Transmission shaft jumping detection system and detection method

Abstract

The invention discloses a jumping detection system and detection method for a transmission shaft, belongs to the technical field of transmission shaft detection, and aims to solve the problem that a traditional detection scheme is only based on a general standard and has no self-detection mechanism and process tracing ability. Comprising a multi-dimensional detection adaptation module which collects multi-dimensional data, screens a core detection dimension, calculates a dynamic qualification judgment interval and an updating mechanism, and constructs an adaptation database; the detection self-inspection module identifies false inspection through first inspection and reinspection, records false inspection data, draws a curve, calculates a self-inspection value, compares the self-inspection value with a threshold value, and starts instrument self-inspection when the threshold value is exceeded; the exception tracing module confirms an out-of-tolerance piece, calculates a process exception value, and positions and regulates a to-be-regulated process by means of a core dimension-machining process mapping table; according to the invention, the detection precision and reliability are improved, a detection-tracing-optimization closed loop is formed, and the qualified rate and the production detection efficiency of the transmission shaft are improved.

Description

Technical Field

[0001] This invention belongs to the field of transmission shaft testing technology, specifically relating to a transmission shaft runout detection system and method. Background Technology

[0002] In the field of drive shaft inspection, existing assembly line solutions rely solely on general industrial standards, failing to integrate multi-dimensional data such as the application scenario, shaft structure, and service conditions of the drive shaft under inspection, nor utilizing big data such as industry fault statistics. The selection of inspection dimensions and the pass / fail judgment range remain fixed in the long term, unable to be dynamically adjusted according to workpiece characteristics and industry needs, thus becoming disconnected from practical applications. Specifically, the following technical problems are raised: 1. The core detection dimensions were not selected based on multi-dimensional data, the dynamic judgment interval was not calculated, and the equipment update mechanism was not established, which failed to meet actual needs; 2. When an abnormality is detected, there is no step to re-inspect, identify false positives, plot curves, and calculate self-test values, which makes it easy to misjudge and difficult to locate the instrument problem; 3. The lack of ability to trace abnormal situations leads to low efficiency in problem cycles. To address this, we propose a drive shaft runout detection system and method. Summary of the Invention

[0003] The purpose of this invention is to provide a drive shaft runout detection system and method to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a drive shaft runout detection system and method, comprising: Multi-dimensional detection and adaptation module: For the type of drive shaft to be inspected for the first time, multi-dimensional basic data is collected, the core runout detection dimension set is filtered, the qualified judgment interval of the core dimensions is calculated, the update cycle of the qualified judgment interval and the update trigger condition of the core dimension set are set, and the drive shaft runout detection adaptation database is constructed; for the drive shaft to be inspected that has been recorded in the database, its core runout detection dimension set and the qualified judgment interval of each core dimension are retrieved for detection; The self-test module determines the pass / fail status of each core runout dimension and the drive shaft. Non-conforming products are re-tested. If the re-test is successful, it is determined that the first inspection was a false positive. The false positive time-deviation data of the core dimension is recorded and the change curve is plotted. Data is extracted using a sliding window to calculate the self-test evaluation value. If the threshold is exceeded, the self-test of the corresponding testing instrument for the core dimension is initiated. Anomaly tracing module: For each core runout detection dimension, drive shafts that still fail re-inspection are marked as confirmed out-of-tolerance parts, and out-of-tolerance deviation time change curves are plotted; process anomaly judgment values ​​are calculated, and if the threshold is exceeded, the process is judged as abnormal; a core dimension-processing process mapping table is constructed to determine the processes to be controlled and adjust the process parameters.

[0005] Preferably, the specific process for selecting the core jump detection dimension set is as follows: For the type of drive shaft to be inspected for the first time, collect multi-dimensional basic data; Obtain all potential runout detection dimensions of the drive shaft under inspection and organize them into a runout detection dimension set; For each potential detection dimension in the set, the importance coefficient of the potential detection dimension is obtained by combining the demand weight of the application scenario of the drive shaft under inspection, its correlation with the basic parameters of the service condition of the drive shaft under inspection, the proportion of failures caused by its deviation in the big data of industry and social applications, the degree of influence of extreme parameters of service condition on its deviation, and the preset weight coefficient. Compare the importance coefficient of each potential detection dimension with the corresponding preset threshold. Dimensions that are not lower than the threshold are marked as core fluctuation detection dimensions. All core fluctuation detection dimensions are then compiled to form a core fluctuation detection dimension set.

[0006] Preferably, the specific process for calculating the pass / fail judgment interval of the core dimension is as follows: For each core runout detection dimension in the core runout detection dimension set, the pass / fail judgment range of the core runout detection dimension is determined by combining the general industrial standard benchmark value corresponding to the core runout detection dimension, the structural parameters of the drive shaft under test related to the core dimension, the basic parameters of the service conditions of the drive shaft under test related to the core dimension, the extreme parameters of the service conditions of the drive shaft under test related to the core dimension, as well as the preset safety factor and preset correction factor.

[0007] Preferably, the specific process for setting the update cycle of the qualified judgment interval and the update trigger conditions of the core dimension set is as follows: For each core runout detection dimension, the stability coefficient of that core runout detection dimension is obtained by combining the fluctuation range of the pass / fail judgment value of that dimension in the past six months, the rate of change of the importance of that dimension in industry demand, the average design iteration cycle of the transmission shaft adapter product to be inspected, and the preset weight coefficient. Multiple stability coefficient ranges are set, and each stability coefficient range corresponds to a qualified judgment range update cycle. The stability coefficients of the core jump detection dimension are matched with these ranges, and the corresponding update cycle is output. This cycle is used as the interval between the current update and the next update of the qualified judgment range of this dimension. When the cycle is reached, the qualified judgment range of this dimension is updated. Calculate the pass / fail interval update cycle for all dimensions in the core runout detection dimension set, select the maximum value and mark it as the update reference time. When the update reference time is reached, calculate the standard deviation of the deviation value of each core runout detection dimension in the core runout detection dimension set of the drive shaft under inspection, and record it as the stability index of the core runout detection dimension set. If the stability index of the core runout detection dimension set of the drive shaft under inspection is greater than or equal to the corresponding threshold, then update the core runout detection dimension set and the pass / fail judgment interval of each core runout detection dimension.

[0008] Preferably, the specific process for retrieving the core runout detection dimension set and the pass / fail judgment interval of each core dimension for the recorded drive shaft to be inspected is as follows: The type information of all drive shafts to be inspected, the set of core runout detection dimensions for the corresponding type, the pass / fail judgment range for each core detection dimension, the update cycle of the pass / fail judgment range for each core detection dimension, and the update base time of the core runout detection dimension set are organized to build a drive shaft runout detection adaptation database. For the drive shaft to be inspected that is already recorded in the drive shaft runout detection adapter database, retrieve the core runout detection dimension set and the pass / fail judgment range of each core detection dimension of the drive shaft to be inspected from the drive shaft runout detection adapter database, and send them to the detection module.

[0009] Preferably, the specific process of recording the false detection time-deviation data of the core dimensions and plotting the change curve is as follows: For each drive shaft to be inspected, obtain the first detection value of each core runout detection dimension, and at the same time retrieve the corresponding pass / fail judgment interval for that dimension; If the initial test value is within the acceptable range, the core runout test dimension is deemed acceptable; otherwise, it is deemed unacceptable, and the corresponding drive shaft to be tested is deemed an unacceptable product. All non-conforming products are sent to the re-inspection end, where precise testing is performed on the core non-conforming testing dimensions of the non-conforming products to obtain the re-inspection values ​​for those core testing dimensions. If the retested value is within the acceptable range, the first test of that dimension is determined to be a false test, and the false test deviation value is calculated. For each core detection dimension, record the time of each false detection to form a data pair of false detection time - false detection deviation value; A two-dimensional rectangular coordinate system is constructed with time on the horizontal axis and false detection deviation value on the vertical axis. All false detection time-false detection deviation value data pairs are marked as data points in the coordinate system. All adjacent data points are connected by a curve in chronological order to generate the false detection deviation time change curve for this core detection dimension.

[0010] Preferably, the specific process for calculating the self-test evaluation value and triggering the self-test of the corresponding testing instrument in the core dimension when the value exceeds the threshold is as follows: A preset sliding detection window is used to extract three types of data from the false detection deviation time change curve within the window: the maximum false detection deviation within the window, the average false detection deviation within the window, and the false detection frequency within the window. At the same time, three types of historical data are extracted from the false detection deviation time change curves corresponding to all historical false detection data of this core jump detection dimension: the maximum historical false detection deviation, the average historical false detection deviation, and the historical false detection frequency. By combining preset weighting coefficients, the self-assessment value of this core jitter detection dimension is analyzed; The self-test evaluation value of the core vibration detection dimension is compared with the preset threshold. If the self-test evaluation value is greater than or equal to the preset threshold, the self-test mode of the detection instrument corresponding to the core vibration detection dimension is activated.

[0011] Preferably, the specific process for plotting the time variation curve of the deviation is as follows: For each core runout detection dimension, select the drive shafts that are still deemed unqualified after re-inspection and mark them as confirmed out-of-tolerance parts. Calculate and confirm the out-of-tolerance deviation value of the core runout detection dimension in the out-of-tolerance component; Record the time of occurrence of the out-of-tolerance error in the core inspection dimension for each confirmed out-of-tolerance part, and form a data pair of out-of-tolerance occurrence time and out-of-tolerance deviation value; Construct a two-dimensional rectangular coordinate system: the horizontal axis represents time and the vertical axis represents the deviation value; mark all deviation occurrence time-deviation value data pairs as abnormal feature points in the coordinate system, and connect all adjacent abnormal feature points in chronological order with a curve to generate the deviation time change curve for this core detection dimension.

[0012] Preferably, a core dimension-processing step mapping table is established to determine the process to be controlled. The specific process for controlling the process parameters is as follows: A preset sliding analysis window is used to extract three types of data from the deviation time variation curve within the window: maximum deviation value, average deviation value, and deviation frequency within the window. Simultaneously, three types of historical data are extracted from the deviation time variation curves corresponding to all historical deviation data for this core inspection dimension: historical maximum deviation value, historical average deviation value, and historical deviation frequency. Combined with preset weighting coefficients, the process anomaly judgment value for this core inspection dimension is analyzed. If the abnormality judgment value of the process is greater than or equal to the preset threshold, then the processing process associated with the core detection dimension is determined to be abnormal. Construct a mapping table of core detection dimensions and associated processing procedures, in which each core detection dimension corresponds to an associated processing procedure. Substitute the core detection dimensions that are determined to be abnormal into the above mapping table, output the corresponding processing steps and mark them as processes to be controlled; The name of the process to be controlled, along with the out-of-tolerance data in the window, is sent to the personnel terminal. After receiving the information, the dispatch staff checks and controls the process parameters of the process to be controlled.

[0013] Compared with the prior art, the beneficial effects of the present invention are: (1) The drive shaft runout detection system and detection method integrate application scenarios, shaft structure, service conditions and industry big data to screen core detection dimensions through a multi-dimensional detection adaptation module. The core detection dimensions are more in line with the actual application needs of the drive shaft, reducing redundant detection and omission of key dimensions. Combined with multi-parameter calculation of dynamic qualified judgment interval and setting an update mechanism, the detection scheme accurately matches the actual needs of the workpiece and industry changes.

[0014] (2) The drive shaft runout detection system and detection method identify false detections by relying on the first inspection + accuracy re-inspection mechanism of the detection self-inspection module, so that the first inspection false detections are accurately identified, avoiding the waste of qualified products due to misjudgment and the omission of real defects. Then, by drawing the false detection curve and calculating the self-inspection evaluation value, the corresponding instrument self-inspection is initiated when the threshold is exceeded, reducing the misjudgment and omission, and timely locating the instrument abnormality, thereby ensuring the reliability of the detection results and reducing the risk of continuous detection deviation caused by the abnormality of the detection instrument.

[0015] (3) The drive shaft runout detection system and detection method use the anomaly traceability module to mark and confirm out-of-tolerance parts, calculate the process anomaly judgment value, locate the process to be controlled through the core dimension-processing process mapping table, and provide feedback on out-of-tolerance data to guide parameter adjustment, forming a detection-traceability-optimization closed loop to avoid problem cycle, thereby quickly blocking the recurrence of out-of-tolerance problems and improving the pass rate of drive shaft products and the overall efficiency of detection and production processes. Attached Figure Description

[0016] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0017] 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.

[0018] Example 1; Please see Figure 1 The present invention provides a drive shaft runout detection system and detection method, including: a multi-dimensional detection adaptation module, a detection self-test module and an anomaly tracing module; Multi-dimensional detection and adaptation module: For the type of drive shaft to be inspected for the first time, multi-dimensional basic data is collected, a core runout detection dimension set is filtered, the pass / fail judgment interval of the core dimensions is calculated, the update cycle of the pass / fail judgment interval and the update trigger condition of the core dimension set are set, and a drive shaft runout detection adaptation database is constructed; for the drive shaft to be inspected that is already recorded in the database, its core runout detection dimension set and the pass / fail judgment interval of each core dimension are retrieved for detection. The specific process is as follows: S1. For the type of drive shaft to be inspected that is undergoing runout detection for the first time, collect multi-dimensional basic data, specifically: By combining shaft information input with industry data interface calls, multi-dimensional basic data of the drive shaft is collected, including: Application scenario data: should at least cover the field of the drive shaft to be tested, the core requirements of the end product, and the accuracy requirements for the fit with related components after installation; Shaft structure data: at least covering the material, shape parameters (solid / hollow, shaft length-to-diameter ratio, presence or absence of splines / keyways), and surface treatment process of the drive shaft to be inspected; Service condition data: at least covering the rated speed, maximum axial load, operating temperature range, and vibration frequency range of the drive shaft under inspection; Industry and social application big data: at least covering statistics on similar drive shaft failures over the past three years (including the proportion of failures caused by out-of-tolerance in each testing dimension, industry demand trends (including the demand growth rate of each testing dimension), and update records of customized testing standards for mainstream customers); S2. Based on multi-dimensional basic data, select the core runout detection dimensions of the drive shaft to be inspected, and calculate the pass / fail judgment value for each core runout detection dimension, specifically: Obtain all potential runout detection dimensions of the drive shaft under inspection and organize them into a runout detection dimension set, which includes: radial runout, axial runout, end face circular runout, etc. For each potential detection dimension in the set of jitter detection dimensions, use the formula: We obtain the importance coefficient W of the potential detection dimension; in, The required weights of current potential inspection dimensions for the application scenarios of the drive shaft to be inspected; The correlation between the current potential inspection dimensions and the basic parameters of the service conditions of the drive shaft under inspection; The percentage of faults caused by current potential detection dimension deviations in big data applications for industry and society; The degree of impact of extreme parameters under service conditions on current potential deviations in detection dimensions; a1, a2, a3, and a4 are preset weighting coefficients; All the above parameters were normalized and dimensionless before calculation; For the application scenario of the drive shaft to be inspected, the required weight of the current potential inspection dimensions is as follows: The importance coefficient of potential detection dimensions integrates application scenario data, service condition data, and big data from industry and social applications. Furthermore, the process for obtaining each calculation parameter in the formula for calculating the importance coefficient of the potential detection dimension is as follows: For demand weighting: Determine the basic weight according to the field of adaptation of the drive shaft to be inspected and the corresponding basic weight level preset by experts; adjust the basic weight according to the degree of matching (complete / partial / mismatch) between the core requirements of the end product and the current potential inspection dimensions, and take the corresponding adjustment coefficient preset by experts; adjust the weight according to the precision requirements of the fit with related components after installation, and take the correction coefficient preset by experts (the more stringent the precision requirements, the larger the correction coefficient); normalize the final result so that the value falls in the range of 0-1, and that is the demand weight. Regarding the correlation: From the collected service condition basic parameters of the drive shaft under inspection, match the corresponding associated service condition basic parameters; first, based on the characteristics of the potential detection dimension (such as radial runout corresponding to rated speed, axial runout corresponding to maximum axial load), match the corresponding service condition basic parameters (such as rated speed / maximum axial load); retrieve the historical dataset of "matched service condition basic parameters - corresponding potential detection dimension runout value" of similar drive shafts stored in the system, and use the linear correlation analysis method to calculate the correlation between the two; take the absolute value of the calculated correlation, and if the absolute value exceeds the 0-1 range, truncate it to 0 or 1 to obtain the correlation; Regarding the failure rate: From the statistical data of similar driveshaft failures, the number of failure cases "caused solely by the current potential detection dimension exceeding the tolerance" is selected. The number of these selected failure cases is used as the numerator, and the total number of failure cases of similar driveshafts in the same period is used as the denominator to calculate the preliminary rate. Failure cases caused by factors other than the current potential detection dimension, such as installation errors and external impacts, are excluded. The preliminary rate is then corrected. The corrected result naturally falls within the 0-1 range, which is the failure rate. Regarding the degree of impact: From the collected extreme parameters of the service conditions of the drive shaft under inspection, associate them with the extreme parameters corresponding to the current potential inspection dimension; retrieve the historical inspection data of similar drive shafts under the same extreme parameters, count the number of out-of-tolerance tests for the current potential inspection dimension with the total number of tests, and calculate the out-of-tolerance rate; adjust the out-of-tolerance rate according to the actual frequency of occurrence of the extreme condition using a preset adjustment rule (the more frequent the occurrence, the larger the adjustment coefficient); normalize the adjusted results so that the value falls within the 0-1 range, thus obtaining the degree of impact.

[0019] The importance threshold for the pre-defined dimension jump detection dimension is used to compare the importance coefficient of each potential detection dimension with the corresponding pre-defined threshold. If it is greater than or equal to the corresponding threshold, the potential detection dimension is marked as the core jump detection dimension. All core bounce detection dimensions are organized to form a core bounce detection dimension set; For each core bounce detection dimension in the core bounce detection dimension set, the formula is used: The pass / fail judgment interval [L, U] of the core jitter detection dimension is obtained. Where L and U are the lower and upper bounds of the pass / fail judgment interval for the core detection dimension, respectively. These are the general industry standard benchmark values ​​corresponding to the core runout detection dimensions; The structural parameters of the transmission shaft under test are related to the core runout detection dimension, such as the shaft wall thickness corresponding to axial movement. For the basic service condition parameters related to the core runout detection dimension, such as: the maximum axial load α corresponding to axial movement; Extreme parameters of service conditions related to the core runout detection dimension, such as: axial runout. α and β are preset safety factors, where α takes the value of (0, 1) and β > 1; k1, k2, m1, and m2 are preset correction coefficients; All the above parameters were normalized and dimensionless before calculation; Furthermore, the process for obtaining the parameters in the calculation formulas for the lower and upper bounds of the pass / fail judgment interval in the core bounce detection dimension is as follows: For general industrial standard benchmark values: extract the benchmark values ​​corresponding to the current core runout detection dimensions from the current general industrial standards or national standards for drive shaft testing, and use them as the basis for calculating the pass / fail judgment range; For shaft structure parameters: First associate the current core runout detection dimension with the shaft structure parameter type (such as shaft wall thickness corresponding to axial runout), and then retrieve the corresponding type of structural parameter value from the collected shaft structure data of the transmission shaft to be inspected; For basic service condition parameters: First, match the basic service condition parameter type corresponding to the current core runout detection dimension (such as the maximum axial load corresponding to axial movement), and then extract the corresponding basic parameter values ​​from the collected service condition data of the drive shaft under test.

[0020] Extreme parameters of service conditions: First, determine the type of extreme parameters of service conditions associated with the current core runout detection dimension (such as extreme axial load corresponding to axial movement), and then obtain the corresponding extreme parameter values ​​from the collected service condition data of the drive shaft under test. S3. Set the update frequency of the pass / fail judgment value for each core runout detection dimension and the update trigger condition for the core runout detection dimension set of the drive shaft. The specific process is as follows: For each core bounce detection dimension, the formula is used: The stability coefficient S of the core jitter detection dimension is obtained; in, This represents the fluctuation range of the pass / fail judgment values ​​for the core runout detection dimensions over the past six months. The rate of change in the importance of core bounce detection dimensions in industry demands; The average design iteration cycle for products adapted to the drive shaft under inspection; b1, b2, and b3 are preset weighting coefficients; All the above parameters were normalized and dimensionless before calculation. Furthermore, the process of obtaining each parameter during the calculation of the stability coefficient is as follows: The fluctuation range of the pass / fail judgment value of the core runout detection dimension in the past six months: retrieve the historical pass / fail judgment value (old value) and the current pass / fail judgment value (new value) of the core runout detection dimension in the past six months from the drive shaft runout detection adapter database, calculate the absolute deviation between the two and divide by the old value to obtain the fluctuation range, and normalize and dimensionless process the result; The rate of change of importance of the core fluctuation detection dimension in industry demand: Based on the collected big data of industry and social applications, the demand ratio or demand growth rate of this core dimension in the past six months is extracted. After evaluation and calibration by domain experts, it is normalized and dimensionless to be used as the rate of change. Average design iteration cycle of the drive shaft adapter product under test: Extract the design iteration information of the drive shaft adapter product under test from the collected application scenario data; or call the average design iteration cycle data of similar adapter products in the industry database and normalize and remove the dimensions from the data; Multiple stability coefficient ranges are set, and each stability coefficient range corresponds to a qualified judgment range update cycle. The stability coefficient of the core bounce detection dimension is matched with all stability coefficient intervals, and the corresponding qualified judgment interval update cycle is output. This update cycle is used as the interval between the current time and the next update of the qualified judgment interval of this dimension. When the qualified judgment interval update cycle is reached, step S2 is triggered to update the qualified judgment interval of the core bounce detection dimension. Calculate the update cycle of the pass / fail judgment interval for all dimensions in the core runout detection dimension set, filter out the maximum value and mark it as the update reference time. When the update reference time is reached, calculate the standard deviation of the deviation value of each core runout detection dimension in the core runout detection dimension set of the drive shaft under test based on the standard deviation formula, and record it as the core runout detection dimension set stability index. Preset the core runout detection dimension set stability index threshold. If the core runout detection dimension set stability index of the drive shaft under test is greater than or equal to the corresponding threshold, then trigger steps S1 and S2 to update the core runout detection dimension set and the pass / fail judgment interval of each core runout detection dimension. The type information of all drive shafts to be inspected, the set of core runout detection dimensions for the corresponding type, the pass / fail judgment range for each core detection dimension, the update cycle of the pass / fail judgment range for each core detection dimension, and the update base time of the core runout detection dimension set are organized to build a drive shaft runout detection adaptation database. S4. For the drive shaft to be inspected that is already recorded in the drive shaft runout detection adapter database, retrieve the core runout detection dimension set and the pass / fail judgment range of each core detection dimension of the drive shaft to be inspected from the drive shaft runout detection adapter database, and send them to the detection module.

[0021] It should be noted that the entire process design, from data collection to dimensional filtering, judgment calculation, dynamic updating, and database construction, achieves both precision and dynamism in the drive shaft inspection solution. Specifically: Dual-path acquisition of multi-dimensional data takes into account both the characteristics of the shaft itself and the dynamics of industry applications, providing comprehensive data support for subsequent inspection design and avoiding inspection deviations caused by single data. Based on multi-dimensional data, the importance coefficient of potential detection dimensions is calculated, and the core vibration detection dimensions selected are more in line with the actual application needs of the drive shaft, reducing redundant detection or omission of key dimensions. The pass / fail judgment range is calculated by combining general standards, shaft structure and service conditions, and a safety factor is introduced for optimization, so that the judgment criteria are more in line with the actual service requirements of the drive shaft and the accuracy of the test is improved. The update cycle for the pass / fail judgment interval is set based on the stability coefficient, and the update of the dimension set is triggered by the set stability index to ensure that the testing standards are dynamically adjusted with industry needs and product iterations, thus avoiding the problem of "standard rigidity". A drive shaft runout detection adaptation database is built. Drive shafts with recorded types can directly retrieve detection parameters without repeated calculations, which greatly improves the efficiency of the detection process and reduces manpower and time costs.

[0022] The self-inspection module determines the pass / fail status of each core runout dimension and the drive shaft. Non-conforming products undergo re-inspection. If the re-inspection passes, it is considered a false positive in the initial inspection. The false positive time-deviation data for each core dimension is recorded and a change curve is plotted. Data is extracted using a sliding window to calculate the self-inspection evaluation value. If the value exceeds a threshold, the corresponding testing instrument for that core dimension initiates self-inspection. The specific process is as follows: For each drive shaft to be inspected, obtain the first detection value of each core runout detection dimension, and at the same time retrieve the corresponding pass / fail judgment interval for that dimension; If the initial test value is within the acceptable range, the core runout test dimension is deemed acceptable; otherwise, it is deemed unacceptable, and the corresponding drive shaft to be tested is deemed an unacceptable product. All non-conforming products are sent to the re-inspection end, where precise testing is performed on the core non-conforming testing dimensions of the non-conforming products (re-inspection accuracy rate is not less than 99.9%) to obtain the re-inspection value of the core testing dimension. If the retest value is within the acceptable range, the first test of that dimension is determined to be a false test, and the false test deviation value (i.e., the absolute difference between the first test value and the retest value) is calculated. For each core detection dimension, record the time of each false detection (based on the time when the dimension was first judged to be unqualified in the first detection), forming a data pair of false detection time and false detection deviation value; A two-dimensional rectangular coordinate system is constructed with time on the horizontal axis and false detection deviation value on the vertical axis. All false detection time-false detection deviation value data pairs are marked as data points in the coordinate system. All adjacent data points are connected by a curve in chronological order to generate the false detection deviation time change curve for this core detection dimension. A preset sliding detection window is used to extract the following from the false detection deviation time change curve within the sliding window: maximum false detection deviation WP, average false detection deviation WJ, and false detection frequency P within the window; at the same time, three types of historical data are extracted from the false detection deviation time change curves corresponding to all historical false detection data in this dimension: maximum historical false detection deviation LWP, average historical false detection deviation LWJ, and historical false detection frequency LP. Using the formula: The self-test evaluation value E is obtained, where c1, c2, and c3 are preset weight coefficients. A preset self-test threshold is set, and the self-test value of the core fluctuation detection dimension is compared with the corresponding threshold. If the self-test value is greater than or equal to the corresponding preset threshold, the self-test mode of the detection instrument corresponding to the core detection dimension is activated.

[0023] It should be noted that the closed-loop design of initial inspection judgment – ​​high-precision re-inspection – false detection tracking – instrument self-test triggering effectively solves the problems of misjudgment, missed detection, and instrument status loss of control in the handling of detection anomalies, as detailed below: The two-layer verification mechanism of initial inspection judgment + ≥ accuracy rate re-inspection can accurately identify false initial inspection, avoid misjudging qualified drive shafts as unqualified, reduce product waste, and prevent the omission of real defects due to initial inspection deviation, thus improving the reliability of test results. Record the false detection time and deviation data of core dimensions and plot the change curves to transform scattered false detection information into a visual trend, providing intuitive data support for analyzing the status of testing instruments and avoiding blind judgment of the cause of false detection. By extracting real-time false detection data through a sliding window and combining it with historical data to calculate self-test evaluation values, the current status of the testing instrument can be dynamically assessed. When the threshold is exceeded, only the instrument self-test of the corresponding core dimension is activated, without the need for a complete shutdown, thus reducing the impact on testing efficiency while ensuring testing accuracy. A separate self-testing logic is designed for the core jump detection dimension, which works in synergy with the core dimensions selected by the multi-dimensional detection adaptation module. This focuses on key detection steps, ensures the reliability of core dimension detection, and further strengthens the accuracy of the overall detection solution.

[0024] Anomaly tracing module: For each core runout detection dimension, drive shafts that still fail re-inspection are marked as confirmed out-of-tolerance parts, and out-of-tolerance deviation time change curves are plotted; the process anomaly judgment value is calculated, and if it exceeds the threshold, the process is judged as an anomaly; a core dimension-processing process mapping table is constructed to determine the processes to be controlled and adjust the process parameters. The specific process is as follows: For each core runout detection dimension, select the drive shafts that are still deemed unqualified after re-inspection and mark them as confirmed out-of-tolerance parts. Calculate and confirm the out-of-tolerance deviation value of the core runout detection dimension in the out-of-tolerance part (i.e., the absolute value of the re-inspection value exceeding the boundary of the corresponding pass / fail judgment interval). Record the time of occurrence of the out-of-tolerance error in the core inspection dimension for each confirmed out-of-tolerance part, and form a data pair of out-of-tolerance occurrence time and out-of-tolerance deviation value; Construct a two-dimensional rectangular coordinate system: the horizontal axis represents time and the vertical axis represents the deviation value; mark all deviation occurrence time-deviation value data pairs as abnormal feature points in the coordinate system, and connect all adjacent abnormal feature points with a curve in chronological order to generate the deviation time change curve for this core detection dimension; A preset sliding analysis window is used to extract the maximum value of the out-of-tolerance deviation (CZ), the average value of the out-of-tolerance deviation (CJ), and the frequency of the out-of-tolerance deviation (CP) within the window from the out-of-tolerance deviation time change curve within the sliding window. At the same time, three types of historical data are extracted from the out-of-tolerance deviation time change curves corresponding to all historical out-of-tolerance data in this dimension: the historical maximum value of the out-of-tolerance deviation (LCZ), the historical average value of the out-of-tolerance deviation (LCJ), and the historical frequency of the out-of-tolerance deviation (LCP). Using the formula: The process anomaly judgment value F of the core detection dimension is obtained; where w1, w2, and w3 are preset weight coefficients. A preset threshold for process abnormality is set. If the process abnormality judgment value of the core detection dimension is greater than or equal to the corresponding preset threshold, it is determined that the processing process associated with the core detection dimension is abnormal. Based on the structural characteristics of the drive shaft to be inspected, the inspection objects of the core inspection dimensions, and the processing logic of the drive shaft, a mapping table of core inspection dimensions and associated processing steps is constructed. Each core inspection dimension in this table is associated with a corresponding processing step. For example, radial runout is associated with lathe precision turning process, axial runout is associated with grinding machine axial positioning process, and end face circular runout is associated with milling machine end face milling process. The core detection dimensions that are identified as abnormal are substituted into the mapping table, and the corresponding processing steps are output and marked as the process to be adjusted. The name of the process to be adjusted and the out-of-tolerance deviation data in the window are sent to the personnel terminal. After receiving the information, the terminal dispatchers check and adjust the process parameters (such as cutting speed, positioning accuracy, and processing temperature) of the process to be adjusted.

[0025] It should be noted that through a complete chain design—from out-of-tolerance confirmation to trend visualization, anomaly quantification, process correlation, and parameter adjustment—a precise closed loop from anomaly detection to process improvement is achieved, as detailed below: Parts that still fail the screening and re-inspection are marked as confirmed out-of-tolerance parts and the out-of-tolerance deviation value is calculated to ensure that the out-of-tolerance data included in the analysis is true and reliable, avoid false detections from interfering with process traceability, and lay an accurate data foundation for subsequent analysis; Plotting the time variation curve of deviations transforms scattered deviation information into trend characteristics over time, intuitively presenting the changing patterns of deviation magnitude and frequency, facilitating rapid identification of concentrated periods and patterns of anomalies, and improving the efficiency of anomaly analysis. By combining real-time out-of-tolerance data from the sliding window with historical data, the process anomaly judgment value is calculated, the degree of process anomaly is quantitatively assessed, and misjudgment caused by subjective judgment is avoided, making the process anomaly judgment more scientific and objective. Based on the shaft structure, the object of inspection, and the process logic, a core dimension-processing process mapping table is constructed to establish a direct correlation between inspection anomalies and processing steps, solving the problem of disconnect between traditional inspection and processing. This allows for quick location of the specific process corresponding to the out-of-tolerance error, avoiding blind investigation. The process to be controlled and the out-of-tolerance data are fed back to the terminal in a synchronized manner, guiding the staff to adjust the process parameters in a targeted manner, forming a closed loop of abnormal detection - process traceability - parameter optimization, which effectively reduces the occurrence of out-of-tolerance problems and improves product qualification rate and production efficiency.

[0026] A method for detecting the runout of a drive shaft, comprising: Step 1: For the type of drive shaft to be inspected for the first time, collect multi-dimensional basic data, filter the core runout detection dimension set, calculate the qualified judgment interval of the core dimensions, set the update cycle of the qualified judgment interval and the update trigger condition of the core dimension set, and build a drive shaft runout detection adaptation database; for the drive shafts to be inspected that are recorded in the database, retrieve their core runout detection dimension set and the qualified judgment interval of each core dimension for inspection; Step 2: Determine the pass / fail status of each core runout dimension and the drive shaft. Non-conforming products are re-inspected. If the re-inspection is successful, it is determined that the first inspection was a false positive. Record the false positive time-deviation data of the core dimensions and plot the change curve. Extract data using a sliding window, calculate the self-inspection evaluation value, and start the self-inspection of the corresponding testing instrument for the core dimension if the threshold is exceeded. Step 3: For each core runout detection dimension, select drive shafts that still fail the re-inspection and mark them as confirmed out-of-tolerance parts, and draw the out-of-tolerance deviation time change curve; calculate the process abnormality judgment value, and if it exceeds the threshold, the process is judged to be abnormal; construct a core dimension-processing process mapping table, determine the process to be controlled, and adjust the process parameters.

[0027] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A runout detection system for a propeller shaft, characterized by: Comprise: Multi-dimensional detection adaptation module: for the first time, the type of the transmission shaft to be detected, collect multi-dimensional basic data, screen the core jump detection dimension set, calculate the qualified judgment interval of the core dimension, set the update period of the qualified judgment interval and the update trigger condition of the core dimension set, and construct the transmission shaft jump detection adaptation database; for the transmission shaft to be detected recorded in the database, call the core jump detection dimension set and each core dimension qualified judgment interval for detection; Detection self-checking module: judge the core jump dimension and the transmission shaft qualification, and recheck the unqualified products; if the recheck is qualified, it is judged that the first detection is misdetected, the core dimension misdetected time-bias data is recorded, and the change curve is drawn; the data is extracted by sliding window, the self-checking evaluation value is calculated, and the self-checking of the corresponding detector of the core dimension is started when the threshold value is exceeded; Abnormality tracing module: for each core jump detection dimension, screen the transmission shafts that are still unqualified after rechecking and mark them as confirmed out-of-tolerance parts, and draw the out-of-tolerance bias time change curve; Calculate the process abnormality judgment value, and if the threshold value is exceeded, determine that the process is abnormal; construct the core dimension-machining process mapping table, determine the process to be controlled, and control the process parameters.

2. A runout detection system for a propeller shaft as set forth in claim 1, characterized in that: The specific process of screening the core jump detection dimension set is: For the first time, the type of the transmission shaft to be detected, collect multi-dimensional basic data; Get all potential jump detection dimensions of the transmission shaft to be detected, and organize them into a jump detection dimension set; For each potential detection dimension in the set, combine the demand weight of the potential detection dimension according to the application scene of the transmission shaft to be detected, the correlation between the potential detection dimension and the basic parameters of the service working condition of the transmission shaft to be detected, the proportion of faults caused by the potential detection dimension out-of-tolerance in the industry social application big data, the influence degree of extreme parameters of the service working condition on the potential detection dimension out-of-tolerance, and the preset weight coefficient to analyze the importance coefficient of the potential detection dimension; Compare the importance coefficient of each potential detection dimension with the corresponding preset threshold value, and mark the core jump detection dimension if it is not lower than the threshold value; organize all core jump detection dimensions to form a core jump detection dimension set.

3. A runout detection system for a propeller shaft as set forth in claim 2 wherein: The specific process of calculating the qualified judgment interval of the core dimension is: For each core jump detection dimension in the core jump detection dimension set, combine the corresponding general industrial standard reference value of the core jump detection dimension, the structure parameters related to the core dimension of the transmission shaft to be detected, the basic parameters of the service working condition related to the core dimension of the transmission shaft to be detected, the extreme parameters of the service working condition related to the core dimension of the transmission shaft to be detected, and the preset safety coefficient and the preset correction coefficient to determine the qualified judgment interval of the core jump detection dimension.

4. A runout detection system for a propeller shaft as set forth in claim 3 wherein: The specific process of setting the update period of the qualified judgment interval and the update trigger condition of the core dimension set is: For each core jump detection dimension, combine the fluctuation amplitude of the dimension qualified judgment value in the past six months, the importance change rate of the dimension in the industry demand, the average design iteration period of the transmission shaft to be detected adaptation product, and the preset weight coefficient to obtain the stability coefficient of the core jump detection dimension; A plurality of stability coefficient intervals are set, and each stability coefficient interval is preset to correspond to a qualified judgment interval update period. The stability coefficient of the core bounce detection dimension is matched with the intervals, and the corresponding update period is output. The period is used as the interval time from the current to the next update of the dimension qualified judgment interval. When the period is reached, the dimension qualified judgment interval is updated. The qualified judgment interval update periods of all dimensions in the core bounce detection dimension set are calculated, the maximum value is selected, and the update reference time is marked. When the update reference time is reached, the standard deviation value of the deviation value of each core bounce detection dimension in the core bounce detection dimension set of the to-be-tested transmission shaft is calculated and recorded as the core bounce detection dimension set stability index. If the core bounce detection dimension set stability index of the to-be-tested transmission shaft is greater than or equal to the corresponding threshold value, the core bounce detection dimension set and the qualified judgment interval of each core bounce detection dimension are updated.

5. A system for detecting runout of a propeller shaft as defined in claim 4 wherein: The specific process of detecting the recorded core bounce detection dimension set and the qualified judgment interval of each core dimension of the to-be-tested transmission shaft is as follows: The type information of all to-be-tested transmission shafts, the core bounce detection dimension set corresponding to the type, the qualified judgment interval of each core detection dimension, the qualified judgment interval update period of each core detection dimension, and the core bounce detection dimension set update reference time are sorted to construct a transmission shaft bounce detection adaptation database. For the recorded to-be-tested transmission shaft in the transmission shaft bounce detection adaptation database, the core bounce detection dimension set and the qualified judgment interval of each core detection dimension of the to-be-tested transmission shaft are retrieved from the transmission shaft bounce detection adaptation database and sent to the detection module.

6. A system for detecting runout of a propeller shaft as defined in claim 5, wherein: The specific process of recording the core dimension mis-detection time-bias data and drawing a change curve is as follows: For each to-be-tested transmission shaft, the first detection value of each core bounce detection dimension is obtained, and the qualified judgment interval corresponding to the dimension is retrieved. If the first detection value is within the qualified judgment interval, the core bounce detection dimension is determined to be qualified, otherwise, the to-be-tested transmission shaft is determined to be an unqualified product. All unqualified products are transported to the re-inspection end, and the precision detection amount is performed on the unqualified core detection dimension of the unqualified product to obtain the re-inspection value of the core detection dimension. If the re-inspection value is within the qualified judgment interval, the first detection of the dimension is determined to be mis-detection, and the mis-detection bias value is calculated. For each core detection dimension, the occurrence time of each mis-detection is recorded to form a data pair of mis-detection occurrence time-mis-detection bias value. A two-dimensional rectangular coordinate system is constructed with the horizontal axis representing time and the vertical axis representing mis-detection bias value. All mis-detection occurrence time-mis-detection bias value data pairs are marked as data points in the coordinate system, and all adjacent data points are connected in chronological order to generate a mis-detection bias time change curve of the core detection dimension.

7. A runout detection system for a propeller shaft as set forth in claim 6 wherein: The specific process of calculating the self-test evaluation value and starting the self-detection of the core dimension corresponding detector when the threshold value is exceeded is as follows: presetting a sliding detection window, extracting three types of data from the false detection deviation time curve in the window: the maximum false detection deviation in the window, the average false detection deviation in the window, and the false detection frequency in the window; meanwhile, from the false detection deviation time curve corresponding to all false detection data of the core bounce detection dimension, extracting corresponding three types of historical data: the maximum historical false detection deviation, the average historical false detection deviation, and the historical false detection frequency; combining a preset weight coefficient, analyzing the self-detection evaluation value of the core bounce detection dimension; comparing the self-detection evaluation value of the core bounce detection dimension with a preset threshold value, if the self-detection evaluation value is greater than or equal to the preset threshold value, starting the self-detection mode of the detector corresponding to the core bounce detection dimension. The specific process of drawing the out-of-tolerance deviation time curve is as follows:

8. A runout detection system for a propeller shaft as set forth in claim 7 wherein: For each core bounce detection dimension, screening the to-be-inspected transmission shafts that are still judged as unqualified after re-inspection, marking this type of transmission shaft as a confirmed out-of-tolerance piece; calculating the out-of-tolerance deviation value of the core bounce detection dimension in the confirmed out-of-tolerance piece; recording the out-of-tolerance occurrence time of the core detection dimension in each confirmed out-of-tolerance piece, and forming a data pair of out-of-tolerance occurrence time-out-of-tolerance deviation value; establishing a two-dimensional rectangular coordinate system: the horizontal axis represents time, and the vertical axis represents the out-of-tolerance deviation value; all out-of-tolerance occurrence time-out-of-tolerance deviation value data pairs are marked as abnormal feature points in the coordinate system, and all adjacent abnormal feature points are connected in sequence by a curve to generate the out-of-tolerance deviation time curve of the core detection dimension. The specific process of establishing a core dimension-machining process mapping table, determining a to-be-controlled process, and controlling a process parameter is as follows:

9. A runout detection system for a propeller shaft as set forth in claim 8 wherein: presetting a sliding analysis window, extracting three types of data from the out-of-tolerance deviation time curve in the window: the maximum out-of-tolerance deviation in the window, the average out-of-tolerance deviation in the window, and the out-of-tolerance frequency in the window; meanwhile, from the out-of-tolerance deviation time curve corresponding to all out-of-tolerance data of the core detection dimension, extracting corresponding three types of historical data: the maximum historical out-of-tolerance deviation, the average historical out-of-tolerance deviation, and the historical out-of-tolerance frequency; combining a preset weight coefficient, analyzing the process abnormality judgment value of the core detection dimension; if the process abnormality judgment value is greater than or equal to a preset threshold value, it is determined that the machining process associated with the core detection dimension is abnormal; establishing a core detection dimension-associated machining process mapping table, each core detection dimension in the table corresponds to an associated machining process; putting the core detection dimension judged as abnormal into the above mapping table, outputting the corresponding machining process and marking it as a to-be-controlled process; sending the to-be-controlled process name and the out-of-tolerance deviation data in the window to the personnel terminal, after the terminal receives the information, the working personnel checks and controls the process parameter of the to-be-controlled process.

10. A bounce detection method of a transmission shaft, applied to a bounce detection system of a transmission shaft according to any one of claims 1-9, comprising: ​ Step one: for the first time to carry out the detection of the transmission shaft type, collect multi-dimensional basic data, screen the core jump detection dimension set, calculate the core dimension qualified judgment interval, set the update cycle and set the update trigger condition, and build an adaptive database; for the recorded transmission shaft, call the core dimension and the qualified interval for detection; Step two: determine the eligibility of each core jump detection dimension and the transmission shaft, and recheck the unqualified products; if the recheck is qualified, it is determined that the first inspection is a false inspection, the data is recorded and a curve is drawn, the data is extracted by means of a sliding window, a self-inspection evaluation value is calculated, and if it is over the threshold value, the corresponding detection instrument is started; Step three: for each core detection dimension, screen out the transmission shafts that are still unqualified after rechecking, mark them as confirmed out-of-tolerance parts, draw an out-of-tolerance curve and calculate a process abnormality judgment value, and if it is over the threshold value, determine that the process is abnormal; Feedback to the process to be controlled, send information to dispatch personnel to control process parameters.

Images