Nuclear power oil injection six-axis mechanical arm control method based on depth camera

Abstract

The invention provides a nuclear power oil injection six-axis mechanical arm control method based on a depth camera. The nuclear power oil injection six-axis mechanical arm control method based on the depth camera comprises the steps of environment perception and modeling, target recognition and positioning, path planning and obstacle avoidance, motion control and execution and real-time monitoring and feedback. Nuclear power oil injection is controlled based on the depth camera, so that the problems of low safety, low efficiency and insufficient accuracy in the prior art are solved, and accurate oil injection operation of the mechanical arm in the complex environment of the nuclear power station is achieved.

Description

Technical Field

[0001] The present application belongs to the field of robot control technology, and specifically relates to a method for controlling a six-axis robot arm for nuclear power oil injection based on a depth camera. Background Art

[0002] Nuclear power plants are high-energy environments, so the operational safety and reliability of equipment are of vital importance. Lubrication and oiling of equipment are important links to ensure the normal operation of equipment and extend its life. Traditional manual oiling methods have problems of high safety risks, low efficiency and insufficient accuracy. In terms of high safety risks, the environment of nuclear power plants is complex, with dangerous factors such as radiation, high temperature, and high pressure. Manual operation may expose personnel to dangerous environments, posing serious safety hazards. In terms of low efficiency, the efficiency of manual operation is limited by the proficiency and work intensity of personnel, and it is difficult to meet the efficient requirements of nuclear power plants for equipment maintenance. In terms of insufficient accuracy, manual oiling may have positioning errors and inaccurate oiling amounts, affecting the lubrication effect and operational reliability of the equipment.

[0003] At present, some mechanical equipment has been used in the maintenance work of nuclear power plants, such as bolt tension measurement, plug installation and high-level radioactive waste recovery, but it still has shortcomings in precise operation in complex environments and cannot completely replace manual labor.

[0004] For example, Patent 1 (publication number CN117911359A) discloses a vision-based 6D pose grasping method for a robotic arm, which uses a depth camera to obtain a target point cloud image, and after YOLOv5 network recognition, uses the SVD algorithm to calculate the target pose, and finally controls the robotic arm to grasp through hand-eye calibration. This patent mainly focuses on vision-based 6D pose grasping and does not involve oil injection operations in the nuclear power field.

[0005] Patent 2 (publication number CN117124332A) discloses a robot arm control method and system based on AI visual grasping. The system integrates radar and AI recognition technology to accurately locate the object to be grasped, and plans the operation route of the robot arm to avoid obstacles and improve operation safety. The patent emphasizes the combination of AI vision and radar, and is mainly used for grasping general objects, not for nuclear power oil injection operations.

[0006] Patent 3 (publication number CN112936275A) discloses a robotic arm grasping system and control method based on a depth camera. The system collects images through a depth camera, uses a deep learning algorithm to calculate the optimal grasping point position of the target object, and controls the robotic arm to grasp. This patent focuses on the grasping of general objects and does not involve specific applications in the nuclear power field.

[0007] Patent 4 (CN111881772A) discloses a multi-manipulator collaborative assembly method and system based on deep reinforcement learning: This method uses a multi-source heterogeneous sensor network and combines deep reinforcement learning to achieve collaborative assembly of multiple manipulators. This patent focuses on the collaborative assembly of multiple manipulators and does not involve the specific operation of nuclear power oil injection.

[0008] Patent 5 (CN114263840A) discloses an automatic oil injection device for the nuclear industry and a method for manufacturing an automatic oil injection system for the nuclear industry, and Patent 6 (CN106884923A) discloses a vacuum oil injection system and method for a hydraulic damper in a nuclear power plant. However, these patents mainly focus on the mechanical structure design of the oil injection device and the vacuum oil injection method, and do not involve a control method based on a depth camera. Although the use of a robot for operation is mentioned, the control method of the six-axis robot arm is not described in detail, especially the control strategy based on a visual sensor. Summary of the invention

[0009] In view of this, the present application is committed to providing a six-axis robotic arm control method for nuclear power oil injection based on a depth camera. By designing a method for controlling nuclear power oil injection based on a depth camera, the problems of low safety, low efficiency and insufficient accuracy existing in the prior art are solved, and the precise oil injection operation of the robotic arm in the complex environment of a nuclear power plant is realized.

[0010] The first aspect of the present application provides a nuclear power oil injection six-axis robot control method based on a depth camera, and the nuclear power oil injection six-axis robot control method based on a depth camera includes: an environmental perception and modeling step: using a three-dimensional laser depth camera with an anti-radiation design installed at the end of the robot arm to obtain real-time three-dimensional point cloud data of the robot arm's working area, and the control system uses voxel filtering and statistical filtering algorithms to preprocess the data to build a three-dimensional environment model; a target recognition and positioning step: the control system uses an improved PointNet++ deep learning network based on a self-attention mechanism and multi-scale feature aggregation to classify and segment the preprocessed point cloud data, and combines the CAD model of the equipment with an improved ICP algorithm for precise alignment to obtain the precise position and posture of the oil injection target; a path planning and obstacle avoidance step: the control system uses an improved RRT* algorithm based on the kinematic model and environmental model of the robot arm, combined with an artificial potential field method, to plan the optimal obstacle avoidance path that meets the robot arm's motion constraints; a motion control and execution step: the control system uses a model predictive control algorithm to convert the planned path into joint motion instructions of the robot arm, and combines a force / position hybrid control strategy to control the robot arm in real time to perform oil injection operations. Real-time monitoring and feedback steps: Use the depth camera to continuously obtain environmental data, detect environmental changes, dynamically update the three-dimensional model and path planning, and trigger the redundant safety mechanism to automatically stop the robot arm movement when an abnormality or emergency is detected.

[0011] In a specific implementation of the present application, a self-attention mechanism is introduced into the PointNet++ deep learning network to calculate the importance weight of each point in the point cloud and strengthen the feature expression of key feature points.

[0012] In a specific implementation of the present application, a feature-based weighted matching method is introduced into the improved ICP algorithm, and a KD tree and nearest neighbor search algorithm are used.

[0013] In a specific embodiment of the present application, in the path planning and obstacle avoidance steps, the cost function comprehensively considers the energy consumption, path length, smoothness and obstacle avoidance requirements of the robot arm, and uses a dynamic window method to predict and avoid moving obstacles.

[0014] In a specific implementation of the present application, in the motion control and execution steps, an adaptive control strategy based on force / position hybrid control is adopted to adjust control parameters in real time according to force feedback information.

[0015] The second aspect of the present application provides a computer device, which includes a processor and a memory. The processor is used to execute a nuclear power oil injection six-axis robot control method based on a depth camera in the first aspect of the present application. The memory is used to store executable instructions of the processor.

[0016] The third aspect of the present application provides a computer-readable storage medium on which executable instructions of a computer are stored. When the executable instructions are executed by a processor, a method for controlling a six-axis manipulator for nuclear power oil injection based on a depth camera according to the first aspect of the present application is implemented.

[0017] The fourth aspect of the present application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements a six-axis robot control method for nuclear power oil injection based on a depth camera according to the first aspect of the present application.

[0018] The beneficial effect of the technical solution of this application is that by combining the improved PointNet++ network and the ICP algorithm, the high-precision target recognition and positioning of the manipulator in the complex environment of the nuclear power plant is realized, and the precise oil injection operation is realized. The improved path planning and control algorithm enables the manipulator to complete the oil injection operation safely and efficiently. The control method of the six-axis manipulator for nuclear power oil injection based on the depth camera has the beneficial effects of improving safety, enhancing adaptability, improving efficiency and technical advantages. In terms of improving safety, it reduces the operation of personnel in high-risk environments and reduces the risk of radiation exposure. In terms of enhancing adaptability, the control method of the six-axis manipulator for nuclear power oil injection based on the depth camera can perceive environmental changes in real time, dynamically adjust the path, and adapt to the complex and changeable working environment of the nuclear power plant. In terms of improving efficiency, the control method of the six-axis manipulator for nuclear power oil injection based on the depth camera has a high degree of automation, and the oil injection operation time is reduced by about 50% compared with the manual method, which improves the equipment maintenance efficiency. In terms of technical advantages, the control method of the six-axis manipulator for nuclear power oil injection based on the depth camera is superior to the existing technology in terms of target recognition accuracy, path planning efficiency and control accuracy, and has significant technical advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Shown is a schematic structural diagram of a nuclear power oil injection robot provided in one embodiment of the present application.

[0020] Figure 2 The figure is a simplified flowchart of a method for controlling a six-axis robot arm for nuclear power oil injection based on a depth camera provided in one embodiment of the present application.

[0021] Figure 3 The figure shows an improved PointNet++ network structure diagram provided in one embodiment of the present application.

[0022] Figure 4 The figure shows an improved PointNet++ network structure diagram provided by another embodiment of the present application. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0024] At least one embodiment of the present application provides a method for controlling a six-axis robot arm for nuclear power oil injection based on a depth camera. The method for controlling a six-axis robot arm for nuclear power oil injection based on a depth camera can be implemented using a nuclear power oil injection robot. Figure 1, the nuclear power oil injection robot may include a depth camera, a control system and a robotic arm. Figure 2 , the nuclear power oil injection six-axis robot control method based on depth camera includes the following steps.

[0025] Environmental perception and modeling steps: Use a 3D laser depth camera with anti-radiation design installed at the end of the robotic arm to obtain real-time 3D point cloud data of the robotic arm's working area. The control system uses voxel filtering and statistical filtering algorithms to pre-process the data and construct a 3D environmental model.

[0026] Target recognition and positioning steps: The control system uses an improved PointNet++ deep learning network based on self-attention mechanism and multi-scale feature aggregation to classify and segment the preprocessed point cloud data. Combined with the CAD model of the equipment, the improved ICP algorithm is used for precise alignment to obtain the precise position and posture of the oil injection target.

[0027] Path planning and obstacle avoidance steps: The control system uses an improved RRT* (fast random tree) algorithm combined with an artificial potential field method based on the kinematic model of the robot and the environmental model to plan the optimal obstacle avoidance path that meets the motion constraints of the robot.

[0028] Motion control and execution steps: The control system uses the model predictive control (MPC) algorithm to convert the planned path into joint motion instructions of the robot arm, combined with the force / position hybrid control strategy, to control the robot arm to perform oil injection operations in real time.

[0029] Real-time monitoring and feedback steps: Use the depth camera to continuously obtain environmental data, detect environmental changes, dynamically update the three-dimensional model and path planning, and trigger the redundant safety mechanism to automatically stop the robot arm movement when an abnormality or emergency is detected.

[0030] It should be noted that the three-dimensional environment model may be a high-precision three-dimensional environment model. The three-dimensional laser depth camera may be a high-precision three-dimensional laser depth camera. The three-dimensional environment model may include the position and shape information of the equipment, obstacles and oil injection targets.

[0031] According to the technical solution provided by the embodiment of the present application, by combining the improved PointNet++ network and the ICP algorithm, the high-precision target recognition and positioning of the manipulator in the complex environment of the nuclear power plant is realized, and the precise oil injection operation is realized. The improved path planning and control algorithm enables the manipulator to complete the oil injection operation safely and efficiently. The control method of the six-axis manipulator for nuclear power oil injection based on the depth camera has the beneficial effects of improving safety, enhancing adaptability, improving efficiency and technical advantages. In terms of improving safety, it reduces the operation of personnel in high-risk environments and reduces the risk of radiation exposure. In terms of enhancing adaptability, the control method of the six-axis manipulator for nuclear power oil injection based on the depth camera can perceive environmental changes in real time, dynamically adjust the path, and adapt to the complex and changeable working environment of the nuclear power plant. In terms of improving efficiency, the control method of the six-axis manipulator for nuclear power oil injection based on the depth camera has a high degree of automation, and the oil injection operation time is reduced by about 50% compared with the manual method, which improves the equipment maintenance efficiency. In terms of technical advantages, the control method of the six-axis manipulator for nuclear power oil injection based on the depth camera is superior to the existing technology in terms of target recognition accuracy, path planning efficiency and control accuracy, and has significant technical advantages.

[0032] It should be noted that the improved PointNet++ deep learning network is used to classify and segment the preprocessed point cloud data. Figure 3 and Figure 4 As shown. In view of the characteristics of the nuclear power plant environment, the PointNet++ network has been improved as follows: 1. Data enhancement: CAD models of nuclear power equipment are used to generate diverse training samples to increase the network's adaptability to complex shapes and materials. 2. Multi-scale feature extraction: A multi-scale feature aggregation module is introduced to capture the local and global geometric features of the equipment and improve the recognition ability of complex structures. 3. Attention mechanism: An attention mechanism is added to the network to strengthen the focus on key feature points and improve recognition accuracy.

[0033] Combined with the CAD model of nuclear power plant equipment, the improved ICP (Iterative Closest Point) algorithm is used for precise registration. The improved ICP algorithm includes: 1. Weighted matching: Introducing point cloud feature weights, giving higher weights to key feature points to reduce mismatching. 2. Dynamic threshold adjustment: Dynamically adjust the distance threshold according to the matching error, filter out abnormal points, and improve the robustness of the registration. 3. Accelerated convergence: Use KD tree and nearest neighbor search algorithm to improve calculation efficiency and shorten convergence time.

[0034] The improvements of the improved RRT* (fast random tree) algorithm are: 1. Cost function optimization: Consider the energy consumption, path smoothness and obstacle avoidance requirements of the robot arm to optimize the path quality. 2. Dynamic obstacle avoidance: Introduce the dynamic window method to predict and avoid moving obstacles to ensure the real-time and safety of the path. 3. Constraint conditions: Comprehensively consider the joint angle limit, speed limit and load capacity of the robot arm to ensure the feasibility of the planned path.

[0035] The control system uses the model predictive control (MPC) algorithm to convert the planned path into the joint motion instructions of the robot arm. The specific performance includes: 1. Dynamic modeling: Establish the dynamic and kinematic models of the robot arm, considering factors such as joint friction, inertia and load changes. 2. Adaptive control: According to the force feedback and visual feedback information, the control parameters are adjusted in real time to compensate for the system error. 3. Precise oiling: During the oiling process, combined with the data of the force sensor, the oiling device is controlled to complete the oiling operation with appropriate force and position.

[0036] In at least one embodiment of the present application, a self-attention mechanism is introduced into the PointNet++ deep learning network to calculate the importance weight of each point in the point cloud and strengthen the feature expression of key feature points.

[0037] In at least one embodiment of the present application, in the improved ICP algorithm, a feature-based weighted matching method is introduced, and a KD tree and a nearest neighbor search algorithm are used. Thus, by introducing a feature-based weighted matching method, a higher matching weight is given to points with significant features. The computational efficiency is improved by using a KD tree and a nearest neighbor search algorithm.

[0038] In at least one embodiment of the present application, in the path planning and obstacle avoidance steps, the cost function comprehensively considers the energy consumption, path length, smoothness and obstacle avoidance requirements of the robot arm, and uses a dynamic window method to predict and avoid moving obstacles.

[0039] In at least one embodiment of the present application, in the motion control and execution steps, an adaptive control strategy based on force / position hybrid control is adopted to adjust control parameters in real time according to force feedback information. In this way, control parameters can be adjusted in real time according to force feedback information to ensure the accuracy of the oil injection operation.

[0040] The technical solution of the present invention is described in detail below in conjunction with specific embodiments.

[0041] 1. Experimental environment

[0042] Equipment: Nuclear power plant motor system

[0043] Robotic arm: six-axis industrial robot arm, load capacity 20kg, repeatability ±0.08mm

[0044] Depth camera: Industrial-grade 3D laser depth camera with radiation-resistant design, measurement accuracy ±0.5mm

[0045] Mobile obstacles: simulate mobile equipment or personnel that may appear in a nuclear power plant

[0046] 2. Experimental Procedure

[0047] (1) Environmental perception and modeling steps

[0048] Start the depth camera to scan the motor system and the surrounding environment in real time to obtain high-precision three-dimensional point cloud data.

[0049] The point cloud data is preprocessed using voxel filtering and statistical filtering algorithms to remove noise and outliers.

[0050] Build a high-precision 3D model of the environment including pipes, valves, and support structures.

[0051] (2) Target identification and positioning steps

[0052] The improved PointNet++ network is used to perform target classification and segmentation on the preprocessed point cloud data, and the valve position that needs oil injection is successfully identified.

[0053] By introducing data enhancement and attention mechanism, the recognition accuracy of targets in complex environments is improved.

[0054] Combined with the CAD model of the valve, the improved ICP algorithm is used for precise alignment, and the positioning error is less than ±1mm.

[0055] (3) Path planning and obstacle avoidance steps

[0056] According to the kinematic model of the robot arm, the improved RRT* algorithm is used to plan the optimal path from the initial position to the target valve.

[0057] Combined with the artificial potential field method, it avoids surrounding fixed obstacles such as pipes and supporting structures.

[0058] Since there are moving obstacles in the environment, the dynamic window method is used to predict and avoid them, and the path planning is updated in real time.

[0059] (4) Motion control and execution steps

[0060] The MPC algorithm is used to convert the planned path into joint motion instructions of the robot arm.

[0061] During the movement, the force feedback information of the robotic arm is monitored in real time to perform error compensation and dynamic adjustment.

[0062] Adopt adaptive control strategy based on force / position hybrid control to ensure the accuracy of oil injection operation.

[0063] (5) Real-time monitoring and feedback steps

[0064] The depth camera continuously acquires environmental data and monitors surrounding environmental changes and the location of moving obstacles.

[0065] When new obstacles or emergency situations are detected, the system immediately updates the environment model and path planning.

[0066] The redundant safety mechanism is triggered, the robotic arm automatically stops moving and sounds an alarm to ensure safety.

[0067] 3. Experimental results

[0068] Target recognition accuracy: In complex environments, the target recognition accuracy reaches over 98%.

[0069] Positioning accuracy: The positioning error of the oil injection target is controlled within ±1mm.

[0070] Oil injection operation time: The single oil injection operation time is reduced by about 60% compared with the manual method, which significantly improves the efficiency.

[0071] Radiation safety: The oil filling process no longer requires human involvement.

[0072] At least one embodiment of the present application also provides a computer device, which includes a processor and a memory. The processor is used to execute a method for controlling a six-axis robot arm for nuclear power oil injection based on a depth camera provided in any of the above embodiments of the present application. The memory is used to store executable instructions of the processor, such as application programs. The number of processors can be one or more. The application stored in the memory may include one or more modules, each of which corresponds to a set of instructions. In addition, the processor is configured to execute instructions to execute the above-mentioned method for controlling a six-axis robot arm for nuclear power oil injection based on a depth camera.

[0073] The computer device may also include a power supply component configured to manage power of the computer device, a wired or wireless network interface configured to connect the computer device to a network, and an input / output (I / O) interface. The computer device may operate based on an operating system stored in the memory, such as Windows Server 2003. TM , Mac OSX TM , Unix TM , Linux TM , FreeBSD TM or similar.

[0074] At least one embodiment of the present application further provides a computer-readable storage medium on which computer executable instructions are stored. When the executable instructions are executed by a processor, a method for controlling a six-axis manipulator for nuclear power oil injection based on a depth camera provided in any of the above embodiments of the present application is implemented.

[0075] A non-transitory computer-readable storage medium, when the instructions in the storage medium are executed by the processor of the above-mentioned computer device, enables the above-mentioned computer device to execute the above-mentioned nuclear power oil injection six-axis robot control method based on a depth camera. The nuclear power oil injection six-axis robot control method based on a depth camera is executed by an agent program.

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

[0077] At least one embodiment of the present application further provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements a method for controlling a six-axis robotic arm for nuclear power oil injection based on a depth camera provided in any of the above embodiments of the present application.

[0078] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a computer program product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the nuclear power oil injection six-axis robot control method based on a depth camera in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program check codes.

[0079] It should be noted that the combination of the various technical features in the embodiments of the present application is not limited to the combination described in the embodiments of the present application or the combination described in the specific embodiments, and all technical features described in the present application can be freely combined or combined in any way unless there is a contradiction between them.

[0080] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the term "comprising" only indicates that the steps and elements that have been clearly identified are included, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0081] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features.

[0082] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for controlling a six-axis manipulator for nuclear power oil injection based on a depth camera, characterized in that: include: Environmental perception and modeling steps: Use a radiation-resistant 3D laser depth camera installed at the end of the robotic arm to obtain real-time 3D point cloud data of the robotic arm's working area. The control system uses voxel filtering and statistical filtering algorithms to pre-process the data and build a 3D environmental model. Target recognition and positioning steps: The control system uses an improved PointNet++ deep learning network based on self-attention mechanism and multi-scale feature aggregation to classify and segment the pre-processed point cloud data. Combined with the CAD model of the nuclear power equipment, the improved ICP algorithm is used for precise registration to obtain the precise position and posture of the oil injection target. Path planning and obstacle avoidance steps: The control system uses the improved RRT* algorithm and the artificial potential field method based on the kinematic model and environmental model of the robot to plan the optimal obstacle avoidance path that meets the motion constraints of the robot; Motion control and execution steps: The control system uses a model predictive control algorithm to convert the planned path into joint motion instructions for the robot arm, combined with a force / position hybrid control strategy to control the robot arm in real time to perform oiling operations.

2. According to the method for controlling a six-axis manipulator for nuclear power oil injection based on a depth camera in claim 1, it is characterized in that: The self-attention mechanism is introduced into the PointNet++ deep learning network to calculate the importance weight of each point in the point cloud and strengthen the feature expression of key feature points.

3. According to the method for controlling a six-axis manipulator for nuclear power oil injection based on a depth camera in claim 1, it is characterized in that: In the improved ICP algorithm, a feature-based weighted matching method is introduced, and the KD tree and nearest neighbor search algorithm are used.

4. According to the method for controlling a six-axis manipulator for nuclear power oil injection based on a depth camera in claim 1, it is characterized in that: In the path planning and obstacle avoidance steps, the cost function comprehensively considers the energy consumption, path length, smoothness and obstacle avoidance requirements of the robot arm, and uses the dynamic window method to predict and avoid moving obstacles.

5. A method for controlling a six-axis manipulator for nuclear power oil injection based on a depth camera according to any one of claims 1 to 4, characterized in that: In the motion control and execution steps, an adaptive control strategy based on force / position hybrid control is adopted to adjust the control parameters in real time according to the force feedback information.

6. A computer device, characterized in that: include: A processor, configured to execute the method for controlling a six-axis manipulator for nuclear power oil injection based on a depth camera according to any one of claims 1 to 5; as well as A memory is used to store executable instructions of the processor.

7. A computer-readable storage medium having computer executable instructions stored thereon, characterized in that: When the executable instructions are executed by the processor, a method for controlling a six-axis nuclear power oil injection robot based on a depth camera according to any one of claims 1 to 5 is implemented.

8. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by the processor, a method for controlling a six-axis nuclear power oil injection robot based on a depth camera according to any one of claims 1 to 5 is implemented.

Images