A method for calibrating the metrological performance of a multi-functional tester for road materials

Abstract

This application provides a metrological performance calibration method for a multifunctional testing machine for road materials, comprising: acquiring raw measurement data of the testing machine under different test conditions, wherein the raw measurement data includes force measurement data, displacement measurement data, and modulus test data; comparing and calibrating the force measurement data using a standard force measuring unit to determine the relative indication error and repeatability of the force measurement; comparing and calibrating the displacement measurement data using a precision length reference to determine the displacement measurement indication error; comparing and calibrating the modulus test data using a combination of the standard force measuring unit and a verification unit to determine the modulus test deviation; and constructing a set of metrological performance parameters for the testing machine based on the relative indication error of the force measurement, repeatability, displacement measurement indication error, and modulus test deviation, and determining the overall measurement accuracy.

Description

Technical Field

[0001] This invention relates to the field of interdisciplinary technology of engineering material testing and metrological calibration, and in particular to a metrological performance calibration method for a multifunctional testing machine for road materials. Background Technology

[0002] In the calibration process of the metrological characteristics of multifunctional testing machines for road materials, the calibration of force measurement repeatability faces severe challenges, especially in maintaining a high degree of consistency in loading conditions under complex loading environments. Road material testing machines are commonly used to simulate the dynamic fatigue and strength testing of materials such as asphalt mixtures or cement-stabilized crushed stone under vehicle loads. These tests require accurate recording of force values ​​under long-term, high-frequency cyclic loading. However, after long-term operation, the loading system often exhibits slight deviations. For example, mechanical components such as hydraulic cylinders or servo motors may experience decreased displacement accuracy due to cumulative wear, or temperature fluctuations may cause changes in oil viscosity, resulting in subtle differences in the application of the same standard force value, such as 10 kN, each time. Furthermore, external environmental disturbances such as laboratory floor vibration or unstable air conditioning airflow can subtly affect the smoothness of the loading path. These factors combined result in imperceptible fluctuations in the actual force values ​​during repeated tests, typically within a relative error range of 0.1% to 0.5%, yet sufficient to distort the assessment results of material modulus or fatigue life. Moreover, even slight non-uniformity in loading rate exacerbates the problem. In standard repeatability tests, the system is required to accelerate to peak force at a constant rate, such as 1 mm / min. However, in practice, the system may accelerate slightly faster in the first half and slower in the second half. This unevenness often stems from the lag in the response of the control valves. Especially in high-frequency tests, such as dozens of cycles per minute, the lag in system response time accumulates and amplifies errors, causing a shift in the peak force recording time, thus affecting the synchronicity and accuracy of the readings. For example, in a scenario simulating highway rutting tests, this interference may cause the standard deviation of the same force value applied 10 times to exceed the allowable value in the calibration specification. The limitations of existing calibration procedures further highlight the technical contradictions. On the one hand, performance is evaluated solely through simple statistics of multiple force values, such as the average and repeatability standard deviation, while ignoring real-time consistency monitoring of loading conditions such as rate curves and displacement trajectories, failing to capture dynamic deviations during each application process. On the other hand, while the testing machine covers multiple measurement ranges, such as the low range of 0-50kN for small crack propagation testing and the high range of 50-200kN for overall pavement load testing, significant differences exist in repeatability. Noise amplification is stronger at low ranges, while at high ranges it is more susceptible to inertia. Current methods do not differentiate between these ranges, leading to calibration results that are out of sync with real-world applications. In fields where pavement material performance evaluation heavily relies on force stability, such as high-precision shear tests on airport runway materials, even minute errors can cause deviations in test results, thus affecting the safety margin of engineering designs. This contradiction not only challenges the reliability of calibration but also restricts the long-term performance assurance of the testing machine under complex operating conditions. Summary of the Invention

[0003] This invention provides a metrological performance calibration method for a multifunctional testing machine for road materials, mainly including: The testing machine acquires raw measurement data under different test conditions, including force measurement data, displacement measurement data, and modulus test data. The force measurement data is compared and calibrated using a standard force measuring unit to determine the relative indication error and repeatability of the force measurement. The displacement measurement data is compared and calibrated using a precision length reference to determine the displacement measurement indication error. The modulus test data is compared and calibrated using a combination of the standard force measuring unit and a verification unit to determine the modulus test deviation. Based on the relative indication error of the force measurement, repeatability, displacement measurement indication error, and modulus test deviation, a set of metrological performance parameters for the testing machine is constructed to determine the overall measurement accuracy.

[0004] Furthermore, the comparison and calibration of the force measurement data using a standard force measuring unit includes: obtaining a standard data sequence from the standard force measuring unit and a measurement data sequence from the testing machine; obtaining a point-by-point error value sequence using array subtraction; calculating a relative indication error value sequence based on the point-by-point error value sequence; marking the corresponding measurement point as an anomaly if any value in the relative indication error value sequence exceeds a preset threshold; applying a multi-level standard force value sequence covering the range; collecting measurement response data from the testing machine; obtaining an interval slope parameter by fitting the relationship between the measurement response data and the standard force value for each loading interval using linear regression; marking an interval switching point as an anomaly if the difference in the slope parameter between adjacent loading intervals exceeds a preset threshold; applying the same standard force value multiple times to obtain repeated measurement data; and calculating the standard deviation and mean of the multiple force value outputs to determine the repeatability index.

[0005] Furthermore, the step of comparing and calibrating the displacement measurement data using a precision length reference includes: placing the precision length reference in the displacement detection area of ​​the testing machine to obtain displacement output data; performing time synchronization alignment processing on the displacement output data and the precision length reference; calculating an initial deviation value through subtraction; subtracting a preset offset from the initial deviation value to obtain a first deviation value; fitting multiple first deviation values ​​using linear regression to obtain a first compensation model; compensating for the displacement measurement indication error using the first compensation model to obtain a second indication error; and recording an anomaly if the second indication error exceeds a preset threshold.

[0006] Furthermore, the step of comparing and calibrating the modulus test data by combining the standard force measuring unit and the verification unit includes: integrating the standard force measuring unit and the verification unit into the loading unit of the testing machine to obtain the output data of the standard force measuring unit and the data collected by the verification unit; determining the modulus test deviation value by comparing the output data of the standard force measuring unit and the data collected by the verification unit, and fitting the modulus test deviation value with a linear regression algorithm to obtain a deviation calibration curve; applying preset loading conditions to generate a stress-strain data set, and calculating the modulus output data through a linear regression model; obtaining the force value sequence and displacement value sequence from the standard force measuring unit and combining them with structural parameters to calculate the theoretical modulus reference value, and determining the modulus test deviation by calculating the deviation between the modulus output data and the theoretical modulus reference value through subtraction.

[0007] Furthermore, the acquisition of raw measurement data of the testing machine under different test conditions includes: collecting raw data of road surface materials under different test conditions through the testing equipment, covering multiple data points of force measurement, displacement measurement and modulus testing to obtain an initial dataset; calculating the quartiles of each attribute from the initial dataset, and if the data points exceed the interquartile range, they are marked as invalid data and discarded to obtain a cleaned valid dataset; using a Kalman filter to remove noise interference from the initial displacement dataset to obtain a processed displacement data set, and if the processed displacement data set exceeds the preset measurement accuracy range, the Kalman filter is reapplied to obtain a corrected displacement data value.

[0008] Furthermore, the step of constructing a set of metrological performance parameters for the testing machine based on the relative indication error, repeatability, displacement measurement indication error, and modulus test deviation includes: calculating the relative indication error and repeatability error of the force measurement from the cleaned valid dataset as the standard deviation divided by the mean to obtain a stability index; calculating the modulus deviation from the performance evaluation results of the displacement measurement, and marking the deviation as abnormal if the deviation value exceeds a preset threshold to determine the accuracy status of the modulus test; integrating the performance evaluation results of the force measurement, displacement measurement, and modulus test to construct a set of metrological performance parameters, and using logistic regression to classify and evaluate the overall measurement accuracy to obtain a comprehensive performance level.

[0009] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a calibration method for the metrological characteristics of a multifunctional testing machine for pavement materials. Addressing the problems of traditional calibration methods where multiple metrological characteristics such as force, displacement, and modulus are independent, the calibration environment deviates significantly from actual testing conditions, and data traceability is poor, leading to insufficient overall metrological accuracy and reliability, this method integrates a standard force measurement unit, a precision length benchmark, and a verification unit into the loading unit. This constructs a unified calibration environment that highly simulates actual working conditions, enabling the verification of relative indication errors and repeatability of force values ​​under multi-level standard force application, calibration of indication errors under a precision displacement benchmark, and accurate assessment of modulus test deviations using traceable stress-strain data generated by the combination of standard force and verification units. This method organically integrates multiple metrological characteristic calibration processes, ensuring accurate and consistent data transmission. Ultimately, it significantly improves the testing machine's force measurement accuracy, repeatability stability, displacement detection accuracy, and modulus calculation reliability, providing more reliable metrological assurance for pavement material performance testing. Attached Figure Description

[0010] Figure 1 This is a flowchart illustrating a metrological performance calibration method for a multifunctional testing machine for road materials according to the present invention. Detailed Implementation

[0011] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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.

[0012] like Figure 1 The metrological performance calibration method of a multifunctional testing machine for road materials in this embodiment may specifically include: Step S101: Obtain the metrological characteristic set of the multi-functional testing machine for road materials. The metrological characteristic set includes the relative indication error of force measurement, the repeatability of force measurement, the indication error of displacement measurement, and the deviation of modulus test.

[0013] Raw data of road materials under different test conditions were collected using experimental equipment, covering multiple data points for force measurement, displacement measurement, and modulus testing. This data was stored as an initial dataset, resulting in structured test records. The initial dataset was loaded into a pandas DataFrame from these structured test records. The quartiles of each attribute were calculated. Data points exceeding 1.5 times the interquartile range were marked as invalid and removed, determining the cleaned and valid dataset. Based on the cleaned and valid dataset, the relative indication error of force measurement was calculated using NumPy, i.e., error E equals (measured value M minus true value T) divided by true value T. The repeatability error was calculated as the standard deviation divided by the mean, obtaining the error distribution and stability index for each test condition to assess the metrological performance of force measurement. From the cleaned and valid dataset, the indication error of displacement measurement was calculated using NumPy. If the error value exceeded the preset tolerance range, it was recorded as an outlier, yielding the performance evaluation results of displacement measurement. Based on the performance evaluation results of displacement measurement and the metrological performance of force measurement, NumPy is used to calculate relevant data for modulus testing. Specifically, the modulus deviation D equals (calculated modulus C minus standard modulus S) divided by the standard modulus S. If the deviation exceeds a preset threshold, it is marked as an abnormal deviation, thus determining the accuracy status of the modulus test. By integrating the performance evaluation results and accuracy status of force measurement, displacement measurement, and modulus testing, a set of metrological performance parameters is constructed. Scikit-learn's Logistic Regression is used to classify and evaluate the overall measurement accuracy. First, feature vectors including error distribution, deviation values, and labels representing accuracy levels are prepared. Then, a model is fitted to obtain the comprehensive performance level. Based on the comprehensive performance level, a record of the metrological performance parameter set of the multi-functional testing machine for road materials is generated and stored as a structured document to determine the reliability of the overall test data.

[0014] For example, in the metrological performance evaluation of a multi-functional testing machine for road materials, the collection of the initial dataset is a crucial step. Suppose 100 sets of data are collected using the testing equipment, covering force, displacement, and modulus tests, with each set including measurements under multiple test conditions. This data is organized into structured test records, containing fields such as force measurements, displacement measurements, and calculated modulus. After being loaded into the data analysis tool, the quartiles of each attribute are first calculated, and outliers exceeding 1.5 times the interquartile range are removed. For instance, if five points in the force data are outside the range, these are marked as invalid and removed, resulting in 95 cleaned and valid datasets. This cleaning method effectively improves data reliability and lays the foundation for subsequent analysis.

[0015] For example, in the metrological performance evaluation of force measurement, the relative indication error is calculated using a cleaned dataset. Assuming a measured force of 1020 Newtons under certain test conditions, a true force of 1000 Newtons, and a relative error of 2%, the error distribution is calculated using multiple sets of data. Simultaneously, the repeatability error is obtained as the ratio of the standard deviation to the mean; assuming a standard deviation of 5 Newtons and a mean of 1000 Newtons, the repeatability error is 0.5%. This error distribution and stability index reflect the accuracy and consistency of force measurement, helping to determine whether the equipment meets metrological requirements.

[0016] For example, in the performance evaluation of displacement measurement, the indicated error is calculated and compared with a preset tolerance range. Assuming the tolerance range is ±0.1 mm, if a measurement error is 0.15 mm, and this error exceeds the range, it is recorded as an outlier. After statistically analyzing the proportion of all outliers, if the proportion of outliers is less than 5%, the displacement measurement performance is considered good. This method can intuitively reflect the stability of the equipment in displacement testing.

[0017] For example, in the accuracy assessment of modulus testing, modulus deviation is calculated. Assuming the calculated modulus is 210 MPa, the standard modulus is 200 MPa, and the deviation is 5%, if the preset threshold is 3%, it is marked as an abnormal deviation. By statistically analyzing the proportion of abnormal deviations, the accuracy status of the modulus test is determined. This assessment method helps to identify potential problems with equipment during complex testing.

[0018] For example, in comprehensive performance evaluation, force error distribution, the proportion of displacement anomalies, and modulus deviation are used as feature vectors, and accuracy level is used as a label to construct a classification model. Assuming the feature vectors show low force error, few displacement anomalies, and small modulus deviation, the model predicts an excellent comprehensive performance level. This classification evaluation integrates multi-dimensional data to provide an overall performance assessment.

[0019] For example, when generating a set of metrological performance parameters, the above evaluation results are stored as a structured document, including information such as error distribution, outlier ratio, and overall rating. If the overall rating is excellent, the overall test data is considered reliable. This recording method facilitates subsequent traceability and equipment calibration, ensuring the credibility of the test results.

[0020] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. 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 present invention, and all such modifications and substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for calibrating the metrological performance of a multifunctional testing machine for road materials, characterized in that, include: Acquire raw measurement data of the testing machine under different test conditions. The raw measurement data includes force measurement data, displacement measurement data, and modulus test data. The force measurement data is compared and calibrated using a standard force measurement unit to determine the relative indication error and repeatability of the force measurement. The displacement measurement data is compared and calibrated using a precise length reference to determine the displacement measurement indication error. The modulus test data is compared and calibrated using the standard force measuring unit and the verification unit to determine the modulus test deviation. Based on the relative indication error of the force measurement, repeatability, indication error of the displacement measurement, and deviation of the modulus test, a set of metrological performance parameters for the testing machine is constructed to determine the overall measurement accuracy.

2. The method as described in claim 1, characterized in that, The step of comparing and calibrating the force measurement data using a standard force measuring unit includes: Obtain standard data sequences from the standard force measuring unit and measurement data sequences from the testing machine; The point-by-point error value sequence is obtained by array subtraction. The relative indication error value sequence is calculated based on the point-by-point error value sequence. If any value in the relative indication error value sequence exceeds the preset threshold, the corresponding measurement point is marked as an abnormal point. A multi-level standard force value sequence covering the range is applied, and measurement response data is collected from the testing machine. The relationship between the measurement response data and the standard force value of each loading interval is fitted by linear regression to obtain the interval slope parameter. If the difference between the slope parameters of adjacent loading intervals exceeds a preset threshold, the interval switching point is marked as abnormal. Repeated measurement data are obtained by applying the same standard force value multiple times, and the standard deviation and mean of the multiple force value outputs are calculated to determine the repeatability index.

3. The method as described in claim 1, characterized in that, The step of comparing and calibrating the displacement measurement data using a precision length reference includes: The precision length reference is placed in the displacement detection area of ​​the testing machine to obtain displacement output data; The displacement output data is time-synchronized and aligned with the precision length reference. An initial deviation value is calculated by subtraction. A preset offset is subtracted from the initial deviation value to obtain the first deviation value. A first compensation model is obtained by fitting multiple first deviation values ​​using linear regression. The displacement measurement indication error is then compensated using the first compensation model to obtain a second indication error. If the second indication error exceeds a preset threshold, an anomaly point is recorded.

4. The method as described in claim 1, characterized in that, The step of comparing and calibrating the modulus test data by combining the standard force measuring unit and the verification unit includes: The standard force measuring unit and the verification unit are integrated into the loading unit of the testing machine to obtain the output data of the standard force measuring unit and the data collected by the verification unit; The modulus test deviation value is determined by comparing the output data of the standard force measuring unit with the data collected by the verification unit, and the deviation calibration curve is obtained by fitting the modulus test deviation value with a linear regression algorithm. Preset loading conditions are applied to generate a set of stress-strain data, and the modulus is calculated and output data is generated through a linear regression model. The theoretical modulus reference value is calculated by combining the force value sequence and displacement value sequence obtained from the standard force measuring unit with the structural parameters. The deviation between the modulus output data and the theoretical modulus reference value is determined by subtraction.

5. The method as described in claim 1, characterized in that, The acquisition of raw measurement data of the testing machine under different test conditions includes: The initial dataset was obtained by collecting raw data of pavement materials under different test conditions using test equipment, covering multiple data points of force measurement, displacement measurement and modulus testing; Calculate the quartile of each attribute from the initial dataset. If a data point is outside the interquartile range, it is marked as invalid data and removed to obtain a cleaned valid dataset. A Kalman filter is used to remove noise from the initial displacement dataset to obtain a processed displacement data set. If the processed displacement data set exceeds the preset measurement accuracy range, the Kalman filter is reapplied to obtain a corrected displacement data value.

6. The method as described in claim 1, characterized in that, The set of metrological performance parameters for the testing machine is constructed based on the relative indication error of the force measurement, repeatability, indication error of the displacement measurement, and the deviation of the modulus test, including: The stability index is obtained by calculating the relative indication error and repeatability error of force measurement from the cleaned effective dataset, dividing the standard deviation by the mean; The modulus deviation is calculated from the performance evaluation results of displacement measurement. If the deviation value exceeds the preset threshold, it is marked as an abnormal deviation to determine the accuracy status of the modulus test. The performance evaluation results of force measurement, displacement measurement and modulus testing are integrated to construct a set of metrological performance parameters. Logistic regression is used to classify and evaluate the overall measurement accuracy to obtain the comprehensive performance level.

Images