Equivalent Vehicle Load Identification Method for Cable-Stayed Bridges Based on Cable Stress Time-History Curves

Abstract

The present invention relates to a method for identifying equivalent vehicle loads of a cable-stayed bridge based on the time history curve of cable stress, comprising the following steps: Step S1: Define the maximum time difference between the peaks of adjacent cables in combination with the monitoring cable position of the cable-stayed bridge and the driving speed, and judge the driving direction according to the sequence of the peaks of two adjacent stay cables; Step S2: Identify the peak values of adjacent cables caused by the same vehicle by combining the one-way interval peak identification method in the driving direction; Step S3: Calculate the vehicle speed through the distance between the peak positions and the corresponding time difference between the two outermost cables of the cable-stayed bridge; Step S4: Perform peak staggering elimination based on the vehicle speed in the sub-interval section; Step S5: Identify the equivalent vehicle load in combination with the finite element model; Step S6: According to the equivalent vehicle load identification result obtained in Step S5, directly count the distribution of the equivalent vehicle load to obtain its probability density curve. The present invention can effectively improve the accuracy of equivalent vehicle load identification.

Description

Technical Field

[0001] The present invention relates to a method for identifying equivalent vehicle loads of a cable-stayed bridge based on the time history curve of cable stress. Background Art

[0002] In recent years, with the development of the national economy, the traffic volume on the road network has been increasing, and overweight vehicles often pass through bridges, posing safety hazards. For long-span bridges such as cable-stayed bridges, the safety performance of the bridge system is closely related to the traffic load level. Therefore, the identification of the equivalent load level is an important part of the system safety assessment. However, the traffic conditions in different regions and different bridges vary, so it is particularly necessary to identify the driving load in combination with the measured data. Conventional vehicle load monitoring generally relies on the dynamic weighing system installed on the bridge, which can identify and record information such as vehicle axle weight, number of axles, and axle speed, and can be used for statistical analysis of vehicle loads and the establishment of load models.

[0003] Cable stress monitoring is an important part of the cable-stayed bridge health monitoring system, and the real-time cable stress is a dynamically changing process. As vehicles pass through the bridge deck, a peak value will appear successively in each cable. In the present invention, this wave peak value is described as the "cable stress peak value". The cable stress peak value can intuitively reflect the most unfavorable situation during the vehicle passage of the cable-stayed bridge and can also reflect the magnitude of the vehicle load value to a certain extent. In addition, as a key component in the cable-stayed bridge system, the cable stress is the key monitoring object. Without considering the bending stiffness of the cable, theoretically, the cable forces at each cross-section of the same cable should be equal. On a cable-stayed bridge without a dynamic weighing system, if the information of the driving load can be analyzed from the cable stress time history curve, the construction and maintenance costs of the monitoring equipment can be greatly saved, and it has good practical application prospects.

[0004] The vehicle loads on bridges have high randomness, and it is difficult to predict bridge loads through deterministic analysis methods. Considering the load as a random distribution (load model) is a widely used feasible method. The load model can obtain the random distribution of load extreme values by statistically analyzing the maximum daily vehicle weight load. The most commonly used is the extreme value type I distribution; it can also directly statistically analyze the vehicle weight, vehicle distance, and obtain the random distribution of the corresponding load after probability statistics.

[0005] Currently, in current practice, the identification of vehicle loads on long-span bridges mainly relies on vehicle dynamic weighing systems and vehicle identification methods based on computer vision. The former is expensive, and the reliability and stability during long-term operation depend on the maintenance level. The weighing accuracy is greatly affected by the road surface height and flatness. At the same time, it is difficult to accurately obtain the real-time dynamic response of components during vehicle passage; the latter can locate the real-time position of the vehicle, but it is difficult to accurately identify the vehicle weight and component response. Summary of the Invention

[0006] In view of this, the purpose of the present invention is to provide a method for identifying equivalent vehicle loads of cable-stayed bridges based on the time history curve of cable stress, so as to improve the accuracy of identifying equivalent vehicle loads.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] A method for identifying equivalent vehicle loads of cable-stayed bridges based on the time history curve of cable stress includes the following steps:

[0009] Step S1: Define the maximum time difference between the peaks of adjacent cables in combination with the cable positions for monitoring in the cable-stayed bridge and the driving speed, and judge the driving direction according to the order of the peaks of two adjacent cables.

[0010] Step S2: Identify the peak values of adjacent cables caused by the same vehicle by combining the one-way interval peak recognition method in the driving direction.

[0011] Step S3: Calculate the vehicle speed through the distance between the peak positions and the corresponding time difference between the two outermost cables of the cable-stayed bridge.

[0012] Step S4: Perform peak staggering elimination based on the vehicle speed in the sub-interval section.

[0013] Step S5: Identify the equivalent vehicle load in combination with the finite element model.

[0014] Step S6: According to the equivalent vehicle load identification result obtained in Step S5, directly count the distribution of the equivalent vehicle load to obtain its probability density curve.

[0015] Further, the specific content of Step S1 is as follows:

[0016] Step S11: During the operation of the cable-stayed bridge, monitor the stress of the cables in the preset interval to obtain the stress time history curve of each monitored cable. At the same time, preset the bridge head and bridge tail of the cable-stayed bridge. Suppose a total of n cables are monitored, and the cables are numbered C1, C2,..., Cn in the order from the bridge head to the bridge tail of the cable-stayed bridge.

[0017] Step S12: Identify the local maximum value of the stress within the preset time interval as the peak value, limit the minimum time interval between two peak values to be Δt, and the minimum peak value to be and ignore the peaks less than from being identified anymore; finally, identify the peak values of each wave of the C1 cable and the moments when they appear, and use the one-dimensional matrix [P 11 , P 12 , …, P 1t to represent the identified peak values; [T 11 , T 12 , …, T 1t] represents the time when the peak appears, t represents the number of peaks, that is, the number of equivalent vehicle loads;

[0018] Step S13: Obtain the maximum time difference ΔT of the peaks of adjacent cables through finite element analysis i→i+1 ;

[0019] Step S14: Determine the direction of vehicle travel by determining the time sequence of the appearance of two adjacent cable wave peaks.

[0020] Furthermore, the step S13 is specifically as follows:

[0021] First, the driving process is simulated in the finite element model of the cable-stayed bridge, and the position of the vehicle when the peak appears on each cable is recorded. For n cables, the position of the vehicle on the cable-stayed bridge when the corresponding stress peak appears is expressed as D1, D2, ..., D n ;

[0022] Secondly, set a speed lower limit ν min , and further obtain the maximum time difference ΔT of each adjacent cable i→i+1 =(D i+1 -D i ) / ν min .

[0023] Furthermore, the step S2 is specifically as follows: if the driving direction is forward, then in the time period [T 21 ,T 21 +ΔT 2→3 ] Directedly search for the peak of C3 and locate the time when the peak appears as T 31 On the contrary, if the driving direction in is negative, then in the time period [T 21 -ΔT 2→3 ,T 21 ] Directedly search for the peak of C3 and locate the time when the peak appears as T 31 ;

[0024] Repeat t times until t peaks of n cables are identified, using the matrix express.

[0025] Furthermore, the step S4 is specifically as follows: the cable-stayed bridge is divided into n-1 sub-intervals from cable C1 to cable Cn, the driving speed of each sub-interval is calculated by the ratio of distance to time difference, and secondly, the peak recognition data of adjacent intervals with opposite speed directions are eliminated.

[0026] Furthermore, the step S5 is specifically as follows:

[0027] Step 51: Combine the peak data, divide the vehicle weight load into N equal points in the interval [0 - M]t, where M is the maximum value of the possible equivalent vehicle load. Calculate the peak values of each cable under the condition of each driving load through finite element method respectively, and obtain the driving load - cable stress peak curve of each cable.

[0028] Step 52: Combine the driving load - cable stress peak curves of each cable, substitute the stress peak points of each cable identified in Steps 1 - 4 into the corresponding curves, and calculate the equivalent vehicle load values corresponding to the stress peak points of each cable through interpolation method;

[0029] Step 53: For the first peak group [P 11 , P 21 , …, P n1 , take the average value of the equivalent vehicle loads of each cable calculated as the equivalent vehicle load corresponding to this peak group. Repeat the above process to identify all equivalent vehicle load values.

[0030] The present invention has the following beneficial effects compared with the prior art:

[0031] 1. The effective identification of the equivalent vehicle load of the present invention can be used for the construction of the vehicle load model, without the need to rely on the dynamic weighing system and the camera system, reducing the monitoring cost;

[0032] 2. The present invention can directly obtain the response information of each cable and obtain the random distribution of the equivalent vehicle load, without the need to conduct complex vehicle - induced effect analysis, which is beneficial to the subsequent safety performance analysis of the cable - stayed bridge. Description of the Drawings

[0033] Figure 1 is the schematic flow chart of the method of the present invention. Detailed Embodiment

[0034] The following further describes the present invention in conjunction with the drawings and embodiments.

[0035] Please refer to Figure 1 , the present invention provides a method for identifying the equivalent vehicle load of a cable - stayed bridge based on the cable stress time - history curve, including the following steps:

[0036] Step S1: Define the maximum time difference between the peaks of adjacent cables in combination with the cable position for monitoring of the cable - stayed bridge and the driving speed, and judge the driving direction according to the sequence of the peaks of two adjacent cables.

[0037] In this embodiment, the specific content of Step S1 is as follows:

[0038] Step S11: During the operation of the cable-stayed bridge, the stress of the cable-stayed cables in the preset interval is monitored to obtain the stress time history curve of each monitoring cable. Meanwhile, the bridgehead and the bridge tail of the cable-stayed bridge are pre-set. It is assumed that a total of n cable-stayed cables are monitored, and the cable-stayed cables are numbered in the order from the bridgehead to the bridge tail of the cable-stayed bridge, C1, C2, ..., Cn;

[0039] Step S12: Peak identification of cable C1. It is relatively simple to identify the peak of the time history data of a single cable only. The peak value can be determined by identifying the local maximum value of stress within a certain time interval. The minimum time interval between two peaks is Δt, and the minimum peak value is Ignore less than The peaks of the C1 cable are no longer identified. Finally, the peak values of the C1 cable and their occurrence times are identified and the one-dimensional matrix [P 11 ,P 12 ,…,P 1t ] represents the identified peak value; [T 11 ,T 12 ,…,T 1t ] represents the time when the peak appears, t represents the number of peaks, that is, the number of equivalent vehicle loads;

[0040] Step S13: Obtain the maximum time difference ΔT of the peaks of adjacent cables through finite element analysis i→i+1 First, the driving process is simulated in the finite element model of the cable-stayed bridge, and the position of the vehicle when the peak appears on each cable is recorded. For n cables, the position of the vehicle on the cable-stayed bridge when the corresponding stress peak appears is expressed as D1, D2, ..., D n Next, set a speed lower limit ν min (e.g. 20km / h), and further obtain the maximum time difference ΔT of each adjacent cable i→i+1 =(D i+1 -D i ) / ν min ;

[0041] Step S14: When the same vehicle passes through the bridge in one direction, the stress curves of the two adjacent cables (such as C1 and C2) will have peaks one after another, so the direction of the vehicle can be determined by judging the time sequence of the peaks of the two cables C1 and C2. 11 The time period before and after [T 11 -ΔT 1→2 ,T 11 +ΔT 1→2 ] to find the peak of cable C2 and locate the time when the corresponding peak appears T 21 If the stress peak of C1 appears earlier than that of C2, the driving direction is from the bridge head to the bridge tail, which is defined as the positive direction; otherwise, the driving direction is from the bridge tail to the bridge head, which is defined as the negative direction.

[0042] Step S2: Combine the unidirectional interval wave peak recognition method in the driving direction to identify adjacent cable wave peaks caused by the same vehicle;

[0043] In this embodiment, specifically: If the driving direction is positive, then search for the wave peak of Cable C3 directionally in the time period [T 21 , T 21 +ΔT 2→3 , and locate the moment when the wave peak appears as T 31 . Conversely, if the driving direction is negative, then search for the wave peak of Cable C3 directionally in the time period [T 21 -ΔT 2→3 , T 21 , and locate the moment when the wave peak appears as T 31 .

[0044] Repeat t times until t wave peaks of n cables are identified, which are represented by the matrix .

[0045] Step S3: Calculate the vehicle speed by the distance between the wave peak positions and the corresponding time difference between the two outermost cables of the cable-stayed bridge. For the t-th wave peak corresponding to the equivalent vehicle load, its vehicle speed ν t =(D n -D1) / (T nt -T 1t ).

[0046] Step S4: Eliminate the wave peak data with opposite speed directions based on the vehicle speeds in the sub-interval segments; First, divide the cable-stayed bridge into n-1 sub-interval segments from Cable C1 to Cable Cn. Calculate the driving speeds of each sub-interval segment by the ratio of the distance to the time difference. Secondly, eliminate the wave peak recognition data with opposite speed directions in adjacent interval segments.

[0047] Step S5: Combine the finite element model to identify the equivalent vehicle load;

[0048] In this embodiment, preferably, Step S5 is specifically:

[0049] Step 51: Combine the wave peak data, divide the vehicle weight load into N equal points in the interval [0-M]t, where M is the maximum value of the possible equivalent vehicle load. Calculate the cable peak values under the traffic conditions of each driving load through finite element respectively, and obtain the driving load - cable stress peak value curve of each cable.

[0050] Step 52: Combine the driving load - cable stress peak value curves of each cable, substitute the cable stress peak value points identified in Steps 1-4 into the corresponding curves, and calculate the equivalent vehicle load values corresponding to the cable stress peak value points by interpolation;

[0051] Step 53: For the first wave peak group [P 11,P 21 ,…,P n1 , the mean value of the equivalent vehicle loads calculated for each cable is taken as the equivalent vehicle load corresponding to this peak group. Repeating the above process can identify all the equivalent vehicle load values.

[0052] Step S6: According to the identification result of the equivalent vehicle load obtained in Step S5, directly conduct statistics on the distribution of the equivalent vehicle load to obtain its probability density curve.

[0053] Embodiment 1:

[0054] The method for identifying the equivalent vehicle load of a cable-stayed bridge in this embodiment in combination with the cable stress time history curve has the following usage process:

[0055] First, number the stay cables in the order from the bridge head to the bridge tail of the cable-stayed bridge, which are C1, C2, …, Cn respectively; secondly, group and match the peaks of each stay cable according to the "peak group", including identifying the peaks of the C1 cable, obtaining the maximum time difference ΔT between the peaks of adjacent cables through finite element analysis i→i+1 , steps such as driving direction discrimination, one-way interval peak identification, vehicle speed identification, and peak staggering elimination based on the vehicle speed in the sub-interval section.

[0056] Subsequently, identify the equivalent vehicle load for each peak group in combination with the finite element model; then further identify the equivalent vehicle load that causes each peak group, including sub-steps such as obtaining the vehicle load - cable stress peak curve through finite element analysis, calculating the equivalent vehicle load value corresponding to the peak point of each cable stress, and taking the mean value of the equivalent vehicle loads calculated for each cable as the equivalent vehicle load corresponding to this peak group;

[0057] Finally, directly conduct statistics on the distribution of the equivalent vehicle load to obtain the random distribution curve of the equivalent vehicle load.

[0058] The above are only the preferred embodiments of the present invention. All equivalent changes and modifications made according to the scope of the patent application of the present invention shall fall within the scope covered by the present invention.

Claims

1. A method for identifying equivalent vehicle loads of a cable-stayed bridge based on the cable stress time history curve, characterized in that, It includes the following steps: Step S1: Define the maximum time difference for the peaks to appear between adjacent cables by combining the cable positions for cable-stayed bridge monitoring and the driving speed, and determine the driving direction based on the order of the peaks appearing in two adjacent cables; Step S2: Combine the one-way interval peak recognition method for the driving direction to identify the peak values of adjacent cables caused by the same vehicle; Step S3: Calculate the vehicle speed through the distance between the peak positions and the corresponding time difference between the two outermost cables of the cable-stayed bridge; Step S4: Perform peak staggering elimination based on the vehicle speeds in sub-intervals; Step S5: Combine the finite element model to identify the equivalent vehicle loads; Step S6: According to the equivalent vehicle load identification result obtained in Step S5, directly count the distribution of the equivalent vehicle loads to obtain its probability density curve; The specific content of Step S1 is as follows: Step S11: During the operation of the cable-stayed bridge, monitor the stress of the cables in the preset interval to obtain the stress time history curves of each monitored cable. At the same time, preset the bridge head and the bridge tail of the cable-stayed bridge. Suppose a total of n cables are monitored, and the cables are numbered C1, C2, …, Cn in the order from the bridge head to the bridge tail of the cable-stayed bridge; Step S12: By identifying the local maximum stress within a preset time interval as the peak value, the minimum time interval between two peaks is defined as Δt, and the minimum peak value is and ignore those less than of the peak and no longer identify them; finally, identify the peak values of each cable C1 and the moments when they occur, and use a one-dimensional matrix [P 11 , P 12 , …, P 1t to represent the identified peak values; [T 11 , T 12 , …, T 1t represents the moments when the peaks occur, t represents the number of peaks, that is, the number of equivalent vehicle loads; Step S13: Obtain the maximum time difference ΔT between the wave peaks of adjacent cables through finite element analysis i→i+1 ; Step S14: Determine the vehicle passing direction by judging the time sequence of the peaks appearing in two adjacent cables; The specific steps of step S2 are as follows: If the driving direction is positive, then within the time period [T 21 , T 21 +ΔT 2→3 , search for the peak of the C3 cable wave in a directed manner, and locate the moment when the peak appears as T 31 ; conversely, if the driving direction in 21 -ΔT 2→3 , T 21 , search for the peak of the C3 cable wave in a directed manner, and locate the moment when the peak appears as T 31 ; Repeat t times until t wave peaks of n cables are identified, represented by the matrix ; The specific content of Step S4 is as follows: Divide the cable-stayed bridge into n - 1 sub-intervals from Cable C1 to Cable Cn. Calculate the driving speed of each sub-interval through the ratio of the distance to the time difference. Secondly, eliminate the peak recognition data with opposite speed directions in adjacent intervals; The specific content of Step S5 is as follows: Step 51: Combine the peak data and divide the vehicle weight load into N equal points in the interval [0 - M]t, where M is the maximum value of the possible equivalent vehicle load; Respectively calculate the peak values of each cable under the driving conditions of each driving load through finite element calculation to obtain the driving load - cable stress peak curve of each cable; Step 52: Combine the driving load - cable stress peak curves of each cable, substitute the peak stress points of each cable identified in Steps 1 - 4 into the corresponding curves, and calculate the equivalent vehicle load values corresponding to the peak stress points of each cable through interpolation; Step 53: For the first peak group [P 11 , P 21 , …, P n1 , take the mean value of the calculated equivalent vehicle loads of each cable as the equivalent vehicle load corresponding to this peak group. Repeat the above process to identify all equivalent vehicle load values.

2. The method for identifying the equivalent vehicle load of a cable-stayed bridge based on the time history curve of cable stress according to claim 1, wherein The specific content of Step S13 is as follows: First, simulate the driving process in the finite element model of the cable-stayed bridge, and record the positions of the vehicles when the wave crests appear on each cable. For n stay cables, the positions of the vehicles on the cable-stayed bridge when the corresponding stresses reach the wave crests are denoted as D1, D2, …, D n ; Secondly, set a lower speed limit ν min , and further obtain the maximum time difference ΔT between adjacent cables i→i+1 =(D i+1 -D i ) / ν min .

Images