A method, device and medium for self-calibration of a press-fit process based on quality detection

Abstract

The application discloses a kind of based on quality detection's press mounting process self-calibration method, equipment and medium, it is related to press mounting process control technical field, including loading process parameter set, execute calibration press mounting height measurement cycle, generate summary sample set, generate process record and detection record, write in traceability queue;Listening to traceability queue, identify detection record, obtain alignment reference package, trace process record, execute validity screening, execute discretization processing, calculate alignment code distance, obtain candidate set;Integrity check is executed to candidate set, calculate physical consistency score, obtain candidate pairing score table, form confirmation pairing relationship, obtain mismatch self-checking mark;Gate determination is carried out, obtain the passing mark confirmation pairing relationship set, form deviation sample set, calculate calibration increment and update termination displacement setting value, obtain updated monotonic table, improve pairing certainty, reduce mismatch propagation, improve press mounting quality consistency.

Description

Technical Field

[0001] This invention relates to the field of press-fitting process control technology, and in particular to a self-calibration method, equipment and medium for press-fitting process based on quality inspection. Background Technology

[0002] In the assembly process, press fitting is often used for interference fits of parts such as bearings and sleeves. Typically, an industrial control device loads parameters such as the target equivalent press-in amount and the set value of the termination displacement according to the process document. During the press fitting cycle, the press-in force and displacement data are collected to form a process record. At the quality inspection station, the end face of the part is measured at multiple points to obtain inspection records such as the average end face height and the end face range. The press fitting parameters are calibrated and verified to ensure stable control of assembly quality.

[0003] However, conventional methods still have room for further optimization. On the one hand, conventional methods usually use timestamps, fixed sequences, or workstation cycles as the basis for linking process records and inspection records. In scenarios with multiple workstations running in parallel and workpieces queuing, temporal drift between records may lead to mispairing, resulting in unclear sources of calibration samples and affecting the reliability and stability of subsequent parameter updates. On the other hand, if the termination displacement setpoint or monotonic table parameters are updated directly based on deviations in the absence of consistency constraints and gating screening, they are easily affected by abnormal samples, causing amplified parameter fluctuations and thus affecting the stability of the calibration chain. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a self-calibration method for press-fitting process based on quality inspection to solve the problems of easy mismatch in process record quality inspection record pairing and decreased stability of calibration parameter updates in the prior art.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a self-calibration method for press-fitting process based on quality inspection, comprising: loading a set of process parameters; executing a calibration press-fitting height measurement cycle; generating a summary quantity sample set; calculating the discrete step length and dimensional quantity; generating an alignment parameter set; performing a press-fitting cycle; generating process records and inspection records; and writing them into a traceability queue; monitoring the traceability queue; identifying inspection records; obtaining an alignment reference package; tracing back the process records; performing validity screening; extracting a set of valid process records; performing discretization processing; calculating the alignment code distance; and obtaining a candidate set; performing integrity verification on the candidate set; calculating a physical consistency score; obtaining a candidate pairing score table; determining the optimal pairing; forming a confirmed pairing relationship; and obtaining a mismatch self-inspection mark; performing gating judgment based on the mismatch self-inspection mark value; obtaining a set of confirmed pairing relationships through the mark; forming a deviation sample set; calculating the calibration increment and updating the termination displacement set value; updating the monotonic table; and obtaining the updated monotonic table.

[0007] As a preferred embodiment of the self-calibration method for press fitting process based on quality inspection described in this invention, the generation of the alignment parameter set includes: reading the process file, obtaining the process parameter set, executing the calibration press fitting height measurement cycle, collecting the press-in force displacement sequence and the measurement point height sequence, determining the end face reference height, fixing the measurement point configuration, calculating the equivalent press-in amount, and generating a summary quantity sample set; based on the summary quantity sample set, calculating the deviation length and dimensional quantity, generating a monotonic table, determining the alignment code threshold and mismatch threshold, writing them into the parameter area, and generating the alignment parameter set.

[0008] As a preferred embodiment of the self-calibration method for press-fitting process based on quality inspection described in this invention, the step of writing to the traceability queue includes: generating process records in the press-fitting cycle and generating inspection records in the height measurement cycle according to the alignment parameter set and the end face reference height; and writing the process records and inspection records to the traceability queue.

[0009] As a preferred embodiment of the self-calibration method for press-fitting process based on quality inspection described in this invention, the step of obtaining the alignment reference package includes: listening to the traceability queue write event, identifying the inspection record, locking the inspection record index, performing discretization processing on the equivalent press-in amount and end face range, generating the inspection alignment code, associating the inspection record index, and obtaining the alignment reference package.

[0010] As a preferred embodiment of the self-calibration method for press-fitting process based on quality inspection described in this invention, the step of obtaining the candidate set includes: reading process records back along the traceability queue according to the alignment reference package, extracting the average pressing force and press-fitting energy, performing validity screening, and obtaining a set of valid process records; performing discretization processing on the average pressing force and press-fitting energy according to the set of valid process records and the alignment reference package, calculating the alignment code distance, and generating a candidate set.

[0011] As a preferred embodiment of the self-calibration method for press-fitting process based on quality inspection described in this invention, the step of obtaining the mismatch self-inspection mark includes: performing integrity verification on the candidate set, obtaining the candidate set that passes the verification, reading the monotonic table from the parameter area, and generating a scoring calculation context; based on the scoring calculation context, combining the candidate set, traversing the candidate pairs, calculating the physical consistency score, and obtaining the candidate pairing score table; performing a minimum value search in the candidate pairing score table, determining the optimal pair, binding the process record index and the detection record index, forming a confirmed pairing relationship, performing mismatch self-inspection judgment, and generating a mismatch self-inspection mark.

[0012] As a preferred embodiment of the self-calibration method for press-fitting process based on quality inspection described in this invention, the step of forming a deviation sample set includes: receiving mismatch self-inspection tags and confirming pairing relationships, performing gating judgment, writing blocking tags for suspected mismatches, writing passing tags through mismatch self-inspection, obtaining a set of confirmed pairing relationships through passing tags; and reading the equivalent press-in amount of the detection record and the target equivalent press-in amount based on the set of confirmed pairing relationships through passing tags, calculating the deviation value, and obtaining the deviation sample set.

[0013] As a preferred embodiment of the self-calibration method for press-fitting process based on quality inspection described in this invention, the step of obtaining the updated monotonic table includes: generating a calibration increment based on the deviation sample set, performing an update threshold limit, deducting the update termination displacement setting value and writing it back to the parameter area, and obtaining the updated termination displacement setting value; collecting the equivalent press-in amount according to the pairing relationship set confirmed by marking and the updated termination displacement setting value, updating the monotonic table, performing monotonicity correction and writing it back to the parameter area, and obtaining the updated monotonic table.

[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the self-calibration method for press-fitting process based on quality inspection as described in the first aspect of the present invention.

[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the self-calibration method for press-fitting process based on quality inspection as described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: by generating an alignment parameter set and establishing a traceability queue, process records and inspection records can be traced and associated under a unified benchmark; by obtaining a candidate set, candidate convergence at the pressing granularity is achieved, improving pairing determinism, reducing mismatch propagation, calculating physical consistency scores and mismatch self-checking marking mechanisms, forming a strict credible sample screening and admission control; gating judgment is performed on credible samples, driving the calculation of calibration increments and updating the termination displacement setpoint and monotonic table, making the self-calibration closed loop stable in the long term, and improving the consistency of pressing quality. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a self-calibration method for press-fitting process based on quality inspection.

[0019] Figure 2 A flowchart for obtaining the candidate set.

[0020] Figure 3 A flowchart for generating mismatch self-check tags.

[0021] Figure 4 The flowchart for obtaining the updated monotonic table. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4This is one embodiment of the present invention, which provides a self-calibration method for press-fitting process based on quality inspection, including the following steps: S1. Load the process parameter set, execute the calibration press fitting height measurement cycle, generate a summary quantity sample set, calculate the deviation length and dimensional quantity, generate an alignment parameter set, perform the press fitting cycle, generate process records and inspection records, and write them to the traceability queue.

[0026] Read the process document, obtain the set of process parameters, execute the calibration press fitting height measurement cycle, collect the press force displacement sequence and the measuring point height sequence, determine the end face reference height, solidify the measuring point configuration, calculate the equivalent press amount, and generate a summary quantity sample set.

[0027] Furthermore, when the industrial control device enters the production preparation state, it reads the process file, obtains the target equivalent pressing amount, the termination displacement setting value, the update threshold, and the contact force threshold, and writes them into the parameter area for storage. During writing, the fixed fields in the parameter area are used to store them separately. After writing is completed, a read back check is performed to confirm that all parameters in the parameter area can be read normally.

[0028] Place the qualified calibration part at the press-fit station and position it according to the mass production positioning method. Perform three consecutive calibration press-fit height measurement cycles, and complete the press-fit height measurement according to the mass production rhythm in each cycle.

[0029] During each pressing operation, the industrial control device synchronously collects the pressing force displacement sequence. When a termination action is issued, the termination displacement point is obtained. The industrial control device performs a moving average filter on the pressing force in the pressing force displacement sequence, with a moving average filter window length of five sampling points. Starting from the initial sampling point of the pressing force displacement sequence, the device detects point by point and identifies the earliest sampling point where the three consecutive sampling points of the pressing force after moving average filtering are not less than the contact force threshold. The displacement corresponding to the earliest sampling point is the contact displacement point. The pressing force curve with displacement is integrated between the contact displacement point and the termination displacement point to obtain the pressing energy. The effective stroke is the displacement corresponding to the termination displacement point minus the displacement corresponding to the contact displacement point. The ratio of the pressing energy to the effective stroke is used as the average pressing force. The average pressing force and pressing energy are obtained after the pressing operation is completed.

[0030] It should be noted that, specifically, to obtain the contact force threshold, the industrial control device uses the contact point indentation force as a benchmark, continues to descend at the same low speed while simultaneously collecting the indentation force and displacement, collects the indentation force at 20 consecutive sampling points, calculates the difference between the maximum indentation force at the 20 sampling points and the indentation force at the contact point, and obtains the safety increment; the industrial control device uses the sum of the contact point indentation force and the safety increment as the contact force threshold, writes the contact force threshold into the parameter area, and completes a back-check confirmation.

[0031] The lower limit of the contact force threshold is the infeed force at the contact starting point, and the upper limit is 0.2 times the minimum average infeed force of three calibration press-fit cycles. This range ensures that the infeed force corresponding to the contact displacement point exceeds the no-load noise and zero drift, avoiding false triggering of the contact displacement point in the no-load section. Specifically, the upper limit of the contact force threshold, which is 0.2 times the minimum average infeed force of three calibration press-fit cycles, ensures that the contact displacement point is located in the initial contact stage of press-fitting, avoiding the contact displacement point falling into the stable press-fitting stage, which would lead to a shortened press-fitting energy integration interval, an excessively large average infeed force, discretization, and increased fluctuations in the monotonic table index. If the contact force threshold exceeds the upper limit, the contact displacement point will shift backward, leading to an increase in the deviation of press-fitting energy and effective stroke calculations, and a decrease in the stability of the candidate set selection.

[0032] Within each height measurement, the quality inspection station performs multi-point height measurement of the end face according to the fixed measurement point configuration to obtain the average height and range of the end face.

[0033] After three calibration press fitting height measurement cycles, three sets of summary samples of average pressing force, press fitting energy, end face range, and end face height were obtained.

[0034] The median of the three sets of end face heights obtained from three calibration pressure fitting height measurement cycles is taken to determine the end face reference height and write it into the parameter area; at the same time, the number and location of the measuring points are fixed so that subsequent mass production height measurement and calibration height measurement are comparable under the same reference.

[0035] For each of the three sets of end face average heights, the end face reference height is subtracted from the end face average height to obtain three sets of equivalent pressing amount samples. The three sets of equivalent pressing amount, the three sets of end face range, end face average height, average pressing force, and pressing energy are used as the summary sample set of three calibration pressing height measurement cycles.

[0036] Based on the aggregated sample set, calculate the step size and scale, generate a monotonic table, determine the alignment code threshold and mismatch threshold, write them into the parameter area, and generate the alignment parameter set.

[0037] Furthermore, the difference between the maximum and minimum values ​​of the average pressing force, pressing energy, equivalent pressing amount, and end face range for the three calibration pressing height measurement cycles is calculated respectively. The deviation distance of the pressing force, the deviation distance of the pressing energy, the deviation distance of the equivalent pressing amount, and the deviation distance of the end face range are determined in sequence. The equivalent pressing amount scale is set as the equivalent pressing amount deviation distance, and the end face range scale is set as the end face range deviation distance. Both the deviation distance and the scale are written into the parameter area.

[0038] The average pressing force of three calibration pressing and height measurement cycles is discretized into bins based on the step length of the pressing force. The median of the corresponding equivalent pressing amount in each bin is taken to generate the initial content of the monotonic table and write it into the parameter area. The parameter area is also written with the maximum bin index of the monotonic table to limit the indexable range of the monotonic table.

[0039] Specifically, uncrunted binning indexes and endpoint-truncated binning indexes are represented as follows: ; ; in, This indicates that the binning index was not truncated. This indicates an endpoint-truncated binning index. This represents the maximum bin index of the monotonic table.

[0040] It should be noted that empty bin filling is performed before the initial content of the monotonic table is written to the parameter area. Specifically, the index range of the monotonic table is limited to bin index 0 to bin index 1. When there are binning indexes within the monotonic table index range that have not formed a median, an invalid value is marked at the binning index of the monotonic table. For each invalid value-marked binning index, the nearest valid binning index is searched simultaneously along the decreasing and increasing directions of the binning index. The value at the smallest valid binning index is selected to fill the invalid value-marked binning index. If the distances are the same, the value at the smaller valid binning index is selected to complete the filling.

[0041] After the empty bin filling process is completed, monotonicity correction is performed. Specifically, the monotonicity table is traversed in ascending order of bin index; the value corresponding to each bin index is compared with the value corresponding to the previous bin index; if the value corresponding to the bin index is less than the value corresponding to the previous bin index, the value corresponding to the bin index is replaced with the value corresponding to the previous bin index.

[0042] Specifically, the expected equivalent push in the monotonic table is expressed as: ; in, This represents the expected equivalent push in the monotonic table. Indicates the average pressing force. Indicates the equivalent pressurization amount. This indicates the distance between the pressure applied and the walking distance. It indicates the distance from the walk.

[0043] It should be noted that when constructing the monotonic table, the average pressure in the process records is used as the grouping basis. The average pressure is discretized according to the pressure dispersion step size to obtain the pressure discrete index. For process records with the same pressure discrete index, the corresponding equivalent pressure samples are collected. The median of the equivalent pressure samples is used as the expected equivalent pressure at the pressure discrete index. The median is used to suppress the influence of a small number of abnormal process records, such as transient impacts, measurement noise, and local jamming, on the expected equivalent pressure, thereby improving the robustness of the monotonic table.

[0044] The alignment code distance was calculated for the three sets of samples from the three calibration pressure test cycles, and the largest alignment code distance was written into the parameter area as the alignment code threshold.

[0045] Specifically, the alignment code distance is expressed as: ; in, Indicates the alignment code distance. This indicates the distance between the equivalent pressure and the walking distance. Indicates the pressure loading energy. This indicates the distance between the pressurization energy and the walking distance. Indicates extreme end-face defects. This indicates the distance between the end face and the walking distance.

[0046] It should be noted that the alignment code threshold is a non-negative integer, and the upper limit of the alignment code threshold is preferably no more than 8. The alignment code threshold is determined by the maximum value of the alignment code distance of three sets of samples in three calibration press-fit height measurement cycles. The alignment code distance is composed of the larger of two values: the absolute value of the difference between the discrete bin index of the average press-in force and the discrete bin index of the equivalent press-in amount, and the absolute value of the difference between the discrete bin index of the press-fit energy and the discrete bin index of the end face range. The alignment code distance is a non-negative integer. The lower limit of the alignment code threshold is 0 to ensure the existence of the gating condition, and the upper limit of the alignment code threshold is 8 to cover the calibration press-fit. The maximum alignment code distance of the cyclic sample is measured and the timing alignment error, discrete binning quantization error, and discrete binning index offset caused by mass production stage detection data and control side are absorbed to avoid excessive rejection of valid process records. The upper limit of the alignment code threshold is set to 8 to limit the size of the candidate set and constrain the full backtracking computation load. If the alignment code threshold exceeds 8, the shrinkage capability of the valid process record set decreases, the size of the candidate set increases the risk of mismatch and increases the full backtracking computation load. When the alignment code threshold exceeds 8, the alignment code threshold is set to 8 and written to the parameter area.

[0047] Furthermore, physical consistency scores were calculated for the three sets of samples from the three calibration pressure testing cycles, and the highest physical consistency score was written into the parameter area as the mismatch threshold.

[0048] Specifically, the physical consistency score is expressed as: ; in, Indicates the physical consistency score. Represents the equivalent pressurization scale. This represents the dimensional measure of the end face range.

[0049] It should be noted that the mismatch threshold is a non-negative number, and the upper limit of the mismatch threshold is preferably no more than 2. The range of the mismatch threshold is [0, 2]. The physical consistency score is determined by taking the maximum value of the equivalent pressure deviation normalization term and the end face range normalization term. The physical consistency score is non-negative. The lower limit of the mismatch threshold is 0 to ensure that the mismatch self-check comparison relationship exists. The upper limit of the mismatch threshold is 2 to cover the maximum value of the physical consistency scores of the three sets of samples in the three calibration pressure test cycles, and to absorb the fluctuations of the normalization term exceeding 1 caused by the detection noise, discrete quantization error, and timing alignment error in the mass production stage, so as to avoid excessive blocking of the confirmation of the pairing relationship. The upper limit of the mismatch threshold is 2 to maintain the sensitivity of mismatch blocking and limit the proportion of mismatched samples entering the update link. If the mismatch threshold exceeds 2, the sensitivity of mismatch blocking decreases, the proportion of mismatched samples entering the calibration incremental update and monotonic table update increases, and the risk of abnormal adjustment of the termination displacement setting value increases. When the mismatch threshold exceeds 2, the mismatch threshold is 2 and written into the parameter area.

[0050] Based on the alignment parameter set and end face reference height, the pressing cycle generates process records, the height measurement cycle generates inspection records, and the process records and inspection records are written into the traceability queue.

[0051] Furthermore, after entering mass production, the industrial control device synchronously collects the pressing force and displacement in each pressing cycle, records the termination displacement point, and performs a moving average filter on the pressing force in the pressing force-displacement sequence. The moving average filter window length is five sampling points. The industrial control device detects point by point from the starting sampling point of the pressing force-displacement sequence, identifies the earliest sampling point where the three consecutive sampling points of the pressing force after moving average filtering are not less than the contact force threshold, and takes the displacement corresponding to the earliest sampling point as the contact displacement point. The average pressing force and pressing energy are calculated, and the sum of the average pressing force and pressing energy is used to generate a process record, which is written to the traceability queue. The process record fields include the average pressing force and pressing energy, but are not written to the original sampling sequence point column. After writing, the loop buffer is cleared to reduce the writing load and queuing delay of the traceability queue.

[0052] The displacement is based on the displacement coordinates in the pressing direction. An increase in the displacement in the pressing direction indicates deeper pressing. The displacement corresponding to the contact displacement point is taken as the zero point of displacement, and the displacement corresponding to the termination displacement point is a positive value.

[0053] After the workpiece enters the quality inspection station, the quality inspection station performs multi-point height measurement on the end face under the measurement benchmark consistent with the qualified calibration part, obtains the height sequence of the measurement points, calculates the average height and range of the end face from the height sequence of the measurement points, and calculates the equivalent pressing amount by subtracting the average height of the end face from the end face benchmark height. The quality inspection station generates an inspection record with the sum of the equivalent pressing amount and the end face range, and writes the inspection record into the traceability queue. The fields of the inspection record are fixed and include the equivalent pressing amount and the end face range, and are not written into the point-by-point height sequence. The deeper the pressing, the smaller the average height of the end face and the larger the equivalent pressing amount.

[0054] After the inspection record is written to the traceability queue, the industrial control device uses the write event as a trigger signal to enter the alignment code generation and candidate filtering process, so that the four summaries of the workpiece in the traceability queue, namely the average pressing force, pressing energy, equivalent pressing amount, and end face range, can be directly read.

[0055] S2. Monitor the trace queue, identify detection records, obtain alignment reference packets, backtrack process records, perform validity screening, extract the set of valid process records, perform discretization processing, calculate alignment code distance, and obtain candidate sets.

[0056] Listen for write events in the traceability queue, identify detection records, lock the detection record index, perform discretization processing on the equivalent push amount and end face range, generate detection alignment code, associate with the detection record index, and obtain the alignment reference package.

[0057] Furthermore, during runtime, write listening logic is set for the traceability queue. Specifically, when new data is written to the traceability queue, the industrial control device reads the type identifier of the written data. If the written data is a detection record, the index position of the detection record is locked as the object to be processed, and the equivalent pressing amount and end face range in the detection record are read. The industrial control device reads the offset length of the pressing force, the offset length of the pressing energy, the offset length of the equivalent pressing amount, the offset length of the end face range, and the alignment code threshold from the parameter area, and caches them in the current processing context to ensure that all subsequent discretization and threshold judgment use the same set of parameters.

[0058] Furthermore, the industrial control device performs discretization processing on the ratio of the calculated value to the corresponding discrete step size and rounds it according to the equivalent pressing amount and end face range in the detection record, to obtain the two components of the detection alignment code, and caches the detection alignment code as the alignment reference packet for this full backtracking.

[0059] Specifically, the two components of the alignment code are represented as follows: ; ; in, This indicates the detection of the first component of the alignment code. This indicates the detection of the second component of the alignment code.

[0060] It should be noted that discretization converts continuous measurements into discrete codes, enabling detection records to be quickly and stably matched with a large number of process records, regardless of timestamp order.

[0061] The industrial control device caches the detection alignment code as the alignment reference package for this full backtracking, and simultaneously saves the detection record index in the cache so as to form candidate pairs with the selected process records and detection records.

[0062] Based on the alignment reference package, process records are read back along the traceability queue to extract the average pressing force and pressing energy, and validity screening is performed to obtain a set of valid process records. Based on the set of valid process records and the alignment reference package, the average pressing force and pressing energy are discretized, the alignment code distance is calculated, and a candidate set is generated.

[0063] Furthermore, starting from the index position of the detection record, the process records are read backwards one by one in the traceability queue; for each process record, the average pressing force and pressing energy are extracted and a validity screening is performed. Specifically, if the average pressing force or pressing energy is invalid, the process record is determined to be invalid and skipped, and the process continues to trace back to the next process record; this achieves full backtracking to extract valid process records, avoiding invalid data from entering the alignment code distance calculation and causing false filtering.

[0064] Furthermore, for each valid process record, the average pressing force and pressing energy are discretized by rounding down the discretization distance of the pressing force and the discretization distance of the pressing energy to obtain two components of the process alignment code; the alignment code detection is used as a reference to calculate the alignment code distance, and the alignment code distance is compared with the alignment code threshold. If the alignment code distance does not exceed the alignment code threshold, the process record is retained; if the alignment code distance exceeds the alignment code threshold, the process record is discarded.

[0065] Specifically, the two components of the process alignment code are represented as follows: ; ; in, The first component of the process alignment code is indicated. The second component of the process alignment code is indicated.

[0066] Furthermore, the process record index that satisfies the alignment code distance not exceeding the alignment code threshold and the corresponding alignment code distance are written into the candidate set, and after the full backtracking is completed, the candidate set and the detection record index are output together for consistency constraint self-check.

[0067] It should be noted that if the candidate set is empty, an empty candidate set will be output to trigger mismatch blocking. Through the candidate set generation mechanism, candidate shrinkage can be achieved using only the summaries in the traceability queue without relying on the timestamp order. This is the key to solving the mismatch of the calibration command target caused by the timing alignment error between the detection data and the control side.

[0068] S3. Perform integrity verification on the candidate set, calculate the physical consistency score, obtain the candidate pairing score table, determine the optimal pairing, form a confirmed pairing relationship, and obtain the mismatch self-check mark.

[0069] Perform integrity checks on the candidate set, obtain the candidate set that passes the checks, read the monotonic table from the parameter area, and generate a scoring calculation context; based on the scoring calculation context and the candidate set, traverse the candidate pairs, calculate the physical consistency score, and obtain the candidate pairing score table.

[0070] Furthermore, the industrial control device receives the candidate set and the corresponding detection record index, performs integrity verification, specifically, determines whether the candidate set is empty; if the candidate set is empty, a scoreable candidate pair cannot be formed, generates a suspected mismatch self-check mark, associates the mismatch self-check mark with the detection record index and writes it into the traceability queue, ends the current processing and returns to wait for the next detection record to enter the candidate set generation process.

[0071] If the candidate set is not empty, check each candidate pair in the candidate set to see if it includes the process record index and alignment code distance. Check if the detection record index exists in the traceability queue and if the equivalent push amount and end face range can be read. If any candidate pair field is missing or the detection record is unreadable, output the suspected mismatch self-check flag and write it to the traceability queue.

[0072] It should be noted that the industrial control device reads the monotonic table, equivalent indentation scale, end face range scale, and mismatch threshold from the parameter area only when both the candidate set and the detection record pass the integrity check, and caches them in the current processing context.

[0073] The candidate set is traversed one by one. For each candidate pair, data retrieval, alignment, and indexing are performed. Specifically, based on the process record index of the candidate pair, the average push force of the process record is read from the traceability queue; based on the detection record index, the equivalent push force and end face range of the detection record are read from the traceability queue; the push force deviation length and the maximum bin index of the monotonic table are read from the parameter area, and the bin index calculation and endpoint truncation rules are performed to obtain the bin index.

[0074] The industrial control device reads the expected equivalent push amount at the bin index of the monotonic table based on the bin index. If the value at the bin index of the monotonic table is marked as invalid, then an empty bin is backtracked to retrieve the value. The nearest valid bin index is searched simultaneously along the decreasing and increasing directions of the bin index, and the value at the minimum valid bin index is selected as the expected equivalent push amount. If the distances are the same, the value at the smaller valid bin index is selected.

[0075] It should be noted that an invalid value marker indicates that there is no valid value available for calculating the expected equivalent push at the binning index of the monotonic table; the invalid value marker uses the data type reserved code of the equivalent push field of the monotonic table; the invalid value marker is given by the invalid value marker parameter stored in the parameter area; when the value of the expected equivalent push field at the binning index of the monotonic table is equal to the invalid value marker parameter, the binning index of the monotonic table is determined to be an empty binning index; when the value of the expected equivalent push field at the binning index of the monotonic table is not equal to the invalid value marker parameter, the binning index of the monotonic table is determined to be a valid binning index; during the initialization phase of the monotonic table, the invalid value marker parameter is written to the parameter area, and the expected equivalent push fields at each binning index of the monotonic table are filled with the invalid value marker parameter; during the update phase of the monotonic table, when generating valid expected equivalent push field values, the invalid value marker parameter is overwritten with the valid expected equivalent push field values; the invalid value marker parameter selects a reserved code that does not fall into the valid value range of the equivalent push.

[0076] The back-off values ​​from the empty separation box are only used for the calculation of the physical consistency score of the candidate pairing and are not written back to the parameter area.

[0077] Read the expected equivalent push amount from the monotonic table output and calculate the physical consistency score for each candidate pair.

[0078] It should be noted that any significant inconsistency in any constraint will lead to an increase in the physical consistency score. The industrial control device writes the process record index and physical consistency score of each candidate pair into the candidate pairing score table, and retains the detection record index corresponding to the candidate pairing in the candidate pairing score table to ensure that the confirmed pairing relationship is traceable.

[0079] Perform a minimum value search in the candidate pairing scoring table to determine the optimal pairing, bind the process record index and the detection record index to form a confirmed pairing relationship, perform mismatch self-check judgment, and generate a mismatch self-check mark.

[0080] Furthermore, a minimum value search is performed in the candidate pairing score table, and the process record index corresponding to the candidate pair with the minimum physical consistency score is taken as the optimal pairing.

[0081] The process record index and the current test record index are bound together to form a confirmed pairing relationship. The confirmed pairing relationship is written into the traceability queue so that the calibration incremental update and monotonic table update can accurately retrieve data and perform gated updates based on the confirmed pairing relationship.

[0082] Read the physical consistency score of the optimal pair and compare it with the mismatch threshold. Specifically, if the physical consistency score is less than or equal to the mismatch threshold, output a mismatch self-check mark that has passed the mismatch self-check, and write the mismatch self-check mark and the confirmed pairing relationship into the traceability queue as the basis for allowing entry into the calibration incremental update and monotonic table update. If the physical consistency score is greater than the mismatch threshold, output a mismatch self-check mark that is suspected of being mismatched, write the mismatch self-check mark into the traceability queue, and mark the detection record as not allowed to enter the subsequent update link, blocking the mismatch sample before the self-calibration update.

[0083] S4. Based on the mismatch self-test mark value, perform gating judgment, obtain the set of pairing relationships confirmed by the mark, form a deviation sample set, calculate the calibration increment and update the termination displacement set value, update the monotonic table, and obtain the updated monotonic table.

[0084] Receive mismatch self-test flags and confirmed pairing relationships, perform gating judgment, write blocking flags for suspected mismatches, write passing flags for mismatch self-tests, and obtain the set of confirmed pairing relationships by passing flags; based on the set of confirmed pairing relationships by passing flags, read the equivalent push-in amount of the detection record and the target equivalent push-in amount, calculate the deviation value, and obtain the deviation sample set.

[0085] Furthermore, upon receiving the mismatch self-test flag and confirming the pairing relationship, a gating decision is executed. Specifically, the industrial control device determines the value of the mismatch self-test flag. If the mismatch self-test flag indicates a suspected mismatch, a blocking flag is written into the traceability queue for the test record, and the test record and its corresponding process record are kept in an unconfirmed state. The calibration increment calculation, termination of displacement setpoint update, and monotonic table update are stopped for the test record. If the mismatch self-test flag indicates a successful mismatch self-test, the confirmed pairing relationship is written into the traceability queue, and a pass flag is written onto the confirmed pairing relationship. The pass flag is used to screen and update samples.

[0086] Furthermore, the confirmed pairing relationships in the traceability queue are traversed, and the set of confirmed pairing relationships with passing marks is selected. The equivalent press-in amount of the corresponding detection record is read according to the detection record index, and the target equivalent press-in amount is read from the parameter area. The industrial control device calculates the deviation sample of the passing sample, subtracts the equivalent press-in amount from the target equivalent press-in amount to obtain the deviation value, and writes the deviation value into the deviation sample set. The deviation value is repeatedly calculated for all passing marked samples in the traceability queue to construct the deviation sample set.

[0087] Furthermore, based on the deviation sample set, a calibration increment is generated, the updated threshold amplitude is executed, the updated termination displacement setting value is deducted and written back to the parameter area, and the updated termination displacement setting value is obtained; according to the pairing relationship set confirmed by the marker and the updated termination displacement setting value, the equivalent push amount is collected, the monotonicity table is updated, the monotonicity correction is executed and written back to the parameter area, and the updated monotonicity table is obtained.

[0088] It should be noted that the update threshold is generated by the process document. Specifically, the process engineer collects trial operation quality test data based on the allowable pressure changes, displacement changes, cycle time requirements and safety boundaries of the press-fitting process, calculates the statistical distribution of calibration increments, selects the threshold value that meets the stability and responsiveness requirements, writes it into the parameter area, and reads the update threshold from the parameter area at the start of calibration and uses it to update the threshold limit.

[0089] The update threshold range can be set according to the actual process. For example, the update threshold range can be set to (0.01, 0.50) mm. The lower limit of the update threshold is aligned with the displacement data acquisition resolution and assembly repeatability error to ensure that the calibration increment after the limit is maintained as a non-zero update amplitude when there is a quality inspection deviation, thus avoiding the update link from stalling. The upper limit of the update threshold is aligned with the press-fitting process safety boundary and process cycle constraints to ensure that the adjustment of the single termination displacement setting value is limited, reduce the risk of parameter jump caused by abnormal samples, and control the impact of a single update on the process cycle and assembly safety boundary.

[0090] Furthermore, the update process is executed. Specifically, the median of the deviation sample set is used as the calibration increment. The update threshold is read from the parameter area, and the calibration increment is limited by the update threshold to ensure that the limited calibration increment does not exceed the range defined by the update threshold, thus limiting the impact of a single update on the process cycle time and assembly safety boundary. The current termination displacement setting value in the parameter area is read. The termination displacement setting value uses the displacement coordinate of the pressing direction. An increase in the termination displacement setting value indicates that the termination displacement point moves along the pressing direction and the pressing is deeper. The calibration increment is the equivalent pressing amount minus the target equivalent pressing amount. A positive calibration increment indicates that the equivalent pressing amount is too large, and a negative calibration increment indicates that the equivalent pressing amount is too small. The termination displacement setting value update is the termination displacement setting value minus the limited calibration increment to ensure that the termination displacement setting value update forms negative feedback. The limited calibration increment is subtracted from the termination displacement setting value to obtain the updated termination displacement setting value, and the updated termination displacement setting value is written back to the parameter area.

[0091] Specifically, the calibration increment is expressed as: ; ; ; in, Indicates the calibration increment. The target equivalent pressure amount This represents the calibration increment after clipping. Indicates the need to update the threshold. This indicates the termination displacement setting value.

[0092] Furthermore, the monotonic table is updated by marking samples. Specifically, the confirmed pairings in the traceability queue are traversed again to filter out confirmed pairings that have passed the marking. For each filtered confirmed pairing, the average pressure of the corresponding process record is read according to the process record index, and the equivalent pressure of the corresponding detection record is read according to the detection record index. The deviation of the pressure and the maximum bin index of the monotonic table are read from the parameter area. The bin index calculation and endpoint truncation rules are executed to obtain the bin index. The endpoint truncation limits the bin index to the range of bin index zero to the maximum bin index of the monotonic table.

[0093] The equivalent pushes belonging to the same bin index are aggregated into a bin sample set. After traversal, the median of each bin sample set is taken as the expected equivalent push value of the monotonic table at the bin index. When the bin sample set is empty, an invalid value is written at the bin index of the monotonic table to form an empty bin marker.

[0094] When an empty bin marker exists, empty bin filling is performed. Specifically, for each invalid value, a bin index is marked, and the nearest valid bin index is searched simultaneously along both the decreasing and increasing directions of the bin index. The value at the smallest valid bin index is selected to fill the invalid value marking bin index. If the distances are the same, the value at the smallest valid bin index is selected to complete the filling, and the updated monotonic table is written back to the parameter area.

[0095] It should be noted that before writing back the monotonic table, a monotonicity correction is performed on the monotonic table to ensure that the monotonic table as a whole remains monotonic and maintains the physical rationality and long-term stability of the consistency constraint. Specifically, the values ​​of the monotonic table are traversed in ascending order of binning index, and the values ​​of adjacent binning indices are compared. If the value of a subsequent binning index is less than the value of the previous binning index, the value of the subsequent binning index is replaced with the value of the previous binning index to eliminate local non-monotonic segments. The updated monotonic table is then written to the parameter area and kept in effect.

[0096] Both the termination displacement setpoint and the monotonic table have been updated and solidified through the closed loop of the marked sample. The updated termination displacement setpoint will be used when the next pressing cycle terminates the control, and the updated monotonic table will be used when the next mismatch self-check calculates the physical consistency score, forming a self-calibration closed loop of mismatch self-check gating, trusted sample updating, and parameter write-back.

[0097] This embodiment also provides a computer device applicable to the self-calibration method of press-fitting process based on quality inspection, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the self-calibration method of press-fitting process based on quality inspection as proposed in the above embodiment.

[0098] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0099] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the self-calibration method for press-fitting process based on quality inspection as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0100] In summary, this invention generates an alignment parameter set and establishes a traceability queue, enabling traceable association of process records and inspection records under a unified benchmark; it achieves candidate convergence at the pressing granularity by acquiring a candidate set, improving pairing determinism, reducing mismatch propagation, and calculating physical consistency scores and mismatch self-checking marking mechanisms to form a strict credible sample screening and admission control; it performs gating judgment on credible samples, drives the calculation of calibration increments and updates the termination displacement setpoint and monotonic table, ensuring long-term stability of the self-calibration closed loop and improving the consistency of pressing quality.

[0101] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A self-calibration method for press-fitting process based on quality inspection, characterized in that: include, Load the process parameter set, execute the calibration press fitting height measurement cycle, generate the summary quantity sample set, calculate the deviation length and dimensional quantity, generate the alignment parameter set, perform the press fitting cycle, generate process records and inspection records, and write them into the traceability queue; Listen to the trace queue, identify detection records, obtain alignment reference packets, backtrack process records, perform validity screening, extract the set of valid process records, perform discretization processing, calculate alignment code distance, and obtain candidate set; Perform integrity checks on the candidate set, calculate physical consistency scores, obtain candidate pairing score tables, determine the optimal pairing, form confirmed pairing relationships, and obtain mismatch self-check flags; Gating judgment is performed based on the mismatch self-test mark value, the set of pairing relationships confirmed by the mark is obtained, a deviation sample set is formed, the calibration increment is calculated and the termination displacement set value is updated, the monotonic table is updated, and the updated monotonic table is obtained.

2. The self-calibration method for press-fitting process based on quality inspection as described in claim 1, characterized in that: The process of generating the alignment parameter set includes reading the process file, obtaining the process parameter set, executing the calibration press fitting height measurement cycle, collecting the press force displacement sequence and the measurement point height sequence, determining the end face reference height, solidifying the measurement point configuration, calculating the equivalent press amount, and generating a summary quantity sample set. Based on the aggregated sample set, calculate the step size and scale, generate a monotonic table, determine the alignment code threshold and mismatch threshold, write them into the parameter area, and generate the alignment parameter set.

3. The self-calibration method for press-fitting process based on quality inspection as described in claim 2, characterized in that: The process of writing to the traceability queue includes generating process records in a pressing cycle and generating detection records in a height measurement cycle, based on the alignment parameter set and the end face reference height. The process records and detection records are then written to the traceability queue.

4. The self-calibration method for press-fitting process based on quality inspection as described in claim 1, characterized in that: The process of obtaining the alignment reference package includes: listening to the write event of the trace queue, identifying the detection record, locking the detection record index, performing discretization processing on the equivalent push amount and the end face range, generating the detection alignment code, associating the detection record index, and obtaining the alignment reference package.

5. The self-calibration method for press-fitting process based on quality inspection as described in claim 4, characterized in that: The process of obtaining the candidate set includes reading process records back along the traceability queue according to the alignment reference package, extracting the average pressing force and pressing energy, performing validity screening, and obtaining a set of valid process records. Based on the valid process record set and alignment reference package, the average pressing force and pressing energy are discretized, the alignment code distance is calculated, and a candidate set is generated.

6. The self-calibration method for press-fitting process based on quality inspection as described in claim 2, characterized in that: The process of obtaining the mismatch self-check flag includes performing an integrity check on the candidate set, obtaining the candidate set that passes the check, reading the monotonic table from the parameter area, and generating a scoring calculation context. Based on the scoring calculation context and combined with the candidate set, the candidate pairs are traversed to calculate the physical consistency score and obtain the candidate pairing score table. Perform a minimum value search in the candidate pairing scoring table to determine the optimal pairing, bind the process record index and the detection record index to form a confirmed pairing relationship, perform mismatch self-check judgment, and generate a mismatch self-check mark.

7. The self-calibration method for press-fitting process based on quality inspection as described in claim 6, characterized in that: The process of forming the deviation sample set includes receiving mismatch self-check tags and confirming pairing relationships, performing gating judgment, writing blocking tags for suspected mismatches, writing passing tags through mismatch self-checks, and obtaining a set of confirmed pairing relationships through passing tags. Based on the pairing relationship set confirmed by marking, the equivalent indentation of the detection record and the target equivalent indentation are read, the deviation value is calculated, and the deviation sample set is obtained.

8. The self-calibration method for press-fitting process based on quality inspection as described in claim 7, characterized in that: The process of obtaining the updated monotonic table includes generating a calibration increment based on the deviation sample set, performing an update threshold limit, deducting the update termination displacement setting value and writing back to the parameter area, and obtaining the updated termination displacement setting value. Based on the pairing relationship set confirmed by the markers and the updated termination displacement setting, the equivalent press-in amount is collected, the monotonicity table is updated, the monotonicity correction is performed and the parameter area is written back, and the updated monotonicity table is obtained.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the self-calibration method for press-fitting process based on quality inspection as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the self-calibration method for press-fitting process based on quality inspection as described in any one of claims 1 to 8.

Images