A Precision Attitude Control Method for Positioning Posts Based on Grating Ruler Feedback

Abstract

This invention discloses a precision attitude control method for positioning columns based on grating ruler feedback, specifically relating to the field of precision attitude control technology for industrial equipment. It addresses the challenges of stable online sensing, reliable inversion, and controllable adjustment of the positioning column's end attitude under continuous production and frequent model changeover conditions. By deploying three non-collinear grating rulers circumferentially around the positioning column and establishing a geometric anchoring system with the installation reference plane as the origin, combined with synchronous acquisition under a unified control cycle, attitude observation vector generation, constraint equation inversion, and observability determination of equation condition numbers and residuals, multi-source displacements are uniformly anchored to the same spatial reference frame, forming a closed-loop attitude adjustment link. This achieves continuous online sensing and precision attitude control of the positioning column's end attitude without relying on manual alignment, ensuring consistent and stable assembly accuracy and repeatability even under continuous production and frequent model changeover conditions.

Description

Technical Field

[0001] This invention relates to the field of precision attitude control technology for industrial equipment, specifically a precision attitude control method for positioning columns based on grating ruler feedback. Background Technology

[0002] In automated assembly and machining production lines, locating posts typically serve as critical components for assembly positioning, tool setting references, or workpiece support. Their end posture directly affects assembly clearance, consistency tolerances, and the stability of repeated positioning across multiple workstations. In existing production lines, the height and posture of locating posts often rely on manual alignment, dial indicators, or levels for offline calibration. In some cases, one or two displacement sensors or simple encoders are added near the locating post to monitor unidirectional displacement and roughly determine whether there is a deviation from the set position. These practices generally lack a unified spatial reference system, and the sensor installation geometry is incompletely described. They typically only monitor axial displacement or a certain lateral deflection, making it difficult to fully characterize the coupling relationship between end-position lifting, pitch, and tilt in three-dimensional space. Furthermore, there is a lack of verifiable boundary definitions for installation drift and non-slip states, resulting in a lack of objective and quantifiable criteria for whether the positioning reference has "quietly drifted" after a period of operation.

[0003] With increased production cycle time and model changeover frequency, some digital transformation solutions have begun to use multi-sensor combinations to monitor the status of positioning columns. However, these solutions often only establish simple linear conversion relationships under ideal geometric assumptions, directly mapping multiple displacement readings to height and angle estimates. They lack systematic verification of whether the sensor measurement directions meet the conditions for observable attitude, and do not constrain or evaluate indicators such as equation condition number and inversion residuals. Furthermore, most existing solutions do not pay sufficient attention to the time consistency, missing data compensation, and jump suppression of acquired data. Industrial bus delays, short-term packet loss, or occasional anomalies often directly enter the attitude calculation and attitude control links, easily introducing misjudgments under complex operating conditions such as off-center load, temperature drift, and amplified guide clearance. In addition, existing technologies generally do not versionize or track calibration parameters, threshold configurations, and operating records, making it difficult to trace parameter adjustment processes and operational behaviors. This makes it difficult to reconstruct the attitude control context when quality disputes or equipment malfunctions occur.

[0004] Against this backdrop, the core technical challenges faced on the production site can be summarized as follows: how to achieve continuous, stable, and traceable online sensing and closed-loop attitude control of the positioning column end attitude under continuous production and frequent model change conditions, without relying on manual alignment, based on multi-source displacement sensing and digital control methods, and maintain the observability of attitude inversion and the safe and reliable execution of control commands even when disturbances such as installation drift, load changes, guide clearance, and communication jitter exist. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a precision attitude control method for positioning posts based on grating ruler feedback, thereby solving the problems mentioned in the background section.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for precise attitude control of a positioning post based on grating ruler feedback, comprising:

[0007] S1. Arrange three non-collinear optical scales at different positions around the positioning column, establish a geometric anchor system with the installation reference plane as the origin, and write the measurement direction and zero offset parameters of each optical scale.

[0008] S2. Synchronously collect the displacement of three grating rulers according to a unified control cycle, form an attitude observation vector and add the same cycle identifier;

[0009] S3. Based on the geometric anchor system, establish a set of constraint equations from the displacement to the end attitude parameters, invert to obtain the axial displacement, pitch angle, and roll angle, and perform observability determination on the equation condition number and residuals to output the effective attitude solution.

[0010] S4. Compare the effective attitude solution with the target attitude to obtain the attitude error. Decompose the error into offset, load and clearance terms to generate corresponding compensation weights.

[0011] S5. Based on the attitude response mapping, convert the attitude error into the control quantity of the execution unit, add travel constraints, rate constraints, and synchronization constraints to solve the control instruction set and issue it for execution.

[0012] S6. Repeatedly perform acquisition, inversion, and solution until the attitude error meets the convergence threshold within a preset number of consecutive cycles, and solidify the geometric anchor parameters, compensation weights, and control constraint configurations for this working condition for reset and recall.

[0013] Furthermore, S1 includes:

[0014] Three optical grating rulers are installed at different positions around the positioning column. The measurement directions of the three optical grating rulers are not collinear with each other under the geometric anchoring system with the installation reference surface as the origin.

[0015] The measurement direction parameters and zero-point offset parameters of each grating ruler are determined and written into the field through calibration.

[0016] The geometric anchor system includes the zero-position definition of the mounting reference surface and the spatial orientation relationship of each grating ruler relative to the mounting reference surface.

[0017] Furthermore, when the positioning column is in the reference state, the readings of each grating ruler are collected according to a unified control cycle;

[0018] The zero-point offset parameter is defined as the offset between the actual reading of each grating ruler under the reference state and the theoretical zero point.

[0019] Within a preset load range and preset runtime, the drift of the zero-point offset parameter is detected. When the drift exceeds the drift threshold, recalibration is triggered to generate a new version.

[0020] The calibration parameters and drift threshold configurations are written to the parameter storage area of ​​the equipment control system in the form of version records.

[0021] Furthermore, S2 includes:

[0022] The control system synchronously acquires the displacement of three grating rulers through the industrial bus according to a unified control cycle, and aligns the displacement to the reference time of the unified control cycle based on the acquisition time marker returned by the grating acquisition unit.

[0023] Alignment processing is performed when the difference in acquisition time exceeds the preset alignment threshold.

[0024] When there is a short-term missing component and the preset completion threshold and preset missing window period limit are met, the missing displacement is generated by interpolation.

[0025] Set a low confidence flag when a significant jump value is detected;

[0026] When performing interpolation on short-term missing data, maintain a low-confidence marker and write the attitude observation vector, period identifier, and low-confidence marker into the run record area and associate them with the evidence chain trace area.

[0027] Furthermore, S3 includes:

[0028] The control system establishes a set of constraint equations that relate the displacement of the three grating rulers to the axial displacement, pitch angle, and tilt angle of the positioning column based on the geometric anchor system.

[0029] The attitude observation vector is read by the period identifier and substituted into the constraint equation system to obtain the attitude parameter combination. The equation condition number and inversion residual are calculated and compared with the corresponding preset threshold.

[0030] When any index exceeds the corresponding preset threshold, retain the most recent valid attitude solution and set an observation degradation flag;

[0031] When the number of equation conditions and the inversion residual do not exceed the corresponding preset threshold within a consecutive preset number of recovery cycles, the effective attitude solution is updated and the observation degradation marker is cleared. The observation degradation event and the effective parameter version identifier are written into the operation record area and associated with the evidence chain trace area.

[0032] Furthermore, S4 includes:

[0033] After obtaining an effective attitude solution, the control system calculates the attitude error through the target attitude and decomposes the error into three attitude error components—bias term, load term, and clearance term—within a preset observation window. The error characteristic quantities are calculated separately and then nonnegated and normalized to obtain the compensation weight.

[0034] The compensation weights are smoothly adjusted under the constraints of a fixed smoothing coefficient and a limited range. The compensation weights are frozen under the freeze condition and restricted under the row limit condition.

[0035] Compensation weights, error decomposition rules, observation window configurations, and related threshold parameters are generated and solidified into configurations with version identifiers. These are then stored along with applicable working condition tags and target attitude source identifiers to ensure version consistency and ease of traceability when called under different working conditions.

[0036] Furthermore, S5 includes:

[0037] The control system acquires the attitude error and weights the attitude error according to a preset compensation weight;

[0038] By mapping the attitude response, the attitude error is converted into the control quantity of the execution unit, and the control instruction set is solved to drive the positioning column to complete the lifting and tilting adjustment within a unified control cycle.

[0039] The control quantity of each execution unit represents the amplitude and direction of the action it should achieve in the current cycle, ensuring that the executed action meets the target posture requirements;

[0040] During the control process, stroke, speed, and synchronization constraints are incorporated to ensure that the amplitude, speed, and relative deviation of each execution unit do not exceed the set limits.

[0041] Furthermore, the control system performs constraint verification and amplitude limiting on the control quantities of the candidate execution units to ensure that the action amplitude, rate and relative deviation of each execution unit meet the set constraints.

[0042] If the constraints are not met, adjust the control quantity and perform idempotency and deduplication processing on the control instruction set to ensure that the control instructions are consistent during transmission, retry, and readback, and avoid repeated execution.

[0043] Furthermore, S6 includes:

[0044] The control system uses a unified control cycle as its rhythm to cyclically execute attitude observation vector generation, effective attitude solution inversion, attitude error calculation, compensation weight update, and control command set solution and issuance until the attitude error meets the convergence threshold within a preset number of consecutive cycles, triggering a stop judgment and completing the attitude adjustment closed loop.

[0045] At the end of each unified control cycle, the control system reads the attitude error and the status flag of the effective attitude solution. When all components of the attitude error do not exceed the convergence threshold, the continuous count is incremented.

[0046] When the continuous counting reaches the preset number of cycles, the posture adjustment is determined to be complete and the issuance of new control command sets is stopped;

[0047] After the attitude adjustment is completed, the geometric anchor parameters, compensation weights and control constraint configurations for this working condition are fixed, and the corresponding fixed configurations are loaded through the working condition identification index to enter closed-loop adjustment;

[0048] The control system generates operation records for the attitude adjustment process and results, and associates them with cycle and version identifiers for subsequent traceability and continuous improvement.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] 1. By deploying three non-collinear optical grating rulers around the positioning column and establishing a geometric anchoring system with the installation reference plane as the origin, combined with synchronous acquisition under a unified control cycle, attitude observation vector generation, constraint equation inversion, and observability determination of equation condition number and residual, the multi-source displacement is uniformly anchored to the same spatial reference frame and forms a closed-loop attitude adjustment link. This enables continuous online sensing and precise attitude adjustment control of the positioning column end attitude without relying on manual alignment, ensuring that the consistency and stability of assembly accuracy and repeatable positioning can still be maintained under continuous production and frequent model change conditions.

[0051] 2. By decomposing attitude errors into offset, load, and clearance terms and generating compensation weights, supporting mechanisms such as acquisition alignment, missing data completion, observation degradation judgment, constraint verification, and protection strategies are introduced. Version locking and evidence chain recording are implemented for geometric anchor parameters, zero-position offset parameters, compensation weights, control constraints, and related thresholds. This enables adaptive compensation and safety concession for complex disturbances such as zero-position drift, load changes, guide clearance, and communication jitter, improving the convergence stability of the attitude adjustment process and the consistency and traceability of cross-batch operation. Attached Figure Description

[0052] Figure 1 This is a flowchart illustrating a precision attitude control method for a positioning post based on grating ruler feedback according to the present invention. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] Example: Figure 1 A flowchart illustrating a precise attitude control method for a positioning post based on grating ruler feedback is provided. The method includes:

[0055] S1. Arrange three non-collinear optical scales at different positions around the positioning column, establish a geometric anchor system with the installation reference plane as the origin, and write the measurement direction and zero offset parameters of each optical scale.

[0056] S2. Synchronously collect the displacement of three grating rulers according to a unified control cycle, form an attitude observation vector and add the same cycle identifier;

[0057] S3. Based on the geometric anchor system, establish a set of constraint equations from the displacement to the end attitude parameters, invert to obtain the axial displacement, pitch angle, and roll angle, and perform observability determination on the equation condition number and residuals to output the effective attitude solution.

[0058] S4. Compare the effective attitude solution with the target attitude to obtain the attitude error. Decompose the error into offset, load and clearance terms to generate corresponding compensation weights.

[0059] S5. Based on the attitude response mapping, convert the attitude error into the control quantity of the execution unit, add travel constraints, rate constraints, and synchronization constraints to solve the control instruction set and issue it for execution.

[0060] S6. Repeatedly perform acquisition, inversion, and solution until the attitude error meets the convergence threshold within a preset number of consecutive cycles, and solidify the geometric anchor parameters, compensation weights, and control constraint configurations for this working condition for reset and recall.

[0061] The technical connections and implementation logic of the six steps are as follows:

[0062] In the precision attitude control of the positioning post, firstly, the control system deploys three non-collinear optical grating rulers at different positions around the positioning post and establishes a geometric anchor system with the installation reference plane as the origin. Simultaneously, the measurement direction and zero-position offset parameters S1 of each optical grating ruler are written into the system. Next, the system synchronously acquires the displacement of the three optical grating rulers according to a unified control cycle, forming an attitude observation vector, and adds a period identifier to ensure the uniformity and continuity of the data S2. Then, based on the geometric anchor system, the control system establishes a set of constraint equations between the displacement and the end attitude parameters. Through inversion, the axial displacement, pitch angle, and roll angle are obtained, and observability is determined, outputting the desired values. The system obtains the effective attitude solution S3; next, it compares the effective attitude solution with the target attitude, calculates the attitude error, and decomposes the error according to the bias term, load term, and clearance term to generate the corresponding compensation weights S4; then, through attitude response mapping, the control system converts the attitude error into the control quantity of the execution unit, and solves and issues the control instruction set S5 by combining the travel constraint, rate constraint, and synchronization constraint; finally, the system cyclically executes the acquisition, inversion, and solution until the attitude error meets the convergence threshold within a preset number of consecutive cycles, and solidifies the geometric anchor parameters, compensation weights, and control constraint configurations for this working condition for subsequent reset calls S6.

[0063] S1. Three non-collinear optical scales are installed at different positions around the positioning column to establish a geometric anchoring system with the installation reference plane as the origin. The measurement direction and zero-position offset parameters of each optical scale are written in. The specific implementation is as follows:

[0064] In actual production lines, the positioning post serves as a key component for assembly positioning, tool setting reference, or workpiece bearing. Its end posture directly determines the subsequent assembly accuracy, machining consistency, and reliability of repeated positioning. To achieve continuous sensing and controllable adjustment of the positioning post's end posture without relying on manual alignment, this embodiment deploys three non-collinear grating rulers at different circumferential positions on the positioning post. Each grating ruler is a displacement measuring device used to measure linear displacement along its sensitive axis. The three grating rulers are spatially non-collinear, thus making the spatial posture of the positioning post's end observable. Here, "different circumferential positions" means that the installation positions of the three grating rulers are distributed along the circumference of the positioning post and do not overlap with each other. "Non-collinear" means that under the geometric anchoring system, the measurement directions of the three grating rulers are not collinear with each other and meet the independent observation criteria.

[0065] The independent observation criterion prioritizes the response determination criterion, which states that when the positioning post is in its reference state and axial micro-displacement and two mutually orthogonal micro-tilt movements are applied respectively, the changes in the three grating readings are not proportional and are not simultaneously zero. When micro-tilt movements cannot be applied due to unmet interlocking conditions, the degradation relationship determination criterion is adopted, which states that when the stability and unidirectionality of the ratio changes among the three grating readings are simultaneously met within the preset observation window, the current arrangement relationship is determined to be a degradation relationship and is considered to fail to meet the preset consistency threshold, thus ensuring that the subsequent attitude inversion relationship does not have a degradation problem.

[0066] Both axial micro-displacement and micro-tilt are controlled small-amplitude movements applied within a range not exceeding a preset safety boundary. The preset safety boundary is jointly determined by the rated stroke, rated load, and anti-collision limit of the positioning column. The amplitude of the movement must not trigger mechanical limit or safety interlock. The amplitude of the movement is jointly determined by the structural stiffness of the positioning column, the displacement resolution of the grating ruler, and the on-site process accuracy requirements. Preferably, the amplitude of the axial micro-displacement can be set to several tens of micrometers to several hundreds of micrometers, and the micro-tilt movement can be set to several arcseconds to several arcminutes. Each movement can be repeated a preset number of times, and the preset number of times is locked with the version. Displacement readings are collected during the stable time period of each movement to obtain the statistical value of the stable segment for judgment confirmation. Based on this, a reproducible calibration process is formed and a corresponding version record is generated.

[0067] Preferably, the three grating rulers can be arranged in a combination of positions 120 degrees apart circumferentially to meet installation space constraints and ensure that the attitude observation sensitivity meets the process accuracy requirements. This arrangement is suitable under the conditions that the different circumferential positions, the non-collinearity of the measurement directions, and the independent observation criteria are met. The grating rulers operate continuously within the effective stroke range of the positioning column, used to sense in real time the axial displacement changes of the positioning column and the attitude changes caused by off-center loading, structural deflection, or guide clearance during equipment operation. The effective stroke refers to the entire displacement range within which the positioning column is allowed to move and participate in the operation during actual process.

[0068] Based on the aforementioned layout of the grating rulers, this embodiment establishes a geometric anchor system with the installation reference plane as the origin. The installation reference plane is the plane that is fixedly connected to the positioning column and the main body of the equipment and serves as the overall installation reference. Its physical meaning is the spatial zero reference for the attitude measurement and control process. The geometric anchor system is a reference system used to uniformly describe the spatial relationship between each grating ruler and the positioning column. It includes at least the zero-position definition of the installation reference plane, the spatial orientation relationship of each grating ruler relative to the installation reference plane, the fixed expression of each measurement direction under the reference system, the unit diameter of the displacement reading, and the data acquisition sequence rules and periodic identifier generation rules.

[0069] To ensure the reproducibility of the project, the positioning column is adjusted to the reference state and calibrated on-site before the equipment is put into production. The reference state refers to the position of the positioning column at the design reference height and the end posture meeting the factory or installation acceptance requirements. It is obtained by moving the positioning column to the design reference height and locking it under no workpiece load conditions, provided that the preset safety conditions are met. Then, the height and tilt of the end reference surface are inspected using standard measuring tools. The standard measuring tools can be set as a combination of measuring tools that can read the height and tilt. The inspection action can be set to read the data at multiple measuring points on the end reference surface and determine whether the preset allowable deviation is met, so as to ensure that the inspection criteria are executable and consistent.

[0070] When the acceptance result meets the preset allowable deviation, it is determined to be the baseline state; otherwise, the adjustment is repeated until it meets the requirement. The preset allowable deviation is given by the on-site process accuracy requirements and can be consistent with the subsequent attitude convergence threshold. For ease of unified management, preferably, the drift threshold can be set as a certain percentage of the preset allowable deviation, the preset running time can be consistent with the changeover cycle or maintenance cycle, and the preset load range can be determined by the mass range of the workpiece and the fixture, so that the source of the threshold has objective constraints rather than arbitrary settings.

[0071] The measurement direction of each grating ruler is its sensitive axis direction and is obtained through on-site calibration. On-site calibration refers to collecting grating readings no less than a preset number of times under a unified control cycle while the positioning column is in the reference state, and taking the stable segment statistical value as the calibration reading. At the same time, when the above-mentioned axial micro-displacement and micro-tilt actions are applied, the corresponding reading changes are recorded to confirm that the measurement direction expression and independent observation criteria are valid. The preset number of times is set to no less than a preset lower limit and locked with the version to reduce the influence of on-site vibration and electromagnetic interference on the readings. The stable segment statistical value is obtained by taking the median value within a time period in which the reading fluctuation does not exceed a preset fluctuation threshold within a control cycle of no less than a preset number of consecutive control cycles. The preset fluctuation threshold is jointly determined by the grating ruler displacement resolution and the on-site vibration and noise level, and is preferably taken as several times the displacement resolution.

[0072] The zero-position offset parameter is the offset between the actual reading of each grating ruler and the theoretical zero position when the positioning column is in the reference state. It is recorded in millimeters as the displacement unit and saved together with the measurement direction parameter in the same version. The aforementioned measurement direction, zero-point offset parameters, and spatial orientation parameters of the grating ruler in the geometric anchor system are all compiled with a unified standard and written into the parameter storage area of ​​the equipment control system, and version locking management is implemented. Version locking means that a unique version identifier is generated each time a parameter is established or updated, forming an unwriteable historical version record. The version record includes at least the version identifier, generation time, operator identifier, installation reference surface identifier, grating ruler identifier, measurement direction parameters, zero-point offset parameters, independent observation criterion results, preset allowable deviation configuration, drift threshold configuration, preset running time configuration, and preset load range configuration. The aforementioned version records are written into the evidence chain traceability area in chronological order by field and value and are associated with the operation record area. After writing, only new records are allowed to be added, and existing records are not allowed to be modified. Modification requirements are reflected in the new version record and associated with the previous version identifier. The writing timing can be set to write immediately after calibration acceptance is passed. When recalibration is triggered during operation, a new version is written and associated with the triggering reason, without overwriting the original version for rollback comparison, thereby meeting the traceability and review requirements.

[0073] To define "stable installation position without relative slippage" as a verifiable boundary, this embodiment defines the non-slippage acceptance criterion as follows: within a preset load range and preset operating time, the drift of the zero-position offset parameter does not exceed a preset drift threshold and no consecutive over-limit events occur. A consecutive over-limit event is defined as the drift exceeding the drift threshold in a preset number of consecutive drift tests. The drift detection criterion can be set by resetting the positioning column to the reference state under no-load conditions and with the positioning column locked, and reading the readings of each grating ruler. The difference is compared with the zero-position offset parameter of the corresponding version. Alternatively, it can be obtained by repeatedly resetting under the same attitude target and statistically analyzing the changes in the zero-position offset parameter. Drift detection can be set to be triggered once per fixed shift or fixed number of cycles, using the most recent reference state reading as the comparison baseline. After initial installation, a reference state reset can be performed once under no-load conditions and once under conditions close to the upper limit of the preset load range. If the zero-position offset drift meets the drift threshold, the installation is considered secure.

[0074] When drift exceeds the limit, recalibration is triggered and a new version is generated, while the original version is retained for rollback comparison. The trigger condition for rollback comparison can be set to the new version failing to meet the preset allowable deviation after a preset number of consecutive resets. After triggering, the historical version is loaded and the new version record is retained for traceability. To ensure on-site safety and compliance boundaries, the above calibration and reset actions are only allowed to be executed when preset safety conditions are met, and can be interlocked with the equipment's emergency stop and protection states. When the interlock conditions are not met, micro-displacement and micro-tilt actions are prohibited, and the corresponding version identifier and triggering reason are recorded, thus ensuring that safety constraints have clear execution rules and traceability.

[0075] Through the above-mentioned deployment, acceptance and parameter processing methods, the purpose of this step is to anchor the multi-source displacement sensing information to the same spatial reference frame, so that the attitude of the positioning column end has stable and repeatable observation conditions, and to provide a consistent basis for subsequent establishment of the constraint equation set of displacement to end attitude parameters based on the geometric anchor system, inversion to obtain axial displacement, pitch angle, roll angle and perform observability judgment.

[0076] The scope of application of this embodiment includes production cycle scenarios with continuous production and frequent model changes. The applicable boundary is that the grating ruler installation meets the independent observation criteria and meets the non-slip acceptance caliper during operation, and the parameter version is in a valid state. Under the premise of meeting this boundary, this method can be applied to positioning column devices with different structural forms, different size specifications and different process uses.

[0077] In a preferred embodiment, the displacement resolution of the grating ruler can be set to the micrometer level; the three grating rulers are preferably arranged in a combination of positions circumferentially spaced 120 degrees apart; the preset number of on-site calibrations can be set to several tens of times and the statistical value of the stable segment that meets the fluctuation threshold is selected; the drift threshold can be set to a certain percentage of the preset allowable deviation and the number of consecutive exceedances within the specified running time is zero; as an alternative, the grating rulers can also be arranged in a non-collinear combination at different heights of the positioning column. As long as the independent observation criterion is met under the geometric anchor system and the non-slip acceptance is passed, the technical effect equivalent to the circumferential arrangement can be achieved.

[0078] S2. Synchronously acquire the displacement of three grating rulers according to a unified control cycle, form an attitude observation vector, and add the same cycle identifier. The specific implementation is as follows:

[0079] After the equipment is put into operation, the control system synchronously collects the displacement of the three grating rulers according to a unified control cycle. This ensures that the multi-source displacement data formed by the three non-collinear grating rulers belongs to the same time slice in time and can be used as attitude observation vectors. The unified control cycle is a fixed scheduling rhythm within the control system. Its physical meaning is the common time reference for the control system to collect grating ruler data, allocate cycle identifiers, and call subsequent links. The setting of the unified control cycle is used to ensure the consistency of data collection under the upper limit of delay and meet the production line cycle requirements, so that the three displacement values ​​in the same cycle can correspond to the same cycle identifier.

[0080] Synchronous acquisition refers to the control system initiating acquisition to the grating acquisition unit within the same unified control cycle and registering the displacement of the three grating rulers in a predetermined order after returning, so that they logically belong to the same cycle; the control system and the grating acquisition unit communicate through an industrial bus, which is a field communication link used to transmit acquisition commands and displacement records between the control system and the acquisition unit. Its communication method can be set to periodic polling or event-triggered feedback, as long as the acquisition registration can be completed within the unified control cycle and the delay upper limit requirement is met.

[0081] To reduce the impact of communication delays and field disturbances on the consistency of displacement measurements, alignment processing is performed on the original displacement measurements before registration. The physical meaning of alignment processing is to align the displacement measurements of the three grating rulers to the reference time of the unified control cycle under the same cycle identifier, thereby eliminating the phase difference caused by the inconsistency of the acquisition start and end times. To ensure the reproducibility of the alignment processing, the grating acquisition unit attaches an acquisition time mark when transmitting each displacement measurement record. The acquisition time mark is a time mark relative to the unified time base of the control system and the granularity meets the scheduling requirements of the unified control cycle. The control system performs alignment of the displacement measurement records with the reference time of the unified control cycle. When the maximum difference between the acquisition time marks of the three displacement measurement records exceeds the preset alignment threshold, alignment processing is performed. If the difference does not exceed the preset alignment threshold, the records can be directly registered to reduce additional delays. The preset alignment threshold can be jointly determined by the unified control cycle and the industrial bus jitter level. The determination method of the preset alignment threshold is defined as taking the statistical upper bound of the stable segment under the reference state and adding a preset margin. The preset margin is subject to version locking constraints.

[0082] To improve the continuity and availability of attitude observation vectors, significant jump values ​​are suppressed. Significant jump values ​​refer to abrupt changes in displacement readings that exceed the reasonable range between adjacent unified control cycles. The reasonable range can be determined by converting the maximum allowable displacement change from the upper limit of the positioning column actuator's speed to the unified control cycle, combined with the noise level obtained from the statistics of the no-load stable segment. When the change of a certain displacement in an adjacent cycle exceeds the reasonable range, the reading of that cycle is marked as low confidence and replaced by the valid reading of the previous cycle or a smoothed reading, in order to avoid the propagation of jump readings to subsequent attitude inversion.

[0083] For short-term missing data, padding is performed within the boundary. A short-term missing data point refers to a situation where the displacement of a grating ruler is not obtained or the displacement record is incomplete within a single unified control cycle. The boundary refers to the allowed number of consecutive missing cycles and the maximum allowable displacement change range. The number of consecutive missing cycles is counted based on the number of consecutive cycles in which the grating ruler displacement record is not obtained. The padding method can be set to generate the missing cycle displacement using interpolation when the missing data does not exceed the boundary and the displacement change rate of adjacent valid cycles does not exceed a preset padding threshold. The preset padding threshold is defined as taking the upper bound of the stable segment statistics under the baseline state and representative operating conditions, adding a preset margin, and is locked with the version number. When the missing data does not exceed the boundary but does not meet the interpolation conditions, the valid reading of the previous cycle is retained and the low confidence mark is maintained. If the data exceeds the boundary, the cycle is marked as unavailable and a data acquisition retry or degradation strategy is triggered. The maximum allowable displacement change and noise level are used to form the sampling boundary for representative working conditions and stable segment statistics, thereby ensuring that the completion behavior has a unique field judgment caliber. The degradation strategy can be set to retain the attitude observation vector of the previous cycle and set the low confidence mark in the current cycle. After the data acquisition is restored and the integrity and consistency requirements are continuously met to reach the preset number of recovery cycles, the low confidence mark is removed and the normal acquisition link is restored. The preset number of recovery cycles is locked with the version.

[0084] The three displacement values, after being processed, are combined in the order of the grating rulers to form an attitude observation vector. The attitude observation vector is a set of three grating ruler displacement values ​​acquired and aligned within the same unified control cycle. Its physical meaning is to be used to establish the constraint relationship between the displacement values ​​and the end attitude parameters based on the geometric anchor system and to invert and obtain the set of observations for axial displacement, pitch angle and roll angle. To ensure cross-cycle traceability and cross-link consistency, the attitude observation vector is attached with the same cycle identifier. The cycle identifier is an identifier generated by the control system within the unified control cycle and bound to the acquisition record of that cycle. It is used for deduplication and traceability of acquisition records. The sequence identifier is the instruction number of the control instruction link. The two are established one-to-one association within the same unified control cycle to ensure that the acquisition records have a definite sequential relationship and can be deduplicated during transmission, retry and storage.

[0085] The period identifier is used to achieve idempotent multiplexing. Idempotent multiplexing means that for the same period identifier, the control system maintains consistency in the registration, storage, and retrieval of the acquired records, and there is no superposition effect due to repeated reception or registration. To this end, the control system uses the period identifier as a unique key to perform a deduplication strategy on the acquired records. When duplicate records with the same period identifier are received, only one record that meets the requirements of integrity and consistency is retained, and the remaining records are marked as duplicate records. Integrity means that all three displacement quantities under the same period identifier are complete and all have the acquisition time mark. Consistency means that the difference between the acquisition time marks of the three displacement quantities does not exceed the preset alignment threshold or meets the threshold after alignment processing, so as to ensure that subsequent attitude inversion and control decisions only refer to a single, determined version of the data.

[0086] To meet production line cycle time and resource constraints, the control system performs a limited number of retries for data acquisition timeouts. Data acquisition timeout refers to the failure to receive three complete displacement records within a preset delay limit within a unified control cycle. The preset delay limit can be set to no more than a certain percentage of the unified control cycle, and its value is determined by the industrial bus bandwidth and the concurrent load on site. Exceeding the preset delay limit indicates a data acquisition timeout and initiates the retry process. The limited number of retries refers to repeatedly requesting the same data acquisition request a preset maximum number of times within a single unified control cycle or within a consecutive preset retry window period. If the retry is successful, the success record is registered according to the cycle identifier and proceeds to the next step. If the retry still fails, the cycle is marked as unavailable and written into the log record to maintain consistency with the aforementioned version locking and evidence chain logging standards and to provide a basis for subsequent traceability.

[0087] The low-confidence marker and the cycle marker are written together into the operation record area of ​​the control system and associated with the evidence chain trace area. The operation record area is a storage area for storing the cycle marker, attitude observation vector, acquisition time marker and low-confidence marker in chronological order. The association method can be set to write the corresponding cycle marker and the current valid parameter version marker into the trace record at the same time, so that subsequent attitude inversion, control decision and retrospective call can locate the unique data record and version background under the same terminology.

[0088] The combination of acquisition, alignment, suppression, and completion described above enables this method to stably generate attitude observation vectors that can be used for attitude inversion even under continuous production and frequent model changes in field conditions, and to maintain the feasibility and traceability of the data link under communication jitter, short-term loss, and occasional jumps.

[0089] In a preferred embodiment, the unified control cycle can be set to milliseconds to adapt to high-speed cycle scenarios; the preset alignment threshold can be set to a small proportion of the unified control cycle to suppress phase differences; the maximum allowable displacement change can be obtained by converting the speed limit and the unified control cycle; the preset completion threshold can be set to a small proportion of the maximum allowable displacement change and combined with noise level correction; the boundary of the number of consecutive missing cycles can be set to a preset number of missing window cycles, which is jointly determined by the on-site process cycle and data integrity requirements and locked with the version number; the preset delay limit can be set to not exceed a certain proportion of the unified control cycle, and the number of retries can be set to not exceed a certain number to achieve a balance between real-time performance and reliability; the attitude observation vector and its cycle identifier are stored in the operation record area of ​​the control system in the form of fields and values ​​and are associated with the evidence chain trace area, so that subsequent attitude inversion, control decision and trace call have a stable and consistent data source.

[0090] S3. Based on the geometric anchor system, establish a set of constraint equations for the displacement to the end attitude parameters, invert to obtain the axial displacement, pitch angle, and roll angle, and perform observability checks on the equation condition number and residuals to output the effective attitude solution. The specific implementation is as follows:

[0091] On the actual production line, the control system, after generating the attitude observation vector and adding a period identifier, establishes a set of constraint equations from the displacement to the end attitude parameters based on the geometric anchor system. It then reads the attitude observation vector of the corresponding period using the period identifier as an index and substitutes it into the set of constraint equations to invert and obtain the axial displacement, pitch angle and roll angle to form the attitude parameter combination.

[0092] The constraint equations are a set of equations that characterize the consistency between the displacement of the three grating rulers and the attitude parameters of the positioning post under the geometric anchor system. Physically, they interpret the displacement changes of the three non-collinear grating rulers in their respective measurement directions as the overall rise, fall, and tilt changes of the positioning post end in space. The axial displacement represents the overall rise and fall of the positioning post along its own axis, the pitch angle represents the tilt of the positioning post in the first tilt direction, and the side tilt angle represents the tilt of the positioning post in the second tilt direction, which is orthogonal to the first tilt direction. The three equations together characterize the attitude of the positioning post end and constitute candidate results for the effective attitude solution.

[0093] An effective attitude solution refers to the combination of attitude parameters that can be used for control decisions after observability determination. The purpose of observability determination is to avoid attitude inversion instability caused by insufficient observation information, degraded readings, or excessive supplementation, which could lead to malfunctions in the execution unit. To this end, the control system simultaneously calculates the equation condition number and inversion residual for the inversion results within the same cycle and performs observability determination. The equation condition number is used to characterize the stability of the constraint equation set under the current geometric anchor and the current attitude observation vector. The inversion residual is used to characterize the degree of inconsistency after the attitude parameter combination is substituted back into the constraint relationship. A preset threshold is used to determine whether the above stability and inconsistency are within an acceptable range.

[0094] To ensure that the preset threshold has a unique and reproducible caliber, it is determined by the measured statistical results under the baseline state and representative operating conditions. The representative operating conditions include at least the no-load state and the state close to the upper limit of the preset load range. The control system collects attitude observation vectors for several consecutive preset statistical periods under each representative operating condition and completes inversion and index calculation. The statistical upper bound of the stable segment of the equation condition number and the inversion residual is extracted and a preset margin is added as the threshold. The stable segment refers to the time period within the above statistical period range, within a consecutive unified control period of no less than the preset statistical period, where the attitude observation vector is not marked with a low confidence mark and the acquisition time mark meets the alignment requirements. The statistical upper bound refers to the upper quantile or maximum value of the index value within the stable segment. The preset margin is used to cover the impact of tolerance fluctuations and minor disturbances on the index. The above threshold, geometric anchor parameters, and zero-position offset parameters are locked with the same caliber and included in the evidence chain, so that subsequent calls can locate the unique threshold version background.

[0095] The observability determination rule is defined as parallel determination, and observation degradation is determined if either exceeds the limit. That is, when the equation condition number exceeds the corresponding threshold or the inversion residual exceeds the corresponding threshold, the observation degradation of the period is determined. When neither of them exceeds the corresponding threshold, the valid attitude solution of the period is output. The trigger cause markers include at least the condition number exceeding the limit marker and the residual exceeding the limit marker, so as to avoid the ambiguity of the trigger cause.

[0096] To ensure that the control behavior under degraded conditions has a unique caliber, when observation degradation is determined, the control system retains the most recent valid attitude solution as the attitude solution for the current cycle and continues to provide this attitude solution to subsequent stages. However, an observation degradation mark is added to the attitude solution for this cycle, and updating the archived version of the valid attitude solution is prohibited to avoid unreliable inversion results replacing historical valid values. At the same time, observation degradation events are recorded to form a chain of evidence. Observation degradation events preferably include at least a cycle identifier, a trigger cause mark, an equation condition number determination result, an inversion residual determination result, a low-confidence mark state of the attitude observation vector, and a current valid parameter version identifier, so that tracing can locate a unique data record and version background.

[0097] To adapt to continuous operation and cycle constraints on site, the aforementioned acquisition link is used to continuously update the attitude observation vector during the period of the most recent valid attitude solution. When the recovery condition is met, a new valid attitude solution is automatically output. The recovery condition is defined as the recovery is triggered when the observability judgment condition is met continuously for a preset number of recovery cycles. The continuous satisfaction is counted according to the continuous unified control cycle. The preset number of recovery cycles is jointly determined by the production line cycle and the disturbance level and locked with the version number. After the recovery is triggered, the valid attitude solution is updated with the inversion result of the first cycle that meets the condition and the observation degradation mark is removed. Thus, the closed-loop operation of degradation suppression and automatic regression is achieved without the introduction of manual intervention.

[0098] This step is applicable to normal assembly, tolerance fluctuations, and minor disturbance scenarios. The applicable boundary is that the grating ruler arrangement meets the attitude observation conditions and the geometric anchor parameters are in the valid version state. Under the premise of meeting this boundary, equivalent replacement forms, including circumferential non-collinear arrangements or non-collinear arrangements at different heights, can be executed.

[0099] In a preferred embodiment, the equation condition number threshold can be set to the range of values ​​after adding a preset margin to the statistical upper bound of the stable segment under the baseline state and representative operating conditions. The inversion residual threshold can be set to the same order of magnitude as the preset allowable deviation or a certain proportion thereof. The recovery cycle number can be set to a preset number of unified control cycles to achieve a balance between real-time performance and robustness. The preset recovery cycle number is jointly determined by the production line cycle time and disturbance level and locked with the version number. When reading degradation, excessive short-term missing data filling, or observation degradation caused by communication jitter occur, the system can reliably maintain the most recent valid attitude solution and record it. After continuously meeting the observability judgment conditions, it will automatically recover and output a new valid attitude solution.

[0100] S4. Compare the effective attitude solution with the target attitude to obtain the attitude error. Decompose the error into bias, load, and clearance terms to generate corresponding compensation weights. The specific implementation is as follows:

[0101] On the actual production line, after obtaining the effective attitude solution, the control system compares the effective attitude solution with the target attitude to obtain the attitude error. The target attitude is the spatial state of the positioning column end required by the current process step. Its physical meaning is the reference height and tilt state that the end reference surface should be in under this process step. The target attitude is determined by the process recipe, the work station task sheet or the process parameter table built into the control system, and is associated with the cycle identifier after being mapped by the process step identifier to obtain the working condition mark, so that the attitude error calculation caliber is consistent and traceable within the same cycle. The attitude error is used to characterize the degree of deviation of the end attitude from the target, preferably including the deviation in the axial displacement direction and the deviation in the pitch and roll directions, so that the subsequent control quantity allocation can simultaneously cover the lifting and tilting adjustments.

[0102] To adapt to the nonlinear effects of off-center loading, temperature drift, and guide wear, the control system performs error decomposition on the attitude error. Error decomposition is defined as attributing the attitude error to offset, load, and clearance terms, and generating corresponding compensation weights. The offset term refers to the attitude error component caused by drift of the zero-position offset parameter, minor changes in the installation reference surface, or slow offset of the geometric anchor parameters. Its discrimination criterion is defined as follows: within the observation window with the same target attitude and unchanged working condition markings, the attitude error changes slowly in the same direction within the stable segment, and the rate of change does not exceed a preset offset change rate threshold. The load term refers to the attitude error component caused by changes in the workpiece mass, clamping force, or forces during assembly, leading to changes in the force state of the positioning column. Its discrimination criterion is defined as follows: when the working condition markings indicate the entry into the load change stage... When a change in the load status mark is detected, the attitude error shows a step or slope change synchronized with the load status mark in the observation window and exceeds the preset load sensitivity threshold; the gap item refers to the attitude error component caused by guide pair gap, reverse gap, or start-stop switching. Its discrimination criterion is defined as when a direction switching event or start-stop event is detected, the attitude error shows a hysteresis-type abrupt change before and after the direction switching and its amplitude exceeds the preset gap hysteresis threshold. The direction switching event is defined as the movement direction of the execution unit control command changing from forward to reverse or from reverse to forward. The start-stop event is defined as the execution unit control command changing from a moving state to a stopped state or from a stopped state to a moving state. The above events are marked by the control system when the control command is issued and associated with the cycle identifier.

[0103] To ensure consistency in statistical standards across the entire system, the observation window and stable segment for attitude error adopt the same configuration as the unified control cycle. The observation window is a time window consisting of several unified control cycles with consecutive preset observation window periods. The number of observation window periods is determined by the process cycle time and statistical requirements and locked with the version number. The stable segment is a continuous periodic segment within the observation window where the low-confidence marker is not set and the acquisition time marker meets the alignment requirements. At the end of each observation window, the control system calculates the error characteristic quantities for the bias, load, and clearance items respectively and updates the compensation weights accordingly. The error characteristic quantities are defined as the mean amplitude, variation amplitude, and synchronization degree with the event marker or load status marker of the corresponding error component within the stable segment. The synchronization degree is used to characterize the correspondence between the error component and the direction switching event, start / stop event, or load status change.

[0104] The compensation weight is a dimensionless coefficient that measures the contribution of different error terms under the current operating condition. Its purpose is to convert the error decomposition results into a weighting basis for subsequent control variable allocation. In order to make the compensation weight update uniquely implementable, the control system first denegates the three types of error characteristics and obtains candidate weights respectively. Then, it normalizes the sum of the three types of candidate weights to obtain the compensation weight, so that the sum of the three types of compensation weights is kept to one. When the sum of the candidate weights is zero, the compensation weight is set to keep the compensation weight of the previous observation window unchanged and set the weight freeze flag to avoid uncertain updates due to the lack of significant error characteristics.

[0105] To suppress and control oscillations, the compensation weights are smoothly adjusted within a limited range. The limited range refers to the allowable value range of each type of compensation weight and the maximum allowable adjustment range within a single observation window. The smoothing adjustment is defined as taking the normalized compensation weight as a candidate new weight and merging it with the compensation weight of the previous observation window using a fixed smoothing coefficient to obtain the updated weight. The fixed smoothing coefficient remains unchanged within the same parameter version and is subject to version locking constraints. Subsequently, the updated weight is subject to a limit. The limit is defined as truncating to the boundary value when the updated weight exceeds the allowable value range, and truncating to the value corresponding to the maximum adjustment range when the change of the updated weight relative to the previous observation window exceeds the maximum adjustment range. A weight limit marker is attached for traceability.

[0106] Preferably, in assembly conditions with significant load changes, the assembly condition is mapped from the condition marker and its determination criteria are defined as the process step identifier being in the assembly load stage or the load status marker changing within the observation window. The load item compensation weight can be set higher than the bias item compensation weight so that the control quantity allocation prioritizes offsetting the attitude deviation caused by the load change. In changeover conditions with frequent starts and stops, the changeover condition is mapped from the condition marker and its determination criteria are defined as the process step identifier being in the changeover stage or the number of direction switching events and start / stop events within the observation window exceeding a preset start / stop frequency threshold. The gap item compensation weight can be set to prioritize participation in subsequent control quantity allocation to reduce the impact of hysteresis error on attitude stability. The priority participation is defined as applying a constraint of not less than a preset lower limit to the weighting coefficient corresponding to the gap item during subsequent control quantity allocation and prioritizing the satisfaction of the constraint under the limiting condition.

[0107] To ensure consistency and traceability during reset calls, compensation weights, error decomposition rules, observation window configurations, stable segment judgment criteria, nonnegativity and normalization rules, weight freeze marking rules, fixed smoothing coefficients, limited ranges, and related threshold parameters are locked along with the version number and written into the fixed configuration. The fixed configuration is a storage area used by the control system to store traceable control configurations. It includes at least the version identifier, generation time, applicable operating condition marker, target attitude source identifier, and the values ​​of the aforementioned weights and threshold parameters, and is associated with the evidence chain traceability area, so that a unique configuration version background can be located when called under the same or similar operating conditions.

[0108] This step is applicable to normal assembly, tolerance fluctuations, and minor nonlinear disturbances caused by off-center loading, temperature drift, and guide wear. The applicable boundary is that the effective attitude solution is in an observable state and the target attitude source is valid. Under the premise of satisfying this boundary, the attitude error decomposition and compensation weight update can converge stably and provide a basis for subsequent control variable allocation.

[0109] S5. Based on the attitude response mapping, the attitude error is converted into the control quantity of the execution unit. Travel constraints, rate constraints, and synchronization constraints are added to solve the control instruction set, which is then issued for execution. The specific implementation is as follows:

[0110] On the actual production line, after the control system completes the attitude error calculation and obtains the compensation weight, it converts the attitude error into the control quantity of the execution unit according to the attitude response mapping and solves the control instruction set so as to drive the positioning column to complete the coordinated adjustment of lifting and tilting in a unified control cycle.

[0111] The execution unit is a drive mechanism that drives the positioning column to move or make fine adjustments. Its physical meaning is that it is a power source that can generate axial displacement and attitude change of the positioning column under the action of control system commands. The number and arrangement of execution units are matched with the structural form of the positioning column. Preferably, it can be multiple distributed drive mechanisms to support the precise attitude adjustment of the end posture. The execution unit control quantity is a quantitative command quantity used to characterize the action amplitude and action direction that each execution unit should achieve in the current cycle. Its dimension corresponds to the drive form of the execution unit and is consistent with the caliber of the control system parameter table.

[0112] Attitude response mapping, under the constraints of the current geometric anchor system and equipment structural parameters, maps the deviations of attitude error in the three components of axial displacement, pitch angle, and roll angle to the corresponding relationships of control quantities of each actuator. Its physical meaning is to convert "attitude adjustment requirements" into a consistent rule for "actuator action allocation". To ensure the reproducibility of the mapping relationship, the attitude response mapping is version-locked along with the geometric anchor system parameters and actuator arrangement parameters. It can be set to obtain response characteristics through controlled micro-displacement and micro-tilt actions under the reference state and representative working conditions to form a mapping parameter table. The mapping parameter table at least includes the correspondence parameters between attitude error components and actuator action components, as well as the applicable working condition flags and is associated with the evidence chain traceability area. The calling rule of the mapping parameter table is defined as the control system matching the applicable working condition flags to select the current mapping parameter table. If no match is found, the default mapping parameter table is called and the mapping degradation flag is set. At the same time, the current working condition flag, the default mapping version identifier, and the triggering reason are recorded for traceability.

[0113] The control system weights the attitude error according to the compensation weight and then inputs it into the attitude response mapping to obtain the control quantity of the candidate execution unit. Under the constraint conditions, the control instruction set is obtained by solving the control instruction set. The control instruction set is a set of coordinated actions of multiple execution units generated and issued under the constraint conditions. Its physical meaning is the combination of executable actions given to each execution unit in the same cycle, so that the attitude of the end of the positioning column advances in a closed loop along the target attitude direction.

[0114] To ensure that the attitude adjustment action always stays within the mechanical capacity and safety boundaries of the equipment, stroke constraints, rate constraints, and synchronization constraints are added during the solution process. Stroke constraints limit the single adjustment range from the remaining available stroke of the actuator to the preset safety boundary. Rate constraints limit the speed of change of the actuator's action from the preset maximum rate and meet the on-site cycle time and delay limits. Synchronization constraints limit the relative deviation between multiple actuators from the preset synchronization deviation threshold to avoid excessive stretching of the end effector. The above constraint parameters are jointly determined by the rated capacity of the actuator, the structural strength of the positioning column, and process safety requirements, and are locked and written into the fixed configuration along with the version number. The fixed configuration and the aforementioned compensation weight configuration maintain the same traceability caliber.

[0115] To ensure a unique understanding of the solution rules, the control system sequentially performs constraint checks and amplitude limiting on the control quantities of candidate execution units. When any execution unit control quantity fails to meet the travel or rate constraints, the control quantity is truncated to the corresponding constraint boundary and an amplitude limiting flag is set. When the relative deviation of multiple execution units exceeds the synchronization constraint, the control quantities of the relevant execution units are coordinated and converged according to the synchronization convergence rule, and a synchronization correction flag is set. The synchronization convergence rule is defined as follows: while keeping the attitude adjustment direction unchanged, the control quantities of execution units that contribute significantly to the relative deviation are reduced proportionally, and the remaining execution units are coordinated proportionally to ensure that the relative deviation of each execution unit after adjustment does not exceed the synchronization deviation threshold, and the number of iterations for synchronization convergence does not exceed the preset iteration limit. When the synchronization constraint cannot be met within the preset iteration limit, or when a control instruction set that satisfies all constraints cannot be formed after constraint convergence, the constraint is determined to be unsatisfactory. The most recent valid attitude solution is retained, and the constraint unsatisfactory flag is set, and a protection strategy is entered to avoid forced action under infeasible conditions.

[0116] After the control command set satisfies the constraints, it is issued and executed through the industrial bus. Idempotency and deduplication strategies are implemented using sequence identifiers. The sequence identifier is the instruction number generated by the control system and associated with the cycle identifier. It is used to ensure that the control command set within the same cycle has a definite order and can be deduplicated during transmission, retry, and readback. The idempotency and deduplication strategy is defined as follows: the control system registers the issued records using the sequence identifier as a unique key. When a readback confirmation or duplicate issuance request with the same sequence identifier is received, only one valid record is retained and execution of the already executed record is prohibited. At the same time, the execution unit executes the idempotency strategy according to the sequence identifier. That is, when an already executed sequence identifier is received, the action is not executed again, but only confirmation information is sent back, thereby avoiding the superposition of duplicate actions caused by communication jitter.

[0117] To meet the constraints of on-site resources and the upper limit of control delay, the control system triggers a limited number of retries for timeouts or inconsistencies in readbacks. Timeout refers to the failure to receive execution confirmation within the preset delay limit. Inconsistency in readbacks refers to the sequence identifier returned by the execution unit being inconsistent with the sequence identifier registered by the control system, or the returned control quantity being inconsistent with the registered control quantity outside the preset readback tolerance. The preset readback tolerance is jointly determined by the quantization resolution of the execution unit and the transmission accuracy of the industrial bus. The limited number of retries refers to repeatedly sending or requesting confirmation for the control command set with the same sequence identifier within a preset maximum number of times. If the retry is successful, the closed-loop confirmation is completed with the same sequence identifier and the next cycle continues. If the retry still fails, the sending event is recorded and the protection strategy is activated.

[0118] The protection strategy is a set of conservative control actions to be taken when safety and compliance boundaries are exceeded, constraints are not met, or retry fails. The triggering sources for safety and compliance boundaries include at least mechanical limit trigger flags, safety interlock trigger flags, emergency stop trigger flags, industrial bus communication disconnection flags, and control delay upper limit continuous exceedance flags. The protection strategy includes at least maintaining the most recent valid attitude solution as the attitude basis for the current cycle, restricting the issuance of new control quantities, and writing relevant events, sequence identifiers, cycle identifiers, and current valid parameter version identifiers into the evidence chain traceability area to ensure stable operation and traceability within the control delay upper limit under continuous production and frequent changeover cycles.

[0119] In a preferred embodiment, the travel constraint can be set to not exceed a certain proportion of the remaining available travel and be limited by a preset safety boundary; the rate constraint can be set to not exceed a certain proportion of the rated maximum rate of the execution unit; the synchronization deviation threshold can be set to be of the same order of magnitude as the attitude convergence threshold or a certain proportion thereof; the sequence identifier can be set to increase with the unified control cycle and correspond one-to-one with the cycle identifier; the preset delay upper limit can be set to not exceed a certain proportion of the unified control cycle; the preset iteration upper limit can be set to a preset number of iterations; and the number of retries can be set to not exceed a preset number of retries. The above numbers are determined by the on-site requirements for balancing real-time performance and reliability and are locked with the version number to achieve a balance between real-time performance and reliability.

[0120] S6. The process of acquiring, inverting, and solving data is repeated until the attitude error meets the convergence threshold within a preset number of consecutive cycles. The geometric anchor parameters, compensation weights, and control constraint configurations for this working condition are then fixed for reset and recall. Specifically, the implementation is as follows:

[0121] On the actual production line, the control system cyclically executes attitude observation vector generation, effective attitude solution inversion, attitude error calculation, compensation weight update, and control command set solution and issuance in a unified control cycle until the attitude error meets the convergence threshold within a preset number of consecutive cycles, triggering a stop judgment and completing the attitude adjustment closed loop.

[0122] The preset number of consecutive cycles is used to eliminate the interference of transient fluctuations on the stop judgment. Its physical meaning is that the attitude error must continuously meet the convergence threshold within a unified control cycle of the preset number of consecutive cycles before the attitude adjustment can be judged as complete. The continuous counting caliber is defined as counting cycle by cycle of adjacent unified control cycles, and counting is restarted if the condition is not met in any cycle. The convergence threshold is the judgment threshold that characterizes the end attitude to meet the process accuracy requirements. Its physical meaning is that allowable ranges are set for axial displacement error, pitch angle error and roll angle error respectively, and they must be met simultaneously when the stop judgment is made. The convergence threshold is determined by the process accuracy requirements and is consistent with the target attitude parameter caliber. It is also locked and written into the fixed configuration along with the version number to ensure consistency across batches.

[0123] To ensure consistency in stopping decisions, the control system reads the attitude error and valid attitude solution status flags at the end of each unified control cycle. When the valid attitude solution passes the observability test and none of the attitude error components exceed the convergence threshold, the continuous count is incremented; if the conditions are not met, the count is reset and adjustment continues. When the continuous count reaches a preset number of cycles, the attitude adjustment is determined to be complete, the issuance of control command sets stops, and the current valid attitude solution is retained as the basis, with a completion flag written.

[0124] The completion marker is a status marker used to indicate that the attitude adjustment closed loop has reached the stopping condition under this working condition. It is bound to the cycle identifier, working condition identifier, and current valid parameter version identifier and written together into the operation record area. The operation record area is associated with the evidence chain traceability area to ensure traceability consistency. To avoid misjudging completion in the event of observation degradation or execution anomaly, if an observation degradation event, constraint unsatisfaction marker, mechanical limit trigger marker, safety interlock trigger marker, emergency stop trigger marker, industrial bus communication loss marker, or communication retry exhaustion marker occurs during the closed loop process, the most recent valid attitude solution is retained as the attitude basis for this cycle and the protection strategy is entered. At the same time, the cause of degradation and the associated cycle identifier are recorded, so that both the stop judgment and the anomaly handling have traceability.

[0125] After the attitude adjustment is completed, the control system solidifies the geometric anchoring parameters, compensation weights, and control constraint configurations of this working condition for reset and recall. Solidification is the process of writing the set of parameters directly related to the stable operation of this working condition into the configuration storage area of ​​the equipment control system in an overwriteable manner and associating it with the evidence chain trace area. This working condition refers to the production line working state corresponding to the working condition identifier obtained by mapping the process step identifier. The working condition identifier is used as an index for locating and loading the solidified configuration.

[0126] To ensure the solidified content has a unique definition, it must include at least the version identifiers for the geometric anchor parameters, zero-position offset parameters, attitude response mapping parameter table, compensation weights and their update rules, travel constraint parameters, rate constraint parameters, synchronization constraint parameters, preset number of consecutive cycles, convergence threshold, and related threshold configurations. It must also include a solidification timestamp, generation time, applicable operating condition identifier, and generation reason marker. The solidification writing method is defined as append-only, retaining historical versions. Update requirements are reflected by generating a new version and associating it with the previous version identifier to ensure consistency between rollback and retrospection.

[0127] The fixed configuration is stored in the configuration storage area of ​​the equipment control system. This storage area is dedicated to storing traceable control configurations and prioritizes loading the corresponding configurations for closed-loop adjustment through the operating condition identifier index. This process ensures rapid loading in each adjustment and guarantees consistency between different batches. When entering the same operating condition again, the fixed configuration calling rule is set to prioritize matching the latest valid version corresponding to the operating condition identifier and loading its parameter set. The valid version is defined as a version that has not been invalidated and has passed integrity verification. The integrity verification includes at least the complete version identifier, complete fixed fields, and consistency with the corresponding parameter version identifier. If no match is found, it reverts to the default version and sets the configuration downgrade flag. At the same time, the downgrade reason, the default version identifier, and the current operating condition identifier are written to the evidence chain traceability area.

[0128] To meet the requirements of on-site inspection and continuous improvement, the control system generates operation records for the attitude adjustment process and results, and associates them with cycle and version identifiers. The operation records include at least the attitude adjustment start and end time, reset time, attitude stability rate, number of out-of-limits, number of retries, number of observed degradations, number of constraints that cannot be met, and number of protection strategies triggered. The statistical definition of attitude stability rate is defined as the percentage of cycles that meet the convergence threshold and whose effective attitude solutions are in the observability judgment state within a specified sample size. The specified sample size is the number of consecutive cycles counted according to the unified control cycle or the number of records counted according to the process batch. Its value is determined by the on-site inspection system and locked with the version number. The above operation records are written into the evidence chain traceability area for subsequent traceability and comparison, thereby forming a sustainable improvement closed loop for continuous production and frequent changeover cycles.

[0129] All calculations involved in the embodiments are dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0130] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0131] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center containing one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0132] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0133] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0134] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0135] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0136] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0137] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0138] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for precise attitude control of a positioning post based on grating ruler feedback, characterized in that, include: S1. Arrange three non-collinear grating rulers at different positions around the positioning column, establish a geometric anchor system with the installation reference plane as the origin, and write the measurement direction and zero offset parameters of each grating ruler. S2. Synchronously collect the displacement of three grating rulers according to a unified control cycle, form an attitude observation vector and add the same cycle identifier; S3. Based on the geometric anchor system, establish a set of constraint equations from the displacement to the end attitude parameters, invert to obtain the axial displacement, pitch angle, and roll angle, and perform observability determination on the equation condition number and residuals to output the effective attitude solution. S4. Compare the effective attitude solution with the target attitude to obtain the attitude error composed of the deviation in axial displacement direction, the deviation in pitch angle direction, and the deviation in roll angle direction. Within a preset observation window, the attitude error is attributed to the offset term, the load term, and the clearance term. Based on the error characteristic quantity of each error component within the preset observation window, the corresponding compensation weight is generated after non-negation and normalization. Under the constraints of a fixed smoothing coefficient and a limited range, the compensation weight is smoothed and limited. S5. Attitude response mapping is to map the deviation relationship of attitude error in the three components of axial displacement, pitch angle and roll angle to the corresponding relationship of each execution unit control quantity under the current geometric anchor system and equipment structure parameter constraints. The mapping parameter table includes at least the corresponding relationship parameters between attitude error components and execution unit action components, as well as the applicable working condition mark and is associated with the evidence chain trace area. The attitude error is weighted according to the compensation weight and then entered into the attitude response mapping to obtain the candidate execution unit control quantity. S6. Repeatedly perform acquisition, inversion, and solution until the attitude error meets the convergence threshold within a preset number of consecutive cycles, and solidify the geometric anchor parameters, compensation weights, and control constraint configurations for this working condition for reset and recall.

2. The method for precise attitude control of a positioning post based on grating ruler feedback according to claim 1, characterized in that, S1 includes: Three optical grating rulers are installed at different positions around the positioning column. The measurement directions of the three optical grating rulers are not collinear with each other under the geometric anchoring system with the installation reference surface as the origin. The measurement direction parameters and zero-point offset parameters of each grating ruler are determined and written into the field through calibration. The geometric anchor system includes the zero-position definition of the mounting reference surface and the spatial orientation relationship of each grating ruler relative to the mounting reference surface.

3. The method for precise attitude control of a positioning post based on grating ruler feedback according to claim 2, characterized in that: When the positioning column is in the reference state, the readings of each grating ruler are collected according to a unified control cycle; The zero-point offset parameter is defined as the offset between the actual reading of each grating ruler under the reference state and the theoretical zero point. Within a preset load range and preset runtime, the drift of the zero-point offset parameter is detected. When the drift exceeds the drift threshold, recalibration is triggered to generate a new version. The calibration parameters and drift threshold configurations are written to the parameter storage area of ​​the equipment control system in the form of version records.

4. The method for precise attitude control of a positioning column based on grating ruler feedback according to claim 1, characterized in that, S2 include: The control system synchronously acquires the displacement of three grating rulers through the industrial bus according to a unified control cycle, and aligns the displacement to the reference time of the unified control cycle based on the acquisition time marker returned by the grating acquisition unit. Alignment processing is performed when the difference in acquisition time exceeds a preset alignment threshold. When there is a short-term missing component and the preset completion threshold and preset missing window period limit are met, the missing displacement is generated by interpolation. Set a low confidence flag when a significant jump value is detected; When performing interpolation on short-term missing data, maintain a low-confidence marker and write the attitude observation vector, period identifier, and low-confidence marker into the run record area and associate them with the evidence chain trace area.

5. The method for precise attitude control of a positioning column based on grating ruler feedback according to claim 1, characterized in that, S3 include: The control system establishes a set of constraint equations that relate the displacement of the three grating rulers to the axial displacement, pitch angle, and tilt angle of the positioning column based on the geometric anchor system. The attitude observation vector is read by the period identifier and substituted into the constraint equation system to obtain the attitude parameter combination. The equation condition number and inversion residual are calculated and compared with the corresponding preset threshold. When any index exceeds the corresponding preset threshold, retain the most recent valid attitude solution and set an observation degradation flag; When the number of equation conditions and the inversion residual do not exceed the corresponding preset threshold within a consecutive preset number of recovery cycles, the effective attitude solution is updated and the observation degradation marker is cleared. The observation degradation event and the effective parameter version identifier are written into the operation record area and associated with the evidence chain trace area.

6. The method for precise attitude control of a positioning column based on grating ruler feedback according to claim 1, characterized in that, S4 include: After obtaining an effective attitude solution, the control system calculates the attitude error through the target attitude and decomposes the error into three attitude error components—bias term, load term, and clearance term—within a preset observation window. The error characteristic quantities are calculated separately and then nonnegated and normalized to obtain the compensation weight. The compensation weights are smoothly adjusted under the constraints of a fixed smoothing coefficient and a limited range. The compensation weights are frozen under the freeze condition and restricted under the row limit condition. Compensation weights, error decomposition rules, observation window configurations, and related threshold parameters are generated and solidified into configurations with version identifiers. These are then stored along with applicable working condition tags and target attitude source identifiers to ensure version consistency and ease of traceability when called under different working conditions.

7. The method for precise attitude control of a positioning column based on grating ruler feedback according to claim 1, characterized in that, S5 include: The control system acquires the attitude error and weights the attitude error according to a preset compensation weight; By mapping the attitude response, the attitude error is converted into the control quantity of the execution unit, and the control instruction set is solved to drive the positioning column to complete the lifting and tilting adjustment within a unified control cycle. The control quantity of each execution unit represents the amplitude and direction of the action it should achieve in the current cycle, ensuring that the executed action meets the target posture requirements; During the control process, stroke, speed, and synchronization constraints are incorporated to ensure that the amplitude, speed, and relative deviation of each execution unit do not exceed the set limits.

8. The method for precise attitude control of a positioning post based on grating ruler feedback according to claim 7, characterized in that: The control system performs constraint verification and amplitude limiting on the control quantities of candidate execution units to ensure that the action amplitude, rate and relative deviation of each execution unit meet the set constraints. If the constraints are not met, adjust the control quantity and perform idempotency and deduplication processing on the control instruction set to ensure that the control instructions are consistent during transmission, retry, and readback, and avoid repeated execution.

9. The method for precise attitude control of a positioning post based on grating ruler feedback according to claim 1, characterized in that, S6 include: The control system uses a unified control cycle as its rhythm to cyclically execute attitude observation vector generation, effective attitude solution inversion, attitude error calculation, compensation weight update, and control command set solution and issuance until the attitude error meets the convergence threshold within a preset number of consecutive cycles, triggering a stop judgment and completing the attitude adjustment closed loop. At the end of each unified control cycle, the control system reads the attitude error and the status flag of the effective attitude solution. When all components of the attitude error do not exceed the convergence threshold, the continuous count is increased. When the continuous count reaches the preset number of cycles, the posture adjustment is determined to be complete and the issuance of new control command sets is stopped; After the attitude adjustment is completed, the geometric anchor parameters, compensation weights and control constraint configurations for this working condition are fixed, and the corresponding fixed configurations are loaded through the working condition identification index to enter closed-loop adjustment; The control system generates operation records for the attitude adjustment process and results, and associates them with cycle and version identifiers for subsequent traceability and continuous improvement.

Images