Autonomous maintenance method for wheeled mobile laser cladding robot

Abstract

The invention discloses an autonomous maintenance method for a wheeled mobile laser cladding robot. The autonomous maintenance method comprises the following steps: step 1, identifying defects: identifying the defects by a wheeled mobile trolley according to a classification model; 2, after the defects are recognized, rust removal is conducted on the equipment; step 3, constructing a point cloud model; 4, calibrating a coordinate system; 5, generating a mechanical arm motion program based on a path planning algorithm; and 6, cladding path tracking control is conducted, specifically, the six-axis mechanical arm conducts cladding according to the generated motion program, and based on a model predictive control algorithm MPC, it is ensured that the mechanical arm conducts cladding according to the generated motion path. The maintenance efficiency is improved, and the labor cost is reduced.

Description

Technical Field

[0001] The invention belongs to the field of robots and relates to an autonomous maintenance method for a wheeled mobile laser cladding robot. Background Art

[0002] To ensure stable factory operation and promptly identify potential safety hazards, maintenance work is an indispensable part of daily operations. Traditional maintenance methods mainly rely on manual maintenance, which faces many limitations.

[0003] First, traditional maintenance methods are time-consuming. Due to the large number of devices, manual maintenance requires a lot of time and human resources, and is subject to certain subjectivity and the risk of human error.

[0004] Secondly, traditional maintenance methods have safety risks. In high-radiation areas and difficult-to-access equipment, manual maintenance poses safety hazards and can easily expose personnel to dangerous environments.

[0005] Furthermore, manual maintenance faces limitations. For example, some equipment is difficult to observe and inspect directly, requiring disassembly or the use of auxiliary equipment, which increases the difficulty and risk of maintenance.

[0006] The repair of some equipment requires laser cladding technology, and the laser cladding process requires large equipment, which is bulky and difficult to move. In addition, the laser emitted by the laser during the laser cladding process is a high-power, high-energy-density laser. The reflection, refraction and high-temperature radiation of the laser during the processing will cause harm to the health of the workers. At the same time, the air dust and arc light generated by the high temperature of the cladding material will also affect the nearby working environment.

[0007] Existing laser cladding technology requires large, bulky equipment, which can't complete the cladding process quickly. The cladding process involves complex steps. For example, manual inspection for defects is required, followed by disassembly. Some large equipment is difficult to disassemble and requires the use of auxiliary equipment, making the process even more complex.

[0008] Before laser cladding, the equipment needs to be derusted to make the surface smoother and easier to clad. Existing laser cladding technology uses sandpaper polishing to remove rust from equipment. For some larger equipment, it takes a lot of time to polish the equipment, which consumes too much time and is inefficient. Summary of the Invention

[0009] In order to overcome the shortcomings of the existing technology, the present invention provides an autonomous maintenance method for a wheeled mobile laser cladding robot, which improves maintenance efficiency and reduces labor costs.

[0010] The technical solution adopted by the present invention to solve its technical problem is:

[0011] A wheeled mobile laser cladding robot autonomous maintenance method comprises the following steps:

[0012] Step 1: Defect Identification: Collect images of equipment damage within the factory and classify them into different types. Train the classified defect samples using an extreme learning machine (ELM), forming a classification model. The wheeled mobile cart identifies the defects based on the classification model.

[0013] Step 2: After identifying the defects, the equipment is rusted. The process is as follows:

[0014] A wheeled mobile cart enters the factory and inspects the equipment in the factory according to the classification model described in step 1. When the cart identifies a defect in a piece of equipment, the six-axis robotic arm uses a laser rust removal head to remove rust from the surface of the equipment. After rust removal is completed, the end of the six-axis robotic arm switches to a laser cladding head.

[0015] Step 3: Construct a point cloud model: The wheeled mobile vehicle collects point cloud data of the equipment to be repaired based on the laser radar. The six-degree-of-freedom robotic arm moves linearly at a constant speed, allowing the laser to smoothly sweep the surface of the component. The coordinate position sent by the laser radar is recorded, and the distance is calculated based on the time or phase difference of the laser reflection back to obtain the point cloud data of the object. The pre-processed point cloud data is then gridded and modeled to generate a point cloud model.

[0016] Step 4: Calibrate the coordinate system: Take an edge point of the point cloud model as the origin and determine a small square area in the point cloud model. The square area has the origin as the vertex and is located within the point cloud model. Determine the X-axis and Y-axis points along the two sides of the small square along the origin. Move the laser cladding head at the end of the six-axis robot to the above three points and record the position and posture of the six-axis robot at the three points to calibrate the coordinate system of the six-axis robot.

[0017] Step 5: Generate the robot arm motion program based on the path planning algorithm. The process is as follows:

[0018] Based on the constructed point cloud model, the starting and ending points of the cladding path are determined by the path planning algorithm. The shortest path is generated based on the algorithm to form the cladding path. The cladding path is converted into a 3D cladding path through coordinate transformation operations. The cladding path is then converted into a calibrated 6-axis robot motion coordinate system to convert the spatial motion trajectory into a 6-axis robot motion program.

[0019] Step 6: Cladding path tracking control: The six-axis robot performs cladding according to the generated motion program. Based on the model predictive control algorithm MPC, it ensures that the robot performs cladding according to the generated motion path.

[0020] Furthermore, in step three, the preprocessing method includes noise filtering and outlier removal.

[0021] Furthermore, in step five, the path planning algorithm is an A* algorithm or an RRT algorithm.

[0022] Furthermore, in step six, the motion program of the robot described in step five is input and imported into the controller, and the prediction module in the controller outputs the predicted cladding path; the decision module in the controller controls the cladding path of the robot and generates the actual cladding path. After generation, the cladding path is output and compared with the predicted path. If there is any deviation, it is fed back to the control decision module to output the accurate path.

[0023] The technical concept of the present invention is to combine the industrial robot system with the laser cladding technology and integrate the equipment required for the laser cladding technology on a wheeled mobile vehicle. Figure 3 As shown. It can omit the step of disassembling the equipment during equipment maintenance, improve maintenance efficiency, and make laser cladding technology free from bulky equipment and realize movable laser cladding. Figure 3 As shown, a laser cladding head and a laser rust removal head are installed at the end of a six-degree-of-freedom robotic arm. Before laser cladding, the laser rust removal head is used to remove rust, which improves rust removal efficiency compared to manual rust removal. After rust removal is completed, the laser cladding head is switched to perform laser cladding.

[0024] The beneficial effects of the present invention are mainly manifested in: in the scenario where the equipment in the factory needs to be repaired by laser cladding, the step of disassembling certain equipment is saved, the robotic arm and laser cladding technology are integrated into one, and movable laser cladding repair is realized. Manual inspection of equipment defects, manual rust removal, manual equipment repair, etc. are handed over to the wheeled mobile laser cladding robot to complete, thereby improving maintenance efficiency and reducing labor costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is the technical roadmap for cladding repair;

[0026] Figure 2 This is the schematic diagram of model predictive control;

[0027] Figure 3 This is a model of a mobile cladding maintenance robot, in which 1 is a laser cladding head; 2 is a laser rust removal head; 3 is a camera; 4 is a laser radar; 5 is a six-degree-of-freedom robotic arm; 6 is a water-cooling pipeline; 7 is a water-cooling mechanism; 8 is a shock-absorbing mechanism; 9 is a large-size explosion-proof off-road wheel; 10 is a chassis; 11 is a laser generator; 12 is a steering mechanism; 13 is a wheel-side motor; and 14 is a wire feeder. DETAILED DESCRIPTION

[0028] The present invention will be further described below with reference to the accompanying drawings.

[0029] Reference Figures 1 to 3 , a wheeled mobile laser cladding robot autonomous maintenance method, comprising the following steps:

[0030] Step 1: Defect Identification: Collect images of recent defect types of equipment damage in the factory and classify them. Based on the classified defect samples, the extreme learning machine (ELM) is trained on the defect samples to form a classification model. For example, if two pieces of equipment in the factory each have two defect types, when the wheeled mobile cart identifies defect No. 2 on equipment 2 according to the classification model, it will execute the corresponding program based on the identified equipment defect type.

[0031] Step 2: After identifying the defects, the equipment is rusted. The process is as follows:

[0032] A wheeled mobile cart enters the factory and inspects the equipment in the factory according to the classification model described in step 1. When the cart identifies a defect in a piece of equipment, the six-axis robotic arm uses a laser rust removal head to remove rust from the surface of the equipment. After rust removal is completed, the end of the six-axis robotic arm switches to a laser cladding head.

[0033] Step 3: Construct a point cloud model: A wheeled mobile cart uses the LiDAR to collect point cloud data from the equipment being repaired. A six-degree-of-freedom robotic arm moves linearly at a constant speed, allowing the laser to smoothly sweep across the surface of the component. The coordinates transmitted by the LiDAR are recorded, and the distance is calculated based on the time or phase difference of the laser's return reflection, thereby generating point cloud data for the object. The collected point cloud data often contains noise and outliers, requiring preprocessing to improve data quality. Common point cloud preprocessing methods include noise filtering and outlier removal. This preprocessed point cloud data can then be further processed using methods such as meshing to generate a point cloud model.

[0034] Step 4: Calibrate the coordinate system: Take an edge point of the point cloud model as the origin and determine a small square area in the point cloud model. The square area has the origin as the vertex and is located within the point cloud model.

[0035] The X-axis and Y-axis points are determined along the two sides of the small square along the origin. The laser cladding head at the end of the six-axis robotic arm moves to the above three points and records the position and posture of the six-axis robotic arm at the three points to calibrate the coordinate system of the six-axis robotic arm.

[0036] Step 5: Generate the robot arm motion program based on the path planning algorithm. The process is as follows:

[0037] According to the constructed point cloud model, based on the path planning algorithm (such as A* algorithm, RRT algorithm, etc.), the starting point and end point of the cladding path are determined through the point cloud model, and the shortest path is generated based on the algorithm to form a cladding path; through operations such as coordinate transformation, the cladding path is converted into a cladding path in three-dimensional space, and then converted with the calibrated six-axis robot arm motion coordinate system to convert the spatial motion trajectory into the motion program of the six-axis robot arm.

[0038] Step 6: Cladding path tracking control: The six-axis robot performs cladding according to the generated motion program. Based on the model predictive control algorithm (MPC), it ensures that the robot performs cladding according to the generated motion path. Figure 2 As shown in Figure 2, the specific steps are as follows: Input the robot motion program described in step 5 and import it into the controller. The prediction module in the controller outputs the predicted cladding path. The decision module in the controller controls the robot's cladding path and generates the actual cladding path. After the generation, the cladding path is output and compared with the predicted path. Any deviation is fed back to the control decision module, which outputs the correct path.

[0039] In this embodiment, a wheeled mobile laser cladding robot is used to repair cracks in equipment. In the prior art, in some places where equipment needs to be repaired, manual laser cladding repair requires cumbersome steps, resulting in the equipment being unable to be repaired on the spot or the repair cycle being too long. The wheeled mobile laser cladding robot proposed in the present invention can replace manual repair. The maintenance robot trolley considered in this embodiment repairs cracks in the following scenario: a wheeled mobile laser cladding robot and a cracked device. According to the above steps one to six, the crack is repaired.

[0040] The mobile cladding maintenance robot of this embodiment includes a laser cladding head 1, a laser rust removal head 2, a camera 3, a laser radar 4, a six-degree-of-freedom robotic arm 5, a water-cooling pipeline 6, a water-cooling mechanism 7, a shock-absorbing mechanism 8, large-size explosion-proof off-road wheels 9, a chassis 10, a laser generator 11, a steering mechanism 12, a wheel-side motor 13 and a wire feeder 14. The shock-absorbing mechanism 8 is installed on the chassis 10 and is connected to the large-size explosion-proof off-road wheels 9; a wire feeder 14 is installed at the front end of the chassis 10, and a laser generator 11 is installed in the middle part; a steering mechanism The mechanism 12 is installed at the lower end of the chassis 10 and adopts an Ackerman steering structure; the wheel-side motor 13 is connected to the large-size explosion-proof off-road wheel 9 and provides power for the large-size explosion-proof off-road wheel 9; the six-degree-of-freedom robotic arm 5 is installed at the rear half of the chassis 10, and the laser generator 11 and the wire feeder 14 are connected to the laser cladding head 1 and the laser rust removal head 2 at the end of the robotic arm through the six-degree-of-freedom robotic arm 5; one end of the water-cooling pipe 6 is connected to the water-cooling mechanism 7, and the other end is connected to the laser cladding head 1; a laser radar 4 and a camera 3 are installed on the top of the six-degree-of-freedom robotic arm 5.

[0041] The embodiments of this specification are merely examples of implementations of the invention and are provided for illustrative purposes only. The scope of protection of the present invention should not be considered limited to the specific embodiments described in these embodiments. The scope of protection of the present invention also extends to equivalent technical means that can be conceived by a person of ordinary skill in the art based on the invention.

Claims

1. A wheeled mobile laser cladding robot autonomous maintenance method, characterized in that: The method comprises the following steps: Step 1: Defect Identification: Collect images of equipment damage within the factory and classify them into different types. Train the classified defect samples using an extreme learning machine (ELM), forming a classification model. The wheeled mobile cart identifies the defects based on the classification model. Step 2: After identifying the defects, the equipment is rusted. The process is as follows: A wheeled mobile cart enters the factory and inspects the equipment in the factory according to the classification model described in step 1. When the cart identifies a defect in a piece of equipment, the six-axis robotic arm uses a laser rust removal head to remove rust from the surface of the equipment. After rust removal is completed, the end of the six-axis robotic arm switches to a laser cladding head. Step 3: Construct a point cloud model: The wheeled mobile vehicle collects point cloud data of the equipment to be repaired based on the laser radar. The six-degree-of-freedom robotic arm moves linearly at a constant speed, allowing the laser to smoothly sweep the surface of the component. The coordinate position sent by the laser radar is recorded, and the distance is calculated based on the time or phase difference of the laser reflection back to obtain the point cloud data of the object. The pre-processed point cloud data is then gridded and modeled to generate a point cloud model. Step 4: Calibrate the coordinate system: Take an edge point of the point cloud model as the origin and determine a small square area in the point cloud model. The square area has the origin as the vertex and is located within the point cloud model. Determine the X-axis and Y-axis points along the two sides of the small square along the origin. Move the laser cladding head at the end of the six-axis robot to the above three points and record the position and posture of the six-axis robot at the three points to calibrate the coordinate system of the six-axis robot. Step 5: Generate the robot arm motion program based on the path planning algorithm. The process is as follows: Based on the constructed point cloud model, the starting and ending points of the cladding path are determined by the path planning algorithm. The shortest path is generated based on the algorithm to form the cladding path. The cladding path is converted into a 3D cladding path through coordinate transformation operations. The cladding path is then converted into a calibrated 6-axis robot motion coordinate system to convert the spatial motion trajectory into a 6-axis robot motion program. Step 6: Cladding path tracking control: The six-axis robot performs cladding according to the generated motion program. Based on the model predictive control algorithm MPC, it ensures that the robot performs cladding according to the generated motion path.

2. The autonomous maintenance method of a wheeled mobile laser cladding robot according to claim 1, characterized in that: In the step 3, the preprocessing method includes noise filtering and outlier removal.

3. The autonomous maintenance method of a wheeled mobile laser cladding robot according to claim 1 or 2, characterized in that: In step 5, the path planning algorithm is the A* algorithm or the RRT algorithm.

4. The autonomous maintenance method of a wheeled mobile laser cladding robot according to claim 1 or 2, characterized in that: In step six, the motion program of the robot described in step five is input and imported into the controller. The prediction module in the controller outputs the predicted cladding path; the decision module in the controller controls the cladding path of the robot and generates the actual cladding path. After generation, the cladding path is output and compared with the predicted path. If there is any deviation, it is fed back to the control decision module to output the accurate path.

Images