Road space-time compaction quality inspection method and system based on sound-structure interaction

Abstract

The invention relates to a road space-time compaction quality inspection method and system based on sound-structure interaction. The method comprises the following steps: receiving a surface wave signal leaked in air from a compaction sampling point by using a sound pressure sensor in a non-contact manner; constructing a VMD-LSTM denoising module for denoising and reconstructing the surface wave signal to obtain a reconstructed signal; generating a two-dimensional time-frequency diagram from the reconstructed signal, and performing feature extraction on the two-dimensional time-frequency diagram by using a classification module to obtain time-frequency features; obtaining compaction influence factors and corresponding compaction degrees under different stage categories and different structure layer categories, and constructing a road space-time fusion data set; training an XGBoost machine learning algorithm by using the road space-time fusion data set; a VMD-LSTM denoising module, a classification module and a trained XGBoost machine learning algorithm are combined together to form a fusion intelligent model, and the fusion intelligent model is utilized to quickly complete compaction quality detection of any structural layer at any stage. According to the invention, full-thickness full-life-cycle road compaction quality detection is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of nondestructive testing, and in particular to a road spatiotemporal compaction quality inspection method and system based on acoustic-solid coupling. Background Art

[0002] With the rapid development of infrastructure construction in my country, the total mileage of roads has increased year by year. The detection of the quality of road structure layer compaction has become particularly important, which is directly related to the service quality and service life of the highway. Traditional compaction quality detection methods, such as sand filling method and core drilling method, are not only time-consuming and labor-intensive, but also destructive to the road structure. In the context of intelligent transportation infrastructure construction, a road quality detection method combining stress waves and intelligent models has emerged, aiming to improve the shortcomings of traditional methods and achieve rapid and non-destructive detection.

[0003] At present, on-site road compaction quality detection mainly uses a dynamic signal acquisition instrument with an acceleration sensor or a multi-channel acquisition instrument to collect artificially excited surface wave signals and then conduct quality assessment. However, the surface waves generated by these methods propagate in the structural layer and are captured by sensors in direct contact, which requires the sensors to be arranged on the surface of the road structure. Although non-destructive testing has been achieved, the flexibility of sensor arrangement still needs to be improved. Existing studies have shown that ground penetrating radar can be used for road compaction quality detection, but the cost of ground penetrating radar is relatively high. As an economical and convenient sensor, the acoustic pressure sensor can receive surface wave signals leaked into the air, which meets the requirements of flexible and fast surface wave signal acquisition. In addition, the road compaction intelligent model established by existing research can only be used for specific structural layers and specific stages. It is urgent to provide a road compaction quality detection method using acoustic pressure sensors that is suitable for full-thickness and full-life cycle roads. Summary of the invention

[0004] In view of the shortcomings of the prior art, the technical problem to be solved by the present invention is to provide a road spatiotemporal compaction quality inspection method and system based on acoustic-solid coupling. The method uses the data obtained by a non-contact sensor, the acoustic pressure sensor, to establish a fusion intelligent model suitable for compaction quality inspection of different structural layers of the road at different stages, thus realizing full-thickness and full-life cycle road compaction quality inspection.

[0005] The technical solution adopted by the present invention to solve the technical problem is: In a first aspect, the present invention provides a road spatiotemporal compaction quality inspection method based on acoustic-solid coupling, comprising the following contents: The acoustic pressure sensor is used to contactlessly receive the surface wave signal excited by external force and leaked into the air at the compaction sampling point; The VMD algorithm is used to decompose the surface wave signal into several modes, and multiple modes with smooth waveforms and consistent with the main frequency range of the surface wave signal are selected for reconstruction to obtain the reconstructed signal; all modes obtained by decomposing the surface wave signal by the VMD algorithm and the corresponding reconstructed signals are combined into denoising samples to construct a denoising data set; The input of the LSTM network is all the modes obtained by decomposition of the VMD algorithm, and the output is the reconstructed signal. The LSTM network is trained using the denoising data set to obtain a trained LSTM network. The VMD algorithm is connected to the trained LSTM network to form a VMD-LSTM denoising module. The compaction degree of the compaction sampling point is obtained, and several compaction degree intervals are divided according to the compaction degree distribution. Each compaction degree interval is used as a compaction degree category. The reconstructed signal obtained by the VMD-LSTM denoising module is processed to generate a two-dimensional time-frequency graph. The two-dimensional time-frequency graph generated by the reconstructed signal is classified according to the compaction degree interval to which the corresponding compaction degree belongs. The two-dimensional time-frequency graph and its corresponding compaction degree category form a classification sample, and a time-frequency graph classification data set is constructed; A classification network is trained using a time-frequency graph classification data set to obtain a trained classification network, wherein the output layer of the classification network is composed of a fully connected layer and a Softmax function, and the trained classification network is used as a classification module after removing the Softmax function to extract features from the two-dimensional time-frequency graph to obtain time-frequency features; the input of the classification module is the two-dimensional time-frequency graph of the reconstructed signal, and the output is the feature vector output by the fully connected layer; Obtain the compaction influencing factors and corresponding compaction degrees under different stage categories and different structural layer categories, and construct a road spatiotemporal fusion dataset; the different stage categories are construction stage, maintenance stage and acceptance and opening stage, the different structural layer categories are road base layer, cement stabilized gravel layer and asphalt surface layer, and the compaction influencing factors are moisture content, soil particle gradation, surface wave velocity, time-frequency characteristics, gravel gradation, cement content, gross volume density, Marshall stability, flow value, asphalt content, theoretical maximum density of surface layer and surface layer porosity; if a compaction influencing factor is missing, it is set to 0; The surface wave velocity is determined by using the reconstructed signal; the time-frequency feature is obtained by extracting the features of the two-dimensional time-frequency graph generated by the reconstructed signal using the classification module; Train the XGBoost machine learning algorithm with the road spatiotemporal fusion dataset to obtain the trained XGBoost machine learning algorithm for compaction detection; Combine the VMD-LSTM denoising module, the classification module and the trained XGBoost machine learning algorithm to form a fusion intelligent model; When used on site, data on compaction influencing factors other than surface wave velocity and time-frequency characteristics are input, and surface wave signals on site are collected. The fusion intelligent model is then used to quickly complete the detection of compaction quality at any stage of any structural layer.

[0006] Furthermore, the classification network is implemented using networks such as ViT, VGG, ResNet, MobileNet or Swin Transformer.

[0007] Furthermore, the XGBoost machine learning algorithm is optimized using an intelligent optimization algorithm before training, and the intelligent optimization algorithm is at least one of a PSO algorithm, a GWO algorithm, a GA algorithm, and the like.

[0008] In a second aspect, the present invention provides a road spatiotemporal compaction quality inspection system based on acoustic-solid coupling, which adopts the road spatiotemporal compaction quality inspection method based on acoustic-solid coupling, including a mobile vehicle, a dynamic signal collector installed on the mobile vehicle, and a control terminal, a pair of sound pressure sensors are installed at the rear of the mobile vehicle through a bracket and two sleeves, the two sleeves are symmetrically installed on the bracket, the sound pressure sensor is electrically connected to the dynamic signal collector, and the dynamic signal collector communicates with the control terminal; A pair of acoustic pressure sensors are arranged at both ends of the compaction sampling point and are not in contact with the compaction sampling point, and are used to receive surface wave signals excited by external force and leaked into the air; A wind shield structure is arranged around the pair of sound pressure sensors; The control terminal obtains the surface wave signal collected by the sound pressure sensor through a dynamic signal collector; The fusion intelligent model is loaded in the control terminal. When used on site, the data of compaction influencing factors other than surface wave velocity and time-frequency characteristics are input, and the surface wave signal on site is collected. The fusion intelligent model is then used to quickly complete the detection of compaction quality at any stage of any structural layer.

[0009] Furthermore, the road spatiotemporal compaction quality inspection system based on acoustic-solid coupling also includes a rubber hammer.

[0010] Furthermore, the road spatiotemporal compaction quality inspection system based on acoustic-solid coupling also includes a cloud, which is used to store a denoised data set, a time-frequency graph classification data set and a road spatiotemporal fusion data set, and the cloud communicates with a control terminal.

[0011] Furthermore, during sampling, the vertical distance between the sound pressure sensor and the compaction sampling point was controlled within 20 cm.

[0012] Furthermore, the two sleeves are movably connected to the bracket, the sleeves can move up and down relative to the bracket, and the wind shielding structure is fixed on the two sleeves and can move up and down with the sleeves.

[0013] Furthermore, the two sleeves are electrically connected to the lifting unit, and the lifting unit is electrically connected to the control terminal, so that the sleeves can be automatically lifted and lowered under the control of the lifting unit.

[0014] Compared with the prior art, the present invention has the following beneficial effects: Based on the principle of acoustic-solid coupling, the present invention uses an acoustic pressure sensor to collect surface waves propagated through the road structure and leaked into the air, and is applied to on-site road compaction quality detection. As a non-contact sensor, the acoustic pressure sensor is used to receive surface wave signals leaked into the air. The acoustic pressure sensor can capture flexibly and is economical.

[0015] The system of the present invention is provided with a windshield structure, and utilizes the dual protection of the VMD-LSTM denoising module and the windshield structure to reduce the influence of the air environment on the surface wave signal, thereby improving the accuracy and reliability of signal denoising.

[0016] The method of the present invention forms a road spatiotemporal fusion data set to train a machine learning algorithm, and then obtains a fusion intelligent model, which can be used to detect the compaction quality of full-thickness roads throughout the entire life cycle, including: road base, cement-stabilized crushed stone layer and asphalt surface layer, in different operating states throughout the entire life cycle, including: construction stage, maintenance stage and acceptance and opening to traffic stage. When the present invention is applied to new scenarios, it can quickly, non-destructively and continuously detect the surface wave signals of the road base, cement-stabilized crushed stone layer and asphalt surface layer, saving manpower and material resources and meeting actual engineering needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The present invention is a flow chart of an embodiment of a road spatiotemporal compaction quality inspection method based on acoustic-solid coupling.

[0018] Figure 2 Schematic diagram of the VMD-LSTM denoising module.

[0019] Figure 3 The flowchart of the classification module is shown in Figure 2.

[0020] Figure 4 The present invention is a schematic structural diagram of an embodiment of a road spatiotemporal compaction quality inspection system based on acoustic-solid coupling.

[0021] In the figure, 1. road structure layer, 2. mobile vehicle, 3. dynamic signal collector, 4. control terminal, 5. sound pressure sensor, 6. wind shield structure, 7. bracket, 8. sleeve. DETAILED DESCRIPTION

[0022] The present invention is further explained below in conjunction with embodiments and drawings, but the protection scope of the present application is not limited thereto.

[0023] The road time-space compaction quality inspection system based on acoustic-solid coupling of the present invention (see Figure 4 ), comprising a mobile vehicle 2, a dynamic signal collector 3 installed on the mobile vehicle and a control terminal 4, a pair of sound pressure sensors are installed at the rear of the mobile vehicle through a bracket 7 and two sleeves 8, the two sleeves 8 are symmetrically installed on the bracket 7, the bracket 7 is fixed to the mobile vehicle, the sound pressure sensor 5 is electrically connected to the dynamic signal collector, and the dynamic signal collector communicates with the control terminal 4; A pair of acoustic pressure sensors are arranged at both ends of the compaction sampling point and are not in contact with the compaction sampling point, and are used to receive surface wave signals artificially excited and leaked into the air; A wind shield structure is arranged around the pair of sound pressure sensors, and the wind shield structure is used to isolate the sound pressure sensors from the outside air and reduce the influence of the air on the signal; The control terminal obtains the surface wave signal collected by the sound pressure sensor through a dynamic signal collector; The fusion intelligent model is loaded in the control terminal. When used on site, the data of compaction influencing factors other than surface wave velocity and time-frequency characteristics are input, and the surface wave signal on site is collected. The fusion intelligent model is then used to quickly complete the detection of compaction quality at any stage of any structural layer.

[0024] During the acquisition process, the sound pressure sensor is placed at both ends of the compaction sampling point as a non-contact sensor. There are two sound pressure sensors in total, which are used to receive the surface wave signal that is artificially excited and leaked in the air. The sound pressure sensor is connected to the dynamic signal acquisition instrument, and the control terminal controls the acquisition and storage.

[0025] The windshield structure in the present invention can be a windshield cover or a windshield plate, etc., to isolate the two sound pressure sensors from the outside air to reduce the influence of air on the signal. The specific implementation structure can be implemented according to the existing technology. The lower end of the windshield structure should be lower than the lowest installation point of the sound pressure sensor. When the lower end of the windshield structure contacts the road structure layer 1, the influence of air on the signal can be reduced to the greatest extent.

[0026] Figure 2It is a flow chart of the VMD-LSTM denoising module. In the present invention, the surface wave signal is further denoised by combining the VMD algorithm and the LSTM network. During signal processing, the surface wave signal is decomposed into several modes using the VMD algorithm, and the appropriate mode is manually selected, and the selected appropriate mode combination is reconstructed to obtain the reconstructed signal and save it. The appropriateness described here can be understood as: the decomposed mode is smooth in waveform and similar in shape to the original signal (i.e., the surface wave signal before decomposition); in frequency, it is close to the main frequency interval of the original signal, and the mode is considered to be a suitable mode. That is to say, the mode with a smooth waveform and consistent with the main frequency interval of the surface wave signal is selected as a suitable mode. The consistency here can be understood as close. This process can be selected manually, and it is considered to be consistent within the error range of the naked eye.

[0027] In the early stage of training, surface wave signals at compaction sampling points in different stage categories and different structural layer categories are obtained. All modes obtained by decomposing the surface wave signals through the VMD algorithm and the corresponding reconstructed signals are composed of denoised samples. A large number of denoised samples constitute the denoised data set.

[0028] All modes decomposed by the VMD algorithm are used as the input of the LSTM network, and the reconstructed signal is used as the target output. The LSTM network is trained using the denoising data set, and the VMD algorithm is connected with the trained LSTM network to form a VMD-LSTM denoising module, so that the VMD-LSTM denoising module can automatically realize mode selection and signal reconstruction, thus completing the automatic denoising process of the surface wave signal. The VMD-LSTM denoising module effectively reduces the residual noise, and the processed surface wave signal is smoother and has less noise, which can meet the needs of actual engineering.

[0029] The VMD algorithm can divide the surface wave signal into several intrinsic mode functions, each of which contains local signal features of different scales of the original signal. The process of the VMD algorithm is: first, the original signal is decomposed using VMD, and then each mode is fast Fourier transformed. By observing the time domain curve and frequency response of the mode, the mode with less noise is selected to reconstruct the signal.

[0030] Figure 3It is a flow chart of the classification module, which obtains the compaction degree and surface wave signal at the compaction sampling points under different stage categories and different structural layer categories, divides the compaction degree interval according to the compaction degree distribution, and each compaction degree interval is used as a compaction degree category. The reconstructed signal obtained by processing the surface wave signal through the VMD-LSTM denoising module generates a two-dimensional time-frequency diagram. The two-dimensional time-frequency diagram generated by the reconstructed signal is classified according to the compaction degree interval to which the corresponding compaction degree belongs, that is, each two-dimensional time-frequency diagram is annotated with a compaction degree category classification label, and the two-dimensional time-frequency diagram and its corresponding classification label form a classification sample. A large number of classification samples constitute a time-frequency diagram classification data set; The classification network is trained using the time-frequency graph classification data set. The output layer of the classification network consists of a fully connected layer and a Softmax function. The trained classification network is removed from the Softmax function and used as a classification module to extract features from the two-dimensional time-frequency graph to obtain time-frequency features. The input of the classification module is the two-dimensional time-frequency graph of the reconstructed signal, and the output is the feature vector output by the fully connected layer, that is, the time-frequency feature.

[0031] The present invention is a road spatiotemporal compaction quality inspection method based on acoustic-solid coupling, and selects an acoustic pressure sensor based on the acoustic-solid coupling principle to collect surface wave signals leaked in the air. The acoustic-solid coupling principle is: a rubber hammer is used to hit the ground on one side of the compaction sampling point. The rubber hammer excites stress waves on the outside of the acoustic pressure sensor, wherein transverse waves and longitudinal waves propagate in the depth direction, and surface waves propagate along the solid surfaces of the road base layer, cement-stabilized gravel layer, and asphalt surface layer. During the propagation process, part of the surface waves will diffuse into the air medium to form leakage surface waves. The leakage surface waves propagate in the air medium to form an acoustic-solid coupling model. The leakage surface waves are received by the two acoustic pressure sensors placed above, and the waveforms are recorded by the dynamic signal acquisition instrument. The acoustic pressure sensor close to the hammering side is recorded as the first acoustic pressure sensor, and the acoustic pressure sensor far from the hammering side is recorded as the second acoustic pressure sensor. The first acoustic pressure sensor first receives the surface waves leaked in the air, and the propagation time of the first acoustic pressure sensor is is the propagation time of the surface wave from the hammering point to the first sound pressure sensor in a single structural layer Plus the time it takes for the leak to propagate into the air and reach the first sound pressure sensor The value of The second sound pressure sensor then receives the leakage surface wave, and the propagation time of the second sound pressure sensor is is the propagation time of the surface wave from the hammering point to the second acoustic pressure sensor in a single structural layer. Plus the time it takes for the leak to travel into the air and reach the second sound pressure sensor ,Right now The time difference between the two sound pressure sensors is Since the two sound pressure sensors are placed horizontally to the ground and installed at the same height, the propagation speed of the surface wave leaked in the air is considered to be the same, so it is considered that Therefore, the time difference of the surface wave propagating from the hammer point to the two sound pressure sensors in the corresponding structural layer is the time difference of the two sound pressure sensors propagating. Surface wave velocity = the distance L between the two sound pressure sensors divided by the time difference ,Right now .

[0032] The present invention uses an acoustic pressure sensor to collect surface wave signals for different structural layers at different stages, that is, surface wave signals are collected at compaction sampling points for the road base layer, cement stabilized gravel layer and asphalt surface layer in the construction stage, maintenance stage and acceptance and opening to traffic stage. The surface wave signal is processed by a VMD-LSTM denoising module to obtain two reconstructed signals, and the time difference is calculated based on the two reconstructed signals, and then the surface wave velocity is calculated. The reconstructed signal processed by the VMD-LSTM denoising module is used to generate a two-dimensional time-frequency diagram, and a classification module is used to extract features from the two-dimensional time-frequency diagram to obtain time-frequency features. The classification module of this embodiment uses the Vision Transformer (ViT) classification network as the main architecture, and the classification network can also use VGG, ResNet, MobileNet, Swin Transformer, etc.

[0033] For the roadbed, the moisture content at the compaction sampling point, soil particle gradation (such as the content of large, medium and small particle sizes), surface wave velocity, time-frequency characteristics and compaction degree are used together to form a roadbed compaction quality detection data set.

[0034] For cement-stabilized gravel layer, the moisture content, gravel gradation, cement content, surface wave velocity, time-frequency characteristics and compaction degree at the compaction sampling point are used together to form a cement-stabilized gravel layer compaction quality detection data set.

[0035] For asphalt pavement, the gross volume density, Marshall stability, flow value, asphalt content, theoretical maximum density of the pavement, porosity of the pavement, surface wave velocity, time-frequency characteristics and compaction degree obtained from the compaction sampling points are used together to form an asphalt pavement compaction quality detection data set.

[0036] The compaction influencing factors in the roadbed compaction quality detection dataset, cement stabilized macadam layer compaction quality detection dataset, and asphalt surface layer compaction quality detection dataset are combined, and the structural layer categories (different layers) and stage categories (construction stage, maintenance stage, and acceptance and opening stage) and the corresponding compaction degree are added to obtain the road spatiotemporal fusion dataset. In the road spatiotemporal fusion dataset, when sampling the road base layer, the values ​​of the data of other compaction influencing factors except the compaction influencing factors of the structural layer are all zero. The same is true for the other two layers, that is, when a compaction influencing factor is missing, it is set to 0. In the roadbed compaction quality detection dataset, cement stabilized macadam layer compaction quality detection dataset, and asphalt surface layer compaction quality detection dataset, all data except the compaction degree are compaction influencing factors.

[0037] In the process of building the above data set, after each structural layer is compacted, a compaction sampling point is set at a certain distance (about 10-20m). In the later on-site use, spot checks or continuous inspections can be carried out, and the area between the two sound pressure sensors is the compaction sampling point. During spot checks and inspections, compaction sampling points can be set according to the specifications.

[0038] The present invention uses the XGBoost machine learning algorithm for compaction prediction, and uses an intelligent optimization algorithm to optimize the XGBoost machine learning algorithm. The intelligent optimization algorithm can be at least one of the PSO algorithm, the GWO algorithm or the GA algorithm, and can establish a PSO-XGBoost regression prediction model, a GA-XGBoost regression prediction model, a GWO-XGBoost regression prediction model, and the like.

[0039] The road spatiotemporal fusion dataset is divided into a training set and a validation set with a ratio of 7:3 for training and validating the above regression prediction model. The PSO-XGBoost regression prediction model and the GA-XGBoost regression prediction model are both set with 50 particles and 3000 iterations, and the GWO-XGBoost regression prediction model is set with 50 wolves and 3000 iterations. MAPE is used as the fitness function, R 2 , RMSE, and MAPE are used to evaluate the regression prediction model results, and the optimal regression prediction model is selected for saving.

[0040] The VMD-LSTM denoising module, the classification module and the optimal regression prediction model are connected in sequence to obtain a fusion intelligent model. In this embodiment, the fusion intelligent model can be recorded as VMD-LSTM-ViT-XGB. When used on site, the data of compaction influencing factors other than the surface wave velocity and time-frequency characteristics are input, and the surface wave signal collected in real time on site is used. The fusion intelligent model is then used to quickly complete the compaction quality detection of any structural layer at any stage, and the compaction degree at which structural layer and which stage can be output.

[0041] During on-site sampling of the present invention, the vertical distance between the sound pressure sensor and the compaction sampling point is within 20 cm, the sound pressure sensor cannot be close to the ground, and the closer to the ground the better, and the distance should not be too large.

[0042] Example 1: This example is a road time-space compaction quality inspection method based on acoustic-solid coupling (see Figure 1 ) process is: 1. Surface wave signal acquisition: A road time-space compaction quality inspection system based on acoustic-solid coupling is established, including a mobile vehicle, a dynamic signal collector installed on the mobile vehicle, and a control terminal. A pair of sound pressure sensors are installed at the rear of the mobile vehicle through a bracket and two sleeves. The two sleeves are symmetrically fixed on the bracket. The sound pressure sensor is electrically connected to the dynamic signal collector, and the dynamic signal collector communicates with the control terminal. The vertical distance between the sound pressure sensor and the compaction sampling point is 15 cm. A windshield structure is set around a pair of sound pressure sensors.

[0043] Before collecting data, the sensitivity of the sound pressure sensor is calibrated. A 1000Hz, 1Pa standard signal is excited by an acoustic calibrator to calibrate the sensitivity of the sound pressure sensor in the environment. After completing the sensitivity calibration, the surface wave signals are collected from the compaction sampling points during the construction, maintenance and acceptance stages of the road base, cement stabilized gravel layer and asphalt surface layer, and saved.

[0044] 2. Signal denoising: Cut the signal output by the sound pressure sensor to obtain a signal containing a mutation signal (in the acquisition process, it is a stable signal when no surface wave is received, and a mutation signal is formed after receiving the surface wave) and a length of 1000 (1000 represents the number of data points). This is the surface wave signal, and then use the VMD algorithm for modal decomposition. Set the number of decomposed modes to 10, decompose the surface wave signal, select appropriate modes for combination, and obtain the reconstructed signal, which is the denoised signal. Use all modes obtained by VMD decomposition as the input of the LSTM network, and the reconstructed signal as the target output. Use the denoising data set to train the LSTM network, connect the VMD algorithm with the trained LSTM network to form a VMD-LSTM denoising module, and save the VMD-LSTM denoising module.

[0045] 3. Constructing a spatiotemporal fusion dataset of roads: The reconstructed signal is used to obtain the surface wave velocity and generate a two-dimensional time-frequency graph. The ViT classification module is used to extract the time-frequency features of each two-dimensional time-frequency graph. The surface wave velocity and time-frequency features are matched one by one with the material properties (such as moisture content, soil particle gradation, gravel gradation, cement content, gross volume density, Marshall stability, flow value, asphalt content, theoretical maximum density of the surface layer, porosity of the surface layer, etc.) and compaction degree at the compaction sampling point, respectively forming a roadbed compaction quality detection dataset, a cement-stabilized gravel layer compaction quality detection dataset, and an asphalt surface layer compaction quality detection dataset. The compaction influencing factors in the roadbed compaction quality detection dataset, the cement-stabilized gravel layer compaction quality detection dataset, and the asphalt surface layer compaction quality detection dataset are combined, and the structural layer category (different layers) and stage category (construction stage, maintenance stage, and acceptance and opening stage) and the corresponding compaction degree are added to obtain a road spatiotemporal fusion dataset.

[0046] Among them, the area between the two sound pressure sensors is the compaction detection interval, and the surface wave velocity and time-frequency characteristics correspond one-to-one to the compaction and the material properties at that location.

[0047] 4. Road compaction quality evaluation: The PSO-XGBoost regression prediction model is used to train and predict the road spatiotemporal fusion dataset, and the road spatiotemporal fusion dataset is divided into a training set and a validation set. The ratio of the training set to the validation set is 7:3. The trained PSO-XGBoost regression prediction model is used for compaction detection.

[0048] The VMD-LSTM denoising module, ViT classification module and PSO-XGBoost regression prediction model are combined to form the VMD-LSTM-ViT-XGB fusion intelligent model. The fusion intelligent model is used to evaluate the road compaction quality, and the surface wave signal is combined with the compaction influencing factors of the corresponding road base layer, cement stabilized gravel layer and asphalt surface layer, as well as the structural layer category and stage category, to finally achieve a fast and lightweight inspection of the road spatiotemporal compaction quality. Once the fusion intelligent model is built, there is no need to construct a time-frequency graph classification dataset, a denoising dataset, and a road spatiotemporal fusion dataset in actual use. When used on site, compaction detection can be achieved by only receiving the surface wave signal excited by external force and leaking into the air in real time.

[0049] Embodiment 2: The road spatiotemporal compaction quality inspection system based on acoustic-solid coupling in this embodiment also includes a cloud, which is used to store denoised data sets, time-frequency graph classification data sets and road spatiotemporal fusion data sets. The cloud communicates with the control terminal, and the cloud can obtain field data in a timely manner and update the corresponding data sets in a timely manner.

[0050] Embodiment 3: In this embodiment, a drop hammer unit is provided, and the drop hammer unit can also be fixed to the outside of the outermost sound pressure sensor through a bracket. The drop hammer unit is electrically connected to the control terminal, and the force and frequency of the drop hammer unit's striking can be set so that it can be automatically excited at fixed time intervals when used on-site to meet the purpose of automatic continuous sampling or intermittent sampling.

[0051] Embodiment 4: In this embodiment, two sleeves are movably connected to the bracket, and the sleeves can move up and down relative to the bracket and can be fixed at the target height. The windshield structure is fixed on the two sleeves and can move up and down with the sleeves. When in use, the sleeve can be manually lowered at the compaction sampling point until the windshield structure on the sleeve contacts the surface to be tested, and then hammered to collect surface acoustic waves.

[0052] Embodiment 5: In this embodiment, the sleeve is electrically connected to the lifting unit, and the lifting unit is electrically connected to the control terminal. The sleeve can be automatically lifted and lowered under the control of the lifting unit. The lifting unit uses a motor and a chain to adjust the height of the windshield structure and the sound pressure sensor from the surface to be tested. The horizontal spacing between the two sound pressure sensors and the vertical distance between the sound pressure sensor and the compaction sampling point can be set in the control terminal. The windshield structure always covers the surrounding of the sound pressure sensor and moves up and down with the sound pressure sensor, so that the windshield structure and the sound pressure sensor can be automatically lifted and lowered and detected on site without affecting the movement of the mobile vehicle, thereby realizing automated inspection.

[0053] Any matters not described in the present invention are applicable to the prior art.

Claims

1. A road spatiotemporal compaction quality inspection method based on acoustic-solid coupling, characterized in that: The road spatiotemporal compaction quality inspection method based on acoustic-solid coupling includes the following contents: The acoustic pressure sensor is used to contactlessly receive the surface wave signal excited by external force and leaked into the air at the compaction sampling point; The VMD algorithm is used to decompose the surface wave signal into several modes, and multiple modes with smooth waveforms and consistent with the main frequency range of the surface wave signal are selected for reconstruction to obtain the reconstructed signal; all modes obtained by decomposing the surface wave signal by the VMD algorithm and the corresponding reconstructed signals are combined into denoising samples to construct a denoising data set; The input of the LSTM network is all the modes obtained by decomposition of the VMD algorithm, and the output is the reconstructed signal. The LSTM network is trained using the denoising data set to obtain a trained LSTM network. The VMD algorithm is connected to the trained LSTM network to form a VMD-LSTM denoising module. The compaction degree of the compaction sampling point is obtained, and several compaction degree intervals are divided according to the compaction degree distribution. Each compaction degree interval is used as a compaction degree category. The reconstructed signal obtained by the VMD-LSTM denoising module is processed to generate a two-dimensional time-frequency graph. The two-dimensional time-frequency graph generated by the reconstructed signal is classified according to the compaction degree interval to which the corresponding compaction degree belongs. The two-dimensional time-frequency graph and its corresponding compaction degree category form a classification sample, and a time-frequency graph classification data set is constructed; The classification network is trained using the time-frequency graph classification data set to obtain a trained classification network, wherein the output layer of the classification network is composed of a fully connected layer and a Softmax function, and the trained classification network is used as a classification module after removing the Softmax function to extract features from the two-dimensional time-frequency graph to obtain time-frequency features; Obtain the compaction influencing factors and corresponding compaction degrees under different stage categories and different structural layer categories, and construct a road spatiotemporal fusion dataset; the different stage categories are construction stage, maintenance stage and acceptance and opening stage, the different structural layer categories are road base layer, cement stabilized gravel layer and asphalt surface layer, and the compaction influencing factors are water content, soil particle gradation, surface wave velocity, time-frequency characteristics, gravel gradation, cement content, gross volume density, Marshall stability, flow value, asphalt content, theoretical maximum density of surface layer and surface porosity; Train the XGBoost machine learning algorithm with the road spatiotemporal fusion dataset to obtain the trained XGBoost machine learning algorithm for compaction detection; Combine the VMD-LSTM denoising module, the classification module and the trained XGBoost machine learning algorithm to form a fusion intelligent model; When used on site, data on compaction influencing factors other than surface wave velocity and time-frequency characteristics are input, and surface wave signals on site are collected. The fusion intelligent model is then used to quickly complete the detection of compaction quality at any stage of any structural layer.

2. The road time-space compaction quality inspection method based on acoustic-solid coupling according to claim 1 is characterized in that: The classification network is implemented using a ViT, VGG, ResNet, MobileNet or Swin Transformer network.

3. The road time-space compaction quality inspection method based on acoustic-solid coupling according to claim 1 is characterized in that: The XGBoost machine learning algorithm is optimized using an intelligent optimization algorithm before training, and the intelligent optimization algorithm is at least one of a PSO algorithm, a GWO algorithm, or a GA algorithm.

4. A road time-space compaction quality inspection system based on acoustic-solid coupling, characterized in that: The road time-space compaction quality inspection system based on acoustic-solid coupling adopts the road time-space compaction quality inspection method based on acoustic-solid coupling according to claim 1, including a mobile vehicle, a dynamic signal collector installed on the mobile vehicle, and a control terminal. A pair of sound pressure sensors are installed at the rear of the mobile vehicle through a bracket and two sleeves, and the two sleeves are symmetrically installed on the bracket. The sound pressure sensor is electrically connected to the dynamic signal collector, and the dynamic signal collector communicates with the control terminal; A pair of acoustic pressure sensors are arranged at both ends of the compaction sampling point and are not in contact with the compaction sampling point, and are used to receive surface wave signals excited by external force and leaked into the air; A wind shield structure is arranged around the pair of sound pressure sensors; The control terminal obtains the surface wave signal collected by the sound pressure sensor through a dynamic signal collector; Load the fusion intelligent model in the control terminal.

5. The road time-space compaction quality inspection system based on acoustic-solid coupling according to claim 4 is characterized in that: The road spatiotemporal compaction quality inspection system based on acoustic-solid coupling also includes a rubber hammer.

6. The road time-space compaction quality inspection system based on acoustic-solid coupling according to claim 4 is characterized in that: The road spatiotemporal compaction quality inspection system based on acoustic-solid coupling also includes a cloud, which is used to store a denoised data set, a time-frequency graph classification data set, and a road spatiotemporal fusion data set, and the cloud communicates with a control terminal.

7. The road time-space compaction quality inspection system based on acoustic-solid coupling according to claim 4 is characterized in that: During sampling, the vertical distance between the sound pressure sensor and the compaction sampling point was controlled within 20 cm.

8. The road time-space compaction quality inspection system based on acoustic-solid coupling according to claim 4 is characterized in that: The two sleeves are movably connected to the bracket, the sleeves can move up and down relative to the bracket, and the wind shielding structure is fixed on the two sleeves and can move up and down with the sleeves.

9. The road time-space compaction quality inspection system based on acoustic-solid coupling according to claim 8 is characterized in that: The two sleeves are electrically connected to the lifting unit, and the lifting unit is electrically connected to the control terminal. The sleeves can be automatically lifted and lowered under the control of the lifting unit.

Images