located at the end of the multi-level spatial positioning mechanism. The hierarchical positioning controller is further used for: The measurement confidence level of the point cloud data acquired by the 3D topography scanner is monitored; If the measurement confidence level is lower than a preset threshold, switch to the backup positioning mode guided by the visual imaging module; In the backup positioning mode, the deviation between the current image features acquired by the visual imaging module and the target image features is obtained; The deviation is converted into joint speed control commands for the multi-level spatial positioning mechanism; Based on the joint speed control command, each joint in the multi-level spatial positioning mechanism is driven to move so that the composite detection terminal is aligned with the surface of the turbine runner to be tested.

6. The in-situ inspection robot according to claim 1, characterized in that, The robot also includes a dual-redundant energy supply module, which comprises: The primary power source is configured to connect to an external power supply via a drag cable; Backup power is configured as an onboard energy storage battery; The hierarchical positioning controller is further used for: Based on the motor output torque and angular velocity of each joint in the multi-level spatial positioning mechanism, the preset safe return time, the power consumption of the robot's control system, and the preset safety margin coefficient, the minimum backup power required for the backup energy source is determined.

7. An in-situ detection method, characterized in that, Applied to the in-situ detection robot as described in any one of claims 1 to 6; the in-situ detection method includes: The movable platform is controlled to perform global coarse positioning based on a preset path and environmental visual features to determine the current position of the movable platform on the bearing surface; Based on the first pose deviation between the current position and the preset detection window, the movable platform is controlled to move to the preset detection window; After the movable platform reaches the preset detection window, the multi-level spatial positioning mechanism is controlled to perform end-point fine positioning based on the point cloud data of the turbine runner surface to be tested collected by the three-dimensional topography scanner, and to determine the second pose deviation between the current pose of the composite detection terminal and the turbine runner surface to be tested. Based on the second pose deviation, the movement of the multi-level spatial positioning mechanism is controlled so that the composite detection terminal is aligned with the surface of the turbine runner to be tested. The composite detection terminal is controlled to perform in-situ detection on the turbine runner under test.

8. The in-situ detection method according to claim 7, characterized in that, The control mechanism for the multi-level spatial positioning system performs end-effector precision positioning based on the point cloud data of the turbine runner surface acquired by the 3D topography scanner, and determines the second pose deviation between the current pose of the composite detection terminal and the turbine runner surface, including: Based on the current pose of each moving component in the multi-level spatial positioning mechanism, the current pose of the composite detection terminal in the global coordinate system is determined by positive kinematic mapping; wherein, the current pose is determined by the target extension of the multi-level lifting platform, the circumferential position of the movable platform, and the current joint angle of the multi-degree-of-freedom robotic arm; the target extension is determined based on at least one of the following: the platform height of the movable platform, the vertical projection height of the multi-degree-of-freedom robotic arm under the current joint angle, and the distance between the composite detection terminal and the surface of the turbine runner under test; Based on the current pose and the point cloud data, the second pose deviation between the current pose and the desired pose of the turbine runner surface under test is calculated.

9. The in-situ detection method according to claim 8, characterized in that, The step of controlling the movement of the multi-level spatial positioning mechanism based on the second pose deviation, so that the composite detection terminal is aligned with the surface of the turbine runner under test, includes: The position compensation amount is determined based on the second position deviation in the second pose deviation; The attitude compensation amount is determined based on the second attitude deviation in the second pose deviation. Based on the position compensation amount and the attitude compensation amount, the compensation amounts of each joint in the position control group and the compensation amounts of each joint in the attitude control group of the multi-level spatial positioning mechanism are solved by inverse kinematics. Based on the compensation amounts of each joint in the position control group and the compensation amounts of each joint in the attitude control group, target position commands for each joint in the multi-level spatial positioning mechanism are generated respectively. According to the target position command, each joint in the multi-level spatial positioning mechanism is driven to move so that the composite detection terminal is aligned with the surface of the turbine runner to be tested.

10. The in-situ detection method according to claim 9, characterized in that, The step of driving each joint in the multi-level spatial positioning mechanism to move according to the target position command, so that the composite detection terminal is aligned with the surface of the turbine runner to be tested, includes: Each joint in the drive position control group is compensated for position one by one according to a preset sequence, and the joint in the position control group that has completed the position compensation is locked. After all joints in the position control group are locked, each joint in the drive attitude control group is adjusted one by one to adjust the attitude of the composite detection terminal so that the composite detection terminal is aligned with the surface of the turbine runner to be tested.

11. The in-situ detection method according to claim 7, characterized in that, The in-situ detection method further includes: The measurement confidence level of the point cloud data acquired by the 3D topography scanner is monitored; If the measurement confidence level is lower than a preset threshold, switch to a backup positioning mode guided by the visual imaging module; In the backup positioning mode, the deviation between the current image features acquired by the visual imaging module and the target image features is obtained; The deviation is converted into joint speed control commands for the multi-level spatial positioning mechanism; Based on the joint speed control command, each joint in the multi-level spatial positioning mechanism is driven to move so that the composite detection terminal is aligned with the surface of the turbine runner to be tested.

12. A computer-readable storage medium, characterized in that, It stores a computer program or instructions that, when executed by a processor, implement the steps in the in-situ detection method as described in any one of claims 7 to 11.

Images