Optimization method for multi-actuator loading of complex nodes in bent chords under joint disturbance calibration

By constructing a perturbation calibration method for the loading matrix and constraint matrix, and optimizing the actuator force value, the problem of load distribution and joint control of rigid body motion of complex nodes in bent string members under the condition of variable number of actuators of multiple web members was solved, realizing accurate loading and efficient test control.

CN122859042BActive Publication Date: 2026-11-17POWERCHINA HUADONG ENG CORP LTD
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202611328519.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-08-31
Publication Date
2026-11-17
Estimated Expiration
2046-08-31

AI Technical Summary

Technical Problem

Existing technologies struggle to achieve coordinated control of load distribution and rigid body motion at nodes in multiaxial mechanical performance tests of complex nodes in bent chord members, especially in cases where the number of actuations of multiple web members in a bent chord is variable, where there are shortcomings in the coordinated control of load distribution and rigid body motion at nodes.

Method used

By acquiring loading information of complex nodes in bent string members, a loading matrix and constraint matrix are constructed, initial force values ​​are generated and perturbation calibration is performed, actuator force values ​​are optimized, and a bounded solution and iterative optimization model for redundant loading parameters is established by combining DIC image acquisition and LVDT displacement monitoring. The loading geometric relationship is updated in real time, and node motion is precisely controlled.

Benefits of technology

It enables precise loading of complex nodes of bent chord members, reduces test errors, improves the adaptability and automation of loading schemes, and is suitable for test scenarios with 2 to 4 web members and 3 to 5 actuators. It also reduces the errors caused by the mismatch between empirical loading schemes and the actual boundaries of the specimen.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122859042B_ABST
    Figure CN122859042B_ABST
Patent Text Reader

Abstract

This invention discloses a method for optimizing the loading of multiple actuators at complex nodes of bent chords using joint perturbation calibration, relating to the field of coordinated loading technology for multiple actuators at complex nodes. This invention collects loading information from specimens, equipment, and assemblies; inputs the target axial force of the chord at the staged loading end and selects equilibrium constraints; constructs loading and constraint matrices through loading geometry to generate initial force projection solutions; obtains the perturbation basis by solving the constraint null space; completes positive and negative small-amplitude perturbations under noise and equipment boundary constraints; establishes force-node motion and force-actuator displacement influence matrices; optimizes actuator force values ​​with the goal of minimizing node rigid body motion and the secondary goal of smoothing adjacent-stage force values; constructs an activity set when the actuator reaches the boundary and performs iterative recalculation; and determines subsequent operations based on the target constraint residual and motion prediction residual. This invention solves the shortcomings of conventional loading schemes, such as excessive node rigid body displacement, force distribution relying on experience, and multi-channel coupling interference, and has wide application value.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of coordinated loading technology for complex node multi-actuator systems, specifically to an optimized loading method for complex node multi-actuator systems of bent chords with joint perturbation calibration. Background Technology

[0002] With the continuous expansion of the scale of large-span spatial steel structure and truss building projects, the demand for multi-axis mechanical performance testing of complex nodes of bent chords, as core force-transmitting components of the structure, is increasing year by year. In order to accurately reproduce the true stress state of the nodes, the industry generally adopts multiple independent actuators to apply axial loads to the chords and multiple web members at the loading end simultaneously, and carries out graded quasi-static loading tests in combination with DIC full-field optical measurement and LVDT displacement acquisition equipment. Related load distribution, multi-channel decoupling, and full-field deformation measurement technologies have become key research directions in the field of structural testing. Existing technologies have formed technical solutions from the dimensions of mechanical load distribution mechanism, multi-channel servo decoupling algorithm, DIC displacement feedback, redundant actuator zero-space distribution, and multi-axis loading device for truss nodes. These technologies can realize multi-point loading and deformation monitoring of foundations and are widely used in mechanical verification tests of truss nodes in exhibition halls, high-speed rail stations, and large-span roofs. However, under the condition of fixed D end of bent chord and variable number of actuators of multiple web members, there are still shortcomings in the coordinated control of load distribution and rigid body motion of nodes.

[0003] Chinese patent (publication number CN105758658A) discloses a loading device for multi-point load distribution test of a large-size model. The scheme adopts a spatial radial tensioned beam mechanical distribution structure to divert the single-point load to multiple loading points of the specimen. The load is distributed by relying on the stiffness of the distribution beam itself, reducing the number of jacks to be set up on site. However, the scheme is based on fixed mechanical components to achieve load distribution. It cannot update the geometric relationship in real time with the offset of the clamp and the change of the eccentricity of the loading point during the loading process. It cannot be adapted to the test scenario of variable number of actuation of multiple web members of a bent chord.

[0004] Chinese patent (publication number CN107097975B) discloses a multi-channel feedback decoupling method for structural testing. This method constructs a multi-channel force signal transmission compensation matrix to cancel the force coupling interference between each servo actuator and improve the accuracy of multi-channel synchronous loading. However, this decoupling logic only corrects the electrical signal of the controller and does not combine the resultant force of the nodal plane and the mechanical balance of bending moment to establish a constraint system, so it cannot synchronously control the test error caused by the overall translation and rotation of the nodal.

[0005] Chinese patent (publication number CN110457797B) discloses a method for distributing the tightening torque of bolts for testing bending moment bearing capacity. This method calculates the pre-tightening torque of each group of bolts through the torque balance equation and the deformation coordination relationship of the component, so as to achieve uniform distribution of load on multiple bolts. However, this scheme is designed for the pre-tightening scenario of bolt connection and is not suitable for truss node tests in which multiple independent actuators control the axial force of chords and web members respectively.

[0006] Chinese patent (publication number CN117388065A) discloses a method for processing complex stress loading of materials based on DIC technology. The device uses digital image correlation technology to obtain the full-field strain of the sample and corrects the measurement deviation caused by the rotation of the sample. However, this scheme is only used for the strain acquisition of small material samples and cannot use the DIC measurement results for iterative optimization of actuator redundancy force distribution.

[0007] In summary, existing multi-point loading, multi-channel decoupling, DIC measurement, and redundant load allocation technologies are independent and difficult to form an integrated hierarchical loading system adapted to complex nodes of fixed bending chords at the D end. There is an urgent need for a multi-actuator loading optimization method that integrates real-time geometric updates, entity disturbance calibration, and dual residual hierarchical judgment. Summary of the Invention

[0008] Based on the aforementioned technical problems, this application discloses a multi-actuator loading optimization method for complex nodes of bent chords with joint disturbance calibration, specifically including: Obtain the loading information of the fixed end chord, the loading end chord, each web member and the corresponding independent actuator in the complex node of the bent chord; input the target axial force of the loading end chord of the current loading level; and select at least one of the transverse resultant force, axial resultant force or in-plane bending moment as the equilibrium component. Based on the loading information of each independent actuator, the loading relationship between each independent actuator and the target axial force and balance component of the loading end chord is analyzed to generate the initial actuator force value. While keeping the target axial force and balance component of the loading end chord unchanged, analyze the disturbance adjustment direction of each independent actuator, and perform small incremental disturbances in the positive and negative directions respectively along each disturbance adjustment direction; During the small incremental disturbance, the overall translation and rotation of the node are acquired by the DIC image acquisition device, and the displacement changes of each independent actuator are acquired by the LVDT. Based on the measurement results before and after the disturbance, the correspondence between the force changes of the independent actuators and the node motion and actuator displacement is analyzed. Based on the obtained correspondence, under the conditions that the target axial force and equilibrium components remain unchanged and the actuator force value, predicted displacement and nodal motion boundary are satisfied, the redundant loading parameters are solved in a bounded manner. The optimized actuator force value is obtained with the minimum translation and rotation angle of the nodal rigid body as the primary objective and the smoothness of the force value of adjacent levels as the secondary objective. When any independent actuator reaches the boundary of the force or displacement range, the corresponding boundary value is fixed and the contribution of that actuator to the target constraint is deducted. The remaining independent actuators then take over the correction of the target constraint and recalculate the loading force value. Calculate the target axial force and target constraint residual of the chord at the loading end and the equilibrium component based on the measured actuator force value at the current loading stage. Then, determine whether to continue loading, recalibrate, or stop loading based on the target constraint residual, measured node translation / rotation angle, actuator force value, and displacement boundary.

[0009] Preferably, the initial actuator force value is specifically: based on the actual coordinates of each loading point in the current loading stage and the actuator direction angle, a column vector of contributions of each independent actuator to the resultant force in two directions and the in-plane bending moment of the node is constructed. Combining the contribution column vectors of the loaded end chord actuator and each web member actuator, we obtain the first... The planar loading matrix of the loading stage; wherein, the fixed end chord is a fixed constraint end and is not included in the number of independent actuators, and the constraint reaction force generated by the fixed support of the fixed end chord is determined by the planar loading matrix and the force value of the actuator in the current loading stage; Using the target axial force of the chord at the loading end as a mandatory constraint, and combining it with the selected equilibrium components, construct the target constraint matrix and the target constraint vector; Calculate the initial prior weights based on the section modulus of each web member; The prior weights of the web members are extended to a full-channel vector, and the prior total coefficients that make the selected balance component closest to the target value are solved. By combining the target axial force of the loading end chord and the prior total coefficient, an initial actuator force vector is generated and projected onto the target constraint within the actuator force range. If a feasible projection solution exists, the solution is used as the initial actuator force value and the target of the current loading stage is determined to be reachable. If no solution exists, the target of the current loading stage is determined to be unreachable.

[0010] Preferably, the perturbation adjustment direction specifically involves: extracting the first... Load-level free actuator set Corresponding target constraint matrix Solve for the condition that satisfies perturbation basis matrix And extend it to all actuator channels to obtain the first Loading-level perturbation basis matrix ,in, The column vectors represent the adjustment directions for each independent disturbance.

[0011] Preferably, the small incremental perturbation is specifically based on the set of free actuators. and target constraint matrix Calculate the number of redundant degrees of freedom; When the number of redundant degrees of freedom is equal to 0, the perturbation calibration is skipped, the motion prediction residual is not calculated or used, and the target constraint residual, the measured node translation / rotation angle of DIC or LVDT, the actuator force value and displacement boundary are directly checked; when the number of redundant degrees of freedom is greater than 0, the subsequent perturbation steps are executed, and each perturbation basis is subjected to small incremental perturbations in the positive and negative directions along each perturbation adjustment direction to obtain the perturbed actuator force vector.

[0012] Preferably, the analysis of the correspondence between the force change of the independent actuator and the node motion and actuator displacement specifically involves: analyzing the correspondence between the force change of the independent actuator and the node motion and actuator displacement using the central difference method to obtain the column vectors of influence of node motion and actuator displacement, as shown in the formula:

[0013]

[0014] in, For the first Loading level The column vector of the influence of each perturbation basis on the node motion , These are the node motion vectors after the forward and reverse perturbations, respectively. For the first Loading level The column vector of the influence of each perturbation basis on the actuator displacement , These are the actuator displacement vectors after positive and negative disturbances, respectively. The magnitude is a small increment. By combining the influence column vectors, we obtain the influence matrix. and .

[0015] Preferably, the optimized actuator force value specifically involves: predicting node motion and actuator displacement based on the influence matrix; Based on the node motion vectors, the node motion objective function is established, and the formula is as follows:

[0016] in, Let the objective function be the node motion. This is the node motion normalization matrix, with the diagonal terms being the reciprocals of the allowable thresholds for node translation and rotation. For the first Predict node motion vectors at the loading level; Within the actuator loading range, solve for the redundant parameters that minimize the node motion objective function to obtain the minimum node motion index; if multiple near-optimal solutions exist, establish a secondary objective function, the formula of which is:

[0017] in, For secondary objective functions, When =1, the initial projection solution of this level is used. For reference; For the first Initial projection solution of loading level, This is the force vector of the actuator from the previous loading stage. For the first Loading level perturbation basis matrix, This is the normalized matrix for the actuator force values; Solve for the condition The optimal solution of the secondary objective ,in, For the near-optimal tolerance of the main target of node motion, For the first The bounded feasible domain of the loading-level redundancy parameters includes at least the actuator force boundary, the predicted actuator displacement boundary, and the allowable boundaries for overall node translation and overall rotation. Based on the optimal redundancy parameters, the adjusted actuator force value is calculated using the following formula:

[0018] in, For the first The actuator force vector optimized at the loading stage. This is the optimal solution for the secondary objective.

[0019] Preferably, the actuator loading range is specifically defined as follows: the actuator loading range consists of the actuator force range, the actuator displacement range, and the allowable ranges of overall node translation and overall node rotation; wherein, the actuator force range is determined by the intersection of the actuator capability, the force sensor range, the fixture bearing capacity, and the specimen safety capability; the actuator displacement range is determined by the intersection of the actuator stroke, the LVDT range, the fixture allowable displacement, and the test safety displacement.

[0020] Preferably, the recalculation of the loading force value specifically involves: adding the actuators that have reached the boundary to the active set, fixing their boundary values, and updating the set of free actuators; calculating the contribution of the locked actuators to the target constraint based on their boundary values, and subtracting this contribution from the original target constraint vector to obtain the corrected target constraint vector that the remaining free actuators need to bear; re-extracting the target constraint matrix based on the remaining free actuators, and recalculating the disturbance basis matrix, node motion influence matrix, and actuator displacement influence matrix under the corrected target constraint vector to solve for the loading force value of the free actuators in the current loading level; merging the boundary values ​​of the locked actuators and the loading force values ​​of the free actuators into the actuator force value of the current loading level; if the constraint conditions still cannot be met after re-solving, the current loading level is determined to be infeasible.

[0021] Preferably, the target constraint residual and motion prediction residual are specifically calculated as follows: the target constraint residual is calculated based on the measured actuator force value of the current loading stage, when the... When the number of redundant degrees of freedom in the loading stage is greater than 0, the motion prediction residual is calculated based on the node motion information collected by DIC; when .... When the number of redundant degrees of freedom in the loading stage is equal to 0, motion prediction residuals are not calculated or used, and the formula is:

[0022] in, For the first Load-level target constraint normalized residuals For the first The target constraint matrix of the loading level, For the first Measured force vector of the actuator at the loading stage. For the first Number of load-level objective equality constraints For the first Loading level The allowable residuals of each constraint component, For the first The target constraint vector of the loading level, This is the calculation to obtain the maximum normalized deviation. The motion prediction residual is calculated based on the node motion information collected by DIC, using the following formula:

[0023] in, For the first Loading-level motion prediction normalized residuals For the first Load-level DIC measured node motion vectors For the first Predict node motion vectors at the loading level. It is an infinite norm.

[0024] Preferably, the determination of whether to continue loading, recalibrate, or stop loading specifically involves: if the first... The number of redundant degrees of freedom at the loading level is greater than 0. , If the measured actuator force, displacement and nodal motion all meet the corresponding range, then the current loading level is deemed feasible and the next loading level is entered. If the first If the number of redundant degrees of freedom in the loading stage is equal to 0, then motion prediction residuals are not calculated or used. In the target constraint residual If the measured node translation, rotation, actuator force and displacement by DIC or LVDT all meet the corresponding range, the current loading level is deemed feasible and the next loading level is entered. If any actuator reaches the force or displacement range boundary, the current loading level is paused and the active set is redistributed. If the first The number of redundant degrees of freedom in the loading stage is greater than 0 and the motion prediction residuals If no actuator boundary anomaly is found, the current loading level is paused and the disturbance calibration is re-executed; If the objective constraint residual If there are no boundary actuators that can be redistributed for the activity set, or if the feasibility conditions are still not met after redistribution and recalibration, then the current target constraint is determined to be infeasible, loading is stopped, and manual intervention is awaited.

[0025] Compared with the prior art, the technical solution of this application has the following technical effects: This invention constructs an integrated solution system that incorporates real-time updated loading geometry, a variable number of independent actuators, target axial force of the loading end chord and multiple types of equilibrium constraints, rigid body motion of DIC nodes, LVDT actuator displacement, actuator force / displacement boundaries, and node motion thresholds into the same hierarchical loading optimization model. It can be adapted to various bending chord node test scenarios with 2 to 4 web members and 3 to 5 actuators, significantly improving its versatility.

[0026] This invention uses the section modulus of the web members to construct the initial force distribution prior and projects it onto the core constraint for correction. It takes into account both the stiffness characteristics of the members themselves and the forced balance requirements of the test, which greatly reduces the test error caused by the mismatch between the empirical loading scheme and the actual boundary of the specimen, and eliminates the need for multiple manual adjustments of the distribution ratio.

[0027] This invention establishes a quantitative correlation between redundancy allocation and node translation, rotation, and actuation displacement through measured data of physical specimens without changing the axial force and equilibrium components of the chord at the loading end. The calibration results closely match the mechanical response caused by the actual eccentricity of the specimen and the offset of the fixture, and accurately suppress the non-target rigid body motion of the node.

[0028] This invention constructs an iterative redistribution logic for the activity set, automatically locks the actuators that reach the force / displacement boundary, and resolves the loading scheme based on the remaining free channels. It fully considers the multiple safety boundary constraints of the equipment, fixtures, and specimens, and the feasibility of each loading scheme can be predicted, avoiding the experimental risks of actuator overload and local overload of the specimen.

[0029] This invention sets up two sets of judgment indicators: target constraint residual and motion prediction residual. By strictly verifying various deviations through the infinite norm, it can accurately distinguish three types of faults: inaccurate balance components, calibration model failure, and equipment boundary exceeding limits. These faults trigger corresponding handling processes such as reallocation, recalibration, and shutdown for manual intervention. The fault location is clear, and the degree of automation of graded loading is higher.

[0030] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the preferred embodiments of this application are described in detail below with reference to the accompanying drawings.

[0031] The above and other objects, advantages and features of this application will become more apparent to those skilled in the art from the following detailed description of specific embodiments in conjunction with the accompanying drawings. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In all drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0033] Based on the description of the figures and their corresponding technical content in the document, the titles of the figures are as follows: Figure 1 A flowchart of the overall process for optimizing the multi-actuator loading of complex nodes in a bent chord under joint disturbance calibration; Figure 2 Architecture diagram of a multi-actuator loading optimization method for complex nodes of bent chords for joint disturbance calibration; Figure 3 This is a schematic diagram of the node multi-channel loading and global measurement layout of this application. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. In the following description, specific details such as specific configurations and components are provided merely to help fully understand the embodiments of this application. Therefore, those skilled in the art should understand that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this application. In addition, for clarity and brevity, descriptions of known functions and structures are omitted in the embodiments.

[0035] It should be understood that the phrase "an embodiment" or "this embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "an embodiment" or "this embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0036] Furthermore, reference numerals and / or letters may be repeated in different examples within this application. Such repetition is for the purpose of simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or settings discussed.

[0037] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, B exists alone, and A and B exist simultaneously. The term " / and" in this article describes another type of relationship between related objects, indicating that two relationships can exist. For example, A / and B can mean: A exists alone, and A and B exist alone. In addition, the character " / " in this article generally indicates that the related objects before and after it are in an "or" relationship.

[0038] In this article, the term "at least one" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, "at least one of A and B" can mean: A exists alone, A and B exist simultaneously, or B exists alone.

[0039] It should also be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion.

[0040] Example 1 mainly describes the multi-actuator loading optimization method for complex nodes of bent chords with joint disturbance calibration, such as... Figure 1 As shown, it specifically includes: Obtain the loading information of the fixed end chord, the loading end chord, each web member and the corresponding independent actuator in the complex node of the bent chord; input the target axial force of the loading end chord of the current loading level; and select at least one of the transverse resultant force, axial resultant force or in-plane bending moment as the equilibrium component. Obtain the loading information of the fixed end chord, the loading end chord, each web member and the corresponding independent actuator in the complex node of the bent chord; input the target axial force of the loading end chord of the current loading level; and select at least one of the transverse resultant force, axial resultant force or in-plane bending moment as the equilibrium component. Based on the loading information of each independent actuator, the loading relationship between each independent actuator and the target axial force and balance component of the loading end chord is analyzed to generate the initial actuator force value. While keeping the target axial force and balance component of the loading end chord unchanged, analyze the disturbance adjustment direction of each independent actuator, and perform small incremental disturbances in the positive and negative directions respectively along each disturbance adjustment direction; During the small incremental disturbance, the overall translation and rotation of the node are acquired by the DIC image acquisition device, and the displacement changes of each independent actuator are acquired by the LVDT. Based on the measurement results before and after the disturbance, the correspondence between the force changes of the independent actuators and the node motion and actuator displacement is analyzed. Based on the obtained correspondence, under the conditions that the target axial force and equilibrium components remain unchanged and the actuator force value, predicted displacement and nodal motion boundary are satisfied, the redundant loading parameters are solved in a bounded manner. The optimized actuator force value is obtained with the minimum translation and rotation angle of the nodal rigid body as the primary objective and the smoothness of the force value of adjacent levels as the secondary objective. When any independent actuator reaches the boundary of the force or displacement range, the corresponding boundary value is fixed and the contribution of that actuator to the target constraint is deducted. The remaining independent actuators then take over the correction of the target constraint and recalculate the loading force value. Based on the obtained correspondence, under the premise that the target axial force and balance component constraints remain unchanged, the redundant loading parameters are solved in a bounded manner so that the adjusted actuator force value simultaneously satisfies the force value range, displacement range, and allowable range of overall node translation and overall rotation angle. The optimization is carried out with minimizing the rigid body translation and rotation angle of the node as the primary objective and smoothing the force values ​​of adjacent loading stages as the secondary objective, and the optimized actuator force value is obtained. When any independent actuator reaches the boundary of the force or displacement range, the corresponding boundary value is fixed and the contribution of that actuator to the target constraint is deducted. The remaining independent actuators then take over the correction of the target constraint and recalculate the loading force value. Calculate the target constraint residuals of the target axial force and equilibrium components of the chord at the loading end based on the measured actuator force values ​​of the current loading stage; when the... When the number of redundant degrees of freedom in the loading stage is greater than 0, the motion prediction residual is calculated based on the node motion information collected by DIC; when .... When the number of redundant degrees of freedom in the loading stage is equal to 0, the motion prediction residual is not calculated or used, and the decision to continue loading, recalibrate, or stop loading is made based on the target constraint residual, the measured node translation / rotation angle, the actuator force value, and the displacement boundary.

[0041] Furthermore, the loading information of the fixed end chord, the loading end chord, each web member and the corresponding independent actuator in the complex node specifically includes the fixed boundary parameters of the fixed end chord, the basic parameters of the loading end chord, the basic parameters of each web member, the performance parameters of each independent actuator and the assembly parameters of each actuator with the chord and web members; Among them, the fixed boundary parameters of the fixed end chord include the constraint form of the fixed support, the spatial coordinates of the fixed point, and the constraint stiffness; the fixed end chord is a fixed constraint end, which only provides constraint reaction force and constraint reaction moment, and does not participate in the load distribution as an independent actuator; its reaction force is determined by the balance of the resultant force and resultant moment formed by the loading end chord and each web member actuator in the current loading stage. The basic parameters of the loading end chord include the geometric dimensions of the loading end chord, the coordinates of the loading point position, and the connection method with the corresponding independent actuator; The basic parameters of each web member include the geometric dimensions, section modulus, load point coordinates, and connection method with the corresponding independent actuator for each web member; The performance parameters of each independent actuator include the actuator model, rated loading force range, rated displacement range, force sensor accuracy, and LVDT displacement acquisition device measurement accuracy; The assembly parameters of each actuator and chord and web member include the actuator's installation direction angle, loading point eccentricity, and assembly clearance compensation parameters; After acquiring the loading information, input the target axial force of the loading end chord of the current loading level, and select at least one of the following as the equilibrium component: transverse resultant force, axial resultant force, or in-plane bending moment.

[0042] Furthermore, the number of web rods is 2 to 4, and the number of independent actuators is [missing information]. ,in 1 represents the number of web members, and 1 represents the number of independent actuators corresponding to the loading end chord.

[0043] Furthermore, the initial actuator force values ​​are specifically as follows: based on the actual coordinates of each loading point in the current loading stage and the actuator direction angle, a column vector is constructed to represent the contribution of each independent actuator to the resultant force in the two directions and the in-plane bending moment at the node. The formula is:

[0044] in, For the first Loading level The contribution column vector of each actuator For the first Loading level The actual force direction angle of the actuator , For the first Loading level The x and y coordinates of each loading point relative to the node reference point; Combining the contribution column vectors of the loaded end chord actuator and each web member actuator, we obtain the first... The planar loading matrix for the loading level is given by the following formula:

[0045]

[0046] in, For the first Loading level plane loading matrix, This represents the total number of independent actuators. The generalized planar reaction force provided by the fixed end chord to the specimen. For the first Force vector of stage actuator; Using the target axial force of the loaded end chord as a mandatory constraint, and combining it with the selected equilibrium components, the target constraint matrix and target constraint vector are constructed as follows:

[0047]

[0048] in, For the first Loading level target constraint matrix, To select the unit column vector of the force values ​​of the chord actuator at the loading end, This is a selection matrix for choosing the components to be retained from the resultant forces and moments in the plane. The number of rows corresponds to the number of selected equilibrium components. For the first Target axial force of the loading end chord in the loading stage. For the first The target constraint vector of the loading level, The selected target balance component vector; The initial prior weights are calculated based on the section modulus of each web member, using the following formula:

[0049] in, For the first Loading level Prior weights of the section modulus of the root and web members. For the first Section modulus of the root web member, For the collection of strut channels; Extend the prior weights of the abdominal struts to a full-channel vector. Solve for the prior total coefficients that make the selected equilibrium component closest to the target value. The formula is:

[0050] in, To select the dimensionless normalization matrix for the balanced components, the diagonal terms are taken as the reciprocals of the corresponding allowable residuals. To find the parameters that minimize the subsequent objective function The value of , It is an L2 norm; Combining the target axial force of the loading end chord and the prior total coefficient, an initial actuator force vector is generated and projected onto the target constraint to obtain the initial projection solution, which serves as the initial actuator force value. The formula is as follows:

[0051]

[0052] The constraints are , ,in, For the first Initial prior force vector of loading stage For the first Initial projection solution of loading level, This is the normalized matrix for the actuator force values. The first Loading-level actuator force lower and upper limit vectors.

[0053] Furthermore, based on the above bounded projection constraints, target reachability is determined: if there exists a constraint that satisfies... and If the actuator force vector is obtained, the target of the current loading stage is determined to be achievable; if there is no bounded solution that satisfies the above conditions, the target of the current loading stage is determined to be unachievable, and the loading of that stage is stopped or the target value is adjusted and recalculated by the test personnel.

[0054] Furthermore, the specific direction of the perturbation adjustment is as follows: extract the first... Load-level free actuator set Corresponding target constraint matrix Solve for the condition that satisfies perturbation basis matrix And extend it to all actuator channels to obtain the first Loading-level perturbation basis matrix ,in, The column vectors represent the adjustment directions for each independent disturbance.

[0055] Furthermore, the small incremental perturbation specifically refers to: based on the set of free actuators. and target constraint matrix The formula for calculating the number of redundant degrees of freedom is:

[0056] in, For the first Number of redundant degrees of freedom at the loading level For the first The current set of free actuators at the current level. for rank; like If the disturbance calibration is skipped, the motion prediction residual is not calculated or used, and the target constraint residual, DIC or LVDT measured node translation / rotation, actuator force value and displacement boundary are directly checked; if If the value is greater than 0, then the subsequent perturbation step is executed. For each perturbation basis, small incremental perturbations are applied in both the forward and reverse directions along the respective perturbation adjustment directions to obtain the actuator force vector after perturbation. The formula is:

[0057]

[0058] in, The first Loading level The actuator force vectors after positive and negative perturbations of the perturbation base. For the first Initial projection solution of loading level, For small increments, For the first Loading level A perturbation basis column vector.

[0059] Furthermore, the small increment amplitude must simultaneously meet the requirements of noise identifiability and actuator boundary requirements. For the first... Loading level Given a perturbation basis, first calculate the boundary margin of each independent actuator channel under forward and reverse perturbations, using the following formula:

[0060] The upper bound of the boundary magnitude is obtained from the boundary margins of all non-zero disturbance components, using the following formula:

[0061] The lower and upper limits of the disturbance amplitude are determined based on the noise identifiability requirements and the safety reduction factor, using the following formula:

[0062]

[0063] in, The upper bound of the boundary amplitude, The noise is identifiable lower limit. This is the upper limit of the amplitude after safety reduction; This is the noise amplification factor. For the first Loading level Equivalent measurement and control noise amplitude of each disturbance parameter For safety reduction factor, Perturbation basis vectors In the Components on each independent actuator channel , The first Loading level The upper and lower limits of the force of each independent actuator. For the first Loading level Initial projected force value of each independent actuator; like Greater than If the value is not found, it indicates that there is no identifiable and safe symmetrical disturbance amplitude within the device boundary. The disturbance base is marked as invalid and the process proceeds to the decision of whether the activity set redistribution or the current loading level is infeasible.

[0064] Furthermore, the effectiveness of the disturbance is determined by the difference between the DIC node motion and the LVDT actuator displacement after positive and negative disturbances, using the following formula:

[0065] in, For identifiable indicators of disturbance response, The threshold for motion noise at DIC nodes. LVDT displacement noise threshold; when When less than 1, in Internal adaptive increase And re-execute the positive and negative perturbations to achieve If the value is still less than 1, the perturbation basis is marked as invalid.

[0066] Furthermore, obtaining the overall translation and overall rotation of a node is specifically as follows: The initial coordinates are obtained by selecting two stable reference feature points on the node using the DIC image acquisition device. , and the current coordinates after the disturbance , ; The formula for calculating the overall rotation angle of a node is:

[0067] in, For the first Overall rotation angle of the loading-level node (unit: rad) Constructing a two-dimensional rotation matrix Based on a two-dimensional rotation matrix The formula for calculating the overall translation vector of a node is:

[0068]

[0069] in, For the first The overall translation vector of the load-level node, its components These represent the horizontal and vertical translations, respectively; the node motion vector is... .

[0070] Furthermore, the specific analysis of the correspondence between the force change of the independent actuator and the nodal motion and actuator displacement is as follows: The correspondence between the force change of the independent actuator and the nodal motion and actuator displacement is analyzed using the central difference method, yielding the column vectors of influence from nodal motion and actuator displacement, as shown in the following formulas:

[0071]

[0072] in, For the first Loading level The column vector of the influence of each perturbation basis on the node motion , These are the node motion vectors after the forward and reverse perturbations, respectively. For the first Loading level The column vector of the influence of each perturbation basis on the actuator displacement , These are the actuator displacement vectors after positive and negative disturbances, respectively. The magnitude is a small increment. By combining the influence column vectors, we obtain the influence matrix. and .

[0073] Furthermore, optimizing the actuator force involves predicting the node motion and actuator displacement based on the influence matrix, using the following formula:

[0074]

[0075] in, For the first Predict node motion vectors at the loading level. For the first The node motion vector of the loading-level disturbance calibration reference state. For the first Load-level redundancy allocation parameter vector, For the first Predict the actuator displacement vector at the loading stage. For the first The actuator displacement vector under the load-level disturbance calibration reference state; Based on the node motion vectors, the node motion objective function is established, and the formula is as follows:

[0076] in, Let the objective function be the node motion. This is the node motion normalization matrix, with the diagonal terms being the reciprocals of the allowable thresholds for node translation and rotation. Within the actuator loading range, solve for the redundant parameters that minimize the node motion objective function to obtain the minimum node motion index; if multiple near-optimal solutions exist, establish a secondary objective function, the formula of which is:

[0077] in, For secondary objective functions, When =1, the initial projection solution of this level is used. For reference; For the first Initial projection solution of loading level, This is the force vector of the actuator from the previous loading stage. For the first Loading level perturbation basis matrix, This is the normalized matrix for the actuator force values; Solve for the condition The optimal solution of the secondary objective ,in, For the near-optimal tolerance of the main target of node motion, For the first The bounded feasible domain of the loading-level redundancy parameters includes at least the actuator force boundary, the predicted actuator displacement boundary, and the allowable boundaries for overall node translation and overall rotation. Based on the optimal redundancy parameters, the adjusted actuator force value is calculated using the following formula:

[0078] in, For the first The actuator force vector after loading level optimization.

[0079] Furthermore, the actuator loading range is specifically defined as follows: the actuator loading range consists of the actuator force range, the actuator displacement range, and the allowable ranges of overall node translation and overall node rotation; wherein, the actuator force range is determined by the intersection of the actuator capability, the force sensor range, the fixture bearing capacity, and the specimen safety capability; the actuator displacement range is determined by the intersection of the actuator stroke, the LVDT range, the fixture allowable displacement, and the test safety displacement.

[0080] Furthermore, the recalculation of the loading force value specifically involves: adding the actuators that have reached the boundary to the active set, fixing their boundary values, and updating the set of free actuators; calculating the contribution of the locked actuators to the target constraint based on their boundary values, and subtracting this contribution from the original target constraint vector to obtain the modified target constraint vector that the remaining free actuators need to bear; re-extracting the target constraint matrix based on the remaining free actuators, and recalculating the perturbation basis matrix, nodal motion influence matrix, and actuator displacement influence matrix under the modified target constraint vector to solve for the loading force value of the free actuators in the current loading level; merging the boundary values ​​of the locked actuators and the loading force values ​​of the free actuators into the actuator force value of the current loading level; if the constraint conditions still cannot be met after re-solving, the current loading level is determined to be infeasible.

[0081] Furthermore, the target constraint residual and motion prediction residual are specifically calculated as follows: the target constraint residual is calculated based on the measured actuator force value of the current loading stage, using the following formula:

[0082] in, For the first Load-level target constraint normalized residuals For the first The target constraint matrix of the loading level, For the first Measured force vector of the actuator at the loading stage. For the first Number of load-level objective equality constraints For the first Loading level The allowable residuals of each constraint component, For the first The target constraint vector of the loading level; The motion prediction residual is calculated based on the node motion information collected by DIC, using the following formula:

[0083] in, For the first Loading-level motion prediction normalized residuals For the first Load-level DIC measured node motion vectors For the first Predict node motion vectors at the loading level. It is an infinite norm.

[0084] Furthermore, determining whether to continue loading, recalibrate, or stop loading specifically involves: if the first... The number of redundant degrees of freedom at the loading level is greater than 0. , If the measured actuator force, displacement and nodal motion all meet the corresponding range, then the current loading level is deemed feasible and the next loading level is entered. If the first If the number of redundant degrees of freedom in the loading stage is equal to 0, then motion prediction residuals are not calculated or used. In the target constraint residual If the measured node translation, rotation, actuator force and displacement by DIC or LVDT all meet the corresponding range, the current loading level is deemed feasible and the next loading level is entered. If any actuator reaches the force or displacement range boundary, the current loading level is paused and the active set is redistributed. If the first The number of redundant degrees of freedom in the loading stage is greater than 0 and the motion prediction residuals If no actuator boundary anomaly is found, the current loading level is paused and the disturbance calibration is re-executed; If the objective constraint residual If there are no boundary actuators that can be redistributed for the activity set, or if the feasibility conditions are still not met after redistribution and recalibration, then the current target constraint is determined to be infeasible, loading is stopped, and manual intervention is awaited.

[0085] This embodiment details a method for optimizing the loading of multiple actuators at complex nodes of a bent chord using joint perturbation calibration. It involves collecting loading information from the specimen, equipment, and assembly; inputting the target axial force of the chord at the loading end and selecting equilibrium constraints; constructing the loading and constraint matrices through loading geometry to generate initial force projection solutions; obtaining the perturbation basis by solving the constraint null space; performing positive and negative small-amplitude perturbations under noise and equipment boundary constraints; establishing force-node motion and force-actuator displacement influence matrices; optimizing actuator force values ​​with minimizing node rigid body motion as the primary objective and smoothing adjacent-level force values ​​as the secondary objective; constructing an activity set when the actuator reaches the boundary and performing iterative recalculation; and determining subsequent operations based on the target constraint residuals and motion prediction residuals.

[0086] Example 2, based on Example 1, describes in detail the entity loading system built based on this method, as follows: The physical loading system used in this embodiment includes a specimen fixing and reaction unit, a loading execution unit, a measurement and acquisition unit, and a calculation and control unit. The specimen fixing and reaction unit includes an external rigid reaction frame and a D-end fixing clamp or a fixing support. The D-end fixing clamp is used to apply a fixed boundary constraint to the fixed end chord of the bent chord node specimen and bear the generalized planar reaction force and bending moment transmitted by the specimen during the test. The D-end fixing clamp does not participate in the loading as an independent actuator.

[0087] The loading execution unit includes several independent actuators. The number of independent actuators is determined according to the number of web members of the specimen. When the number of web members is m, m is 2 to 4. The loading execution unit includes one independent actuator connected to the loading end chord and m independent actuators connected to each web member respectively. Therefore, the number of independent actuators is n=m+1, that is, 3 to 5. One end of each independent actuator is installed or supported on the outer rigid reaction frame, and the other end is connected to the corresponding loading end chord or web member through a loading clamp, hinged connector or adjustable connector, and is used to apply axial loading force to the node specimen along the loading direction of the corresponding member.

[0088] The measurement and acquisition unit includes force and displacement acquisition devices for each independent actuator, an LVDT (Low-Level Digital Transmission Device) for monitoring slippage or constraint deformation at the fixed end boundary, and a DIC (Digital Indexed Conversion) image acquisition device positioned facing the core region of the node. The force acquisition devices acquire the real-time loading force of each independent actuator, the displacement acquisition devices acquire the output displacement of each independent actuator, the LVDT determines whether slippage or abnormal constraint deformation has occurred at the D-end fixed boundary, and the DIC image acquisition device acquires the overall translation and rotation of the core region of the node.

[0089] The calculation and control unit is communicatively connected to the loading execution unit and the measurement and acquisition unit. It is used to receive the target rod force, the upper and lower limits of the actuator force value, the actuator displacement, the real-time force value, the LVDT measurement value, and the nodal motion measured by DIC. According to this method, it performs target constraint construction, bounded load solution, joint disturbance calibration, nodal motion minimization, activity set weight distribution, and feasibility residual judgment to obtain the target loading force or loading correction amount of each independent actuator in the current loading level, and outputs control commands to the corresponding independent actuator.

[0090] This embodiment describes in detail the physical loading system built based on this method. The system is divided into four main units: specimen fixing reaction force, loading execution, measurement and acquisition, and calculation and control. The specimen is fixed and bears all reaction force bending moments by relying on the rigid reaction frame and the D-end clamp. At the same time, 3 to 5 independent actuators are matched according to 2 to 4 web members, which are respectively connected to the loading end chord and each web member to apply axial load. Through a standardized arrangement scheme in which the number of actuators matches the number of web members, a multi-actuator physical connection structure adapted to the number of web members is formed, and the system provides the conditions for the acquisition and execution of force, displacement and nodal motion data for the method described in Embodiment 1.

[0091] Example 3, based on Example 1 or 2, uses the specimen parameters, equipment parameters, and measurement arrangement of a double-web member bent chord node as an example to illustrate the calculation process and parameter substitution method of this approach, as detailed below: A double-web bending truss node specimen with rigid fixed support at end D was selected. The number of web members was m=2, and three independent hydraulic servo actuators were used (one for the chord at the loading end and one for each of the two web members). The chord at the fixed end was fully constrained.

[0092] The experiment was equipped with an LVDT displacement acquisition device and a DIC optical full-field measurement system to collect actuator displacement and nodal motion data, such as Figure 3 The diagram shown is a schematic of the multi-channel loading and full-domain measurement arrangement of the nodes. Its outer arc-shaped rigid reaction frame serves as the support base for all actuators. A total of 3 sets of independent servo actuators are arranged. Actuator C corresponds to the loading point of the chord section at the loading end, and actuators A and B correspond to the loading points of the two web members respectively. The fixed clamp or fixed support at end D is used to constrain the chord section at the fixed end. The figure shows that two types of measuring devices are used to collect data in a coordinated manner. The legend distinguishes each measuring element: the red rectangular element is the LVDT displacement sensor, and the grid-filled area is the DIC optical measurement observation area. Two sets of LVDTs are arranged along the axial direction of the loading rod of each actuator. LVDTs No. 1 and No. 2 are installed at the loading end of the loading end chord; LVDTs No. 7 and No. 8 are installed on the upper web rod corresponding to actuator A, and LVDTs No. 5 and No. 6 are installed on the lower web rod corresponding to actuator B; two sets of LVDTs No. 3 and No. 4 are installed at the constraint end of the D chord to monitor the slippage of the D end boundary or the deformation of the fixture; except for the D end boundary monitoring point, the other LVDTs are used to synchronously collect the output displacement of each independent actuator, providing raw data of actuator displacement changes for the disturbance calibration process; A grid-like DIC observation area is defined in the core intersection region of the nodes. The DIC equipment selects two stable feature points in the observation area and collects the lateral and longitudinal translations and overall rotation of the nodes under graded loading and small positive and negative disturbance conditions throughout the entire process. This provides full-field deformation measurement data for calculating the node motion vector, constructing the node motion influence matrix, and solving the motion prediction residual in this invention. The entire measurement setup is fully adapted to the data acquisition requirements of the joint disturbance calibration loading optimization method of this invention. LVDT and DIC synchronously acquire actuator displacement and nodal rigid body motion information, fully supporting the entire process of initial force value solution, disturbance influence relationship calibration, and dual residual determination of test data input. The C-chord section of the specimen has a cross-sectional area of ​​8400 mm², and is made of Q355 steel (yield strength 355 MPa). The target axial force of the C-chord at the final loading stage is calculated based on the 8400 mm² cross-sectional area of ​​the specimen and the 355 MPa yield strength of the steel. The target axial force of the C-chord at each intermediate loading stage is set independently by the test personnel in combination with the safety boundary of the specimen. No fixed empirical loading gradient is used throughout the process. When the section moduli of the two web members are the same, an equal initial weight will be generated. If there is a difference in the section modulus, a different weight will be generated accordingly. The resultant lateral force at the node is selected as zero as the equilibrium component. This component, together with the target axial force of the C chord, is used to construct a set of constraint equations. The rank of the constraint matrix is ​​2. The three actuators form a set of constraints to maintain redundant degrees of freedom. This degree of freedom can reduce the overall translation and rotation of the node obtained by DIC measurement without changing the axial force and resultant lateral force of the C chord. The generalized reaction force of the D end on the specimen is obtained by solving the load of all actuators in reverse. Before the formal graded loading, positive and negative small incremental perturbations are applied to all constraints to maintain the perturbation basis. The perturbation amplitude simultaneously meets the two requirements of being higher than the equivalent noise of the equipment and not touching the force boundary of the actuator throughout the process. After the perturbation is completed, the two types of influence matrices, node motion and actuator displacement, are obtained by the central difference algorithm. Then, the optimal loading force of the three actuators at this stage is solved by combining the primary objective of minimizing node deformation, the secondary objective of smoothing inter-stage loads, and all equipment constraints. After replacing the actuator or adjusting the fixture, you only need to re-intersect and update the upper and lower limits of force and displacement for each channel by combining the equipment parameters, sensor range, fixture and specimen safety limits.

[0093] This embodiment, based on the actual specimen parameters and measurement arrangement of the double-web member bent chord node, explains the calculation process and parameter substitution method of this approach. The process involves determining the final target axial force through chord section parameters, generating initial load weights using web member section modulus, constructing a constraint system by selecting the lateral resultant force as the equilibrium component, establishing the deformation influence matrix through constraint redundancy degrees of freedom combined with positive and negative small-amplitude disturbances, and solving the actuator load using a primary and secondary dual-objective approach. Furthermore, after equipment replacement, only the channel boundary is updated without altering the core solution logic, thus forming a loading regime calculation process applicable to the double-web member node working condition of three independent actuators.

[0094] The above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. For those skilled in the art, the present invention can have various modifications and variations. Any changes, modifications, substitutions, integrations, and parameter changes made to these embodiments within the spirit and principles of the present invention, without departing from the principles and spirit of the present invention, through conventional substitutions or to achieve the same function, fall within the scope of protection of the present invention.

Claims

1. A method for loading optimization of a complex joint of a curved chorded beam with multiple actuators in joint perturbation calibration, characterized in that, include: Obtain the loading information of the fixed end chord, the loading end chord, each web member and the corresponding independent actuator in the complex node of the bent chord; input the target axial force of the loading end chord of the current loading level; and select at least one of the transverse resultant force, axial resultant force or in-plane bending moment as the equilibrium component. Based on the loading information of each independent actuator, the loading relationship between each independent actuator and the target axial force and balance component of the loading end chord is analyzed to generate the initial actuator force value; While keeping the target axial force and balance component of the loading end chord unchanged, analyze the disturbance adjustment direction of each independent actuator, and perform small incremental disturbances in the positive and negative directions respectively along each disturbance adjustment direction; The small incremental perturbation specifically involves: calculating the number of redundant degrees of freedom based on the set of free actuators and the target constraint matrix; When the number of redundant degrees of freedom is equal to 0, the perturbation calibration is skipped, the motion prediction residual is not calculated or used, and the target constraint residual, DIC or LVDT measured node translation, rotation, actuator force value and displacement boundary are directly checked; when the number of redundant degrees of freedom is greater than 0, the subsequent perturbation steps are executed, and each perturbation basis is subjected to small incremental perturbations in the positive and negative directions along each perturbation adjustment direction to obtain the perturbed actuator force vector; During the small incremental disturbance, the overall translation and rotation of the node are acquired by the DIC image acquisition device, and the displacement changes of each independent actuator are acquired by the LVDT. Based on the measurement results before and after the disturbance, the correspondence between the force changes of the independent actuators and the node motion and actuator displacement is analyzed. Based on the obtained correspondence, under the conditions that the target axial force and equilibrium components remain unchanged and the actuator force value, predicted displacement and nodal motion boundary are satisfied, the redundant loading parameters are solved in a bounded manner. The optimized actuator force value is obtained with the minimum translation and rotation angle of the nodal rigid body as the primary objective and the smoothness of the force value of adjacent levels as the secondary objective. When any independent actuator reaches the boundary of the force or displacement range, the corresponding boundary value is fixed and the contribution of that actuator to the target constraint is deducted. The remaining independent actuators then take over the correction of the target constraint and recalculate the loading force value. Calculate the target axial force and target constraint residual of the chord at the loading end and the equilibrium component based on the measured actuator force value at the current loading stage. Then, determine whether to continue loading, recalibrate, or stop loading based on the target constraint residual, measured node translation, rotation angle, actuator force value, and displacement boundary.

2. The method for optimizing the loading of multiple actuators at complex nodes of a bent chord based on joint disturbance calibration according to claim 1, characterized in that, The initial actuator force value is specifically: based on the actual coordinates of each loading point in the current loading stage and the actuator direction angle, a column vector of contributions of each independent actuator to the resultant force in two directions and the in-plane bending moment of the node is constructed. Combining the contribution column vectors of the loaded end chord actuator and each web member actuator, we obtain the first... The planar loading matrix of the loading stage; wherein, the fixed end chord is a fixed constraint end and is not included in the number of independent actuators, and the constraint reaction force generated by the fixed support of the fixed end chord is determined by the planar loading matrix and the force value of the actuator in the current loading stage; Using the target axial force of the chord at the loading end as a mandatory constraint, and combining it with the selected equilibrium components, construct the target constraint matrix and the target constraint vector; Calculate the initial prior weights based on the section modulus of each web member; The prior weights of the web members are extended to a full-channel vector, and the prior total coefficients that make the selected balance component closest to the target value are solved. By combining the target axial force of the loading end chord and the prior total coefficient, an initial actuator force vector is generated and projected onto the target constraint within the actuator force range. If a feasible projection solution exists, the solution is used as the initial actuator force value and the target of the current loading stage is determined to be reachable. If no solution exists, the target of the current loading stage is determined to be unreachable.

3. The method for optimizing the loading of multiple actuators at complex nodes of a bent chord based on joint disturbance calibration according to claim 1, characterized in that, The perturbation adjustment direction is specifically as follows: extracting the first... Load-level free actuator set Corresponding target constraint matrix Solve for the condition that satisfies perturbation basis matrix And extend it to all actuator channels to obtain the first Loading-level perturbation basis matrix ,in, The column vectors represent the adjustment directions for each independent disturbance.

4. The method for optimizing the loading of multiple actuators at complex nodes of a bent chord based on joint disturbance calibration according to claim 1, characterized in that, The analysis of the correspondence between the force change of the independent actuator and the nodal motion and actuator displacement specifically involves: analyzing the correspondence between the force change of the independent actuator and the nodal motion and actuator displacement using the central difference method to obtain the column vectors of influence of nodal motion and actuator displacement, as shown in the following formula: in, For the first Loading level The column vector of the influence of each perturbation basis on the node motion , These are the node motion vectors after the forward and reverse perturbations, respectively. For the first Loading level The column vector of the influence of each perturbation basis on the actuator displacement , These are the actuator displacement vectors after positive and negative disturbances, respectively. The magnitude is a small increment. By combining the influence column vectors, we obtain the influence matrix. and .

5. The method for optimizing the loading of multiple actuators at complex nodes of a bent chord based on joint disturbance calibration according to claim 1, characterized in that, The optimized actuator force value is specifically: based on the influence matrix, the node motion and actuator displacement are predicted. Based on the node motion vectors, the node motion objective function is established, and the formula is as follows: in, Let the objective function be the node motion. This is the node motion normalization matrix, with the diagonal terms being the reciprocals of the allowable thresholds for node translation and rotation. For the first Predict node motion vectors at the loading level; Within the actuator loading range, solve for the redundant parameters that minimize the node motion objective function to obtain the minimum node motion index; if multiple near-optimal solutions exist, establish a secondary objective function, the formula of which is: in, For secondary objective functions, When =1, the initial projection solution of this level is used. For reference; For the first Initial projection solution of loading level, This is the force vector of the actuator in the previous loading stage. For the first Loading level perturbation basis matrix, This is the normalized matrix for the actuator force values. For the first Load-level redundancy allocation parameter vector; Solve for the satisfaction The optimal solution of the secondary objective ,in, For the near-optimal tolerance of the main target of node motion, For the first The bounded feasible domain of the loading-level redundancy parameters includes at least the actuator force boundary, the predicted actuator displacement boundary, and the allowable boundaries for overall node translation and overall rotation. Based on the optimal redundancy parameters, the adjusted actuator force value is calculated using the following formula: in, For the first The actuator force vector optimized at the loading stage. This is the optimal solution for the secondary objective.

6. The method for optimizing the loading of multiple actuators at complex nodes of a bent chord based on joint disturbance calibration according to claim 5, characterized in that, The actuator loading range is specifically defined as follows: the actuator loading range consists of the actuator force range, the actuator displacement range, and the allowable ranges of overall node translation and overall node rotation; wherein, the actuator force range is determined by the intersection of the actuator capability, the force sensor range, the fixture bearing capacity, and the specimen safety capability; the actuator displacement range is determined by the intersection of the actuator stroke, the LVDT range, the fixture allowable displacement, and the test safety displacement.

7. The method for optimizing the loading of multiple actuators at complex nodes of a bent chord based on joint disturbance calibration according to claim 1, characterized in that, The recalculation of the loading force value specifically involves: adding the actuators that have reached the boundary to the active set, fixing their boundary values, and updating the set of free actuators; calculating the contribution of the locked actuators to the target constraint based on their boundary values, and subtracting the contribution from the original target constraint vector to obtain the corrected target constraint vector that the remaining free actuators need to bear; Based on the remaining free actuators, the target constraint matrix is ​​re-extracted, and the disturbance basis matrix, nodal motion influence matrix, and actuator displacement influence matrix are recalculated under the modified target constraint vector to solve for the free actuator loading force value of the current loading level. The locked actuator boundary value and the free actuator loading force value are merged into the actuator force value of the current loading level. If the constraint conditions still cannot be met after resolving, the current loading level is determined to be infeasible.

8. The method for optimizing the loading of multiple actuators at complex nodes of a bent chord based on joint disturbance calibration according to claim 1, characterized in that, The target constraint residual and motion prediction residual are specifically calculated as follows: the target constraint residual is calculated based on the measured actuator force value of the current loading stage; when the... When the number of redundant degrees of freedom in the loading stage is greater than 0, the motion prediction residual is calculated based on the node motion information collected by DIC; when .... When the number of redundant degrees of freedom in the loading stage is equal to 0, motion prediction residuals are not calculated or used. The formula for the residual of the objective constraint is: in, For the first Load-level target constraint normalized residuals The first of the target constraint deviation vectors The absolute value of each component, For the first The target constraint matrix of the loading level, For the first Measured force vector of the actuator at the loading stage. For the first Number of load-level objective equality constraints For the first Loading level The allowable residuals of each constraint component, For the first The target constraint vector of the loading level, This is the calculation to obtain the maximum normalized deviation. The motion prediction residual is calculated based on the node motion information collected by DIC, using the following formula: in, For the first Loading-level motion prediction normalized residuals For the first Load-level DIC measured node motion vectors For the first Predict node motion vectors at the loading level. It is an infinite norm. This is the normalized matrix for node motion.

9. The method for optimizing the loading of multiple actuators at complex nodes of a bent chord based on joint disturbance calibration according to claim 8, characterized in that, The determination of whether to continue loading, recalibrate, or stop loading specifically involves: if the first... The number of redundant degrees of freedom at the loading level is greater than 0. , If the measured actuator force, displacement and nodal motion all meet the corresponding range, then the current loading level is deemed feasible and the next loading level is entered. If the first If the number of redundant degrees of freedom in the loading stage is equal to 0, then motion prediction residuals are not calculated or used. In the target constraint residual If the measured node translation, rotation, actuator force and displacement by DIC or LVDT all meet the corresponding range, the current loading level is deemed feasible and the next loading level is entered. If any actuator reaches the force or displacement range boundary, the current loading level is paused and the active set is redistributed. If the first The number of redundant degrees of freedom in the loading stage is greater than 0 and the motion prediction residuals If no actuator boundary anomaly is found, the current loading level is paused and the disturbance calibration is re-executed; If the objective constraint residual If there are no boundary actuators that can be redistributed for the activity set, or if the feasibility conditions are still not met after redistribution and recalibration, then the current target constraint is determined to be infeasible, loading is stopped, and manual intervention is awaited.

Citation Information

Patent Citations

  • Experiment loading device for realizing large-size model multipoint load distribution

    CN105758658A

  • Multi-signal feedback loading decoupling method and apparatus for structural testing systems

    CN107097975B

  • A method for distributing the tightening torque of a test moment bearing bolt.

    CN110457797B

  • Complex stress loading device and processing method based on DIC technology

    CN117388065A

  • Force-displacement hybrid control method for three-degree-of-freedom loading system

    CN108241289A