Large steel structure field welding integrated robot and operation method

Abstract

The invention belongs to the technical field of automatic welding of large steel structures. According to the large steel structure field welding integrated robot and the operation method, through the design of the quick change mechanism, integration of various tail end tools is achieved, and integration of multiple functions such as welding, grinding, flaw detection and spraying is achieved; a real-time online analysis-processing-sensing system of multi-source data space-time fusion interaction is constructed, a multi-terminal working path autonomous online planning strategy in a complex field environment is provided, and a high-precision, high-speed and anti-disturbance control method for a mechanical arm and a self-learning optimal control method for terminal precise positioning are provided. The problem that the on-site steel structure welding procedure is simplified is solved, the installation precision and the construction efficiency are improved, and intelligent building development is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of large-scale steel structure automated welding, and in particular to a large-scale steel structure on-site welding integrated robot and an operation method. Background Art

[0002] The statements in this section merely provide background art related to the present invention and do not necessarily constitute prior art.

[0003] On-site welding and assembly of large steel structures include a series of steps such as welding of steel components, non-destructive testing of joints, grinding of welds and painting of welding areas. As a key step in building construction, the quality of steel structure welding directly affects the safety and durability of the building.

[0004] Existing processes face many challenges and problems: (1) Welding defects are one of the most common problems in steel structure welding, mainly including cracks, pores, lack of fusion and weld discontinuity. These defects are usually caused by improper welding technology, improper welding parameters or poor quality of welding materials. (2) The lack of standardization of steel structure welding processes is another major problem. Due to the lack of unified welding standards and specifications in different project projects, the quality of welding operations often depends on the skills and experience of welders. In this case, the welding quality is difficult to guarantee. In actual construction, due to lack of professional training or insufficient experience, the skill level of welders will also have an important impact on the welding quality. Summary of the invention

[0005] In order to address the shortcomings of the prior art, the present invention provides an integrated robot and operation method for on-site welding of large steel structures. Through the design of a quick-change mechanism, the integration of multiple terminal tools is realized, and multi-functional integration such as welding, grinding, flaw detection, and spraying is realized; a real-time online analysis-processing-perception system with spatiotemporal fusion and interaction of multi-source data is constructed, and an autonomous online planning strategy for multi-terminal work paths in complex on-site environments is proposed. A high-precision, high-speed, anti-disturbance control method for robotic arms and a self-learning optimal control method for precise terminal positioning are proposed. The problem of simplifying the on-site steel structure welding process has been solved, the installation accuracy and construction efficiency have been improved, and the development of intelligent construction has been promoted.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: In a first aspect, the present invention provides an integrated robot for on-site welding of large steel structures.

[0007] A large steel structure on-site welding integrated robot, comprising: a robot chassis, an industrial computer, a mechanical arm, and a quick-change mechanism at the end of the mechanical arm, wherein the robot chassis is equipped with an end tool library, and the end tool library is used to carry different types of end tools; The quick-change mechanism comprises a main tray and a tool tray, wherein the main tray is fixedly connected to the end of the robot arm, the tool tray is connected to the end tool, and the tool tray and the corresponding end tool are placed in the end tool library together, and the main tray and the tool tray are connected by a locking mechanism; The tool tray is connected to a plurality of sensor elements that communicate with the industrial computer. The industrial computer is used to control the movement of the robot arm to replace the end tool according to the work task and the received sensor data, and to plan the path of the work task after the end tool is replaced to complete the work task.

[0008] As a further limitation of the first aspect of the present invention, the end tool includes: a welding tool, a spraying tool, a detection tool and a grinding tool.

[0009] As a further limitation of the first aspect of the present invention, each of the main trays is connected to a plurality of the tool trays, and each tool tray corresponds to an end tool.

[0010] As a further limitation of the first aspect of the present invention, the sensing element includes: a photoelectric camera, a laser radar, a temperature sensing unit and a gyroscope.

[0011] As a further limitation of the first aspect of the present invention, the locking mechanism comprises: a locking sleeve, a locking cam, a locking steel ball and a locking ring located at the end of the tool disk, the locking sleeve is installed at the front end of the main disk body, and six evenly distributed holes are provided in the cylindrical protruding part of the locking sleeve, and the locking steel ball is arranged in the hole; The locking cam is fixed to the piston rod by a set screw. The matching surfaces of the locking cam and the locking steel ball are designed with a slope. The locking cam can slide in the hollow cavity in the middle of the locking sleeve, push the locking steel ball to move, and form a rigid contact or separation with the locking ring to achieve the locking or loosening of the main disk and the tool disk.

[0012] As a further limitation of the first aspect of the present invention, the main body of the main tray is equipped with a positioning pin for connecting with the tool tray and playing a guiding role, and the end of the tool tray is provided with a tapped hole that cooperates with the positioning pin.

[0013] In a second aspect, the present invention provides an operating method of an integrated robot for on-site welding of large steel structures.

[0014] An operation method of a large steel structure on-site welding integrated robot, using the large steel structure on-site welding integrated robot according to the first aspect of the present invention, comprises the following process: Acquire the sensor data collected by each sensor element during the movement of the robot chassis; Analyze the support level of each sensor data as evidence for a certain decision, and combine the support levels of different sensor data to take the decision with the highest support level of combined evidence as the information fusion result; According to the information fusion result and the preset parameter anti-disturbance controller, the anti-disturbance motion control of the robot chassis is performed.

[0015] As a further limitation of the second aspect of the present invention, the parametric active disturbance rejection controller includes: a tracking differentiator, an extended state observer and a nonlinear state error feedback control law, and the nonlinear state error feedback control law is: ,in, is the error between the system output and the expected output, Is the error The derivative of , , is the control rate parameter, is the exponential of the nonlinear term.

[0016] As a further limitation of the second aspect of the present invention, a working path planning of the robot arm is performed according to the information fusion result and the received work task, and the working path planning includes: The robot workspace is decomposed into multiple irrelevant populations, and the shortest robot path, minimum time and collision prevention are used as multi-objective evaluation functions. The fitness functions of multiple sub-populations are evaluated and a cooperative relationship is established. The global optimal solution of the robot welding path planning is obtained through cooperative communication, and the global optimal solution is used as the work path planning result.

[0017] As a further limitation of the second aspect of the present invention, setting the heuristic function , number of individuals and number of tasks ,in: , Indicates the task point , The distance between two , ; The tasks of each sub-population are numbered and the starting point of all ant colonies' paths and the initial pheromone concentration are set to be the same. The next task point is selected according to the pheromone concentration and the corresponding distance until all tasks are completed. Multiple paths are obtained based on the principle of visiting each task point only once. Individual at the moment From the mission point Move to the mission point The probability of moving It is expressed as: ; Where: Indicates the current individual The set of task points to be visited, represents pheromone factor, represents the heuristic function factor, represents the task point selected from the set to be visited, , Respectively Individual at the moment From the mission point Move to the mission point and mission points The pheromone concentration, and express Individual at the moment From the mission point Move to the mission point and mission points The heuristic function of Taking the path distance of each subpopulation and the total distance of the cooperation path with other subpopulations as the objective function, the weighted multi-objective fitness function is established as follows: , is the weighting coefficient: ; Each sub-population is updated independently without affecting each other, satisfying: ; in, For the pheromone volatilization factor, Indicates pheromone residues. Indicates that pheromones have been added. , Increase the coefficient of pheromone, and For individuals The total length of the path and the time at the task point With mission points Increased pheromone concentrations during the If the output task is satisfied, the task is assigned and the total distance of each task path is calculated, and the minimum value is taken from all the results: , represents the total distance between robots in the task allocation result, , ,···, Respectively represent the shortest total distance when each subpopulation is the initiator and other populations are the collaborators; If the output task is not met, the cycle will repeat until the output task is met.

[0018] Compared with the prior art, the present invention has the following beneficial effects: The present invention uses a large unmanned vehicle as a moving vehicle and a large robotic arm as an operating body, and focuses on the development of automatic quick-change mechanisms at the ends of different tools, overcoming the problem of barrier-free integration of different functional equipment in a limited space, forming the "torso" of the integrated welding robot; establishing a collaborative detection system for multiple types of sensors such as visible light, infrared, ultrasound, pressure, and temperature, and constructing a real-time online analysis-processing-perception system for multi-source data spatiotemporal fusion interaction, forming the "five senses" of the integrated welding robot; proposing a high-precision, high-speed, anti-disturbance control method for the robotic arm and a self-learning optimal control method for precise terminal positioning, forming the "brain" of the integrated welding machine, and through the intelligent collaboration of the "torso-five senses-brain" of the integrated welding robot, breaking through the problem of simplifying the on-site steel structure welding process, improving installation accuracy and construction efficiency, and promoting the development of intelligent construction.

[0019] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0021] Figure 1 A schematic diagram of the large steel structure on-site welding integrated robot provided by the present invention; Figure 2 A schematic diagram of the quick-change mechanism provided by the present invention; Figure 3 A schematic diagram of a tool tray provided by the present invention; Figure 4 A schematic diagram of a main disk provided by the present invention; Figure 5 A schematic diagram of the principle of the parameter active disturbance rejection controller provided by the present invention; Among them, 1. Robot chassis; 2. Industrial computer; 3. Paint storage tank; 4. Central control touch screen; 5. Air compressor; 6. Robot arm; 7. Quick change mechanism; 701. Positioning pin; 702. Twisted hole; 703. Locking sleeve; 704. Locking steel ball; 705. Locking ring; 706. Locking cam; 707. Piston rod; 708. Main disk; 709. Tool disk; 8. Welding gun; 9. Ultrasonic sensor. DETAILED DESCRIPTION

[0022] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0023] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0024] In this implementation, an integrated robot for on-site welding of large steel structures is proposed, including: a robot chassis 1, a robotic arm 6, a quick-change mechanism 7, an end tool library, an industrial computer 2, a paint storage tank 3, a central control touch screen 4, an air compressor 5, a welding gun 8 and an ultrasonic sensor 9.

[0025] like Figure 1 It can be seen that the robot chassis 1 not only carries the working system of the entire robot, but also acts as a walking mechanism (with walking wheels or walking tracks at the bottom) to ensure that the robot can move stably on different terrains.

[0026] A lifting platform is specially designed on the robot chassis 1 to adapt to the non-fixed height during steel structure construction, so that the robot can flexibly adapt to various working heights and significantly improve work efficiency; in addition, the robot chassis 1 is also equipped with an air compressor 5, a terminal tool library, a paint storage tank 3 and an industrial computer 2 with a central control touch screen 4.

[0027] The air compressor 5 is responsible for providing high-pressure gas to the nozzle to meet the needs of the spraying operation; the end tool library stores a plurality of different types of end tools, which may include welding tools (such as welding guns 8), spraying tools, detection tools and grinding tools, so that the robot can quickly switch to the corresponding end tool through the quick-change mechanism 7 according to the task requirements, and realize welding, spraying, detection, grinding and other working modes; the paint storage tank 3 is used to store the paint required for spraying; the industrial computer 2 is used to manage and control the operation of the robot, adjust the action position of the robot chassis 1 and the robotic arm 6, and the central control touch screen 4 is used to monitor the working status of the robot in real time.

[0028] In this implementation, the industrial computer 2 obtains work instructions through the interaction of the central control touch screen 4, and switches different work modes according to different work instructions. When the received instruction is a welding command, the industrial computer 2 switches the end tool to the welding gun 8 according to the information returned by the end tool of the robot 6. At the same time, the perception module identifies the functional working position and feeds back the information to the industrial computer 2 in real time. The industrial computer 2 controls the motors of each joint of the robot 6, moves the welding gun 8 to the specified position, and then plans the welding path to control the movement of the robot arm to perform welding work and complete the welding task; similarly, when the received command is other commands, the industrial computer 2 switches to the corresponding end tool according to the task and the information returned by the end of the robot 6, controls the movement path of the robot 6, and completes the corresponding work task.

[0029] In this implementation, the robot arm 6 is exquisitely designed and can achieve multi-angle posture adjustment in space to ensure flexibility and accuracy of the operation. A high-resolution camera is equipped above the flange of the robot arm 6 to identify and finely locate the task area, and the quick-change mechanism 7 is the core of realizing the multifunctional integration of the robot. It enables the robot arm 6 to automatically replace the actuator and easily realize function conversion, further enhancing the robot's operating ability and application scope.

[0030] The quick-change mechanism 7 can replace multiple robots for assembly line operation or the original bulky multifunctional tooling actuators, allowing a single robot to automatically change the end-effector tools within seconds to complete a variety of operation tasks, increase the robot's operational flexibility, and greatly reduce downtime. During the welding process of steel structure construction, multiple processes are independent of each other, affecting construction efficiency. To solve this problem, this implementation method innovatively designs a quick-change mechanism 7 to quickly replace multiple tools and ensure that the robot's end-of-line quick switching device works reliably and stably.

[0031] like Figure 2 , Figure 3 and Figure 4 As shown, the quick-change mechanism 7 includes a main tray 708 and a tool tray 709. The main tray 708 is fixed to the robot arm 6 through a connecting flange, and the tool tray 709 is placed on a tool rack, and an execution tool is configured at its lower end. The main tray 708 and the tool tray 709 can be connected by a locking mechanism disposed in the two trays. In actual application, one main tray 708 can be configured with multiple tool trays 709 and end execution tools to complete a variety of operation tasks.

[0032] More specifically, the locking mechanism includes a locking sleeve 703, a locking cam 706, a locking steel ball 704 and a locking ring 705 located at the end of the tool disk. The locking sleeve 703 is installed at the front end of the main body of the main disk 708. Six evenly distributed holes are provided in the cylindrical protruding part of the locking sleeve 703, and the locking steel balls 704 are arranged in the holes; the locking cam 706 is fixed to the piston rod 707 by a set screw, and the matching surface of the locking cam 706 and the locking steel ball 704 is designed with a slope. The locking cam 706 can slide in the hollow cavity in the middle of the locking sleeve 703, pushing the locking steel ball 704 to move, forming a rigid contact or separation with the locking ring 705, thereby realizing the locking or loosening of the main disk 708 and the tool disk 709. The main body of the main disk 708 is equipped with a positioning pin 701 connected to the tool disk 709, which plays a guiding role in the docking process and plays a positioning role after the docking is locked; the tool disk 709 is provided with a twisted hole 702 that cooperates with the positioning pin 701.

[0033] In complex tasks such as the operation of multifunctional integrated welding robots, building intelligent control systems is of great significance in modern industry. It can significantly improve work efficiency and quality, enhance operational safety, optimize resource utilization, achieve multifunctional integration, and support stable operation in complex environments, taking into account the on-site construction environment and technical requirements of large steel structures. This implementation method proposes an operation method for an integrated robot for on-site welding of large steel structures. The core of the intelligent perception-control-decision-making system to be established online includes: an intelligent perception module to provide a multi-sensor multi-source heterogeneous data foundation, an intelligent control module to give the robot arm the underlying motion control capability of precise anti-interference, and an intelligent decision-making module to optimize the path planning scheme of the welding-flaw detection-grinding-painting process.

[0034] The intelligent perception module is part of the intelligent control system and is responsible for collecting the system status information. The multifunctional integrated welding robot perception module is implemented by various sensors (gyroscopes, photoelectric sensors, etc.), which collects the system status and target information of the integrated equipment and provides effective decision-making information for the control system to make control decisions; the intelligent control strategy can select the best control strategy according to the system status and target information to achieve efficient and reliable control effects. The control strategy of the integrated welding robot in this implementation method includes the control method of the welding robot body based on the anti-disturbance control and the welding, grinding and spraying path control based on path planning.

[0035] The working environment of the integrated welding robot is complex, and the information obtained by a single sensor is very limited. In addition, it is also affected by its own quality and performance. Therefore, the integrated welding robot is equipped with a large number of different types of sensors to meet the needs of detection and data collection. If the information collected by each sensor is processed separately and in isolation, it will not only increase the workload of information processing, but also cut off the inherent connection between the information of each sensor, lose the relevant environmental characteristics that may be contained in the organic combination of the information, cause a waste of information resources, and may even lead to wrong decisions.

[0036] To solve this problem, the present invention proposes a multi-sensor fusion method based on evidence combination. The evidence combination method believes that completing a certain intelligent task is to make several possible decisions based on information about a certain aspect of the environment, and the multi-sensor data information reflects the situation of this aspect of the environment to a certain extent. Therefore, the support degree of each data as evidence supporting a certain decision is analyzed, and the support degrees of different sensor data are combined, that is, evidence combination, and the decision with the largest support degree of the existing combined evidence is analyzed as the information fusion result.

[0037] DS evidence theory (Dempster-Shafer evidence theory) is also called DS theory. In DS evidence theory, a sample space is called an identification frame. express, A series of objects The objects are mutually exclusive and contain all the objects to be identified. , The set of all of itself becomes its power set, denoted as .

[0038] Basic probability assignment (BPA) refers to the process of calculating the basic probability of each piece of evidence in the identification framework Θ. This process uses the basic probability assignment function, also called Function or Function to complete (note: not equal to probability), The function reflects the size of the proposition reliability, denoted as And satisfy: (1); in, represents evidence in the computational identification framework Θ.

[0039] The core of DS theory is the Depster synthesis rule, which assumes that all evidence is completely independent and reliable. In the Depster synthesis rule, if any piece of evidence denies a basic hypothesis, no matter how much confidence other evidence supports the basic hypothesis, the synthesis result will completely deny the basic hypothesis. The Depster synthesis rule is as follows: for , The finite number of function The synthesis rule is: (2).

[0040] In this implementation, the output of each sensor is considered as a source of evidence. , ··· The evidence representing the output of different sensors provides support for the decision. In evidence theory, this evidence is expressed as a confidence function. The final output uses Dempster's combination rule to combine the evidence (confidence function) from different sensors. This rule calculates the final confidence and possibility by considering the consistency and conflict between the evidence. Based on the synthesized confidence function, the total confidence or probability of each hypothesis can be calculated. The final decision is based on the hypothesis with the highest confidence. The final output of this task is the system state estimate.

[0041] This implementation also proposes a robot dynamics control method based on parameter self-correction anti-disturbance control strategy. The steel structure welding robot frequently starts, stops, and has non-stationary stages such as high-speed motion conversion during its working process, which affects the robot's working efficiency and positioning accuracy. To address this problem, a welding robot dynamics control method based on parameter self-correction anti-disturbance control strategy is studied, and a parameter self-disturbance controller is designed. The internal and external uncertainties of the robot system are regarded as "total disturbances", and they are actively and fully estimated and compensated.

[0042] The parametric ADRC consists of three important parts: tracking differentiator, extended state observer and nonlinear error feedback control law. The tracking differentiator provides a method for differentiating noisy or discontinuous signals to avoid unnecessary noise, making the robot system more efficient and robust; the extended state observer expands the disturbance that affects the robot output into a new state variable, observes the expanded total disturbance signal through a feedback mechanism, and eliminates the influence of the disturbance through compensation. The feedback control law is a control law based on error feedback and error elimination. The error obtained from the output of the tracking differentiator and the extended state observer is fed back to the control input to eliminate the system error.

[0043] Active Disturbance Rejection Control (ADRC) is an advanced control strategy. The core idea of ​​ADRC is to treat the internal and external uncertainties of the system as total disturbances, and estimate and compensate them through the Extended State Observer (ESO). This method does not rely on the precise mathematical model of the controlled object, making ADRC perform well in dealing with nonlinear, time-varying and complex system control with unknown disturbances.

[0044] The basic structure of ADRC includes three main parts: Tracking Differentiator (TD), Extended State Observer (ESO) and Nonlinear State Error Feedback (NLSEF). TD is used to extract the continuous signal and differential signal of the system input; ESO is used to estimate the system state and total disturbance; NLSEF designs the control law according to the output of ESO to achieve precise control of the system state.

[0045] The Extended State Observer (ESO) is one of the key components in ADRC. The main task of ESO is to estimate the system state and the total disturbance. In ADRC, the total disturbance is regarded as part of the system state and is observed and estimated together with the system state. The design of ESO is usually based on the linearization or approximate model of the system dynamic equations, and its output includes the estimated value of the system state and the estimated value of the total disturbance.

[0046] In this implementation, the tracking differentiator input is the robot's expected position or expected trajectory, output , is the smoothed expected value and its derivative, the extended state observer input and The output is the system input signal (control quantity, such as joint torque and motor torque) and the system output signal (feedback signal such as actual position). and is the state estimate and the total disturbance estimate; the nonlinear error feedback control rate input and is the state error and the disturbance estimated by the extended state observer, and the output is the control signal (control quantity, such as joint torque and motor torque).

[0047] The general form of ESO can be expressed as: (3); in, , and are the estimates of the system state and the total disturbance, respectively, is the measured output of the system, and is the parameter of ESO.

[0048] The Tracking Differentiator (TD) is used to generate a tracking signal and its derivative. The TD is designed to solve the problem of traditional differentiators when dealing with discontinuous or noisy measurement signals. The TD can provide a continuous and noise-free signal while maintaining a fast response to step inputs.

[0049] The mathematical expression of TD is: (4); in, is the control input, is the system output, is the output of TD, indicating the tracking signal, yes The derivative of , which represents the differential of the tracking signal.

[0050] The nonlinear state error feedback control law (NLSEF) is the last link of ADRC, which is responsible for generating control signals based on the information provided by ESO and TD. The design of NLSEF is usually based on the fastest control theory or nonlinear control theory to achieve fast and accurate control of the system state.

[0051] The control law of NLSEF can be expressed as: (5); in, is the error between the system output and the expected output, is the derivative of the error, , , is the control rate parameter, is the exponent of the nonlinear term, which is used to enhance the robustness of the control rate.

[0052] Welding robots can complete a large number of welding tasks in a short time, and have the advantages of high efficiency, stable quality, safety and reliability. However, the welding path planning of welding robots is a complex and difficult problem. In order to solve this problem, based on Figure 5 The collaborative evolution method shown in the figure proposes a welding robot path planning method, which decomposes the robot workspace into multiple unrelated populations, and evaluates the fitness functions of multiple sub-populations with the shortest robot path, minimum time and collision prevention as multi-objective evaluation indicators, and establishes a cooperative relationship. In other cases, each sub-population evolves independently and is unrelated to each other. Through cooperative communication, the global optimal solution of the robot welding path planning is obtained, and the efficient and detailed allocation of the integrated robot welding path is completed.

[0053] The specific algorithm flow is as follows: Step (1) Algorithm initialization: Parameter initialization: Heuristic function , number of individuals , number of tasks ,in: (6); Where: Indicates the task point , The distance between two .

[0054] Step (2) Select the own variables according to the pheromone: Solution: Number the tasks of each sub-population and set the starting point of all population paths to be the same and the initial pheromone concentration to be the same. Select the next task point according to the pheromone concentration and the corresponding distance until all are completed. It represents the task point to be visited, and multiple paths are obtained based on the principle of visiting each task point only once. Individual at the moment From the mission point Move to the mission point The probability of moving It is expressed as: (7); Where: Indicates the current individual The set of task points to be visited, represents pheromone factor, Represents the heuristic function factor represents the task point selected from the set to be visited, , Respectively Individual at the moment From the mission point Move to the mission point and mission points The pheromone concentration, , express Individual at the moment From the mission point Move to the mission point and mission points The heuristic function of .

[0055] Step (3) Fitness evaluation: Based on the idea of ​​cooperation and collaboration, the path distance of each sub-population and the total distance of the cooperation paths with other subpopulations As the objective function, the weighted multi-objective fitness function is established as follows, where: and is the weighting coefficient: (8); Step (4) Update pheromones and population evolution: Update pheromones. Each sub-population updates independently without affecting each other. The following formula must be satisfied.

[0056] (9); Where: For the pheromone volatilization factor, Indicates pheromone residues. express Pheromones added at the moment, express Pheromones of the moment, express Moment Individual The pheromones have been added.

[0057] (10); In the formula, Increase the coefficient of pheromone, and For individuals The total length of the path and the time at the task point With mission points Increased pheromone concentration.

[0058] Step (5) Output result: Determine whether the end condition is met: If the output task is met, then the task is assigned and the total distance of each task path is calculated, and the minimum value is taken from all the results, as shown below; otherwise, return to step (2) and loop again.

[0059] (11); In the formula, Indicates the total distance between robots in the task allocation result; , ,···, They respectively represent the shortest total distance when each sub-population is the initiator and other populations are the collaborators. The task corresponding to the shortest total distance is the path planning result.

[0060] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A large steel structure on-site welding integrated robot, characterized in that: include: A robot chassis, an industrial computer, a robot arm, and a quick-change mechanism at the end of the robot arm. The robot chassis is equipped with an end tool library, and the end tool library is used to carry different types of end tools; The quick-change mechanism comprises a main tray and a tool tray, wherein the main tray is fixedly connected to the end of the robot arm, the tool tray is connected to the end tool, and the tool tray and the corresponding end tool are placed in the end tool library together, and the main tray and the tool tray are connected by a locking mechanism; The tool tray is connected to a plurality of sensor elements that communicate with the industrial computer. The industrial computer is used to control the movement of the robot arm to replace the end tool according to the work task and the received sensor data, and to plan the path of the work task after the end tool is replaced to complete the work task.

2. The large steel structure on-site welding integrated robot according to claim 1, characterized in that: The end tools include: welding tools, spraying tools, testing tools and grinding tools.

3. The large steel structure on-site welding integrated robot according to claim 1, characterized in that: Each of the main trays is connected to a plurality of the tool trays, and each tool tray corresponds to an end tool.

4. The large steel structure on-site welding integrated robot according to claim 1, characterized in that: The sensing element includes: a photoelectric camera, a laser radar, a temperature sensing unit, a gyroscope and an ultrasonic sensor.

5. The large steel structure on-site welding integrated robot according to any one of claims 1 to 4, characterized in that: The locking mechanism comprises: a locking sleeve, a locking cam, a locking steel ball and a locking ring located on the tool disk, the locking sleeve is installed at the front end of the main disk body, and a plurality of evenly distributed holes are provided in the cylindrical protruding part of the locking sleeve, and the locking steel ball is arranged in the hole; The locking cam is fixed to the piston rod by a set screw. The matching surfaces of the locking cam and the locking steel ball are designed with a slope. The locking cam can slide in the hollow cavity in the middle of the locking sleeve, push the locking steel ball to move, and form a rigid contact or separation with the locking ring to achieve the locking or loosening of the main disk and the tool disk.

6. The large steel structure on-site welding integrated robot as claimed in claim 5, characterized in that: The main body of the main disk is equipped with a positioning pin connected to the tool disk, which plays a guiding role during the docking process and a positioning role after the docking is locked. The tool disk is provided with a twisting hole that cooperates with the positioning pin.

7. A method for operating a large steel structure on-site welding integrated robot, characterized in that: Using the large steel structure on-site welding integrated robot as described in any one of claims 1 to 4, The process includes: Acquire the sensor data collected by each sensor element during the movement of the robot chassis; Analyze the support level of each sensor data as evidence for a certain decision, and combine the support levels of different sensor data to take the decision with the highest support level of combined evidence as the information fusion result; According to the information fusion result and the preset parameter anti-disturbance controller, the anti-disturbance motion control of the robot chassis is performed.

8. The operation method of the large steel structure on-site welding integrated robot according to claim 7, characterized in that: The parameter anti-disturbance controller includes: a tracking differentiator, an extended state observer and a nonlinear state error feedback control law, and the nonlinear state error feedback control law is: ,in, is the error between the system output and the expected output, Is the error The derivative of , , is the control rate parameter, is the exponential of the nonlinear term.

9. The operation method of the large steel structure on-site welding integrated robot according to claim 7, characterized in that: According to the information fusion result and the received work task, the work path planning of the robot arm is performed, and the work path planning includes: The robot workspace is decomposed into multiple irrelevant populations, and the shortest robot path, minimum time and collision prevention are used as multi-objective evaluation functions. The fitness functions of multiple sub-populations are evaluated and a cooperative relationship is established. The global optimal solution of the robot welding path planning is obtained through cooperative communication, and the global optimal solution is used as the work path planning result.

10. The operation method of the large steel structure on-site welding integrated robot according to claim 9, characterized in that: Setting the heuristic function , number of individuals and number of tasks ,in: , Indicates the task point and The distance between two , ; The tasks of each sub-population are numbered and the starting point of all ant colonies' paths and the initial pheromone concentration are set to be the same. The next task point is selected according to the pheromone concentration and the corresponding distance until all tasks are completed. Multiple paths are obtained based on the principle of visiting each task point only once. Individual at the moment From the mission point Move to the mission point The probability of moving It is expressed as: ； Where: Indicates the current individual The set of task points to be visited, represents pheromone factor, Represents the heuristic function factor represents the task point selected from the set to be visited, , Respectively Individual at the moment From the mission point Move to the mission point and mission points The pheromone concentration, , express Individual at the moment From the mission point Move to the mission point and mission points The heuristic function of The path distance of each subpopulation and the total distance of the cooperation paths with other subpopulations As the objective function, establish the weighted multi-objective fitness function: ； in, , is the weighted coefficient, each sub-population is updated independently without affecting each other, satisfying: ； in, For the pheromone volatilization factor, Indicates pheromone residues. Indicates that pheromones have been added. , Q is the pheromone increase coefficient, and For individuals The total length of the path and the time at the task point With mission points Increased pheromone concentrations for The pheromone concentration at the moment; If the output task is satisfied, the task is assigned and the total distance of each task path is calculated, and the minimum value is taken from all the results: , represents the total distance between robots in the task allocation result, , ,···, Respectively represent the shortest total distance when each subpopulation is the initiator and other populations are the collaborators; If the output task is not met, the cycle will repeat until the output task is met.

Images