CN103344395A - Determination method and device of bridge reinforcing target bearing capacity - Google Patents

Abstract

The invention discloses a determination method and device of bridge reinforcing target bearing capacity, and relates to the field of safety of bridges. The determination method of bridge reinforcing target bearing capacity comprises the step of measuring the parameters of each car passing through a target bridge within a set time threshold value, the step of processing the parameters to obtain the load effect value when each car passes through the most adverse position of the target bridge, wherein the most adverse position is the position corresponding to the highest point of a load effect influence line corresponding to the target bridge, the step of processing all the load effect values to obtain a sample space satisfying generalized Paratore distribution and further determine a load effect representative value of the target bridge according to the generalized Paratore distribution rule, and the step of obtaining the target bearing capacity of reinforcing the target bridge according to the load effect representative value. The determination method of bridge reinforcing target bearing capacity can accurately deduce bridge load effect extreme values under the condition of a small calculation cost and reinforce the bridge by utilizing the bridge load effect extreme values.

Description

A kind of confirmation method and device of bridge strengthening target bearing capacity

Technical field

The present invention relates to the bridge security field, especially a kind of confirmation method and device of bridge strengthening target bearing capacity.

Background technology

Along with the fast development of China's communication, the appearance of automobile storage superimposed truck on the trend, particularly some main lines of communication that maximize has gradually caused grave danger to the safety of bridge just under arms along the line.Recent years, superimposed truck presses the accident of the bridge that collapses to occur repeatedly, for example: Heilungkiang Hulanhe River bridge in 2009, Jilin Jin Jiang River bridge in 2010, Beijing in 2011 Huairou Bai Heqiao, and Harbin Yang Mingtan bridge in 2012, collapse because overloading wagon seriously causes.People press for and know that the continuous wagon flow of dealing every day has much to the end that is applied to of bridge, administrative authority also wishes to understand under current wagon flow effect, whether bridge just under arms exists potential safety hazard in future in being on active service the phase, thereby the control measures of science are more reinforced or worked out to bridge.

To accurately answer the problems referred to above, need be to the analysis that becomes more meticulous of current wagon flow on the bridge, related work comprises:

(1) the WIM system is installed, measures current car flow information in real time, comprising: the number of axle, axle weight, distance between axles, the speed of a motor vehicle, accompany the vehicle time interval etc.

(2) wagon flow as stochastic process, the car flow information data that record are carried out statistical study, and describe its probability distribution situation with suitable probability model.

(3) simulation wagon flow stochastic process is analyzed wagon flow to the effect of bridge, obtains the structural response time-histories of bridge under the wagon flow effect.

(4) analyze the regularity of distribution of a period of time internal loading effect maximum value, and to future the peak load effect in more over a long time predict, thereby obtain the load effect typical value of bridge in certain specific assessment cycle.

Half a century in the past, the researchers of science of bridge building circle are purpose to improve the Bridge Design standard, have carried out a large amount of investigation about carload, and have all set up corresponding load standard as the case may be in each country.But what the carload standard reflected is the thing of " bulking property " of certain country carload, and how correlative study adopts relative simple loads model to represent the essential characteristic of carload if also concentrating on.But with regard to concrete a certain bridge block, the carload that it is born has its " singularity ", therefore, assesses the security of certain concrete bridge, and what should pay close attention to is the singularity of its load, but not bulking property.Therefore on science, the carload standard can not be used directly in the assessment of labour bridge, and this has reached common recognition substantially in educational circles.Need determine reasonably assessment load at concrete carload, concrete bridge.And will study the true load process of concrete certain bridge block and obtain the load effect typical value of longer assessment cycle, can only be by research actual measurement car flow information, the algorithm that exploitation more becomes more meticulous could realize, present also do not have a suitable method.

Carload effect analysis based on actual measurement wagon flow data mainly adopts the MCS method, that is: statistical informations such as the car weight that reflects according to the WIM data, vehicle commander, set up corresponding probability model, simulation wagon flow process is also calculated load effect, obtains bridge peak load effect in assessment cycle according to the statistical law of load effect then.

The MCS method also is the method the most accurately of the generally acknowledged analysis carload process of present educational circles.But, actual car weight, the probability distribution of growth are difficult to describe with an ideal model, Fig. 1 is that a multimodal distributes for the car weight statistic histogram of driving vehicle on China's certain bridge of south that provides according to correlation technique as can be seen, can not be expressed as the form of any one existing probability model.In addition, carload research the more important thing is to provide following contingent peak load of long period that this need predict that this is very difficult to the growth pattern of following load except analyzing the load that has taken place.For example: a bridge passes through 1000 cars every day, then can obtain 1000 bridge load effect datas, by selected suitable probability Distribution Model, can describe the probability distribution situation of bridge load effect at an easy rate with a probability function.As the maximal value that will predict 1 day, then need probability distribution function find 1/1000 on quantile, as the maximal value that will predict 1 year, then need find 1/365000 last quantile.On so high quantile, probability distribution function is very inaccurate, and this is referred to as " afterbody susceptibility " problem statistically.

Fig. 2 is the statistical law figure of the load effect of the described bridge of Fig. 1.In the main part of statistic histogram, the probability distribution function that curve is represented and real data are very approaching, and have arrived the high score site as can be seen, and it is very big that the two difference then becomes, and probability distribution function is no longer accurate.About the offset issue of statistical probability model at afterbody, the solution that present research is not also found.Because statistical probability model inaccurate on high branch, analyzing over a long time, the load effect extreme value has difficulties.Also Just because of this, all approximate with traffick number replacement in a day 1 year traffick number in the research of various countries' carload standard, probability model is extrapolated on the too high quantile avoiding.Yet during the load effect problem under the concrete actual wagon flow of certain military service bridge of research, it is too random that such hypothesis seems.

Cited two problems above: high branch is inaccurate during inaccurate, the load effect extrapolation of MCS method stochastic parameter model, and topmost two problems that run into when being research actual measurement carload effect also are two problems that the present invention wants the emphasis solution.

In the prior art, head it off mainly is the method for using Nowak method and MCS.

Prior art one is the Nowak method.The Nowak method is at first put forward by Nowak, the method that adopts when becoming various countries' load specification revision afterwards.Fig. 3 determines the process flow diagram of assessment load effect typical value for the Nowak method of utilizing that provides according to correlation technique.The basic way of Nowak method as shown in Figure 3.Comprise the steps:

Step 302 obtains a series of vehicle datas;

Step 304 acts on vehicle on the bridge of different spans;

Step 306, corresponding each car analysis obtains a load effect value;

Step 308 is carried out statistical study to all load effect values;

Step 310 obtains the load effect typical value.

After in the step 308 all load effect values being carried out statistical study, obtain probability distribution equation F ().(countries such as the U.S., Canada utilize normal distribution simulating vehicle load effect, and China adopts extreme value I type to distribute to simulate to the maximal value of load effect in the middle of a day).Step 310 is specially, and supposes that T is assessment cycle, N _dBe day traffick number, according to the hypothesis of " representing a year in a day ", total traffick is counted N=T * N in the T _d, determine the typical value S of load effect by 95% fraction _{T, max}For:

S _T,max＝F ^-1(0.95 ^1/N) （1）

The method has following shortcoming:

(1) one of technology is paid close attention to an independent car to the effect of bridge, can not solve many cars situation on bridge simultaneously, the therefore load analysis that is not suitable for becoming more meticulous.

(2) technology one is too simple about the hypothesis of vehicular load effect normal distribution, is theoretically unsound.

(3) technology one determines that the Extrapolation method of assessment cycle load effect typical value supposes " replacing the vehicle number in 1 year with one day vehicle number " in practice, is theoretically unsound.Relatively reasonable when the load effect of assessment long period (decades), but cycle safety more inadequately.

In order to analyze wagon flow to the load effect of certain bridge, some scholar adopts prior art two, i.e. the method for MCS with becoming more meticulous.Fig. 4 determines the process flow diagram of assessment load effect typical value for the MCS method of utilizing that provides according to correlation technique, the basic procedure of the method for MCS as shown in Figure 4, concrete steps are as follows:

Step 402 records the wagon flow data.Described data can comprise: the number of axle, heavy, the distance between axles of axle, and car weight, vehicle commander, following distance.

Step 404, statistics obtains model.Described model refers to the statistical probability distributed model of each parameter.

Step 406 will be simulated wagon flow and act on the bridge.

Step 408, the load effect value of calculating bridge.The calculating of described load effect value is to obtain by analysis of bridge structure.The vehicle number that the simulation wagon flow comprises is more many, and the load effect numerical value number of acquisition is more many.

Step 410 judges whether the simulation wagon flow reaches threshold value by the time.If, jump to step 412, if not, jump to step 406.Described threshold value can be certain fate, as one day, a week.

Step 412 is carried out statistical study to the load effect numerical value that obtains.Through statistical study, can obtain one day and even the statistical law of all peak load effects.

For assessment cycle be the situation of T because vehicle flowrate is too big, can't direct modeling, still adopt the typical value that obtains the assessment cycle load effect with the similar way of Nowak method.

The method has following shortcoming:

(1) technology two is carried out the wagon flow simulation according to the statistical law of wagon flow parameter.Because the wagon flow parameter model is more complicated often, as shown in Figure 1, cause simulation wagon flow and actual wagon flow to have deviation to a certain extent.

(2) technology two is difficult to finish macrocyclic simulation.For example, a day traffick number is 10000, if assessment cycle is 50 years, then need simulate 182500000 cars, calculates 182500000 subordinate load effects, and calculation cost is too big consequently can't to carry out.For fear of too big analog computation amount, in actual applications, taked similarly to be extrapolated to technology one method of high branch.Because the existence of " probability distribution is inaccurate on high quantile " problem, and " representing 1 year in one day " suppose unreasonable, the result of extrapolation is often not accurate enough.

Summary of the invention

The confirmation method and the device that the purpose of this invention is to provide a kind of bridge strengthening target bearing capacity, thus under the situation of less calculation cost, accurately derive the purpose that bridge load effect typical value is reinforced bridge according to described load effect typical value to reach.

For reaching this purpose, the invention provides a kind of confirmation method of bridge strengthening target bearing capacity, comprising:

Fix time in the threshold value parameter of each vehicle that passes through from the target bridge of measuring gage;

Handle described parameter, the load effect value when obtaining the least favorable position of each vehicle by described target bridge, wherein, the least favorable position refers to that the load effect of described target bridge correspondence influences the corresponding position of peak of line;

Handle all described load effect values, obtain obeying the sample space of broad sense para holder distribution (GPD distribution), and then determine the load effect typical value of described target bridge according to the rule that the holder of broad sense para distributes;

Draw the target bearing capacity that the described target bridge of reply is reinforced according to described load effect typical value.

Preferably, the measurement mechanism of described parameter is dynamic weighing system.

Preferably, described parameter comprises: vehicle pass-through date, vehicle pass-through time, track, vehicle place, axletree number, each axle weight, distance between axles, first-to-last of axle dimension, gross combination weight and Vehicle Speed.

Preferably, the step that obtains obeying the sample space that the holder of broad sense para distributes comprises:

Get all described load effect values as sample;

In 0 described sample of every N continuous, get maximum sample;

With all described maximum sample descending sorts, form sample space { x ₁, x ₂..., x _n, wherein, n is the sample number of sample space, x ₁, x ₂..., x _nSample for described sample space.

Preferably, the step of determining the load effect typical value is specially:

A, the amount of fetching data are k, and k is the positive integer less than n-1; Get u=x _K+1Construct k data point

K data point carried out linear fitting, obtain slope σ;

B, to the cumulative distribution function of sample space

$F\left(x\right)=\left\{\begin{array}{cc}\frac{n-k}{n+1}& x\≤u\\ \frac{n-k}{n+1}+\frac{k+1}{n+1}[1-{(1+\mathrm{\ξ}\frac{x-u}{\mathrm{\σ}})}^{-1/\mathrm{\ξ}}]& x>u\end{array}\right.$ Carry out the KS check, wherein, the KS inspected number is: $D(u,\mathrm{\ξ},\mathrm{\σ})=\underset{l\≤i\≤k}{\max }\left(\right|\frac{n+1-i}{n+1}-F\left(x_i\right)\left|\right);$

C, with k, ξ continuous transformation value on " k-ξ " space, repeat the process of a, b, obtain to make on described " k-ξ " space k and the ξ of described KS inspected number minimum to be designated as k respectively ^*And ξ ^*, and obtain and k ^*And ξ ^*Corresponding u and σ are designated as u ^*, σ ^*

D, the load effect typical value of getting described target bridge are

Wherein, N is per day traffick number, and T represents that the described predetermined estimation cycle is T, at this moment k, ξ, σ, u difference value k ^*, ξ ^*, u ^*, σ ^*

Preferably, described " k-ξ " space is space { 20＜k＜n/5 ,-0.2＜ξ＜0.2}.

Preferably, described method also comprises:

According to following formula

Calculate the overload coefficient of described target bridge; Wherein, S _CodeCarload standard value for the target bridge.

Preferably, described N0 is the integer more than or equal to 20.

The invention also discloses a kind of affirmation device of bridge strengthening target bearing capacity, comprising:

The parameter measurement device is used for fix time in the threshold value parameter of each vehicle that passes through from the target bridge of measuring gage;

Parameter Processor, for the treatment of described parameter, the load effect value when obtaining the least favorable position of each vehicle by described target bridge, wherein, the least favorable position refers to that the load effect of described target bridge correspondence influences the corresponding position of peak of line; And handle all described load effect values, obtain obeying the sample space that the holder of broad sense para distributes;

Counter is used for utilizing described sample space to determine the load effect typical value of described target bridge;

Target bearing capacity fallout predictor is used for doping the target bearing capacity that reply target bridge is reinforced according to described load effect typical value.

Preferably, described parameter Processor is used for: get all described load effect values as sample, get maximum sample in 0 described sample of every N continuous, with all described maximum sample descending sorts, form sample space { x ₁, x ₂..., x _n, wherein, n is the sample number of sample space, x ₁, x ₂..., x _nSample for described sample space.

Compared with prior art, the present invention has following technique effect at least:

(1) adopt actual measurement wagon flow data to carry out the structural load effect analysis.Both avoid load effect of the prior art to calculate too coarse problem, also avoided simulation wagon flow and the inconsistent problem of actual conditions in the prior art.Can accurately estimate to make the safe class of scientifically assessing bridge accordingly in the true load situation of labour bridge, in time find potential safety hazard, bridge is reinforced, avoid the unnecessary wasting of resources again.

(2) The extreme value distribution of employing GPD model description vehicular load effect.The variation of GPD model by afterbody parameter ξ reaches the best-fit result to sample data, avoided in the prior art load effect probability distribution inaccurate, and prior art determines that by extrapolating the load effect extreme value causes inaccurate problem.

(3) utilize the GPD model that the extreme value of load effect is predicted by the method for extrapolation, do not need " hypothesis that the day car flow represents 1 year vehicle flowrate ", overcome the theoretical incomplete problem of prior art, also overcome the problem that the simulation number of times was too many when prior art was longer in assessment cycle, calculation cost is excessive simultaneously.

(4) be that { 20＜k＜n/5 ,-0.2＜ξ＜0.2} have significantly reduced operand, have shortened operation time in the space with " k-ξ " space boundary.

(5) the overload coefficient of the present invention's proposition is the parameter of the most direct reaction overload situation, and the numerical value of overload coefficient is more big, and the representative overload is more serious.Administrative authority can be according to the overload coefficient of every bridge block, thereby works out the more decision-making of the improvement overload of science.

Above-mentioned explanation only is the general introduction of technical solution of the present invention, for can clearer understanding technological means of the present invention, and can be implemented according to the content of instructions, and for above and other objects of the present invention, feature and advantage can be become apparent, below especially exemplified by the specific embodiment of the present invention, and conjunction with figs., be described in detail as follows.

Description of drawings

The car weight statistic histogram of driving vehicle on China's certain bridge of south that provides according to correlation technique is provided Fig. 1;

Fig. 2 is the statistical law figure of the load effect of the described bridge of Fig. 1;

Fig. 3 determines the process flow diagram of assessment load effect typical value for the Nowak method of utilizing that provides according to correlation technique;

Fig. 4 determines the process flow diagram of assessment load effect typical value for the MCS method of utilizing that provides according to correlation technique;

The process flow diagram of the confirmation method of the bridge strengthening position that provides according to one embodiment of the invention is provided Fig. 5;

The data statistic of Fig. 6 after for the WIM systematic survey vehicle that provides according to one embodiment of the invention;

Fig. 7 is that the mid span moment of the free beam of L influences the line synoptic diagram for the span that provides according to one embodiment of the invention;

Fig. 8 influences the synoptic diagram that line calculates the load effect of vehicle for what provide according to one embodiment of the invention according to load effect;

Fig. 9 for provide according to one embodiment of the invention according to 1 day the comparison diagram with actual 14 day data computation structures of predicting the outcome;

The load extreme value order statistics feature synoptic diagram of Figure 10 for providing according to one embodiment of the invention;

The KS inspected number that provides according to the one embodiment of the invention variation synoptic diagram with tail data amount and afterbody parameter is provided Figure 11;

The synoptic diagram that concerns of assessment cycle of providing according to one embodiment of the invention and load effect typical value is provided Figure 12;

The synoptic diagram that concerns of the overload coefficient that provides according to one embodiment of the invention and assessment cycle is provided Figure 13;

Synoptic diagram when Figure 14 is respectively exponential distribution, long-tail distribution, truncation distribution for the GPD that provides according to one embodiment of the invention;

The vehicular load effect stochastic process synoptic diagram of Figure 15 for providing according to one embodiment of the invention;

Figure 16 is the structured flowchart of affirmation device that bridge strengthening target bearing capacity is provided according to one embodiment of the invention.

Embodiment

Below in conjunction with accompanying drawing the specific embodiment of the present invention is elaborated.

The process flow diagram of the confirmation method of the bridge strengthening position that provides according to one embodiment of the invention is provided Fig. 5.In the corresponding embodiment of Fig. 5, method of the present invention can comprise that step 502 is to step 510.

Fix time in the threshold value parameter of each vehicle that passes through from the target bridge of step 502, measuring gage;

Step 504 is handled described parameter, the load effect value when obtaining the least favorable position of each vehicle by described target bridge, and wherein, the least favorable position refers to that the load effect of described target bridge correspondence influences the corresponding position of peak of line;

Step 506 is handled all described load effect values, obtains obeying the sample space that the holder of broad sense para distributes;

Step 508, the rule that distributes according to the holder of broad sense para is determined the load effect typical value of described target bridge;

Step 510 draws the target bearing capacity that the described target bridge of reply is reinforced according to described load effect typical value.。

Wherein, described target bearing capacity comprises maximum shear, maximal bending moment on bridge pier greatest axis power, the beam.

In step 502, can measure the parameter of the vehicle that from the target bridge, passes through by dynamic weighing system (WIM system).Kilogram), distance between axles (unit: mm), first-to-last of axle dimension, gross combination weight and the speed of a motor vehicle (unit: kilometer/hour) parameter of WIM systematic survey can comprise the heavy (unit: of date, time, track, vehicle place, axletree number, each axle as shown in Figure 6 at least.

In the corresponding embodiment of Fig. 6, can from the data that WIM measures, extract the concrete parameter information of each car, and calculate following distance (S), that is: S=Δ T * V by the accompany two car time intervals (Δ T) and the front truck speed of a motor vehicle (V).Like this, obtain the analysis that actual wagon flow is used for the vehicular load effect.

The load effect analysis of certain concrete bridge need influence line by means of the load effect of this bridge under actual wagon flow effect.Load effect influences line and represents when the unit concentrated force acts on certain position of bridge the size of load effect.Load effect influences the feature that line has reflected bridge itself, and it describes the process that load effect changes with power effect position.Fig. 7 represents that namely span is that the mid span moment of the free beam of L influences line, and as can be seen, when the unit concentrated force was moved by the left end span centre of beam, mid span moment was increased to L/4 by 0 linearity; When moving to right-hand member by span centre, mid span moment is reduced to 0 by the L/4 linearity.

The actual wagon flow of WIM systematic survey acted on to be influenced on the line.As shown in Figure 8, simultaneously have two cars on the bridge this moment, and the axle of 5 axles heavily is respectively F ₁～F ₅, the line numerical value that influences of 5 axle position correspondences is respectively I ₁～I ₅, the load effect value Z that causes then can calculate by following formula:

$Z=\underset{i=1}{\overset{5}{\mathrm{\Σ}}}{F}_i{\×I}_i---\left(2\right)$

In actual computation, each car when namely influencing the peak position of line among Fig. 8, calculates peak load effect value Z by influencing the least favorable position of line one time.Therefore, the WIM system log (SYSLOG) data of what cars, just can calculate what peak load effect value.

Said process is the refinement of step 502, has realized by utilizing described dynamic weighing system WIM to measure in the default time threshold information data of each vehicle that passes through from described target bridge.Subsequently, add up and handle described information data, the process of the load effect value when obtaining the least favorable position of each vehicle by described target bridge.

Embodiment one

In one embodiment of the invention, record 14 days vehicle parameter, utilize step 506,508 according to first day vehicle parameter 14 days carload effect to be predicted, and the value that will predict and the actual value that records compare, find that the result is close substantially.It is very little that this explanation utilizes step 506,508 to try to achieve the error of the peak load effect value (load effect typical value) in the longer cycle, the accuracy rate height.Detailed process is as follows.

The simply supported girder bridge of certain 60m span has the massive overload car to pass through through preliminary observation, needs to calculate the carload effect, and then the security of structure is carried out comprehensive assessment, bridge is reinforced.In this embodiment, after the WIM system had surveyed 14 days traffick data, totally 697233 cars passed through, and a traffick number was 49802 in average day.

Utilize first day data (data of 49802 automobiles) can calculate each car from bridge by the time the peak load effect value, thereby obtain 49802 peak load effect value; In these 49802 peak load effect value, per 20 maximal values that (being N0=20) gets, the maximal value reorganization with getting obtains 2990 sample values (being n=2990) altogether; With these 2990 sample value descending sorts, form the sample space that comprises n sample.N0 can also value be the integer greater than 20, this is because when the NO value is greater than 20 integer, the correlativity of the load effect that the interrelation two cars produces is less, and sample is more near independent same distribution, thereby the load effect typical value of utilizing step 506,508 to calculate is more accurate.

Said process has been set up the GPD model relevant with all described load effect values, can try to achieve the load effect typical value by following step a, b, c, d.

A, the amount of fetching data are k, and k is the positive integer less than n-1; Get u=x _K+1Construct k data point

K data point carried out linear fitting, obtain slope σ;

B, to the cumulative distribution function of sample space

$F\left(x\right)=\left\{\begin{array}{cc}\frac{n-k}{n+1}& x\≤u\\ \frac{n-k}{n+1}+\frac{k+1}{n+1}[1-{(1+\mathrm{\ξ}\frac{x-u}{\mathrm{\σ}})}^{-1/\mathrm{\ξ}}]& x>u\end{array}\right.$ Carry out the KS check, wherein, the KS inspected number is: $D(u,\mathrm{\ξ},\mathrm{\σ})=\underset{l\≤i\≤k}{\max }\left(\right|\frac{n+1-i}{n+1}-F\left(x_i\right)\left|\right);$

C, with k, ξ continuous transformation value on " k-ξ " space, repeat the process of a, b, obtain to make on described " k-ξ " space k and the ξ of described KS inspected number minimum to be designated as k respectively ^*And ξ ^*, and obtain and k ^*And ξ ^*Corresponding u and σ are designated as u ^*, σ ^*

" k-ξ " space can be limited to space { 20＜k＜n/5 ,-0.2＜ξ＜0.2}.According to early stage a large amount of load analyses and data accumulation can draw, will " k-ξ " space boundary in that { 20＜k＜n/5 ,-0.2＜ξ＜0.2} can significantly reduce operand, shortening operation time.Only be a preferable range herein, do not cause restriction for the concrete scope in " k-ξ " space.

D, with k ^*, ξ ^*, u ^*, σ ^*Substitution $F\left(x\right)=\left\{\begin{array}{cc}\frac{n-k}{n+1}& x\≤u\\ \frac{n-k}{n+1}+\frac{k+1}{n+1}[1-{(1+\mathrm{\ξ}\frac{x-u}{\mathrm{\σ}})}^{-1/\mathrm{\ξ}}]& x>u\end{array}\right..$

Carry out finding behind the space search according to said process, when tail data amount k=425, obtain optimum GPD model parameter value, be respectively: u=9738kN-m, ξ=-0.054, σ=2010kN-m.Therefore afterbody GPD probability distribution function expression formula is:

$F\left(x\right)=\frac{2990-425}{2991}+\frac{426}{2991}(1-{(1-0.054\frac{x-9738}{2010})}^{1/0.054})x>9738\mathrm{kNm}---\left(3\right)$

Formula (3) is utilized 1 day wagon flow data, has provided every N ₀(N ₀The cumulative probability function of the peak load effect when=20) car is by bridge.

The validity of following verification expression (3).Total N(N=697233 in 14 days) car passes through, and the load effect of suppose generation is arranged from big to small and is respectively S ₁, S ₂, S ₃According to extreme value theory, S _iCan calculate by following formula:

$S_i=F^{-1}\left(\frac{i}{N/{\mathrm{<figure-callout\; id="0"\; label="N"\; filenames="HDA00003477183400011.png,HDA00003477183400012.png"\; state="\{\{state\}\}">N</figure-callout>}}_0+1}\right)---\left(4\right)$

Wherein F () provides in formula (3).

Calculate S according to formula (4) _iValue, and compare with the result who calculates according to measured data.Fig. 9 for provide according to one embodiment of the invention according to 1 day the comparison diagram with actual 14 day data computation structures of predicting the outcome, as shown in Figure 9, Fig. 9 has compared the situation of maximum preceding 50 load effects.As can be seen, the load effect value of prediction and the actual load effect value that records are very approaching, show: the method for this patent can be predicted the peak load effect of bridge in longer cycle exactly according to the wagon flow situation of short period.

Embodiment two

In another embodiment of the present invention, calculate by 14 days the vehicle parameter that the WIM system is recorded, obtain T peak load effect eigenwert (load effect typical value), thereby can utilize the load effect typical value to determine the position that reinforce bridge.In the present embodiment, also proposed overload coefficient η, administrative authority can be according to the overload coefficient of every bridge block, thereby works out the more decision-making of the improvement overload of science.

After 14 days the wagon flow data that the WIM system is recorded were handled, step was as follows:

(1) load effect calculates and samples resemble

At first according to influence line computation go out each car from bridge by the time maximum mid span moment, (be N by per 50 then ₀=50) mid span moment maximal value reorganization sample obtains 13944 sample values (being n=13944) altogether.Figure 10 has provided descending order sample value (x ₁ ^*..., x ₅₀ ^*) and order (k) equation

Between relation.If be straight line, show ξ=0; If fovea superior shows ξ〉0; Otherwise ξ＜0.As can be seen, the main body of load effect is straight line, i.e. ξ=0 is that extreme value I type distributes, but has arrived the place of high tail, and ξ＜0 has demonstrated the characteristics of truncation.In addition, the load effect that calculates according to national carload standard " highway bridge and culvert design general specification, JTGD60-2004 " only is 18125kN-m, and the automobile overload of this bridge is comparatively serious as can be seen.

(2) determine the GPD model, describe the afterbody distribution characteristics of load effect extreme value

{ space search is carried out in 20＜k＜2788 in-0.2＜ξ＜0.2} in " k-ξ " space.Search Results as shown in figure 11, when k=1121, ξ=-0.071, D (1121 ,-0.071)=0.0081 is minimum value.So think that tail data amount optimal value is 1211, afterbody sample optimum shape parameter is-0.071; Corresponding threshold values u=10051kN-m, and calculate σ=2466kN-m according to formula (6).Therefore the afterbody GPD distribution function expression formula of match is:

$F\left(x\right)=\frac{13944-1121}{13945}+\frac{1122}{13945}(1-{(1-0.071\frac{x-10051}{2466})}^{1/0.071})x>10051\mathrm{kNm}---\left(5\right)$

(3) assessment cycle load effect typical value and the determining of overload coefficient

Formula (5) has provided every N ₀The car peak load effect during by bridge (is used S _{1, max}Expression) cumulative probability function.The assessment cycle of supposing this bridge is T, and a day traffick number is N, then the assessment phase total (365 * T * N) car passes through, during the peak load effect (use S _{T, max}Expression) cumulative probability function F _T(x), with S _{1, max}Cumulative probability function F (x) following relation is arranged:

$F_T\left(S_{T,\max }\right)={\left(F\left(S_{l,\max }\right)\right)}^{365\&times;T\&times;\frac{N}{N_0}}---\left(6\right)$

Get F in the standard _T(x)=0.95 the value of correspondence is the load effect eigenwert, so S _{T, max}Can obtain by following formula:

$S_{T,\max }={F_T}^{-1}\left(0.95\right)=F^{-1}\left({0.95}^{\frac{{\mathrm{<figure-callout\; id="0"\; label="N"\; filenames="HDA00003477183400011.png,HDA00003477183400012.png"\; state="\{\{state\}\}">N</figure-callout>}}_0}{365\&times;T\&times;N}}\right)---\left(7\right)$

Said method had both avoided load effect of the prior art to calculate too coarse problem, had also avoided simulation wagon flow and the inconsistent problem of actual conditions in the prior art.Can accurately estimate to make the safe class of scientifically assessing bridge accordingly in the true load situation of labour bridge, in time find potential safety hazard, avoid the unnecessary wasting of resources again.

F () provides in formula (5), so T peak load effect eigenwert can calculate, as shown in figure 12.If this bridge is assessed or reinforced, should carry out according to the load effect among Figure 12.

The mid span moment that utilizes the carload standard to calculate is 18125kN-m.If will carry out more in all directions assessment to bridge, need revise the standard load.Definition overload coefficient (η) is: the ratio between the load effect typical value of inferring according to the actual measurement wagon flow and the load effect value that calculates according to standard.That is:

$\mathrm{\&eta;}=\frac{S_{T,\max }}{S_{\mathrm{code}}}$

Wherein, S _{T, max}Be the load effect typical value that the actual measurement wagon flow is inferred, S _CodeFor standard is calculated the load effect value.

The numerical value of overload coefficient is more big, and the representative overload is more serious.Overload coefficient and the relation of assessment cycle are as shown in figure 13.Administrative authority can be according to the overload coefficient of every bridge block, thereby works out the more decision-making of the improvement overload of science.

(4) to bridge strengthening target bearing capacity determination

The overload coefficient has been arranged, just can revise the standard load easily, bridge has comprehensively been checked assessment and Scheme of Strengthening design.

In the present embodiment, a simply supported girder bridge, span is 20 meters, its expection service life is 20 years.Be 1.9 according to the analysis of the front coefficient that obtains overloading.If this bridge is reinforced its bridge pier greatest axis power (N _Max), maximum shear (Q on the beam _Max), maximal bending moment (M _Max) all be conventional load calculated value (N _Design, Q _Design, M _Design) 1.9 times, that is:

N _Max=1.9 * N _Design=1.9 * 225=427.5kN (9a)

Q _Max=1.9 * Q _Design=1.9 * 225=427.5kN (9b)

M _Max=1.9 * M _Design=1.9 * 1725=3277.5kN-m (9c)

Wherein, conventional load calculated amount N _Design, Q _Design, M _DesignRefer to maximum shear, maximal bending moment on the bridge pier greatest axis power, beam under the conventional design respectively, its numerical value calculates according to the load standard in the Bridge Design standard at present.

The parameter that the embodiment of the invention provides is handled (referring to step 506) and subsequent calculations step (referring to step 508) relates to a large amount of formula and calculating, and existing derivation and corresponding principle with two treatment steps describes, and be specific as follows.

One, at first to be appreciated that the theory relevant with the GPD model.

Extreme value theory is thought, supposes to have with distributing and n separate random sample { X ₁, X ₂..., X _n, its maximum value is Z=max{X ₁, X ₂..., Xn} when n is enough big, exists threshold values u and parameter a, makes

Probability distribution must belong in following three kinds of functions one:

Ⅰ: $F\left(z\right)=\exp \{-\exp [-\left(\frac{z-u}{a}\right)\left]\right\},-\&infin;<z<\&infin;---\left(10a\right)$

Ⅱ: $F\left(z\right)=\left\{\begin{array}{cc}0,& z\&le;a\\ \exp [-{\left(\frac{z-u}{a}\right)}^{-\mathrm{\&alpha;}}],& z>a\end{array}\right.---\left(10b\right)$

Ⅲ: $F\left(z\right)=\left\{\begin{array}{cc}\exp \{-\exp [-{\left(\frac{z-u}{a}\right)}^{\mathrm{\&alpha;}}\left]\right\},& z<a\\ 1,& z\&GreaterEqual;a\end{array}\right.---\left(10c\right)$

Function I, II, III are respectively extreme value I type, extreme value II type and extreme value III type and distribute.

When the u in the formula (10) is enough big, at X〉under the situation of u, make Y=(Z-u), above-mentioned three kinds of Z may distribute and can distribute to describe with GPD uniformly:

$F\left(y\right)=\mathrm{Pr}(z-u<y|z>u)==\left\{\begin{array}{cc}1-{(1+\mathrm{\&xi;}\frac{y}{\mathrm{\&sigma;}})}^{-1/\mathrm{\&xi;}},& \mathrm{\&xi;}\&NotEqual;0\\ 1-\exp (-\frac{y}{\mathrm{\&sigma;}}),& \mathrm{\&xi;}=0\end{array}\right.---\left(11\right)$

ξ is called the afterbody parameter, and u is called threshold values.ξ=0 an o'clock GPD deteriorates to exponential distribution; ξ＞0 o'clock GPD is that long-tail distributes; ξ＜0 o'clock, GPD is that truncation distributes.Synoptic diagram when Figure 14 is respectively exponential distribution, long-tail distribution, truncation distribution for the GPD that provides according to one embodiment of the invention.The synoptic diagram of three kinds of distributions as shown in figure 14.

Find to have only number as extreme value sample Z to reach 1000 when above through experiment, could infer 3 parameters (ξ, u and σ) of GPD model more exactly.Study for vehicular load, if Z represents the extreme value of one day load effect, need 1000 days, millions of times carload analysis so, this is impossible mission at present.

Two, understood the theory relevant with the GPD model after, the POT theory that extreme value is arranged that also will use.

According to the stability theory of POT, for independent identically distributed sample, as long as the threshold value u that provides is enough big, the part that exceeds threshold value in the sample is all obeyed the GPD distribution.

The vehicular load effect is a stochastic process, the vehicular load effect stochastic process synoptic diagram of Figure 15 for providing according to one embodiment of the invention.In the section T, the load effect that wagon flow produces can be represented with stochastic process curve x (t) at a time.And the data of physical record are the data acquisition { X}={x after dispersing by the time ₁, x ₂..., x _n, correspond respectively to t ₁, t ₂..., t _nConstantly, the value x (t of stochastic process curve ₁), x (t ₂) ..., x (t _n).

If will use the POT theory, must guarantee that sample is independent identically distributed.Because the vehicular load effect has comprised the contribution of all cars on the bridge, two samples that the two cars gap bridge that therefore accompanies produces have bigger correlativity.In order to obtain independent identically distributed sample, the maximum value in selected some (N0) samples is recombinated to sample, for example: i sample y after the reorganization _iBe expressed as:

$y_i=\max \{x_{(i-1){\mathrm{<figure-callout\; id="0"\; label="N"\; filenames="HDA00003477183400011.png,HDA00003477183400012.png"\; state="\{\{state\}\}">N</figure-callout>}}_0+1},x_{(i-1){\mathrm{<figure-callout\; id="0"\; label="N"\; filenames="HDA00003477183400011.png,HDA00003477183400012.png"\; state="\{\{state\}\}">N</figure-callout>}}_0+2},......x_{(i-1){\mathrm{<figure-callout\; id="0"\; label="N"\; filenames="HDA00003477183400011.png,HDA00003477183400012.png"\; state="\{\{state\}\}">N</figure-callout>}}_0+N_0}\}$

That is: (i-1) N in the original sample ₀+ 1 to iN ₀Maximal value in the individual sample.

Get N respectively ₀Be 1,10,20,50,100,1000,10000, the load effect sample that typical wagon flow is generated carries out correlation analysis.y _iWith y _I+1Related coefficient see the following form.N in the table ₀=1 not secondary reorganization of expression sample.

Data in the table show: N ₀=1 o'clock, related coefficient showed that up to 0.371 the load effect that the interrelation two cars produces has bigger correlativity; N ₀Be 20 o'clock, related coefficient is reduced to 0.0367 rapidly, afterwards along with N ₀Increase, related coefficient continue to reduce, but obvious during not as beginning.As can be seen from Table 1, N ₀The value minimum can get 20.

Three, understood the POT theory of the theory relevant with the GPD model, extreme value after, just can carry out determining based on the GPD model parameter of trailing space search.

According to the POT theory, obtain after the independent identically distributed load effect sample, just can carry out the parameter of GPD model and determine.The GPD model comprises 3 parameter: u, ξ and σ, and in financial industry, the HILL estimation technique commonly used is determined this three parameters at present.But the HILL estimation technique is more accurate greater than 0.5 o'clock at ξ, and approaches or no longer suitable less than 0 the time as ξ.The ξ value of vehicular load effect correspondence is often near 0, so the HILL method is inapplicable.Hosking and Wallis have also proposed method, regularly can be used for determining ξ and σ when threshold values u gives, but the threshold values u of most critical is not but had good definite way.Therefore at this particular random process of carload effect, need the new GPD parameter determination method of research.

According to above-mentioned theory, can draw the GPD model parameter based on the trailing space search of the present invention and determine method, concrete steps are as follows:

(1) the carload effect sample that obtains is recombinated, in every N continuous ₀Get maximal value in the individual sample as new sample, obtain new sample space { X}={x ₁, x ₂, x ₃... x _n, n is reorganization rear space sample number.Suggestion N ₀Value be at least 20.

(2) { all samples of X} sort, and { X} is converted to new { X to sample space to the space ^*}={ x ₁ ^*, x ₂ ^*X _n ^*, and x ₁ ^*X ₂ ^*... x _n ^*

(3) setting the tail data amount is k, threshold values u=x _K+1 ^*With sample space { X ^*Preceding k value is to the plussage formation sample set { x of its k+1 value ₁ ^*-x _K+1 ^*, x ₂ ^*-x _K+1 ^*... x _k ^*-x _K+1 ^*, be called afterbody POT sample.Determining of tail data amount is very crucial, and the excessive or too small model that all can cause departs from actual conditions, advises that actual value is between 20 to n/5.

(4) setting form parameter is ξ, constructs k data point, shown in (12)

$\left\{\frac{1}{\mathrm{\&xi;}}\right[{\left(\frac{j}{k+1}\right)}^{-\mathrm{\&xi;}}-1],x_j^{*}\},j=\mathrm{1,2},...,k---\left(12\right)$

K data point in the formula (12) carried out linear fitting, and slope is σ.

At this moment, the cumulative probability function of GPD model note is done:

$F\left(x\right)=\left\{\begin{array}{cc}\frac{n-k}{n+1}& x\&le;u\\ \frac{n-k}{n+1}+\frac{k+1}{n+1}[1-{(1+\mathrm{\&xi;}\frac{x-u}{\mathrm{\&sigma;}})}^{-1/\mathrm{\&xi;}}]& x>u\end{array}\right.---\left(13\right)$

(5) the described GPD model of formula (13) is carried out the KS check, KS inspected number D (u, ξ σ) are expressed as:

$D(u,\mathrm{\&xi;},\mathrm{\&sigma;})=\underset{l\&le;i\&le;k}{\max }\left(\right|\frac{n+1-i}{n+1}-F\left({x_i}^{*}\right)\left|\right)---\left(14\right)$

(6) with k from 20 to n/4 continuous transformations, with ξ from-0.2 to 0.2 continuous transformation, repeat above-mentioned steps (3)～(5), at last with make on the whole space KS inspected number D minimum (u, ξ σ) are last GPD parameter, that is:

$(u^{*},{\mathrm{\&xi;}}^{*},{\mathrm{\&sigma;}}^{*})=\{u,\mathrm{\&xi;},\mathrm{\&sigma;}|\underset{-0.2\&le;\mathrm{\&xi;}\&le;0.2}{\underset{20\&le;k\&le;n/5}{\mathrm{<figure-callout\; id="20"\; label="min"\; filenames="HDA00003477183400062.png,HDA00003477183400071.png"\; state="\{\{state\}\}">min</figure-callout>}}}D(u,\mathrm{\&xi;},\mathrm{\&sigma;})\}---\left(15\right)$

Through above-mentioned steps, can access the cumulative distribution function of GPD model, can obtain the load effect typical value by formula (6) (7) again, provide reference for the engineering staff assesses the bridge tolerance degree, the engineering staff can reinforce bridge according to this load effect typical value.

It is to be noted: { 20＜k＜n/5 ,-0.2＜ξ＜0.2} are the scopes that provides according to the experience that early stage, a large amount of load analyses accumulated in the search volume.The search volume is limited in more among a small circle can significantly reduces operand.

The invention also discloses a kind of affirmation device of bridge strengthening target bearing capacity.In the corresponding embodiment of Figure 16, device of the present invention can comprise:

Parameter measurement device 162 is used for fix time in the threshold value parameter of each vehicle that passes through from the target bridge of measuring gage;

Parameter Processor 164, for the treatment of described parameter, the load effect value when obtaining the least favorable position of each vehicle by described target bridge, wherein, the least favorable position refers to that the load effect of described target bridge correspondence influences the corresponding position of peak of line; And handle all described load effect values, obtain obeying the sample space that the holder of broad sense para distributes;

Counter 166 is used for utilizing described sample space to determine the load effect typical value of described target bridge;

Target bearing capacity fallout predictor 168 is used for doping the target bearing capacity that reply target bridge is reinforced according to described load effect typical value.

In one embodiment of the invention, described parameter Processor 164 can be used for: get all described load effect values as sample, get maximum sample in 0 described sample of every N continuous, with all described maximum sample descending sorts, form sample space { x ₁, x ₂..., x _n, wherein, n is the sample number of sample space, x ₁, x ₂..., x _nSample for described sample space.

Technique scheme has following technique effect at least:

(1) adopt actual measurement wagon flow data to carry out the structural load effect analysis.Both avoid load effect of the prior art to calculate too coarse problem, also avoided simulation wagon flow and the inconsistent problem of actual conditions in the prior art.Can accurately estimate to make the safe class of scientifically assessing bridge accordingly in the true load situation of labour bridge, in time find potential safety hazard, bridge is reinforced, avoid the unnecessary wasting of resources again.

(2) The extreme value distribution of employing GPD model description vehicular load effect.The variation of GPD model by afterbody parameter ξ reaches the best-fit result to sample data, avoided in the prior art load effect probability distribution inaccurate, and prior art determines that by extrapolating the load effect extreme value causes inaccurate problem.

(3) utilize the GPD model that the extreme value of load effect is predicted by the method for extrapolation, do not need " hypothesis that the day car flow represents 1 year vehicle flowrate ", overcome the theoretical incomplete problem of prior art, also overcome the problem that the simulation number of times was too many when prior art was longer in assessment cycle, calculation cost is excessive simultaneously.

(4) be that { 20＜k＜n/5 ,-0.2＜ξ＜0.2} have significantly reduced operand, have shortened operation time in the space with " k-ξ " space boundary.

(5) the overload coefficient of the present invention's proposition is the parameter of the most direct reaction overload situation, and the numerical value of overload coefficient is more big, and the representative overload is more serious.Administrative authority can be according to the overload coefficient of every bridge block, thereby works out the more decision-making of the improvement overload of science.

By reference to the accompanying drawings the present invention has been carried out exemplary description above; obviously specific implementation of the present invention is not subjected to the restriction of aforesaid way; as long as the various improvement of having adopted method design of the present invention and technical scheme to carry out; or directly apply to other occasion without improvement, all within protection scope of the present invention.

Claims (10)

1. the confirmation method of a bridge strengthening target bearing capacity is characterized in that, comprising:

Fix time in the threshold value parameter of each vehicle that passes through from the target bridge of measuring gage;

Handle described parameter, the load effect value when obtaining the least favorable position of each vehicle by described target bridge, wherein, the least favorable position refers to that the load effect of described target bridge correspondence influences the corresponding position of peak of line;

Handle all described load effect values, obtain obeying the sample space that the holder of broad sense para distributes, and then determine the load effect typical value of described target bridge according to the rule that the holder of broad sense para distributes;

Draw the target bearing capacity that the described target bridge of reply is reinforced according to described load effect typical value.

2. the method for claim 1 is characterized in that, the measurement mechanism of described parameter is dynamic weighing system.

3. method as claimed in claim 1 or 2 is characterized in that, described parameter comprises: vehicle pass-through date, vehicle pass-through time, track, vehicle place, axletree number, each axle weight, distance between axles, first-to-last of axle dimension, gross combination weight and Vehicle Speed.

4. as any described method in the claim 1 to 3, it is characterized in that the step that obtains obeying the sample space that the holder of broad sense para distributes comprises:

Get all described load effect values as sample;

In 0 described sample of every N continuous, get maximum sample;

With all described maximum sample descending sorts, form sample space { x ₁, x ₂..., x _n, wherein, n is the sample number of sample space, x ₁, x ₂..., x _nSample for described sample space.

5. as any described method in the claim 1 to 4, it is characterized in that, determine that the step of load effect typical value is specially:

A, the amount of fetching data are k, and k is the positive integer less than n-1; Get u=x _K+1Construct k data point

K data point carried out linear fitting, obtain slope σ;

B, to the cumulative distribution function of sample space

$F\left(x\right)=\left\{\begin{array}{cc}\frac{n-k}{n+1}& x\&le;u\\ \frac{n-k}{n+1}+\frac{k+1}{n+1}[1-{(1+\mathrm{\&xi;}\frac{x-u}{\mathrm{\&sigma;}})}^{-1/\mathrm{\&xi;}}]& x>u\end{array}\right.$ Carry out the KS check, wherein, the KS inspected number is: $D(u,\mathrm{\&xi;},\mathrm{\&sigma;})=\underset{1\&le;i\&le;k}{\max }\left(\right|\frac{n+1-i}{n+1}-F\left(x_i\right)\left|\right);$

C, with k, ξ continuous transformation value on " k-ξ " space, repeat the process of a, b, obtain to make on described " k-ξ " space k and the ξ of described KS inspected number minimum to be designated as k respectively ^*And ξ ^*, and obtain and k ^*And ξ ^*Corresponding u and σ are designated as u ^*, σ ^*

D, the load effect typical value of getting described target bridge are

Wherein, N is per day traffick number, and T represents that the described predetermined estimation cycle is T, at this moment k, ξ, σ, u difference value k ^*, ξ ^*, u ^*, σ ^*

6. method as claimed in claim 5 is characterized in that, described " k-ξ " space is space { 20＜k＜n/5 ,-0.2＜ξ＜0.2}.

7. the method described in claim 5 or 6 is characterized in that, also comprises:

According to following formula

Calculate the overload coefficient of described target bridge; Wherein, S _CodeCarload standard value for the target bridge.

8. as any described method in the claim 4 to 8, it is characterized in that described N0 is the integer more than or equal to 20.

9. the affirmation device of a bridge strengthening target bearing capacity is characterized in that, comprises

Parameter measurement device (162) is used for fix time in the threshold value parameter of each vehicle that passes through from the target bridge of measuring gage;

Parameter Processor (164), for the treatment of described parameter, load effect value when obtaining the least favorable position of each vehicle by described target bridge, wherein, the least favorable position refers to that the load effect of described target bridge correspondence influences the corresponding position of peak of line; And handle all described load effect values, obtain obeying the sample space that the holder of broad sense para distributes;

Counter (166) is used for utilizing described sample space to determine the load effect typical value of described target bridge;

Target bearing capacity fallout predictor (168) is used for doping the target bearing capacity that reply target bridge is reinforced according to described load effect typical value.

10. device as claimed in claim 9 is characterized in that,

Described parameter Processor is used for: get all described load effect values as sample, get maximum sample in 0 described sample of every N continuous, with all described maximum sample descending sorts, form sample space { x ₁, x ₂..., x _n, wherein, n is the sample number of sample space, x ₁, x ₂..., x _nSample for described sample space.

Images