Rail transit equipment welding robot motion precision reliability evaluation method based on confidence reliability

Abstract

The application relates to a track transportation equipment welding robot motion precision reliability evaluation method based on confirmed reliability, which comprises the following steps: acquiring nominal kinematic parameters and D-H parameter errors of a welding robot; solving robot joint angles according to a target pose of an end effector; establishing a mapping relationship between the D-H parameter errors and end position errors; establishing a position precision performance margin equation based on the mapping relationship and a preset allowed position error threshold; calculating single-point position precision confirmed reliability based on an uncertain first-order reliability method, and performing double tangent plane correction on the reliability result; performing discretization processing on a continuous welding seam path, and obtaining whole-seam welding trajectory precision confirmed reliability according to the reliability of each discrete path point, so that the reliability evaluation of the welding robot motion precision is realized. Compared with the prior art, the application has the advantages of reducing calculation complexity, improving position precision reliability and the like.

Description

Technical Field

[0001] This invention relates to the field of robot motion, and in particular to a method for evaluating the motion accuracy reliability of welding robots for rail transit equipment based on assured reliability. Background Technology

[0002] Welding robots are now widely used in welding rail vehicle bodies, bogie frames, and components. However, in actual welding processes, if the actual position of the robot's end effector deviates from the predetermined weld position, it can cause quality defects such as incomplete penetration, undercut, and unstable weld formation. Especially for equipment with high structural safety requirements, such as rail transit vehicles, the accuracy of the welding position directly affects the quality of component connections and the reliability of vehicle service.

[0003] The common approach to solving robot positioning accuracy problems is to first establish kinematic equations based on the DH model, and then use matrix methods, differential methods, or numerical identification methods to establish a geometric error model for compensation. This approach only addresses error modeling and compensation; however, there is no clear criterion for determining whether the robot can stably meet welding accuracy requirements after compensation.

[0004] In robot position accuracy reliability analysis, some studies treat DH parameter errors, joint errors, or end-effector pose errors as random variables and use probabilistic reliability methods to calculate the probability that the robot meets accuracy requirements. This approach usually requires a large amount of sample data. However, welding robots are high-reliability, long-life equipment, and the number of failure samples obtainable in actual operation is limited, while calibration and testing costs are also high. Furthermore, the number of sampling points, their spatial distribution, the accuracy of the measuring equipment, the convergence result of the parameter identification algorithm, and environmental factors such as temperature and load all affect the estimation results of error parameters. Therefore, robot position accuracy errors include both random fluctuations and uncertainties arising from incomplete understanding. In this situation, directly applying traditional probabilistic reliability methods can easily lead to an overestimation of reliability.

[0005] Furthermore, rail vehicle welding often involves trajectory tracking along a continuous weld seam, rather than single-point positioning. In this case, trajectory accuracy and reliability should be reflected in whether the positional error of the end effector along the entire weld seam path consistently does not exceed the allowable threshold. Some existing methods treat the entire weld seam as a high-dimensional time-varying reliability problem, resulting in high computational costs and low efficiency in engineering applications. However, if only a small number of feature points are selected for evaluation, trajectory areas with significant curvature changes may be missed, making it difficult to reflect the true accuracy error risk of the entire weld seam during the welding process.

[0006] Chinese invention patent CN111985076A discloses a method and apparatus for evaluating the reliability of robot motion. This method considers the influence of parameter interval variables and error interval variables at various points of the robot, including the interval variables and error interval variables of each joint point, and has the advantages of accurate and intuitive reliability evaluation. However, it still has the problems of different working conditions affecting the estimation results of error parameters, large computational load and low efficiency in engineering applications, difficulty in reflecting the true accuracy error risk of the entire weld seam during the welding process, and the problem of overestimating reliability due to linear approximation of a single tangential plane.

[0007] In summary, there is currently a lack of a reliable assessment method for the motion accuracy of welding robots for rail transit equipment based on certainty-based reliability, in order to solve or partially solve the above problems. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the existing technology by providing a reliability assessment method for the motion accuracy of welding robots for rail transit equipment based on certainty reliability. This method aims to solve or partially solve the problems of large computational load and low efficiency in engineering applications of the estimation results of error parameters affected by different working conditions, difficulty in reflecting the actual accuracy error risk of the entire weld seam during the welding process, and the problem of high reliability caused by linear approximation of a single tangent plane.

[0009] The objective of this invention can be achieved through the following technical solutions: According to one aspect of the present invention, a method for evaluating the motion accuracy reliability of a welding robot for rail transit equipment based on assured reliability is provided, the method comprising: S1. Obtain the nominal kinematic parameters and DH parameter errors of the welding robot; S2. Solve for the robot joint angle set based on the target pose of the end effector, and select an optimal set of joint angles based on the constraints in the actual application; S3. Introduce the DH parameter error into the transformation matrix of adjacent links to obtain the linear relationship between the link parameter error and the pose error of adjacent links. Transmit the link error to the robot base coordinate system step by step, and establish the mapping relationship between the DH parameter error and the end position error based on the nominal kinematic parameters and joint angles. S4. Based on the mapping relationship between the DH parameter error and the end position error, establish a position accuracy performance margin equation according to the preset allowable position error threshold; S5. Calculate the single-point position accuracy assurance reliability based on the uncertain first-order reliability method, and perform double-tangent plane correction on the single-point position accuracy assurance reliability result to obtain the corrected single-point position accuracy assurance reliability. S6. Discretize the continuous weld path and obtain the reliability of the entire weld trajectory accuracy based on the reliability of the single-point position accuracy of each discrete path point, thereby realizing the reliability assessment of the motion accuracy of the welding robot.

[0010] As a preferred technical solution, the process of obtaining the DH parameter error includes: Sampling points are generated within the workspace of the welding robot, and the theoretical pose and actual pose of the robot end effector are obtained for each sampling point. The difference between the actual pose and the theoretical pose is identified using the LM method and expressed as a normal uncertain variable to obtain the DH parameter error.

[0011] As a preferred technical solution, the mapping relationship between the DH parameter error and the end position error is expressed as follows: in, Indicates the position error of the robot's end effector. This represents the error coefficient matrix determined by the robot's nominal kinematic parameters. This indicates the error in the DH parameter.

[0012] As a preferred technical solution, the position accuracy performance margin equation is expressed as: in, For performance margin, For DH parameter error, To set the allowable position error threshold, For the robot end-effector position error, Represents the transpose of a matrix; when When the robot's end-effector position error is within the allowable range, it is considered to be within the acceptable range. If this happens, the current target point's position accuracy is considered invalid.

[0013] As a preferred technical solution, the process of obtaining the reliability assurance of the single-point position accuracy includes: The most likely failure point is defined as the point on the limit state surface with the smallest Chebyshev distance from the origin. A first-order Taylor expansion of the limit state function is performed at the most likely failure point to obtain the linearized limit state function. The single-point position accuracy failure confidence and single-point position accuracy certainty reliability are calculated based on the standard normal uncertain distribution function.

[0014] As a preferred technical solution, the most likely failure point is obtained by solving a constrained optimization problem or a convex multicell bisection method.

[0015] As a preferred technical solution, the double-tangent plane correction includes: A first tangent plane is established at the most likely failure point of the single-point position accuracy reliability result, and the first tangent plane is translated to the corresponding tangent position on the other side of the limit state surface by utilizing the symmetry of the limit state surface to obtain the second tangent plane. The total failure confidence is approximated by the upper bound of the confidence of the failure domain outside the region enclosed by the two tangent planes. Then, the corrected single-point location accuracy assurance reliability is obtained by the following formula: in, To ensure the reliability of the corrected single-point position accuracy, This is the approximate total failure confidence.

[0016] As a preferred technical solution, the steps for obtaining the reliability assurance of the accuracy of the entire weld seam trajectory include: The continuous weld path is discretized to obtain discrete path points, and the reliability of the single-point position accuracy of each discrete path point is calculated. The entire continuous weld is regarded as a series system composed of discrete path points. The reliability of the accuracy of the entire weld trajectory is taken as the minimum value among the reliability of each discrete path point, and the reliability of the accuracy of the entire weld trajectory is output.

[0017] As a preferred technical solution, if the position error of any path point in the discrete path points exceeds a preset allowable threshold, it is considered that the accuracy of the entire weld trajectory has failed.

[0018] According to another aspect of the present invention, a reliability assessment device for the motion accuracy of a welding robot for rail transit equipment based on assured reliability is provided. The device is used to implement the aforementioned reliability assessment method for the motion accuracy of a welding robot for rail transit equipment based on assured reliability. The device comprises: The parameter error acquisition module is used to acquire the robot's nominal kinematic parameters and obtain the DH parameter error distribution characteristics through measurement and identification. The kinematics solution module is used to build the DH kinematic model and solve the robot joint angles based on the end-effector pose. The position error modeling module is used to establish the mapping relationship between DH parameter errors and end position errors; The performance margin construction module is used to establish the position accuracy performance margin equation based on the allowable position error threshold. The module for solving the most likely failure point is used to solve the most likely failure point in a standard normal uncertainty space using constrained optimization problems or the convex multicell bisection method. The single-point reliability calculation module is used to calculate the certainty reliability of single-point location accuracy based on the uncertain first-order reliability method. The dual-tangent plane correction module is used to correct the reliability of single-point position accuracy and reduce the overestimation caused by single-tangent plane approximation. The trajectory reliability assessment module is used for trajectory planning, path discretization, and reliability calculation of series systems for continuous weld seam paths. The results output module is used to output the reliability of single-point position accuracy, the reliability of the entire weld trajectory accuracy, and the path point information with the lowest reliability.

[0019] Compared with the prior art, the present invention has at least one of the following beneficial effects: (1) This invention introduces the DH parameter error into the transformation matrix of adjacent links to obtain the linear relationship between the link parameter error and the pose error of adjacent links. The error of each link is passed to the robot base coordinate system step by step to establish the mapping relationship between the DH parameter error and the end position error. This enables the reliability calculation to be directly related to the robot joint angle and configuration, thereby reflecting the position error change under different welding poses. This solves the problem of different working conditions affecting the error parameter estimation results and achieves the technical effect that the robot position accuracy reliability evaluation is more in line with the actual working conditions and the credibility is more accurate.

[0020] (2) This invention discretizes the continuous weld trajectory into multiple path points and calculates the reliability of the entire trajectory according to the rules of a series system. This solves the problem that the computational workload is large and the efficiency of engineering applications is low. However, if only a small number of feature points are selected for evaluation, the trajectory area with large curvature changes may be missed, making it difficult to reflect the actual accuracy error risk of the entire weld during the welding process. This invention achieves the goal of reducing computational complexity while avoiding the omission of weak positions in the weld trajectory. (3) The present invention corrects the reliability result of single-point position accuracy by using the double-tangent plane correction method, making the evaluation result more conservative and reliable, solving the problem of high reliability caused by single-tangent plane linear approximation, and achieving the technical effect of improving the accuracy of the reliability evaluation result of welding robot motion accuracy.

[0021] (4) The present invention treats the DH parameter error as a normal uncertain variable, which can take into account the cognitive uncertainty caused by parameter identification, measurement error and working condition changes when the calibration sample is limited, making the robot position accuracy reliability evaluation more in line with the actual working conditions.

[0022] (5) The present invention adopts an uncertain first-order reliability method based on the Chebyshev distance of the most likely failure point, which can avoid the problem that the traditional mean point expansion method is sensitive to the expression form of the limit state function and improve the rationality of single-point position accuracy reliability calculation.

[0023] (6) The present invention uses the convex multicell bisection method to solve the most likely failure point, transforming the high-dimensional non-smooth optimization problem into the intersection judgment of the hypercube and the limit state surface, thereby improving the computational stability and convergence efficiency. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating the steps of the method of the present invention; Figure 2 A schematic diagram of the limit state surface and its tangent plane; Figure 3 A schematic diagram of the process for evaluating the reliability of continuous weld trajectory accuracy. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0026] Example 1 To address the aforementioned problems in the existing technology, this embodiment provides a reliability assessment method for the motion accuracy of welding robots for rail transit equipment based on certainty reliability. This method is implemented by a processor executing a computer program to assess the certainty reliability of the single-point positioning accuracy and continuous weld trajectory accuracy of the welding robot end effector under limited sample conditions.

[0027] This method takes the nominal kinematic parameters of the welding robot, the error distribution characteristics of the DH parameters, the end-effector pose or weld path, and the allowable position error threshold as inputs. It outputs the single-point position accuracy assurance reliability, the overall weld trajectory accuracy assurance reliability, and the path point position with the lowest reliability and its corresponding robot configuration. Specific steps are as follows: Figure 1 As shown, it specifically includes: S1. Obtain the nominal kinematic parameters and DH parameter error of the welding robot.

[0028] Specifically, the nominal kinematic parameters of the welding robot are obtained. For a six-axis tandem welding robot, the parameters include link lengths. Linkage torsion angle Linkage offset Joint angle When adjacent joint axes are parallel or nearly parallel, a winding is introduced between adjacent links. Parameters of axis rotation This is to reduce the impact of parallel joints on the stability of parameter identification.

[0029] Several sampling points are generated within the welding robot's workspace. These sampling points can be generated using a Latin hypercube sampling method to ensure that they cover as much of the robot's reachable workspace as possible, avoiding concentration of sampling points in localized areas.

[0030] For each sampling point, the theoretical pose and actual pose of the robot's end effector are obtained. The theoretical pose is calculated from the nominal kinematic parameters and the forward kinematic equations; the actual pose is measured by a laser tracker, a vision measurement system, or other pose measurement equipment.

[0031] The difference between the actual pose and the theoretical pose is used as the identification input, and the LM algorithm is employed to identify the DH parameter errors of the robot. After multiple identifications, the distribution characteristics of each parameter error are statistically analyzed.

[0032] The identified DH parameter error is expressed as a normally uncertain variable, resulting in: in, Indicates the first One DH parameter error, Represents a normal uncertain distribution. Indicates the first The mean of the errors of each parameter. Indicates the first The standard deviation of the error of each parameter.

[0033] The DH parameter error vector is denoted as: in, The number of parameter errors involved in the identification. This is the matrix transpose.

[0034] For a six-axis welding robot, when each link is considered , , , , When the parameters are fixed, a 30-dimensional parameter error vector can be formed; when some fixed parameters are not involved in the identification, a lower-dimensional parameter error vector can also be formed.

[0035] S2. Solve for the robot joint angle set based on the target pose of the end effector, and select an optimal set of joint angles according to the constraints in the actual application.

[0036] Specifically, the target pose of the end effector is first given. For single-point positioning, the target pose is a definite end position and attitude; for continuous welding, the target pose is jointly determined by the weld path and the welding torch attitude requirements.

[0037] When the input is a continuous weld, trajectory planning is performed first. For straight welds, linear interpolation is used to generate path points; for circular welds, the three-point circle method is used to determine the center and radius, and then the circular path points are generated. The welding torch posture can be determined by quaternion spherical linear interpolation, or by establishing a local coordinate system based on the tangential, radial, and normal directions of the weld to ensure the continuity of the welding torch posture.

[0038] Solve the robot joint angles based on the target pose. For a six-axis welding robot that satisfies Pieper's criterion, when the axes of the 4th, 5th, and 6th joints intersect at a point, the position and attitude can be decoupled. First, the first three joint angles can be solved from the end position, and then the last three joint angles can be solved from the end attitude.

[0039] The multiple solutions obtained from inverse kinematics are screened. First, solutions that exceed joint limits, are close to singular configurations, or pose an interference risk are excluded. Then, the solution with the smallest change in joint angles relative to the current robot configuration is selected from the remaining solutions as the joint angles corresponding to the target point, in order to ensure the continuity of the welding trajectory and avoid abrupt joint changes during the welding process.

[0040] S3. Introduce the DH parameter error into the transformation matrix of adjacent links to obtain the linear relationship between the link parameter error and the pose error of adjacent links. Transmit the error of each link to the robot base coordinate system step by step, and establish the mapping relationship between the DH parameter error and the end position error based on the nominal kinematic parameters and joint angles.

[0041] Specifically, a homogeneous transformation matrix between adjacent link coordinate systems is established based on nominal kinematic parameters, and the forward kinematic equation of the robot's end effector relative to the base coordinate system is obtained through matrix multiplication. Small errors in the DH parameters are introduced into the transformation matrix of adjacent links. Based on differential kinematics, second-order and higher-order terms are ignored, and first-order differential terms are retained to obtain the linear relationship between link parameter errors and adjacent link pose errors. The errors of each link are then progressively propagated to the robot's base coordinate system to obtain the mapping relationship between the end effector position error and the DH parameter error. in, Indicates the position error of the robot's end effector. This represents the error coefficient matrix determined by the robot's nominal kinematic parameters. This indicates the error in the DH parameter.

[0042] Calculate the magnitude of the end position error based on the end position error vector. : because As the robot's joint angles change, the same set of DH parameter errors will produce different end-effector position errors under different robot configurations.

[0043] S4. Based on the mapping relationship between DH parameter error and end position error, establish a position accuracy performance margin equation according to the preset allowable position error threshold.

[0044] Set the allowable position error threshold This threshold can be determined based on the allowable deviation of weld positioning, welding process documents, or product quality requirements.

[0045] A position accuracy performance margin equation is established with the allowable position error threshold as the boundary. The quadratic form is preferred: The following equivalent form can also be used: in, For performance margin, For DH parameter error, To set the allowable position error threshold, For the robot end-effector position error, This represents the transpose of a matrix.

[0046] when When the robot's end-effector position error is within the allowable range, it is considered to be within the acceptable range. If this occurs, the current target point's position accuracy is considered invalid.

[0047] S5. Calculate the certainty reliability of single-point location accuracy based on the uncertain first-order reliability method.

[0048] Specifically, the DH parameter error variable is transformed to the standard normal uncertainty space. The transformation relationship is as follows: in, Indicates the first A standard normal uncertain variable.

[0049] After the transformation, the performance margin equation is expressed as the limit state function in the standard normal uncertainty space: in, , For a standard normal uncertain variable, The number of DH parameter error variables involved in identification and reliability calculation. , .

[0050] Define single-point location accuracy failure confidence For the event Uncertainty measure, defining the single-point position accuracy assurance reliability. For the event The uncertainty measure, the two satisfy: because Typically, these are non-monotonic and nonlinear functions, and their distribution cannot be directly solved using the uncertainty rules for strictly monotonic functions. Therefore, this method employs an uncertain first-order reliability method based on the most likely failure point to calculate the certainty reliability of single-point location accuracy.

[0051] In the standard normal uncertainty space, the most likely failure point is defined as the limit state surface. The point on the map whose Chebyshev distance is minimized from the origin. This Chebyshev distance is: in, It represents the infinite norm.

[0052] The most likely failure point is obtained through the following optimization problem: because Since it is a non-smooth norm, the most likely failure point can be obtained by solving the constrained optimization problem described above. Alternatively, convex optimization methods, feasible region search methods, or other numerical optimization methods that meet the requirements of constrained optimization can be used to solve for the most likely failure point. When using the convex multicell bisection method, the specific methods include: Set a reliable index for searching the lower bound. And set an upper bound for the search. Calculate the midpoint between the two: Constructing a convex multicell in the standard normal uncertain space: Determine the limit state surface Is it related to convex multicellular bodies? If they intersect, then let If the two do not intersect, then let Repeat the above steps until the condition is met. ,in This is the preset calculation precision.

[0053] Find the most likely failure point Afterwards, Regarding the limit state function A first-order Taylor expansion is performed to obtain the linearized limit state function, and the single-point location accuracy failure confidence is calculated based on the standard normal uncertainty distribution function. and single-point position accuracy assurance reliability The calculation results are as follows: in, It is the distribution function of a standard normal uncertain variable.

[0054] S6. Perform double-tangent plane correction on the single-point position accuracy assurance reliability result to obtain the corrected single-point position accuracy assurance reliability.

[0055] Considering that when a single tangent plane approximates the limit state surface, some actual failure areas may be misjudged as reliable areas, resulting in an overestimation of reliability, this method corrects the single-point reliability calculation results with a double tangent plane.

[0056] For the limiting state surface formed by the position error threshold, it usually appears as a high-dimensional elliptical cylinder or a similar closed surface in the standard normal uncertainty space, and the safe region is located inside this surface.

[0057] like Figure 2 As shown, a first tangent plane is established at the most likely failure point, and by utilizing the symmetry of the limit state surface, the first tangent plane is translated to the corresponding tangent position on the other side of the limit state surface to obtain a second tangent plane.

[0058] The total failure confidence is approximated by the upper bound of the confidence of the failure domain outside the region enclosed by the two tangent planes. Then, the corrected single-point location accuracy assurance reliability is obtained by the following formula: in, This represents the L1 norm.

[0059] This correction can reduce the problem of overly high reliability caused by the single-tangent plane approximation.

[0060] S7. Discretize the continuous weld path and obtain the reliability of the entire weld trajectory accuracy based on the reliability of the single-point position accuracy of each discrete path point, thereby realizing the reliability assessment of the motion accuracy of the welding robot.

[0061] like Figure 3 As shown, specifically, the continuous weld path is discretized to obtain... The discrete path points are determined based on the weld type, curvature variation, and error sensitivity. Larger discrete intervals can be used for straight lines or low-curvature regions; smaller discrete intervals are used for corners, arc transition regions, or regions with significant robot configuration changes.

[0062] For each discrete path point, perform the above steps sequentially to obtain the single-point position accuracy assurance reliability of that path point. .

[0063] The entire continuous weld seam is considered as being A series system consisting of discrete path points. If the positional error of any path point exceeds the allowable threshold, the accuracy of the entire weld trajectory is considered to have failed.

[0064] According to the reliability rules for series systems, the certain reliability of the entire weld seam trajectory is taken as the minimum value among the reliability values ​​of each discrete path point: Output accuracy and reliability of the entire weld trajectory It outputs the path point location that achieves minimum reliability, the corresponding joint angle, and the robot configuration. This output can be used to identify weak sections in the weld trajectory and can also serve as a basis for subsequent optimization of robot configuration, adjustment of welding parameters, or error compensation.

[0065] This study focuses on the ER6-730MI six-joint serial industrial robot. A Latin hypercube sampling method is employed to generate sampling points within the robot's workspace, and the actual end-effector pose of each sampling point is acquired using a measuring device. Based on the difference between the theoretical and actual poses, the LM algorithm is used to identify the robot's DH parameter errors.

[0066] Because the axis of the second joint of this robot is parallel to the axis of the third joint, a winding mechanism is introduced between adjacent links during modeling. Parameters of axis rotation For some fixed parameters, they are not included in the error identification. After multiple identifications, the error distribution characteristics of the DH parameters involved in the identification are obtained and expressed as normally distributed uncertain variables.

[0067] In the single-point position accuracy evaluation embodiment, given the robot joint angle as... .

[0068] The robot's end-effector position error is 0.522 mm, and the allowable position error threshold is 1 mm. The most likely failure point is determined using the convex multicell bisection method, with a computational accuracy of 10. -6When the required accuracy was achieved, iterated 22 times. After applying the double-tangent plane correction, the reliability of the single-point position accuracy decreased by 0.0591 compared to before the correction, indicating that the double-tangent plane correction can reduce the problem of high reliability caused by the single-tangent plane approximation.

[0069] In the continuous weld trajectory accuracy evaluation embodiment, a discrete interval of 2mm was used for straight welds, and the allowable position error threshold was set to 1mm. The confidence reliability of the trajectory accuracy before correction was 0.7355, and the confidence reliability of the trajectory accuracy after correction was 0.6068. For circular arc welds, an angle discretization method was used, and weak areas with low reliability were further refined. The confidence reliability of the trajectory accuracy before correction was 0.6789, and the confidence reliability of the trajectory accuracy after correction was 0.5910.

[0070] Example 2 Regarding the aforementioned method for assessing the motion accuracy reliability of a rail transit equipment welding robot based on assured reliability, this embodiment provides a device for assessing the motion accuracy reliability of a rail transit equipment welding robot based on assured reliability. This device is used to execute the aforementioned method for assessing the motion accuracy reliability of a rail transit equipment welding robot based on assured reliability. The device includes a processor and a memory, the memory storing a computer program executable by the processor. The device includes: The parameter error acquisition module is used to acquire the robot's nominal kinematic parameters and obtain the DH parameter error distribution characteristics through measurement and identification. The kinematics solution module is used to build the DH kinematic model and solve the robot joint angles based on the end-effector pose. The position error modeling module is used to establish the mapping relationship between DH parameter errors and end position errors; The performance margin construction module is used to establish the position accuracy performance margin equation based on the allowable position error threshold. The module for solving the most likely failure point is used to solve the most likely failure point in the standard normal uncertainty space using the convex multiple cell bisection method. The single-point reliability calculation module is used to calculate the certainty reliability of single-point location accuracy based on the uncertain first-order reliability method. The dual-tangent plane correction module is used to correct the reliability of single-point position accuracy and reduce the overestimation caused by single-tangent plane approximation. The trajectory reliability assessment module is used for trajectory planning, path discretization, and reliability calculation of series systems for continuous weld seam paths. The results output module is used to output the reliability of single-point position accuracy, the reliability of the entire weld trajectory accuracy, and the path point information with the lowest reliability.

[0071] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A reliability assessment method for the motion accuracy of welding robots in rail transit equipment based on assured reliability, characterized in that, The method includes: S1. Obtain the nominal kinematic parameters and DH parameter errors of the welding robot; S2. Solve for the robot joint angle set based on the target pose of the end effector, and select an optimal set of joint angles based on the constraints in the actual application; S3. Introduce the DH parameter error into the transformation matrix of adjacent links to obtain the linear relationship between the link parameter error and the pose error of adjacent links. Transmit the link error to the robot base coordinate system step by step, and establish the mapping relationship between the DH parameter error and the end position error based on the nominal kinematic parameters and joint angles. S4. Based on the mapping relationship between the DH parameter error and the end position error, establish a position accuracy performance margin equation according to the preset allowable position error threshold; S5. Calculate the single-point position accuracy assurance reliability based on the uncertain first-order reliability method, and perform double-tangent plane correction on the single-point position accuracy assurance reliability result to obtain the corrected single-point position accuracy assurance reliability. S6. Discretize the continuous weld path and obtain the reliability of the entire weld trajectory accuracy based on the reliability of the single-point position accuracy of each discrete path point, thereby realizing the reliability assessment of the motion accuracy of the welding robot.

2. The reliability assessment method for motion accuracy of welding robots in rail transit equipment based on assured reliability as described in claim 1, characterized in that, The process of obtaining the DH parameter error includes: Sampling points are generated within the workspace of the welding robot, and the theoretical pose and actual pose of the robot end effector are obtained for each sampling point. The difference between the actual pose and the theoretical pose is identified using the LM method and expressed as a normal uncertain variable to obtain the DH parameter error.

3. The reliability assessment method for motion accuracy of welding robots in rail transit equipment based on assured reliability as described in claim 1, characterized in that, The mapping relationship between the DH parameter error and the end position error is expressed as follows: in, Indicates the position error of the robot's end effector. This represents the error coefficient matrix determined by the robot's nominal kinematic parameters. This indicates the error in the DH parameter.

4. The reliability assessment method for motion accuracy of welding robots in rail transit equipment based on assured reliability as described in claim 1, characterized in that, The position accuracy performance margin equation is expressed as follows: in, For performance margin, For DH parameter error, To set the allowable position error threshold, For the robot end-effector position error, Represents the transpose of a matrix; when When the robot's end-effector position error is within the allowable range, it is considered to be within the acceptable range. If this happens, the current target point's position accuracy is considered invalid.

5. The reliability assessment method for motion accuracy of a welding robot for rail transit equipment based on assured reliability as described in claim 1, characterized in that, The process of obtaining the reliability assurance of the single-point position accuracy includes: The most likely failure point is defined as the point on the limit state surface with the smallest Chebyshev distance from the origin. A first-order Taylor expansion of the limit state function is performed at the most likely failure point to obtain the linearized limit state function. The single-point position accuracy failure confidence and single-point position accuracy certainty reliability are calculated based on the standard normal uncertain distribution function.

6. The reliability assessment method for motion accuracy of a welding robot for rail transit equipment based on assured reliability, as described in claim 5, is characterized in that... The most likely failure point is obtained by solving a constrained optimization problem or a convex multicell bisection method.

7. The method for evaluating the motion accuracy reliability of a welding robot for rail transit equipment based on assured reliability, as described in claim 5, is characterized in that... The double-tangent plane correction includes: A first tangent plane is established at the most likely failure point of the single-point position accuracy reliability result, and the first tangent plane is translated to the corresponding tangent position on the other side of the limit state surface by utilizing the symmetry of the limit state surface to obtain the second tangent plane. The total failure confidence is approximated by the upper bound of the confidence of the failure domain outside the region enclosed by the two tangent planes. Then, the corrected single-point location accuracy assurance reliability is obtained by the following formula: in, To ensure the reliability of the corrected single-point position accuracy, This is the approximate total failure confidence.

8. The reliability assessment method for motion accuracy of a welding robot for rail transit equipment based on assured reliability as described in claim 1, characterized in that, The steps for obtaining the reliability assurance of the accuracy of the entire weld trajectory include: The continuous weld path is discretized to obtain discrete path points, and the reliability of the single-point position accuracy of each discrete path point is calculated. The entire continuous weld is regarded as a series system composed of discrete path points. The reliability of the accuracy of the entire weld trajectory is taken as the minimum value among the reliability of each discrete path point, and the reliability of the accuracy of the entire weld trajectory is output.

9. A reliability assessment method for motion accuracy of a welding robot for rail transit equipment based on assured reliability, as described in claim 8, is characterized in that... If the position error of any path point in the discrete path points exceeds a preset allowable threshold, it is considered that the accuracy of the entire weld trajectory has failed.

10. A reliability assessment device for the motion accuracy of welding robots in rail transit equipment based on assured reliability, characterized in that, The device is used to implement the reliability assessment method for motion accuracy of welding robots for rail transit equipment based on certainty reliability as described in any one of claims 1-9, and the device includes: The parameter error acquisition module is used to acquire the robot's nominal kinematic parameters and obtain the DH parameter error distribution characteristics through measurement and identification. The kinematics solution module is used to build the DH kinematic model and solve the robot joint angles based on the end-effector pose. The position error modeling module is used to establish the mapping relationship between DH parameter errors and end position errors; The performance margin construction module is used to establish the position accuracy performance margin equation based on the allowable position error threshold. The module for solving the most likely failure point is used to solve the most likely failure point in a standard normal uncertainty space using constrained optimization problems or the convex multicell bisection method. The single-point reliability calculation module is used to calculate the certainty reliability of single-point location accuracy based on the uncertain first-order reliability method. The dual-tangent plane correction module is used to correct the reliability of single-point position accuracy and reduce the overestimation caused by single-tangent plane approximation. The trajectory reliability assessment module is used for trajectory planning, path discretization, and reliability calculation of series systems for continuous weld seam paths. The results output module is used to output the reliability of single-point position accuracy, the reliability of the entire weld trajectory accuracy, and the path point information with the lowest reliability.

Images