Road surface flatness detection device and detection method based on vehicle vibration response
Through the generative adversarial neural network combined with vehicle vibration response data, optimize the discrimination and generation of losses, the problem of dependence on theoretical models based on vehicle vibration response detection methods in the prior art is solved, and low-cost and efficient road flatness detection is achieved, improving the improvement of The detection accuracy and robustness are provided to provide rapid decision-making support for road network maintenance.
Patent Information
- Application Number
- CN202411264518.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-10
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-09-10
AI Technical Summary
The existing road flatness detection method based on vehicle vibration response is highly dependent on the theoretical model. When there are measurement errors and vehicle parameters are uncertain, it is difficult to accurately identify the road flatness, and the cost is high.
Generative adversarial neural network (GAN) is used to combine vehicle vibration response data, and the vibration response of the vehicle's front and rear axles is measured through acceleration sensors. The discriminator and generator of the generative adversarial neural network are used to optimize the generation loss and discrimination. Loss, reduce dependence on vehicle parameters and improve identification accuracy.
It realizes low-cost and efficient road flatness detection, reduces dependence on traditional direct measurement tools, improves detection accuracy and robustness, is suitable for large-scale road network maintenance management, and provides timely decision-making basis.
Smart Images

Figure CN118999435A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of road and track engineering detection, and in particular to a road surface smoothness detection device and a detection method based on vehicle vibration response. Background Art
[0002] Flatness is an important indicator of road performance and driving comfort. Identifying road flatness is of great decision-making significance for timely road maintenance and management. At present, there are three main methods for identifying road flatness: the first method is to directly measure through equipment such as profilometers, rulers, and levels. Although the accuracy is guaranteed to a certain extent, it is time-consuming and laborious; the second method is to identify road flatness through laser profilometers, ground penetrating radars, optical imaging, point cloud models, etc., although the accuracy is considerable, these instruments and equipment are very expensive; the third method is to indirectly identify road flatness through vehicle vibration response. Although the accuracy is relatively low, it has the characteristics of low cost and high measurement efficiency. In recent years, with the development of big data technology, the method of indirectly identifying road flatness based on vehicle vibration response has become more and more popular. However, the existing indirect identification method based on vehicle vibration is highly dependent on the vehicle vibration theoretical model. When there is a measurement error or the vehicle parameters cannot be measured, the method based on the theoretical model often fails to achieve satisfactory identification results.
[0003] Therefore, in order to reduce the cost of road surface roughness detection and avoid the problem that the existing method of indirect identification of road surface roughness based on vehicle vibration is highly dependent on theoretical models, it is urgently necessary to combine the current advanced machine learning models and use vehicle vibration response to indirectly measure road surface roughness through big data analysis. Summary of the invention
[0004] The purpose of the present invention is to provide a road surface roughness detection device and detection method based on vehicle vibration response, so as to reduce the detection cost of road surface roughness, while avoiding the problem that the existing method of indirectly identifying road surface roughness based on vehicle vibration is highly dependent on theoretical models.
[0005] The present invention is implemented by adopting the following technical scheme: a road surface flatness detection method based on vehicle vibration response, characterized in that it comprises the following steps:
[0006] S1, attach acceleration sensors to the front and rear axles of a dual-axle test vehicle without a shock absorber system, and drive the test vehicle over an uneven road surface to measure and record the vibration acceleration response of the front and rear axles. and
[0007] S2, establish a generative adversarial neural network to generate the vibration acceleration response of the front and rear axles of the vehicle and and road surface roughness sample r;
[0008] S3, Generate responses through the discriminator of the generative adversarial neural network and With measurement response and Perform consistency judgment and construct discrimination loss;
[0009] S4, using the generated acceleration response and The corresponding displacement response is obtained by numerical integration and And the vibration equation of the two-axle vehicle is established and and The physical constraint equation of r constructs the generation loss including the physical constraint loss and the adversarial loss;
[0010] S5, first reduce the discriminative loss to optimize the discriminator, and then reduce the generative loss to optimize the generator;
[0011] S6, the road roughness sample r generated by the generative adversarial neural network after optimization is the roughness of the road to be identified.
[0012] Furthermore, the generative adversarial neural network includes a generator and a discriminator, and the generator and the discriminator are both composed of a five-layer feedforward neural network.
[0013] Furthermore, the physical constraint equation in step S4 is expressed as follows:
[0014]
[0015] Where: d 1 and d 2 are the distances between the center of gravity of the vehicle and the front and rear axles, respectively, and d is the distance between the front and rear axles of the vehicle, and d = d 1 +d 2 , k 1 is the spring stiffness of the tire on the front axle of the vehicle, k 2 is the spring stiffness of the tire on the rear axle of the vehicle, M v is the vehicle mass, J v is the moment of inertia of the vehicle's center of gravity.
[0016] Furthermore, the discrimination loss in step S3 is expressed as follows:
[0017]
[0018] Where: E(.) is the mathematical expectation function, D(.) is the discriminator of the generative adversarial neural network, t represents the time, and z is the input variable of the generative adversarial neural network, which is generated by the standard normal distribution.
[0019] Furthermore, the generation loss in step S4 is expressed as follows:
[0020] Gen loss =Adv loss +β·Phy loss
[0021]
[0022]
[0023] Where: Adv loss represents the adversarial loss of the generative adversarial neural network, Phy loss represents the physical constraint loss of the physical constraint equation, and β represents the weight coefficient of the physical constraint loss.
[0024] A road surface flatness detection device based on vehicle vibration response, characterized in that it includes a front axle, a rear axle, a frame, a traction frame, an acceleration sensor and a data acquisition system; the frame is fixedly supported on the front axle and the rear axle, and the traction frame and a ballast box are installed on the frame; the acceleration sensor is respectively installed on the front axle and the rear axle, and the data acquisition system is arranged in the ballast box of the frame and connected to the acceleration sensor.
[0025] Furthermore, the front axle and the rear axle are respectively formed by connecting two left and right wheels and a steel rod, the frame is fixedly supported on the steel rods of the front axle and the rear axle, and the acceleration sensors are respectively installed at the center positions of the steel rods of the front axle and the rear axle.
[0026] The road surface roughness detection device and detection method based on vehicle vibration response described in the present invention have the following beneficial effects:
[0027] Significant cost-effectiveness: This method indirectly evaluates the road surface flatness through vehicle vibration response, which greatly reduces the detection cost compared to the traditional method of using direct measurement tools such as profilometers, rulers, and levels. At the same time, compared with high-precision but expensive equipment such as laser profilometers and ground-penetrating radars, this method achieves higher economic efficiency while ensuring a certain detection accuracy.
[0028] Efficient and convenient: Using vehicle vibration response to detect road surface flatness does not require complicated installation and calibration processes, and can quickly and continuously obtain road surface status data, significantly improving detection efficiency. This is particularly important for road maintenance management of large-scale road networks, and can quickly provide a basis for decision-making.
[0029] Data-driven and model fusion: Combining advanced generative adversarial neural network (GAN) technology, this method not only utilizes the powerful data processing capabilities of machine learning, but also incorporates the vehicle vibration theory model to achieve joint drive of physical data and mathematical models. This method reduces excessive reliance on vehicle parameters and improves recognition accuracy and robustness.
[0030] Strong applicability: The detection device is designed as a simple dual-axle detection vehicle with a compact structure, easy to install and deploy. By adjusting the weight of the ballast box, the vehicle driving state under different load conditions can be simulated, further enhancing the universality and reliability of the detection results.
[0031] Rapid decision support: The road surface flatness recognition results provided by the present invention can provide timely and accurate decision-making basis for road repair and maintenance. By quickly evaluating the flatness of different sections in the road network, the repair work of severely damaged sections can be arranged first, thereby improving the traffic quality and driving safety of the entire road network. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. The drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.
[0033] Figure 1 It is a flow chart of the detection method of the present invention;
[0034] Figure 2 It is a schematic diagram of the detection device of the present invention;
[0035] Figure 3 It is a schematic diagram of the axle of the detection device of the present invention;
[0036] Figure 4 It is a schematic diagram of the vibration acceleration response of the front axle of a vehicle actually measured when the detection vehicle passes through an uneven road surface according to an embodiment of the present invention;
[0037] Figure 5 It is a schematic diagram of the vibration acceleration response of the rear axle of a vehicle actually measured when the detection vehicle passes through an uneven road surface according to an embodiment of the present invention;
[0038] Figure 6 A schematic diagram of a feedforward neural network structure of a generator and a discriminator of a generative adversarial neural network according to an embodiment of the present invention;
[0039] Figure 7 A schematic diagram showing the variation of the physical constraint loss of a generative adversarial neural network with training samples according to an embodiment of the present invention;
[0040] Figure 8 A schematic diagram showing how the adversarial loss of a generative adversarial neural network changes with training samples according to an embodiment of the present invention;
[0041] Fig. 9 Schematic diagram of road surface roughness indirectly identified through vehicle vibration response according to an embodiment of the present invention.
[0042] In the figure, 1, front axle; 2, rear axle; 3, wheel; 4, steel rod; 5, frame; 6, traction frame; 7, ballast box; 8, acceleration sensor; 9, data acquisition system. DETAILED DESCRIPTION
[0043] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.
[0044] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0045] like Figure 1-9 As shown, a road surface flatness detection device based on vehicle vibration response includes a front axle 1, a rear axle 2, a frame 5, a traction frame 6, an acceleration sensor 8 and a data acquisition system 9. The front axle 1 and the rear axle 2 are respectively formed by connecting two left and right wheels 3 and a steel rod 4. The frame 5 is fixedly supported on the steel rods 4 of the front axle 1 and the rear axle 2, and the traction frame 6 and the ballast box 7 are installed on the frame 5. The acceleration sensor 8 is respectively installed at the center position of the steel rods 4 of the front axle 1 and the rear axle 2, and the data acquisition system 9 is arranged in the ballast box 7 of the frame 5 and connected to the acceleration sensor 8. The detection device is connected to a vehicle with a power system through the traction frame 6, and is towed forward by the vehicle.
[0046] A road surface roughness detection method based on vehicle vibration response includes the following main steps:
[0047] S1, attach acceleration sensors to the front and rear axles of the test vehicle to collect the vibration acceleration response of the vehicle passing through uneven roads and like Figure 4 and Figure 5 ;
[0048] S2, establish a generative adversarial neural network to generate the vibration acceleration response of the front and rear axles of the vehicle and and road surface roughness sample r;
[0049] S3, Generate responses through the discriminator of the generative adversarial neural network and With measurement response and Perform consistency judgment and construct discrimination loss;
[0050] S4, using the generated acceleration response and The corresponding displacement response is obtained by numerical integration and And the vibration equation of the two-axle vehicle is established and and The physical constraint equation of r constructs the generation loss including the physical constraint loss and the adversarial loss;
[0051] S5, first reduce the discriminant loss to optimize the discriminator, and then reduce the generation loss to optimize the generator; S6, the road roughness sample r generated by the generative adversarial neural network after optimization is the roughness of the road to be identified.
[0052] The generative adversarial neural network includes a generator and a discriminator, and the generator and the discriminator are both composed of a five-layer feedforward neural network, such as Figure 6 ;
[0053] The physical constraint equation is expressed as follows:
[0054]
[0055] Where: d 1 and d 2 are the distances between the center of gravity of the vehicle and the front and rear axles, respectively, and d is the distance between the front and rear axles of the vehicle, and d = d 1 +d 2 , k 1 is the spring stiffness of the tire on the front axle of the vehicle, k 2 is the spring stiffness of the tire on the rear axle of the vehicle, Mv is the vehicle mass, and Jv is the moment of inertia of the vehicle's center of gravity.
[0056] The discriminant loss of the discriminator is expressed as follows: by reducing the discriminant loss, the discriminator cannot distinguish the similarities and differences between the generated data and the measured data;
[0057]
[0058] Where: E(.) is the mathematical expectation function, D(.) is the discriminator of the generative adversarial neural network, t represents the time, and z is the input variable of the generative adversarial neural network, which is generated by the standard normal distribution.
[0059] The generation loss of the generator is expressed as follows:
[0060] Gen loss =Adv loss +β·Phy loss
[0061]
[0062] Where: Adv loss represents the adversarial loss of the generative adversarial neural network, Phy loss represents the physical constraint loss of the physical constraint equation, and β represents the weight coefficient of the physical constraint loss.
[0063] The changes of the generator's generation loss and physical constraint loss in the iterative process are as follows: Figure 7 and Figure 8 ;
[0064] Finally, the identified road surface roughness profile is as follows Fig. 9 .
[0065] The vehicle parameter values of the detection vehicle in this embodiment are shown in Table 1.
[0066] Table 1 Vehicle parameter values of the test vehicle
[0067]
[0068] The above embodiments describe the basic principles and main features of the present invention and the advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the changes and modifications made by those skilled in the art shall be within the scope of protection of the appended claims of the present invention without departing from the spirit and scope of the present invention.
Claims
1. A road surface roughness detection method based on vehicle vibration response, It is characterized in that The following steps are involved: S1, attach acceleration sensors to the front and rear axles of a dual-axle test vehicle, and drive the test vehicle over uneven roads to measure and record the vibration acceleration responses of the front and rear axles. and S2, establish a generative adversarial neural network to generate the vibration acceleration response of the front and rear axles of the vehicle and and road surface roughness sample r; S3, Generate responses through the discriminator of the generative adversarial neural network and With measurement response and Perform consistency judgment and construct discrimination loss; S4, using the generated acceleration response and The corresponding displacement response is obtained by numerical integration and And the vibration equation of the two-axle vehicle is established and and The physical constraint equation of r constructs the generation loss including the physical constraint loss and the adversarial loss; S5, first reduce the discriminative loss to optimize the discriminator, and then reduce the generative loss to optimize the generator; S6, the road roughness sample r generated by the generative adversarial neural network after optimization is the roughness of the road to be identified.
2. A road surface roughness detection method based on vehicle vibration response according to claim 1, It is characterized in that The generative adversarial neural network includes a generator and a discriminator, and the generator and the discriminator are both composed of a five-layer feedforward neural network.
3. A road surface roughness detection method based on vehicle vibration response according to claim 1, It is characterized in that The physical constraint equation in step S4 is expressed as follows: Where: d 1 and d 2 are the distances between the center of gravity of the vehicle and the front and rear axles, respectively, and d is the distance between the front and rear axles of the vehicle, and d = d 1 +d 2 , k 1 is the spring stiffness of the tire on the front axle of the vehicle, k 2 is the spring stiffness of the tire on the rear axle of the vehicle, M v is the vehicle mass, J v is the moment of inertia of the vehicle's center of gravity.
4. A road surface roughness detection method based on vehicle vibration response according to claim 1, It is characterized in that The discriminant loss in step S3 is expressed as follows: Where: E(.) is the mathematical expectation function, D(.) is the discriminator of the generative adversarial neural network, t represents the time, and z is the input variable of the generative adversarial neural network, which is generated by the standard normal distribution.
5. A road surface roughness detection method based on vehicle vibration response according to claim 1, It is characterized in that The generation loss in step S4 is expressed as follows: Gen loss =Adv loss +β·Phy loss Where: Adv loss represents the adversarial loss of the generative adversarial neural network, Phy loss represents the physical constraint loss of the physical constraint equation, and β represents the weight coefficient of the physical constraint loss.
6. A road surface roughness detection device based on vehicle vibration response, used to implement a road surface roughness detection method based on vehicle vibration response as described in any one of claims 1 to 5, It is characterized in that The vehicle comprises a front axle (1), a rear axle (2), a vehicle frame (5), a traction frame (6), an acceleration sensor (8) and a data acquisition system (9); the vehicle frame (5) is fixedly supported on the front axle (1) and the rear axle (2), and the traction frame (6) and a ballast box (7) are installed on the vehicle frame (5); the acceleration sensor (8) is installed on the front axle (1) and the rear axle (2), respectively, and the data acquisition system (9) is arranged in the ballast box (7) of the vehicle frame (5) and connected to the acceleration sensor (8).
7. A road surface roughness detection device based on vehicle vibration response according to claim 6, It is characterized in that The front axle (1) and the rear axle (2) are respectively formed by connecting two left and right wheels (3) and a steel rod (4); the frame (5) is fixedly supported on the steel rods (4) of the front axle (1) and the rear axle (2); and the acceleration sensor (8) is respectively installed at the center position of the steel rods (4) of the front axle (1) and the rear axle (2).
Citation Information
Patent Citations
Flatness detection method and detection device based on vibration detection
CN105421200A
Road surface flatness detection method based on driving vibration data
CN111058360A
Pavement flatness prediction method and device
CN113932758A
Convolutional neural networks for pavement roughness assessment using calibration-free vehicle dynamics
US20240193407A1
