Pavement noise attenuation performance test system and method

Abstract

The invention relates to the technical field of pavement noise attenuation performance testing, and discloses a pavement noise attenuation performance testing system and method, and the method comprises the steps: executing a transverse and longitudinal sound pressure distribution strategy, respectively calculating the transverse and longitudinal sound pressure distribution of an under-bridge channel, and optimizing the number of measurement points of an on-bridge road and an under-bridge road; executing a period segmentation weighting strategy, obtaining a traffic signal period of a road under the bridge, and optimizing sound pressure time characteristics under the bridge; executing a noise transfer matrix estimation strategy, calculating a noise transfer matrix from the under-bridge road to the on-bridge road according to the sound pressure time characteristics, and estimating under-bridge noise received by the on-bridge road; executing an on-bridge noise correction strategy, calculating an on-bridge real sound pressure spectrum according to the on-bridge sound pressure time characteristics and the under-bridge noise, and weighting the frequency band of the real sound pressure spectrum; and executing an on-bridge noise attenuation fitting strategy, calculating the equivalent sound pressure level of each measurement point of the on-bridge road, and linearly fitting on-bridge noise attenuation in combination with the reference position, thereby improving the accuracy of the bridge noise test.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of road surface noise attenuation performance testing, in particular to a road surface noise attenuation performance testing system and method. BACKGROUND

[0002] The road surface noise attenuation performance testing of newly built bridges can accurately reflect the noise propagation characteristics of the bridge under the action of traffic load, providing a scientific basis for bridge design and sound insulation facility arrangement. By quantifying the noise attenuation on the bridge, the bridge structure and traffic organization scheme are optimized to ensure that the bridge traffic environment meets the environmental noise standard, while providing reliable data support for urban traffic noise management and resident quality of life protection.

[0003] The existing bridge road surface noise testing method relies on single-point or a small number of measurement points for collection, lacks systematic spatial measurement point planning, and cannot fully reflect the actual impact of bridge underpass noise on the bridge road. In the traditional method of bridge sound pressure measurement, the effective technical differentiation or removal of bridge underpass noise is usually not used, and the background noise during the red light period and the vehicle dynamics interfere with the bridge noise, which is difficult to remove, resulting in that the bridge noise data contains a large amount of non-target noise, and it is difficult to truly reflect the bridge surface noise level. The existing technology usually uses a simple distance attenuation model or static linear estimation in the calculation of noise transmission from the bridge underpass to the bridge, ignores the frequency dependence and spatial correlation, and also lacks ear sensitivity weighting processing for different frequency bands, which cannot accurately quantify the actual contribution of the bridge underpass noise to the bridge sound field, thereby affecting the accuracy of the equivalent sound pressure level calculation and noise attenuation evaluation. In summary, the existing technology cannot effectively remove the bridge underpass noise in the bridge noise, and cannot meet the needs of fine noise control and scientific evaluation of newly built bridges.

[0004] The present application provides a road surface noise attenuation performance testing system and method to solve the problems mentioned in the background art. SUMMARY

[0005] The present application provides a road surface noise attenuation performance testing system and method to promote the solution to the problems mentioned in the background art.

[0006] The present application provides the following technical solution: a road surface noise attenuation performance testing method, comprising:

[0007] Testing the road surface noise attenuation performance of the bridge road of the newly built bridge, wherein the bridge road intersects with the underpass road, and the bridge road is not in traffic, and the underpass road is in traffic;

[0008] The underpass road is a one-way vehicle road under the bridge;

[0009] A plurality of horizontal sections are set in the underpass, and a plurality of measurement points are set on the horizontal sections;

[0010] A three-dimensional coordinate system is established, and a plurality of measuring points are set on the bridge road along the driving direction;

[0011] A horizontal and vertical sound pressure distribution strategy is executed, and the horizontal and vertical sound pressure distributions of the bridge passage are calculated respectively, so as to optimize the number of measuring points on the bridge road and the bridge passage;

[0012] According to the optimized measuring points of the bridge passage, the bridge passage sound pressure time characteristics are collected;

[0013] A period segmentation weighting strategy is executed to obtain the traffic signal period of the bridge passage, and the bridge passage sound pressure time characteristics are optimized in combination with the phases divided by the traffic signal period;

[0014] A reference position is set on the bridge road to fix the noise source, and the bridge sound pressure time characteristics are collected according to the optimized measuring points of the bridge road;

[0015] A noise transfer matrix estimation strategy is executed to calculate the noise transfer matrix from the bridge passage to the bridge road according to the sound pressure time characteristics, and to estimate the bridge passage noise received by the bridge road;

[0016] A bridge noise correction strategy is executed to calculate the real sound pressure spectrum of the bridge road according to the bridge sound pressure time characteristics and the bridge passage noise, and to weight the frequency bands of the real sound pressure spectrum;

[0017] A bridge noise attenuation fitting strategy is executed to calculate the equivalent sound pressure level of each measuring point of the bridge road, and to linearly fit the bridge noise attenuation in combination with the reference position.

[0018] Optionally, the plurality of horizontal sections are set on the bridge passage, and a plurality of measuring points are set on the horizontal sections, comprising:

[0019] Each horizontal section is equally spaced;

[0020] For any horizontal section, a plurality of local point positions are set on the horizontal section, the vertical distance from each local point position to the bridge road is measured, and the average value of the vertical distance is calculated as ;

[0021] The grid spacing is set ;

[0022] , is an empirical coefficient for adjusting the empirical parameter of the grid density, is a sound wave attenuation coefficient, which is the degree of attenuation per unit distance, and is normalized and updated to a normalized value;

[0023] is a natural exponential, for enhancing the sound wave attenuation coefficient;

[0024] The length of the bridge road is The square is denoted as the grid.

[0025] Divide the horizontal section into grids and use the midpoints of the grids on the horizontal section as measurement points.

[0026] Optionally, the implementation of the transverse and longitudinal sound pressure distribution strategy, which calculates the transverse and longitudinal sound pressure distribution of the underpass respectively, to optimize the number of measurement points on and under the bridge, includes:

[0027] The direction perpendicular to the horizontal plane is denoted as the longitudinal direction, and the direction parallel to the horizontal plane is denoted as the transverse direction.

[0028] Obtain measurement points for the underpass Measurement points on the road and bridge ;

[0029] Will Normalize and update to the normalized values;

[0030] Calculate longitudinal sound pressure distribution , ,in, The set reference distance, For measurement points and measurement points Euclidean distance, The sound pressure attenuation index is... For high attenuation coefficient, This indicates that the sound pressure decreases with distance;

[0031] Obtain any measurement point on the outermost side of the underpass along its longitudinal direction. ;

[0032] Calculate transverse sound pressure distribution , ,in, The highly correlated weights reflect the measurement points. For measurement points The proportion of contribution For the average index, control measurement Point synthesis measurement points The sound pressure level;

[0033] Measurement points The sound pressure at that location;

[0034] Count the number of measurement points under the bridge Number of measurement points on the road on the bridge ;

[0035] Solve the objective function using the branch and bound method. for maximizing the sound pressure distribution formed by the measurement points, a coefficient for penalizing the number of measurement points, used to adjust the number of measurement point arrangements.

[0036] Optionally, the execution of the period segmentation weighting strategy obtains the traffic signal period of the road under the bridge, and optimizes the sound pressure time feature under the bridge in combination with the phases divided according to the traffic signal period, including:

[0037] Obtaining the sound pressure collected at the same moment by each measurement point under the bridge, calculating the mean value, and taking the mean value as the sound pressure under the bridge at the same moment;

[0038] Obtaining the sound pressure under the bridge at each moment to form the sound pressure time feature under the bridge

[0039] The traffic signal period includes a red light phase and a green light phase.

[0040] The sound pressure time feature under the bridge is divided into multiple segments according to the red light phase and the green light phase, specifically: , , represents the th traffic signal period, , is the traffic signal period length, is the red light phase length, is the green light phase length, is the initial moment;

[0041] Setting a traffic state weight function ;

[0042] When located in the green light phase, , when located in the red light phase, ;

[0043] Segmented weighting of the sound pressure time feature under the bridge , updated to the segmented weighted value;

[0044] Hilbert transform is performed on the sound pressure time feature of the green light phase to obtain ;

[0045] The mean value is calculated , and the mean value is used to update the sound pressure time feature under the bridge.

[0046] Optionally, the execution of the noise transfer matrix estimation strategy calculates the noise transfer matrix from the road under the bridge to the road above the bridge according to the sound pressure time feature, and estimates the noise received by the road above the bridge, including:

[0047] ​The bridge under sound pressure time characteristics and the bridge over sound pressure time characteristics are time-aligned and corrected, and STFT is performed to obtain a bridge over complex spectrum matrix and a bridge under complex spectrum matrix . wherein, is a frequency point;

[0048] When the frequency point is fixed, the bridge over complex spectrum matrix and the bridge under complex spectrum matrix are expressed as matrices respectively corresponding to and in time frames;

[0049] A noise transfer matrix is set;

[0050] A target function is defined, and among all the matrices that can linearly map the bridge under sound pressure time characteristics into the bridge over sound pressure time characteristics, the best matrix is selected as the noise transfer matrix, for calculating the square root of the sum of squares of the matrix elements, is a regularization coefficient;

[0051] The target function is derived with respect to the noise transfer matrix to obtain ;

[0052] The solution is , is a conjugate transpose, is an identity matrix; represents a complex gain from a bridge under channel measurement point to a bridge over road measurement point at a frequency point ;

[0053] The bridge under noise received by the bridge over road is estimated .

[0054] Optionally, the bridge over noise correction strategy is executed, the bridge over real sound pressure spectrum is calculated according to the bridge over sound pressure time characteristics and the bridge under noise, and the frequency band of the real sound pressure spectrum is weighted, including:

[0055] The bridge over real sound pressure spectrum is calculated; ;

[0056] Three types of frequency bands are set based on human ear hearing sensitivity, including low frequency, medium frequency and high frequency;

[0057] A frequency band weight function is set, when the frequency point is low frequency, when the frequency point is medium frequency, when the frequency point is high frequency, ;

[0058] Segmented weighting , and the bridge real sound pressure spectrum is updated as the result of the segmental weighting.

[0059] Optionally, the bridge noise attenuation fitting strategy is executed to calculate the equivalent sound pressure level of each measurement point on the bridge road, and the bridge noise attenuation is linearly fitted in combination with the reference position, including:

[0060] For the measurement points on the bridge road , the bridge real sound pressure spectrum at the frequency point is obtained ;

[0061] The equivalent sound pressure level of the measurement point is calculated ;

[0062] The sound pressure of the reference position is obtained ;

[0063] The sound pressure at each time point is squared and averaged to obtain the equivalent sound pressure level of the measurement point , for degradation, for modulo length, is the total number of time frames;

[0064] The Euclidean distance of the measurement point to the reference point is calculated ;

[0065] The bridge noise attenuation is linearly fitted in combination with the reference position, specifically:

[0066] Then , wherein is the bridge noise attenuation coefficient, is the fitting residual, used to measure the error of the fitting;

[0067] The residual sum of squares is minimized , the bridge noise attenuation coefficient is output, and the equivalent sound pressure level is updated according to the output bridge noise attenuation coefficient.

[0068] A road surface noise attenuation performance test system, comprising:

[0069] A spatial measurement point planning module establishes a three-dimensional coordinate system of the bridge underpass channel and the bridge road, divides a horizontal cross-section grid of the bridge underpass, calculates the grid points and arranges measurement points on the bridge road along the driving direction;

[0070] The sound pressure distribution optimization module calculates the transverse and longitudinal sound pressure distribution of the passage under the bridge, calculates the distance and height attenuation between the measurement points, and determines the number and position of the measurement points on and under the bridge by using the branch and bound method.

[0071] The time period weighting module collects the time sequence of the sound pressure under the bridge, divides each phase according to the traffic light cycle, applies segmented weighting and performs Hilbert transform to generate the weighted time characteristics of the sound pressure under the bridge;

[0072] The noise transfer matrix calculation module aligns the time sequence of the sound pressure on and under the bridge and performs STFT analysis to calculate the regularized noise transfer matrix and estimate the complex gain of the noise under the bridge on the road;

[0073] The sound spectrum correction and attenuation fitting module calculates the real sound spectrum on the bridge according to the sound pressure on the bridge and the noise under the bridge, performs weighted processing on the low, medium and high frequency bands, and performs linear fitting with the reference point to obtain the noise attenuation coefficient on the bridge.

[0074] The present application has the following beneficial effects:

[0075] 1. The road surface noise attenuation performance test method realizes the quantitative analysis of the noise attenuation on the bridge by establishing measurement points on and under the bridge, collecting sound pressure time characteristics, calculating noise transfer matrix and real sound spectrum on the bridge, and finally fitting the noise attenuation curve. Compared with the traditional single-point measurement or simple noise recording method, the dynamic collection of the sound pressure under the bridge combined with the traffic signal cycle can truly reflect the influence of the traffic noise under the bridge on the road on the bridge, and the noise attenuation coefficient on the bridge can be accurately calculated through linear fitting. The method can be measured at the bridge construction stage, and can provide a scientific basis for the design of bridge pavement, the optimization of sound insulation facilities and the evaluation of noise environment, improve the accuracy and reliability of bridge noise control, and meet the engineering application requirements.

[0076] 2. The road surface noise attenuation performance test method realizes the scientization and standardization of the measurement point planning by setting multiple horizontal sections and local measurement points of the passage under the bridge, and calculating the grid spacing according to the vertical distance. The grid density is adjusted by using empirical coefficients and sound wave attenuation coefficients to adapt to different bridge structures and acoustic conditions, and the uniformity and measurement accuracy of the measurement point coverage are improved. The transverse and longitudinal sound pressure distribution of the passage under the bridge is calculated, and the number of measurement points on and under the bridge is optimized by using the branch and bound method, so that the optimization of the measurement point arrangement is realized. This method not only considers the distance and sound pressure attenuation law between the measurement points, but also accurately calculates the sound pressure contribution through height correlation weight and average index, thereby improving the measurement accuracy. At the same time, the sound pressure distribution formed by the measurement points is maximized by using the objective function, and a measurement point number penalty coefficient is introduced, so that the measurement accuracy and cost efficiency are considered, and the measurement point arrangement on and under the bridge is more scientific and reasonable. The strategy can ensure that the noise measurement results on the bridge accurately reflect the distribution characteristics of the traffic noise under the bridge, and improve the reliability of the noise attenuation evaluation.

[0077] 3、The road surface noise attenuation performance test method realizes dynamic noise characteristic optimization by combining the bridge traffic signal period and segmenting and weighting the sound pressure time characteristics according to the phases of the traffic light. The method effectively eliminates the background noise generated during the red light period without vehicles, and only retains the effective noise signals generated by vehicles during the green light period, thereby improving the representativeness and authenticity of the sound pressure characteristics under the bridge. At the same time, the sound pressure data during the green light period is reduced in dimension and compressed in characteristics through Hilbert transform and mean value calculation, only the key characteristic information is retained, so that the amount of noise time series data under the bridge is significantly reduced, thereby reducing the data storage requirement. The optimized time characteristics are more reliable in subsequent noise transfer matrix calculation and bridge noise correction, and adapt to the fluctuation of actual traffic conditions, improve the accuracy of bridge noise attenuation evaluation, and provide a scientific basis for bridge environmental noise management and noise control strategy.

[0078] 4、The road surface noise attenuation performance test method realizes quantitative estimation of the bridge under traffic noise on the bridge road by establishing a noise transfer matrix from the bridge under to the bridge. The sound pressure time series on the bridge and under the bridge is analyzed by STFT, and the regularization matrix H is solved to accurately reflect the noise transfer characteristics at different frequencies. The method avoids the rough estimation of the traditional simple distance attenuation model, makes the calculation of the noise contribution on the bridge more accurate, and quantifies the influence of the noise under the bridge on the sound field on the bridge. Combined with matrix regularization and conjugate transpose solution, the calculation stability and error are effectively controlled to ensure the scientificity and repeatability of the noise estimation results. The equivalent sound pressure level of each measuring point on the bridge is calculated, and the noise attenuation on the bridge is linearly fitted combined with the reference point to realize quantitative analysis of the noise distribution on the bridge. The equivalent sound pressure level is calculated using the real sound spectrum and time frame accumulation on the bridge, and linear fitting is performed combined with the distance between the measuring point and the reference point to output the noise attenuation coefficient on the bridge, which can accurately reflect the change rule of the noise on the bridge with distance.

[0079] 5、The road surface noise attenuation performance test method realizes quantitative analysis of the noise distribution on the bridge by calculating the equivalent sound pressure level of each measuring point on the bridge and linearly fitting the noise attenuation on the bridge combined with the reference point. The equivalent sound pressure level is calculated using the real sound spectrum and time frame accumulation on the bridge, and linear fitting is performed combined with the distance between the measuring point and the reference point to output the noise attenuation coefficient on the bridge, which can accurately reflect the change rule of the noise on the bridge with distance. The method improves the accuracy of noise attenuation evaluation, ensures the scientificity and reliability of the noise data on the bridge, and provides quantitative basis for bridge pavement design, noise isolation measure optimization and environmental noise management. BRIEF DESCRIPTION OF DRAWINGS

[0080] Figure 1 It is a flowchart of the method of the present application.

[0081] Figure 2 It is a schematic diagram of the system module of the present application.

[0082] Figure 3Fig. 1 is a schematic view of the bridge and the road below the bridge according to the present application. DETAILED DESCRIPTION

[0083] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0084] Embodiment one, refer to Figure 1 A road surface noise attenuation performance test method, comprising:

[0085] In the process of testing the road surface noise attenuation performance of a newly built bridge, the test needs to cover multiple traffic signal periods and different traffic flow states throughout the day, and the collection time length can reach 2 to 4 hours. The road below the bridge is continuously passed through during the test, and the vehicle passing density is about 800 to 1200 vehicles / hour / lane, and the intermittent traffic flow is generated by the alternation of red and green light phases, and the red light is about 30 to 60 seconds, and the green light is about 60 to 90 seconds. Under this dynamic traffic environment, the noise of the road below the bridge changes with the vehicle passing and the signal phase, and the influence of its interference on the noise on the bridge is inevitable. The traditional measurement method is difficult to distinguish the effective traffic noise from the background noise during the red light period, so in the long-period noise collection process, the interference of the noise below the bridge is objectively existing, which directly affects the authenticity of the sound pressure data on the bridge and the subsequent analysis accuracy.

[0086] Testing the road surface noise attenuation performance of the road on the bridge of a newly built bridge, wherein the road on the bridge intersects with the road below the bridge, and the road on the bridge is not passed through, and the road below the bridge is passed through;

[0087] The road below the bridge is a one-way road for vehicles below the bridge;

[0088] A plurality of horizontal sections are set in the passage below the bridge, and a plurality of measurement points are set on the horizontal sections;

[0089] Each horizontal section is equally spaced;

[0090] For any horizontal section, a plurality of local points are set on the horizontal section, the vertical distance from each local point to the road on the bridge is measured, and the average value of the vertical distance is calculated as ;

[0091] The grid spacing is set ;

[0092] , is an empirical coefficient, used to adjust the empirical parameter of the grid density, and the value range is 0.5 to 2.0, is the sound wave attenuation coefficient, is the degree of attenuation per unit distance, and is the normalized value of is normalized and updated to be the normalized value of 0.15;

[0093] The formula is used to determine the distance between measurement points that should be arranged on the horizontal section according to the speed of sound wave attenuation with distance. When the speed of sound wave attenuation with distance is fast, the grid spacing value is small, and the measurement points are biased to be dense. When the speed of sound wave attenuation with distance is slow, the grid spacing value is large, and the measurement points are biased to be sparse.

[0094] For long-span bridges, the number of measurement points is effectively reduced (30-50% of human arrangement work is reduced), while the spatial integrity of the measurement data is maintained, and the sound pressure distribution curve is smoothed. In the noise prediction model, the sampling grid generated according to the sound wave attenuation law can significantly improve the modeling accuracy. Compared with the traditional uniform point arrangement, the test error is reduced by about 12% on average, which is suitable for high-speed bridge areas with large wind and complex vehicle traffic.

[0095] is a natural exponential, for enhancing the sound wave attenuation coefficient;

[0096] a square with a length of is divided into grids;

[0097] The midpoint of the divided grid on the horizontal section is used as the measurement point.

[0098] A three-dimensional coordinate system is established, and a plurality of measurement points are arranged on the bridge along the driving direction;

[0099] The horizontal and vertical sound pressure point arrangement strategies are executed to calculate the sound pressure distribution in the horizontal and vertical directions of the bridge passage, respectively, for optimizing the number of measurement points on the bridge road and the bridge passage;

[0100] The direction perpendicular to the horizontal plane is referred to as the vertical direction, and the direction parallel to the horizontal plane is referred to as the horizontal direction;

[0101] The measurement points of the bridge passage and the measurement points of the bridge road are obtained;

[0102] is normalized and updated to be the normalized value of ;

[0103] The vertical sound pressure distribution , is calculated, where is a set reference distance, is the Euclidean distance between the measurement point and the measurement point , is the sound pressure attenuation index, for the height attenuation coefficient, represents the sound pressure decay with distance;

[0104] The formula takes into account the inverse square law of sound pressure decay with distance, the gradual absorption of energy by air when sound waves propagate upwards, and the mutual influence of sound pressures at the two measurement points and The inverse square law of sound pressure decay with distance is one of the core laws of acoustic propagation theory.

[0105] In actual bridge testing, the traditional method requires a large number of measurement points to invert the sound energy decay law, while the formula calculates the overall distribution with fewer samples. Taking a river-crossing bridge as an example, only 20 measurement points are needed to fit the complete sound field, with a fitting error of less than 5 dB. The longitudinal distribution formula is used to calculate the intensity distribution of noise transmitted from the lower traffic to the bridge, which is the basis for road noise reduction evaluation and sound barrier design on the bridge.

[0106] Obtain any measurement point on the outermost side of the bridge passage in the longitudinal direction.

[0107] Calculate the transverse sound pressure distribution , where is the height correlation weight, reflecting the proportion of the contribution of the measurement point to the measurement point , is the average index, controlling the sound pressure of the measurement point synthesized by the measurement point ;

[0108] In the bridge environment, sound waves not only propagate longitudinally, but also reflect and scatter laterally. The transverse propagation path is shorter, and the height absorption effect still exists. Therefore, the height weight term is introduced. The greater the correlation coefficient value between different points in the transverse direction under the bridge, the stronger the sound energy propagates in the transverse direction, and the more uniform the sound field.

[0109] The transverse sound pressure distribution is used to estimate the sound energy diffusion range and help determine the left and right layout density of the measurement points under the bridge. If the value is low, the measurement points on both sides of the bridge passage can be appropriately reduced.

[0110] are the sound pressures at the measurement points ,

[0111] Count the number of measurement points under the bridge passage and the number of measurement points on the bridge road.

[0112] Use the existing branch and bound method to solve the objective function for maximizing the sound pressure distribution formed by the measurement points. The measurement point number penalty coefficient is used to adjust the number of measurement point layout, and the objective function is used to balance the sound field distribution accuracy and the measurement point number cost, The measurement point can increase the resolution of the sound pressure distribution, The measurement point increase causes the measurement cost, the measurement point number penalty coefficient is introduced, if The measurement point is less if The measurement point is more if

[0113] The solution can use the existing technology integer programming, the solution process is to define the objective function and the constraint matrix, discretize the variable space, enumerate or linear programming solution: use the branch and bound method to select the current optimal solution node, prune the branch that makes the target value decrease, and obtain the optimal combination;

[0114] In the cross-river bridge experiment: originally planned to layout 30 (down) + 20 (up) = 50 measurement points, through the optimization model, finally layout 18 (down) + 12 (up) = 30 measurement points, the fitting error of the measured data is reduced from ± 5.6 dB to ± 3.2 dB, and the test time is reduced from 2 days to 1 day.

[0115] According to the measurement point of the optimized bridge under the road, the sound pressure time characteristics under the bridge are collected;

[0116] The period segmentation and weighting strategy is executed, the traffic signal period of the bridge under the road is obtained, and the phase divided by the traffic signal period is combined to optimize the sound pressure time characteristics under the bridge;

[0117] The sound pressure collected at the same time of each measurement point of the bridge under the road is obtained, the mean value is calculated, and the mean value is taken as the sound pressure of the bridge under the road at the same time;

[0118] The sound pressure of the bridge under the road at each time is obtained to form the sound pressure time characteristics under the bridge ;

[0119] The traffic signal period includes a red light phase and a green light phase;

[0120] The vehicle is stationary when the light is red, and accelerates through when the light is green, the noise energy difference is significant, and the noise data is matched with the real traffic state by weighting the time sequence according to the signal phase.

[0121] The sound pressure time characteristics under the bridge are cut into multiple segments according to the red light phase and the green light phase, specifically: , , indicates the th traffic signal period, , is the traffic signal period length, is the red light phase length, the green light phase length, the initial moment;

[0122] set the traffic state weight function ;

[0123] when located in the green light phase, when located in the red light phase, ;

[0124] segmented weighting of the sound pressure time characteristics under the bridge , updated to the segmented weighted value;

[0125] the existing technology Hilbert transform is performed on the sound pressure time characteristics of the green light phase to obtain ;

[0126] the average value is calculated , and the average value is used to update the sound pressure time characteristics under the bridge.

[0127] A reference position is set on the road above the bridge for fixing the noise source, and the sound pressure time characteristics above the bridge are collected according to the measured points on the road above the bridge after optimization;

[0128] In the bridge pavement noise attenuation test, the vehicles almost stop during the red light phase, and the noise under the bridge is mainly idle speed, wind noise or environmental background noise, which has little contribution to the noise on the road above the bridge, and belongs to invalid background noise. If the red light noise is directly included in the analysis, it will lead to underestimation of the traffic noise intensity of the sound pressure time characteristics under the bridge, affecting the accuracy of the calculation and spectrum correction of the noise transmission matrix above the bridge. At the same time, removing the red light noise can improve the representativeness and stability of the sound pressure characteristics under the bridge, so that the bridge noise attenuation evaluation is more in line with the actual traffic conditions. It can reduce the amount of redundant data, reduce the subsequent calculation pressure, improve the analysis efficiency and accuracy.

[0129] The measurement process of the noise under the bridge is transformed from average noise evaluation to traffic cycle response noise evaluation, which significantly reduces the data sampling and storage burden. Through the division of the red and green light cycle and the phase weighting, the continuously sampled sound pressure time sequence is divided and aggregated according to the traffic signal cycle, the data in the red light phase is given a low weight or directly removed, and only the effective sound field response under the green light phase is retained. Under the premise of not affecting the statistical characteristics of the sound pressure, the storage amount of the original time sequence can be reduced by about 40% to 60%.

[0130] The weighted sound pressure time sequence is segmented and smoothed in the time domain, which improves the stability of the Hilbert energy envelope and coherence analysis, and is more easily to identify periodic noise patterns and abnormal pulse events (such as heavy vehicle sudden acceleration or braking). The time consistency of the sound pressure time characteristics under the bridge is improved by about 20%;

[0131] The noise transfer matrix estimation strategy is performed to calculate the noise transfer matrix from the road under the bridge to the road on the bridge according to the sound pressure time characteristics, and to estimate the noise received by the road on the bridge under the bridge;

[0132] The sound pressure time characteristics under the bridge and the sound pressure time characteristics on the bridge are time-aligned and corrected, and the short-time Fourier transform (STFT) of the prior art is performed to obtain the complex spectrum matrix on the bridge and the complex spectrum matrix under the bridge , wherein, is the frequency point, the frequency point is the sampling point of the frequency in the discrete Fourier analysis, the unit is Hz, and the frequency points of the signals on the bridge and under the bridge must be kept consistent in the noise transfer matrix, because the propagation and attenuation of sound waves are frequency-dependent processes. Only by comparing the sound pressure amplitude and phase at the same frequency, a true and physically interpretable noise transfer matrix can be obtained;

[0133] If the comparison is not made at the same frequency point, the main noise of the vehicle under the bridge is concentrated in 50-500 Hz (engine, tire, air vibration), and the road noise on the bridge is concentrated in 200-800 Hz (tire noise, plate vibration), which will result in an incorrect comparison of high frequency to low frequency. Therefore, the same frequency point must be used, which is equivalent to performing energy matching on each independent frequency band channel.

[0134] When the frequency point is fixed, the complex spectrum matrix on the bridge and the complex spectrum matrix under the bridge are expressed as matrices according to the time frame, which correspond to and respectively;

[0135] The noise transfer matrix is set;

[0136] The objective function is defined, among all the matrices that can linearly map the sound pressure time characteristics under the bridge to the sound pressure time characteristics on the bridge, the best matrix is selected as the noise transfer matrix, for calculating the square root of the sum of the square of the matrix elements, is the regularization coefficient;

[0137] The derivative of the objective function with respect to the noise transfer matrix is obtained ;

[0138] The solution is , is the conjugate transpose, is the unit matrix; represents the complex gain from the measurement point on the road under the bridge to the measurement point on the road on the bridge at the frequency point ;

[0139] The noise received by the road on the bridge under the bridge is estimated .

[0140] The noise transfer matrix H establishes the mathematical mapping relationship between the sound field on the bridge and the sound field under the bridge. The noise at any time under the bridge is predicted to the response position on the bridge to realize the prediction of the noise on the bridge. Instead of the traditional empirical coefficient method, the prediction accuracy is greatly improved (the error is improved from ± 7 dB to ± 2 dB);

[0141] 1000 frames of sound spectrum (X) are collected at the measurement point under the bridge, and 1000 frames (Y) are collected on the bridge.

[0142] If the calculation result |H(f)| is 0.15, 0.08 and 0.03 at 100 Hz, 250 Hz and 500 Hz respectively, it means that the sound pressure transmission rate of the noise from under the bridge to the bridge is 15%, 8% and 3% respectively;

[0143] The noise correction strategy on the bridge is executed, the real sound pressure spectrum on the bridge is calculated according to the time characteristics of the sound pressure on the bridge and the noise under the bridge, and the frequency band of the real sound pressure spectrum is weighted;

[0144] The real sound pressure spectrum on the bridge is calculated , ;

[0145] From the measured , the noise component under the bridge mapped to the bridge by H is subtracted to obtain the real sound pressure spectrum on the bridge , that is, the net noise after removing the interference under the bridge, and then the frequency band is weighted to highlight the influence of low frequency (50-500 Hz).

[0146] The noise power spectrum measured on the bridge is 72 dB(A), the contribution of the noise under the bridge is estimated by the matrix to be 10 dB(A), and the real sound pressure on the bridge is 62 dB(A). If the weighting Ω(f)=[1.5,1.0,0.8] (low, medium and high frequency), the corrected noise dominant frequency band is concentrated in 125 Hz~250 Hz.

[0147] Based on the sensitivity of human ear hearing, three types of frequency bands are set, including low frequency, medium frequency and high frequency;

[0148] The frequency band weight function is set , when the frequency point is low frequency, , when the frequency point is medium frequency, , when the frequency point is high frequency, ;

[0149] The segmented weighting , and the real sound pressure spectrum on the bridge is updated to the result of the frequency band weighting.

[0150] The noise results of the bridge deck are only composed of the noise of the pavement material and the vehicle contact by deducting the noise of the passage under the bridge accurately. It is particularly important for the performance verification of noise barriers or damping layers. The results show that the noise reduction effect of low-noise asphalt is improved by about 4 dB(A) after removing the interference under the bridge, which is consistent with the laboratory roller test.

[0151] The noise attenuation fitting strategy on the bridge is performed to calculate the equivalent sound pressure level of each measurement point on the bridge, and the noise attenuation on the bridge is linearly fitted combined with the reference position.

[0152] For the measurement points on the bridge road , the real sound pressure spectrum on the bridge at the frequency point is obtained ;

[0153] The equivalent sound pressure level of the measurement point is calculated ;

[0154] The sound pressure of the reference position is obtained ;

[0155] The equivalent sound pressure level of the measurement point is obtained by squaring and averaging the sound pressure at each time point, used for degradation, used for modulo length, is the total number of time frames;

[0156] The Euclidean distance of the measurement point to the reference point is calculated ;

[0157] The noise attenuation on the bridge is linearly fitted combined with the reference position, specifically:

[0158] then , where is the noise attenuation coefficient on the bridge, is the fitting residual, used to measure the error of the fitting;

[0159] The residual sum of squares is minimized , and the noise attenuation coefficient on the bridge is output, and the equivalent sound pressure level is updated according to the output noise attenuation coefficient on the bridge

[0160] The reference point sound pressure = 72 dB(A), the measurement point distance is 5 m, 10 m, 15 m, and the corresponding sound pressure is 69, 65, 62 dB, and the linearly fitted noise attenuation coefficient on the bridge is ≈0.65 dB / m. It means that the noise is attenuated by about 0.65 dB per 1 m of propagation.

[0161] Noise attenuation coefficient is the core index of bridge acoustic design, which is used to evaluate the overall performance of the bridge pavement and sound insulation structure. Through linear fitting, the actual attenuation rate on the bridge is objectively obtained, and whether the pavement layer meets the design requirements (≥0.5 dB / m) is verified accordingly. In the bridge measurement, the method calculates the noise attenuation coefficient of 0.68 dB / m, which has an error of less than 6% compared with the laboratory test value, proving the reliability of the model.

[0162] Embodiment two, refer to Figure 2 A road surface noise attenuation performance test system, comprising:

[0163] A spatial measurement point planning module, which establishes a three-dimensional coordinate system of the bridge underpass and the bridge road, divides the horizontal cross-section grid of the bridge underpass, calculates the grid points and arranges the measurement points on the bridge road along the driving direction;

[0164] A sound pressure distribution optimization module, which calculates the transverse and longitudinal sound pressure distribution of the bridge underpass, counts the distance and height attenuation between the measurement points, and determines the number and position of the measurement points on the bridge and under the bridge by using the branch and bound method;

[0165] A time period weighting module, which collects the sound pressure time sequence under the bridge, divides each phase according to the traffic light cycle, applies segmented weighting and performs Hilbert transform to generate the weighted sound pressure time characteristics under the bridge;

[0166] A noise transfer matrix calculation module, which aligns the sound pressure time sequence on the bridge and under the bridge and performs STFT analysis to calculate the regularized noise transfer matrix and estimate the complex gain of the noise under the bridge on the bridge road;

[0167] A sound spectrum correction and attenuation fitting module, which calculates the real sound spectrum on the bridge according to the sound pressure on the bridge and the noise under the bridge, performs weighted processing on the low, medium and high frequency bands, and performs linear fitting with the reference point to obtain the noise attenuation coefficient on the bridge.

[0168] It should be noted that in this paper, relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. Moreover, the term "includes", "contains" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or equipment.

[0169] ​The above merely describes the preferred embodiments of the present application, and it should be pointed out that, for those skilled in the art, several improvements and refinements can be made without departing from the technical principles of the present application, and these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A method for testing road surface noise attenuation performance, characterized in that, include: The road surface noise reduction performance of the newly built bridge was tested. The road on the bridge intersected with the road under the bridge, and the road on the bridge was not open to traffic, while the road under the bridge was open to traffic. The road under the bridge is a one-way road for vehicles passing under the bridge. Multiple horizontal sections are set in the underpass, and multiple measurement points are set on the horizontal sections; Establish a three-dimensional coordinate system and set up multiple measurement points along the driving direction on the road on the bridge; Implement a transverse and longitudinal sound pressure distribution strategy, calculate the transverse and longitudinal sound pressure distribution of the underpass separately, and use this to optimize the number of measurement points on and under the bridge. The temporal characteristics of sound pressure under the bridge were collected based on the measurement points of the optimized road under the bridge. The execution cycle segmentation weighting strategy is used to obtain the traffic signal cycle of the road under the bridge, and the phase of the traffic signal cycle is combined to optimize the time characteristics of the sound pressure under the bridge. Reference locations were set on the road on the bridge to fix the noise source, and the time characteristics of the sound pressure on the bridge were collected based on the measurement points on the optimized road. A noise transfer matrix estimation strategy is implemented to calculate the noise transfer matrix from the road under the bridge to the road on the bridge based on the sound pressure time characteristics, and to estimate the noise received by the road on the bridge from the road under the bridge. Implement a noise correction strategy on the bridge, calculate the true sound pressure spectrum on the bridge based on the time characteristics of the sound pressure on the bridge and the noise under the bridge, and weight the frequency bands of the true sound pressure spectrum. A noise attenuation fitting strategy for the bridge is implemented, the equivalent sound pressure level at each measurement point on the road on the bridge is calculated, and the noise attenuation on the bridge is linearly fitted in combination with the reference position.

2. The method for testing road noise attenuation performance according to claim 1, characterized in that, The method of setting multiple horizontal cross-sections in the underpass and setting multiple measurement points on the horizontal cross-sections includes: Each horizontal section is equally spaced; For any horizontal cross-section, multiple local points are set on the horizontal cross-section, and the vertical distance from each local point to the road on the bridge is measured. The mean of the vertical distances is calculated and denoted as . ; Set grid spacing ; , These are empirical coefficients, empirical parameters used to adjust the mesh density. Let be the sound wave attenuation coefficient, and be the degree of attenuation per unit distance. Normalize and update to the normalized values; For natural index, Used to enhance the sound wave attenuation coefficient; The length is The square is denoted as the grid division; Divide the horizontal section into grids and use the midpoints of the grids on the horizontal section as measurement points.

3. The method for testing road noise attenuation performance according to claim 1, characterized in that, The implementation of the transverse and longitudinal sound pressure distribution strategy involves calculating the transverse and longitudinal sound pressure distribution of the underpass, respectively, to optimize the number of measurement points on and under the bridge, including: The direction perpendicular to the horizontal plane is denoted as the longitudinal direction, and the direction parallel to the horizontal plane is denoted as the transverse direction. Obtain measurement points for the underpass Measurement points on the road and bridge ; Will Normalize and update to the normalized values; Calculate longitudinal sound pressure distribution , ,in, The set reference distance, For measurement points and measurement points Euclidean distance, The sound pressure attenuation index is... For high attenuation coefficient, This indicates that the sound pressure decreases with distance; Obtain any measurement point on the outermost side of the underpass along its longitudinal direction. ; Calculate transverse sound pressure distribution , ,in, Highly correlated weights reflect the measurement points For measurement points The proportion of contribution For the average index, control measurement Point synthesis measurement points The sound pressure level; Measurement points The sound pressure at that location; Count the number of measurement points under the bridge Number of measurement points on the road on the bridge ; Solve the objective function using the branch and bound method. Used to maximize the sound pressure distribution at the measurement point This is a penalty coefficient for the number of measurement points, used to adjust the number of measurement points deployed.

4. The method for testing road noise attenuation performance according to claim 1, characterized in that, The execution cycle segmentation weighting strategy obtains the traffic signal cycle of the road under the bridge, and optimizes the time characteristics of the sound pressure under the bridge by combining the phase division of the traffic signal cycle, including: The sound pressure levels at each measurement point on the road under the bridge were collected simultaneously, the average value was calculated, and the average value was used as the sound pressure level on the road under the bridge at that moment. Obtain the temporal characteristics of the sound pressure under the bridge at each moment. ; The traffic signal cycle includes a red light phase and a green light phase; The sound pressure time characteristics under the bridge were divided into multiple segments according to the red light phase and the green light phase, specifically: , , indicating the first One traffic signal cycle, , Traffic signal cycle duration The duration of the red light phase. The duration of the green light phase. This is the initial time. Define the traffic state weighting function ; When in the green light phase When in the red light phase, ; Piecewise weighted summation of the time characteristics of sound pressure under the bridge Updated to the segmented weighted value; Hilbert transform was performed on the sound pressure time characteristics of the green light phase to obtain ; right Calculate the mean The time characteristics of the sound pressure under the bridge were updated using the mean.

5. The method for testing road noise attenuation performance according to claim 1, characterized in that, The noise transfer matrix estimation strategy involves calculating the noise transfer matrix from the road under the bridge to the road on the bridge based on the sound pressure time characteristics, and estimating the noise received by the road on the bridge from the road under the bridge, including: Time alignment correction was performed on the sound pressure time characteristics under the bridge and on the bridge, and STFT was simultaneously performed to obtain the complex spectrum matrix on the bridge. and the complex spectrum matrix under the bridge ,in, Frequency point; When frequency When fixed, the complex spectrum matrix on the bridge is... and the complex spectrum matrix under the bridge Represented as matrices according to time frames, respectively and ; Set the noise transfer matrix ; Define the objective function Among all matrices that can linearly map the time characteristics of sound pressure under the bridge to the time characteristics of sound pressure on the bridge, the optimal matrix is ​​selected as the noise transfer matrix. Used to calculate the square root of the sum of squares of matrix elements. The regularization coefficient is used. The derivative of the objective function with respect to the noise transfer matrix is ​​obtained. ; Solving , This is the conjugate transpose. It is the identity matrix; This indicates the frequency point of the measurement point from the underpass to the road on the bridge. The complex gain; Estimate the noise received by the road above the bridge from below. .

6. The method for testing road noise attenuation performance according to claim 1, characterized in that, The implementation of the bridge noise correction strategy involves calculating the true sound pressure spectrum on the bridge based on the time characteristics of the sound pressure on the bridge and the noise under the bridge, and weighting the frequency bands of the true sound pressure spectrum, including: Calculate the true sound pressure spectrum on the bridge , ; Based on the sensitivity of human hearing, three frequency bands are set, including low frequency, mid frequency and high frequency; Setting the frequency band weighting function When frequency point Low frequency, When frequency point For intermediate frequency, When frequency point High frequency, ; Segmented weighting The actual sound pressure spectrum on the bridge was updated to a frequency band weighted result.

7. The method for testing road noise attenuation performance according to claim 1, characterized in that, The implementation of the bridge noise attenuation fitting strategy involves calculating the equivalent sound pressure level at each measurement point on the road above the bridge, and then linearly fitting the bridge noise attenuation based on the reference location, including: Measurement points for the road on the bridge , obtain at frequency point The true sound pressure spectrum on the bridge below ; Calculate measurement points equivalent sound pressure level ; Obtain the sound pressure at the reference location ; The average of the squared sound pressure levels at each time point is used to obtain the measurement points. The equivalent sound pressure level, Used for downgrading Used for taking mold length This represents the total number of time frames. Calculate measurement points Euclidean distance to the reference point ; The noise attenuation on the bridge is linearly fitted using a reference position, specifically as follows: but ,in, This is the noise attenuation coefficient on the bridge. The fitting residual is used to measure the fitting error. Minimize the sum of squared residuals Noise attenuation coefficient on output bridge The equivalent sound pressure level is updated based on the output bridge noise attenuation coefficient. .

8. A road surface noise attenuation performance testing system, applied to the road surface noise attenuation performance testing method according to claims 1-7, characterized in that, include: The spatial measurement point planning module establishes a three-dimensional coordinate system between the underpass and the road on the bridge, divides the horizontal cross-section grid under the bridge, calculates the midpoint of the grid, and lays out measurement points along the driving direction on the road on the bridge. The sound pressure distribution optimization module calculates the lateral and longitudinal sound pressure distribution in the underpass, statistically analyzes the distance and height attenuation between measurement points, and uses the branch and bound method to determine the number and location of measurement points on and under the bridge. The time period weighting module collects the sound pressure time series under the bridge, divides each phase according to the traffic light cycle, applies segmented weighting and performs Hilbert transform to generate the weighted sound pressure time features under the bridge. The noise transfer matrix calculation module aligns the sound pressure time series on and under the bridge and performs STFT analysis to calculate the regularized noise transfer matrix and estimate the complex gain of the noise under the bridge on the road above the bridge. The sound spectrum correction and attenuation fitting module calculates the true sound spectrum of the bridge based on the sound pressure on the bridge and the noise under the bridge, performs weighted processing on the low, mid and high frequency bands, and performs linear fitting with reference points to obtain the noise attenuation coefficient on the bridge.

Images