CN121403409A - An automatic coordinate system calibration method and system for ultrasonic testing terminal actuators - Google Patents

Abstract

A coordinate system automatic calibration method and system for an ultrasonic detection terminal executor relates to the technical field of executor calibration, and comprises the steps of replacing a nozzle of the ultrasonic detection terminal executor with a vector target seat, arranging a robot on a base, driving the ultrasonic detection terminal executor to move in a measuring space of a laser tracker by a plurality of different postures by the robot, synchronously acquiring joint angle vectors of the robot and three-dimensional coordinates of two targets under each posture, converting the three-dimensional coordinates of the two targets into a robot flange coordinate system, constructing rough estimation of a pose transformation matrix from the ultrasonic detection terminal executor to the flange coordinate system based on the coordinates of the two targets under the robot flange coordinate system, calculating to obtain accurate estimation of the pose transformation matrix by adopting a data fusion algorithm based on the rough estimation of the pose transformation matrix under all the postures of the robot, and solving the problems of low precision and poor robustness of the traditional calibration method.

Description

Automatic calibration method and system for coordinate system of ultrasonic detection terminal actuator

Technical Field

The invention relates to the technical field of actuator calibration, in particular to an automatic calibration method and system for a coordinate system of an ultrasonic detection terminal actuator.

Background

In the field of robot ultrasonic nondestructive detection, the Tool Center Point (TCP) calibration precision of an ultrasonic terminal actuator is a core premise for determining the reliability and consistency of a detection result, and a single-arm ultrasonic terminal actuator is generally composed of a quick-change disc, an anti-collision device, a water cavity, a nozzle and other complex structures, wherein a TCP coordinate system is defined as the center position of a nozzle outlet, the coordinate axis direction of the TCP coordinate system, namely a X, Y, Z axis, is strictly aligned with a robot terminal flange coordinate system, however, in multiple links such as machining, component assembly, field installation and the like, dimensional tolerance, shape and position errors and assembly gaps are inevitably introduced, so that small but non-negligible deviation exists between the actual physical pose of the TCP coordinate system and the theoretical design pose of the TCP coordinate system relative to a flange. If the deviation cannot be accurately measured and compensated, the ultrasonic beam is directly led to be misdirected, and the positioning accuracy of defect detection and the accuracy of quantitative evaluation are seriously reduced, so that the actuator must be calibrated with high accuracy before being put into use.

Traditional mainstream calibration methods are highly dependent on manual operation and contact measurement tools. Typically, the technician will use calipers, altimeters, dial gauges, or even three-coordinate measuring machines to directly measure the external geometry of the actuator nozzle in an attempt to infer the TCP position from a physical reference. The method has the advantages that a series of systematic defects are exposed in practice, firstly, the tail end of an actuator, especially a nozzle area, is compact in structure and narrow in space, and internal water channels and external installation features are staggered, so that a measuring probe is difficult to approach a real TCP physical reference point (namely a nozzle hole center), an operator is forced to conduct indirect measurement and geometric conversion, and accordingly multiple error accumulation is introduced, secondly, the whole process is strongly dependent on personal experience, skill and even subjective judgment of the operator, random errors caused by manual intervention exist in all links from selection of a measuring reference and control of a tool to reading and recording of data, and repeatability and reproducibility of a calibration result are poor, and requirements of standardized and large-scale production on process consistency cannot be met.

More importantly, the traditional method usually finishes calibration under the condition that the robot is in a single or very few static postures, and the single-point calibration mode has the principle limitation that the influence of different error sources cannot be effectively separated and identified, various system errors such as positioning errors, joint backlash, flexible deformation of connecting rods, errors of flange mounting surfaces and the like of the robot body are mixed in measurement results, and the errors are different in performances under different postures and different loads of the robot, so that the TCP precision is often obviously reduced and the robustness is insufficient when the robot moves to other positions or postures of a working space based on calibration parameters obtained from a few postures.

In addition, the traditional contact type calibration method lacks an effective and independent in-process verification mechanism, the calibration process cannot provide closed-loop verification on the accuracy of the result, an operator can only select 'trust' of the measurement, and once deviation is found in the subsequent detection process, the deviation is difficult to trace back and distinguish whether the deviation is caused by calibration error, robot track error or workpiece positioning error, so that great difficulty is brought to process debugging and problem tracing.

Thus, we propose a calibration method that can improve the calibration accuracy, as well as ensure robustness.

Disclosure of Invention

The invention aims to provide an automatic calibration method and an automatic calibration system for a coordinate system of an ultrasonic detection terminal actuator, which are used for solving the problems of low precision and poor robustness of the traditional calibration method.

The invention is realized by the following technical scheme:

the automatic calibration method for the coordinate system of the ultrasonic detection terminal executor specifically comprises the following steps:

replacing a nozzle of an ultrasonic detection terminal executor with a vector target seat, and installing a robot on the base;

The robot drives the ultrasonic detection terminal executor to move in the measuring space of the laser tracker in a plurality of different postures;

Synchronously acquiring joint angle vectors of the robot under each gesture and three-dimensional coordinates of two targets on a vector target seat measured by a laser tracker under a base coordinate system;

According to the pose transformation matrix of the robot base coordinate system relative to the base coordinate system and the pose of the robot, calculating the pose transformation matrix of the robot flange coordinate system relative to the base coordinate system under each pose;

converting three-dimensional coordinates of the two targets into a robot flange coordinate system by using a pose transformation matrix;

based on two target coordinates in a robot flange coordinate system, constructing rough estimation of a pose transformation matrix of a vector target seat actuator coordinate system relative to the robot flange coordinate system;

Based on rough estimation of the pose transformation matrix under all robot poses, calculating by adopting a data fusion algorithm to obtain accurate estimation of the pose transformation matrix.

Further, the position and posture transformation matrix relative to the base coordinate system according to the robot base coordinate systemCalculating a pose transformation matrix of a robot flange coordinate system relative to a base coordinate system under each poseThe specific calculation formula is as follows:

;

In the formula, Is the joint angle vectorAnd the pose of the flange coordinate system is obtained through positive kinematic calculation.

Further, the three-dimensional coordinates of the two targets are converted into a robot flange coordinate system by using the pose transformation matrix, and a specific calculation formula is as follows:

;

;

In the formula, Is the three-dimensional coordinates of the first target point,Is the three-dimensional coordinates of the second target.

Further, the constructing a rough estimate of the pose transformation matrix of the vector target holder actuator coordinate system relative to the robot flange coordinate system based on the two target coordinates in the robot flange coordinate system specifically includes:

Determining the z-axis direction of a vector target holder actuator coordinate system by two collinear target coordinates;

defining a point after the first target point deviates from the design distance d along the z-axis direction as an origin;

Projecting an x-axis direction vector under the flange coordinate system onto a plane perpendicular to the z-axis direction, and normalizing to determine the x-axis direction of the vector target holder actuator coordinate system;

determining the y-axis direction of a vector target actuator coordinate system according to the right-hand Cartesian coordinate system rule;

Based on the z-axis direction, the origin coordinates, the x-axis direction, and the y-axis direction, a rough estimate of a pose transformation matrix of the vector target holder actuator coordinate system relative to the robot flange coordinate system is constructed.

Further, the data fusion algorithm is a mean method, including:

For a pair of Origin coordinatesAveraging to obtain translation vector;

;

For a pair ofA plurality of rotation matricesPerforming rotation average, and calculating by quaternion average method or lie algebraic average method to obtain average rotation matrix;

Combining the averaged translation vectors with an average rotation matrixThe method is the accurate estimation of the pose transformation matrix:

。

further, the data fusion algorithm is a nonlinear optimization method, and includes:

taking a result obtained by a mean value method as an initial value, constructing a nonlinear least square optimization problem, wherein an objective function of the problem is as follows;

;

In the formula, For the rotation matrix to be optimized,For the translation vector to be optimized,AndRespectively fixing coordinates of the first target point and the second target point in a target seat coordinate system;

solving an objective function through a Levenberg-Marquardt algorithm or a Gaussian-Newton algorithm to obtain the optimal result AndAnd forming accurate estimation of the pose transformation matrix.

A coordinate system auto-calibration system for an ultrasonic testing end effector, comprising:

A robot having known kinematic parameters;

the vector target seat is detachably arranged at the nozzle of the ultrasonic terminal actuator to be calibrated, and is provided with two collinear nozzles with the center distance of Spherical reflection targets of (2);

The laser tracker is used for capturing the three-dimensional coordinates of the spherical reflection target in the measurement space;

A computing and control unit, which is respectively connected with the controller of the robot and the laser tracker in a communication way, and is configured to execute the following operations:

controlling a robot to drive an ultrasonic detection terminal executor to move to a plurality of preset postures, and synchronously collecting pose data and target point coordinate data of the robot;

Performing conversion and calculation of a coordinate system;

According to the coordinate system automatic calibration method for the ultrasonic detection terminal executor, constructing rough estimation of the pose transformation matrix under each pose;

And executing a data fusion algorithm, and calculating to obtain and output the accurate estimation of the final pose transformation matrix.

Furthermore, the computing and control unit also pre-stores a pose transformation matrix of the robot base coordinate system relative to the base coordinate system and the design distance.

Furthermore, the vector target seat is installed in a matched mode through the base of the vector target seat and the center of the nozzle installation position of the ultrasonic terminal actuator, and the fact that the connecting line direction of the two target points is collinear with the central axis of the nozzle is guaranteed.

Further, the calculation and control unit is an independent computer or a data processing module integrated in the robot controller.

The technical scheme of the invention has at least the following advantages and beneficial effects:

The invention discloses a coordinate system automatic calibration method and a system for an ultrasonic detection terminal executor, which radically eliminates errors caused by contact pressure and alignment deviation of a measuring probe by adopting a laser tracker to carry out non-contact optical measurement, and on the other hand, obtains space coordinate data of far-exceeding manual measurement resolution by measuring a standard spherical reflection target point arranged on a vector target seat, and lays a data foundation for high-precision calibration.

The method changes the role of an operator from measurement execution and judgment to flow start and monitor, eliminates random errors caused by personal skills and subjective judgment, and ensures that calibration results of different personnel and different time have high consistency and reproducibility.

In addition, by fusing a plurality of groups of spatially distributed data and adopting a mean value method or a nonlinear optimization algorithm to carry out overall solution, random noise in single measurement can be effectively averaged, the fluctuation influence of systematic errors such as robot positioning errors, joint gaps and the like under different poses can be obviously restrained, so that the finally calibrated TCP parameters can be kept high in precision in the whole robot working space, and the robustness of the calibration result is greatly improved.

In addition, by adopting two data fusion algorithms of a mean value method and a nonlinear optimization method, wherein the mean value method is efficient in calculation and can quickly obtain stable estimation, the nonlinear optimization method takes the result of the mean value method as an initial value, and the optimal transformation is integrally solved by minimizing the re-projection errors of all measurement points, so that the data potential is further squeezed, a calibration result which has higher theoretical precision and is insensitive to noise can be obtained, and the method is suitable for application scenes with higher precision.

It should be noted that the vector target seat designed by the invention is detachably arranged at the position of the actuator nozzle, and two collinear targets on the vector target seat accurately simulate the axis of the nozzle, so that the same target seat can be adapted to a plurality of actuators with the same type due to the design of 'plug and play'; meanwhile, the target seat has definite installation requirement (ensuring collineation), and reduces the difficulty of installation and adjustment.

Drawings

FIG. 1 is a schematic flow chart of an automatic calibration method for an ultrasonic detection terminal actuator according to the present invention;

FIG. 2 is a schematic diagram of an automatic calibration system for an ultrasonic testing end effector in accordance with the present invention;

fig. 3 is a schematic structural diagram of an electronic device according to the present invention.

Reference numeral 1, a base, 2, a robot, 3, a vector target seat, 4, a target spot, 5, and a laser tracker.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.

Example 1

1-2, A coordinate system automatic calibration method for an ultrasonic detection terminal actuator specifically includes:

The nozzle of the ultrasonic detection terminal executor is replaced by a vector target seat 3, and a robot 2 is arranged on a base 1;

The vector target seat 3 comprises a base, two spherical reflection targets 4 are fixedly arranged on the base, a connecting column with the same diameter as that of the nozzle is arranged at the head end of the base, and the connecting column is connected with an ultrasonic detection terminal executor in a matching way; in addition, the spherical reflection targets 4 comprise spherical shells, the tops of the spherical shells are provided with observation windows, and the positions of spherical centers in the spherical shells are fixedly provided with reflectors, and the attention is paid to the fact that the spherical center connecting line of the two spherical reflection targets 4 is strictly coincident with the theoretical central axis of the nozzle in mechanical design;

So that the laser tracker 5 can measure the three-dimensional coordinates of the center of the target spot 4 with extremely high precision by receiving the light rays emitted by the reflecting mirror in the target spot 4, and can reversely deduce the defined straight line direction by calculation and according to the known design distance by measuring the three-dimensional coordinates of the centers of the two target spots 4 Namely, the mechanical offset distance from the target point 4, which is close to the connecting column, on the vector target seat 3 to the theoretical outlet of the nozzle, and precisely calculating the position of the theoretical outlet of the nozzle on the straight line;

in addition, the base 1 and the base of the robot 2 are rigidly connected, and the pose transformation matrix of the base coordinate system of the robot relative to the base coordinate system can be accurately calibrated by off-line and used as a known constant, in particular, the pose transformation matrix The method comprises the steps of after a calibration tool is grabbed by a robot 2, fixedly placing a target ball at the tail end of the tool, driving the tool to move in multiple postures by controlling the robot 2, synchronously recording joint data of the robot 2 under each posture, measuring the position of the target ball on the tool relative to a coordinate system of a base 1 by using a laser tracker 5, finally calibrating kinematic parameters of the robot 2, conversion relation between the coordinate system of the tool and the coordinate system of a flange at the tail end of the robot 2 and pose relation between the base coordinate system of the robot 2 and the coordinate system of the base 1, namely pose conversion matrix, by a joint optimization method based on the acquired data。

In conclusion, the vector target seat 3 provides standard characteristic points which can be measured by the limit precision of the laser tracker 5, the base 1 scheme ensures that the pose data of the robot 2 and the measured data of the tracker can be associated with each other with high precision under a unified reference system, a high-quality data input source is provided for the whole algorithm by combining the pose data and the measured data of the tracker, the target seat is a detachable calibration tool, the base 1 is a preset platform, the whole calibration process can be controlled in a programmable manner, the robot 2 can automatically drive the target seat to move to a plurality of preset poses, and the system synchronously acquires data, so that the radical transition from manual operation to program automatic execution is realized;

And furthermore, the measurement range brought by the base 1 is expanded, so that the system can acquire robot 2 posture data with wider and more diverse spatial distribution, fusion calculation is carried out by utilizing the data, random errors can be more fully averaged and systematic errors can be compensated, and the final calibration result is independent of a specific position, thereby keeping stable and reliable in the whole working space.

The robot 2 drives the ultrasonic detection terminal executor to move in the measuring space of the laser tracker 5 in a plurality of different postures;

because the positioning error, the flexible deformation of the connecting rod and the like of the robot 2 can change along with the gesture and are not fixed deviation, after the multi-gesture data fusion of the method, the changed errors tend to be counteracted with each other in statistics instead of being accumulated in the calibration result, so that the calibrated TCP parameters are more accurate and more stable in the whole working space of the robot 2 and are not accurate only in the calibration gesture;

And after nonlinear optimization, the re-projection error of the measurement point under each gesture can be analyzed, and if the residual error of a certain gesture is abnormal, the robot 2 under the gesture can be prompted to move abnormally, the measurement of the tracker is blocked or rough difference occurs, so that the basis for data quality inspection and process diagnosis is provided.

Synchronously acquiring the joint angle vector of the robot 2 under each gesture and the three-dimensional coordinates of two targets 4 on the vector target seat 3 measured by the laser tracker 5 under a base coordinate system;

Wherein the joint angle vector of the robot 2 Is the minimum complete input set of the robot 2 kinematic model, compared with the method of directly reading the flange pose calculated by the robot 2 controller, the joint angle is more original and reliable bottom data, and the pose is calculated by positive kinematicsConsistency of the kinematic model can be ensured;

The first target is a target 4 which is close to a connecting column on a base, the second target is another target 4, the coordinates of the two targets 4 which are originally output by the laser tracker 5 are under the coordinate system of the laser tracker, the measured value is converted to a base coordinate system in real time through the conversion relation of the laser tracker coordinate system and the base coordinate system which are obtained through the pre-calibration, the coordinate system conversion which is frequently needed in the follow-up calculation can be finished in advance through the operation, under the base coordinate system, the relation between all data and the robot base coordinate system is described by a known constant, the calculation chain is clearer, and if the world coordinate system of the laser tracker 5 has tiny drift due to thermal deformation or slight disturbance in the calibration process, but as long as the relative relation between the laser tracker 5 and the base 1 is accurate through the recalibration of the fixed target 4, all the measured values are stable under the base coordinate system.

Furthermore, synchronous acquisition ensures the inherent consistency of each set of data, and any "blooming" phenomenon due to dyssynchrony is eliminated, so that subsequent computation based on rigid body transformations has a strict mathematical basis.

Pose transformation matrix relative to base coordinate system according to robot base coordinate systemCalculating a pose transformation matrix of a robot flange coordinate system relative to a base coordinate system under each poseThe specific calculation formula is as follows:

;

In the formula, Is the joint angle vectorThe position and the posture of the flange coordinate system are obtained through positive kinematic calculation;

For reliably correlating high-precision external measurements with a model of imperfections inside the robot 2, wherein the pose transformation matrix of the robot base coordinate system relative to the base coordinate system The method is a constant transformation matrix obtained through off-line precise calibration, accurately describes the rigid connection relation between two physical entities of the robot 2 (Base) and the Base 1 (world), and because the calibration process can be separated from the production beat and is repeatedly measured and optimized by using a high-precision means, the precision and the reliability of the matrix are far higher than any pose data reported by the robot 2 on line in real time;

Is the real-time output of the positive kinematic model of the robot 2, which is based on the current joint angle The theoretical flange pose is calculated, and the calculation is based on design parameters (DH parameters) of the robot 2, but is influenced by all internal errors such as positioning errors, gear clearances, connecting rod flexibility and the like of the robot 2, so that the relation of the flange to the base is described to have certain uncertainty;

And the step breaks down the complex "robot 2-tracker" system calibration problem into two simpler, more independent sub-problems.

Converting the three-dimensional coordinates of the two targets 4 into a robot flange coordinate system by using a pose transformation matrix, wherein a specific calculation formula is as follows:

;

;

In the formula, Is the three-dimensional coordinates of the first target point,Three-dimensional coordinates of the second target point;

Based on two target point 4 coordinates in a robot flange coordinate system, the rough estimation of a pose transformation matrix of an actuator coordinate system of a vector target seat 3 relative to the robot flange coordinate system is mainly used for unambiguously defining a coordinate system according to the general rules of three-dimensional space rigid body kinematics and a right-hand Cartesian coordinate system by measuring data of a target seat with two collinear target points 4, and specifically comprises the following steps:

The z-axis direction of the vector target holder 3 actuator coordinate system is determined by two collinear target 4 coordinates, and the calculation formula is as follows:

;

Along the z-axis direction, defining a point after the first target point is offset by a design distance as an origin, wherein a calculation formula is as follows:

;

vector of x-axis direction under flange coordinate system Projecting the vector target seat 3 to a plane perpendicular to the z-axis direction, carrying out normalization, and determining the x-axis direction of an actuator coordinate system of the vector target seat 3, wherein the calculation formula is as follows:

;

The y-axis direction of the vector target holder 3 actuator coordinate system is determined according to the right-hand Cartesian coordinate system rule, and the calculation formula is as follows:

based on the z-axis direction, the origin coordinate, the x-axis direction and the y-axis direction, the rough estimation of the pose transformation matrix of the vector target holder 3 actuator coordinate system relative to the robot flange coordinate system is constructed, and the calculation formula is as follows:

;

by means of the model, imperfect and discrete external measurement data can be related to an ideal and to-be-solved tool coordinate system parameter, and the implementation effect is that the model successfully decomposes a complex 6-degree-of-freedom space calibration problem into a series of executable and verifiable deterministic calculation steps, and produces high-quality and structured intermediate results, so that a solid foundation is laid for finally realizing high-precision and high-robustness optimal estimation.

Based on rough estimation of the pose transformation matrix under all robot poses, calculating to obtain accurate estimation of the pose transformation matrix by adopting a data fusion algorithm;

Specifically, the data fusion algorithm is a mean method, including:

For a pair of Origin coordinatesTaking an average value;

;

For a pair of A plurality of rotation matricesPerforming rotation average, and calculating by adopting a quaternion average method or a lie algebra average method because the rotation matrix belongs to a special orthogonal group SO (3) and cannot be directly subjected to arithmetic average in Euclidean space;

Wherein the quaternion averaging method converts each rotation matrix into unit quaternions The average quaternion is calculated on the unit quaternion manifold by adopting a spherical average (SPHERICAL AVERAGE) or weighted average methodReconverted back to the average rotation matrix;

And lie algebra averaging by converting each rotation matrix to its lie algebra through logarithmic mappingCalculating a number average in a tangent spaceThrough exponential mappingObtaining an average rotation matrix。

The above method can ensure average rotation matrixSatisfying orthogonality constraints of a rotation matrixWhereinIs a unitary matrix, and。

Combining the averaged translation vectors with a rotation matrixThe method is the accurate estimation of the pose transformation matrix:

。

in addition, the data fusion algorithm can also be a nonlinear optimization method, and the optimization is performed on the basis of the result of the mean method, and specifically comprises the following steps:

taking a result obtained by a mean value method as an initial value, constructing a nonlinear least square optimization problem, wherein an objective function of the problem is as follows;

;

In the formula, For the rotation matrix to be optimized,For the translation vector to be optimized,AndRespectively fixing coordinates of the first target point and the second target point in a target seat coordinate system;

solving an objective function through a Levenberg-Marquardt algorithm or a Gaussian-Newton algorithm to obtain the optimal result AndThe precise estimation of the optimized pose transformation matrix is formed, namely the pose of the vector target seat coordinate system obtained by calibration, and the origin of the pose is the originAnd the z-axis direction of the vector target seat (3) actuator coordinate system is the position of the ultrasonic end effector outlet TCP and the main direction of the ultrasonic beam.

Example 2

A coordinate system auto-calibration system for an ultrasonic testing end effector, comprising:

a robot 2 having known kinematic parameters;

A vector target seat 3 detachably mounted at the nozzle of the ultrasonic end effector to be calibrated, on which two collinear and central distances are arranged Spherical reflection target 4 of (a);

the laser tracker 5 is used for capturing the three-dimensional coordinates of the spherical reflection target 4 in a measurement space;

a calculation and control unit, respectively in communication with the controller of the robot 2 and the laser tracker 5, configured to perform the following operations:

controlling the robot 2 to drive the ultrasonic detection terminal executor to move to a plurality of preset postures, and synchronously collecting pose data of the robot 2 and coordinate data of the target point 4;

Performing conversion and calculation of a coordinate system;

According to the coordinate system automatic calibration method for the ultrasonic detection terminal executor, constructing rough estimation of the pose transformation matrix under each pose;

And executing a data fusion algorithm, and calculating to obtain and output the accurate estimation of the final pose transformation matrix.

In addition, the calculation and control unit also pre-stores a pose transformation matrix of the robot base coordinate system relative to the base coordinate system and the design distance.

In addition, the vector target seat 3 is installed through the matching of the base of the vector target seat and the center of the nozzle installation position of the ultrasonic end effector, so that the direction of the connecting line of the two targets 4 is ensured to be collinear with the central axis of the nozzle.

The calculation and control unit is an independent computer or a data processing module integrated in the controller of the robot 2, as required.

Example 3

An electronic device as shown in fig. 3, comprising:

A processor, a memory, a communication interface;

The memory is used for storing executable instructions of the processor;

wherein the processor is configured to perform the above-described coordinate system auto-calibration method for an ultrasonic testing end effector via execution of the executable instructions.

A readable storage medium having stored thereon a computer program which when executed by a processor implements the above-described coordinate system auto-calibration method for an ultrasonic detection terminal actuator.

The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. The automatic calibration method for the coordinate system of the ultrasonic detection terminal executor is characterized by comprising the following steps of:

The method comprises the steps that a nozzle of an ultrasonic detection terminal executor is replaced by a vector target seat (3), and a robot (2) is arranged on a base (1);

an ultrasonic detection terminal executor is driven by a robot (2) to move in a measuring space of a laser tracker (5) in a plurality of different postures;

synchronously acquiring joint angle vectors of the robot (2) under each gesture and three-dimensional coordinates of two targets (4) on the vector target seat (3) measured by the laser tracker (5) under a base coordinate system;

According to the pose transformation matrix of the robot base coordinate system relative to the base coordinate system and the pose of the robot, calculating the pose transformation matrix of the robot flange coordinate system relative to the base coordinate system under each pose;

converting three-dimensional coordinates of the two targets (4) into a robot flange coordinate system by using a pose transformation matrix;

based on the coordinates of two targets (4) under the robot flange coordinate system, obtaining rough estimation of a pose transformation matrix of the vector target seat (3) actuator coordinate system relative to the robot flange coordinate system;

Based on rough estimation of the pose transformation matrix under all robot poses, calculating by adopting a data fusion algorithm to obtain accurate estimation of the pose transformation matrix.

2. The method for automatically calibrating a coordinate system of an ultrasonic inspection end effector according to claim 1, wherein the matrix of the pose transformation of the robot base coordinate system with respect to the base coordinate systemCalculating a pose transformation matrix of a robot flange coordinate system relative to a base coordinate system under each poseThe specific calculation formula is as follows:

;

In the formula, Is the joint angle vectorAnd the pose of the flange coordinate system is obtained through positive kinematic calculation.

3. The method for automatically calibrating the coordinate system of the ultrasonic detection terminal executor according to claim 2, wherein the three-dimensional coordinates of the two targets (4) are converted into the robot flange coordinate system by using a pose transformation matrix, and the specific calculation formula is as follows:

;

;

In the formula, Is the three-dimensional coordinates of the first target point,Is the three-dimensional coordinates of the second target.

4. The method for automatically calibrating the coordinate system of the ultrasonic detection terminal executor according to claim 1, wherein the constructing the rough estimation of the pose transformation matrix of the vector target seat (3) executor coordinate system relative to the robot flange coordinate system based on the coordinates of two targets (4) under the robot flange coordinate system specifically comprises the following steps:

determining the z-axis direction of an actuator coordinate system of the vector target seat (3) by two collinear target point (4) coordinates;

Defining a first target offset design distance along the z-axis direction The point behind is the origin;

Projecting an x-axis direction vector under a flange coordinate system onto a plane perpendicular to the z-axis direction, and normalizing to determine the x-axis direction of an actuator coordinate system of the vector target seat (3);

determining the y-axis direction of an actuator coordinate system of the vector target seat (3) according to the right-hand Cartesian coordinate system rule;

Based on the z-axis direction, the origin coordinates, the x-axis direction and the y-axis direction, a rough estimate of the pose transformation matrix of the vector target (3) actuator coordinate system relative to the robot flange coordinate system is constructed.

5. The method for automatically calibrating a coordinate system of an ultrasonic detection terminal executor according to claim 1, wherein the data fusion algorithm is a mean method, and comprises the following steps:

For a pair of Origin coordinatesAveraging to obtain translation vector;

;

For a pair ofA plurality of rotation matricesPerforming rotation average, and calculating by quaternion average method or lie algebraic average method to obtain average rotation matrix;

Combining the averaged translation vectors with an average rotation matrixThe method is the accurate estimation of the pose transformation matrix:

。

6. The method for automatically calibrating a coordinate system of an ultrasonic detection terminal executor according to claim 5, wherein the data fusion algorithm further calculates an accurate estimate of an optimized pose transformation matrix by a nonlinear optimization method based on a mean method, and the method comprises the following steps:

taking a result obtained by a mean value method as an initial value, constructing a nonlinear least square optimization problem, wherein an objective function of the problem is as follows;

;

In the formula, For the rotation matrix to be optimized,For the translation vector to be optimized,AndRespectively fixing coordinates of the first target point and the second target point in a target seat coordinate system;

solving an objective function through a Levenberg-Marquardt algorithm or a Gaussian-Newton algorithm to obtain the optimal result AndAn accurate estimate of the optimized pose transformation matrix is constructed.

7. A coordinate system automatic calibration system for implementing the coordinate system automatic calibration method for an ultrasonic testing end effector according to any one of claims 1 to 6, characterized by comprising:

A robot (2) having known kinematic parameters;

The vector target seat (3) is detachably arranged at the nozzle of the ultrasonic end effector to be calibrated, and is provided with two collinear nozzles with the center distance of Spherical reflection target (4) of (a);

the laser tracker (5) is used for capturing the three-dimensional coordinates of the spherical reflection target point (4) in a measurement space;

A calculation and control unit, in communication with the controller of the robot (2) and the laser tracker (5), respectively, configured to perform the following operations:

Controlling the robot (2) to drive the ultrasonic detection terminal executor to move to a plurality of preset postures, and synchronously collecting pose data of the robot (2) and coordinate data of the target point (4);

Performing conversion and calculation of a coordinate system;

According to the coordinate system automatic calibration method for the ultrasonic detection terminal executor, constructing rough estimation of the pose transformation matrix under each pose;

And executing a data fusion algorithm, and calculating to obtain and output the accurate estimation of the final pose transformation matrix.

8. The system for automatic calibration of a coordinate system according to claim 7, wherein the calculation and control unit further pre-stores a pose transformation matrix of the robot base coordinate system with respect to the base coordinate system, and the design distance.

9. The automatic calibration system for a coordinate system according to claim 7, wherein the vector target seat (3) is mounted by matching the base of the vector target seat with the center of the nozzle mounting position of the ultrasonic end effector, so that the connecting line direction of the two targets (4) is ensured to be collinear with the central axis of the nozzle.

10. The system for automatic calibration of a coordinate system according to claim 7, wherein the calculation and control unit is a stand-alone computer or a data processing module integrated in the controller of the robot (2).