Bridge structure-oriented near / through strike-slip fault ground motion synthesis method

Abstract

The application discloses a kind of near / cross strike-slip fault ground motion synthesis methods for bridge structure. The method comprises the following steps: step one, according to the bridge site and the selected wave principle, select the near-fault pulse type seismic record matched with the target response spectrum. Step two, rotate the seismic record to parallel and perpendicular fault direction, and decompose it into high-frequency component and low-frequency component using a filter. Step three, based on continuous wavelet transform, the low-frequency component is further decomposed, the pulse component is simulated using pulse function, and the fitting parameters are determined by introducing particle swarm optimization algorithm. Step four, the residual component, the artificial fitting pulse and the high-frequency component are superimposed to obtain the wide-frequency near-fault ground motion. Step five, select the pulse waveform with slip-off effect and direction effect to replace the fitting pulse, extract the next high-frequency component in the residual component, and finally superimpose the two with the high-frequency component to obtain the across-fault ground motion, to realize the accurate simulation of strike-slip fault ground motion.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of bridge anti-seismic, in particular to a near / cross strike-slip fault ground motion synthesis method for bridge structure. BACKGROUND

[0002] China is located between two major seismic belts in the world (i.e. the circum-Pacific seismic belt and the Eurasian seismic belt), is affected by the extrusion of the Pacific plate, the Indian plate and the Philippine Sea plate, and has strong new geological structure activity and widely developed seismic fault zones. There are more than 500 seismic fault zones identified in the country, among which large active fault zones are hundreds of kilometers or even thousands of kilometers long, and almost cover the entire territory. Multiple earthquake disasters have shown that under the dual action of ground vibration and surface rupture, the seismic response of cross-fault bridges is significantly different from that of far-field bridges, and bridges directly crossing the fault are more likely to suffer devastating damage.

[0003] Although the principle of avoiding the seismogenic fault is adopted in the anti-seismic specifications of various countries, it is inevitable for highway and railway bridges as traffic hubs to pass through active fault regions with high seismic risk. With the continuous expansion of China's highway transportation network, more and more highway bridges are built in high seismic intensity regions such as northwest Yunnan, southeast Tibet, west Sichuan and southeast coastal areas, resulting in a continuous increase in the total number and density of bridges. Due to the limitations of topography, construction cost, engineering period and regional economic development, etc., some newly built bridges and existing bridges have to be built near the fault or directly cross the active fault.

[0004] At present, there are problems such as unclear understanding of the seismic action of bridges crossing the fault, unknown disaster-causing mechanism and insufficient seismic resilience, and there is a lack of systematic research on the influence of fault dislocation, which exposes the deficiencies and limitations of the seismic work of cross-fault bridges. As the first step of seismic assessment of cross-fault bridges, reasonable ground motion input is a prerequisite for the above research work. It is urgent to adopt different methods, including theoretical analysis, numerical simulation and model test, to systematically and deeply study the synthesis of cross-fault ground motion, reveal its pulse characteristics, and input accurate and reasonable cross-fault bridge ground motion. However, cross-fault ground motion is significantly different from far-field ground motion, and different researchers do not use the same method to synthesize cross-fault ground motion, and the randomness and contingency factors of ground motion selection in the synthesis process are large, so it is necessary to further break through and improve the convenience, applicability and accuracy. SUMMARY

[0005] The purpose of the present application is to provide a near / cross strike-slip fault ground motion synthesis method for bridge structure, which is not affected by the actual number of fault earthquakes and can realize accurate simulation of strike-slip fault ground motion, providing strong support for seismic assessment of bridge structures in strong earthquake regions, to improve the rationality of near / cross strike-slip fault bridge ground motion input.

[0006] The application adopts the technical solutions below.

[0007] A near / cross-strike-slip fault seismic motion synthesis method for bridge structures, according to the bridge site type and fault mechanism, a near-fault pulse-type seismic record matched with the current specification response spectrum is selected. The selected seismic record is rotated to be parallel and perpendicular to the fault strike, and a Butterworth non-causal filter is used to decompose it into high-frequency and low-frequency components; the low-frequency component is further decomposed based on continuous wavelet transformation, a pulse function is used to accurately simulate the pulse component, and a particle swarm optimization algorithm is introduced to determine the fitting parameters; the residual component, the artificially fitted velocity pulse and the high-frequency component are superimposed to obtain the global broadband near-fault seismic motion; further synthesis of cross-fault seismic motion, selecting a pulse waveform with a slip-off effect or a directional effect to replace the fitted pulse, and eliminating the sub-low frequency component in the residual component, and finally superimposing the high-frequency component to obtain the cross-fault seismic motion.

[0008] The method specifically comprises the following steps:

[0009] Step one: near-fault seismic record selection. According to the bridge site conditions, the horizontal acceleration response spectrum specified in the Highway Bridge Seismic Design Specification (JTG / T 2231-01—2020) is established; and the seismic record is selected from the PEER Ground Motion Database according to the selected wave principle;

[0010] Step two: rotation and high-low frequency decomposition. According to the parallelogram rule, the selected seismic record is rotated to be parallel to the fault strike and perpendicular to the fault strike; a Butterworth non-causal filter and a filter frequency formula are used to decompose the parallel or perpendicular fault seismic motion into high-frequency and low-frequency components;

[0011] Step three: simulation of velocity pulse component. Based on continuous wavelet transformation, the low-frequency component of the parallel and perpendicular fault seismic motion is decomposed, the pulse component is extracted, an artificial velocity pulse function (Dickinson function) is used to fit the pulse component, and a particle swarm optimization algorithm is introduced to determine the fitting parameters;

[0012] Step four: near-fault seismic motion synthesis. The residual component in step three, the artificially fitted velocity pulse and the high-frequency component in step two are superimposed to obtain the global broadband near-fault seismic motion;

[0013] Step five: synthesis of cross-fault seismic motion. A pulse waveform with a slip-off effect and a directional effect is selected to replace the artificially fitted pulse, and further the sub-low frequency component in the residual component is eliminated, and finally superimposed with the high-frequency component to obtain the cross-fault seismic motion.

[0014] Further, the determination of the target response spectrum in step one is as follows:

[0015] The horizontal acceleration response spectrum specified in the Code for Seismic Design of Highway Bridges (JTG / T 2231-01-2020) (hereinafter referred to as the Code) is taken as the target spectrum, and the calculation formula is:

[0016]

[0017] S max =2.5C i C s C d A (2)

[0018] In the formula, T is the period; T0 is the maximum period of the linear rising section of the response spectrum, which is taken as 0.1 s; T g is the characteristic period; S max is the maximum value of the horizontal design acceleration response spectrum; C i is the seismic importance coefficient; C s is the site coefficient; C d is the damping adjustment coefficient; and A is the horizontal basic ground motion peak acceleration.

[0019] The definition of the selection principle is as follows:

[0020] ① The selected strong motion records are pulse records from a free field;

[0021] ② The selected strong motion records are classified according to the fault mechanism and site conditions at the bridge site;

[0022] ③ The minimum available frequency of the selected strong motion records is less than or equal to 0.2 Hz;

[0023] ④ The fault distance of the selected strong motion records is not less than 10 km, which requires that the strong motion records should have the minimum velocity pulse effect before the superposition of the pulse effect;

[0024] ⑤ The selected magnitude is converted from the intensity corresponding to the target spectrum, with a variation range of ±0.5 magnitude grade. The conversion is performed according to the relationship between magnitude and epicenter intensity M = 0.58I0 + 1.5.

[0025] Further, the rotation of the original seismic record in step two is as follows:

[0026] Linear combination of the two original horizontal orthogonal components f1(t) and f2(t) is performed to obtain the seismic records in the parallel fault direction and the perpendicular fault direction, respectively:

[0027]

[0028] In the formula, θ is the angle between the fault strike and the station record; f FP (t, θ) and fFN (t, θ) is the record component of parallel and vertical fault direction after rotating θ angle.

[0029] Further, the high and low frequency decomposition of the parallel and vertical fault direction seismic records in step two is as follows:

[0030] Butterworth non-stationary filter is used to filter the parallel and vertical fault components. The filter frequency response function H(f) is as follows:

[0031]

[0032] In the formula, f is the input frequency; f c is the cut-off frequency; n is the filter order, which is 4.

[0033] The filter cut-off frequency f c is calculated as follows:

[0034] f c = 1 / (αT p -dt) (5)

[0035] In the formula, T p is the pulse period of the seismic record; α is an empirical coefficient, and dt is the time step of ground motion.

[0036] Where the pulse period T p is determined by the peak point method, and the value is the time interval of the wave peak or wave trough adjacent to the peak value PGV in the velocity time history. Since the parallel fault direction of strike-slip fault has the effect of slip-off, and the vertical fault direction has the effect of directionality, α is taken as 0.25 for the parallel fault direction, and α is taken as 0.8 for the vertical fault direction seismic record.

[0037] Further, the extraction of the velocity pulse component in step three is as follows:

[0038] The base function expression in wavelet analysis is:

[0039]

[0040] In the formula, Φ(·) represents the mother wavelet function; s is the scale factor for expanding the wavelet; and l is the translation factor for translating the wavelet in time.

[0041] In the linear combination representing the original signal f(t), the combination coefficient C s,l can be obtained by the following formula:

[0042]

[0043] Further, the fitting of the velocity pulse component in step three is as follows:

[0044] The characteristics of ground motion are statistically sampled from real earthquake records. The Gabor wavelet is used to model the low-frequency pulse in the velocity time history v p The model is simulated. The velocity pulse function in the model is given by

[0045]

[0046] where V p is the peak velocity; T p is the pulse period; T pk is the peak time; N c is the number of cycles; and φ is the phase angle.

[0047] The Gabor wavelet model contains five unknown parameters, V p , T p , T pk , N c , and φ. The particle swarm optimization algorithm is introduced to determine the pulse parameters to be fitted. The fitness function is the objective function, as follows:

[0048] f(t) = min |v(t) - v p (t)| (9)

[0049] The velocity update formula of the dth dimension of particle i is:

[0050]

[0051] The position update formula of the dth dimension of particle i is:

[0052]

[0053] where i is the particle index; d is the dimension of the space in which the particle is located; k is the evolution number of the particle; v is the particle velocity; p is the best position of particle i in the dth dimension; x is the particle position; g is the best position of all particles in the dth dimension; c1 and c2 are learning factors used to adjust the maximum step length of learning; r1 and r2 are two random numbers with values in the range [0-1]; and ω is the inertia weight.

[0054]

[0055] Further, the synthesis of the cross-fault ground motion in step five is as follows:

[0056] For strike-slip faults, the ground motion parallel to the fault direction has a clear slip effect. The frequency f p of the pulse component is determined according to the following formula:

[0057] ​lg(1 / f p ) = -2.127 + 0.386M w (12)

[0058] Permanent ground displacement D site and fault plane slip D fault of strike-slip fault are calculated as follows:

[0059] ln(D fault ) = 1.15M W - 3.28 (13)

[0060]

[0061] where M W is the moment magnitude; a0, a2, a3 are regression parameters; R rup is the fault distance; δ is the fault dip angle.

[0062] Permanent ground displacement D site and maximum ground displacement D max satisfy the following relationship:

[0063] D max ≈ D site / λ, λ ∈ [0.2, 0.8] (16)

[0064] For strike-slip fault, the ground motion perpendicular to the fault direction has obvious directional effect. The peak velocity PGV and the pulse period T p (T p = 1 / f p ) are as follows:

[0065] lg(T p ) = -1.067 + 0.299M w (17)

[0066] ln(PGV) = 2.17 + 0.34M W - 0.11 ln R rup (18)

[0067] where PGV is the peak velocity, which determines the value of pulse amplitude V p , V p can be taken as [0.85, 1.00] PGV; R rup is the fault distance.

[0068] For the residual component, high-pass filtering with a cutoff frequency of 1 Hz is used to filter out the low-frequency components and retain the high-frequency components.

[0069] The high frequency component is superimposed with the second high frequency component to obtain a final high frequency peak velocity, the pulse peak time of the artificial pulse function with the slide shock effect and the direction effect is adjusted to the corresponding time of the final high frequency peak velocity, and finally, the two are superimposed to obtain the cross-fault ground motion in the parallel fault direction and the vertical fault direction.

[0070] Beneficial effects:

[0071] (1) The target response spectrum and the wave selection principle are established, so that the influence of the fault mechanism and the site condition of the bridge site can be fully considered.

[0072] (2) The rotation change and the filtering processing are adopted, so that the high frequency and the low frequency components of the parallel and vertical fault direction ground motions can be accurately obtained respectively.

[0073] (3) The low frequency component is further decomposed through the continuous wavelet change, so that the velocity pulse characteristics and the residual component can be fully considered, and the actual ground motion is more consistent.

[0074] (4) The effective synthesis of the near / cross strike-slip fault ground motion can be realized, and accurate ground motion input is provided for the bridge seismic evaluation in the strong earthquake area. BRIEF DESCRIPTION OF DRAWINGS

[0075] Figure 1 It is a flowchart of the present application;

[0076] Figure 2 It is a schematic diagram of the target response spectrum and the selected seismic wave response spectrum;

[0077] Figure 3 It is a schematic diagram of the extraction and synthesis of the pulse component; wherein (a) extraction, (b) synthesis;

[0078] Figure 4 It is a comparison diagram of the synthesized near-fault ground motion and the original record;

[0079] Figure 5 It is a velocity pulse waveform with typical slide shock effect and direction effect; wherein (a) slide shock effect, (b) direction effect.

[0080] Figure 6 It is a schematic diagram of the parallel and vertical fault direction of the synthesized cross-fault ground motion; wherein (a) parallel fault direction, (b) vertical fault direction. DETAILED DESCRIPTION

[0081] In order to make the purpose and technical scheme of the embodiments of the present application clearer, the technical scheme of the embodiments of the present application will be described clearly and completely below in combination with the drawings of the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments.

[0082] Embodiment 1

[0083] As Figures 1-5 shown, a near / cross strike-slip fault ground motion synthesis method for bridge structure includes the following steps:

[0084] Step one: near-fault seismic record selection. According to the site conditions of the bridge, the horizontal acceleration response spectrum specified in the Code for Seismic Design of Highway Bridges (JTG / T 2231-01-2020) is established; According to the selected wave principle, select the wave from the PEER Ground Motion Database of the United States Pacific Earthquake Engineering Research Center;

[0085] Step two: rotation change and high-low frequency decomposition. According to the parallelogram rule, the selected seismic record is rotated to the parallel fault strike and the vertical fault strike according to the fault strike; The Butterworth non-causal filter and the filter frequency formula are used to decompose the parallel or vertical fault ground motion into high-frequency components and low-frequency components;

[0086] Step three: simulation of velocity pulse component. Based on the continuous wavelet change, the low-frequency component of the parallel and vertical fault ground motion is decomposed, the pulse component is extracted, the artificial velocity pulse function (Dickinson function) is used to fit the pulse component, and the particle swarm optimization algorithm is introduced to determine the fitting parameters;

[0087] Step four: near-fault ground motion synthesis. The residual component in step three, the artificial fitted velocity pulse, and the high-frequency component in step two are superimposed to obtain the full-bandwidth near-fault ground motion;

[0088] Step five: cross-fault ground motion synthesis. Select an impulse waveform with a slip-off effect and a directional effect to replace the artificial fitting pulse, and further remove the sub-low frequency component in the residual component, and finally superimpose it with the high-frequency component to obtain the cross-fault ground motion.

[0089] As Figure 2 shown, the determination of the target response spectrum and the wave selection principle in step one is as follows:

[0090] The horizontal acceleration response spectrum specified in the Code for Seismic Design of Highway Bridges (JTG / T 2231-01-2020) (hereinafter referred to as the Code) is taken as the target spectrum, and the calculation formula is:

[0091]

[0092] S max =2.5C i C s C d A (2)

[0093] Where, T is the period; T0 is the maximum period of the rising section of the response spectrum straight line, 0.1 s; T g is the characteristic period; S max is the maximum value of the horizontal design acceleration response spectrum; C i is the seismic importance coefficient; C s is the site coefficient; C d is the damping adjustment coefficient; A is the horizontal basic ground motion peak acceleration.

[0094] The selected wave principle is defined as follows:

[0095] ①The selected strong motion record is a pulse record from a free field;

[0096] ②The selected strong motion record is selected according to the fault mechanism and site conditions at the bridge site;

[0097] ③The minimum available frequency of the selected strong motion record is less than or equal to 0.2 Hz;

[0098] ④The fault distance of the selected strong motion record is not less than 10 km, which requires that the strong motion record should have the minimum velocity pulse effect before the superposition of the pulse effect;

[0099] ⑤The selected magnitude is converted from the target spectrum corresponding intensity, with a range of ±0.5 magnitude grade. The conversion is performed according to the relationship between magnitude and epicenter intensity M = 0.58I0 + 1.5.

[0100] Further, in step two, the rotation of the original seismic record is as follows:

[0101] Linear combination of the two original horizontal orthogonal components f1(t) and f2(t) is performed to obtain the parallel fault direction and the perpendicular fault direction seismic records:

[0102]

[0103] Where: θ is the angle between the fault strike and the station record; f FP (t, θ), f FN (t, θ) are the parallel and perpendicular fault direction record components after rotating θ degrees.

[0104] Further, the high and low frequency decomposition of the parallel and perpendicular fault direction seismic records in step two is as follows:

[0105] Butterworth non-stationary filter is used to filter the parallel and perpendicular fault components. The filter frequency response function H(f) is as follows:

[0106]

[0107] Where, f is the input frequency; f cis the cut-off frequency; n is the filter order, which is 4.

[0108] Filter cut-off frequency f c is calculated as follows:

[0109] f c = 1 / (aT p -dt) (5)

[0110] In the formula, T p is the pulse period of the seismic record; a is an empirical coefficient, and dt is the time step of the ground motion.

[0111] where the pulse period T p is determined by the peak point method, which is the time interval of the wave peak or wave trough adjacent to the peak value PGV in the velocity time history; since the parallel fault direction of the strike-slip fault has a thrust effect, and the vertical fault direction has a directional effect, a is taken as 0.25 for the parallel fault direction, and a is taken as 0.8 for the vertical fault direction of the seismic record.

[0112] As shown in FIG. 3, the extraction and fitting of the velocity pulse component in step three are as follows: Figure 3

[0113] The base function expression in wavelet analysis is:

[0114]

[0115] In the formula, Φ(·) represents the mother wavelet function; s is a scale factor for expanding the wavelet, and l is a translation factor for translating the wavelet in time.

[0116] In the linear combination representing the original signal f(t), the combination coefficient C s,l can be obtained by the following formula:

[0117]

[0118] By statistically analyzing the ground motion characteristics from actual seismic record samples, the Gabor wavelet is used to analyze the low-frequency pulse velocity time history v p of the simulated model. The following formula gives the velocity pulse function in the model:

[0119]

[0120] In the formula, V p is the pulse peak velocity; T p is the pulse period; T pk is the pulse peak time; N c is the pulse cycle number; and φ is the pulse phase angle.

[0121] The Gabor wavelet model contains V p ​T p T pk N c φtotal 5 unknown parameters, the particle swarm optimization algorithm is introduced to determine the pulse parameters to be fitted. The fitness function is the objective function, as follows:

[0122] f(t) = min |v(t) - v p (t) | (9)

[0123] The velocity update formula of the dth dimension of particle i:

[0124]

[0125] The position update formula of the dth dimension of particle i:

[0126]

[0127] Where: i is the particle label; d is the dimension of the space where the particle is located; k is the evolution number of the particle; is the particle velocity; is the best position of particle i in the dth dimension space; is the particle position; is the best position of all particles in the dth dimension space; c1, c2 are learning factors used to adjust the maximum step size of learning; r1, r2 are two random numbers with a value range of [0-1]; ω is the inertia weight.

[0128] As Figure 4 shown, the residual component in step three, the artificially fitted velocity pulse, and the high-frequency component in step two are superimposed to obtain the global broadband near-fault ground motion.

[0129] As Figure 5 shown, the pulse waveform of the slip-off effect and the directivity effect in step five is selected as follows:

[0130] For strike-slip faults, the ground motion parallel to the fault direction has obvious slip-off effect. The frequency f p in the pulse component is determined according to the following formula:

[0131] ln(1 / f p ) = -2.127 + 0.386M w (12)

[0132] The calculation formulae of the ground permanent displacement D site and the fault plane slip amount D fault of strike-slip faults are as follows:

[0133] ln(D fault ) = 1.15M W - 3.28 (13)

[0134]

[0135] In the formula, M W The moment magnitude is represented by a0, a2, and a3; these are regression parameters; R0 rup δ is the fault distance; δ is the fault dip angle.

[0136] Ground permanent displacement D site and maximum ground displacement D max The following relationship must be satisfied:

[0137] D max ≈D site / λ,λ∈[0.2,0.8](16)

[0138] For strike-slip faults, ground motions perpendicular to the fault direction exhibit a significant directional effect. Peak velocity (PGV) and pulse period (T) are also relevant. p (T p =1 / f p The value of ) is shown in the following formula:

[0139] lg(T p = -1.067 + 0.299M w (17)

[0140] ln(PGV) = 2.17 + 0.34M W -0.11lnR (18)

[0141] In the formula, PGV is the peak velocity, which determines the pulse amplitude V. p The value of V p A range of [0.85, 1.00] PGV can be taken; R rup This is the fault distance.

[0142] By repeatedly adjusting other parameters, a result exhibiting the slippage effect can be obtained. Figure 5 (a) directional effect ( Figure 5 (b) Trans-fault artificial pulse.

[0143] like Figure 6 As shown, the synthesis of trans-fault ground motions in step five is as follows:

[0144] The residual components are filtered using a high-pass filter with a cutoff frequency of 1Hz to remove the second-lowest frequency components and retain the second-highest frequency components.

[0145] The high frequency component is superimposed with the sub-high frequency component to obtain a final high frequency peak velocity, and the pulse peak time of the artificial pulse function with the slide and direction effects is adjusted to the corresponding time of the final high frequency peak velocity, and finally the two are superimposed to obtain the cross-fault ground motion in the parallel fault direction and the vertical fault direction Figure 6 (a), (b).

[0146] The above merely describes preferred specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any changes or replacements easily thought of by those skilled in the art within the technical scope disclosed by the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for synthesizing near / across strike-slip fault ground motion for a bridge structure, characterized by, The method comprises the following steps: Step one: selection of near-fault seismic records: select near-fault pulse-type seismic records matching the current specification response spectrum according to the bridge site conditions and the relevant wave selection principles; Step two: rotation and high-low frequency decomposition: according to the parallelogram rule, rotate the selected seismic records to the parallel fault direction and the vertical fault direction according to the fault strike; use a filter to decompose the parallel or vertical fault ground motion into high-frequency components and low-frequency components; Step three: simulation of velocity pulse components: further decompose the low-frequency components of the parallel and vertical fault ground motion based on continuous wavelet variation, extract the pulse components, use an artificial velocity pulse function to fit the pulse components, and introduce a particle swarm optimization algorithm to determine the related fitting parameters, thereby obtaining the artificial velocity pulse component; Step four: near-fault ground motion synthesis: superimpose the residual component in step three, the artificial fitted velocity pulse, and the high-frequency component in step two to obtain the full-domain broadband near-fault ground motion; Step five: further synthesis of cross-fault ground motion: select a pulse waveform with a pulse effect and a directionality effect to replace the artificial fitted pulse, and further eliminate the sub-low frequency components in the residual component, and finally superimpose the high-frequency components to obtain the cross-fault ground motion.

2. The method of claim 1, wherein the method is a method of synthesizing near- or across-bridge structure strike-slip fault ground motion, characterized by, The wave selection principles in step one are as follows: ① The selected strong motion records are pulse records from a free-field site; ② The selected strong motion records are selected according to the fault mechanism at the bridge site and the site conditions; ③ The minimum available frequency of the selected strong motion records is less than or equal to 0.2 Hz; ④ The fault distance of the selected strong motion records is not less than 10 km, which requires that the strong motion records have the minimum velocity pulse effect before the pulse effect is superimposed; ⑤ The selected magnitude is converted from the target spectrum corresponding intensity, ranging from ± 0.5 magnitude grade; the conversion is carried out according to the magnitude and epicenter intensity relationship M W = 0.58I0+1.

5.

3. The method of claim 1, wherein the method is a method of synthesizing near- or across-bridge structure strike-slip fault ground motion, characterized by, The rotation and decomposition in step two are as follows: Linearly combine two original horizontal orthogonal components f1(t) and f2(t) to obtain parallel fault direction and vertical fault direction records, respectively: where: θ is the angle between the fault strike and the station record; f FP (t, θ), f FN (t, θ) are the corresponding record components parallel and perpendicular to the fault direction after rotating the angle θ. Use a Butterworth non-stationary filter to filter the parallel and vertical fault components; the filter frequency response function H(f) is as follows: In the formula, f is an input frequency; f c is a cutoff frequency; n is a filter order, and is 4. Filter cutoff frequency f c is calculated as follows: f c = 1 / (aT p -dt) (3) In the formula, T p is the pulse period of the seismic record; a is an empirical coefficient, and dt is the time step of the ground motion. where the pulse period T p The peak point method is used to determine the value of the time interval of the adjacent wave peak or wave trough of the peak value PGV in the velocity time history; due to the thrust effect of the strike-slip fault in the parallel fault direction and the directional effect in the vertical fault direction, 0.25 is taken for the parallel fault direction α, and 0.8 is taken for the vertical fault direction α of the seismic record.

4. The method of claim 1, wherein the method is a method of synthesizing near- or across-bridge structure strike-slip fault ground motion, characterized by, The pulse component extraction in step three is as follows: The base function expression in wavelet analysis is: In the formula, Φ(·) represents the mother wavelet function; s is the scale factor for expanding the wavelet; l is the translation factor for translating the wavelet in time; In the linear combination representing the original signal f(t), the combination coefficients C s,l may be obtained by 5. The method of claim 1, wherein the method is a method of synthesizing near- or across-bridge structure strike-slip fault ground motion, and The pulse component fitting in step three is as follows: By statistically analyzing the characteristics of ground motion from real seismic records, the low-frequency pulse velocity time history v p Analogous model; the velocity pulse function in this model is given by In the formula, V p is the pulse peak velocity; T p is the pulse period; T pk is the pulse peak time; N c is the number of pulse cycles; φ is the pulse phase angle; The Gabor wavelet model contains V p , T p , T pk , N c , and φ, a total of 5 unknown parameters, and a particle swarm optimization algorithm is introduced to determine the pulse parameters to be fitted. The fitness function is the objective function, as follows: f(t) = min | v(t) - v p (t) | (7) The velocity update formula of the dth dimension of particle i: The position update formula of the dth dimension of particle i: where: i is the particle index; d is the dimension of the space where the particle is located; k is the evolution number of the particle; is the velocity of the particle; is the best position of the particle i in the dth dimension of the space; is the position of the particle i; is the best position of all particles in the dth dimension of the space; c1, c2 are learning factors used to adjust the maximum step of learning; r1, r2 are two random numbers with a value range of [0-1]; ω is the inertia weight.

6. The method of claim 1, wherein the method is a method of synthesizing near- or across-bridge structure strike-slip fault ground motion, characterized by, The artificial fitted pulse replacement in step five is as follows: For strike-slip faults, the ground motion parallel to the fault direction has a significant pulse effect; the frequency f p is determined according to the following formula: lg(l / f p ) = -2.127 + 0.386M w (10) Ground permanent displacement D of strike-slip fault site and fault plane slip amount D fault The calculation formula is as follows: ln(D fault ) = 1.15M W -3.28 (11) where M W is the moment magnitude; a0, a2, a3 are regression parameters; R rup is the fault distance; δ is the fault dip; Ground permanent displacement D site and maximum ground displacement D max satisfies the following relationship: D max ≈D site / λ,λ∈[0.2,0.8](14)For strike-slip faults, the vertical fault direction ground motion has a significant directional effect; the peak ground velocity PGV and the pulse period T p (T p =1 / f p ) values, as shown in the following formula: lg(T p ) = -1.067 + 0.299M w (15) In (PGV) = 2.17 + 0.34M W -0.11 In R rup (16) In the formula, PGV is the peak velocity, which determines the value of the pulse amplitude V p V p may take [0.85, 1.00]PGV; R rup is the fault distance.

7. The method of synthesizing near-fault / cross-strike-slip fault ground motion for a bridge structure-oriented bridge structure according to claim 1, wherein The cross-fault ground motion synthesis in step five is as follows: For the residual component, use a high-pass filter with a cutoff frequency of 1 Hz for filtering processing to eliminate the sub-low frequency components and retain the sub-high frequency components; Superimpose the high-frequency components and the sub-high frequency components to obtain the final high-frequency peak velocity, adjust the pulse peak time of the artificial pulse function with the slide-off effect and the directionality effect to the corresponding time of the final high-frequency peak velocity, and finally superimpose the two to obtain the cross-fault ground motion in the parallel fault direction and the vertical fault direction.

Images