Temperature gradient fatigue load spectrum and construction method for flat wide steel box girder bridges

Abstract

A temperature gradient fatigue load spectrum and construction method for flat wide steel box girder bridges are presented. The temperature gradient fatigue load spectrum consists of vertical and transverse temperature gradient fatigue load spectra. Long-term temperature field monitoring data for flat wide steel box girder bridges is used to analyze the temperature field distribution characteristics. A mean shift cluster analysis method is applied to temperature sample points at typical locations to determine the number of clusters of daily temperature gradient loads for flat wide steel box girder bridges. Combined with probability and statistical analysis theory, vertical and transverse temperature gradient fatigue load spectra for flat wide steel box girder bridges with design service lives of 100, 150, and 200 years are proposed. This temperature gradient fatigue load spectrum can provide a technical basis for fatigue damage analysis of flat steel box girder bridges.

Description

Technical Field

[0001] The present invention belongs to the technical field of bridge engineering, and in particular relates to a temperature gradient fatigue load spectrum and a construction method for a flat wide steel box girder bridge. Background Art

[0002] Flat wide steel box girders are a typical closed section type. Under solar radiation, not only vertical temperature gradients will be generated in the cross section, but also significant transverse temperature gradients will exist. Flat wide steel box girders will produce longitudinal temperature fatigue stress under the action of vertical nonlinear temperature gradients. Due to the frame effect of the closed section, the transverse temperature gradient will produce transverse temperature fatigue stress. Previous studies have shown that the coupled cyclic effect of temperature loads and vehicle loads will significantly increase the fatigue cumulative damage of steel box girder bridge details. The fatigue damage caused by temperature and vehicle loads alone cannot be simply superimposed. Since temperature effects include the cyclic effects of various types of temperature differences, such as daily temperature differences and seasonal temperature differences, it is necessary to use the temperature field monitoring data of flat steel box girder bridges to establish their temperature fatigue load spectrum, calculate the temperature fatigue stress history, and couple it with the vehicle fatigue stress sequence to conduct fatigue damage analysis. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a temperature gradient fatigue load spectrum and a construction method for a flat wide steel box girder bridge.

[0004] The technical solution adopted to solve the above technical problems is: a temperature gradient fatigue load spectrum for flat wide steel box girder bridges, which includes a vertical temperature gradient fatigue load spectrum and a transverse temperature gradient fatigue load spectrum; the vertical temperature gradient fatigue load spectrum is composed of six vertical temperature sub-gradient fatigue load models and the probability of occurrence of the sub-gradients within the design service life, specifically:

[0005]

[0006] In formula (1), T Gi (y) is the temperature of the ith vertical temperature sub-gradient at position y, h is the beam height of the flat steel box girder bridge, the unit is m, Gi is the ith vertical temperature sub-gradient, T Gi,1 、T Gi,2 、T Gi,3 They are the top plate and the distance h between the nozzle and the bottom plate in the i-th vertical temperature sub-gradient. f The temperature at the bottom plate is represented by T G1,1 、T G1,2 、T G1,3 They are respectively the top plate in the first vertical temperature sub-gradient, the distance between the nozzle and the bottom plate h f The temperature at the bottom plate is represented by T G1,1 The value range is [4.8,7.8], the unit is ℃, TGi,j is the temperature at the jth typical height in the ith vertical sub-temperature gradient, N d is the design service life, P(T G1 (y))、P(T G2 (y))、P(T G3 (y))、P(T G4 (y))、P(T G5 (y))、P(T G6 (y)) are the occurrence probabilities of the six vertical temperature sub-gradients within the design service life, T G1,1 、T G2,1 、T G3,1 、T G4,1 、T G5,1 、T G6,1 These are the representative temperature values ​​of the top plate for the six vertical temperature sub-gradients.

[0007] The transverse temperature gradient fatigue load spectrum is composed of six transverse temperature sub-gradients and the probability of occurrence of the sub-gradients within the design service life, specifically:

[0008]

[0009] In formula (2), T Hq (x) is the temperature of the qth transverse temperature sub-gradient at the transverse position x, L is the width of the flat steel box girder, the unit is m, T Hq,w is the temperature representative value at the wth typical transverse position in the qth transverse temperature sub-gradient, T Hq,1 、T Hq,2 、T Hq,3 、T Hq,4 are the representative temperature values ​​at the edge of the sunny side, the position close to the 1 / 4L position of the sunny side, the position close to the 1 / 2L position of the sunny side, and the edge of the shady side in the qth transverse temperature sub-gradient, T H1,1 、T H1,2 、T H1,3 、T H1,4 are the representative temperature values ​​at the edge of the sun side, the position close to the 1 / 4L position of the sun side, the position close to the 1 / 2L position of the sun side, and the edge of the sun side in the first transverse temperature sub-gradient, T H1,1 The value range is [12.0,14.0], the unit is ℃, T H1,2 The value range is [12.4,14.4], and the unit is ℃. H1 (y))、P(T H2 (y))、P(T H3 (y))、P(T H4 (y))、P(T H5 (y))、P(T H6(y)) are the occurrence probabilities of the six transverse temperature sub-gradients within the design service life, T H1,1 、T H2,1 、T H3,1 、T H4,1 、T H5,1 、T H6,1 They are the representative values ​​of the temperature at the edge of the sunny side in the six lateral temperature sub-gradients.

[0010] The present invention also provides a method for constructing vertical and transverse temperature gradient fatigue load spectra of a flat wide steel box girder bridge, comprising the following steps:

[0011] Step 1. Arrange temperature measuring points on the top plate, bottom plate, and diaphragm of the flat wide steel box girder bridge. The number of measuring points is c. The temperature values ​​of each measuring point are collected at regular intervals to obtain the measured temperature gradient time history curve. The measured temperature curve of the flat steel box girder bridge is extended to the design service life N through the long-short memory recurrent neural network. d The temperature history at a typical location is selected, and the maximum temperature at each measuring point on each day is formed into a set of daily temperature extreme value vectors to obtain the N d ×365 groups of daily temperature extreme value vectors, recorded as the design service life daily temperature vector set M;

[0012] Step 2. Randomly select a set of daily temperature extreme value vectors from the design service life daily temperature vector set M as the cluster center C, use the mean shift method to establish a cluster D based on the cluster center C, and expand the cluster D according to the cluster expansion method;

[0013] The cluster expansion method is:

[0014] 1) Calculate the Euclidean distance between other daily temperature extreme value vectors in the daily temperature vector set M of the design service life and the cluster center C, and select the Euclidean distance d e The daily temperature extreme value vectors with a mean shift radius of R are classified into cluster D, and the frequency of each selected group of daily temperature extreme value vectors being classified into the current cluster D is recorded as an increase of 1;

[0015] 2) Calculate the vector differences between the daily temperature extreme vectors and the cluster center in cluster D except the cluster center, add the vector differences and calculate the average to get vector S. Then the cluster center drifts to the next position along the direction of vector S, and the drift distance is the modulus of vector S.

[0016] 3) After the cluster center arrives at the new location, the Euclidean distance d between the design service life daily temperature vector set M and the cluster center is again calculated. e The daily temperature extreme value vectors with ≤R are classified into cluster D, and the frequency of each group of selected daily temperature extreme value vectors being classified into cluster D is recorded as an increase of 1;

[0017] 4) Repeat steps 2) and 3) until no new daily temperature extreme value vectors are added to cluster D, completing the expansion of cluster D. Count the total frequency of each group of daily temperature extreme value vectors in cluster D being classified into cluster D, and record it as the cluster frequency.

[0018] Step 3. Randomly select a set of daily temperature extreme value vectors from the daily temperature vector set M of the design service life except the cluster cluster as the virtual cluster center C / And establish a virtual cluster D with the virtual cluster center / , expand the virtual cluster D according to the cluster expansion method in step 2 / ;

[0019] Step 4. Calculate virtual clusters D / The Euclidean distance between the virtual cluster center and the cluster center in cluster D, if the Euclidean distance d e >Set threshold d th , then the virtual cluster D / Add as a new cluster; if the Euclidean distance d e <Set threshold d th , then the virtual cluster D / Merge the cluster with the shortest Euclidean distance;

[0020] Step 5. According to steps 3 and 4, traverse all the daily temperature extreme value vectors in the daily temperature vector set M of the design service life, compare the clustering frequency of each group of daily temperature extreme value vectors in each cluster, and classify each group of daily temperature extreme value vectors into the cluster with the highest corresponding clustering frequency. The obtained cluster is the sub-gradient of the temperature gradient fatigue load spectrum, and the cluster center of the cluster is the temperature representative value of the sub-gradient. The ratio of the clustering frequency of each group of daily temperature extreme value vectors in the same cluster to the sum of the clustering frequencies of all vectors is the probability of occurrence of the sub-gradient in the design service life.

[0021] Preferably, the European distance d in step 1) of step 2 e According to the following formula:

[0022]

[0023] Where, d e is the Euclidean distance, T Nd (y k ,t C ) is t C Date k The temperature maximum at the location, y k is the measuring point position, t C is the date of the cluster center, and t is the date of the daily temperature extreme value vector other than the cluster center.

[0024] Preferably, the vector S in step 2) of step 2 is:

[0025]

[0026] Where z is the number of groups of daily temperature extreme value vectors in the cluster, t l The date of the extreme temperature vector of the lth group.

[0027] Preferably, the drift radius R is in the range of [1.2, 2.8]; the threshold d th The value range is [1.0,4.5].

[0028] Preferably, the arrangement method of the measuring points in step 1 is: arrange one measuring point on the upper surface of the bottom plate of the flat wide steel box girder bridge, and take the measuring point as the coordinate origin, arrange measuring points on the diaphragm and the top plate along the vertical height direction, and arrange a total of c=8 measuring points. The vertical distance between the measuring point position and the lower surface of the bottom plate of the flat wide steel box girder bridge can be 0.00m, h-1.52m, h-0.82m, h-0.3m, h-0.2m, h-0.1m, h-0.05m, h.

[0029] Preferably, the arrangement method of the measuring points in step 1 is: arrange one measuring point at the edge of the positive side of the top plate, and use the measuring point as the coordinate origin, and arrange a total of c=7 measuring points along the width direction of the top plate, and the positions of the measuring points are expressed as horizontal distances from the coordinate origin as 0.0m, L-27.4m, L-20.3m, L-16.45m, L-12.6m, L-5.5m, L.

[0030] Preferably, the temperature values ​​of each measuring point are collected at intervals of 60 to 1200 seconds in step 1.

[0031] The beneficial effects of the present invention are as follows:

[0032] 1. The temperature gradient fatigue load spectrum for a flat wide steel box girder bridge constructed in this invention consists of a temperature fatigue load sub-gradient model and the occurrence probability of the sub-gradient, and can be used to calculate the temperature stress history within a 200-year design service life.

[0033] 2. The method for constructing vertical and transverse temperature gradient fatigue load spectra for flat steel box girders proposed in this invention uses a mean-shift algorithm to self-identify the number of temperature sub-gradient clusters, overcoming the problem of relying on experience to divide the number of clusters in previous cluster analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is a cross-sectional view of the monitored flat, wide steel box girder cable-stayed bridge.

[0035] Figure 2 It is the vertical temperature sub-gradient of the flat wide steel box girder bridge.

[0036] Figure 3 This is the probability map of occurrence of each vertical temperature sub-gradient.

[0037] Figure 4 This is a flow chart of the method for constructing a vertical temperature gradient fatigue load spectrum for a flat wide steel box girder bridge in Example 1.

[0038] Figure 5 Schematic diagram of the temperature measurement point layout of the monitored flat wide steel box girder cable-stayed bridge.

[0039] Figure 6 This is the vertical temperature history curve of a flat wide steel box girder cable-stayed bridge.

[0040] Figure 7 Clustering results of vertical temperature gradient data for flat wide steel box girder cable-stayed bridge.

[0041] Figure 8 is the transverse temperature sub-gradient of a flat wide steel box girder bridge.

[0042] Figure 9 This is the probability map of occurrence of each lateral temperature sub-gradient.

[0043] Figure 10 Clustering results of transverse temperature sub-gradients of a flat wide steel box girder cable-stayed bridge. DETAILED DESCRIPTION

[0044] The present invention will be further described in detail below with reference to the accompanying drawings and examples, but the present invention is not limited to the following embodiments.

[0045] Example 1

[0046] This embodiment takes a flat steel box girder cable-stayed bridge with a beam height h of 3.5m in a coastal area as an example. Figure 1 The temperature gradient fatigue load spectrum of the bridge includes the vertical temperature gradient fatigue load spectrum and the transverse temperature gradient fatigue load spectrum. The vertical temperature gradient fatigue load spectrum consists of 6 vertical temperature sub-gradients and the probability of occurrence of the sub-gradients within the design service life, specifically:

[0047]

[0048] In formula (1), T Gi (y) is the temperature of the ith vertical temperature sub-gradient at position y, h is the beam height of the flat steel box girder bridge, the unit is m, Gi is the ith vertical temperature sub-gradient, T Gi,1 、T Gi,2 、T Gi,3 They are the top plate and the distance h between the nozzle and the bottom plate in the i-th vertical temperature sub-gradient. f The temperature at the bottom plate is represented by T G1,1、T G1,2 、T G1,3 They are respectively the top plate in the first vertical temperature sub-gradient, the distance between the nozzle and the bottom plate h f The temperature at the bottom plate is represented by T G1,1 The value range is [4.8,7.8], the unit is ℃, T Gi,j is the temperature at the jth typical height in the ith vertical sub-temperature gradient, N d is the design service life, P(T G1 (y))、P(T G2 (y))、P(T G3 (y))、P(T G4 (y))、P(T G5 (y))、P(T G6 (y)) are the occurrence probabilities of the six vertical temperature sub-gradients within the design service life, T G1,1 、T G2,1 、T G3,1 、T G4,1 、T G5,1 、T G6,1 These are the representative temperature values ​​of the top plate for the six vertical temperature sub-gradients.

[0049] The temperature value T at the top plate in the first vertical temperature sub-gradient of the flat steel box girder cable-stayed bridge is G1,1 =6.3℃, where the temperature representative values ​​at typical positions of each vertical temperature sub-gradient of the temperature gradient fatigue load spectrum of the flat steel box girder bridge with a design service life of 100 years, 150 years, and 200 years are shown in Table 1. The linear temperature sub-gradient composed of the typical measuring points is shown in Figure 2 As shown, 150 years and 200 years are the long-life design service life, and the vertical temperature sub-gradient T G1 (y)~T G6 The frequencies of (y) are 24.5%, 20.9%, 14.5%, 18.2%, 17.3%, and 4.6%, see Figure 3 .

[0050] Table 1 Representative values ​​of vertical temperature gradient fatigue load spectrum of flat wide steel box girder

[0051]

[0052] exist Figure 4 The method for constructing the vertical temperature gradient fatigue load spectrum of the flat steel box girder cable-stayed bridge comprises the following steps:

[0053] Step 1: Arrange temperature measuring points on the top plate, bottom plate and diaphragm of the flat wide steel box girder bridge, establish a coordinate system with the center of the bottom plate as the coordinate origin, and arrange 8 measuring points in total. The vertical distance from the lower surface of the bottom plate of the flat wide steel box girder is y. The positions of the 8 measuring points can be recorded as 0.00m, 1.98m, 2.68m, 3.20m, 3.30m, 3.40m, 3.45m and 3.50m respectively. Among them, the measuring point y=0.0m is arranged on the bottom plate of the steel box girder, and the measuring point y=3.14m is arranged on the top plate of the steel box girder. The measuring points are arranged as follows: Figure 5 As shown. The temperature values ​​of each measuring point are collected at intervals of 10s to obtain the measured temperature gradient time history curve; Figure 6 The measured temperature gradient time history curve is extended to the design service life N through the long short memory recurrent neural network. d The temperature history at a typical location is selected, and the maximum temperature at each measuring point on each day is formed into a set of daily temperature extreme value vectors, so as to obtain the N d ×365 groups of daily temperature extreme value vectors, recorded as the design service life daily temperature vector set M. In this embodiment, the collection time of the temperature value of the measuring point can also be 1 second interval or 1800 seconds interval;

[0054] Step 2. Randomly select a set of daily temperature extreme value vectors from the design service life daily temperature vector set M as the cluster center C, use the mean shift method to establish a cluster D based on the cluster center C, and expand the cluster D according to the cluster expansion method;

[0055] The cluster expansion method is:

[0056] 1) Calculate the Euclidean distance between other daily temperature extreme value vectors in the daily temperature vector set M of the design service life and the cluster center C, and select the Euclidean distance d e The daily temperature extreme value vectors with a mean shift radius of R are classified into cluster D. At the same time, the frequency of each group of selected daily temperature extreme value vectors being classified into the current cluster D is recorded as an increase of 1;

[0057] Euclidean distance d e According to the following formula:

[0058]

[0059] Where, d e is the Euclidean distance, T Nd (y k ,t C ) is t C Date k The temperature maximum at the location, y k is the measuring point position, t C is the date of the cluster center, and t is the date of the daily temperature extreme value vector other than the cluster center.

[0060] 2) Calculate the vector differences between the daily temperature extreme vectors and the cluster center in cluster D except the cluster center, add the vector differences and calculate the average to get vector S. Then the cluster center drifts to the next position along the direction of vector S, and the drift distance is the modulus of vector S.

[0061] The above vector S is:

[0062]

[0063] Where z is the number of groups of daily temperature extreme value vectors in the cluster, t l The date of the extreme temperature vector of the lth group.

[0064] 3) After the cluster center arrives at the new location, the Euclidean distance d between the design service life daily temperature vector set M and the cluster center is again calculated. e The daily temperature extreme value vectors with ≤R are classified into cluster D, and the frequency of each group of selected daily temperature extreme value vectors being classified into cluster D is recorded as an increase of 1;

[0065] 4) Repeat steps 2) and 3) until no new daily temperature extreme value vectors are added to cluster D, completing the expansion of cluster D. Count the total frequency of each group of daily temperature extreme value vectors in cluster D being classified into cluster D, and record it as the cluster frequency.

[0066] Step 3. Randomly select a set of daily temperature extreme value vectors from the daily temperature vector set M of the design service life except the cluster cluster as the virtual cluster center C / And establish a virtual cluster D with the virtual cluster center / , expand the virtual cluster D according to the cluster expansion method in step 2 / ;

[0067] Step 4. Calculate virtual clusters D / The Euclidean distance between the virtual cluster center and the cluster center in cluster D, if the Euclidean distance d e >Set threshold d th , then the virtual cluster D / Add as a new cluster; if the Euclidean distance d e <Set threshold d th , then the virtual cluster D / Merge the cluster with the shortest Euclidean distance;

[0068] Step 5. According to steps 3 and 4, all the daily temperature extreme value vectors in the daily temperature vector set M of the design service life are traversed, and the cluster frequency of each group of daily temperature extreme value vectors in each cluster is compared. Each group of daily temperature extreme value vectors is classified into the cluster with the highest corresponding cluster frequency. The obtained cluster is the sub-gradient T of the vertical temperature gradient fatigue load spectrum. Gi (y), the clustering results are as follows Figure 7 The cluster center of the cluster is the temperature representative value of the sub-gradient. The ratio of the cluster frequency of each group of daily temperature extreme value vectors in the same cluster to the total cluster frequency of all vectors is the probability of occurrence of the sub-gradient in the design service life P(T Gi ), thus obtaining the vertical temperature gradient fatigue load spectrum of the long-life flat steel box girder bridge.

[0069] The transverse temperature gradient fatigue load spectrum of the bridge consists of six transverse temperature sub-gradients and the probability of occurrence of the sub-gradients within the design service life, specifically:

[0070]

[0071] In formula (2), T Hq (x) is the temperature of the qth transverse temperature sub-gradient at the transverse position x, L is the width of the flat steel box girder, the unit is m, T Hq,w is the temperature representative value at the wth typical transverse position in the qth transverse temperature sub-gradient, T Hq,1 、T Hq,2 、T Hq,3 、T Hq,4 are the representative temperature values ​​at the edge of the sunny side, the position close to the 1 / 4L position of the sunny side, the position close to the 1 / 2L position of the sunny side, and the edge of the shady side in the qth transverse temperature sub-gradient, T H1,1 、T H1,2 、T H1,3 、T H1,4 are the representative temperature values ​​at the edge of the sun side, the position close to the 1 / 4L position of the sun side, the position close to the 1 / 2L position of the sun side, and the edge of the sun side in the first transverse temperature sub-gradient, T H1,1 The value range is [12.0,14.0], the unit is ℃, T H1,2 The value range is [12.4,14.4], and the unit is ℃. H1 (y))、P(T H2 (y))、P(T H3 (y))、P(T H4 (y))、P(T H5 (y))、P(T H6 (y)) are the occurrence probabilities of the six transverse temperature sub-gradients within the design service life, T H1,1 、T H2,1 、TH3,1 、T H4,1 、T H5,1 、T H6,1 They are the representative values ​​of the temperature at the edge of the sunny side in the six lateral temperature sub-gradients.

[0072] Transverse temperature gradient fatigue load spectrum T H1,1 =13.3℃,T H1,2 =13.8℃, the temperature gradient fatigue load spectra of the flat steel box girder bridge with design service life of 100 years, 150 years and 200 years are shown in Table 2. The temperature linear gradient model of the typical position is as follows: Figure 8 As shown, 150 years and 200 years are the long-life design service life, and the transverse temperature sub-gradient T H1 ~T H6 The frequencies are 21.8%, 18.2%, 15.5%, 24.5%, 15.5%, and 4.5%. Figure 9 shown.

[0073] Table 2 Representative values ​​of gradient temperature for transverse temperature gradient fatigue load spectrum of flat wide steel box girder

[0074]

[0075] A method for constructing a transverse temperature gradient fatigue load spectrum for a flat steel box girder cable-stayed bridge includes the following steps:

[0076] Step 1: Arrange temperature measuring points on the top plate of the flat steel box girder cable-stayed bridge. Take the positive side edge of the top plate as the coordinate origin and arrange temperature measuring points in the width direction of the top plate. The measuring point positions are expressed as 0.00m, 5.60m, 12.70m, 16.55m, 20.40m, 27.50m, and 33.00m in horizontal distance from the coordinate origin. The measuring points are arranged as follows: Figure 5 As shown in Figure 2, the other steps are the same as the method for constructing the vertical temperature gradient fatigue load spectrum.

[0077] Steps 2 to 5 are the same as the method for constructing the vertical temperature gradient fatigue load spectrum of the flat steel box girder cable-stayed bridge. The clustering results are as follows: Figure 10 shown.

Claims

1. A temperature gradient fatigue load spectrum for a flat wide steel box girder bridge, characterized by: The temperature gradient fatigue load spectrum includes a vertical temperature gradient fatigue load spectrum and a transverse temperature gradient fatigue load spectrum; The vertical temperature gradient fatigue load spectrum consists of six vertical temperature sub-gradients and the probability of occurrence of the sub-gradients within the design service life, specifically: In formula (1), T Gi (y) is the temperature of the ith vertical temperature sub-gradient at position y, h is the beam height of the flat steel box girder bridge, the unit is m, Gi is the ith vertical temperature sub-gradient, T Gi,1 、T Gi,2 、T Gi,3 They are the top plate and the distance h between the nozzle and the bottom plate in the i-th vertical temperature sub-gradient. f The temperature at the bottom plate is represented by T G1,1 、T G1,2 、T G1,3 They are respectively the top plate in the first vertical temperature sub-gradient, the distance between the nozzle and the bottom plate h f The temperature at the bottom plate is represented by T G1,1 The value range is [4.8,7.8], the unit is ℃, T Gi,j is the temperature at the jth typical height in the ith vertical sub-temperature gradient, N d is the design service life, P(T G1 (y))、P(T G2 (y))、P(T G3 (y))、P(T G4 (y))、P(T G5 (y))、P(T G6 (y)) are the occurrence probabilities of the six vertical temperature sub-gradients within the design service life, T G1,1 、T G2,1 、T G3,1 、T G4,1 、T G5,1 、T G6,1 are the representative temperature values ​​of the six vertical temperature sub-gradients at the top plate; The transverse temperature gradient fatigue load spectrum is composed of six transverse temperature sub-gradients and the probability of occurrence of the sub-gradients within the design service life, specifically: In formula (2), T Hq (x) is the temperature of the qth transverse temperature sub-gradient at the transverse position x, L is the width of the flat steel box girder, the unit is m, T Hq,w is the temperature representative value at the wth typical transverse position in the qth transverse temperature sub-gradient, T Hq,1 、T Hq,2 、T Hq,3 、T Hq,4 are the representative temperature values ​​at the edge of the sunny side, the position close to the 1 / 4L position of the sunny side, the position close to the 1 / 2L position of the sunny side, and the edge of the shady side in the qth transverse temperature sub-gradient, T H1,1 、T H1,2 、T H1,3 、T H1,4 are the representative temperature values ​​at the edge of the sun side, the position close to the 1 / 4L position of the sun side, the position close to the 1 / 2L position of the sun side, and the edge of the sun side in the first transverse temperature sub-gradient, T H1,1 The value range is [12.0,14.0], the unit is ℃, T H1,2 The value range is [12.4,14.4], the unit is ℃; P(T H1 (y))、P(T H2 (y))、P(T H3 (y))、P(T H4 (y))、P(T H5 (y))、P(T H6 (y)) are the occurrence probabilities of the six transverse temperature sub-gradients within the design service life, T H1,1 、T H2,1 、T H3,1 、T H4,1 、T H5,1 、T H6,1 They are the representative values ​​of the temperature at the edge of the sunny side in the six lateral temperature sub-gradients.

2. The method for constructing the temperature gradient fatigue load spectrum of a flat wide steel box girder bridge according to claim 1 is characterized in that: The following steps are involved: Step 1. Arrange c temperature measuring points on the top plate, bottom plate, and diaphragm of the flat wide steel box girder bridge, and collect the temperature values ​​of each measuring point at regular intervals to obtain the measured temperature gradient time history curve; extend the measured temperature curve of the flat steel box girder bridge to the design service life N through the long-short memory recurrent neural network. d The temperature history at a typical location is selected, and the maximum temperature at each measuring point on each day is formed into a set of daily temperature extreme value vectors to obtain the N d ×365 groups of daily temperature extreme value vectors, recorded as the design service life daily temperature vector set M; Step 2. Randomly select a set of daily temperature extreme value vectors from the design service life daily temperature vector set M as the cluster center C, use the mean shift method to establish a cluster D based on the cluster center C, and expand the cluster D according to the cluster expansion method; The cluster expansion method is: 1) Calculate the Euclidean distance between other daily temperature extreme value vectors in the daily temperature vector set M of the design service life and the cluster center C, and select the Euclidean distance d e The daily temperature extreme value vectors with a mean shift radius of R are classified into cluster D, and the frequency of each selected group of daily temperature extreme value vectors being classified into the current cluster D is recorded as an increase of 1; 2) Calculate the vector differences between the daily temperature extreme vectors and the cluster center in cluster D except the cluster center, add the vector differences and calculate the average to get vector S. Then the cluster center drifts to the next position along the direction of vector S, and the drift distance is the modulus of vector S. 3) After the cluster center arrives at the new location, the Euclidean distance d between the design service life daily temperature vector set M and the cluster center is again calculated. e The daily temperature extreme value vectors with ≤R are classified into cluster D, and the frequency of each group of selected daily temperature extreme value vectors being classified into cluster D is recorded as an increase of 1; 4) Repeat steps 2) and 3) until no new daily temperature extreme value vectors are added to cluster D, completing the expansion of cluster D. Count the total frequency of each group of daily temperature extreme value vectors in cluster D being classified into cluster D, and record it as the cluster frequency. Step 3. Randomly select a set of daily temperature extreme value vectors from the daily temperature vector set M of the design service life except the cluster cluster as the virtual cluster center C / And establish a virtual cluster D with the virtual cluster center / , expand the virtual cluster D according to the cluster expansion method in step 2 / ; Step 4. Calculate virtual clusters D / The Euclidean distance between the virtual cluster center and the cluster center in cluster D, if the Euclidean distance d e >Set threshold d th , then the virtual cluster D / Add as a new cluster; if the Euclidean distance d e <Set threshold d th , then the virtual cluster D / Merge the cluster with the shortest Euclidean distance; Step 5. According to steps 3 and 4, traverse all the daily temperature extreme value vectors in the daily temperature vector set M of the design service life, compare the clustering frequency of each group of daily temperature extreme value vectors in each cluster, and classify each group of daily temperature extreme value vectors into the cluster with the highest corresponding clustering frequency. The obtained cluster is the sub-gradient of the temperature gradient fatigue load spectrum, and the cluster center of the cluster is the temperature representative value of the sub-gradient. The ratio of the clustering frequency of each group of daily temperature extreme value vectors in the same cluster to the sum of the clustering frequencies of all vectors is the probability of occurrence of the sub-gradient in the design service life.

3. The method for constructing the temperature gradient fatigue load spectrum of a flat wide steel box girder bridge according to claim 2 is characterized in that: The Euclidean distance d in step 1) of step 2 e According to the following formula: Where, d e is the Euclidean distance, T Nd (y k ,t C ) is t C Date k The temperature maximum at the location, y k is the measuring point position, t C is the date of the cluster center, and t is the date of the daily temperature extreme value vector other than the cluster center.

4. The method for constructing the temperature gradient fatigue load spectrum of a flat wide steel box girder bridge according to claim 2 is characterized in that: In step 2), the vector S is: Where z is the number of groups of daily temperature extreme value vectors in the cluster, t l The date of the extreme temperature vector of the lth group.

5. The method for constructing a temperature gradient fatigue load spectrum for a flat wide steel box girder bridge according to claim 2 is characterized by: The drift radius R ranges from [1.2, 2.8]; the threshold d th The value range is [1.0,4.5].

6. The method for constructing a temperature gradient fatigue load spectrum for a flat wide steel box girder bridge according to claim 2 is characterized in that: The c temperature measuring points described in step 1 are used to construct a vertical temperature gradient fatigue load spectrum. The arrangement method is as follows: one measuring point is arranged on the upper surface of the bottom plate of the flat wide steel box girder bridge, and the measuring point is used as the coordinate origin. Measuring points are arranged on the diaphragm and the top plate along the vertical height direction. A total of c = 8 measuring points are arranged. The vertical distance between the measuring point position and the lower surface of the bottom plate of the flat wide steel box girder bridge can be 0.00m, h-1.52m, h-0.82m, h-0.30m, h-0.20m, h-0.10m, h-0.05m, h.

7. The method for constructing a temperature gradient fatigue load spectrum for a flat wide steel box girder bridge according to claim 2 is characterized in that: The c temperature measuring points described in step 1 are used to construct the transverse temperature gradient fatigue load spectrum. The arrangement method is as follows: one measuring point is arranged at the edge of the positive side of the top plate, and this measuring point is used as the coordinate origin. A total of c = 7 measuring points are arranged along the width direction of the top plate. The positions of the measuring points are expressed as 0.00m, L-27.40m, L-20.30m, L-16.45m, L-12.60m, L-5.50m, and L in terms of the horizontal distance from the coordinate origin.

8. The method for constructing the temperature gradient fatigue load spectrum of a flat wide steel box girder bridge according to claim 2 is characterized in that: The temperature values ​​of each measuring point are collected at intervals of 60 to 1200 seconds in step 1.

Images