A fast calibration method for industrial robots based on continuous motion measurement

Abstract

The present invention relates to a rapid calibration method for an industrial robot based on continuous motion measurement. The method comprises: based on a pool of alternative configurations of an industrial robot workspace, selecting the optimal measurement configuration to determine the waypoints through which the industrial robot moves; based on the distribution of the waypoints, determining the planning trajectory of the smooth continuous motion with the best time; measuring the actual trajectory of the industrial robot, aligning the planned trajectory with the measured trajectory to determine the actual data of the selected measurement configuration; constructing a kinematic model of the industrial robot, and calibrating the industrial robot using a rapid calibration method based on a CMM strategy. The present invention can not only improve the absolute positioning accuracy of the industrial robot, but also reduce the data acquisition time of the kinematic calibration of the industrial robot, thereby improving the calibration efficiency of the actual production line robot.

Description

Technical Field

[0001] The invention relates to a rapid calibration method for an industrial robot based on continuous motion measurement, and belongs to the technical field of robot calibration. Background Art

[0002] For industrial robots, absolute positioning accuracy is an important criterion for completing complex and precise tasks. Industrial robots are composed of many parts and components with complex combinations. The processing errors of the parts and the assembly errors between the parts are the main factors affecting the absolute positioning accuracy. The kinematic calibration of industrial robots is an effective and cost-effective means to improve the absolute positioning accuracy of industrial robots by identifying the kinematic parameter errors through measurement.

[0003] However, at present, the time cost of the measurement phase of industrial robot calibration is still too high and not flexible enough. Among them, the classic configuration acquisition can be described as a cyclic measurement process. First, the robot is controlled to move to the first measurement configuration. After the end is stable and stationary, the measuring instrument (program) is turned on to obtain the end posture and joint displacement data. Then the robot is controlled to move to the next measurement configuration, and the cycle is repeated until the data under all measurement configurations are acquired. This is a measurement strategy of "Move-Settle-Measure" (Loop of Move-Settle-Measure, LMSM) that executes a loop. The characteristic of this strategy is that there is no need to consider the motion error and measurement delay problems during the measurement process. The robot is in a stationary state during the measurement process. After the instrument acquires the data, the robot executes the motion program. However, the measurement time under this measurement strategy increases linearly with the increase of configurations, which is not an efficient calibration method. For the online calibration or regular recalibration needs of robots in the production line due to collisions, wear of transmission components, replacement of end effectors, and addition of robots of the same type in the production line, it is expected that the calibration system will have higher efficiency. Therefore, a calibration method that can reduce the time of calibration mid-position measurement and at the same time meet the original error modeling, identification and other technical conditions to ensure the positioning accuracy requirements of industrial robots can be of great significance to improving the accuracy and efficiency of industrial robot production. Summary of the invention

[0004] The present invention provides an industrial robot rapid calibration method based on continuous motion measurement, aiming to solve at least one of the technical problems existing in the prior art.

[0005] The technical solution of the present invention relates to a rapid calibration method for an industrial robot based on continuous motion measurement. The method according to the present invention comprises the following steps:

[0006] S100, based on the pool of candidate configurations of the industrial robot workspace, selecting an optimal measurement configuration to determine a waypoint through which the industrial robot moves;

[0007] S200, determining a planning trajectory of a smooth continuous motion with optimal time based on the distribution of waypoints;

[0008] S300, measuring the actual trajectory of the industrial robot, aligning the planned trajectory with the measured trajectory to determine actual data of the selected measurement configuration;

[0009] S400, construct a kinematic model of the industrial robot, and calibrate the industrial robot using a fast calibration method based on a continuous motion measurement strategy.

[0010] Further, the step S100 includes:

[0011] S110, preparing alternative configurations in the working space of the industrial robot, and initializing a certain number of measurement configurations;

[0012] S120, using a search method of DETMAX plus simulated annealing to search among candidate configurations to determine the optimal measurement configuration.

[0013] Further, the step S200 includes:

[0014] S210, connecting waypoints using a spline interpolation method to plan a smooth path passing through the waypoints, and obtaining a movement distance between any two adjacent waypoints;

[0015] S220, according to the movement distance between any two adjacent waypoints, use the segmented S model to plan the trajectory between every two waypoints, and obtain the time-related movement parameters and movement time between the two adjacent waypoints;

[0016] S230: Determine the passing order of the waypoints with the shortest moving time based on a genetic algorithm according to the moving time between any adjacent waypoints.

[0017] Furthermore, in step S300, the measured velocity curve and the planned velocity curve are aligned according to the trough characteristics of the velocity curve, so as to obtain measurement data of the measurement configuration required for calibration from the measured velocity curve.

[0018] Further, the step S400 includes:

[0019] S410, constructing a kinematic model of the industrial robot to establish an error mapping model;

[0020] S420. Establish an inverse kinematics error model for industrial robots;

[0021] S430, building a measurement data acquisition platform to respectively collect the calibration geometry sets required in the CMM strategy and the “Loop of Move-Settle-Measure, LMSM” strategy;

[0022] S440, respectively using the calibration positioning data obtained by the CMM strategy and the LMSM strategy to identify kinematic error parameters and compensate for the kinematic parameters of the industrial robot.

[0023] Furthermore, the industrial robot is a SPM-RP parallel robot, and in step S410,

[0024] Based on the moving platform coordinate system In the base coordinate system The rotation matrix in the transformation relationship and translation vectors , a single branch of the SPM-RP parallel robot The error equation is expressed as follows:

[0025] ,

[0026] In the formula, i =1,2,3,4, indicating the sequence number of the branch chain;

[0027] Indicates branch chain The center of the lower revolute joint R in the PRPaR structure is relative to The position of the coordinate system, yes The position error vector of ;

[0028] Indicates that SPM-RP is in zero position, branch chain The center of the upper rotation joint R in the PRPaR structure is relative to The position of the coordinate system, yes The position error vector of ; Indicates branch chain The displacement of the translation joint P in the PRPaR structure; It represents the joint displacement deviation, which characterizes the deviation between the measured value and the calculated value of the kinematic error parameter in the joint displacement direction;

[0029] R Q,i Indicates branch chain Connected to Joint coordinate system of points Relative to The rotation matrix is ​​related to the axis direction of the moving pair; represents the differential rotation matrix, the axis angle parameters are included in ,and By fixed to Joint coordinate system of points Relative to the base coordinate system Around Axis rotation obtained by angle;

[0030] Indicates branch chain The parallelogram structure in the PRPaR structure Equivalent rod length, error of equivalent rod length ,and . Represents an intermediate variable , Represents an intermediate variable The error term of . e z express Unit vector in direction ;

[0031] In the formula, the superscript represents the nominal terms of kinematic parameters.

[0032] Furthermore, in step S420, the end position information of the industrial robot is used as input, and the joint displacement deviation of the industrial robot is used as output. The error mapping model is expressed as:

[0033]

[0034] in,

[0035]

[0036] In the formula, the subscript Indicates the first Branched chain sequence, subscript Indicates A sequence of measured configurations, A j Indicates The error mapping matrix under the measurement configuration is: and Indicates the measured end position of the industrial robot;

[0037] Among them, the first The joint displacement of a joint is expressed as follows:

[0038] ,

[0039] According to the joint encoder measurement The joint displacement of the joint is , get the joint displacement deviation It is expressed as follows:

[0040] .

[0041] Further, the step S440 includes:

[0042] S441, obtaining a set of equations for kinematic parameter identification according to the established inverse kinematics error model and the actual data of the measurement configuration collected based on the CMM strategy;

[0043] S442. Use the extended Kalman filter method to obtain an estimated value of the kinematic parameter error under the CMM strategy, and update the nominal kinematic parameters with the estimated value of the parameter error.

[0044] The technical solution of the present invention also relates to a computer-readable storage medium on which program instructions are stored, and the above-mentioned method is implemented when the program instructions are executed by a processor.

[0045] The technical solution of the present invention also relates to a rapid calibration system for an industrial robot based on continuous motion measurement, wherein the system comprises a computer device, and the computer device comprises the above-mentioned computer-readable storage medium.

[0046] The beneficial effects of the present invention are as follows:

[0047] The present invention is based on the rapid calibration method of industrial robots based on the continuous-moving measurement (CMM) strategy, which can eliminate the waiting time for measurement and complete the calibration measurement work during the robot movement process. In the CMM measurement strategy, the system obtains the required configuration data for calibration from a continuous movement of the robot, and the quality of the obtained data is basically equivalent to the data quality obtained using the traditional "movement-stabilization-measurement" (Loop of Move-Settle-Measure, LMSM) measurement strategy. Therefore, the use of the CMM-based rapid calibration method of the present invention can not only improve the absolute positioning accuracy of the industrial robot, but also reduce the data collection time for the kinematic calibration of the industrial robot, thereby improving the calibration efficiency of the actual production line robot.

[0048] The method of the present invention is applied to the SPM-RM parallel robot, and the efficiency and accuracy of the rapid calibration scheme based on continuous motion measurement are verified. Compared with the traditional calibration method based on the classic LMSM measurement strategy, the rapid calibration based on the CMM strategy achieves a 93.13% improvement in calibration efficiency at the expense of only 1.77% in positioning accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is an overall flow chart of the CMM-based measurement strategy according to the method of the present invention.

[0050] Figure 2 It is a flow chart of configuration selection for calibration optimization measurement according to the method of the present invention.

[0051] Figure 3 Schematic diagram of curves of jerk, acceleration, velocity and displacement of the segmented S-shaped model according to the method of the present invention.

[0052] Figure 4 3D model structure schematic diagram of the SPM-RP robot according to the method of the present invention.

[0053] FIG. 5( a ) is a schematic diagram of the overall structure of the SPM-RP robot according to the method of the present invention.

[0054] FIG5( b ) is a schematic diagram of a single-branched chain structure of a SPM-RP robot according to the method of the present invention.

[0055] FIG. 6( a ) is a schematic diagram of a physical model of data acquisition for calibration according to the method of the present invention.

[0056] FIG. 6( b ) is a simplified structural diagram of calibration data acquisition according to the method of the present invention.

[0057] FIG. 7( a ) is a spatial curve diagram of the SPM-RP continuous trajectory calibrated according to the method of the present invention.

[0058] FIG. 7( b ) is a planned speed curve diagram and a measured speed curve diagram of the SPM-RP continuous trajectory calibrated according to the method of the present invention.

[0059] Figure 8 It is a schematic diagram of 100 independent verification configuration data sets randomly collected in the SPM-RP workspace for verifying the fast calibration method of the present invention and the traditional calibration method.

[0060] FIG9( a ) is the absolute positioning error value of the SPM-RP before calibration according to the method of the present invention.

[0061] FIG9( b ) shows the absolute positioning error values ​​of the SPM-RP after calibration using the CMM strategy and the LMSM strategy respectively according to the method of the present invention. DETAILED DESCRIPTION

[0062] The concept, specific structure and technical effects of the present invention will be clearly and completely described below in conjunction with the embodiments and drawings to fully understand the purpose, scheme and effect of the present invention.

[0063] It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed" or "connected" to another feature, it may be directly fixed or connected to another feature, or it may be indirectly fixed or connected to another feature. The singular forms "a", "said" and "the" used herein are also intended to include the plural forms, unless the context clearly indicates otherwise. In addition, unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art. The terms used in this specification are intended only to describe specific embodiments and are not intended to limit the invention. The term "and / or" used herein includes any combination of one or more of the related listed items.

[0064] It should be understood that, although the term first, second, third etc. may be adopted to describe various elements in the present disclosure, these elements should not be limited to these terms. These terms are only used to distinguish the same type of elements from each other. For example, without departing from the scope of the present disclosure, the first element may also be referred to as the second element, and similarly, the second element may also be referred to as the first element. The use of any and all examples or exemplary language ("for example", "such as" etc.) provided herein is only intended to better illustrate embodiments of the present invention, and unless otherwise required, will not impose limitations on the scope of the present invention.

[0065] Reference Figures 1 to 3 In some embodiments, the rapid calibration method of an industrial robot based on continuous motion measurement according to the present invention comprises at least the following steps:

[0066] S100, based on the pool of candidate configurations of the industrial robot workspace, selecting an optimal measurement configuration to determine a waypoint through which the industrial robot moves;

[0067] S200, determining a planning trajectory of a smooth continuous motion with optimal time based on the distribution of waypoints;

[0068] S300, measuring the actual trajectory of the industrial robot, aligning the planned trajectory with the measured trajectory to determine actual data of the selected measurement configuration;

[0069] S400, construct the kinematic model of the industrial robot and calibrate the industrial robot using a fast calibration method based on the CMM strategy.

[0070] The method of the invention can realize rapid calibration under the CMM strategy and is an optimal selection method for calibration measurement configuration. It adopts a CMM measurement strategy for obtaining configuration data required for calibration from one continuous motion of the robot and can be applied to the kinematic calibration platform of the new SPM-RP parallel robot.

[0071] It should be noted that the above steps S100 to S300 of the present invention belong to a general strategy for obtaining the configuration data required for calibration from a continuous motion of the robot, which is applicable to the measurement in the kinematic calibration of general industrial robots. The process of continuous motion measurement is as follows: Figure 1 As shown, steps S100 to S300 correspond to Part 1 to Part 3 in the figure respectively. Step S400 belongs to a rapid calibration method for an industrial robot based on the continuous motion measurement of steps S100 to S300, which uses an SPM-RP parallel industrial robot as a specific embodiment to verify the calibration method of the continuous motion measurement in steps S100 to S300.

[0072] Specific implementation of step S100

[0073] The present invention selects a set of optimal configurations that can improve the accuracy of kinematic parameter error identification as much as possible from the pool of candidate configurations of the workspace of the industrial robot. The selected optimal configurations are the waypoints that the robot moves through. Figure 2 , the optimal configuration selection process includes at least the following sub-steps:

[0074] S110, preparing a candidate configuration pool and determining an initial measurement configuration.

[0075] Specifically, in the workspace of the industrial robot, N number of alternative configurations are prepared at equal intervals or randomly. It should be noted that the above-mentioned alternative configurations can be represented by the posture of the end Cartesian space or the displacement of the joint space. The initial measurement configuration is randomly selected from the alternative configurations in the configuration pool. Specifically, the measurement configuration is a subset of the alternative configurations, and a certain number of measurement configurations are randomly selected from the alternative configurations as the initialization measurement configurations. After determining a set of initial measurement configurations, the DETMAX method begins to iterate and selects the optimal measurement configuration from the alternative configurations. Furthermore, in order to obtain a higher precision identification effect, the number M of measurement configurations is at least equal to 2 to 3 times the integer part of the quotient obtained by dividing the number of kinematic error parameters by the dimension of a single configuration. Therefore, the measurement configuration is selected from the alternative configurations in the configuration pool, and the number is .

[0076] S120 and DETMAX are used in combination with simulated annealing method to search for the optimal configuration.

[0077] The DETMAX (DETerminant MAXimization method) used in the present invention iteratively searches for the optimal configuration from the candidate configuration pool to replace the same number of initial measurement configurations, and each iteration includes an AddConfiguration subroutine and a MinusConfiguration subroutine in turn. After multiple AddConfiguration and MinusConfiguration iterations, when the observability index tends to be stable or converged, the optimal set of measurement configurations can be obtained. Furthermore, in order to prevent the iteration from falling into the local optimum, the present invention adds simulated annealing to the above AddConfiguration and MinusConfiguration iterations, so that the measurement configuration has the probability of jumping out of the local optimum.

[0078] Here is a specific example to illustrate. For example, 1) add a first-in first-out (FIFO) bit stack in the DETMAX method, whose capacity is ; 2) In each DETMAX search iteration, after the AddConfiguration subroutine is executed, the optimal configuration obtained by the AddConfiguration subroutine is placed in the configuration stack. Since the number of iterations is much larger than the stack capacity, After the first iteration, the stack executes the FIFO rule; 3) In each DETMAX search iteration, the MinusConfiguration subroutine introduces a simulated annealing algorithm to increase the randomness and accept the deletion of the previous 4) Finally, it helps the DETMAX method to jump out of the local optimal measurement configuration during the iterative optimization process to increase the probability of finding the global optimal measurement configuration.

[0079] Specific implementation of step S200

[0080] Based on the waypoint distribution obtained in the above step S100, the present invention automatically plans a smooth spatial path curve, which satisfies both the continuity of speed planning and the time optimization of the overall curve. Segmented S-type speed trajectory planning is carried out for the curve between two waypoints, and the motion parameters and motion time of each segment are determined, wherein the determined motion parameters include jerk, acceleration, speed and displacement, etc. The crossing sequence of the waypoints is reorganized based on the genetic algorithm to ensure the time optimization of the spatial path curve. The determination of the planning trajectory of the smooth and continuous motion with the optimal time includes at least the following steps: S210, smooth path planning through the waypoints.

[0081] Using the spline interpolation method to connect waypoints, the connection between two adjacent waypoints can be achieved, that is,

[0082]

[0083] In the formula, Indicates Segment curve (located at With The interpolation function between waypoints, is a position variable, indicating the Segment curve interpolation points , or , and It is Waypoint , or The shape of the interpolation function is determined by the parameter Control, easy to get the waypoint by function integration and waypoints The absolute distance between the interpolation paths .

[0084] It should be noted that the above path planning of the present invention can determine the feasible route from the starting measurement configuration to the target measurement configuration, which does not directly consider the time factor. The path planning not only generates a smooth path for continuous movement and passing through all waypoints, ensuring that the speed of the robot does not change drastically at the measurement configuration, but also determines the distance between each pair of adjacent waypoints during trajectory planning for the next step of planning.

[0085] S220: Segmented S-shaped trajectory planning between waypoints.

[0086] In order to weaken the chattering problem between waypoints and at waypoints and ensure the first-order differentiability of the path velocity curve, the present invention designs a segmented S-shaped trajectory between every two waypoints with respect to time, and optimizes the time for calculating the trajectory motion under the conditions of known maximum speed, maximum acceleration, maximum jerk and distance constraints.

[0087] Specifically, see Figure 3 The symmetrical 7-segment S-shaped trajectory shown in the figure has a and waypoints The curve changes of jerk, acceleration, velocity and displacement of the trajectory between and their minimum and maximum values ​​are defined as , , and Among them, the first three sections of the S-shaped speed For the acceleration stage, The last three sections are uniform speed sections. It is the deceleration section. Since it is a symmetrical curve, its various time periods have the following relationship:

[0088]

[0089] and, its various periods of time The corresponding movement distance relationship is as follows:

[0090]

[0091] Waypoint and waypoints The moving distance of the trajectory between is:

[0092]

[0093] Among them, due to the above periods of time Not the motion planning cycle of the robot controller If the planning is directly calculated according to the above time period, the rounding error will be introduced, which will affect the accurate judgment of the waypoints and the smoothness of the movement. and waypoints The distance constraint of the trajectory between the two is to appropriately increase the total planning time and adjust the acceleration parameter, acceleration parameter and speed parameter to carry out Time round planning:

[0094]

[0095] S230, Optimization of the waypoint sequence with the least time

[0096] Under the condition that the path connecting any waypoints, the moving distance between any two adjacent waypoints, and the moving time between any two adjacent waypoints are known, the present invention uses a genetic algorithm to select the passing order of the waypoints so that the final moving time is the shortest. Figure 1 , since the travel time between waypoints needs to be recalculated according to the changes in the continuous path, see Figure 1 In the "update traversal sequence of the selected configuration" module, S230 will determine whether to stop iterating S210 and S220 according to whether the overall planning time of the spline path is reduced, that is, after a certain number of iterations, if the overall planning time of the spline path is not reduced, the iteration is stopped and the next step is entered.

[0097] Specific implementation of step S300

[0098] According to the S-curve between the measured configurations The planning values ​​of the time segment can obtain the planning speed curve of the end of the industrial robot as the planning trajectory data. Among them, the maximum speed of the end of the industrial robot is the peak of a series of speed curves, and the minimum speed is the trough between them. And, the end of the industrial robot is tracked and measured by a measuring instrument, and then the speed data of the end of the industrial robot is acquired through coordinate transformation as the actual trajectory data. According to the consistency characteristics of the measured speed curve and the planned speed curve, the duration of the highest point of the speed (i.e., the peak) is relatively long, the speed at the highest point is significantly different from the speed at the lowest point (i.e., the trough), and the duration of the speed curve in the trough is short. The trough characteristics of the speed curve are used to align the measurement curve and the planning curve of the speed, and finally the actual measurement data of the measurement configuration required for calibration is obtained from the continuous motion measurement curve. It should be noted that although there is a certain fluctuation in the consistency of the speed curve due to the influence of the measurement error, coordinate transformation error and servo control error of the instrument, the method of the present invention can ignore the above influence.

[0099] Specific implementation of step S400

[0100] The CMM-based rapid calibration method proposed in the present invention can not only improve the absolute positioning accuracy of industrial robots, but also reduce the data acquisition time of kinematic calibration of industrial robots, thereby improving the calibration efficiency of actual production line robots. The rapid calibration method based on the above measurement of the present invention at least comprises the following steps:

[0101] S410, constructing a kinematic model of the industrial robot to establish an error mapping model;

[0102] S420. Establish an inverse kinematics error model for industrial robots;

[0103] S430, build a measurement data collection platform to collect the calibration geometry set required in the CMM strategy;

[0104] S440, using the calibration positioning data obtained by the CMM strategy to identify kinematic error parameters and compensate for the kinematic parameters of the industrial robot.

[0105] Furthermore, the present invention is applied to a new type of SPM-RM parallel industrial robot. By building a measurement data acquisition platform, the terminal posture data set and verification data set required in the CMM strategy and LMSM strategy are respectively collected, and then the calibration positioning data obtained by the CMM strategy and the LMSM strategy are used to identify the kinematic error parameters, and the kinematic parameters of the SPM-RP parallel robot are compensated. The verification data set is used to verify the effectiveness of the rapid calibration method of the present invention. Compared with the traditional calibration method based on the classic LMSM measurement strategy, the rapid calibration based on the CMM strategy has achieved a 93.13% improvement in calibration efficiency at the expense of only 1.77% of the positioning accuracy compared to the traditional calibration method. Specifically, the present invention uses the rapid calibration method of the SPM-RM parallel robot and its verification, which at least includes the following steps:

[0106] S410, SPM-RP parallel robot structure model construction and kinematic parameter error determination:

[0107] See also Figure 4 The method of the present invention can be applied to Figure 4 The model of the Schonflies-motion parallel manipulator with rotational pitch motion, SPM-RP parallel robot shown in the figure. The SPM-RP parallel robot consists of an end moving platform, two PRPaRR branches ( Figure 4 Shown and ) and two PRPaR branches ( Figure 4 Shown and ), where P, R, and Pa represent active translation joints, passive rotation joints, and passive parallelogram structures, respectively. The topological structures of the PRPaR parts of the four branches are consistent and symmetrical. The joint axis of P intersects with the axis of the upper rotation joint R and the axis directions are perpendicular to each other. The lower rotation joint R is connected to the end moving platform. and The axes of the two upper revolute joints R of the branch chain are parallel. and The axes of the two lower revolute joints R of the branch chain are parallel. Similarly, and The axes of the two upper revolute joints R of the branch chain are parallel. and The axes of the two lower rotation joints R of the branch chain are also parallel. Figure 4 The axes of the four rotating joints R on the right side of the moving platform always remain coplanar regardless of how the SPM-RP parallel robot moves. , and The center points of the lower revolute joints R of the branch chains are collinear.

[0108] The error mapping model is established using the closed-loop vector kinematics model of the SPM-RP parallel robot. Figure 5a and Figure 5b The SPM-RP as a whole and any single branch Structural diagram, moving platform coordinate system In the base coordinate system The rotation matrix in the transformation relationship and translation vectors It can be expressed as:

[0109]

[0110] According to Structural characteristics of the branched chain, equivalent rod length of the parallelogram structure Pa in the PRPaR structure is fixed, the above formula is transformed into:

[0111] .

[0112] The error mapping model can be obtained based on the first-order perturbation principle for any single branch The error equation is:

[0113] ,

[0114] In the formula, i =1,2,3,4, indicating the serial number of the branch chain. Indicates branch chain The center of the lower revolute joint R in the PRPaR structure is relative to The position of the coordinate system, yes The position error vector of . Indicates that SPM-RP is in zero position, branch chain The center of the upper rotation joint R in the PRPaR structure is relative to The position of the coordinate system, yes The position error vector of ; Indicates branch chain The displacement of the translation joint P in the PRPaR structure; It represents the joint displacement deviation, which characterizes the deviation between the measured value and the calculated value of the kinematic error parameter in the joint displacement direction. R Q,iIndicates branch chain Connected to Joint coordinate system of points Relative to The rotation matrix is ​​related to the axis direction of the moving pair; represents the differential rotation matrix, the axis angle parameters are included in ,and By fixed to Joint coordinate system of points Relative to the base coordinate system Around Axis rotation Obtained from angle. Indicates branch chain The parallelogram structure in the PRPaR structure Equivalent rod length, error of equivalent rod length ,and . Represents an intermediate variable , Represents an intermediate variable The error term of . e z express Unit vector in direction In the formula, the superscript represents the nominal terms of kinematic parameters.

[0115] Among them, the kinematic parameter errors include:

[0116] 1) Position error of the kinematic pair (R pair). yes The position error vector of ; yes The position parameter error vector of the zero position, and .

[0117] 2) Axis error of the kinematic pair (P pair). The axis angle parameter is included in ,and By fixed to Joint coordinate system of points Relative to the base coordinate system Around Axis rotation The differential rotation matrix is ​​obtained by for:

[0118] .

[0119] Therefore, the angular error parameter of the axis is expressed as , and according to the antisymmetric matrix transformation theory, the term in the error equation for a single branch is can be rewritten as , the operation factor × represents the antisymmetric operation of the vector - for any 3D vector ,exist

[0120] .

[0121] In the formula, v 1, v 2, v 3 represents 3-dimensional vector The three scalar elements in .

[0122] 3) Length error, that is, the error of the equivalent rod length parameter ,and .

[0123] The above kinematic error parameters can be collectively expressed as the error parameter vector , superscript represents the nominal value of kinematic parameters, which can also be expressed as .

[0124] S420. Establish the inverse kinematics error model of the SPM-RP parallel robot:

[0125] Establish the inverse kinematics parameter error equation, the measured position and the measured rotation matrix containing attitude information ,by and To mark the measured end position of the industrial robot. In the inverse kinematics error model, the end position information of the industrial robot is taken as the input and the joint displacement deviation of the industrial robot is taken as the output. The joint displacement of the joint is:

[0126] ,

[0127] According to the joint encoder measurement The joint displacement of the joint is , so the joint displacement deviation It can be expressed as:

[0128]

[0129] The error mapping model can be obtained, subscript Indicates The measurement configuration sequence is

[0130]

[0131] in,

[0132]

[0133] In the formula, A j No. The error mapping matrix under the measurement configuration.

[0134] S430, build a measurement data collection platform to collect the calibration geometry sets required in the CMM strategy and the LMSM strategy respectively.

[0135] Specifically, referring to the data acquisition platform of the SPM-RP parallel robot shown in Figure 6 (a) and Figure 6 (b), a Leica AT960 laser tracker is used to measure the spherical reflector (SMR) installed on the surface of the moving platform and calculate the terminal pose. and Figure 7(a) is the spatial trajectory of the motion. The blue dots in the figure are the positions between the measurement configurations in the trajectory, the green dots are the positions of the measurement configurations, and the red direction arrows indicate the posture direction under the measurement configuration. Figure 7(b) is the velocity scalar curve along the motion direction under continuous motion. The circular intersection after aligning the velocity planning curve is the actual measurement configuration data required for calibration in the CMM strategy. The measurement configuration data collected after the SPM-RP is stopped at the measurement configuration corresponding to the trough of the planning velocity curve in Figure 7(b) using the LMSM strategy.

[0136] Among them, the measurement efficiency improvement of CMM strategy compared with LMSM strategy is:

[0137]

[0138] in, represents the motion measurement time between each adjacent measurement configuration, and Indicates the waiting time for measuring the configuration. The motion measurement time of the measured configuration The waiting time for measuring the configuration under the LMSM strategy is 10.7s, and the average value of the experimental waiting time is 5s. It is 155.7s, which is the measurement efficiency improvement rate .

[0139] S440. Use the calibration positioning data obtained by the CMM strategy and the LMSM strategy to identify the kinematic error parameters, compensate the kinematic parameters of the SPM-RP parallel robot, and use the verification data set to verify the effectiveness of the fast calibration method.

[0140] First, using the error model in step S420 and the measured configuration data collected based on the CMM strategy and the LMSM strategy in step S430, the following two sets of equations for kinematic parameter identification are obtained, with superscript represents the data collected by the LMSM strategy, and the superscript Indicates the data collected by the CMM strategy.

[0141]

[0142] Then, the same identification method, the extended Kalman filter method, is used for the two sets of identification equations to obtain the estimated values ​​of the kinematic parameter errors under the CMM strategy and the LMSM strategy. and , and update the nominal kinematic parameters with the estimated values ​​of the parameter errors

[0143]

[0144] See also Figure 8 , using 100 independent validation configuration data sets randomly collected in the SPM-RP workspace, respectively verify and The accuracy of the end of its nominal kinematic parameters , and The absolute positioning error in the direction is shown in Figure 9(a). and The absolute positioning error after 3D is shown in Figure 9(b), where the red dotted line indicates the absolute positioning error after 3D CMM strategy identification. , the blue dotted line indicates the identified .

[0145] The present invention obtains the following by statistically analyzing the absolute positioning error data of 100 verification configurations:

[0146] 1) Using the LMSM measurement strategy, the maximum value "Maximum", average value "Mean" and standard deviation "STD" of the terminal positioning error after calibration are improved by 92.03%, 95.76% and 87.78% compared with the accuracy before calibration; the corresponding improvement rates of the terminal attitude accuracy are 92.19%, 97.12% and 79.33% respectively.

[0147] 2) Using the CMM strategy, the maximum value "Maximum", average value "Mean" and standard deviation "STD" of the terminal positioning error after calibration are improved by 90.30%, 93.99%, and 84.44% compared with the accuracy before calibration, and the corresponding terminal posture accuracy improvement rates are 85.97%, 94.76%, and 65.33% respectively. Compared with the LMSM strategy, the average position accuracy loss of the CMM strategy is 1.77%, and the average posture accuracy loss is 2.36%.

[0148] It should be appreciated that the method steps in the embodiments of the present invention can be implemented or implemented by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer readable memory. The method can use standard programming techniques. Each program can be implemented in a high-level process or object-oriented programming language to communicate with a computer system. However, if necessary, the program can be implemented in an assembly or machine language. In any case, the language can be a compiled or interpreted language. In addition, the program can be run on a programmed ASIC for this purpose.

[0149] In addition, the operations of the processes described herein may be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes described herein (or variations and / or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) that is executed collectively on one or more processors, by hardware, or a combination thereof. The computer program includes a plurality of instructions that may be executed by one or more processors.

[0150] Further, the method can be implemented in any type of computing platform that is operably connected to a suitable computer, including but not limited to a personal computer, a minicomputer, a mainframe, a workstation, a network or distributed computing environment, a separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, etc. Various aspects of the present invention can be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, an optical read and / or write storage medium, an RSM, a ROM, etc., so that it can be read by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the process described herein. In addition, the machine-readable code, or portions thereof, can be transmitted via a wired or wireless network. When such media includes instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor, the invention described herein includes these and other different types of non-transitory computer-readable storage media. When programmed according to the methods and techniques of the present invention, the present invention can also include the computer itself.

[0151] The computer program can be applied to input data to perform the functions described herein, thereby converting the input data to generate output data stored in a non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the present invention, the converted data represents physical and tangible objects, including specific visual depictions of physical and tangible objects produced on the display.

[0152] The above is only a preferred embodiment of the present invention. The present invention is not limited to the above implementation. As long as the technical effect of the present invention is achieved by the same means, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of protection of the present invention. Within the scope of protection of the present invention, its technical scheme and / or implementation method may have various modifications and changes.

Claims

1. A rapid calibration method for industrial robots based on continuous motion measurement, characterized in that: The method comprises the following steps: S100, based on the pool of candidate configurations of the industrial robot workspace, selecting an optimal measurement configuration to determine a waypoint through which the industrial robot moves; S200, determining a planning trajectory of a smooth continuous motion with optimal time based on the distribution of waypoints; S300, measuring the actual trajectory of the industrial robot, aligning the planned trajectory with the measured trajectory to determine actual data of the selected measurement configuration; S400, constructing a kinematic model of an industrial robot, and calibrating the industrial robot using a fast calibration method based on a continuous motion measurement strategy, namely, a CMM strategy; Wherein, the step S400 includes: S410, constructing a kinematic model of the industrial robot to establish an error mapping model; S420. Establish an inverse kinematics error model for industrial robots; S430, building a measurement data collection platform to collect the calibration geometry sets required in the continuous motion measurement strategy and the LMSM strategy respectively; S440, respectively using the calibration positioning data obtained by the CMM strategy and the LMSM strategy to identify kinematic error parameters, and compensate the kinematic parameters of the industrial robot; wherein, the step S440 includes: S441, obtaining a set of equations for kinematic parameter identification according to the established inverse kinematics error model and the actual data of the measurement configuration collected based on the CMM strategy; S442. Use the extended Kalman filter method to obtain an estimated value of the kinematic parameter error under the CMM strategy, and update the nominal kinematic parameters with the estimated value of the parameter error.

2. The method according to claim 1, characterized in that The step S100 includes: S110, preparing alternative configurations in the working space of the industrial robot, and initializing a certain number of measurement configurations; S120, using a search method of DETMAX plus simulated annealing to search among candidate configurations to determine the optimal measurement configuration.

3. The method according to claim 2, characterized in that The step S200 includes: S210, connecting waypoints using a spline interpolation method to plan a smooth path passing through the waypoints, and obtaining a movement distance between any two adjacent waypoints; S220, according to the movement distance between any two adjacent waypoints, use the segmented S model to plan the trajectory between every two waypoints, and obtain the time-related movement parameters and movement time between the two adjacent waypoints; S230: Determine the passing order of the waypoints with the shortest moving time based on a genetic algorithm according to the moving time between any adjacent waypoints.

4. The method according to claim 1, characterized in that In the step S300, the velocity measurement curve and the planning curve are aligned according to the trough characteristics of the velocity curve, so as to obtain measurement data of the measurement configuration required for calibration from the measurement curve.

5. The method according to claim 1, characterized in that The industrial robot is a SPM-RP parallel robot. In step S410, Based on the rotation matrix R and translation vector in the transformation relationship between the moving platform coordinate system {rm} and the base coordinate system {rb}, the single branch chain Z of the SPM-RP parallel robot i The error equation is expressed as follows: l i Δl i =(the i -d i R Q,i e z ) T (See i -Δd i R Q,i e z -d i R Q,i ΔR Q,i e z ) See you later. i =RΔs i -Δq i In the formula, i = 1, 2, 3, 4, indicating the sequence number of the branch chain; s i represents branch i P The position of the center of the lower revolute joint R in the RPaR structure relative to the moving platform coordinate system {rm}, Yes i The position error vector of q i Indicates that SPM-RP is in the zero position, and the branch i P The position of the center of the upper revolute joint R in the RPaR structure relative to the base coordinate system {rb}, Yes i The position error vector of d i represents branch i P Translational joints in the RPaR structure P The displacement of i It represents the joint displacement deviation, which characterizes the deviation between the measured value and the calculated value of the kinematic error parameter in the joint displacement direction; R Q,i Indicates that branch i is firmly connected to Q i The joint coordinate system of the point {rq i } is the rotation matrix relative to the base coordinate system {rb}, which is related to the axis direction of the moving pair; ΔR Q,i represents the differential rotation matrix, the axis angle parameters are included in and By being fixed to Q i The joint coordinate system of the point {rq i } Rotate around the z-axis relative to the base coordinate system {rb} obtained by angle; l i represents branch i P The equivalent rod length of the parallelogram structure Pa in the RPaR structure and the error of the equivalent rod length and η i represents the intermediate variable η i =p+Rs i -q i , Δη i represents the intermediate variable η i The error term; e z express z The unit vector e in the direction z =[0,0,1] T ; Where the superscript n represents the nominal term of the kinematic parameters.

6. The method according to claim 5, characterized in that In step S420, the end position information of the industrial robot is used as input, and the joint displacement deviation of the industrial robot is used as output. The error mapping model is expressed as: in, Where, subscript i represents the i-th branch sequence of the SPM-RP parallel robot, subscript j represents the j-th measured configuration sequence, and A j represents the error mapping matrix under the jth measurement configuration, p m and R m represents the measured end position of the industrial robot; the superscript m represents the physical quantity obtained by measurement; [e z ×] represents the unit vector e z =[0,0,1] T Transform into its corresponding antisymmetric matrix; Among them, the joint displacement of the i-th joint calculated according to the measured end posture is expressed as follows: The joint displacement of the i-th joint measured by the joint encoder is Get the joint displacement deviation Δd i It is expressed as follows: Where, the superscript c represents the joint displacement d i It is calculated based on the inverse kinematics function inv() in the above formula. 7 . A computer-readable storage medium having program instructions stored thereon, wherein the program instructions, when executed by a processor, implement the method according to any one of claims 1 to 6 .

8. Industrial robot rapid calibration system based on continuous motion measurement, characterized in that: include: A computer device comprising a computer readable storage medium according to claim 7.

Images