Optimization method for multi-station laser tracker network measurement based on instrument error model
By introducing an instrument geometric error model and multi-station common point constraints, and employing a bounded nonlinear least squares optimization method, the problem of pose relationship coupling in multi-station laser tracker network measurement was solved, enabling on-site correction of common measurement point coordinates and key error parameters, thereby improving the stability and efficiency of the measurement.
Patent Information
- Application Number
- CN202611304515.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-26
- Publication Date
- 2026-11-17
AI Technical Summary
Existing multi-station laser tracker network measurement methods require the pre-establishment of pose relationships between each station. Furthermore, there is coupling between station attitude parameters, common measurement point coordinates, and instrument error parameters, which affects the stability and solution efficiency of the joint solution model and makes it impossible to achieve on-site correction of key error parameters.
A forward measurement model for instrument geometric error is introduced, and the coordinates of common measuring points and instrument geometric error parameters are used as optimization variables. Through the consistency constraint of multiple stations' common points, a bounded nonlinear least squares optimization method is adopted to obtain the optimized coordinates of common measuring points and active error parameters.
Under the same optimization framework, the coordinates of common measuring points and key error parameters of the laser tracker were corrected on-site, reducing the coupling between station attitude parameters and instrument error parameters, and improving the stability and efficiency of multi-station network measurement.
Smart Images

Figure CN123115577A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of precision measurement and laser tracker data processing technology, specifically to a method for optimizing multi-station laser tracker network measurement based on an instrument error model. Background Technology
[0002] Laser trackers typically acquire the spatial coordinates of a target point through distance measurements, azimuth angles, and elevation angles. In large-scale component measurement and industrial assembly measurement tasks, a single laser tracker may be limited by the measurement range, obstruction relationships, measurement viewing angle, and on-site deployment conditions. Multiple laser tracker stations observing the same set of common measurement points can expand the measurement coverage and improve measurement geometry.
[0003] Existing multi-station laser tracker network measurement methods typically require pre-establishing the pose relationships between each station, or using the rotation and translation parameters of each station relative to a unified coordinate system as explicit unknowns for joint solution during the network solution process. As the number of stations increases, the number of pose unknowns also increases. Coupling exists between station attitude parameters, common measurement point coordinates, and instrument internal error parameters, affecting the stability and solution efficiency of the joint solution model.
[0004] Laser trackers also suffer from geometric errors such as ranging offset, ranging scaling factor, horizontal axis offset, beam offset, horizontal axis angle error, beam axis angle error, azimuth scale error, and elevation scale error. For example, initial values for the instrument's geometric error parameters can be obtained using a fixed target network method. However, in actual use, changes in operating temperature, transportation vibration, installation conditions, and the site environment can cause some geometric error parameters to shift. Fixing all pre-calibrated parameters for multi-station network measurements may cause field parameter drift to be transferred to the coordinate calculation results of common measuring points.
[0005] In view of the above, this application is hereby submitted. Summary of the Invention
[0006] The technical problem this invention aims to solve is that existing multi-station laser tracker network measurement methods typically require pre-establishing the pose relationships between stations, or using the rotation and translation parameters of each station relative to a unified coordinate system as explicit unknowns for joint solution during the network solution process. As the number of stations increases, the number of pose unknowns also increases. Furthermore, the coupling between station attitude parameters, common measurement point coordinates, and instrument error parameters affects the stability and solution efficiency of the joint solution model, and prevents the implementation of corrections. This invention provides a multi-station laser tracker network measurement optimization method based on an instrument error model. It introduces a forward measurement model of instrument geometric error into the multi-station laser tracker network measurement process, and uses common measurement point coordinates and instrument geometric error parameters as optimization variables. This method obtains the common measurement point coordinates through multi-station common point consistency constraints and obtains the on-site correction results of the instrument geometric error parameters, i.e., the optimized common measurement point coordinates and active error parameters. This invention obtains the correction results of common measurement point coordinates and key laser tracker error parameters within the same optimization framework, thereby solving the above problems.
[0007] This invention is achieved through the following technical solution:
[0008] In a first aspect, the present invention provides a method for optimizing networked measurement of multi-station laser trackers based on an instrument error model, the method comprising:
[0009] Acquire raw observation data from multiple stations and obtain initial values for the instrument geometric error parameters of the laser tracker;
[0010] Substitute the initial values of the instrument geometric error parameters into the forward measurement model containing the instrument geometric error parameters, and combine them with the original observation data from multiple stations to generate the initial values of the coordinates of the common measuring points.
[0011] Select the geometric error parameters that have a significant impact on the multi-station network measurement results and can be corrected by the constraints of the common points of the multi-stations from the instrument geometric error parameters as the active error parameters;
[0012] The coordinates of the common measuring points and the activity error parameters are jointly constructed as joint optimization variables, and the initial values of the joint optimization variables are composed of the initial values of the coordinates of the common measuring points and the initial values of the activity error parameters.
[0013] The set of coordinates of each station is calculated based on the forward measurement model. The set of coordinates of each station is then registered with the current common set of coordinates of the stations without scale. The multi-station consistency residual is constructed based on the difference between the registered coordinates of the stations and the current common coordinates of the stations.
[0014] Based on the multi-station consistency residuals, the joint optimization variables are optimized using bounded nonlinear least squares to obtain the optimized coordinates of the common measurement points and the optimized activity error parameters.
[0015] Furthermore, the original observation data from multiple stations consist of the distance, azimuth, and elevation angles measured by four or more laser tracker stations for the same set of common measuring points.
[0016] Furthermore, the forward measurement model adopts the NPL geometric error model;
[0017] The forward measurement model uses ranging offset, ranging scaling factor, horizontal axis offset, beam offset, horizontal axis angle error, beam axis angle error, azimuth scale error and elevation scale error to correct the laser tracker observations. Based on the corrected ranging value, corrected azimuth angle, corrected elevation angle, beam origin and beam direction, the coordinates of the measurement point are calculated, and the initial value of the common measurement point coordinates is determined based on the measurement point coordinates.
[0018] Furthermore, the activity error parameters include ranging offset, horizontal axis offset relative to vertical axis, y-axis offset of beam relative to origin, z-axis offset of beam relative to origin, angular error of beam axis in xy plane, and pitch angle zero offset.
[0019] Furthermore, the expression for the joint optimization variables is:
[0020] ;
[0021] in, Denotes the joint optimization variables. Represents the coordinate matrix of common measurement points. This represents the vector obtained by expanding the coordinate matrix of common measurement points column-wise or row-wise. This represents the activity error parameters for each station other than the reference station. This indicates the number of laser tracker stations participating in the network measurement.
[0022] Furthermore, the steps for constructing multi-station consistent residuals are as follows:
[0023] Based on the coordinates of the measuring points and the number of common measuring points calculated by the forward measurement model, the first centroid of the output point set of the non-reference station model is calculated.
[0024] Based on the coordinates of common measurement points in the current joint optimization variables, calculate the second centroid of the current set of common measurement point coordinates;
[0025] Based on the first and second centroids, construct the covariance matrix between the output point set of the non-reference station model and the current common measurement point coordinate set;
[0026] Singular value decomposition is performed on the covariance matrix to obtain the singular value decomposition results;
[0027] Based on the singular value decomposition results, the rigid body registration parameters are calculated, including the rigid body registration rotation matrix and translation vector.
[0028] Calculate the coordinates of the measuring points after registration at the non-reference station based on the rigid body registration parameters;
[0029] Based on the registered coordinates of the measuring points and the current coordinates of the common measuring points, a multi-station consistency residual is constructed.
[0030] Furthermore, the formula for calculating the multi-station consistency residual is as follows:
[0031] ;
[0032] ;
[0033] In the formula, Indicates the first The first station The consistency residuals of multiple stations corresponding to each measurement point; Represents the first among the current joint optimization variables. Coordinates of one common measuring point; Indicates the first After registration of the first non-reference station Coordinates of each measuring point; Indicates the first Rigid body registration rotation matrices corresponding to each non-reference station. Indicates the first Rigid body registration translation vectors corresponding to each non-reference station; Indicates the first The first station was calculated using the forward measurement model. Coordinates of the measuring points.
[0034] Furthermore, based on the multi-station consistency residuals, the joint optimization variables are optimized using bounded nonlinear least squares to obtain the optimized coordinates of the common measurement points and the optimized activity error parameters, including:
[0035] Based on the multi-station consistency residuals, a first-stage unweighted bounded nonlinear least squares optimization is performed on the joint optimization variables to obtain the first-stage optimization results and the first-stage residuals; the least squares optimization model is as follows: ,in, Represents the optimized joint optimization variables. Denotes the joint optimization variables to be optimized. This indicates the number of laser tracker stations participating in the network measurement. Indicates the number of common measuring points. Indicates the joint variables The determined first The first station Multi-station consistency residuals at individual measurement points Represents the Euclidean norm;
[0036] Calculate the Tukey robust weights based on the residuals from the first stage;
[0037] The optimization results of the first stage are used as the initial values for the second stage optimization. Based on the Tukey robust weight, the multi-station consistency residuals are weighted and the second stage bounded nonlinear least squares optimization is performed to obtain the coordinates of the common measuring points and the activity error parameters after robust weighted optimization.
[0038] The optimized coordinates of the common measuring points are used as the measurement results of the multi-station laser tracker network, and the optimized active error parameters are used as the on-site correction results of the key geometric error parameters of the laser tracker.
[0039] Furthermore, the initial values of the instrument's geometric error parameters are obtained through pre-calibration; the pre-calibration method includes at least one of the following: fixed target network method, offline calibration method, on-site coarse calibration method, historical parameter recall method, or manufacturer calibration parameter recall method.
[0040] Furthermore, the initial values of the instrument geometric error parameters of the laser tracker are obtained using the fixed target network method, including:
[0041] Multiple fixed target points are set up in the measurement space, and the fixed target points maintain their spatial position during the coarse calibration measurement process.
[0042] The laser tracker is sequentially deployed at a certain coarse calibration station. The fixed target point is repeatedly measured at each coarse calibration station to obtain the range, azimuth and elevation angles of the fixed target point.
[0043] Using the fixed target point coordinates, coarse calibration station configuration parameters, and instrument geometric error parameters as the parameters to be estimated, a fixed target network observation equation is established to minimize the difference between the model-predicted observations and the actual observations.
[0044] The initial values of the instrument geometric error parameters are obtained by solving the observation equation of the fixed target network using nonlinear least squares.
[0045] The initial values of the instrument geometric error parameters will be used as the initial instrument parameters for subsequent multi-station network joint optimization.
[0046] Secondly, this invention provides a multi-station laser tracker network measurement optimization system based on an instrument error model, the system comprising:
[0047] The acquisition unit is used to acquire raw observation data from multiple stations.
[0048] The calibration unit is used to obtain the initial values of the instrument geometric error parameters of the laser tracker;
[0049] The forward measurement model calculation unit is used to substitute the initial values of the instrument geometric error parameters into the forward measurement model containing the instrument geometric error parameters, and combine the original observation data from multiple stations to generate the initial values of the coordinates of common measuring points.
[0050] The active error parameter selection unit is used to select geometric error parameters from the instrument geometric error parameters that have a significant impact on the multi-station network measurement results and can be corrected by the constraints of the multi-station common point as active error parameters;
[0051] The joint optimization variable construction unit is used to construct joint optimization variables by combining the coordinates of common measuring points and the activity error parameters, and the initial value of the joint optimization variable is composed of the initial value of the coordinates of common measuring points and the initial value of the activity error parameters.
[0052] The multi-station consistency residual construction unit is used to calculate the set of measurement point coordinates of each station based on the forward measurement model, perform scale-free rigid body registration between the set of measurement point coordinates and the current common measurement point coordinate set, and construct the multi-station consistency residual based on the difference between the registered measurement point coordinates and the current common measurement point coordinates.
[0053] The optimization unit is used to optimize the joint optimization variables based on the multi-station consistency residuals using bounded nonlinear least squares, to obtain the optimized coordinates of the common measurement points and the optimized activity error parameters.
[0054] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0055] This invention presents an optimization method for multi-station laser tracker network measurement based on an instrument error model. It introduces a forward measurement model of instrument geometric error into the multi-station laser tracker network measurement process, and uses the coordinates of common measurement points and instrument geometric error parameters as optimization variables. This method obtains the coordinates of common measurement points through multi-station common point consistency constraints and obtains the on-site correction results of the instrument geometric error parameters, i.e., the optimized coordinates of common measurement points and active error parameters. This invention obtains the correction results of common measurement point coordinates and key laser tracker error parameters within the same optimization framework.
[0056] (1) This invention introduces the positive measurement model of instrument geometric error into the multi-station laser tracker network measurement process, and uses the coordinates of common measuring points and key instrument error parameters as optimization variables, which is beneficial to complete the on-site correction of key geometric error parameters of laser tracker while obtaining multi-station network measurement results.
[0057] (2) This invention does not use the external parameters of multi-station pose as explicit joint optimization variables. Instead, it handles the coordinate system difference between the coordinates of non-reference station measurement points and the coordinates of common measurement points through rigid body registration during the residual construction process. This helps to reduce the unknowns in explicit optimization and reduce the coupling between station pose parameters and instrument error parameters.
[0058] (3) The present invention obtains the initial value of the instrument geometric error parameter based on the fixed target network method, and uses the initial value to generate the initial value of the common measuring point coordinate through the forward measurement model. This is beneficial to keep the initial value generation process consistent with the subsequent joint optimization model and improve the rationality of the initial state of nonlinear optimization.
[0059] (4) The present invention selects parameters that have a significant impact on the measurement results of multi-station network and can be corrected by the constraints of the common points of multiple stations as activity error parameters, which is beneficial to compensate for the shift of key error parameters caused by changes in field conditions on the basis of controlling the optimization scale.
[0060] (5) The present invention also adopts a two-stage robust weighted optimization. The weight is calculated based on the residual of the first stage and the second stage weighted optimization is performed. This is beneficial to reduce the impact of abnormal residuals on the joint solution results of common measuring point coordinates and activity error parameters, and improve the adaptability of the multi-station joint optimization process to local abnormal observations. Attached Figure Description
[0061] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:
[0062] Figure 1 This is a flowchart of the multi-station laser tracker network measurement optimization method based on the instrument error model of the present invention;
[0063] Figure 2 This is a block diagram of the multi-station laser tracker network measurement optimization system based on the instrument error model of the present invention. Detailed Implementation
[0064] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.
[0065] To reduce the reliance on pre-established pose networking in multi-station laser tracker network measurements, decrease the number of optimization variables resulting from treating multi-station pose extrinsic parameters as explicit unknowns, reduce the coupling between station attitude parameters, common measurement point coordinates, and instrument error parameters, and correct the offset of key geometric error parameters of the laser tracker during field use, this invention aims to obtain the corrected results of common measurement point coordinates and key laser tracker error parameters within the same optimization framework, thereby reducing the impact of abnormal observation residuals on the multi-station joint optimization results. This invention requires a multi-station laser tracker network measurement optimization method that, without pre-establishing multi-station pose networking or treating multi-station pose extrinsic parameters as explicit joint optimization variables, utilizes an instrument geometric error model and multi-station common measurement point constraints to jointly optimize the coordinates of common measurement points and key instrument geometric error parameters, completing multi-station network measurements and on-site correction of key parameters.
[0066] The method of this invention acquires multi-station raw observation data (distance, azimuth, and elevation angles) from four or more laser tracker stations for the same set of common measuring points. It uses the fixed target network method to obtain the initial values of the instrument geometric error parameters of the laser tracker. The initial values of the instrument geometric error parameters are substituted into the forward measurement model containing the instrument geometric error parameters to generate the initial values of the coordinates of the common measuring points. The geometric error parameters that have a significant impact on the multi-station network measurement results and can be corrected by the constraints of the common points of the multi-stations are selected as the active error parameters.
[0067] This invention uses the coordinates of common measurement points and activity error parameters as joint optimization variables. During residual construction, this invention does not use the extrinsic parameters of multi-station pose as explicit joint optimization variables. For non-reference stations, this invention calculates the model output point set for that station based on the current activity error parameters, and performs scale-free rigid body registration between this model output point set and the current common measurement point coordinate set to obtain the rotation matrix and translation vector required for residual calculation. The rotation matrix and translation vector are determined by the current point set in each residual calculation process and are not used as global explicit optimization variables in the joint solution. Based on the difference between the registered measurement point coordinates and the current common measurement point coordinates, a multi-station consistent residual is constructed. Through bounded nonlinear least squares optimization, the optimized common measurement point coordinates and activity error parameters are obtained.
[0068] This invention can be used for on-site measurement of large-size components, three-dimensional coordinate measurement in industrial sites, joint measurement with multi-station laser trackers, calculation of coordinates of common measuring points, on-site correction of key geometric error parameters of laser trackers, and fusion processing of multi-station measurement data.
[0069] Example 1
[0070] like Figure 1 and Figure 2 As shown, the present invention provides a multi-station laser tracker network measurement optimization method based on an instrument error model. This method includes:
[0071] S1. Obtain raw observation data from multiple stations and obtain initial values of the instrument geometric error parameters of the laser tracker suitable for the forward measurement model. The initial values of the instrument geometric error can be obtained by using the fixed target network method, or by using the sub-item calibration method, external reference auxiliary method (reference instrument such as a coordinate measuring machine), etc.
[0072] In this embodiment, the multi-station raw observation data consists of raw observation data from four or more laser tracker stations for the same set of common measuring points. Specifically, for the first... The station and the first The original observation data from the public measuring points includes distance, azimuth, and elevation angles.
[0073]
[0074] in, Indicates the first The station position for the first The set of observations from a common measuring point, i.e., the raw observation data; Indicates the first The station position for the first Distance values of a common measuring point Indicates the first The station position for the first Azimuth angle of a common measuring point Indicates the first The station position for the first The elevation angle of a common measuring point , , This indicates the number of laser tracker stations. This indicates the number of common measurement points.
[0075] In this embodiment, the fixed target network method, i.e., the laser tracker geometric error parameter calibration method based on a fixed target network, is used to obtain the initial values of the instrument geometric error parameters of the laser tracker corresponding to each station, including:
[0076] S11, Multiple fixed target points are set up in the measurement space, and the fixed target points maintain their spatial position during the coarse calibration measurement;
[0077] S12, the laser tracker is deployed sequentially at a certain coarse calibration station, and the fixed target point is repeatedly measured at each coarse calibration station to obtain the range, azimuth and elevation angles of the fixed target point;
[0078] S13, using the fixed target point coordinates, coarse calibration station configuration parameters and instrument geometric error parameters as parameters to be estimated, establish the fixed target network observation equation to minimize the difference between the model-predicted observations and the actual observations;
[0079] S14. The initial values of the instrument geometric error parameters are obtained by solving the observation equation of the fixed target network through nonlinear least squares.
[0080]
[0081] in, This represents the initial or estimated coordinates of a fixed target point. This represents the estimated values of the station configuration parameters in the coarse calibration network. This represents the initial value of the instrument's geometric error parameter. Represents a fixed set of target point coordinates. This represents the set of station configuration parameters in the coarse calibration network. This represents the set of instrument geometric error parameters, and S represents the number of coarse calibration stations. Indicates the fixed number of target points. Indicates the coarsely calibrated station number. Indicates the fixed target point number. Indicates the first The first coarse calibration station position for the first Distance measurement values of a fixed target point Indicates the first The first coarse calibration station position for the first Azimuth observations of a fixed target point Indicates the first The first coarse calibration station position for the first The elevation angle observations of a fixed target point This represents the distance measurement value predicted by the geometric error model. This represents the azimuth angle predicted by the geometric error model. This represents the pitch angle predicted by the geometric error model.
[0082] S15, and use the initial values of the instrument geometric error parameters as the initial instrument parameters for subsequent multi-station network joint optimization.
[0083] In one implementation, the initial value of the instrument geometric error parameter is expressed as:
[0084]
[0085] in, Indicates the first Initial values of instrument geometric error parameters for each station Indicates the ranging offset. Indicates the ranging scale factor. This indicates the offset of the horizontal axis relative to the vertical axis. This indicates the y-axis offset of the light beam relative to the origin. This indicates the z-axis offset of the light beam relative to the origin. This represents the angular error of the horizontal axis in the yz plane. This represents the angular error of the beam axis in the xy plane. This represents the first-order cosine scale error coefficient of the azimuth angle. This represents the first-order sine scale error coefficient of the azimuth angle. This represents the second-order cosine scale error coefficient of the azimuth angle. This represents the second-order sine scale error coefficient of the azimuth angle. This indicates that the pitch angle is zero-offset. This represents the first-order cosine scale error coefficient for the pitch angle. This represents the first-order sine scale error coefficient for the pitch angle. This represents the second-order cosine scale error coefficient for the pitch angle. This represents the second-order sine scale error coefficient for the pitch angle.
[0086] S2. Substitute the initial values of the instrument geometric error parameters into the forward measurement model containing the instrument geometric error parameters, and combine them with the original observation data from multiple stations to generate the initial values of the coordinates of the common measuring points.
[0087] In this embodiment, the forward measurement model adopts the NPL geometric error model. The forward measurement model uses ranging offset, ranging scale factor, horizontal axis offset, beam offset, horizontal axis angle error, beam axis angle error, azimuth scale error and elevation scale error to correct the laser tracker observations. Based on the corrected ranging value, corrected azimuth angle, corrected elevation angle, beam origin and beam direction, the coordinates of the measurement point are calculated, and the initial value of the common measurement point coordinates is determined based on the measurement point coordinates.
[0088] In one implementation, the first The initial values of the instrument geometric error parameters for each station are expressed as follows:
[0089]
[0090] in, Indicates the first Initial values of instrument geometric error parameters for each station Indicates the first Distance offset at each station Indicates the first Distance scaling factor for each station Indicates the first The offset of the horizontal axis relative to the vertical axis for each station. Indicates the first The beam at each station is offset in the y-direction relative to the origin. Indicates the first The beam at each station is offset in the z-direction relative to the origin. Indicates the first The angular error of the horizontal axis of each station in the yz plane. Indicates the first The angular error of the beam axis in the xy plane at each station. Indicates the first The first-order cosine scale error coefficient of the azimuth angle for each station. Indicates the first The first-order sine scale error coefficient of the azimuth angle for each station. Indicates the first The second-order cosine scale error coefficient of the azimuth angle for each station. Indicates the first The second-order sine scale error coefficient of the azimuth angle for each station. Indicates the first Pitch angle zero offset for each station Indicates the first The first-order cosine scale error coefficient of the pitch angle at each station Indicates the first The first-order sine scale error coefficient of the pitch angle at each station Indicates the first The second-order cosine scale error coefficient of the pitch angle at each station. Indicates the first The second-order sine scale error coefficient of the pitch angle at each station.
[0091] For the The observation of the first station There are several common measuring points, and the input to the forward measurement model is the distance measurement value. Azimuth Pitch angle and initial values of instrument geometric error parameters The output of the forward measurement model is the three-dimensional coordinates of the measuring point in the local coordinate system of the station, i.e., the measuring point coordinates. .
[0092]
[0093] in, Indicates the first The first station was calculated based on the initial value of the instrument geometric error parameter. Coordinates of each measuring point This represents a forward measurement model that includes instrument geometric error parameters. Indicates the first The station position for the first Distance values of a common measuring point Indicates the first The station position for the first Azimuth angle of a common measuring point Indicates the first The station position for the first The elevation angle of a common measuring point.
[0094] Specifically, the forward measurement model includes the following calculation steps:
[0095] S21, perform error correction on the ranging value to obtain the corrected ranging value. In the implementation including a ranging scaling factor, the corrected ranging value is:
[0096]
[0097] in, This indicates the corrected distance measurement value. Indicates the first The station position for the first Distance values of a common measuring point Indicates the first Distance offset at each station Indicates the first Distance scaling factor for each station.
[0098] In a simplified implementation, the ranging scale factor remains fixed or has been incorporated into the ranging scale correction, and the corrected ranging value is:
[0099]
[0100] in, This indicates the corrected distance measurement value. Indicates the first The station position for the first Distance values of a common measuring point Indicates the first Distance offset at each station.
[0101] S22, perform scale error correction on the azimuth angle to obtain the corrected azimuth angle. The corrected azimuth angle is:
[0102]
[0103] in, This indicates the corrected azimuth angle. Indicates the first The azimuth angle of each station to the i-th common measuring point Indicates the first The first-order cosine scale error coefficient of the azimuth angle for each station. Indicates the first The first-order sine scale error coefficient of the azimuth angle for each station. Indicates the first The second-order cosine scale error coefficient of the azimuth angle for each station. Indicates the first The second-order sine scale error coefficient of the azimuth angle of each station.
[0104] S23, perform zero-position offset and scale error correction on the pitch angle to obtain the corrected pitch angle. The corrected pitch angle is:
[0105]
[0106] in, This indicates the corrected pitch angle. Indicates the first The station position for the first The elevation angle of a common measuring point Indicates the first Pitch angle zero offset for each station Indicates the first The first-order cosine scale error coefficient of the pitch angle at each station Indicates the first The first-order sine scale error coefficient of the pitch angle at each station Indicates the first The second-order cosine scale error coefficient of the pitch angle at each station. Indicates the first The second-order sine scale error coefficient of the pitch angle at each station.
[0107] S24, establish the initial axis vector based on the horizontal axis offset and beam offset. Define the unit vector:
[0108]
[0109] in, This represents the unit vector along the x-axis.
[0110] No. The horizontal axis positioning vector for each station is:
[0111]
[0112] in, Indicates the first The horizontal axis positioning vector of each station in the zero position state. Indicates the first The offset of the horizontal axis relative to the vertical axis for each station. This represents the unit vector along the x-axis.
[0113] No. The beam initial offset vector for each station is:
[0114]
[0115] in, Indicates the first The beam initial offset vector at each station in the zero-position state. This indicates the y-axis offset of the light beam relative to the origin. This indicates the z-axis offset of the light beam relative to the origin.
[0116] No. The zero-position beam direction vector for each station is:
[0117]
[0118] in, Indicates the first The beam direction vector of each station in the zero position state. Indicates rotation about the z-axis The rotation matrix, Indicates the first The angular error of the beam axis in the xy plane at each station. This represents the unit vector along the x-axis.
[0119] S25, calculate the horizontal axis positioning vector based on the corrected azimuth angle.
[0120]
[0121] in, Indicates the first The station points to the first The horizontal axis positioning vector when there are multiple common measurement points Indicates rotation about the z-axis The rotation matrix, This indicates the corrected azimuth angle. Indicates the first The horizontal axis positioning vector of each station in the zero position state.
[0122] S26. Construct a composite rotation matrix based on the corrected azimuth angle, corrected pitch angle, and transverse axis angle error.
[0123]
[0124] in, Indicates the first The station points to the first The composite rotation matrix when there are multiple common measurement points. Indicates rotation about the x-axis The rotation matrix, Indicates rotation about the y-axis The rotation matrix, Indicates rotation about the x-axis The rotation matrix, Indicates rotation about the z-axis The rotation matrix, Indicates the first The angular error of the horizontal axis of each station in the yz plane.
[0125] S27, calculate the... The station points to the first The beam origin at a common measurement point.
[0126]
[0127] in, Indicates the first The station points to the first The beam origin at a common measurement point Indicates the first The station points to the first The horizontal axis positioning vector when there are multiple common measurement points Represents a composite rotation matrix. Indicates rotation about the z-axis The rotation matrix, Indicates the first The beam initial offset vector at each station in the zero position state.
[0128] S28, calculate the... The station points to the first The direction of the beam at a common measurement point.
[0129]
[0130] in, Indicates the first The station points to the first The beam direction vector at a common measurement point Represents a composite rotation matrix. Indicates rotation about the z-axis The rotation matrix, Indicates the first The beam direction vector of each station in the zero position state.
[0131] S29, calculate the coordinates of the measuring point based on the beam start point, the corrected distance measurement value, and the beam direction.
[0132]
[0133] in, Indicates the first The first station was calculated based on the initial value of the instrument geometric error parameter. Coordinates of each measuring point Indicates the first The station points to the first The beam origin at a common measurement point This indicates the corrected distance measurement value. Indicates the first The station points to the first The beam direction vector at a common measurement point.
[0134] S210, determine the initial values of the coordinates of the common measuring points based on the coordinates of the measuring points obtained from the forward calculation.
[0135] In one implementation, the forward calculated coordinates of a designated station are selected as the initial values of the common measuring point coordinates.
[0136]
[0137] in, Indicates the first Initial values of the coordinates of a common measuring point. Indicates the specified station position The first value calculated based on the initial value of the instrument geometric error parameter. Coordinates of each measuring point This indicates the station number used to generate the initial coordinates of the common measurement points.
[0138] In another implementation, the forward calculated coordinates of multiple stations are initially registered to the same coordinate relationship to generate initial values of common measuring point coordinates.
[0139]
[0140] in, Indicates the first Initial values of the coordinates of a common measuring point. This represents the set of stations that participated in the initial value generation. Represents a set The number of stations in the middle. Indicates the first The rigid body registration rotation matrix for each station during the initial value generation process. Indicates the first The rigid body registration translation vector of each station during the initial value generation process. Indicates the first The first station obtained by forward calculation Coordinates of the measuring points.
[0141] Through the above steps, the instrument geometric error parameters obtained from the coarse calibration are substituted into the forward measurement model, and the initial values of the common measuring point coordinates are calculated point by point from the original distance, azimuth, and elevation angles. These initial values of the common measuring point coordinates serve as the initial input for subsequent joint optimization of the common measuring point coordinates and the active error parameters.
[0142] S3. Select the geometric error parameters that have a significant impact on the multi-station network measurement results and can be corrected by the constraints of the common points of the multi-stations from the instrument geometric error parameters as the active error parameters; keep the other parameters at their initial calibration values.
[0143]
[0144] in, Indicates the first The set of activity error parameters for each station participating in joint optimization Indicates the first A complete set of instrument geometric error parameters for each station.
[0145] The selection criteria for activity error parameters include:
[0146] Basis 1: The selected parameters have a strong influence on the measurement results of multi-station network.
[0147] Basis 2: The selected parameters are modifiable under the constraint of multiple station common points.
[0148] Criterion 3: The number of selected parameters should keep the size of the joint optimization variables within a solvable range.
[0149] Criterion 4: Unselected geometric error parameters retain the initial calibration values obtained by the fixed target network method.
[0150] In one implementation, the activity error parameter This includes range offset, horizontal axis offset relative to vertical axis, y-axis offset of beam relative to origin, z-axis offset of beam relative to origin, angular error of beam axis in xy plane, and pitch angle zero offset.
[0151]
[0152] in, Indicates the ranging offset. This indicates the offset of the horizontal axis relative to the vertical axis. This indicates the y-axis offset of the light beam relative to the origin. This indicates the z-axis offset of the light beam relative to the origin. This represents the angular error of the beam axis in the xy plane. This indicates that the pitch angle is zero-offset.
[0153] S4. Construct joint optimization variables by combining the coordinates of common measuring points and the activity error parameters of each non-reference station.
[0154] In one implementation, the expression for the joint optimization variables is:
[0155] ;
[0156] in, Denotes the joint optimization variables. Represents the coordinate matrix of common measurement points. This represents the vector obtained by expanding the coordinate matrix of common measurement points column-wise or row-wise. This represents the activity error parameters for each station other than the reference station. This indicates the number of laser tracker stations participating in the network measurement.
[0157] One of the multiple laser tracker stations is selected as the reference station. The reference station is used to constrain the overall degrees of freedom during the joint optimization process.
[0158] The initial values of the joint optimization variables are composed of the initial values of the coordinates of the common measurement points and the initial values of the activity error parameters:
[0159]
[0160] in, Indicates the initial values of the joint optimization variables. This represents the initial value of the coordinates of the common measuring point. Indicates the first Initial values of the activity error parameters for each non-reference station.
[0161] S5. Calculate the set of measurement point coordinates for each station based on the forward measurement model, perform scale-free rigid body registration between the set of measurement point coordinates and the current common set of measurement point coordinates, and construct multi-station consistency residuals based on the difference between the registered measurement point coordinates and the current common measurement point coordinates.
[0162] In this embodiment, the set of coordinates of each station's measurement points is calculated based on the forward measurement model, including:
[0163] For each station, based on the distance measured, azimuth, elevation angle, and corresponding instrument geometric error parameters, the three-dimensional coordinates of the measuring point at that station are calculated using the forward measurement model, i.e., the measuring point coordinates:
[0164]
[0165] in, Indicates the first The first station was calculated based on the instrument geometric error positive measurement model. Coordinates of each measuring point This represents a forward measurement model that includes instrument geometric error parameters. Indicates the first The station position for the first Distance values of a common measuring point Indicates the first The station position for the first Azimuth angle of a common measuring point Indicates the first The station position for the first The elevation angle of a common measuring point Indicates the first The current instrument geometric error parameters at each station.
[0166] In one implementation, the forward measurement model includes the following calculation process:
[0167] S51, performs range offset correction on the range measurement value.
[0168]
[0169] in, This indicates the corrected distance measurement value. Indicates the first The station position for the first Distance values of a common measuring point Indicates the first Distance offset at each station.
[0170] S52, performs scale error correction on the azimuth angle.
[0171]
[0172] in, This indicates the corrected azimuth angle. Indicates the first The station position for the first Azimuth angle of a common measuring point Indicates the first The first-order cosine scale error coefficient of the azimuth angle for each station. Indicates the first The first-order sine scale error coefficient of the azimuth angle for each station. Indicates the first The second-order cosine scale error coefficient of the azimuth angle for each station. Indicates the first The second-order sine scale error coefficient of the azimuth angle of each station.
[0173] S53 performs zero-position offset and scale error correction on the pitch angle.
[0174]
[0175] in, This indicates the corrected pitch angle. Indicates the first The station position for the first The elevation angle of a common measuring point Indicates the first Pitch angle zero offset for each station Indicates the first The first-order cosine scale error coefficient of the pitch angle at each station Indicates the first The first-order sine scale error coefficient of the pitch angle at each station Indicates the first The second-order cosine scale error coefficient of the pitch angle at each station. Indicates the first The second-order sine scale error coefficient of the pitch angle at each station.
[0176] S54. Calculate the beam origin and beam direction based on the corrected range, corrected azimuth, corrected elevation, axis offset parameters, and beam offset parameters.
[0177]
[0178] in, Indicates the first The calculation of the first station position yields the first Coordinates of each measuring point Indicates the first The station points to the first The beam start point at each measurement point This indicates the corrected distance measurement value. Indicates the first The station points to the first The beam direction unit vector at each measurement point.
[0179] In this embodiment, the present invention does not use the pose extrinsic parameters between the various laser tracker stations as explicit joint optimization variables. For non-reference stations, during each residual calculation, the set of measurement point coordinates output by the instrument error model for that station is performed with the current common set of measurement point coordinates without scale to obtain the rotation matrix and translation vector required for residual calculation.
[0180] Specifically, the steps for constructing multi-station consistent residuals are as follows:
[0181] Step A: Based on the coordinates of the measuring points and the number of common measuring points calculated by the forward measurement model, calculate the first centroid of the output point set of the non-reference station model;
[0182]
[0183] in, Indicates the first The first centroid of the output point set of the non-reference station model. Indicates the number of common measuring points. Indicates the first The first station was calculated using the forward measurement model. Coordinates of the measuring points.
[0184] Step B: Based on the coordinates of the common measurement points in the current joint optimization variables, calculate the second centroid of the current set of common measurement point coordinates;
[0185]
[0186] in, This represents the second centroid of the current set of common measurement point coordinates. Represents the first among the current joint optimization variables. Coordinates of a common measuring point.
[0187] Step C: Based on the first and second centroids, construct the covariance matrix between the output point set of the non-reference station model and the current common measurement point coordinate set;
[0188]
[0189] in, Indicates the first The covariance matrix corresponding to each non-reference station.
[0190] Step D: Perform singular value decomposition on the covariance matrix to obtain the singular value decomposition results;
[0191]
[0192] in, This is the result of singular value decomposition; and This represents the orthogonal matrix obtained from singular value decomposition. This represents a singular value matrix.
[0193] Step E: Calculate the rigid body registration parameters based on the singular value decomposition results. The rigid body registration parameters include the rigid body registration rotation matrix and translation vector.
[0194]
[0195]
[0196] in, Indicates the first Rigid body registration rotation matrices corresponding to each non-reference station. Indicates the first Rigid body registration translation vectors corresponding to each non-reference station.
[0197] Step F: Calculate the coordinates of the measuring points after registration at the non-reference station based on the rigid body registration parameters;
[0198]
[0199] in, Indicates the first After registration of the first non-reference station Coordinates of the measuring points.
[0200] Step G: Based on the registered coordinates of the measuring points and the current coordinates of the common measuring points, construct the multi-station consistency residual.
[0201]
[0202] in, Indicates the first The first station The consistency residuals of multiple stations corresponding to each measurement point.
[0203] For reference stations, residuals can be constructed based on the difference between the coordinates of the model output points of the reference stations and the coordinates of the common measurement points.
[0204]
[0205] in, Indicates the reference station number The residuals corresponding to each measuring point This indicates the reference station position calculated using the forward measurement model. Coordinates of the measuring points.
[0206] With the above residual construction method, the pose relationship between stations is not used as an explicit joint optimization variable in the solution, and the difference in station coordinate systems is handled by rigid body registration in the residual calculation process.
[0207] S6. Based on the multi-station consistency residual, the joint optimization variables are optimized by bounded nonlinear least squares to obtain the optimized coordinates of the common measuring points and the optimized activity error parameters, and then output them.
[0208] In this embodiment, step S6 specifically includes:
[0209] Based on the multi-station consistency residuals, the joint optimization variables are optimized using bounded nonlinear least squares. The least squares optimization model is as follows:
[0210]
[0211] in, Represents the optimized joint optimization variables. Denotes the joint optimization variables to be optimized. This indicates the number of laser tracker stations participating in the network measurement. Indicates the number of common measuring points. Indicates the joint variables The determined first The first station Multi-station consistency residuals at individual measurement points Represents the Euclidean norm;
[0212] Based on the least squares optimization model, optimization is performed with the optimization variables satisfying boundary constraints to obtain the optimized coordinates of the common measurement points and the optimized activity error parameters; the boundary constraints are: , Indicates the lower bound of the joint optimization variables. Indicates the upper bound of the joint optimization variables;
[0213] In one implementation, the boundary of the common measuring point coordinates is set according to the measurement space range, and the boundary of the active error parameter is set according to the physical dimensions of the parameter, the coarse calibration results, and the field measurement conditions.
[0214] After optimization, the optimized coordinates of the common measuring points and the activity error parameters are obtained.
[0215]
[0216] in, This represents the optimized coordinate matrix of the common measurement points. Indicates the first The optimized activity error parameters for each non-reference station.
[0217] In the aforementioned bounded nonlinear least squares optimization process, the multi-station consistency residuals are jointly determined by the observation data of each station, the forward measurement model, and the rigid body registration results. When there are obstructions, unstable target ball contact, local measurement disturbances, or abnormal model residuals at some common measurement points, some large residual components may affect the joint optimization results. To improve the adaptability of the joint solution process of common measurement point coordinates and activity error parameters to abnormal residuals, a robust weighting strategy can be introduced based on the aforementioned bounded nonlinear least squares optimization.
[0218] As a further implementation of step S6, in one embodiment, a two-stage robust optimization strategy is used to solve the joint optimization variables. This two-stage robust optimization strategy is based on the multi-station consistency residuals and bounded nonlinear least squares optimization model established in step S6, without changing the composition of the joint optimization variables or the construction method of the multi-station consistency residuals. In the first stage, unweighted bounded nonlinear least squares optimization is performed on the joint optimization variables to obtain the first-stage optimization result and the first-stage residuals. The first-stage optimization is used to obtain preliminary optimization results for the common measurement point coordinates and activity error parameters, and to provide a basis for the residual distribution for subsequent residual weighting. In the second stage, Tukey robust weights are calculated based on the first-stage residuals, and these weights are applied to the corresponding residual components to form weighted residuals. Subsequently, using the first-stage optimization result as the initial value for the second-stage optimization, bounded nonlinear least squares optimization is performed again on the weighted residuals to obtain the common measurement point coordinates and activity error parameters after robust weighting optimization. The weights of the Tukey robust kernel function are:
[0219]
[0220] in, Indicates the first The robust weights corresponding to each residual component This indicates the first stage of optimization. Each residual component Indicates the residual threshold. Indicates the residual component number.
[0221] The second-stage optimization objective is:
[0222]
[0223] in, Denotes the joint variables after robust weighted optimization. Indicates the joint variables The determined first Each residual component.
[0224] The second stage can also be equivalently represented as least-squares optimization of the weighted residuals.
[0225]
[0226] in, Indicates the first Each weighted residual component.
[0227] The two-stage robust optimization method described above involves obtaining the unweighted residual distribution in the first stage and then applying differentiated weights to each residual component based on its magnitude. Observation components with smaller residuals have higher weights, while those with residuals exceeding a threshold have their weights reduced or set to zero. This reduces the impact of abnormal residuals on the joint calculation results of common measurement point coordinates and activity error parameters.
[0228] It should be noted that the two-stage robust optimization is a preferred implementation of the bounded nonlinear least squares optimization in step S6. Without a robust weighting strategy, this invention can still directly perform bounded nonlinear least squares optimization based on multi-station consistency residuals to obtain the optimized common measurement point coordinates and activity error parameters. With a robust weighting strategy, the output is the robustly weighted optimized common measurement point coordinates and activity error parameters.
[0229] After joint optimization, the optimized coordinates of common measuring points and the optimized activity error parameters are output. The optimized coordinates of common measuring points serve as the measurement results of the multi-station laser tracker network, and the optimized activity error parameters serve as the on-site correction results for the key geometric error parameters of the laser tracker.
[0230] The output results include the following:
[0231] (1) Optimized coordinates of common measuring points;
[0232] (2) Optimized activity error parameters for each station;
[0233] (3) The coordinates of the measuring points at each station, calculated using the instrument error model;
[0234] (4) Consistent residuals across multiple stations;
[0235] (5) Evaluation results of point-to-point distance error.
[0236] The core innovations of the above technical solutions of this invention are as follows:
[0237] 1. Joint optimization of multi-station common measuring points and key error parameters based on instrument error model constraints
[0238] This invention introduces a forward measurement model of instrument geometric error into the network measurement process of a multi-station laser tracker, and uses the coordinates of common measuring points and instrument geometric error parameters as optimization variables. This method obtains the coordinates of common measuring points through multi-station common point consistency constraints and obtains the on-site correction results for key geometric error parameters.
[0239] 2. Activity error parameter subset correction based on multi-station observability
[0240] This invention selects error parameters that significantly affect multi-station network measurements and can be effectively corrected by multi-station common point constraints as active error parameters, while keeping the other parameters at their initial calibration values. This subset of parameters includes ranging offset, horizontal axis offset relative to the vertical axis, beam y-axis offset, beam z-axis offset, beam axis angle error, and pitch angle zero-position offset, etc.
[0241] 3. Residual construction method for non-explicit joint solution of multi-station pose
[0242] This invention does not use the pose extrinsic parameters between multiple laser trackers as explicit joint optimization variables. Instead, during each residual construction, it performs rigid body registration between the point set output by the instrument error model for non-reference stations and the current common measurement point coordinate set, and constructs a multi-station consistent residual based on the registered point set. This design reduces the number of explicit optimization variables and decreases the coupling between station pose parameters and instrument error parameters.
[0243] 4. Within the aforementioned joint optimization framework, a two-stage robust optimization strategy can be employed to reduce the impact of abnormal residuals on the joint optimization results. This strategy, as a preferred implementation method in the joint optimization solution process of this invention, works in conjunction with multi-station consistency residual construction and bounded nonlinear least squares optimization to improve the robustness of the joint solution process for common measurement point coordinates and activity error parameters.
[0244] Example 2
[0245] like Figure 2 As shown, the difference between this embodiment and Embodiment 1 is that this embodiment further provides a multi-station laser tracker network measurement optimization system based on an instrument error model. This system corresponds one-to-one with the multi-station laser tracker network measurement optimization method based on an instrument error model in Embodiment 1. The system includes:
[0246] The acquisition unit is used to acquire raw observation data from multiple stations.
[0247] The calibration unit is used to obtain the initial values of the instrument geometric error parameters of the laser tracker using the fixed target network method;
[0248] The forward measurement model calculation unit is used to substitute the initial values of the instrument geometric error parameters into the forward measurement model containing the instrument geometric error parameters, and combine the original observation data from multiple stations to generate the initial values of the coordinates of common measuring points.
[0249] The active error parameter selection unit is used to select geometric error parameters from the instrument geometric error parameters that have a significant impact on the multi-station network measurement results and can be corrected by the constraints of the multi-station common point as active error parameters;
[0250] The joint optimization variable construction unit is used to construct joint optimization variables by combining the coordinates of common measuring points and the activity error parameters, and the initial value of the joint optimization variable is composed of the initial value of the coordinates of common measuring points and the initial value of the activity error parameters.
[0251] The multi-station consistency residual construction unit is used to calculate the set of measurement point coordinates of each station based on the forward measurement model, perform scale-free rigid body registration between the set of measurement point coordinates and the current common measurement point coordinate set, and construct the multi-station consistency residual based on the difference between the registered measurement point coordinates and the current common measurement point coordinates.
[0252] The optimization unit is used to optimize the joint optimization variables based on the multi-station consistency residuals using bounded nonlinear least squares, to obtain the optimized coordinates of the common measurement points and the optimized activity error parameters.
[0253] The execution process of each unit can be carried out according to the steps of the multi-station laser tracker network measurement optimization method based on the instrument error model in Example 1, and will not be described in detail in this example.
[0254] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0255] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0256] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0257] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0258] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for optimizing networked measurement of multi-station laser trackers based on an instrument error model, characterized in that, The method includes: Obtain raw observation data from multiple stations and initial values of the instrument geometric error parameters of the laser tracker; Substitute the initial values of the instrument geometric error parameters into the forward measurement model containing the instrument geometric error parameters, and combine them with the original observation data from the multiple stations to generate the initial values of the coordinates of the common measuring points. Select the geometric error parameters that have a significant impact on the multi-station network measurement results and can be corrected by the constraints of the common points of the multi-stations from the instrument geometric error parameters as the active error parameters; The coordinates of the common measuring point and the activity error parameter are jointly constructed as a joint optimization variable, and the initial value of the joint optimization variable is composed of the initial value of the coordinates of the common measuring point and the initial value of the activity error parameter. Based on the forward measurement model, the set of measurement point coordinates for each station is calculated, and the set of measurement point coordinates is registered with the current common set of measurement point coordinates without scale. Based on the difference between the registered measurement point coordinates and the current common measurement point coordinates, a multi-station consistency residual is constructed. Based on the multi-station consistency residuals, the joint optimization variables are optimized using bounded nonlinear least squares to obtain the optimized coordinates of the common measurement points and the optimized activity error parameters.
2. The multi-station laser tracker network measurement optimization method based on the instrument error model according to claim 1, characterized in that, The multi-station raw observation data consists of the distance, azimuth, and elevation angles measured by four or more laser tracker stations for the same set of common measuring points.
3. The multi-station laser tracker network measurement optimization method based on the instrument error model according to claim 1, characterized in that, The forward measurement model adopts the NPL geometric error model; The forward measurement model uses ranging offset, ranging scaling factor, horizontal axis offset, beam offset, horizontal axis angle error, beam axis angle error, azimuth scale error, and elevation scale error to correct the laser tracker's observations. Based on the corrected ranging value, corrected azimuth angle, corrected elevation angle, beam origin, and beam direction, the coordinates of the measurement points are calculated, and the initial values of the common measurement point coordinates are determined based on the coordinates of the measurement points.
4. The multi-station laser tracker network measurement optimization method based on the instrument error model according to claim 1, characterized in that, The activity error parameters include ranging offset, horizontal axis offset relative to vertical axis, y-axis offset of beam relative to origin, z-axis offset of beam relative to origin, angular error of beam axis in xy plane, and pitch angle zero offset.
5. The multi-station laser tracker network measurement optimization method based on the instrument error model according to claim 1, characterized in that, The expression for the joint optimization variables is: ; in, Denotes the joint optimization variables. Represents the coordinate matrix of common measurement points. This represents the vector obtained by expanding the coordinate matrix of common measurement points column-wise or row-wise. This represents the activity error parameters for each station other than the reference station. This indicates the number of laser tracker stations participating in the network measurement.
6. The multi-station laser tracker network measurement optimization method based on the instrument error model according to claim 1, characterized in that, The steps for constructing the multi-station consistent residual are as follows: Based on the coordinates of the measuring points and the number of common measuring points calculated by the forward measurement model, the first centroid of the output point set of the non-reference station model is calculated. Based on the coordinates of common measurement points in the current joint optimization variables, calculate the second centroid of the current set of common measurement point coordinates; Based on the first centroid and the second centroid, construct the covariance matrix between the output point set of the non-reference station model and the current common measurement point coordinate set; Perform singular value decomposition on the covariance matrix to obtain the singular value decomposition results; Based on the singular value decomposition results, rigid body registration parameters are calculated, including rigid body registration rotation matrix and translation vector. Calculate the coordinates of the measuring points after registration at the non-reference station based on the rigid body registration parameters. Based on the registered coordinates of the measuring points and the current coordinates of the common measuring points, a multi-station consistency residual is constructed.
7. The multi-station laser tracker network measurement optimization method based on the instrument error model according to claim 1, characterized in that, The formula for calculating the multi-station consistency residual is as follows: ; ; In the formula, Indicates the first The first station The consistency residuals of multiple stations corresponding to each measurement point; Represents the first among the current joint optimization variables. Coordinates of one common measuring point; Indicates the first After registration of the first non-reference station Coordinates of each measuring point; Indicates the first Rigid body registration rotation matrices corresponding to each non-reference station. Indicates the first Rigid body registration translation vectors corresponding to each non-reference station; Indicates the first The first station was calculated using the forward measurement model. Coordinates of the measuring points.
8. The multi-station laser tracker network measurement optimization method based on the instrument error model according to claim 1, characterized in that, Based on the multi-station consistency residuals, the joint optimization variables are optimized using bounded nonlinear least squares to obtain the optimized common measurement point coordinates and optimized activity error parameters, including: Based on the multi-station consistency residuals, a first-stage unweighted bounded nonlinear least squares optimization is performed on the joint optimization variables to obtain the first-stage optimization results and the first-stage residuals; the least squares optimization model is as follows: ,in, Represents the optimized joint optimization variables. Denotes the joint optimization variables to be optimized. This indicates the number of laser tracker stations participating in the network measurement. Indicates the number of common measuring points. Indicates the joint variables The determined first The first station Multi-station consistency residuals at individual measurement points Represents the Euclidean norm; Calculate the Tukey robust weights based on the residuals from the first stage; The optimization results of the first stage are used as the initial values for the second stage optimization. Based on the Tukey robust weight, the multi-station consistency residuals are weighted and the second stage bounded nonlinear least squares optimization is performed to obtain the coordinates of the common measuring points and the activity error parameters after robust weighted optimization. The optimized coordinates of the common measuring points are used as the measurement results of the multi-station laser tracker network, and the optimized active error parameters are used as the on-site correction results of the key geometric error parameters of the laser tracker.
9. The multi-station laser tracker network measurement optimization method based on the instrument error model according to claim 1, characterized in that, The initial values of the instrument's geometric error parameters are obtained through pre-calibration; the pre-calibration method includes at least one of the following: fixed target network method, offline calibration method, on-site coarse calibration method, historical parameter recall method, or manufacturer calibration parameter recall method.
10. A multi-station laser tracker network measurement optimization system based on an instrument error model, characterized in that, The system includes: The acquisition unit is used to acquire raw observation data from multiple stations. The calibration unit is used to obtain the initial values of the instrument geometric error parameters of the laser tracker; The forward measurement model calculation unit is used to substitute the initial value of the instrument geometric error parameter into the forward measurement model containing the instrument geometric error parameter, and combine it with the original observation data of the multi-station to generate the initial value of the coordinates of the common measuring point. The active error parameter selection unit is used to select geometric error parameters from the instrument geometric error parameters that have a significant impact on the multi-station network measurement results and can be corrected by the constraints of the multi-station common point as active error parameters; The joint optimization variable construction unit is used to construct a joint optimization variable by combining the coordinates of the common measuring point and the activity error parameter, and the initial value of the joint optimization variable is composed of the initial value of the coordinates of the common measuring point and the initial value of the activity error parameter. The multi-station consistency residual construction unit is used to calculate the set of measurement point coordinates of each station based on the forward measurement model, perform scale-free rigid body registration of the set of measurement point coordinates with the current common measurement point coordinate set, and construct the multi-station consistency residual based on the difference between the registered measurement point coordinates and the current common measurement point coordinates. The optimization unit is used to optimize the joint optimization variables based on the multi-station consistency residuals using bounded nonlinear least squares to obtain the optimized common measurement point coordinates and optimized activity error parameters.
