Method for deducing void position of approach slab of road and bridge transition section based on vehicle acceleration

Abstract

The invention relates to the technical field of road and bridge transition section health monitoring and intelligent traffic, in particular to a method for deducing a road and bridge transition section access slab void position based on vehicle acceleration, and the method specifically comprises the following steps: collecting a vibration signal of a vehicle passing through a road and bridge transition section through a vehicle-mounted acceleration sensor; denoising, noise reduction and normalization processing are carried out on the collected data, and then the real acceleration collected by the sensor is extracted through wavelet transform and fast Fourier transform; a road and bridge transition section physical mechanical model and a vehicle and road coupling mechanical model are constructed, and the road and bridge transition section access slab void condition is obtained; a bayesian probability model is adopted to calibrate the road and bridge transition section access slab void condition, a likelihood function and prior distribution are constructed, posterior distribution inference is carried out according to the likelihood function and the prior distribution, and posterior distribution is calculated through an MCMC method. According to the method, the road and bridge transition section access slab void position is speculated through the vehicle vibration data and the Bayesian method, and accurate quantification and uncertainty evaluation can be carried out on road diseases.

Description

Technical Field

[0001] The present invention relates to the fields of health monitoring of road-bridge transition sections and intelligent transportation technology, and in particular to a method for deducing the position of a slatted slat in a road-bridge transition section based on vehicle acceleration. Background Art

[0002] Bridgehead slat debonding refers to the gap between the bridgehead slat and the roadbed, resulting in the slat losing its effective support. This phenomenon can cause a series of problems, seriously affecting the safety and service life of the road. Problem description: Due to slat debonding, vehicles experience noticeable bumps and bounces when passing through the bridgehead, reducing driving comfort, increasing driver fatigue, causing additional impact on the vehicle suspension system, shortening the vehicle life, and in severe cases, causing the vehicle to lose control and cause traffic accidents. The debonding area causes uneven stress on the slat, local stress concentration, and the slat may crack, break, or even fall off, resulting in insufficient road support and the appearance of cracks, potholes and other defects. Water may accumulate in the debonding area, causing the roadbed to soften, further exacerbating the problem. Bridge supports may be damaged due to uneven stress, followed by uneven settlement, further exacerbating structural damage, increasing road maintenance costs, and affecting driving safety, especially for vehicles traveling at high speeds.

[0003] Existing methods for detecting the location of gaps in the slabs at the transition section of a road bridge rely on manual inspections or fixed sensors, which are costly, inefficient, and unable to be monitored in real time. Furthermore, gaps and settlements in the slabs at the transition section of a road bridge are highly concealed, making it difficult for conventional detection methods to accurately locate and quantify them. Furthermore, existing research lacks sufficient research on the correlation between vehicle vibration signals and pavement defects, and lacks a systematic inversion model.

[0004] Therefore, the present invention proposes a method for deducing the position of the slab clearance in the transition section of a road bridge based on vehicle acceleration to solve the above problem. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention develops a method for deducing the position of the gap between the trestle and the bridge transition section based on vehicle acceleration. The present invention aims to provide a low-cost, high-efficiency, and real-time monitoring method for detecting gap between the trestle and the bridge transition section. By inverting the disease parameters through vehicle acceleration data, the method provides decision support for road maintenance.

[0006] The technical solution to the technical problem of the present invention is a method for deducing the position of the slab clearance in the transition section of a road bridge based on vehicle acceleration, comprising the following steps: S1, collecting vibration signals of a vehicle passing through a road-bridge transition section through a vehicle-mounted acceleration sensor; S2, denoising, noise reduction and normalization are performed on the collected data, and then the real acceleration collected by the sensor is extracted through wavelet transform and fast Fourier transform; S3. Construct a physical and mechanical model of the bridge-road transition section and a vehicle-road coupling mechanical model. Using the vehicle-road coupling mechanical model, determine the relationship between the acceleration of a vehicle passing through the bridge-road transition section and the vehicle-road interaction force. Using the physical and mechanical model of the bridge-road transition section, establish a relationship between the bridge-road transition section trestle emptying condition and the acceleration of a vehicle passing through the bridge-road transition section, thereby determining the bridge-road transition section trestle emptying condition. S4. Use a Bayesian probability model to calibrate the condition of the rafting gap in the transition section of the road bridge. Collect vehicle acceleration data and vehicle type parameters to construct a likelihood function and prior distribution. Based on the likelihood function and prior distribution, infer the posterior distribution and calculate the posterior distribution using the MCMC Markov Chain Monte Carlo method. S5. Use the posterior distribution to estimate the damage caused by the gap, obtain the vehicle type parameters and the deflection value of the road-bridge transition section corresponding to the acceleration, calculate the maximum posterior estimate and confidence interval of the gap position, and continuously update the vehicle type parameters and the deflection value of the road-bridge transition section to update and predict the gap situation of the road-bridge transition section in real time.

[0007] S1 is as follows: Acceleration sensors are installed on different types of vehicles, and the sampling frequency is set. The accelerometer is installed close to the center of gravity of the vehicle, with the X-axis of the accelerometer aligned with the vehicle's forward direction. The vehicle carrying the accelerometer passes through the road-bridge transition section containing the gap at a constant speed, and vibration signals before and after the transition section are extracted. The vibration signals include vertical, lateral, and longitudinal acceleration data.

[0008] S2 is as follows: S2.1. Data preprocessing: Preprocess the collected data by using adaptive Kalman wave to eliminate high-frequency noise and outliers in the collected data, and perform wavelet denoising and normalization on the collected acceleration data; S2.2, Feature Extraction The impact signal in the preprocessed data is extracted by wavelet transform, and then the dominant frequency component is extracted by fast Fourier transform. Finally, the local frequency component of the non-stationary signal is extracted by wavelet transform to separate the real acceleration collected by the sensor.

[0009] S3 is as follows: S3.1. Constructing the physical and mechanical model of the road-bridge transition section: The abutment of the road bridge is determined to be a fixed end, the roadbed of the road bridge is supported by an elastic foundation, and the physical and mechanical model of the road bridge transition section is a Winkler foundation beam with one end simply supported, the other end free, and partly hollow. The mechanical model of the road bridge transition section obeys the Winkler basic assumption. According to Winkler's hypothesis, if the bridge transition section slab is located on a continuous elastic foundation and bends with the vertical load of the road surface along the bridge transition section, when the deflection value is smaller than the thickness of the bridge transition section slab, any point in the vertical direction of the bridge transition section slab will bend. The interaction force It is proportional to the deflection value of the slab; According to the deformation of the slab to reflect the unevenness of the road surface, any point in the vertical direction of the slab in the transition section of the road bridge Deflection value The relationship between the interaction force is as follows: , (1) in, represents the foundation bed coefficient, Indicates the settlement difference between the abutment and the roadbed; The size of the gap in the transition section of the bridge is simplified. The gap in the transition section of the bridge is simplified into a cube. The depth, length and width of the gap are recorded as ; Side length of the hollow surface The relationship between the deflection value is as follows: , The following relationship is further derived: , (2) in, and Represents two different parameters determined by the given road bridge transition section slab thickness, surface layer thickness and corresponding elastic modulus, base layer thickness and corresponding elastic modulus, soil base thickness and corresponding elastic modulus: According to formula (1) and formula (2), the side length of the void surface is obtained and the interaction force The relationship between the bridge transition section and the acceleration of the vehicle passing through the bridge transition section is established. S3.2. Constructing a vehicle-road coupling mechanical model: Vehicle-road interaction involves both forward and inverse problems. The forward problem involves calculating the vehicle acceleration, velocity, and displacement of a specified road section using a vehicle-road coupling model. The inverse problem involves evaluating the debonding of the bridge transition section using the obtained vehicle acceleration. When a vehicle passes through the debonding area, the uneven road surface impacts the vehicle's suspension system, causing a change in acceleration. The vehicle is simplified into a two-degree-of-freedom system, and the bridge transition section is simplified into a Winkler elastic foundation beam model. A multi-body dynamics model of the vehicle suspension system and the road surface is constructed. The vehicle is idealized as a combination of multiple rigid bodies connected by a series of springs and dampers. In the vehicle-road coupled mechanics model, the vehicle-road interaction force is solved according to the vehicle motion equation based on the collected vehicle acceleration data. , and then calculate the deflection value according to the physical and mechanical model of the road-bridge transition section, and finally obtain the gap size of the slab; The vehicle motion equation is established as follows: , , , in, Determine the sprung mass of the vehicle, represents the unsprung mass, represents the damping coefficient of the suspension system, represents the spring stiffness, is the vertical stiffness of the tire, is the vertical acceleration of the sprung mass, is the vertical velocity of the sprung mass, represents the vertical displacement of the sprung mass, Indicates the vertical acceleration of the unsprung mass and the vertical velocity of the unsprung mass , vertical displacement of unsprung mass , pitch angle around the vehicle's center of mass , pitch angular velocity , pitch angular acceleration and road pitch excitation , Indicates half of the wheelbase, represents the pitch moment of inertia, Indicates the road roughness input, represents the vehicle-road interaction force; The vehicle is represented by a quarter car model. The sprung mass and unsprung mass in the car model are connected by springs and viscous dampers. The vertical displacement, vertical velocity and pitch angular velocity of the sprung mass in the dynamic equations are solved. The vehicle and the road are coupled at the tire contact point through the interaction force vector. The Newmark-β method, which generalizes the linear acceleration method, is used to solve the inverse problem. 、 、 、 and Solve, assuming that at time step arrive The internal acceleration changes linearly, , the calculation formula is as follows: , , in, Represents the time step The vertical displacement of the sprung mass is Represents the time step The vertical velocity of the sprung mass is Represents the time step The vertical acceleration of the sprung mass is Represents the time step The vertical displacement of the sprung mass is Represents the time step The vertical velocity of the sprung mass is Represents the time step The vertical acceleration of the sprung mass is represents the time step, Indicates the weight of controlling the acceleration change, set =1 / 4, Indicates the weight of controlling speed change, set =1 / 2; Time step The pitch angular velocity around the vehicle's center of mass is The calculation formula is as follows: , in, Represents the time step Pitch angular velocity, Represents the time step Pitch angular acceleration, Represents the time step Pitch angular acceleration, Represents the time step Pitch angular velocity at ; According to the vertical acceleration of the sprung mass , vertical acceleration of sprung mass , pitch angular velocity and sprung mass , suspension system damping coefficient , spring stiffness , calculate the unsprung mass displacement using the Newmark-β method , the calculation formula is as follows: , in, represents the Laplace variable, , is the real part, is the imaginary part; The displacement of the unsprung mass is calculated by the Newmark-β method. Calculate time steps The acceleration of the unsprung mass and speed , the calculation formula is as follows; , , in, represents the change in time squared, Represents the time step The vertical displacement of the unsprung mass is Represents the time step The vertical displacement of the unsprung mass is Represents the time step The vertical displacement of the unsprung mass when ; Calculate the time-varying interaction forces between the vehicle and the road , the calculation formula is as follows: , (3) S3.3 is obtained by combining formulas (1) and (3): , (4) The relationship between the slat clearance and acceleration can be obtained from formulas (2) and (4): .

[0010] 5. The method for deducing the position of the slab gap in a road-bridge transition section based on vehicle acceleration according to claim 4, wherein S4 is specifically as follows: The Bayesian update method is used to obtain the specific values ​​of the vehicle type parameters collected by the drop test and the acceleration data collected by the on-board acceleration sensor. The deflection value of the bridge transition section is calculated by combining the vehicle-road coupling model, and then the side length of the bridge gap is obtained. , where any point on the road-bridge transition section represents , Indicates the first positions, and the position set on the road-bridge transition section is expressed as , represents the total number of positions, and the corresponding deflection value is expressed as , express Deflection value at ; (1) Construct a likelihood function based on the collected data: Using a simple independent Gaussian error model, The vehicle's acceleration iteration data is as follows: , Among them, the likelihood function represents the possibility of all acceleration data, Represents the acceleration set corresponding to each vehicle type and road deflection value, Indicates the The acceleration of the car, Indicates the A car in The acceleration at represents the deflection value, Indicates the The sprung mass of the vehicle, Indicates the The unsprung mass of the vehicle, Indicates the The damping coefficient of the vehicle suspension system, Indicates the The spring rate of the vehicle, Indicates the The vertical stiffness of the tire of a vehicle, represents the acquisition variance; (2) Constructing prior distribution: Specify a prior distribution for the sprung mass of each vehicle , represents the distribution of vehicle mass among all vehicles used in the road-bridge transition section, assuming that each vehicle has the same mass. represents the set of sprung mass mean and variance, the prior values ​​are independent for each vehicle, and all masses are obtained The prior value of Represents the total number of vehicles, and the calculation formula is as follows: , Similarly, the prior distributions of unsprung mass, suspension damping coefficient, spring stiffness, and tire vertical stiffness are calculated for each vehicle using the following formula: , , , , in, represents the set of mean and variance of unsprung mass, represents the set of mean and variance of the suspension system damping coefficient, represents the set of spring stiffness means and variances, represents the set of mean and variance of tire vertical stiffness; The prior distribution of the pavement deflection value adopts the inherent Gaussian process, and the calculation formula is as follows: , in, represents the smoothing parameter of the road profile, represents the prior distribution of pavement deflection values; (3) Posterior distribution inference and sampling: The posterior distribution is inferred from the likelihood function and the prior distribution: , in, It represents the probability that the vehicle type and transition section deflection are in a certain range when the acceleration data is known. represents the possibility of all acceleration data, express The sprung mass of the vehicle, express The unsprung mass of the vehicle, express The spring rate of the vehicle, express The vertical stiffness of the tire of a vehicle, express The damping coefficient of the vehicle's suspension system, represents the vehicle acceleration, .

[0011] The lag distribution is calculated using the MCMC Markov Chain Monte Carlo method, specifically Gibbs sampling. The operation process is as follows: (1) Define the initial set of vehicle types and initial deflection value , set the number of iterations , , and Respectively The vehicle's initial sprung mass, unsprung mass, spring stiffness, tire vertical stiffness, and suspension damping coefficient; (2) Use the vehicle motion equation of the vehicle-road coupled mechanical model to calculate the initial expected acceleration set corresponding to the vehicle type and deflection value ; (3) Use the current deflection set Alternative initial deflection value (4) Each iteration is based on the acceptance probability and Update vehicle types and road-bridge transition section deflection values; (4-1) Update vehicle type parameters: In the In the iteration, for each vehicle type, new parameters are sampled from the normal distribution to generate candidate parameters. The generated candidate parameters are expressed as follows: , , , , , in, 、 、 、 and Indicates the Candidate parameters of the vehicle, 、 、 、 and denote the variance of sprung mass, unsprung mass, spring stiffness, tire vertical stiffness and suspension damping coefficient respectively. 、 、 、 and Respectively represent The sprung mass, unsprung mass, spring stiffness, tire vertical stiffness, and suspension damping coefficient for the iteration; The acceleration set corresponding to each vehicle type and road deflection value is calculated through the vehicle motion equation of the vehicle-road coupling model of the bridge-road transition section ,…, ,…, ; Calculating the probability of acceptance : , in, Indicates the The deflection value of the iteration, Indicates taking the minimum value; Generate a random number uniformly distributed between 0 and 1 ,like , then accept the candidate parameters, if , then the candidate parameter is not accepted; When accepting, , , , , ; If you do not accept , , , , ; The vehicle type parameter of the vehicle is updated and expressed as (4-2) Update the deflection value of the road bridge transition section: In the In the iteration, for the deflection value of the bridge transition section, candidate deflection values ​​are sampled from the normal distribution, and the generated candidate deflection values ​​are expressed as follows: , in, represents the candidate deflection value, Indicates the The deflection value of the iteration, represents the variance of deflection value; Constructing a new deflection set , , with the new deflection set Override the current deflection set , the acceleration set corresponding to each vehicle type and road deflection value is calculated through the vehicle motion equation of the vehicle-road coupling model of the bridge-road transition section ; Calculating the probability of acceptance : , Generate a random number uniformly distributed between 0 and 1 ,like , then accept the new deflection set ,like , the new deflection set is not accepted ; When accepting, ; If you do not accept ; (4-3) The data set for updating vehicle type and bridge transition section deflection value is expressed as , Indicates the The data set updated in the iteration; Repeat steps (1) to (4) until .

[0012] The effects provided in the summary of the invention are only the effects of the embodiments, rather than all the effects of the invention. The above technical solution has the following advantages or beneficial effects: The present invention uses on-board acceleration sensors to collect vehicle vibration signals in real time. Combined with adaptive Kalman filtering and wavelet noise reduction technology, it can effectively eliminate high-frequency noise and outlier interference, improve detection efficiency, and do not require traffic closures for data collection, which can save time and reduce operation and maintenance costs. The present invention constructs a physical and mechanical model of the road-bridge transition section and a vehicle-road coupling mechanical model to link the interaction force, deflection value, vehicle acceleration, and gap size, and ensures logical rigor through mechanical closed-loop verification. The present invention adopts a Bayesian-MCMC framework to construct a Gaussian error model through a likelihood function to quantify measurement uncertainty. The smoothness constraints of the Gaussian process are determined by constructing a prior distribution. In the posterior inference, vehicle parameters and deflection values ​​are iteratively updated through Gibbs sampling, which can effectively integrate multi-vehicle data and suppress the interference of single vehicle parameter errors on the results. Finally, through adaptive damage quantification and confidence assessment, vehicle parameters and deflection values ​​are dynamically updated to achieve real-time prediction and reliability assessment of gap conditions, providing a probabilistic basis for road and bridge maintenance decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0014] Figure 1 Schematic diagram of the method of the present invention.

[0015] Figure 2 Schematic diagram of comparison between the predicted value and the true value of the method of the present invention. DETAILED DESCRIPTION

[0016] In order to clearly illustrate the technical features of this solution, the present invention is described in detail below through specific implementation methods and in conjunction with the accompanying drawings.

[0017] Example 1 like Figure 1 As shown, a method for deducing the position of the slab gap in the transition section of a road bridge based on vehicle acceleration is characterized by comprising the following steps: S1, collecting vibration signals of a vehicle passing through a road-bridge transition section through a vehicle-mounted acceleration sensor; S2, denoising, noise reduction and normalization are performed on the collected data, and then the real acceleration collected by the sensor is extracted through wavelet transform and fast Fourier transform; S3. Construct a physical and mechanical model of the bridge-road transition section and a vehicle-road coupling mechanical model. Using the vehicle-road coupling mechanical model, determine the relationship between the acceleration of a vehicle passing through the bridge-road transition section and the vehicle-road interaction force. Using the physical and mechanical model of the bridge-road transition section, establish a relationship between the bridge-road transition section trestle emptying condition and the acceleration of a vehicle passing through the bridge-road transition section, thereby determining the bridge-road transition section trestle emptying condition. S4. Use a Bayesian probability model to calibrate the condition of the rafting gap in the transition section of the road bridge. Collect vehicle acceleration data and vehicle type parameters to construct a likelihood function and prior distribution. Based on the likelihood function and prior distribution, infer the posterior distribution and calculate the posterior distribution using the MCMC Markov Chain Monte Carlo method. S5. Use the posterior distribution to estimate the damage caused by the gap, obtain the vehicle type parameters and the deflection value of the road-bridge transition section corresponding to the acceleration, calculate the maximum posterior estimate and confidence interval of the gap position, and continuously update the vehicle type parameters and the deflection value of the road-bridge transition section to update and predict the gap situation of the road-bridge transition section in real time.

[0018] In a specific implementation manner, S1 is specifically as follows: Accelerometers are installed on different types of vehicles, and the sampling frequency is set to 100 Hz. The accelerometer is installed close to the center of gravity of the vehicle, with the X-axis of the accelerometer aligned with the vehicle's forward direction. The vehicle carrying the accelerometer passes through the road-bridge transition section containing the gap at a constant speed, and vibration signals are extracted before and after the transition section. The vibration signals include vertical, lateral, and longitudinal acceleration data, and the location for extracting the vibration signals is 10 meters before and after the transition section.

[0019] In a specific implementation manner, S2 is specifically as follows: S2.1. Data preprocessing: Preprocess the collected data by using adaptive Kalman wave to eliminate high-frequency noise and outliers in the collected data, and perform wavelet denoising and normalization on the collected acceleration data; S2.2, Feature Extraction The impact signal in the preprocessed data is extracted by wavelet transform, and the high-frequency vibration of 10-50 Hz is easily excited by air. Then, the dominant frequency component is extracted by fast Fourier transform, and the local frequency component of the non-stationary signal is extracted by wavelet transform to separate the real acceleration collected by the sensor.

[0020] In a specific implementation, S3 is as follows: S3.1. Constructing the physical and mechanical model of the road-bridge transition section: The abutment of the road bridge is determined to be a fixed end, the roadbed of the road bridge is supported by an elastic foundation, and the physical and mechanical model of the road bridge transition section is a Winkler foundation beam with one end simply supported, the other end free, and partly hollow. The mechanical model of the road bridge transition section obeys the Winkler basic assumption. According to Winkler's hypothesis, if the bridge transition section slab is located on a continuous elastic foundation and bends with the vertical load of the road surface along the bridge transition section, when the deflection value is smaller than the thickness of the bridge transition section slab, any point in the vertical direction of the bridge transition section slab will bend. The interaction force It is proportional to the deflection value of the slab; According to the deformation of the slab to reflect the unevenness of the road surface, any point in the vertical direction of the slab in the transition section of the road bridge Deflection value The relationship between the interaction force is as follows: , (1) in, represents the foundation bed coefficient, Indicates the settlement difference between the abutment and the roadbed; The size of the gap in the transition section of the bridge is simplified. The gap in the transition section of the bridge is simplified into a cube. The depth, length and width of the gap are recorded as ; Side length of the hollow surface The relationship between the deflection value is as follows: , The following relationship is further derived: , (2) in, and Represents two different parameters determined by the given road bridge transition section slab thickness, surface layer thickness and corresponding elastic modulus, base layer thickness and corresponding elastic modulus, soil base thickness and corresponding elastic modulus: When the thickness of the bridge transition section slab is 0.3m, the surface layer thickness is 0.25m, the corresponding elastic modulus is 30000MPa, the base layer thickness is 0.5m, the corresponding elastic modulus is 2500MPa, the soil base thickness is 8m, the corresponding elastic modulus is 50MPa, and then the above parameters are curve fitted according to the finite element abaqus numerical simulation to obtain , ; According to formula (1) and formula (2), the side length of the void surface is obtained and the interaction force The relationship between the bridge transition section and the acceleration of the vehicle passing through the bridge transition section is established. S3.2. Constructing a vehicle-road coupling mechanical model: Vehicle-road interaction involves both forward and inverse problems. The forward problem involves calculating the vehicle acceleration, velocity, and displacement of a specified road section using a vehicle-road coupling model. The inverse problem involves evaluating the debonding of the bridge transition section using the obtained vehicle acceleration. When a vehicle passes through the debonding area, the uneven road surface impacts the vehicle's suspension system, causing a change in acceleration. The vehicle is simplified into a two-degree-of-freedom system, and the bridge transition section is simplified into a Winkler elastic foundation beam model. A multi-body dynamics model of the vehicle suspension system and the road surface is constructed. The vehicle is idealized as a combination of multiple rigid bodies connected by a series of springs and dampers. In the vehicle-road coupled mechanics model, the vehicle-road interaction force is solved according to the vehicle motion equation based on the collected vehicle acceleration data. , and then calculate the deflection value according to the physical and mechanical model of the road-bridge transition section, and finally obtain the gap size of the slab; The vehicle motion equation is established as follows: , , , in, Determine the sprung mass of the vehicle, represents the unsprung mass, represents the damping coefficient of the suspension system, represents the spring stiffness, is the vertical stiffness of the tire, is the vertical acceleration of the sprung mass, is the vertical velocity of the sprung mass, represents the vertical displacement of the sprung mass, Indicates the vertical acceleration of the unsprung mass and the vertical velocity of the unsprung mass , vertical displacement of unsprung mass , pitch angle around the vehicle's center of mass , pitch angular velocity , pitch angular acceleration and road pitch excitation , It represents half of the wheelbase, that is, the distance from the center of mass to the front and rear wheels. represents the pitch moment of inertia, Indicates the road roughness input, represents the vehicle-road interaction force; The vehicle is represented by a quarter car model. The sprung mass and unsprung mass in the car model are connected by springs and viscous dampers. The vertical displacement, vertical velocity and pitch angular velocity of the sprung mass in the dynamic equations are solved. The vehicle and the road are coupled at the tire contact point through the interaction force vector. The Newmark-β method, which generalizes the linear acceleration method, is used to solve the inverse problem. 、 、 、 and Solve, assuming that at time step arrive The internal acceleration changes linearly, , the calculation formula is as follows: , , in, Represents the time step The vertical displacement of the sprung mass is Represents the time step The vertical velocity of the sprung mass is Represents the time step The vertical acceleration of the sprung mass is Represents the time step The vertical displacement of the sprung mass is Represents the time step The vertical velocity of the sprung mass is Represents the time step The vertical acceleration of the sprung mass is represents the time step, Indicates the weight of controlling the acceleration change, set =1 / 4, Indicates the weight of controlling speed change, set =1 / 2; Time step The pitch angular velocity around the vehicle's center of mass is The calculation formula is as follows: , in, Represents the time step Pitch angular velocity, Represents the time step Pitch angular acceleration, Represents the time step Pitch angular acceleration, Represents the time step Pitch angular velocity at ; According to the vertical acceleration of the sprung mass , vertical acceleration of sprung mass , pitch angular velocity and sprung mass , suspension system damping coefficient , spring stiffness , calculate the unsprung mass displacement using the Newmark-β method , the calculation formula is as follows: , in, represents the Laplace variable, , is the real part, is the imaginary part; The displacement of the unsprung mass is calculated by the Newmark-β method. Calculate time steps The acceleration of the unsprung mass and speed , the calculation formula is as follows; , , in, represents the change in time squared, Represents the time step The vertical displacement of the unsprung mass is Represents the time step The vertical displacement of the unsprung mass is Represents the time step The vertical displacement of the unsprung mass when ; Calculate the time-varying interaction forces between the vehicle and the road , the calculation formula is as follows: , (3) S3.3 is obtained by combining formulas (1) and (3): , (4) The relationship between the slat clearance and acceleration can be obtained from formulas (2) and (4): .

[0021] In a specific implementation manner, S4 is specifically as follows: The Bayesian update method is used to obtain the specific values ​​of the vehicle type parameters collected by the drop test and the acceleration data collected by the on-board acceleration sensor. The deflection value of the bridge transition section is calculated by combining the vehicle-road coupling model, and then the side length of the bridge gap is obtained. , where any point on the road-bridge transition section represents , Indicates the first positions, and the position set on the road-bridge transition section is expressed as , represents the total number of positions, and the corresponding deflection value is expressed as , express Deflection value at ; (1) Construct a likelihood function based on the collected data: Using a simple independent Gaussian error model, The vehicle's acceleration iteration data is as follows: , Among them, the likelihood function represents the possibility of all acceleration data, Represents the acceleration set corresponding to each vehicle type and road deflection value, Indicates the The acceleration of the car, Indicates the A car in The acceleration at represents the deflection value, Indicates the The sprung mass of the vehicle, Indicates the The unsprung mass of the vehicle, Indicates the The damping coefficient of the vehicle suspension system, Indicates the The spring rate of the vehicle, Indicates the The vertical stiffness of the tire of a vehicle, represents the acquisition variance; (2) Constructing prior distribution: Specify a prior distribution for the sprung mass of each vehicle , represents the distribution of vehicle mass among all vehicles used in the road-bridge transition section, assuming that each vehicle has the same mass. represents the set of sprung mass mean and variance, the prior values ​​are independent for each vehicle, and all masses are obtained The prior value of Represents the total number of vehicles, and the calculation formula is as follows: , Similarly, the prior distributions of unsprung mass, suspension damping coefficient, spring stiffness, and tire vertical stiffness are calculated for each vehicle using the following formula: , , , , in, represents the set of mean and variance of unsprung mass, represents the set of mean and variance of the suspension system damping coefficient, represents the set of spring stiffness means and variances, represents the set of mean and variance of tire vertical stiffness; The prior distribution of the pavement deflection value adopts the inherent Gaussian process, and the calculation formula is as follows: , in, represents the smoothing parameter of the road profile, represents the prior distribution of pavement deflection values; (3) Posterior distribution inference and sampling: The posterior distribution is inferred from the likelihood function and the prior distribution: , in, It represents the probability that the vehicle type and transition section deflection are in a certain range when the acceleration data is known. represents the possibility of all acceleration data, express The sprung mass of the vehicle, express The unsprung mass of the vehicle, express The spring rate of the vehicle, express The vertical stiffness of the tire of a vehicle, express The damping coefficient of the vehicle's suspension system, represents the vehicle acceleration, .

[0022] In a specific implementation, the lag distribution is calculated using the MCMC Markov Chain Monte Carlo method, specifically Gibbs sampling is selected, and the operation process is as follows: (1) Define the initial set of vehicle types and initial deflection value , set the number of iterations , , and Respectively The vehicle's initial sprung mass, unsprung mass, spring stiffness, tire vertical stiffness, and suspension damping coefficient; (2) Use the vehicle motion equation of the vehicle-road coupled mechanical model to calculate the initial expected acceleration set corresponding to the vehicle type and deflection value ; (3) Use the current deflection set Alternative initial deflection value (4) Each iteration is based on the acceptance probability and Update vehicle types and road-bridge transition section deflection values; (4-1) Update vehicle type parameters: In the In the iteration, for each vehicle type, new parameters are sampled from the normal distribution to generate candidate parameters. The generated candidate parameters are expressed as follows: , , , , , in, 、 、 、 and Indicates the Candidate parameters of the vehicle, 、 、 、 and denote the variance of sprung mass, unsprung mass, spring stiffness, tire vertical stiffness and suspension damping coefficient respectively. 、 、 、 and Respectively represent The sprung mass, unsprung mass, spring stiffness, tire vertical stiffness, and suspension damping coefficient for the iteration; The acceleration set corresponding to each vehicle type and road deflection value is calculated through the vehicle motion equation of the vehicle-road coupling model of the bridge-road transition section ,…, ,…, ; Calculating the probability of acceptance : , in, Indicates the The deflection value of the iteration, Indicates taking the minimum value; Generate a random number uniformly distributed between 0 and 1 ,like , then accept the candidate parameters, if , then the candidate parameter is not accepted; When accepting, , , , , ; If you do not accept , , , , ; The vehicle type parameter of the vehicle is updated and expressed as (4-2) Update the deflection value of the road bridge transition section: In the In the iteration, for the deflection value of the bridge transition section, candidate deflection values ​​are sampled from the normal distribution, and the generated candidate deflection values ​​are expressed as follows: , in, represents the candidate deflection value, Indicates the The deflection value of the iteration, represents the variance of deflection value; Constructing a new deflection set , , with the new deflection set Override the current deflection set , the acceleration set corresponding to each vehicle type and road deflection value is calculated through the vehicle motion equation of the vehicle-road coupling model of the bridge-road transition section ; Calculating the probability of acceptance : , Generate a random number uniformly distributed between 0 and 1 ,like , then accept the new deflection set ,like , the new deflection set is not accepted ; When accepting, ; If you do not accept ; (4-3) The data set for updating vehicle type and bridge transition section deflection value is expressed as , Indicates the The data set updated in the iteration; Repeat steps (1) to (4) until .

[0023] Example 2 In order to prove the beneficial effects of the present invention, the hollow depth result obtained by the method of the present invention is verified. The hollow depth is the side length of the hollow surface calculated in the method of the present invention, such as Figure 2 As shown, the Bayesian prediction value in the present invention is compared with the true value. The experimental vehicle passes through the road bridge transition test section at 60 km / h. The acceleration data of the acceleration sensor when the vehicle passes is extracted. The MCMC method is run for 5000 sampling times. The 95% value of the MCMC sample is used as the calculation result. The posterior mean of the void position is K10+230 (pile number), and the posterior mean of the void depth is d=33.2 mm (95% confidence interval: 32.1-34.6 mm). The actual void depth is 34 mm. The calculated deflection value is very close to the true value. Laboratory verification shows that the error is only 2%, the settlement error is ≤3 mm, and the void depth inversion error is ≤5 mm. The method of the present invention is used to detect the detachment position of the bridge transition section, which can reduce the detection cost by more than 80% and increase the coverage to the entire road section. Compared with the geological radar detection results, the error of the method of the present invention is less than 5%.

[0024] Example 3 When testing multiple vehicles, the actual gap depth was known to be 34 mm, and the vehicle speed was 60 km / h. The posterior mean d after the first five vehicles passed was 32.7 mm (with an 8 mm confidence interval). The updated posterior mean d after the sixth to tenth vehicles passed was 33.4 mm (with a narrowed confidence interval of 4 mm). When the 15th vehicle's data was added, the confidence interval remained ≤3 mm, reaching the maintenance decision threshold. This demonstrates that the invented method can dynamically update the gap depth of a bridge-road transition section when multiple vehicles pass through.

[0025] Although the above describes the specific implementation methods of the invention in conjunction with the accompanying drawings, it does not limit the scope of protection of the invention. Based on the technical solution of the present invention, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present invention.

Claims

1. A method for deducing the position of the slab gap in the transition section of a road bridge based on vehicle acceleration, characterized by: The following steps are involved: S1, collecting vibration signals of a vehicle passing through a road-bridge transition section through a vehicle-mounted acceleration sensor; S2, denoising, noise reduction and normalization are performed on the collected data, and then the real acceleration collected by the sensor is extracted through wavelet transform and fast Fourier transform; S3. Construct a physical and mechanical model of the bridge-road transition section and a vehicle-road coupling mechanical model. Using the vehicle-road coupling mechanical model, determine the relationship between the acceleration of a vehicle passing through the bridge-road transition section and the vehicle-road interaction force. Using the physical and mechanical model of the bridge-road transition section, establish a relationship between the bridge-road transition section trestle emptying condition and the acceleration of a vehicle passing through the bridge-road transition section, thereby determining the bridge-road transition section trestle emptying condition. S4. Use a Bayesian probability model to calibrate the condition of the rafting gap in the transition section of the road bridge. Collect vehicle acceleration data and vehicle type parameters to construct a likelihood function and prior distribution. Based on the likelihood function and prior distribution, infer the posterior distribution and calculate the posterior distribution using the MCMC Markov Chain Monte Carlo method. S5. Use the posterior distribution to estimate the damage caused by the gap, obtain the vehicle type parameters and the deflection value of the road-bridge transition section corresponding to the acceleration, calculate the maximum posterior estimate and confidence interval of the gap position, and continuously update the vehicle type parameters and the deflection value of the road-bridge transition section to update and predict the gap situation of the road-bridge transition section in real time.

2. The method for deducing the position of the slab gap in the transition section of a road bridge based on vehicle acceleration according to claim 1 is characterized in that: S1 is as follows: Acceleration sensors are installed on different types of vehicles, and the sampling frequency is set. The accelerometer is installed close to the center of gravity of the vehicle, with the X-axis of the accelerometer aligned with the vehicle's forward direction. The vehicle carrying the accelerometer passes through the road-bridge transition section containing the gap at a constant speed, and vibration signals before and after the transition section are extracted. The vibration signals include vertical, lateral, and longitudinal acceleration data.

3. The method for deducing the position of the slab gap in the transition section of a road bridge based on vehicle acceleration according to claim 2 is characterized in that: S2 is as follows: S2.

1. Data preprocessing: Preprocess the collected data by using adaptive Kalman wave to eliminate high-frequency noise and outliers in the collected data, and perform wavelet denoising and normalization on the collected acceleration data; S2.2, Feature Extraction The impact signal in the preprocessed data is extracted by wavelet transform, and then the dominant frequency component is extracted by fast Fourier transform. Finally, the local frequency component of the non-stationary signal is extracted by wavelet transform to separate the real acceleration collected by the sensor.

4. The method for deducing the position of the slab gap in the transition section of a road bridge based on vehicle acceleration according to claim 3 is characterized in that S3 The details are as follows: S3.

1. Constructing the physical and mechanical model of the road-bridge transition section: The abutment of the road bridge is determined to be a fixed end, the roadbed of the road bridge is supported by an elastic foundation, and the physical and mechanical model of the road bridge transition section is a Winkler foundation beam with one end simply supported, the other end free, and partly hollow. The mechanical model of the road bridge transition section obeys the Winkler basic assumption. According to Winkler's hypothesis, if the bridge transition section slab is located on a continuous elastic foundation and bends with the vertical load of the road surface along the bridge transition section, when the deflection value is smaller than the thickness of the bridge transition section slab, any point in the vertical direction of the bridge transition section slab will bend. The interaction force It is proportional to the deflection value of the slab; According to the deformation of the slab to reflect the unevenness of the road surface, any point in the vertical direction of the slab in the transition section of the road bridge Deflection value The relationship between the interaction force is as follows: ，（1） in, represents the foundation bed coefficient, Indicates the settlement difference between the abutment and the roadbed; The size of the gap in the transition section of the bridge is simplified. The gap in the transition section of the bridge is simplified into a cube. The depth, length and width of the gap are recorded as ; Side length of the hollow surface The relationship between it and the deflection value is as follows: ， The following relationship is further derived: ，（2） in, and Represents two different parameters determined by the given road bridge transition section slab thickness, surface layer thickness and corresponding elastic modulus, base layer thickness and corresponding elastic modulus, soil base thickness and corresponding elastic modulus: According to formula (1) and formula (2), the side length of the void surface is obtained and the interaction force The relationship between the bridge transition section and the acceleration of the vehicle passing through the bridge transition section is established. S3.

2. Constructing a vehicle-road coupling mechanical model: Vehicle-road interaction involves both forward and inverse problems. The forward problem involves calculating the vehicle acceleration, velocity, and displacement of a specified road section using a vehicle-road coupling model. The inverse problem involves evaluating the debonding of the bridge transition section using the obtained vehicle acceleration. When a vehicle passes through the debonding area, the uneven road surface impacts the vehicle's suspension system, causing a change in acceleration. The vehicle is simplified into a two-degree-of-freedom system, and the bridge transition section is simplified into a Winkler elastic foundation beam model. A multi-body dynamics model of the vehicle suspension system and the road surface is constructed. The vehicle is idealized as a combination of multiple rigid bodies connected by a series of springs and dampers. In the vehicle-road coupled mechanics model, the vehicle-road interaction force is solved according to the vehicle motion equation based on the collected vehicle acceleration data. , and then calculate the deflection value according to the physical and mechanical model of the road-bridge transition section, and finally obtain the gap size of the slab; The vehicle motion equation is established as follows: ， ， ， in, Determine the sprung mass of the vehicle, represents the unsprung mass, represents the damping coefficient of the suspension system, represents the spring stiffness, is the vertical stiffness of the tire, is the vertical acceleration of the sprung mass, is the vertical velocity of the sprung mass, represents the vertical displacement of the sprung mass, Indicates the vertical acceleration of the unsprung mass and the vertical velocity of the unsprung mass , vertical displacement of unsprung mass , pitch angle around the vehicle's center of mass , pitch angular velocity , pitch angular acceleration and road pitch excitation , Indicates half of the wheelbase, represents the pitch moment of inertia, Indicates the road roughness input, represents the vehicle-road interaction force; The vehicle is represented by a quarter car model. The sprung mass and unsprung mass in the car model are connected by springs and viscous dampers. The vertical displacement, vertical velocity and pitch angular velocity of the sprung mass in the dynamic equations are solved. The vehicle and the road are coupled at the tire contact point through the interaction force vector. The Newmark-β method, which generalizes the linear acceleration method, is used to solve the inverse problem. 、 、 、 and Solve, assuming that at time step arrive The internal acceleration changes linearly, , the calculation formula is as follows: ， ， in, Represents the time step The vertical displacement of the sprung mass is Represents the time step The vertical velocity of the sprung mass is Represents the time step The vertical acceleration of the sprung mass is Represents the time step The vertical displacement of the sprung mass is Represents the time step The vertical velocity of the sprung mass is Represents the time step The vertical acceleration of the sprung mass is represents the time step, Indicates the weight of controlling the acceleration change, set =1 / 4, Indicates the weight of controlling speed change, set =1 / 2; Time step The pitch angular velocity around the vehicle's center of mass is The calculation formula is as follows: ， in, Represents the time step Pitch angular velocity, Represents the time step Pitch angular acceleration, Represents the time step Pitch angular acceleration, Represents the time step Pitch angular velocity at ; According to the vertical acceleration of the sprung mass , vertical acceleration of sprung mass , pitch angular velocity and sprung mass , suspension system damping coefficient , spring stiffness , calculate the unsprung mass displacement using the Newmark-β method , the calculation formula is as follows: ， in, represents the Laplace variable, , is the real part, is the imaginary part; The displacement of the unsprung mass is calculated by the Newmark-β method. Calculate time steps The acceleration of the unsprung mass and speed , the calculation formula is as follows; ， ， in, represents the change in time squared, Represents the time step The vertical displacement of the unsprung mass is Represents the time step The vertical displacement of the unsprung mass is Represents the time step The vertical displacement of the unsprung mass when ; Calculate the time-varying interaction forces between the vehicle and the road , the calculation formula is as follows: ，（3） S3.3 is obtained by combining formulas (1) and (3): ，（4） The relationship between the slat clearance and acceleration can be obtained from formulas (2) and (4): 。 5. The method for deducing the position of the slab gap in the transition section of a road bridge based on vehicle acceleration according to claim 4 is characterized in that: S4 is as follows: The Bayesian update method is used to obtain the specific values ​​of the vehicle type parameters collected by the drop test and the acceleration data collected by the on-board acceleration sensor. The deflection value of the bridge transition section is calculated by combining the vehicle-road coupling model, and then the side length of the bridge gap is obtained. , where any point on the road-bridge transition section represents , Indicates the first positions, and the position set on the road-bridge transition section is expressed as , represents the total number of positions, and the corresponding deflection value is expressed as , express Deflection value at ; (1) Construct a likelihood function based on the collected data: Using a simple independent Gaussian error model, The vehicle's acceleration iteration data is as follows: ， Among them, the likelihood function represents the possibility of all acceleration data, Represents the acceleration set corresponding to each vehicle type and road deflection value, Indicates the The acceleration of the car, Indicates the A car in The acceleration at represents the deflection value, Indicates the The sprung mass of the vehicle, Indicates the The unsprung mass of the vehicle, Indicates the The damping coefficient of the vehicle suspension system, Indicates the The spring rate of the vehicle, Indicates the The vertical stiffness of the tire of a vehicle, represents the acquisition variance; (2) Constructing prior distribution: Specify a prior distribution for the sprung mass of each vehicle , represents the distribution of vehicle mass among all vehicles used in the road-bridge transition section, assuming that each vehicle has the same mass. represents the set of sprung mass mean and variance, the prior values ​​are independent for each vehicle, and all masses are obtained The prior value of Represents the total number of vehicles, and the calculation formula is as follows: ， Similarly, the prior distributions of unsprung mass, suspension damping coefficient, spring stiffness, and tire vertical stiffness are calculated for each vehicle using the following formula: ， ， ， ， in, represents the set of mean and variance of unsprung mass, represents the set of mean and variance of the suspension system damping coefficient, represents the set of spring stiffness means and variances, represents the set of mean and variance of tire vertical stiffness; The prior distribution of the pavement deflection value adopts the inherent Gaussian process, and the calculation formula is as follows: ， in, represents the smoothing parameter of the road profile, represents the prior distribution of pavement deflection values; (3) Posterior distribution inference and sampling: The posterior distribution is inferred from the likelihood function and the prior distribution: ， in, It represents the probability that the vehicle type and transition section deflection are in a certain range when the acceleration data is known. represents the possibility of all acceleration data, express The sprung mass of the vehicle, express The unsprung mass of the vehicle, express The spring rate of the vehicle, express The vertical stiffness of the tire of a vehicle, express The damping coefficient of the vehicle's suspension system, represents the vehicle acceleration, .

6. The method for deducing the position of the slab gap in the transition section of a road bridge based on vehicle acceleration according to claim 5 is characterized in that: The lag distribution is calculated using the MCMC Markov Chain Monte Carlo method, specifically Gibbs sampling. The operation process is as follows: (1) Define the initial set of vehicle types and initial deflection value , set the number of iterations , and Respectively The vehicle's initial sprung mass, unsprung mass, spring stiffness, tire vertical stiffness, and suspension damping coefficient; (2) Use the vehicle motion equation of the vehicle-road coupled mechanical model to calculate the initial expected acceleration set corresponding to the vehicle type and deflection value ; (3) Use the current deflection set Alternative initial deflection value (4) Each iteration is based on the acceptance probability and Update vehicle types and road-bridge transition section deflection values; (4-1) Update vehicle type parameters: In the In the iteration, for each vehicle type, new parameters are sampled from the normal distribution to generate candidate parameters. The generated candidate parameters are expressed as follows: ， ， ， ， ， in, 、 、 、 and Indicates the Candidate parameters of the vehicle, 、 、 、 and denote the variance of sprung mass, unsprung mass, spring stiffness, tire vertical stiffness and suspension system damping coefficient, respectively. 、 、 、 and Respectively represent The sprung mass, unsprung mass, spring stiffness, tire vertical stiffness, and suspension damping coefficient for the iteration; The acceleration set corresponding to each vehicle type and road deflection value is calculated through the vehicle motion equation of the vehicle-road coupling model of the bridge-road transition section ,…, ，…， ； Calculating the probability of acceptance : ， in, Indicates the The deflection value of the iteration, Indicates taking the minimum value; Generate a random number uniformly distributed between 0 and 1 ,like , then accept the candidate parameters, if , then the candidate parameter is not accepted; When accepting, , , , , ; If you do not accept , , , , ; The updated vehicle type parameter of the vehicle is expressed as (4-2) Update the deflection value of the road bridge transition section: In the In the iteration, for the deflection value of the bridge transition section, candidate deflection values ​​are sampled from the normal distribution, and the generated candidate deflection values ​​are expressed as follows: ， in, represents the candidate deflection value, Indicates the The deflection value of the iteration, represents the variance of deflection value; Constructing a new deflection set , , with the new deflection set Override the current deflection set , the acceleration set corresponding to each vehicle type and road deflection value is calculated through the vehicle motion equation of the vehicle-road coupling model of the bridge-road transition section ; Calculating the probability of acceptance : ， Generate a random number uniformly distributed between 0 and 1 ,like , then accept the new deflection set ,like , the new deflection set is not accepted ; When accepting, ; If you do not accept ; (4-3) The data set for updating vehicle type and bridge transition section deflection value is expressed as , Indicates the The data set updated in the iteration; Repeat steps (1) to (4) until .

Images