A calibration error correction method and system for a drop hammer type deflectometer

Abstract

The application relates to the technical field of data error correction, in particular to a calibration error correction method and system of a drop hammer type deflectometer, which comprises the following steps: obtaining deflection values of a displacement sensor of the drop hammer type deflectometer to form a deflection time sequence, calculating deflection abnormal values of each deflection value, obtaining each deflection right sequence of each deflection value, further obtaining deflection continuation characteristic values of each deflection value, obtaining load transfer weights of each deflection value through the correlation degree between local deflection sequences of the same bit sequence deflection value in adjacent deflection time sequences, obtaining road loose influence degrees of each deflection value in combination with the deflection continuation characteristic values of the corresponding deflection values, further obtaining deflection vibration characteristic values of each deflection value and smoothing window scales of each deflection value, and performing smoothing treatment on each deflection time sequence to correct deflection data errors of the drop hammer type deflectometer. The application can improve the calibration error correction precision of the drop hammer type deflectometer.

Description

Technical Field

[0001] This application relates to the field of data error correction technology, specifically to a calibration error correction method and system for a falling weight deflectometer. Background Technology

[0002] The Falling Weight Deflectometer (FWD) measures the dynamic deflection and deflection basin generated under impact loads by simulating the impact of moving vehicle wheels on the road surface. This helps to assess the structural bearing capacity and stability of the roadbed and pavement, and facilitates the timely detection of potential safety hazards on the road surface.

[0003] The calibration of falling weight deflectometers mainly involves measuring deflection values ​​using displacement sensors and measuring falling weight impact load values ​​using load sensors. This requires periodic calibration using high-precision sensors. However, in actual measurements, traffic interference inevitably introduces errors into the falling weight deflectometer readings. Therefore, falling weight deflectometers typically use data acquisition software to correct for these errors. However, due to road surface defects such as looseness and voids, the data acquisition software's curve fitting method for error correction struggles to distinguish between deflection data from road defects and traffic interference. This can lead to over-smoothing of deflection data related to road defects, reducing the accuracy of pavement structure parameter inversion and latent defect diagnosis. Summary of the Invention

[0004] To address the aforementioned technical problems, the purpose of this application is to provide a method and system for correcting the calibration error of a falling weight deflectometer. The specific technical solution adopted is as follows:

[0005] This application provides a method for correcting the calibration error of a falling weight deflectometer, comprising the following steps:

[0006] Obtain the deflection values ​​from the displacement sensor of the falling weight deflectometer to form a deflection procedure sequence;

[0007] Based on the average level of each deflection value in all deflection time program sequences, the degree of difference between each deflection value in each deflection time program sequence and the discrete range of each deflection value in all deflection time program sequences, the deflection outlier value of each deflection value in each deflection time program sequence is obtained. The right-hand sequence of each deflection value is extracted. After classifying each right-hand sequence of deflection values, the number of each deflection value and the difference of the corresponding position value of each deflection value are used to obtain the deflection continuity characteristic value of each deflection value in each deflection time program sequence.

[0008] By obtaining all deflection values ​​adjacent to each deflection value, the local deflection sequence of each deflection value is obtained. Based on the correlation between the local deflection sequences of the same deflection value in the program sequence of adjacent deflection values, the load transfer weight of each deflection value in the program sequence of each deflection value is obtained. Combined with the deflection continuity characteristic value of each deflection value, the road loosening influence degree of each deflection value in the program sequence of each deflection value is obtained.

[0009] Based on the degree of disorder in the local deflection sequence of each deflection value and the deviation of each deflection value in the local deflection sequence, the deflection vibration characteristic value of each deflection value in the deflection time program is obtained. Then, combined with the deflection anomaly value of each deflection value and the influence of road loosening, the smoothing window scale of each deflection value in the deflection time program is obtained. The smoothing algorithm is then used to smooth the deflection time program to correct the deflection data error of the falling weight deflectometer.

[0010] Preferably, the method for calculating the deflection anomaly value of each deflection value in the program sequence for each deflection is as follows:

[0011]

[0012] In the formula, w i,j σ represents the deflection anomaly value of the j-th deflection value in the program sequence for the i-th deflection; j h is the standard deviation of the j-th deflection value in the sequence of all deflections; i,j The j-th deflection value in the program column when the i-th deflection occurs; It is the mean of the j-th deflection value in the program sequence for all deflection times.

[0013] Preferably, the method for extracting the right-hand sequence of each deflection value is as follows:

[0014] The deflection values ​​of each deflection value in the program column at each deflection time are arranged into a right-hand sequence of each deflection value, which is the number of adjacent positions to the right of each deflection value. When the number of deflection value data to the right is less than the first preset number, the right-hand sequence of deflection values ​​is all the deflection value data to the right of each deflection value in the program column at each deflection time.

[0015] Preferably, the method for calculating the deflection continuity characteristic value of each deflection value in the program sequence for each deflection is as follows:

[0016]

[0017] In the formula, τ i,j Let be the characteristic value of the deflection continuation of the j-th deflection value in the sequence at the i-th deflection; exp(·) is an exponential function with the natural constant e as its base; m i,jd represents the number of other deflection values ​​within the data level of the right-hand deflection sequence of the j-th deflection value, where each data level is obtained by classifying the right-hand deflection sequence using the natural breakpoint algorithm; i,j is the mean of the differences between the position values ​​of all deflection proximity points in the rightward sequence of the deflection value of the j-th deflection value and the corresponding position value of the j-th deflection value. Here, each deflection proximity point in each rightward sequence of each deflection value is another deflection value within the data level of each deflection value in the rightward sequence of each deflection value; M is the number of elements in the rightward sequence of the deflection.

[0018] Preferably, the method for obtaining the local deflection sequence of each deflection value is further as follows:

[0019] Centered on each deflection value in the program sequence, obtain the second preset number of deflection values ​​closest to this center, and arrange them in ascending order to obtain the local deflection sequence of each deflection value.

[0020] Preferably, the method for calculating the load transfer weight of each deflection value in the program sequence for each deflection is as follows:

[0021]

[0022] In the formula, G i,j For the i-th deflection, assign the load transfer weight to the j-th deflection value in the program sequence; G i-1,j For the (i-1)th deflection, the load transfer weight is assigned to the j-th deflection value in the program sequence; X i,j X is the local deflection sequence of the j-th deflection value in the program sequence when the i-th deflection occurs. i+1,j is the local deflection sequence of the j-th deflection value in the program sequence at the (i+1)-th deflection; pea(·) is the Pearson correlation coefficient.

[0023] Preferably, the road loosening influence degree of each deflection value in the program sequence for each deflection is calculated using the following formula:

[0024]

[0025] In the formula, P i,j τ represents the road loosening effect of the j-th deflection value in the program sequence for the i-th deflection; i,j G is the deflection continuity characteristic value of the j-th deflection value in the program sequence when the i-th deflection occurs; i,j The load transfer weight is the load for the j-th deflection value in the program sequence at the i-th deflection, and α is a constant to avoid the denominator being 0.

[0026] Preferably, the formula for calculating the deflection vibration characteristic value of each deflection value in the program sequence for each deflection is:

[0027] R i,j =ρi,j ×ent i,j ;

[0028] In the formula, R i,j ρ is the deflection vibration characteristic value of the j-th deflection value in the program sequence for the i-th deflection; i,j The number of zero points in the line graph of the local deflection deviation sequence is defined as follows: the local deflection deviation sequence is obtained by sorting the differences between each deflection value and the mean of all deflection values ​​in the local deflection sequence. Adjacent deflection values ​​in the local deflection deviation sequence are then connected to obtain the line graph of the local deflection deviation sequence. i,j The information entropy is the local deflection deviation sequence corresponding to the j-th deflection value in the program sequence when the i-th deflection occurs.

[0029] Preferably, the method for calculating the smoothing window scale of each deflection value in the program sequence for each deflection is as follows:

[0030] λ i,j =fun[W×(s) i,j +β)];

[0031] In the formula, λ i,j is the smoothing window scale for the j-th deflection value in the sequence at the i-th deflection; W is the initial window scale for the Loess smoothing algorithm; s i,j Let be the deflection smoothing score of the j-th deflection value in the program sequence for the i-th deflection time. Here, the ratio of the deflection anomaly value to the road loosening influence degree is used as the deflection noise confidence degree. The deflection vibration characteristic value and the deflection noise confidence degree of each deflection value in the program sequence for the i-th deflection time are combined to form the deflection smoothing evaluation vector of each deflection value in the program sequence for the i-th deflection time. The deflection smoothing score is obtained by the Topsis algorithm. β is the parameter tuning factor. fun[·] is to obtain the closest odd number.

[0032] This application also provides a calibration error correction system for a falling weight deflectometer, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described calibration error correction methods for a falling weight deflectometer.

[0033] As can be seen from the above, the calibration error correction method and system for a falling weight deflectometer provided in this application have at least the following beneficial effects:

[0034] This application calibrates and verifies displacement and load sensors using high-precision sensors at preset intervals, determining calibration coefficients to effectively improve the accuracy of falling weight deflectometer measurements and ensure their normal operation. Based on the continuity characteristics and load transfer characteristics of deflection data in areas of road looseness defects, deflection noise confidence levels are obtained to identify these defects. This helps to minimize deflection anomalies caused by road looseness defects, preventing excessive smoothing of road defect deflection data by data acquisition software. Furthermore, it effectively identifies abnormal deflection data caused by noise interference, leading to a more accurate assessment of road bearing capacity and stability. By combining this with a Loess smoothing algorithm using a dynamic window scale, noise interference and traffic vibration errors are eliminated while preserving local details of both road looseness defects and normal deflection values, improving data correction efficiency and quality. This provides superior data support for subsequent road modulus inversion and hazard identification. Attached Figure Description

[0035] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 A flowchart illustrating the steps of a calibration error correction method for a falling weight deflectometer provided in this application;

[0037] Figure 2 A flowchart illustrating the steps of the method for obtaining the road loosening impact provided in this application. Detailed Implementation

[0038] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a calibration error correction method and system for a falling weight deflectometer proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0039] Unless otherwise specified and limited, terms such as “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a circuit structure, article, or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the article or device that includes said element. Furthermore, the term “and / or” as used herein includes any and all combinations of one or more of the associated listed items. All technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0040] The following, in conjunction with the accompanying drawings, details the specific scheme of the calibration error correction method and system for a falling weight deflectometer provided in this application.

[0041] Please see Figure 1 The document illustrates a flowchart of a calibration error correction method for a falling weight deflectometer according to an embodiment of this application, including the following steps:

[0042] Step 1: Obtain the deflection value from the displacement sensor of the falling weight deflectometer to form a deflection procedure.

[0043] The displacement sensor of the falling weight deflectometer is calibrated using a reference calibration method at preset intervals, employing a standard displacement sensor with an accuracy at least a preset multiple higher than the original displacement sensor. When constructing the reference device, a base is installed on the ground to serve as a support for the displacement sensor. The lower end of a lead screw is fixedly connected to the support, and the upper end extends into the cavity of the standard displacement sensor, ensuring the relative positions of the ground, support, and lead screw are stable, thus forming a stable measurement reference. Preferably, in this embodiment, the preset interval is 3 months, and the preset multiple is 3 times.

[0044] During the calibration of the displacement sensor in a falling weight deflectometer, the falling weight impacts the ground, causing ground deformation and resulting in a displacement of the lead screw relative to the standard displacement sensor, i.e., the deflection value. At this time, the standard displacement sensor and the standard deflectometer synchronously sense the displacement. The control host acquires the data from the standard displacement sensor and reads it through the FWD software. Additionally, a velocity sensor measures the vibration of a reference beam to correct for any interference from beam vibration on the accuracy of the calibration results. Simultaneously, during data comparison and calibration, the standard displacement sensor reading is used as a benchmark, and the readings are compared and analyzed to determine the calibration coefficient of the displacement sensor, thus achieving the calibration of the falling weight deflectometer's displacement sensor.

[0045] In addition, the load cell of the falling weight deflectometer is also calibrated at preset intervals using a standard load cell with an accuracy at least a preset multiple higher than that of the load cell. Calibration is performed on a flat, hard test surface to ensure stable load impact. Guide blocks are used to coaxially connect the load cell's load plate and the calibration load plate, ensuring complete rigid contact. During the calibration of the falling weight deflectometer load cell, the falling weight impacts the test surface. The impact load is synchronously sensed by both the load cell and the standard load cell. The control host collects data from the standard load cell, and the FWD software reads the load cell data. Simultaneously, during data comparison and calibration, the standard load cell reading is used as a benchmark, and the readings are compared and analyzed to determine the calibration coefficient of the load cell, thus achieving the calibration of the falling weight deflectometer load cell. Preferably, in this embodiment, the preset interval is 3 months, and the preset multiple is 3 times.

[0046] This embodiment uses a vehicle-mounted falling weight deflectometer on asphalt roads to test the dynamic deflection of the road surface under impact load, thereby evaluating the bearing capacity of the roadbed and pavement and identifying safety hazards. The vehicle-mounted falling weight deflectometer has a 200kg falling weight, generating a 50kN impact load, and a 300mm diameter bearing plate. The instrument has nine displacement sensors located at 0, 150, 300, 450, 600, 750, 900, 1050, and 1200mm. The testing vehicle is equipped with data acquisition software and a control system. The test locations are represented by station numbers, with a preset interval between adjacent station numbers. Preferably, in this embodiment, the preset interval is 50m.

[0047] When the engineering test vehicle reaches the test position, the station number is entered, the bearing plate is lowered, and the various sensors of the deflection detection device are deployed. Then, the load generating device is activated through the vehicle control system, causing the drop hammer to impact the bearing plate. Subsequently, the deflection time sequence of the nine displacement sensors is acquired using the vehicle data acquisition software. The deflection time sequence is a sequence composed of the deflection values ​​collected by the displacement sensors of the drop hammer deflectometer arranged in ascending order according to their position. Preferably, in this embodiment, the interval between the sequence deflection values ​​is 25 μs, and each station number is tested 5 times.

[0048] Step 2: Based on the average level of each deflection value in all deflection time sequence programs, the degree of difference between each deflection value in each deflection time sequence program and the discrete range of each deflection value in all deflection time sequence programs, obtain the deflection outlier value of each deflection value in each deflection time sequence program. Extract the right-hand deflection sequence of each deflection value. After classifying each right-hand deflection sequence, obtain the number of each deflection value and the difference in the corresponding position value of each deflection value, and obtain the deflection continuity characteristic value of each deflection value in each deflection time sequence program.

[0049] Electromagnetic interference and noise signals may be generated by the electronic components, wires and cables inside the vehicle-mounted equipment of the falling weight deflectometer. These electromagnetic interference noises may be superimposed on the useful signal output by the displacement sensor, causing the signal received by the signal processing system to be distorted. This can lead to random abrupt changes in deflection values ​​during measurement and increased dispersion of deflection data from multiple measurements.

[0050] During the testing process of the vehicle-mounted falling weight deflectometer, since each station number is tested 5 times, each displacement sensor will obtain 5 deflection time program sequences with the same interval length. Each displacement sensor is used as a target sensor, and the deflection values ​​of each deflection time program sequence in each target sensor are obtained. In this embodiment, taking one of the target sensors as an example, the mean of the j-th deflection value in all deflection time program sequences is calculated, and the standard deviation of the j-th deflection value in all deflection time program sequences is obtained.

[0051] Based on the above analysis, and considering the average level of each deflection value in all deflection time program sequences, the degree of difference between the average level of each deflection value and the deflection value in each corresponding deflection time program sequence, and combined with the discrete range of each deflection value in all deflection time program sequences, the deflection anomaly value of each deflection value in each deflection time program sequence is calculated. In this embodiment, the specific calculation formula is as follows:

[0052]

[0053] In the formula, w i,j σ represents the deflection anomaly value of the j-th deflection value in the program sequence for the i-th deflection; j h is the standard deviation of the j-th deflection value in the sequence of all deflections; i,j The j-th deflection value in the program column when the i-th deflection occurs; It is the mean of the j-th deflection value in the program sequence for all deflection times.

[0054] In particular, during repeated tests of the falling weight deflectometer, the greater the deviation ratio of the deflection value and the greater the dispersion of the deflection data from repeated tests, the more abnormal the deflection data is, and the more likely it is electromagnetic interference caused by the vehicle-mounted equipment, which distorts the deflection data received by the signal processing system.

[0055] Furthermore, in road construction projects, an excessively low asphalt-aggregate ratio, excessively large internal gaps in the mixture during paving, and loose mixture can all lead to loose defects in asphalt pavements. Moreover, the internal structure of asphalt pavement in areas with loose defects is not dense enough, and there may even be voids and weakened interlocking between particles, making it impossible to effectively transfer and disperse stress to the surrounding area. This results in increased deviations in deflection values, decreased structural stability, and large fluctuations in repeated test results, making it difficult to distinguish between noise interference and loose defect areas in deflection anomalies. In addition, noise signals caused by electromagnetic interference are superimposed on the useful signal output by displacement sensors. This noise is highly random and often causes random abrupt changes in the deflection data sequence. Furthermore, asphalt roads with loose defect areas exhibit a certain degree of elasticity, causing their deflection data to show a certain continuity over time.

[0056] Therefore, based on the temporal continuity of deflection data in the loose defect area, the continuity to the right of the i-th deflection value is observed. Starting from the j-th deflection value in the program sequence at the i-th deflection time, a first preset number of adjacent deflection values ​​are obtained to the right, forming the rightward deflection sequence of the j-th deflection value. When the number of deflection values ​​to the right is less than the first preset number, the rightward deflection sequence comprises all deflection values ​​to the right of the j-th deflection value in the program sequence at the i-th deflection time. Furthermore, the rightward deflection sequence is classified using the natural breakpoint algorithm, and the number of data levels is automatically obtained using the elbow method to group the deflection data. The number of data levels can also be set as needed. Therefore, other deflection values ​​within the data level of each deflection value in each rightward deflection sequence are considered as the deflection proximity points of each rightward deflection sequence. The number of deflection values ​​at each deflection proximity point in the rightward sequence of each deflection is counted, and the mean of the differences between the positional values ​​of all deflection proximity points in the rightward sequence of the j-th deflection value and the corresponding positional value of the j-th deflection value is calculated. Preferably, in this embodiment, the first preset number is 20, and the number of data levels is set to 3; wherein, the natural breakpoint algorithm and the elbow method are well-known techniques, and the specific process will not be described in detail.

[0057] Based on the above analysis, the number of deflection proximity points in the rightward sequence of each deflection value is counted. Combined with the differences in the positional values ​​of all deflection proximity points in the rightward sequence, the deflection continuity characteristic value of each deflection value in the program sequence is calculated. In this embodiment, the specific calculation formula is as follows:

[0058]

[0059] In the formula, τ i,j Let be the characteristic value of the deflection continuation of the j-th deflection value in the sequence at the i-th deflection; exp(·) is an exponential function with the natural constant e as its base; m i,jd represents the number of other deflection values ​​within the data level of the right-hand deflection sequence of the j-th deflection value, where each data level is obtained by classifying the right-hand deflection sequence using the natural breakpoint algorithm; i,j is the mean of the differences between the position values ​​of all deflection proximity points in the rightward sequence of the deflection value of the j-th deflection value and the corresponding position value of the j-th deflection value. Here, each deflection proximity point in each rightward sequence of each deflection value is another deflection value within the data level of each deflection value in the rightward sequence of each deflection value; M is the number of elements in the rightward sequence of the deflection.

[0060] in, This value reflects the distribution of deflection values ​​within each right-hand deflection sequence. A larger value indicates a higher proportion of deflection values ​​near their starting points within the right-hand deflection sequence, and a greater tendency for local deflection data to converge towards the initial point. This value reflects the positional distribution of deflection proximity points within each right-hand deflection sequence. A larger value indicates that the proximity points are closer to the starting point, suggesting a higher likelihood of continuity in the right-hand deflection sequence at the starting point. This also weakens the possibility of random abrupt changes in the deflection data and makes its continuity over time more significant. The deflection continuity characteristic value τ... i,j The larger.

[0061] Step 3: Obtain the local deflection sequence of each deflection value by using all deflection values ​​adjacent to each deflection value. Based on the correlation between the local deflection sequences of the same deflection value in the program sequence of adjacent deflection values, obtain the load transfer weight of each deflection value in the program sequence of each deflection value. Combined with the deflection continuity characteristic value of each deflection value, obtain the road loosening influence degree of each deflection value in the program sequence of each deflection value.

[0062] Furthermore, due to the poor elastic and strength properties of asphalt pavement in loose defect areas, when a drop weight load is applied to the pavement, the loose defect areas cannot effectively transfer and diffuse the load to the surrounding road structure, resulting in weak load transfer capacity. In this embodiment, the j-th deflection value of the sequence at the i-th deflection is taken as the center, and the deflection values ​​closest to this center by a second preset number are obtained. The local deflection sequence of the j-th deflection value is then obtained in ascending order of position. Preferably, in this embodiment, the second preset number is 20.

[0063] Based on the above analysis, and according to the correlation between the local deflection sequences of the same deflection value in adjacent deflection time sequence programs, the load transfer weight of each deflection value in each deflection time sequence program is obtained. In this embodiment, the specific calculation formula is as follows:

[0064]

[0065] In the formula, G i,j For the i-th deflection, assign the load transfer weight to the j-th deflection value in the program sequence; G i-1,jFor the (i-1)th deflection, the load transfer weight is assigned to the j-th deflection value in the program sequence; X i,j X is the local deflection sequence of the j-th deflection value in the program sequence when the i-th deflection occurs. i+1,j is the local deflection sequence of the j-th deflection value in the program sequence at the (i+1)-th deflection; pea(·) is the Pearson correlation coefficient.

[0066] In this embodiment, the deflection sequence with i=1 corresponds to the central displacement sensor. As the value of i increases, the position of the displacement sensor also increases. +1 is to ensure that the Pearson correlation coefficient is not negative. The deflection sequence with i=9 corresponds to the displacement sensor at the end of the deflection detection device. Since the distance between the end displacement sensor and the central displacement sensor is far and the deflection basin characteristics are not significant, the contribution to modulus back calculation and potential defect identification is relatively low. Therefore, in this embodiment, the load transfer weight with i=8 is used.

[0067] Furthermore, pea(·) is used to reflect the load transfer performance of each displacement sensor corresponding to the road structure when the falling weight load is applied to the road surface. The more synchronized the deflection data between adjacent local deflection sequences under the same position sequence, the smaller the deflection phase difference between adjacent road structures, and the better the load transfer performance.

[0068] Therefore, by using the deflection continuation characteristic value of each deflection value in each deflection time sequence and the load transfer weight of each deflection value in each deflection time sequence, the road loosening influence degree of each deflection value in each deflection time sequence is obtained. In this embodiment, the specific calculation formula is as follows:

[0069]

[0070] In the formula, P i,j τ represents the road loosening effect of the j-th deflection value in the program sequence for the i-th deflection; i,j G is the deflection continuity characteristic value of the j-th deflection value in the program sequence when the i-th deflection occurs; i,j The load transfer weight is the load for the j-th deflection value in the program sequence at the i-th deflection, and α is a constant to avoid the denominator being 0.

[0071] Wherein, the value of α ranges from (0, 0.1), and in this embodiment, α is taken as 0.01; G i,j Reflecting the load transfer performance of the road structure, when a drop weight load is applied to the road surface, loose defect areas cannot effectively transfer and diffuse the load to the surrounding road structure, resulting in weaker load transfer capacity and stronger deflection response hysteresis; τ i,j τ reflects the numerical continuity of the j-th deflection value in the rightward deflection sequence. i,jThe larger the value, the weaker the possibility of random mutations in the deflection data, and the more likely the deflection data anomalies are caused by road loosening defects due to structural inhomogeneity. Therefore, the greater the influence of road loosening defects on the deflection time sequence, the greater the influence of road loosening defects P. i,j The larger the value, the greater the impact. Additionally, the flowchart of the method for obtaining the impact of road loosening provided in this embodiment is as follows: Figure 2 As shown.

[0072] To enhance the construction quality of municipal road projects, promptly identify potential safety hazards, and improve the quality of falling weight deflectometer data, it is necessary to smooth deflection data affected by noise interference while retaining deflection data indicating road loosening defects. Therefore, the ratio of the deflection anomaly value to the road loosening influence degree in each deflection value's program sequence is calculated and used as the deflection noise confidence level for each deflection value in each program sequence, reflecting the likelihood of noise interference affecting the deflection value.

[0073] Therefore, excessively large internal gaps and loose aggregate during paving can lead to road loosening defects. Due to uneven road structure, areas with loosening defects are prone to sudden changes in local deflection, weakening load transfer capacity and generating abnormal deflection data. Deflection noise confidence level can identify road loosening defects, minimize deflection anomalies caused by these defects, and effectively identify abnormal deflection data caused by noise interference. Deflection values ​​with higher deflection noise confidence levels should be smoothed to eliminate noise interference errors and improve the quality of data measured by falling weight deflectometers.

[0074] Step 4: Based on the degree of disorder in the local deflection sequence of each deflection value and the deviation of each deflection value in the local deflection sequence, the deflection vibration characteristic value of each deflection value in the deflection time program is obtained. Then, combined with the deflection anomaly value of each deflection value and the influence of road loosening, the smoothing window scale of each deflection value in the deflection time program is obtained. The smoothing algorithm is then used to smooth the deflection time program to correct the deflection data error of the falling weight deflectometer.

[0075] During actual testing with a falling weight deflectometer, traffic interference is a significant factor that can easily lead to measurement errors in deflection values ​​on the road. For example, there may be other vibration sources around the test site, such as other heavy machinery operations or vehicles passing by. The vibrations generated by these sources will be transmitted to the vehicle-mounted falling weight deflectometer through the ground or air, causing the displacement sensor to shake up and down, thus producing vibration errors in the deflection data.

[0076] Therefore, the mean of all deflection values ​​in the local deflection sequence of the j-th deflection value in the program sequence is calculated when the i-th deflection occurs. The difference between each deflection value in the local deflection sequence and its corresponding mean is also calculated. These differences are then sorted in ascending order according to the position of the deflection values ​​to obtain the local deflection deviation sequence for the j-th deflection value. Furthermore, adjacent deflection values ​​in the local deflection deviation sequence are connected to obtain a line graph, and the number of zero points crossed in the line graph is counted.

[0077] Based on the above analysis, and according to the degree of disorder in the local deflection sequence of each deflection value, as well as the deviation of each deflection value in the local deflection sequence, the deflection vibration characteristic value of each deflection value in the program sequence at each deflection time is obtained. In this embodiment, the specific calculation formula is as follows:

[0078] R i,j =ρ i,j ×ent i,j ;

[0079] In the formula, R i,j ρ is the deflection vibration characteristic value of the j-th deflection value in the program sequence for the i-th deflection; i,j The number of zero points in the line graph of the local deflection deviation sequence is defined as follows: the local deflection deviation sequence is obtained by sorting the differences between each deflection value and the mean of all deflection values ​​in the local deflection sequence. Adjacent deflection values ​​in the local deflection deviation sequence are then connected to obtain the line graph of the local deflection deviation sequence. i,j The information entropy is the local deflection deviation sequence corresponding to the j-th deflection value in the program sequence when the i-th deflection occurs.

[0080] Where, ρ i,j Used to reflect the frequency of vertical oscillation in deflection data, ent i,j The dispersion of numerical distribution in a local deflection deviation sequence is used to determine the intensity of deflection vibration error caused by significant traffic vibration phenomena, specifically the characteristic value R of the deflection vibration. i,j The larger the value, the more smoothing processing should be performed when correcting actual errors in a falling weight deflectometer to eliminate traffic vibration errors.

[0081] Therefore, the deflection noise confidence level and deflection vibration characteristic value of the j-th deflection value in the program sequence for the i-th deflection are normalized by range. Using the normalized deflection noise confidence level and deflection vibration characteristic value, the deflection smoothing evaluation vector of the j-th deflection value in the program sequence for the i-th deflection is obtained, and the deflection smoothing score is obtained using the Topsis algorithm. A higher deflection smoothing score indicates that the data acquisition software of the vehicle-mounted falling weight deflectometer should perform smoothing processing on the deflection values ​​to eliminate noise interference errors and traffic vibration errors, improve the quality of the falling weight deflectometer detection data, and facilitate subsequent road modulus inversion and potential hazard identification. Preferably, in this embodiment, range normalization and the Topsis algorithm are well-known technologies and will not be described in detail.

[0082] Furthermore, in this embodiment, during the actual measurement process, the Loess smoothing algorithm is used to smooth the deflection values ​​in each deflection time program sequence, thereby correcting the calibration error of each deflection time program sequence. Preferably, in this embodiment, the Loess smoothing algorithm is a well-known technology and will not be described in detail.

[0083] Therefore, based on the deflection noise confidence level and deflection vibration characteristic value of each deflection value in each deflection time sequence after normalization, the smoothing window scale of each deflection value in each deflection time sequence is obtained. In this embodiment, the specific calculation formula is as follows:

[0084] λ i,j =fun[W×(s) i,j +β)];

[0085] In the formula, λ i,j is the smoothing window scale for the j-th deflection value in the sequence at the i-th deflection; W is the initial window scale for the Loess smoothing algorithm; s i,j The deflection smoothing score is given for the j-th deflection value in the program sequence for the i-th deflection time. The ratio of deflection anomalies to road loosening influence is used as the deflection noise confidence level. The normalized deflection vibration characteristic value and deflection noise confidence level of each deflection value in the program sequence for the i-th deflection time are combined to form the deflection smoothing evaluation vector for each deflection value in the program sequence for the i-th deflection time. The deflection smoothing score is obtained through the Topsis algorithm. β is a parameter tuning factor; fun[·] is used to obtain the closest odd number. Preferably, in this embodiment, W is 15 and β is 0.4.

[0086] Here, β is used to provide a larger window scale for the deflection values ​​of noise interference error and traffic vibration error, thereby improving the smoothing processing capability of the Loess smoothing algorithm. At the same time, it is necessary to preserve the local details of road loose defects and normal deflection values ​​of falling weight deflectometers as much as possible, thereby improving the accuracy of subsequent road modulus inversion and potential hazard identification.

[0087] Accordingly, according to the method described above in this embodiment, the window scale of each deflection value in the i-th deflection time sequence is obtained, and the Loess smoothing algorithm is used to smooth the i-th deflection time sequence through the window scale of each deflection value in the i-th deflection time sequence, thereby correcting the deflection data error of the i-th deflection time sequence.

[0088] Similarly, the deflection time program sequences of each displacement sensor of the falling weight deflectometer can be smoothed, and the deflection data can be analyzed to correct the deflection data error of the falling weight deflectometer.

[0089] Based on the same inventive concept as the above method, this application embodiment also provides a calibration error correction system for a falling weight deflectometer, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described calibration error correction methods for a falling weight deflectometer.

[0090] It is understood that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0091] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0092] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the protection scope of this application.

Claims

1. A method for correcting the calibration error of a falling weight deflectometer, characterized in that, Includes the following steps: Obtain the deflection values ​​of the displacement sensors of the falling weight deflectometer to form a deflection time program sequence; the falling weight deflectometer has a total of 9 displacement sensors, and each displacement sensor is tested 5 times at each test position to obtain the deflection values ​​of each deflection time program sequence in each displacement sensor. Based on the average level of each deflection value in all deflection time program sequences, the degree of difference between each deflection value in each deflection time program sequence and the discrete range of each deflection value in all deflection time program sequences, the deflection outlier value of each deflection value in each deflection time program sequence is obtained. The right-hand sequence of each deflection value is extracted. After classifying each right-hand sequence of deflection values, the number of each deflection value and the difference of the corresponding position value of each deflection value are used to obtain the deflection continuity characteristic value of each deflection value in each deflection time program sequence. By obtaining all deflection values ​​adjacent to each deflection value, the local deflection sequence of each deflection value is obtained. Based on the correlation between the local deflection sequences of the same deflection value in the program sequence of adjacent deflection values, the load transfer weight of each deflection value in the program sequence of each deflection value is obtained. Combined with the deflection continuity characteristic value of each deflection value, the road loosening influence degree of each deflection value in the program sequence of each deflection value is obtained. Based on the degree of disorder in the local deflection sequence of each deflection value and the deviation of each deflection value in the local deflection sequence, the deflection vibration characteristic value of each deflection value in the program sequence of each deflection time is obtained. Then, combined with the deflection anomaly value of each deflection value and the influence of road loosening, the smoothing window scale of each deflection value in the program sequence of each deflection time is obtained. The program sequence of each deflection time is smoothed by combining the smoothing algorithm to correct the deflection data error of the falling weight deflectometer. For one of the displacement sensors, the method for calculating the deflection anomaly value of each deflection value in the program sequence during each deflection is as follows: ； In the formula, The deflection anomaly value is the j-th deflection value in the program column when the i-th deflection occurs. Let be the standard deviation of the j-th deflection value in the sequence of all deflections; The j-th deflection value in the program column when the i-th deflection occurs; The mean of the j-th deflection value in the program sequence for all deflections; The method for extracting the right-hand sequence of each deflection value is as follows: The deflection values ​​of each deflection value in the program column at each deflection time are arranged into a right-hand sequence of each deflection value, which is the number of adjacent positions to the right of each deflection value. When the number of deflection value data to the right is less than the first preset number, the right-hand sequence of deflection is all the deflection value data to the right of each deflection value in the program column at each deflection time. The method for calculating the deflection duration characteristic value of each deflection value in the program sequence for each deflection is as follows: ； In the formula, The deflection continuation characteristic value is the deflection duration of the j-th deflection value in the program sequence when the i-th deflection occurs. It is an exponential function with the natural constant e as its base; is the number of other deflection values ​​within the data level of the right-hand deflection sequence of the j-th deflection value, where the right-hand deflection sequence of each data level is obtained by classifying the deflection sequence using the natural breakpoint algorithm; is the average of the differences between the position values ​​of all deflection proximity points in the rightward sequence of the j-th deflection value and the corresponding position value of the j-th deflection value. Here, each deflection proximity point in each rightward sequence of each deflection value is another deflection value within the data level of each deflection value in the rightward sequence of each deflection value; M is the number of elements in the rightward sequence of the deflection. The method for obtaining the local deflection sequence of each deflection value is further as follows: Centered on each deflection value in the program sequence at each deflection time, obtain the second preset number of deflection values ​​that are closest to this center, and arrange them in ascending order to obtain the local deflection sequence of each deflection value. The method for calculating the load transfer weight of each deflection value in the program sequence for each deflection is as follows: ； In the formula, The load transfer weight is assigned to the j-th deflection value in the program sequence for the i-th deflection. The load transfer weight is assigned to the j-th deflection value in the program sequence for the (i-1)-th deflection. This is the local deflection sequence of the j-th deflection value in the program sequence when the i-th deflection occurs. This is the local deflection sequence of the j-th deflection value in the program column when the (i+1)-th deflection occurs; It is the Pearson correlation coefficient; The deflection sequence corresponds to the center displacement sensor. The displacement sensor at the end of the program sequence corresponds to the deflection time; The road loosening influence of each deflection value in the program sequence for each deflection time is calculated using the following formula: ； In the formula, The road loosening effect degree of the j-th deflection value in the program sequence when the i-th deflection occurs; It is a constant that avoids a denominator of 0; The formula for calculating the deflection vibration characteristic value of each deflection value in the program sequence for each deflection time is as follows: ； In the formula, The deflection vibration characteristic value of the j-th deflection value in the program sequence at the i-th deflection; The number of zero points in the line graph of the local deflection deviation sequence is defined as follows: the sequence of differences between each deflection value and the mean of all deflection values ​​in the local deflection sequence is used as the local deflection deviation sequence. Adjacent deflection values ​​in the local deflection deviation sequence are connected to obtain the line graph of the local deflection deviation sequence. The information entropy is the local deflection deviation sequence corresponding to the j-th deflection value in the program sequence when the i-th deflection occurs. The method for calculating the smoothing window scale of each deflection value in the program sequence for each deflection is as follows: ； In the formula, The smoothing window scale is the value of the j-th deflection in the program sequence when the i-th deflection occurs. This represents the initial window size for the Loess smoothing algorithm; The deflection smoothing score is the deflection smoothing score of the j-th deflection value in the program sequence for the i-th deflection time. The ratio of the deflection anomaly value to the road loosening influence is used as the deflection noise confidence. The deflection vibration characteristic value and the deflection noise confidence of each deflection value in the program sequence for the i-th deflection time are combined to form the deflection smoothing evaluation vector of each deflection value in the program sequence for the i-th deflection time. The deflection smoothing score is obtained by the Topsis algorithm. For parameter tuning factors; It retrieves the closest odd number.

2. A calibration error correction system for a falling weight deflectometer, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the calibration error correction method for a falling weight deflectometer as described in claim 1.

Images