A multi-station mechanical arm calibration device station arrangement optimization design method

Abstract

The application discloses a kind of multi-station mechanical arm calibration device station arrangement optimization design methods, steps are as follows:1. design multiple station arrangement schemes, select one station arrangement scheme as measured scheme;2. install measuring rod and place 3D measuring device in offline simulation environment;3. corresponding mechanical arm joint angle is solved based on mechanical arm inverse kinematics, and it is screened again through configuration screening, finally obtain a large number of feasible and reasonable mechanical arm joint angle;4. place 3D measuring device on other stations in measured scheme and repeat step 3, finally obtain all feasible and reasonable mechanical arm joint angle;5. optimal measurement joint angle combination is selected based on observability index, and the observability index of optimal measurement joint angle combination is obtained, and is marked as optimal observability index;6. select other station arrangement scheme as measured scheme, repeat steps 2-5, with optimal observability index as evaluation index to determine the optimal station arrangement scheme.

Description

Technical Field

[0001] This invention relates to the field of robotics, and more particularly to a method for optimizing the workstation layout of a multi-station robotic arm calibration device. Background Technology

[0002] Chinese Invention No. 202110278848.0 proposes a robot calibration device based on multi-station measurement, including a calibrated device (1), a calibration device (2), and a multi-station template (3). The multi-station template (3) provides multiple calibration measurement positions to extend the calibration measurement range of the calibration device (2) for the calibrated device (1). The calibrated device (1) is connected to a robotic arm, and the calibration device (2) is connected to the multi-station template (3). The multi-station template (3) is provided with M positioning elements (2). 0) with N clamping elements (21), M greater than 1, N greater than 1, the positioning element (20) is used to position the calibration device (2), the clamping element (21) is used to clamp the calibration device (2), its calibration device can quickly calibrate the robot at any time in the industrial site, and the calibration device can measure at multiple stations through the multi-station template, which can obtain a large measurement range. At the same time, the relative position between each station is a known information, which can be used for robot kinematic parameter identification and improve calibration accuracy. This method uses a multi-station fixture to realize the multi-station clamping of the 3D measuring device. This design of fixing multiple stations on the same substrate makes the device have excellent accuracy retention and portability, and the relative position between multiple stations can be maintained for a long time. It also allows the fixture to be theoretically quickly deployed to any robotic arm for on-site calibration. However, the arrangement of each station of the multi-station fixture currently depends entirely on the designer's design experience and lacks theoretical support. Especially for large robotic arms, since the length of the measuring rod is limited, it is necessary to make full use of the advantages of multi-station measurement to ensure calibration accuracy. This means that the arrangement of multiple stations is crucial to the final calibration effect, so a more scientific optimization design method for station arrangement is urgently needed. Summary of the Invention

[0003] To address the aforementioned technical problems, the present invention aims to provide a method for optimizing the workstation layout of a multi-station robotic arm calibration device. This method can more scientifically complete the design of the workstation layout scheme, thereby fully leveraging the advantages of multi-station measurement.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A method for optimizing the workstation layout of a multi-station robotic arm calibration device includes the following steps:

[0006] (1) Design multiple workstation layout schemes, the workstation layout schemes include the number of workstations, the XYZ coordinates and attitude angles of each workstation, and select one of the workstation layout schemes as the scheme to be tested.

[0007] (2) In the offline simulation environment, the measuring rod is installed at the end of the robotic arm, and the 3D measuring device is placed at one of the workstations in the test solution;

[0008] (3) Maintain the point constraint of the center of the measuring ball at the end of the measuring rod, distribute the attitude angle of the measuring rod evenly in the measurement space of the 3D measuring device, solve the corresponding mechanical arm joint angle based on the inverse kinematics of the mechanical arm, and then screen it through configuration screening to finally obtain a large number of feasible and reasonable mechanical arm joint angles. The configuration screening includes simulation collision detection.

[0009] (4) Place the 3D measuring device on other workstations in the test scheme and repeat step (3) to finally obtain all feasible and reasonable robot arm joint angles;

[0010] (5) Select N sets of robot arm joint angles for calibration measurement from all feasible and reasonable robot arm joint angles. These N sets of robot arm joint angles are called the optimal measurement joint angle combination. Select the optimal measurement joint angle combination based on the observation index and obtain the observation index of the optimal measurement joint angle combination, which is called the optimal observation index.

[0011] (6) Select other workstation layout schemes as the schemes to be tested, repeat steps (2) to (5) to obtain the optimal observation index corresponding to different workstation layout schemes, and use the optimal observation index as the evaluation index to determine the optimal workstation layout scheme.

[0012] Preferably, each workstation layout includes at least three workstations, each located around the robotic arm.

[0013] As a preferred option, grid coordinate points are selected at equal intervals in the workspace of the robotic arm, and all workstation layout schemes require the workstations to be placed on the corresponding grid coordinate points.

[0014] As a preferred method, the rotation method is used to design multiple workstation layout schemes in step (1) and determine the optimal workstation layout scheme in step (6). The specific process is as follows:

[0015] a1. Select an initial multi-station solution;

[0016] b1. Optimize one of the workstations as the workstation to be optimized, keep the other workstations as fixed workstations, move the workstation to be optimized to a new position near its original position, and complete the design of a new workstation layout scheme. Execute steps (2) to (5) to obtain the optimal observation index corresponding to the workstation layout scheme. Repeat the search near the original position of the workstation to be optimized until the optimal observation index corresponding to the multi-workstation scheme is optimal, thereby completing a position change.

[0017] c1. Repeat step b1 by taking the other workstations as workstations to be optimized in turn, thereby completing one round of position rotation.

[0018] d1. Repeat steps b1-c1 continuously for multiple rounds of position rotation. When the optimal observability index no longer increases or the increase is less than the threshold ε1, exit the loop and output the multi-station scheme that obtains the optimal observability index during the rotation process to complete the iterative optimization.

[0019] Preferably, in step b1, to achieve 2D position optimization of the workstation, a stepwise search method is used for local optimization, and the steps are as follows:

[0020] a2. Using the grid coordinates of one of the workstations as the center point, search for all grid points within a local area near the center point and obtain the optimal observation index corresponding to all grid points.

[0021] b2. If the optimal observation index of all grid points in the local area no longer increases relative to the optimal observation index of the center point, or the increase is less than the threshold ε2, then exit the loop and take the grid point with the best optimal observation index in the local area as the optimized workstation position; otherwise, take the grid point with the best optimal observation index in the local area as the center point of the next step, and record it as the third grid point.

[0022] c2. If the number of searches reaches the set upper limit N, then exit the loop and take the third grid point as the optimized workstation position; otherwise, search all grid points within the local range of the third grid point and obtain the optimal observation index corresponding to all grid points, increment the search count by 1, and jump to step b2 to solve.

[0023] As a preferred method, in order to determine the height parameter of the workstation, multiple candidate values ​​are first prepared for the height parameter of the workstation. Then, in step b1, during the repeated search near the original position of the workstation to be optimized, the height parameter of the workstation to be optimized is set to different candidate values ​​in turn until the optimal observability index corresponding to the multi-workstation scheme is optimal, thereby completing one position transformation.

[0024] As a preferred option, for various irregular workspaces, when there are obstacles in the working environment of the robotic arm that prevent the placement of workstations, it is necessary to first mark the grid points in the obstacle area of ​​the workspace, and then during the workstation rotation optimization process, during each local search step, i.e. step a2 and step c2, an obstacle judgment is added. The workstation to be optimized can only be moved to the grid without obstacles in order to complete the design of a new workstation layout scheme. Then, steps (2)-(5) are executed to obtain the optimal observation index corresponding to the workstation layout scheme.

[0025] The present invention, by adopting the above technical solution, has the following beneficial effects:

[0026] 1. This invention is based entirely on iterative optimization of the optimal observation index, and is no longer limited by the designer's experience. It has a theoretical foundation and is easier to find the global optimum.

[0027] 2. This invention adopts an iterative optimization-based design method, which is a heuristic algorithm. Its initial value is selected based on engineering experience and often already has global information, which is close to the global optimum and has high computational efficiency.

[0028] 3. It can meet the optimization needs of different heights, different opening orientations, and irregularly shaped workspaces, greatly improving the diversity of multi-workstation layout schemes. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the structure of the present invention;

[0030] Figure 2 This is a 2D position optimization diagram of the workstation according to the present invention;

[0031] Figure 3 This is a schematic diagram showing the optimization of the workstation height parameters and opening orientation parameters of the present invention;

[0032] Figure 4 This is a schematic diagram illustrating the optimization of workstation layout in the presence of obstacles according to the present invention. Detailed Implementation

[0033] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0034] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0035] Example 1,

[0036] like Figure 1 As shown, a method for optimizing the workstation layout of a multi-station robotic arm calibration device includes the following steps:

[0037] (1) Design multiple workstation layout schemes, the workstation layout schemes include the number of workstations, the XYZ coordinates and attitude angles of each workstation, and select one of the workstation layout schemes as the scheme to be tested.

[0038] (2) In the offline simulation environment, the measuring rod 13 is installed at the end of the robotic arm 11, and the 3D measuring device 8 is placed at one of the workstations in the test scheme.

[0039] (3) Maintain the point constraint of the center 12 of the measuring ball at the end of the measuring rod, distribute the attitude angle of the measuring rod evenly in the measurement space of the 3D measuring device, solve the corresponding mechanical arm joint angle based on the inverse kinematics of the mechanical arm, and then screen it through configuration screening to finally obtain a large number of feasible and reasonable mechanical arm joint angles. The configuration screening includes simulation collision detection.

[0040] (4) Place the 3D measuring device on other workstations in the test scheme and repeat step (3) to finally obtain all feasible and reasonable robot arm joint angles;

[0041] (5) Select N sets of robot arm joint angles for calibration measurement from all feasible and reasonable robot arm joint angles. These N sets of robot arm joint angles are called the optimal measurement joint angle combination. Select the optimal measurement joint angle combination based on the observation index and obtain the observation index of the optimal measurement joint angle combination, which is called the optimal observation index.

[0042] (6) Select other workstation layout schemes as the schemes to be tested, repeat steps (2) to (5) to obtain the optimal observation index corresponding to different workstation layout schemes, and use the optimal observation index as the evaluation index to determine the optimal workstation layout scheme.

[0043] The measurement space of the 3D measurement device refers to the intersection point of the axes of three orthogonally placed displacement sensors defined as O1. Point O1 is used as three reference planes that are parallel to the measurement planes of the three sensor probes. The three reference planes divide the spherical space with the measurement point as the center into eight equal parts. One-eighth of the open area facing outward from the three displacement sensors is taken as the measurement space.

[0044] The specific method for measuring the attitude angle of the uniformly distributed 3D measuring rod in the measurement space of the displacement measuring device is as follows: The coordinates on the sphere are defined using latitude and longitude in geography according to a certain graduation value. The direction of the line connecting the intersection of each meridian and parallel to the center of the sphere is the z-axis direction of the measuring spherical coordinate system at the end of the robotic arm in the measured pose. The x-axis direction is obtained by rotating an initial x-axis around the z-axis by an angle γ, where γ ranges from -180° to 180° according to the graduation value d. γ By traversing the coordinate system and determining the z-axis and x-axis, the entire end-point measurement spherical coordinate system can be determined.

[0045] Alternatively, instead of using latitude and longitude to define coordinate points on the sphere, points can be evenly distributed across the sphere. The direction of the line connecting these points to the center of the sphere is defined as the z-axis direction of the spherical coordinate system measured by the robotic arm's end effector. The x-axis direction is obtained by rotating an initial x-axis around the z-axis by an angle γ, where γ ranges from -180° to 180° in increments d. γ By traversing the coordinate system and determining the z-axis and x-axis, the entire end-point measurement spherical coordinate system can be determined.

[0046] The simulated collision detection includes static collision detection. Static collision detection is performed when the robotic arm moves to a certain joint angle and then collides with itself or with the outside world when it is statically placed. If such a joint angle is detected, it is discarded. The collision detection in the simulation determines whether a self-collision occurs between the joints and links of the robotic arm or a collision with the outside world occurs between the robotic arm and the ground, base, or measuring device.

[0047] The observational index can characterize the ill-conditioned nature of the Jacobian matrix when kinematic parameters are identified using optimization-based methods. Specifically, it is represented by the optimal observational index. There are multiple evaluation methods for the observational index. The larger the value of the optimal observational index, the better the flexibility of the robotic arm at that set of joint angles.

[0048] The joint angle measurement was selected using the DETMAX algorithm, and the specific steps are as follows:

[0049] a. Randomly select m sets of data from all feasible and reasonable joint angle data, and calculate the observation index values ​​corresponding to these m sets of data;

[0050] b. Traverse through the remaining data to find a set of data such that the observed index value of the (m+1) sets of data after adding this set of data to the original m sets of data is maximized;

[0051] c. Iterate through and filter through the (m+1) sets of data to find a set of data that maximizes the observable index value of the m sets of data after deleting this set of data.

[0052] d. Continue this process in a loop until the deleted data set is exactly the same as the added data set. This will give you the m data sets with the largest observational index values, thus enabling you to select the optimal combination of measured joint angles.

[0053] like Figure 1 As shown, each workstation layout includes at least three workstations, with each workstation positioned around the robotic arm.

[0054] In the workspace of the robotic arm, grid coordinate points are selected at equal intervals, and all workstation layout schemes require the workstations to be placed on the corresponding grid coordinate points.

[0055] The rotation method is used to design multiple workstation layout schemes in step (1) and determine the optimal workstation layout scheme in step (6). The specific process is as follows:

[0056] a1. Select an initial multi-station solution;

[0057] b1. Optimize one of the workstations as the workstation to be optimized, keep the other workstations as fixed workstations, move the workstation to be optimized to a new position near its original position, complete the design of a new workstation layout scheme, and then execute steps (2) to (5) to obtain the optimal observation index corresponding to the workstation layout scheme. Repeat the search near the original position of the workstation to be optimized until the optimal observation index corresponding to the multi-workstation scheme is optimal, thereby completing a position change.

[0058] c1. Repeat step b1 by taking the other workstations as workstations to be optimized in turn, thereby completing one round of position rotation.

[0059] d1. Repeat steps b1-c1 continuously for multiple rounds of position rotation. When the optimal observability index no longer increases or the increase is less than the threshold ε1, exit the loop and output the multi-station scheme that obtains the optimal observability index during the rotation process to complete the iterative optimization.

[0060] The setting of the threshold ε1 is crucial. If it is set too small, it will be difficult to break out of the loop, significantly reducing the solution efficiency. Furthermore, the optimal observation index of the DETMAX algorithm itself fluctuates with changes in the initial candidate group, so pursuing a small improvement is not very meaningful. However, if it is set too large, the fluctuation of the optimal observation index may cause premature exit from the loop, thus failing to find the global optimum. After parameter tuning and testing, this embodiment sets ε1 to 0.007.

[0061] Since each position change involves only the optimization of one workstation, the number of optimization parameters is significantly reduced, allowing most existing mature optimization algorithms (such as search-based optimization algorithms, gradient-based optimization algorithms, and heuristic optimization algorithms) to be applied and solve the optimization problem well. Because the solution domain has been meshed in this embodiment, a stepwise search method is used for local optimization for simplification, as follows: Figure 2 As shown, Figure 2 The local area used is 3×3, but other areas can be selected as needed.

[0062] In step b1, to achieve 2D position optimization of the workstation, a stepwise search method is used for local optimization, and the steps are as follows:

[0063] a2. Using the grid coordinates of one of the workstations (e.g., Figure 2 Using the yellow grid in the diagram as the center point, a search is performed on all grid points within a local area near the center point, and the optimal observation index (e.g., ...) is obtained for all grid points. Figure 2 (The green grid in the middle);

[0064] b2. If the optimal observability index of all grid points within the local area no longer increases relative to the optimal observability index of the center point, or the increase is less than the threshold ε2, then exit the loop and take the grid point with the optimal observability index within the local area as the optimized workstation position, denoted as the second grid (e.g., Figure 2 (The red grid in the image) Otherwise, the grid point with the best local observation index is taken as the center point for the next step, and is denoted as the third grid point (e.g., the red grid in the image). Figure 2 (The orange grid in the middle);

[0065] c2. If the number of searches reaches the set upper limit N, then exit the loop and reset the third grid point (e.g., Figure 2 The orange grid in the diagram is used as the optimized workstation location; otherwise, all grid points within the local range of the third grid point are searched, and the optimal observation index corresponding to all grid points is obtained. The search count is incremented by 1, and the process jumps to step b2 to solve the problem.

[0066] Similar to ε1, the value of ε2 needs to balance solution efficiency and accuracy. Moreover, compared to single-round position changes, the result of a single local search is more affected by the random fluctuations of the optimal observation index, so ε2 needs to be smaller. After parameter tuning and testing, this embodiment sets ε2 to 0.002. Furthermore, the upper limit N of the search count also needs to be carefully set. If N is set too small, each search will only search a smaller local range, which will inevitably reduce the probability of finding a local optimum, and may even lead to insufficient improvement in the optimal observation index of a single round of position changes, causing premature exit from the loop and missing the global optimum. On the other hand, this may prevent each position change from getting trapped in a local optimum, thus increasing the probability of the algorithm finding the global optimum. More importantly, reducing N can significantly reduce the amount of iterative computation, thereby improving the overall efficiency of the algorithm.

[0067] Example 2

[0068] The parts that are the same as those in Example 1 will not be repeated here. The differences are as follows:

[0069] To determine the height parameter of the workstation, multiple candidate values ​​are first prepared for the height parameter of the workstation. Then, in step b1, during the repeated search near the original position of the workstation to be optimized, the height parameter of the workstation to be optimized is set to different candidate values ​​in turn until the optimal observability index corresponding to the multi-workstation scheme is optimal, thereby completing one position transformation.

[0070] In a scenario where the open measurement area (opening) of the measuring device is tilted to one side, in order to determine the opening orientation angle of the workstation, multiple candidate values ​​are first prepared for the opening orientation angle of the workstation. Then, in step b1, during the repeated search near the original position of the workstation to be optimized, the opening orientation angle parameter of the workstation to be optimized is set to different candidate values ​​in turn until the optimal observation index corresponding to the multi-workstation scheme is optimal, thereby completing one position transformation.

[0071] like Figure 3 As shown, taking workstation height as an example, before performing 2D position optimization on a workstation, we first fix other parameters unchanged. Thus, with the 2D position parameter and workstation orientation parameter unchanged, we calculate all parameters within the range of the workstation height parameter, finding the parameter that yields the maximum optimal observable index as the initial value of the local optimum parameter, obtaining the first height parameter. Then, we use a stepwise search method (to solve...) Figure 3 When the green area in the first row is selected, the 2D position optimization of the workstation is achieved (i.e., the process of steps a2 to b2 in Example 1). The optimization is performed among the three parameters—left, right, and middle—of the first height parameter. The parameter that achieves the maximum optimal observable index is used as the locally optimal parameter for the next step of the 2D position local search. Figure 3The orange parameters in the first row (indicated by the search parameters) gradually change with each search until the final optimal 2D position is determined. The workstation height parameter that yields the highest optimal observable index is then output as the final optimization result. Figure 3 (The red parameters in the first row).

[0072] It is important to note that, regarding installation height, based on small-batch experimental results, we found that if conventional boundary conditions are used for installation height, it is easy to get trapped in local optima. That is, the installation height remains at a single boundary value (e.g., a minimum of -150mm or a maximum of 450mm) without transitioning to another boundary value. This may be because a multi-station layout containing both the lowest and highest installation heights is highly beneficial for improving the observable optimality. However, during station rotation, once the installation height of several stations falls into one boundary value, the remaining stations will immediately fall into another. This "high-high-low" configuration is difficult to change in subsequent iterations, which is highly detrimental to finding the global optimum.

[0073] Therefore, we directly connect the upper and lower boundary values ​​of the workstation height to form a closed loop. That is, the left parameter of the lowest value -150mm is the highest value 450mm. In this way, during local search, it is still possible to change this "high-high-low" configuration, for example, from "high-high-low" to "low-high-low". This solves the problem very well.

[0074] like Figure 3 As shown in the second row of figures, regarding the orientation angle of the workstation, since this embodiment only considers the rotation angle of the 3D measuring device around the vertical line, its parameter range itself is a closed loop (e.g., Figure 3 As shown in the second row of figures, -180° is equivalent to 180°, so there is no need to consider this issue, and we can directly find the local optimization.

[0075] Example 3

[0076] The parts that are the same as those in Example 1 will not be repeated here. The differences are as follows:

[0077] For various irregular workspaces, when there are obstacles in the working environment of the robotic arm that prevent the placement of workstations, it is necessary to first mark the grid points in the obstacle area of ​​the workspace. Then, during the workstation rotation optimization process, during each local search step, i.e. step a2 and step c2, an obstacle judgment is added. The workstation to be optimized can only be moved to the grid without obstacles in order to complete the design of a new workstation layout scheme. Then, steps (2)-(5) are executed to obtain the optimal observation index corresponding to the workstation layout scheme.

[0078] like Figure 4As shown, because the robotic arm (such as Figure 4 The work environments (purple areas in the diagram) vary and may contain obstacles. Simply mark the grid points within the obstacle area of ​​the workspace (e.g., ...). Figure 4 (The pink area in the image), and based on engineering experience, in feasible areas (such as...) Figure 4 Select a specified number of initial workstations (e.g., within the blue area in the image) Figure 4 (The green grid in the image) Then, during the workstation rotation optimization process, an obstacle detection is added at each step of the local search, so that each workstation will iteratively move to the optimal position within the feasible area (e.g., the green grid in the image). Figure 4 (The yellow grid in the middle).

[0079] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, alterations, deletions of some features, additions of features, or recombinations of features to the above embodiments within the scope of the present invention without departing from the principles and spirit of the present invention. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the innovative principles of the present invention shall still fall within the scope of the technical solutions of the present invention.

Claims

1. A method for optimizing the workstation layout of a multi-station robotic arm calibration device, characterized in that: Includes the following steps: (1) Design multiple workstation layout schemes, the workstation layout schemes include the number of workstations, the XYZ coordinates and attitude angles of each workstation, and select one of the workstation layout schemes as the scheme to be tested. (2) In the offline simulation environment, the measuring rod is installed at the end of the robotic arm, and the 3D measuring device is placed at one of the workstations in the test solution; (3) Maintain the point constraint of the center of the measuring ball at the end of the measuring rod, distribute the attitude angle of the measuring rod evenly in the measurement space of the 3D measuring device, solve the corresponding mechanical arm joint angle based on the inverse kinematics of the mechanical arm, and then screen it through configuration screening to finally obtain a large number of feasible and reasonable mechanical arm joint angles. The configuration screening includes simulation collision detection. (4) Place the 3D measuring device on other workstations in the test scheme and repeat step (3) to finally obtain all feasible and reasonable robot arm joint angles; (5) Select N sets of robot arm joint angles for calibration measurement from all feasible and reasonable robot arm joint angles. These N sets of robot arm joint angles are called the optimal measurement joint angle combination. Select the optimal measurement joint angle combination based on the observation index and obtain the observation index of the optimal measurement joint angle combination, which is called the optimal observation index. (6) Select other workstation layout schemes as the schemes to be tested, repeat steps (2) to (5) to obtain the optimal observation index corresponding to different workstation layout schemes, and use the optimal observation index as the evaluation index to determine the optimal workstation layout scheme. In the workspace of the robotic arm, grid coordinate points are selected at equal intervals, and all workstation layout schemes require the workstations to be placed on the corresponding grid coordinate points; The rotation method is used to design multiple workstation layout schemes in step (1) and determine the optimal workstation layout scheme in step (6). The specific process is as follows: a1. Select an initial multi-station solution; b1. Optimize one of the workstations as the workstation to be optimized, keep the other workstations as fixed workstations, move the workstation to be optimized to a new position near its original position, and complete the design of a new workstation layout scheme. Execute steps (2) to (5) to obtain the optimal observation index corresponding to the workstation layout scheme. Repeat the search near the original position of the workstation to be optimized until the optimal observation index corresponding to the multi-workstation scheme is optimal, thereby completing a position change. c1. Repeat step b1 by taking the other workstations as workstations to be optimized in turn, thereby completing one round of position rotation. d1. Repeat steps b1-c1 continuously for multiple rounds of position rotation. When the optimal observability index no longer increases or the increase is less than the threshold ε1, exit the loop and output the multi-station scheme that obtains the optimal observability index during the rotation process to complete the iterative optimization.

2. The method for optimizing the workstation layout of a multi-station robotic arm calibration device according to claim 1, characterized in that: Each workstation layout includes at least three workstations, each located around the robotic arm.

3. The method for optimizing the workstation layout of a multi-station robotic arm calibration device according to claim 1, characterized in that: In step b1, to achieve 2D position optimization of the workstation, a stepwise search method is used for local optimization, and the steps are as follows: a2. Using the grid coordinates of one of the workstations as the center point, search for all grid points within a local area near the center point and obtain the optimal observation index corresponding to all grid points. b2. If the optimal observation index of all grid points in the local area no longer increases relative to the optimal observation index of the center point, or the increase is less than the threshold ε2, then exit the loop and take the grid point with the best optimal observation index in the local area as the optimized workstation position; otherwise, take the grid point with the best optimal observation index in the local area as the center point of the next step, and record it as the third grid point. c2. If the number of searches reaches the set upper limit N, then exit the loop and take the third grid point as the optimized workstation position; otherwise, search all grid points within the local range of the third grid point and obtain the optimal observation index corresponding to all grid points, increment the search count by 1, and jump to step b2 to solve.

4. The method for optimizing the workstation layout of a multi-station robotic arm calibration device according to claim 1, characterized in that: To determine the height parameter of the workstation, multiple candidate values ​​are first prepared for the height parameter of the workstation. Then, in step b1, during the repeated search near the original position of the workstation to be optimized, the height parameter of the workstation to be optimized is set to different candidate values ​​in turn until the optimal observability index corresponding to the multi-workstation scheme is optimal, thereby completing one position transformation.

5. The method for optimizing the workstation layout of a multi-station robotic arm calibration device according to claim 3, characterized in that: For various irregular workspaces, when there are obstacles in the working environment of the robotic arm that prevent the placement of workstations, it is necessary to first mark the grid points in the obstacle area of ​​the workspace. Then, during the workstation rotation optimization process, during each local search step, i.e. step a2 and step c2, an obstacle judgment is added. The workstation to be optimized can only be moved to the grid without obstacles in order to complete the design of a new workstation layout scheme. Then, steps (2)-(5) are executed to obtain the optimal observation index corresponding to the workstation layout scheme.

Images