Mechanical arm and rebound apparatus linkage control method and system based on impedance

Abstract

The invention discloses a mechanical arm and rebound apparatus linkage control method and system based on adaptive impedance, relates to the crossing field of a transformer substation detection technology and a robot control technology, and is suitable for concrete strength detection of reconstruction and extension projects of 500kV and above transformer substations. According to the method, a dynamic model containing environmental disturbance is established, an admittance control equation with disturbance compensation is constructed, multi-source positioning of a detection point is realized in combination with an adaptive weighted fusion algorithm, impedance parameters are dynamically adjusted based on an adaptive updating law, and a dynamic safety distance threshold model is constructed to realize charged body protection. And meanwhile, an improved A and DWA algorithm is adopted to complete arm-leg collaborative path planning. According to the method, the problems of weak anti-disturbance capability, fixed safety distance, low positioning precision and poor operation efficiency of a traditional linkage control method are solved, the detection efficiency is improved, and the precision, safety and intelligent level of detection operation in a complex environment of the transformer substation are remarkably enhanced.

Description

Technical Field

[0001] This invention belongs to the field of robot detection and control technology, specifically relating to an impedance-based linkage control method and system for a robotic arm and a rebound spring. Background Technology

[0002] In substation renovation and expansion projects, concrete structure strength testing is a key step in ensuring project quality. Traditional testing methods rely on manual operation of rebound hammers, which suffers from low efficiency, high labor intensity, and significant impact of human factors on testing accuracy.

[0003] With the development of robotics technology, the linkage control of robotic arms and rebound springs is gradually being applied to automated inspection. However, existing technologies still have many shortcomings, such as insufficient anti-disturbance capability, rigid positioning fusion method, single safety protection mechanism, inefficient arm-leg collaborative planning, and lack of adaptive optimization strategy, making it difficult to adapt to the complex operating environment of substations.

[0004] In the prior art, for example, Chinese utility model patent with publication number CN217845837U discloses a detection conversion mechanism for a rebound detection robot, which only achieves angle adjustment through mechanical structure and does not involve impedance control and disturbance compensation; another example is Chinese utility model patent with publication number CN218617226U, which discloses a rebound device for a UAV used in tunnel lining, which adopts a fixed safety distance threshold, and the UAV is limited by the space of the substation and electromagnetic interference, and cannot adapt to complex terrain.

[0005] Therefore, there is an urgent need for a linkage control method and system with adaptive impedance control, dynamic safety protection, and multi-source intelligent fusion to address the shortcomings of existing technologies. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and system for linkage control of a robotic arm and a rebound hammer based on adaptive impedance, so as to achieve high precision, high safety, high efficiency and high intelligence in concrete strength testing in substation renovation and expansion projects.

[0007] In a first aspect, the present invention proposes an impedance-based linkage control method for a robotic arm and a rebound spring, the method comprising the following steps: S1. Based on the rigid assembly relationship between the six-degree-of-freedom robotic arm and the rebound spring, a dynamic model with environmental disturbances is established. The initial values ​​of impedance parameters are determined by MATLAB discretization simulation combined with on-site calibration, and the admittance control equation with disturbance compensation is constructed. S2. Multi-source positioning is performed using a dual-light camera and a BeiDou positioning module. The three-dimensional coordinates of the detection point are calculated using an adaptive weighted fusion algorithm. Based on these coordinates, the quadruped robot is controlled to move to the detection point. S3. The robotic arm drives the rebound hammer to approach the detection point, and the concrete contact force is collected in real time by the end torque sensor. and contact torque The impedance parameters are dynamically adjusted based on the adaptive update law. S4. Construct a dynamic safe distance threshold model for energized bodies in substations through collaborative monitoring using lidar and electric field sensors: ,in This is the baseline safety distance for a 500kV substation. This represents the actual voltage of the charged body, in kV. Ambient humidity, in % % The velocity of the robotic arm's end effector is expressed in m / s. , This is a correction factor; when the actual distance When this occurs, the automatic power-off mechanism is triggered, and the robotic arm resets to a safe position within 1 second; and S5, Adopting improved A The algorithm plans the global path of the quadruped robot and introduces a dual-objective cost function of energy consumption and path. The local path of the robotic arm is adjusted through the DWA algorithm, and combined with the arm-leg cooperative constraint, obstacle avoidance and concrete rebound detection are achieved, and the detection data is recorded and uploaded.

[0008] Preferably, the dynamic model containing environmental disturbances in step S1 is: ,in For the equivalent mass of the rebound hammer, This refers to the joint damping coefficient of the robotic arm. The elastic coefficient, For the displacement of the rebound hammer, For the driving force of the robotic arm, The environmental disturbance loads of the substation include electromagnetic interference and equipment vibration; the initial values ​​of the impedance parameters include: inertial characteristics. Damping characteristics Stiffness characteristics The admittance control equation with disturbance compensation is: ,in , For the expected displacement of the rebound hammer, This is the estimated value of the disturbance load.

[0009] More preferably, the disturbance load The EKF state equation, obtained through extended Kalman filter EKF estimation, is as follows: The observation equation is: ,in For state vectors, Here is the state transition matrix. For the input matrix, For the observation matrix, These are process noise and observation noise, respectively.

[0010] Preferably, the three-dimensional coordinates of the detection point in step S2 are: ,in , For visual positioning weights, Weighting for BeiDou positioning The positioning results for the dual-light camera. The result is the BeiDou positioning result; the adaptive weighted fusion weights are updated through real-time variance. , ,in For update rate, , These are the real-time variances of visual positioning and BeiDou positioning, respectively.

[0011] Preferably, the impedance parameter in step S3 is: , , ,in This is an adjustment factor, ranging from 0.01 to 0.05. For the target contact force, make Fluctuation ≤ ±3N, rebound hammer attachment normal deviation ≤ 0.5°.

[0012] Preferably, step S3 further includes a torque compensation strategy: when At that time, the robotic arm's posture correction is triggered, and the correction angle is adjusted. , This ensures that the rebound hammer adheres stably to the concrete wall.

[0013] Preferably, the improved A described in step S5 The heuristic function of the algorithm is: ,in For Euclidean distance, For nodes To the The cost of obstacle avoidance for each obstacle The number of obstacles, The weighting coefficients are used to ensure that the arm-leg coordinated obstacle avoidance response time is ≤500ms and the obstacle avoidance success rate is ≥99%.

[0014] More preferably, the energy consumption-path bi-objective cost function in step S5 is: , For path length, To estimate energy consumption, the arm-leg coordination constraint is as follows: , For the joint angle of the robotic arm, This represents the robot's moving speed.

[0015] Preferably, it also includes a reinforcement learning optimization step: constructing a policy network based on the TD3 algorithm. ,state ,action Design a multidimensional reward function: ,in The function is used to ensure that the rebound hammer has a bonding success rate of ≥98%.

[0016] Secondly, embodiments of the present invention provide an impedance-based linkage control system for a robotic arm and a rebound spring, used to implement the linkage control method as described in any one of the first aspects, comprising: The admittance control equation establishment module is configured to establish a rigid assembly relationship between a six-degree-of-freedom robotic arm and a rebound device, establish a dynamic model with environmental disturbances, determine the initial values ​​of impedance parameters through MATLAB discretization simulation combined with on-site calibration, and construct admittance control equations with disturbance compensation. The mobile control module is configured to perform multi-source positioning using a dual-light camera and a BeiDou positioning module, calculate the three-dimensional coordinates of the detection point using an adaptive weighted fusion algorithm, and control the quadruped robot to move to the detection point based on the coordinates. The impedance parameter adjustment module is configured to enable the robotic arm to move the rebound hammer closer to the detection point, and to collect the concrete contact force in real time via an end torque sensor. and contact torque The impedance parameters are dynamically adjusted based on the adaptive update law. The safety distance monitoring and protection module is configured to construct a dynamic safety distance threshold model for energized bodies in substations through collaborative monitoring using lidar and electric field sensors. ,in This is the baseline safety distance for a 500kV substation. This represents the actual voltage of the charged body, in kV. Ambient humidity, in % % The velocity of the robotic arm's end effector is expressed in m / s. , This is a correction factor; when the actual distance When the automatic power-off mechanism is triggered, the robotic arm will reset to a safe position within 1 second; and; The path planning and data upload module is configured to use improved A The algorithm plans the global path of the quadruped robot and introduces a dual-objective cost function of energy consumption and path. The local path of the robotic arm is adjusted through the DWA algorithm, and combined with the arm-leg cooperative constraint, obstacle avoidance and concrete rebound detection are achieved, and the detection data is recorded and uploaded.

[0017] Compared with the prior art, the beneficial results of the present invention are as follows: (1) Significantly improved detection accuracy: Through disturbance compensation and adaptive impedance adjustment, the contact force fluctuation of the rebound hammer is ≤ ±3N, the attachment normal deviation is ≤ 0.5°, and the rebound value detection error is ≤ ±1, which meets the national standard high precision requirements; adaptive multi-source positioning makes the detection point positioning error ≤ 5cm.

[0018] (2) Comprehensive safety guarantee for operation: The dynamic safety distance threshold model is adapted to different voltages, environments and motion states. The automatic power-off and emergency braking mechanism reduces the risk of accidental contact with live parts to 0, which meets the safety operation standards of 500kV substations.

[0019] (3) Significantly improved work efficiency: Multi-target arm and leg collaborative planning makes obstacle avoidance response time ≤500ms and obstacle avoidance success rate ≥99%; after reinforcement learning optimization, the detection efficiency is improved, which is higher than the existing technology.

[0020] (4) Strong environmental adaptability: The quadruped robot’s obstacle crossing and hill climbing ability is adapted to the complex terrain of the substation; the adaptive algorithm can cope with environmental factors such as electromagnetic interference and humidity changes, and the attachment success rate is ≥98%.

[0021] (5) High level of intelligence: It can complete parameter self-adjustment, path self-planning and safety self-protection without manual intervention, reducing operation and maintenance costs and is suitable for large-scale substation renovation and expansion projects. Attached Figure Description

[0022] The accompanying drawings are included to provide a further understanding of the embodiments and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments and, together with the description, serve to explain the principles of the invention. Other embodiments and many anticipated advantages of the embodiments will be readily recognized as they become better understood through reference to the following detailed description. Elements in the drawings are not necessarily to scale. The same reference numerals refer to corresponding similar parts.

[0023] Figure 1 This is a schematic flowchart of an impedance-based linkage control method for a robotic arm and a rebound spring, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the architecture of the impedance-based linkage control system for a robotic arm and a rebound spring, according to an embodiment of the present invention. Detailed Implementation

[0024] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0025] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0026] This invention relates to the intersection of substation testing technology and robot control technology, specifically to a method and system for the linkage control of a robotic arm and a rebound hammer based on adaptive impedance, which is particularly suitable for automated rebound testing of concrete structure strength in the renovation and expansion projects of 500kV and above substations.

[0027] Firstly, Figure 1 An embodiment of the present invention discloses an impedance-based linkage control method for a robotic arm and a rebound spring, such as... Figure 1 As shown, the method includes the following steps: S1. Based on the rigid assembly relationship between the six-degree-of-freedom robotic arm and the rebound spring, a dynamic model with environmental disturbances is established. The initial values ​​of impedance parameters are determined by MATLAB discretization simulation combined with on-site calibration, and the admittance control equation with disturbance compensation is constructed. S2. Multi-source positioning is performed using a dual-light camera and a BeiDou positioning module. The three-dimensional coordinates of the detection point are calculated using an adaptive weighted fusion algorithm. Based on these coordinates, the quadruped robot is controlled to move to the detection point. S3. The robotic arm drives the rebound hammer to approach the detection point, and the concrete contact force is collected in real time by the end torque sensor. and contact torque The impedance parameters are dynamically adjusted based on the adaptive update law. S4. Construct a dynamic safe distance threshold model for energized bodies in substations through collaborative monitoring using lidar and electric field sensors: ,in This is the baseline safety distance for a 500kV substation. This represents the actual voltage of the charged body, in kV. Ambient humidity, in % % The velocity of the robotic arm's end effector is expressed in m / s. , This is a correction factor; when the actual distance When this occurs, the automatic power-off mechanism is triggered, and the robotic arm resets to a safe position within 1 second; and S5, Adopting improved A The algorithm plans the global path of the quadruped robot and introduces a dual-objective cost function of energy consumption and path. The local path of the robotic arm is adjusted through the DWA algorithm, and combined with the arm-leg cooperative constraint, obstacle avoidance and concrete rebound detection are achieved, and the detection data is recorded and uploaded.

[0028] In one specific embodiment, the method of the present invention optimizes the entire detection process by constructing an impedance control model with disturbance compensation, adaptive multi-source localization fusion, dynamic safety distance threshold, and multi-target arm-leg collaborative planning. The specific steps are as follows: Step 1: Establish the dynamic model and admittance control equations with disturbances. Based on the rigid assembly relationship between the six-degree-of-freedom robotic arm and the rebound spring, and considering environmental disturbances in the substation (electromagnetic interference, equipment vibration), a dynamic model is established: ; in The equivalent mass of the rebound hammer is 0.5 kg. The joint damping coefficient of the robotic arm is 2 N·s / m. The elastic modulus (taken as 100 N / m). For the displacement of the rebound hammer, For the driving force of the robotic arm, This refers to the environmental disturbance load of the substation (range -5N to 5N), including electromagnetic interference and equipment vibration.

[0029] Discretized simulations were performed using the MATLAB R2023b toolbox, combined with on-site calibration at a 500kV substation, to determine the initial values ​​of impedance parameters and inertial characteristics. Damping characteristics Stiffness characteristics .

[0030] To compensate for disturbance loads, admittance control equations with disturbance estimation are constructed: ; in, For displacement deviation, For the expected displacement of the rebound hammer, The disturbance load is estimated using an extended Kalman filter (EKF) to ensure that the disturbance compensation accuracy is ≤0.5N.

[0031] The EKF state equation is: The observation equation is: ,in For state vectors, Here is the state transition matrix. For the input matrix, For the observation matrix, These are process noise and observation noise, respectively.

[0032] Step 2: Adaptive Multi-Source Localization and Quadruped Robot Motion Control Images of the concrete inspection area were acquired using a dual-light camera (1920×1080 visible light resolution, 25x optical zoom), and the contours of the inspection points were extracted using the YOLOv8 algorithm. The original coordinates were obtained using a centimeter-level BeiDou positioning module, and the final 3D coordinates of the inspection points were calculated using an adaptive weighted fusion algorithm. ; in, , For visual positioning weights, Weighting for BeiDou positioning The positioning results for the dual-light camera. The result is the BeiDou positioning information.

[0033] Weight , The update is dynamically updated based on the real-time variance of visual and BeiDou positioning data, and the update rule is: ; ; in, In this embodiment, for the update rate... ; , These are the real-time variances of visual positioning and BeiDou positioning, respectively (calculated using the sliding window method).

[0034] The core control unit controls the quadruped robot (IP67 protection level, maximum obstacle clearance 18cm) to move to the detection point based on the fused coordinates, and avoids obstacles in real time through lidar during the movement.

[0035] Step 3: Adaptive Impedance Parameter Adjustment and Springback Detection The robotic arm moves the rebound spring closer to the detection point, and the end effector force / torque sensor (sampling rate 1kHz) collects data in real time. and .when When the impedance parameters deviate from the target range (50~80N), the impedance parameters are adjusted based on the adaptive update law. , , ; in, This is an adjustment factor, ranging from 0.01 to 0.05. For the target contact force, make Fluctuation ≤ ±3N, rebound hammer attachment normal deviation ≤ 0.5°. In this embodiment, .

[0036] when At that time, the robotic arm's posture correction compensation is triggered, correcting the angle. , This ensures the rebound hammer maintains a stable attachment to the concrete wall. In this embodiment, the correction angle... Ensure that the normal deviation of the rebound hammer is ≤0.5°.

[0037] when Once the data stabilizes within the target range, the rebound hammer automatically triggers detection, with a sampling frequency of 10Hz, and the raw data is transmitted to the core control unit in real time.

[0038] Step 4: Dynamic safety distance monitoring and protection LiDAR (mapping accuracy ±2cm) real-time acquisition of the distance between the robotic arm's end effector and the charged body. Electric field sensor collects electric field strength Combined with ambient humidity End-effector speed Construct a dynamic safe distance threshold model: ; in, This is the baseline safety distance for a 500kV substation. This represents the actual voltage of the charged body, in kV. Ambient humidity, in % % The velocity of the robotic arm's end effector is expressed in m / s. , This is a correction factor.

[0039] When the actual distance When the automatic power-off mechanism is triggered, the robotic arm will reset to a safe position within 1 second.

[0040] Step 5: Arm-Leg Coordination Path Planning and Data Upload Adopting improved A The algorithm plans the global path for a quadruped robot, introducing a dual-objective cost function of energy consumption and path optimization. Improvement A The heuristic function of the algorithm is: ; in, For Euclidean distance, For nodes To the The cost of obstacle avoidance for each obstacle The number of obstacles, The weighting coefficients are used to ensure that the arm-leg coordinated obstacle avoidance response time is ≤500ms and the obstacle avoidance success rate is ≥99%.

[0041] The energy consumption-path bi-objective cost function is: ; in, For path length, To estimate energy consumption. , These are the weighting coefficients. The local path of the robotic arm is adjusted using the DWA algorithm, combined with arm-leg coordination constraints: ; in, The maximum angular velocity of the robotic arm joint. This represents the robot's maximum moving speed.

[0042] After the test is completed, the core control unit records data such as rebound value, test location, and environmental parameters and uploads them to the substation management platform.

[0043] Optional steps: Reinforcement learning optimization A policy network is constructed based on the TD3 algorithm. ,state ,action Based on displacement deviation Contact force Safe distance Using the state as an example, and the impedance parameter adjustment as the action, design a multidimensional reward function: ; in, Using the indicator function, through offline training and online fine-tuning, the rebounder's adhesion success rate is ≥98%.

[0044] Further reference Figure 2 As an implementation of the methods shown in the above figures, this application provides an embodiment of an impedance-based linkage control system for a robotic arm and a rebound spring. This system embodiment is similar to... Figure 1 Corresponding to the method embodiments shown, the system can be specifically applied to various electronic devices.

[0045] Secondly, embodiments of the present invention also disclose an impedance-based linkage control system for a robotic arm and a rebound device, used to implement the linkage control method as described in any one of the first aspects, comprising: an admittance control equation establishment module 21, a movement control module 22, an impedance parameter adjustment module 23, a safety distance monitoring and protection module 24, and a path planning and data uploading module 25.

[0046] In one specific embodiment, the admittance control equation establishment module 21 is configured to establish a dynamic model with environmental disturbances based on the rigid assembly relationship between the six-degree-of-freedom robotic arm and the rebound device. Through MATLAB discretization simulation combined with on-site calibration, the initial values ​​of impedance parameters are determined, and the admittance control equation with disturbance compensation is constructed. The mobile control module 22 is configured to perform multi-source positioning using a dual-light camera and a Beidou positioning module, and calculate the three-dimensional coordinates of the detection point using an adaptive weighted fusion algorithm; based on these coordinates, it controls the quadruped robot to move to the detection point; the impedance parameter adjustment module 23 is configured to enable the robotic arm to drive the rebound spring towards the detection point, and to collect the concrete contact force in real time through an end torque sensor. and contact torque The impedance parameters are dynamically adjusted based on the adaptive update law. The safety distance monitoring and protection module 24 is configured to construct a dynamic safety distance threshold model for energized bodies in a substation through collaborative monitoring using lidar and electric field sensors. ,in This is the baseline safety distance for a 500kV substation. This represents the actual voltage of the charged body, in kV. Ambient humidity, in % % The velocity of the robotic arm's end effector is expressed in m / s. , This is a correction factor; when the actual distance When the automatic power-off mechanism is triggered, the robotic arm will reset to a safe position within 1 second; and; Path planning and data upload module 25, configured to use improved A The algorithm plans the global path of the quadruped robot and introduces a dual-objective cost function of energy consumption and path. The local path of the robotic arm is adjusted through the DWA algorithm, and combined with the arm-leg cooperative constraint, obstacle avoidance and concrete rebound detection are achieved, and the detection data is recorded and uploaded.

[0047] The functions and methods of the above modules correspond to each other, and will not be repeated here.

[0048] The system of the present invention integrates hardware and software to implement the above method, including a mobile carrier unit, an actuator unit, a multi-sensor fusion unit, a core control unit, a security protection unit, and a path planning unit. The functions of each unit are as follows: Mobile carrier unit: It adopts a quadruped robot as a mobile platform, integrates joint odometer, IMU and ground flatness sensor, and provides real-time feedback on motion status and road surface information; it is equipped with WiFi / WAPI dual-mode communication module to support high-speed communication with the core control unit.

[0049] Actuator unit: The end effector of the six-degree-of-freedom robotic arm is fixed with a rebound spring and a force / torque sensor. The robotic arm is equipped with a joint torque sensor to achieve force / position hybrid control. The rebound spring supports automatic trigger detection and has a sampling frequency of ≥10Hz.

[0050] Multi-sensor fusion unit: integrates dual-light camera, Beidou positioning, lidar, electric field sensor and temperature and humidity sensor to achieve multi-source positioning, environmental monitoring and safety perception; lidar supports dynamic obstacle trajectory prediction.

[0051] Core control unit: Station-side robot inspection integrated machine, equipped with 20-core CPU, 128G memory, RTX3060 GPU, built based on k8s and Docker container technology, with built-in adaptive impedance control, multi-source fusion, reinforcement learning and other algorithms, and compatible with domestic hardware and operating system.

[0052] Safety protection unit: includes a multi-dimensional safety monitoring module, an automatic power-off triggering unit, and an emergency braking module. It achieves graded protection based on dynamic safety distance thresholds and records safety warning logs.

[0053] Path planning unit: Integrated improvement A Compared with the DWA algorithm, it supports multi-objective optimization, dynamic obstacle avoidance and detection point sequence genetic algorithm optimization, shortening the overall detection path length.

[0054] The following describes a specific embodiment of the present invention in detail, using the concrete strength testing scenario of a 500kV outdoor substation renovation and expansion project as an example.

[0055] (I) Experimental Environment and Parameter Settings 1. Hardware equipment: Quadruped robot: Cloud Deep X30Pro, weight 59kg, payload 20kg, battery life 3h, protection level IP67; Robotic arm: Six-degree-of-freedom collaborative arm, 90cm reach, 3kg load capacity, and positioning accuracy ±0.05mm; Sensing devices: dual-light camera (Hikvision DS-2TD2617-3 / PA), Beidou BDS-3 positioning module (centimeter level), solid-state lidar (Robotech RS-LiDAR-M1), force / torque sensor (ATI Nano17). Core control unit: Station-end robot inspection integrated machine (20-core CPU, 128G memory, RTX3060 GPU).

[0056] 2. Environmental parameters: Substation live parts voltage: 500kV, ambient humidity 50%~80%, temperature -10℃~35℃; Test object: concrete wall panel (strength grade C30~C50), 100 test points; Target contact force: 60N (intermediate value), safety distance reference value: 0.7m.

[0057] (II) Implementation Steps 1. System Deployment and Initialization: Assemble the quadruped robot, robotic arm, rebound spring, and sensing devices. Deploy the integrated substation terminal in the substation's information management area and establish a connection via 5G+WAPI dual-mode communication. Complete device self-tests and parameter initialization, and calibrate the initial values ​​of impedance parameters. .

[0058] 2. Dynamic Model and Perturbation Estimation: A dynamic model was established using MATLAB simulation, with typical substation disturbance loads (±3N force fluctuations caused by electromagnetic interference) input, and EKF estimation was employed. The estimation error is ≤0.3N; the admittance control equation is constructed and burned into the core control unit.

[0059] 3. Multi-source localization and robot movement: Dual-light cameras acquire images of the concrete wall, the YOLOv8 algorithm identifies the contours of detection points, and the BeiDou module obtains the original coordinates; the final coordinates are calculated through adaptive weighted fusion (positioning error 4cm), and the core control unit sends movement commands to the quadruped robot. The robot then improves its A... The algorithm plans a path, avoiding construction equipment and electrified areas, and moves to the first detection point (taking 20 seconds).

[0060] 4. Adaptive Impedance Adjustment and Springback Detection: The robotic arm moves the springback device closer to the detection point, and the force / torque sensor collects the contact force. (Below the target value), the core control unit adjusts according to the update law. Up to 34.5, making Stabilize to 60N (fluctuation ±2N); during the testing process Trigger attitude correction, correct angle The normal deviation of the adhesive layer is 0.3°. The rebound hammer automatically triggers the detection, collects 10 sets of data, and uploads them.

[0061] 5. Safety Monitoring and Protection: LiDAR monitors the distance between the robotic arm's end effector and a charged object. With an ambient humidity of 70% and a robotic arm end-effector speed of 0.2 m / s, the dynamic safety distance was calculated using a dynamic safety distance model. ,because The warning is triggered, and the core control unit adjusts the robotic arm path to make... Restored to 0.8m.

[0062] 6. Arm-leg coordination and continuous detection: After completing the first detection point, the path planning unit adjusts the local path of the robotic arm through the DWA algorithm, and the quadruped robot moves to the next detection point. The arm-leg coordination obstacle avoidance response time is 300ms. Through reinforcement learning optimization, the attachment success rate of the subsequent 99 detection points reaches 99%, the total detection time is 3.3 hours, and the efficiency is 30 points / hour.

[0063] 7. Data Processing and Report Generation: The core control unit corrects the rebound value using a polynomial fitting algorithm. It generates a visual inspection report, marking the locations of three test points that fail to meet strength standards, providing decision support for project management.

[0064] (III) Verification of Implementation Results This experiment completed the testing of 100 concrete measuring points, and the verification results are as follows: Detection accuracy: rebound value error ≤ ±1, contact force fluctuation ≤ ±3N, attachment normal deviation ≤ 0.5°; Work safety: There are no safety warning upgrades throughout the process, the distance to live parts is always ≥0.7m, and the risk of accidental contact is 0; Operational efficiency: 30 points / hour, a 200% improvement over manual inspection (10 points / hour) and a 50% improvement over traditional robot inspection (20 points / hour); Environmental adaptability: The robot moves stably on gravel roads and 15cm steps, with a 100% success rate in obstacle avoidance.

[0065] In summary, the method and system of this invention can effectively adapt to the complex environment of 500kV substation renovation and expansion projects, and achieve high-precision, high-safety, and high-efficiency testing of concrete strength, possessing good practicality and promotional value.

[0066] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention is not limited to the specific combination of the above-described technical features, but also includes other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in this invention.

Claims

1. A method for linkage control of a robotic arm and a rebound spring based on impedance, characterized in that, The method includes the following steps: S1. Based on the rigid assembly relationship between the six-degree-of-freedom robotic arm and the rebound spring, a dynamic model with environmental disturbances is established. The initial values ​​of impedance parameters are determined by MATLAB discretization simulation combined with on-site calibration, and the admittance control equation with disturbance compensation is constructed. S2. Multi-source positioning is performed using a dual-light camera and a BeiDou positioning module. The three-dimensional coordinates of the detection point are calculated using an adaptive weighted fusion algorithm. Based on these coordinates, the quadruped robot is controlled to move to the detection point. S3. The robotic arm drives the rebound hammer to approach the detection point, and the concrete contact force is collected in real time by the end torque sensor. and contact torque The impedance parameters are dynamically adjusted based on the adaptive update law. S4. Construct a dynamic safe distance threshold model for energized bodies in substations through collaborative monitoring using lidar and electric field sensors: ,in This is the baseline safety distance for a 500kV substation. This represents the actual voltage of the charged body, in kV. Ambient humidity, in % % The velocity of the robotic arm's end effector is expressed in m / s. , This is a correction factor; when the actual distance When the automatic power-off mechanism is triggered, the robotic arm will reset to a safe position within 1 second; as well as S5, Adopting improved A The algorithm plans the global path of the quadruped robot and introduces a dual-objective cost function of energy consumption and path. The local path of the robotic arm is adjusted through the DWA algorithm, and combined with the arm-leg cooperative constraint, obstacle avoidance and concrete rebound detection are achieved, and the detection data is recorded and uploaded.

2. The linkage control method according to claim 1, characterized in that, The dynamic model involving environmental disturbances in step S1 is as follows: ,in For the equivalent mass of the rebound hammer, This refers to the joint damping coefficient of the robotic arm. The elastic coefficient, For the displacement of the rebound hammer, For the driving force of the robotic arm, Environmental disturbance loads in substations, including electromagnetic interference and equipment vibration; Initial values ​​for impedance parameters include: inertial characteristics Damping characteristics Stiffness characteristics The admittance control equation with disturbance compensation is: ,in , For the expected displacement of the rebound hammer, This is the estimated value of the disturbance load.

3. The linkage control method according to claim 2, characterized in that, The disturbance load The EKF state equation, obtained through extended Kalman filter EKF estimation, is as follows: The observation equation is: ,in For state vectors, Here is the state transition matrix. For the input matrix, For the observation matrix, These are process noise and observation noise, respectively.

4. The linkage control method according to claim 1, characterized in that, The three-dimensional coordinates of the detection point in step S2 are: ,in , For visual positioning weights, Weighting for BeiDou positioning. The positioning results for the dual-light camera. The result is the BeiDou positioning result; the adaptive weighted fusion weights are updated through real-time variance. , ,in For update rate, , These are the real-time variances of visual positioning and BeiDou positioning, respectively.

5. The linkage control method according to claim 1, characterized in that, The impedance parameters mentioned in step S3 are: , , ,in This is an adjustment factor, ranging from 0.01 to 0.

05. For the target contact force, make Fluctuation ≤ ±3N, rebound hammer attachment normal deviation ≤ 0.5°.

6. The linkage control method according to claim 1, characterized in that, Step S3 also includes a torque compensation strategy: when At that time, the robotic arm's posture correction is triggered, and the correction angle is adjusted. , This ensures that the rebound hammer adheres stably to the concrete wall.

7. The linkage control method according to claim 1, characterized in that, The improvement A described in step S5 The heuristic function of the algorithm is: ,in For Euclidean distance, For nodes To the The cost of obstacle avoidance for each obstacle The number of obstacles, The weighting coefficients are used to ensure that the arm-leg coordinated obstacle avoidance response time is ≤500ms and the obstacle avoidance success rate is ≥99%.

8. The linkage control method according to claim 7, characterized in that, The energy consumption-path bi-objective cost function mentioned in step S5 is: , For path length, To estimate energy consumption, the arm-leg coordination constraint is as follows: , For the joint angle of the robotic arm, This represents the robot's moving speed.

9. The linkage control method according to claim 1, characterized in that, It also includes reinforcement learning optimization steps: constructing a policy network based on the TD3 algorithm. ,state ,action Design a multidimensional reward function: ,in The function is used to ensure that the rebound hammer has a bonding success rate of ≥98%.

10. An impedance-based linkage control system for a robotic arm and a rebound spring, characterized in that, The method for implementing the linkage control method as described in any one of claims 1-9 includes: The admittance control equation establishment module is configured to establish a rigid assembly relationship between a six-degree-of-freedom robotic arm and a rebound spring, establish a dynamic model with environmental disturbances, determine the initial values ​​of impedance parameters through MATLAB discretization simulation combined with on-site calibration, and construct admittance control equations with disturbance compensation. The mobile control module is configured to perform multi-source positioning using a dual-light camera and a BeiDou positioning module, calculate the three-dimensional coordinates of the detection point through an adaptive weighted fusion algorithm, and control the quadruped robot to move to the detection point based on the coordinates. The impedance parameter adjustment module is configured to enable the robotic arm to move the rebound hammer closer to the detection point, and to collect the concrete contact force in real time via an end torque sensor. and contact torque The impedance parameters are dynamically adjusted based on the adaptive update law. The safety distance monitoring and protection module is configured to construct a dynamic safety distance threshold model for energized bodies in substations through collaborative monitoring using lidar and electric field sensors. ,in This is the baseline safety distance for a 500kV substation. This represents the actual voltage of the charged body, in kV. Ambient humidity, in % % The velocity of the robotic arm's end effector is expressed in m / s. , This is a correction factor; when the actual distance When the automatic power-off mechanism is triggered, the robotic arm will reset to a safe position within 1 second; and; The path planning and data upload module is configured to use improved A The algorithm plans the global path of the quadruped robot and introduces a dual-objective cost function of energy consumption and path. The local path of the robotic arm is adjusted through the DWA algorithm, and combined with the arm-leg cooperative constraint, obstacle avoidance and concrete rebound detection are achieved, and the detection data is recorded and uploaded.

Images