JP2023006142A - Derivation method, derivation device, derivation system and program - Google Patents

Abstract

To calculate the number of movable bodies that constitute a movable body traveling in a structure with a lower load.SOLUTION: An acquisition step for acquiring time sequence data which includes the physical quantity generated at a prescribed observation point in a structure in response to the travel of a formation movable body formed by one or more movable bodies in the structure; a time derivation step for deriving the entry time and the exit time of the formation movable body relative to the structure on the basis of the time sequence data; and a number derivation step for deriving the number of movable bodies included in the formation movable body on the basis of the time sequence data, the entry time and the exit time: are included.SELECTED DRAWING: Figure 19

Description

The present invention relates to a derivation method, a derivation device, a derivation system, and a program.

In recent years, many social infrastructures have deteriorated over time, and there is a need for a method of diagnosing the state of structures that constitute social infrastructures, such as railway bridges.
Patent Literature 1 discloses a method for investigating and evaluating the structural performance of a railway bridge, which makes it possible to appropriately investigate and evaluate the structural performance of a bridge using observation data of the acceleration response of the bridge during train travel. In the method for investigating the structural performance of a railway bridge in Patent Document 1, a train is assumed to be a moving load train, a bridge is a simple beam, and a theoretical analysis model for the dynamic response of the railway bridge when the train is running is formulated, and the acceleration of the bridge when the train is running. is measured, and unknown parameters of the theoretical analysis model are estimated from the acceleration data by inverse analysis.
Further, Patent Literature 2 discloses a method of obtaining an impact coefficient (dynamic response component) of a bridge using a vehicle vertical acceleration response of a running train, especially when passing over the bridge.

Patent No. 6543863 Patent No. 6467304

BACKGROUND ART There is a case where a set of moving bodies, such as a railway train, in which one or more moving bodies are organized moves on a structure such as a bridge. In such a case, for the purpose of generating a motion model of the structure for diagnosis, there is a demand to know how many mobile bodies are organized in the organized mobile body that moves the structure. In Patent Document 1, the amount of calculation in the inverse analysis method for obtaining unknown parameters becomes enormous. Further, in Patent Document 2, it was not possible to find out how many mobile bodies are organized in the mobile body that moves the structure. As described above, in Patent Documents 1 and 2, it was not possible to determine how many moving bodies are organized in moving bodies that move a structure with a lower load.

A derivation method for solving the above problems includes a physical quantity generated at a predetermined observation point in the structure as a response due to the organized moving body in which one or more moving bodies are organized moving through the structure, time series an acquisition step of acquiring data; a time derivation step of deriving an entry time and an exit time to the structure of the mobile organization based on the time-series data; and the time-series data, the entry time, and the exit and a number deriving step of deriving the number of the moving bodies included in the organized moving body based on the time.
A derivation device for solving the above problems includes a physical quantity generated at a predetermined observation point in a structure as a response due to movement of an organized mobile body in which one or more mobile bodies are organized. an acquisition unit that acquires data; a time derivation unit that derives an entry time and an exit time from the structure of the mobile organization based on the time-series data; and the time-series data, the entry time, and the exit. and a number deriving unit that derives the number of the moving bodies included in the organized moving body based on the time.
A lead-out system for solving the above-mentioned problems is a lead-out system comprising a lead-out device and a sensor, wherein the lead-out device moves an organized moving body in which one or more moving bodies are organized to move a structure. A physical quantity generated at a predetermined observation point in the structure as a response to the event, the acquisition unit for acquiring time-series data including the physical quantity measured via the sensor, and based on the time-series data, a time deriving unit for deriving an entry time and a leaving time for the structure of the moving body, and the moving body included in the moving body based on the time-series data, the entering time, and the leaving time; and a number derivation unit for deriving the number of
A program for solving the above problem is stored in a computer as a response to movement of an organized moving body in which one or more moving bodies are organized, and physical quantities occurring at predetermined observation points in the structure, an acquisition step of acquiring time-series data; a time derivation step of deriving an entry time and an exit time of said moving body from said structure based on said time-series data; and said time-series data and said entry time. and a number derivation step of deriving the number of the moving bodies included in the organized moving body based on the leaving time.

It is a block diagram which shows the structure of a derivation|leading-out system. It is a figure which shows the cross section of a bridge. It is a figure which shows the dimension of a unit bridge girder. 1 is a diagram showing the dimensions of a railway vehicle; FIG. It is a figure which shows the outline|summary of a unit bridge girder. It is a figure explaining the bending moment in a unit bridge girder. It is a figure which shows the outline|summary of the bending of the unit bridge girder by a wheel. FIG. 4 is a diagram showing an outline of bending of a unit bridge girder due to a railway vehicle; It is a figure which shows the outline|summary of the bending of a unit bridge girder by a railroad train. It is a figure which shows the bending of a unit bridge girder by a railroad train. It is a figure which shows the FFT result of the bending of a unit bridge girder. FIG. 10 is a diagram showing deflection of a unit bridge girder due to a railroad train after high-pass filtering; FIG. 2 shows details of each element of the derivation system; It is a figure which shows the time-series data of the displacement of a unit bridge girder. It is a figure which shows the FFT result with respect to the time-series data of the displacement of a unit bridge girder. It is a figure explaining the derivation|leading-out process of an approach time and an exit time. It is a figure explaining the derivation|leading-out process of an approach time and an exit time. FIG. 4 is a diagram showing time-series data of displacement of a unit bridge girder after high-pass filtering; 4 is a flowchart showing derivation processing;

Here, embodiments of the present invention will be described according to the following order.
(1) Configuration of derivation system:
(1-1) Overview of derivation system:
(1-2) Flexure model:
(1-3) Verification experiment:
(1-4) Element details:
(2) Derivation process:
(3) Other embodiments:

(1) Configuration of derivation system:
(1-1) Overview of derivation system:
FIG. 1 is a block diagram showing an example of the configuration of a derivation system 10 according to this embodiment. The derivation system 10 calculates the number of railroad vehicles included in the railroad train 6 based on time-series data including physical quantities at predetermined observation points on the bridge 5 on which the railroad train 6 composed of one or more railroad vehicles moves. is a system that derives The railroad train 6 is an example of a mobile organization. Each railroad vehicle included in the railroad train 6 is an example of a moving object. A bridge 5 is an example of a structure on which a moving object moves. Each railroad car of the railroad train 6 moves on the bridge 5 via wheels provided on the axles. A wheel is an example of a contact portion between a railroad vehicle and a bridge. In this embodiment, each of the railway vehicles organized into the railway train 6 has the same structure. As shown in FIG. 1 , the derivation system 10 comprises a measuring device 1 , at least one sensor device 2 provided on the superstructure 7 of the bridge 5 and a server device 3 .

Based on the acceleration data output from each sensor device 2 , the measuring device 1 calculates the deflection displacement of the superstructure 7 due to the running of the railroad train 6 . The measuring device 1 is installed, for example, on the abutment 8b. The measuring device 1 and the server device 3 can communicate via a communication network 4 such as a mobile phone wireless network and the Internet, for example. The measuring device 1 transmits to the server device 3 information on the displacement of the superstructure 7 due to the running of the railroad train 6 . The server device 3 derives the number of railway vehicles organized into the railway train 6 based on the transmitted displacement data.

In this embodiment, the bridge 5 is a railway bridge, such as a steel bridge, a girder bridge, an RC bridge, or the like. RC is an abbreviation for Reinforced-Concrete. Moreover, in this embodiment, the bridge 5 is a structure to which BWIM (Bridge Way In Motion) is applicable. BWIM is a technique for measuring the weight, number of axles, etc. of a moving object passing through a bridge by measuring the deformation of the bridge by using the bridge as a scale. A bridge that can analyze the weight of a moving object passing through from the response of deformation and strain of the bridge is considered to be a structure to which BWIM can be applied. Therefore, the BWIM system, which applies the physical process between action and response to the bridge, makes it possible to measure the weight of moving objects moving across the bridge. Measurement of the weight of a moving object involves measuring the correlation coefficient between displacement and load in advance, and using the correlation coefficient to derive the load of the passing moving object from the measurement results of the displacement of the bridge when the moving object passes. is done in

The bridge 5 includes an upper structure 7, which is a portion on which the moving body moves, and a lower structure 8 that supports the upper structure 7. As shown in FIG. FIG. 2 is a cross-sectional view of the upper structure 7 taken along line AA of FIG. As shown in FIGS. 1 and 2, the superstructure 7 includes a floor 7a including floor slabs F, main girders G, cross girders (not shown), bearings 7b, rails 7c, sleepers 7d, and ballast 7e. and including. Also, as shown in FIG. 1, the substructure 8 includes a bridge pier 8a and an abutment 8b. The superstructure 7 is a structure that spans between adjacent abutments 8b and piers 8a, between two adjacent abutments 8b, or between two adjacent piers 8a. Below, the abutment 8b and the pier 8a are generically called a support part. In the present embodiment, a pair of supporting portions and a bridge girder portion of the superstructure 7 spanning between the pair of supporting portions are collectively referred to as one bridge girder. In other words, a single bridge girder is a simple beam-like structure whose both ends are supported by two support portions. Therefore, the bridge 5 shown in FIG. 1 includes two bridge girders. Each bridge girder included in the bridge 5 is hereinafter referred to as a unit bridge girder.
The measuring device 1 and the sensor device 2 are, for example, wired or wirelessly connected, and communicate via a communication network such as CAN (Controller Area Network).

The sensor device 2 is used to measure predetermined physical quantities used to derive displacement (deflection) at observation points set on the superstructure 7 . In this embodiment, this predetermined physical quantity is acceleration. Moreover, in this embodiment, the sensor device 2 is installed at this observation point. The sensor device 2 also includes an acceleration sensor such as a crystal acceleration sensor or a MEMS (Micro Electro Mechanical Systems) acceleration sensor. The sensor device 2 outputs acceleration data for deriving the displacement of the superstructure 7 due to the movement of the railroad train 6, which is a moving object, at the observation point.

In this embodiment, the sensor device 2 is installed in the longitudinal center of the superstructure 7, specifically, in the longitudinal center of the main girder G. As shown in FIG. However, the sensor device 2 only needs to be able to detect the acceleration for calculating the displacement of the upper structure 7 , and its installation position is not limited to the central portion of the upper structure 7 . If the sensor device 2 is provided on the floor slab F of the superstructure 7, it may be destroyed by the running of the railroad train 6, and the measurement accuracy may be affected by local deformation of the bridge floor 7a. Therefore, in the example of FIGS. 1 and 2 the sensor device 2 is provided on the main girder G of the superstructure 7 .

A floor slab F, a main girder G, and the like of the superstructure 7 are vertically bent by the load of the railroad train 6 running on the superstructure 7 . Each sensor device 2 measures the acceleration of deflection of the floor slab F and the main girder G due to the load of the railroad train 6 running on the superstructure 7 .

(1-2) Flexure model:
Here, a model of bridge deflection when a railroad train moves on one bridge will be described. Here, the model is information such as a formula indicating a correspondence relationship between predetermined information and an estimation result. In this embodiment, the bridge deflection model is a formula based on the structure of the bridge, as described below.

Moreover, below, the number (number) of the railroad vehicles organized into the railroad train which moves a bridge is set to N. Let the entry time, which is the time when the railroad train enters the bridge, be t _i . Here, entry of a railroad train into a bridge means that a wheel of one axle of the railroad vehicle C ₁ (the first railroad vehicle from the head of the railroad train) has entered the bridge. Also, hereinafter, the leaving time, which is the time when the railroad train leaves the bridge, is _{t 0} . Here, the exit of the railroad train from the bridge means that the wheel of the rearmost axle of the railroad vehicle CN (the _rearmost railroad vehicle of the railroad train) has withdrawn from the bridge. Moreover, below, the period (period from the time t _i to the time _{t o} ) when a railroad train passes a bridge is set to ts. Below, N, t _i , t _o , and t _s are collectively referred to as observation information.

Also, hereinafter, the bridge length, which is the length of the bridge in the traveling direction of the railroad train, is defined as _{LB. Let Lx} be the distance from the end of the bridge in the direction in which the railway train enters to the observation point. FIG. 3 shows _LB and _Lx . Hereinafter, the end of the bridge in the direction in which the railroad train enters is referred to as the entry end. Also, hereinafter, the end of the bridge in the direction in which the railroad train is leaving will be referred to as the exit end. In addition, the vehicle length, which is the length in the traveling direction of the m-th railroad vehicle from the head of the railroad train, is defined as _Lc (m). Hereinafter, L _c (1) to L _c (N) are collectively referred to as L _c . Also, let C _m be the m-th railroad car from the head of the railroad train. Also, let the number of axles in the railway vehicle C _m be a _r (m). Hereinafter, a _r (1) to a _r (N) are collectively referred to as a _r .
In the following, the a _r ( _{m) axles in the railway vehicle C m are} referred to as 1st, 2nd, 3rd, _.
In addition, let L _a ( _aw (m, 1)) be the distance from the front end of the railcar C _m in the traveling direction to the first axis. Here, a _w (α, β) indicates the β-th axle from the leading axle of the α-th railway vehicle in the railway train. Also, the distance between the n−1 axis (n: an integer equal to or greater than 2) and the n axis in the railway vehicle C _m is defined as L _a (a _w (m, n)). That is, L _a (a _w (α, β)) for β greater than or equal to 2 indicates the distance between the β axis and the (β−1) axis in the railway train C _α . Also, L _a ( _aw (α, 1)) indicates the distance between one axle of the railway train C _α and the front end of the railway train C _α in the traveling direction. Hereinafter, L _a (a _w (1, 1)) to L _a (a _w (N, a _r (N))) are collectively referred to as L _a . Each L _a indicates the position of the corresponding axle in the corresponding railcar. For example, L _a (a _w (m, 1)) indicates that there is one axis behind the front end of the railcar C _m at a distance of L _a (a _w (m, 1)). In addition, L _a ( _aw (m, 2)) indicates that two axles exist behind the first axle by a distance of L _a ( _aw (m, 2)) in the railway vehicle C _m .
Here, the railway train is composed of railway vehicles having a similar four-axle configuration. That is, a _r (m) (m=1, 2, . . . , N) is 4. L _c ( _m ), _La ( _aw (m, 1)), _La ( _aw (m, 2)), _La (aw (m, 3)), La ( _aw (m, 3)), L _a (a _w (m, 4)) is shown.
Below, L _B , L _x , L _c , a _r , and L _a are collectively referred to as environment information.

t _s is obtained as the difference between t _o and t _i as shown in the following equation (1).

Also, the total number of wheels T _ar of a railroad train is obtained by the following equation (2).

The distance from the 1st axis of the leading railroad vehicle C1 to the n-axis of the _m -th railroad vehicle Cm is represented as _Dwa ( _aw (m,n)). D _wa (a _w (m, n)) is obtained from the following equation (3).

The distance from one axis of the leading rail car C1 to the last axis a _r (N) of the last rail car C _N is D _wa ( _aw (N, ar(N)) ). Using D _wa (a _w (N, a _r (N))), the average speed v _a of the railroad train passing through the bridge is represented by the following equation (4).

The following equation (5) holds from equations (3) and (4).

Next, a description will be given of the deflection that occurs in the bridge when a load is applied to the bridge.
Figure 5 shows a schematic diagram of the bridge. FIG. 5 shows a situation in which a load P is applied to the bridge. Here, the distance between the position on the bridge where the load P is applied and the entry end is represented as a. Also, the distance between the position where the load P is applied on the bridge and the exit end is represented as b. In this case, the bending moment at the position where the load P is applied on the bridge is represented by the following equation (6).

FIG. 6 shows the bending moment at each position of the bridge due to the load P. As shown in FIG. 6, the bending moment generated in the bridge by the load P is 0 at the entry end, and increases proportionally as the position where the load P is applied from the entry end approaches, and the load P is applied. At the position, it becomes the value indicated by equation (6). Further, the bending moment generated in the bridge by the load P decreases proportionally from the position where the load P is applied to the retracted end, and becomes 0 at the retracted end. Therefore, the bending moment at an arbitrary position X on the bridge is represented by the following equation (7).

x in the expression (7) indicates the distance from the entry end to the position X in the traveling direction of the railroad train. Also, Ha in Expression (7) is a value shown in Expression (8) below.

Between the deflection w of the bridge at an arbitrary position X and the bending moment, the relationship represented by the following formula (9) holds.

E in equation (9) is the Young's modulus of the bridge. Also, I in Equation (9) is the second moment of the bridge. θ in equation (9) is the angle formed by the horizontal line at the position X and the bent bridge. The following equation (10) holds from equations (7) and (9).

By integrating both sides of equation (10) with x twice, equation (11) below, which expresses the deflection w at position X, is obtained.

g1 and g2 in equation (11) are constant terms. Here, since the bridge is supported at the entrance end and the exit end, no bending occurs at the entrance end and the exit end. That is, in equation (11), both sides are 0 when x=0 and x= _LB . Therefore, g1 and g2 are given by the following equations (12) and (13).

The following equation (14) representing the deflection w at the position X is obtained from equations (11), (12), and (13).

When the load P is applied to the center of the bridge in the longitudinal direction, the maximum deflection of the deflections caused in the bridge due to the application of the load P occurs at the center of the bridge in the longitudinal direction. Assuming that this maximum deflection is w _0.5l , an equation expressing w _0.5l is obtained. If the load P is applied to the longitudinal center of the bridge, a=b=0.5L _B. Also, since the target position X for which deflection is to be obtained is the center of the bridge in the longitudinal direction, x=0.5L _B. Also, in this case, since x<=a, H _a =0 from equation (8). By substituting x=0.5L _B , a=b=0.5L _B , and H _a =0 into the equation (14), the following equation (15) representing the deflection w _0.51 can be obtained.

w _0.5l is used to normalize the deflection at any position in the bridge represented by equation (14).
When the position of the load P exists on the approach end side of the position X, that is, when x>a, H _a =1 from formula (8), and formula (14) is transformed into the following formula (16). is represented by

Let a=L _B r. Here, r is a real number of 0 or more and 1 or less. Since b=L _B −a, it is expressed as b=L _B (1−r). Substituting a=L _B r and b=L _B (1−r) into equation (16) and normalizing by dividing by w _0.5l yields normalization at position X when x>a Equation (17) below, which indicates the applied deflection w _std , is obtained.

Similarly, when the position of the load P exists on the withdrawal end side of the position X, i.e., when x<=a, H _a = 0 from formula (8), and formula (14) is the following formula ( 18).

Let a=L _B r. Here, r is a real number of 0 or more and 1 or less. Since b=L _B −a, it is expressed as b=L _B (1−r). Substituting a=L _B r and b=L _B (1−r) into equation (18) and normalizing by dividing by w _0.5l yields the standard at position X when x<=a Equation (19) below is obtained, which indicates the normalized deflection w _std .

By substituting L _x for x in equations (17) and (19), the normalized deflection w _std at the deflection observation point is expressed as the following equation (20) as a function of r. be done.

Function R(r) in equation (20) is a function shown in equation (21) below.

Here, using equations (20) and (21), we show the time variation of the deflection caused at the observation point by the load applied to the bridge via the wheels of any one axle a _w (m, n). Find a function.
First, let t _xn be the time it takes for a wheel of one axle of a railroad train to travel from the approach end to the observation point. t _xn is obtained from L _x and v _a by the following equation (22).

Also, let t _ln be the time it takes for one wheel of the railroad train to cross the bridge, that is, the time it takes from the entry end to the exit end. t _ln is obtained from _LB and _va by the following equation (23).

Let t _o (m, n) be the time at which the wheels of the n-axis a _w (m, n) of the m-th rail car of the railroad train reach the entry end. t _o (m, n) is obtained from t _i , v _a , and D _wa (a _w (m, n)) by the following equation (24).

From Equation (22), L _x is expressed as in Equation (25) below.

Also, from equation (23), _LB is expressed as in equation (26) below.

The position of the axle a _w (m, n) is the load position. Therefore, the position of the axle a _w (m, n) is at a distance a=L _B r in the direction from the entry end to the exit end. Also, if a variable indicating time is t, the distance from the approach end of a _w (m, n) at time t is the distance traveled by the railcar from time t _o (m, n) to time t. equal. Therefore, the following formula (27) holds.

From Equation (27), r is expressed as in Equation (28) below.

By replacing L _x , L _B , and r in Equations (20) and (21) using Equations (25), (26), and (28), wheels of axle a _w (m, n) A function w _std (a _w (m, n), t) of the following equation (29) is obtained as a model showing the time variation of the deflection occurring at the observation point due to the load applied to the bridge via . Function R(t) in equation (29) is a function represented by equation (30) below.

Observation information and environmental information (t _i , t _o , N, L _B , L _x , L _c (1) to L _c (N), a _r (1) to a _r (N), L _a (a _w (1, 1)) through L _a (a _w (N, a _r (N)))) are known, using these information, w _std (a _w (m, n), t) is sought. For example, t _s is obtained from t _i and _t 0 using equation (1). From t _s , N, a _r , L _a , and L _c , v _a is obtained using equation (5). From _{va, LB, and Lx, txn and tln are obtained using equations} (22) and (23). From L _a , L _c , and ti, t _o (m, n) is obtained using equations (3) and (24). Then, by substituting the obtained t _xn , t _ln , t _o (m, n) into the equations (29) and (30), the function w _std (a _w (m, n), t) of t is sought.

FIG. 7 shows an example of changes in the deflection amount at the observation point indicated by w _std (a _w (m, n), t). The horizontal axis of the graph in FIG. 7 indicates time, and the vertical axis indicates the amount of deflection. Also, according to the movement of one railroad vehicle C _m , the set of wheels for each of the a _r (m) axles will move on the bridge. Therefore, the function C _std (m, t) as a model representing the time variation of the amount of deflection caused at the observation point due to the movement of one railway vehicle C _m is w _std (a _w (m, n ) and t) are obtained by the following equation (31).

When a _r (m) is 4, that is, when the railway vehicle C _m has a four-axle configuration, FIG. 8 shows how the amount of deflection changes at the observation points indicated by the function C _std (m, t). The horizontal axis of the graph in FIG. 8 indicates time, and the vertical axis indicates the amount of deflection. Also, the solid line graph in FIG. 8 indicates C _std (m, t), and each dotted line graph indicates w _std (a _w (m, n), t) for each axle.

Also, according to the movement of the railway train, N railway vehicles move on the bridge. Therefore, the function T _std (t) as a model showing the time change of the amount of deflection caused at the observation point due to the movement of one railway train is the sum of C _std (m, t) for each railway vehicle, as follows: (32).

When N is 16, that is, when 16 railcars are organized in a railroad train, FIG. 9 shows how the amount of deflection changes at the observation points indicated by the function T _std (t). The horizontal axis of the graph in FIG. 9 indicates time, and the vertical axis indicates the amount of deflection. Further, the solid line graph in FIG. 9 indicates T _std (t), and each dotted line graph indicates C _std (m, t) for each railway vehicle. As shown in the graph of FIG. 9, the waveform is obtained by adding the deflection of each passing railway vehicle, and it can be seen that vibration occurs at the period when successive vehicles pass through the bridge.
The above is the explanation of the deflection model in the bridge.

(1-3) Verification experiment:
The inventors obtained the deflection amount T _std (t) under the conditions that the observation information and the environmental information have the following values. That is, N = 4, t _i = 7.21 [seconds], t _o = 8.777 [seconds], t _s = _1.567 [seconds], LB = 25 [m], L _x = 12.5 [m], L _c each = 25 [m], a _r each = 4, m = 1 to N respectively, L _a (a _w (m, 1)) = 2.5 [m], m = 1 to N L _a (a _w (m, 2)) = 2.5 [m] for each, m = 1 to N L _a (a _w (m, 3)) = 15 [m] for each, m = 1 to N L _a (a _w (m, 4))=2.5 [m] for each.

FIG. 10 shows the deflection amount T _std (t) at this time. The horizontal axis of the graph in FIG. 10 indicates time, and the vertical axis indicates the amount of deflection. In addition, the inventors determined the intensity of each frequency component included in T _std (t) by fast Fourier transforming (FFT) the determined T _std (t). The FFT results for T _std (t) are shown in FIG. The horizontal axis of the graph in FIG. 11 indicates the frequency, and the vertical axis indicates the intensity of the corresponding frequency component. Then, the inventors obtained the fundamental frequency F _f of T _std (t) from the FFT results of T _std (t) as the frequency of vibrations occurring in the bridge according to the continuous movement of railroad vehicles. Here, the fundamental frequency is the frequency of the lowest frequency component included in the signal. Specifically, the inventors identified the peak corresponding to the lowest frequency from the FFT result of T _std (t), excluding the sidelobes caused by the effect of the window function used in the FFT, and identified The resulting peak was obtained as the fundamental frequency. In the example of FIG. 11, as indicated by the dashed-dotted line, there are two sidelobe peaks in the range of less than 2 Hz caused by the window function used in the FFT. The inventors identified the peak surrounded by the dotted line as the lowest frequency peak among the peaks excluding these peaks, and obtained the frequency corresponding to the identified peak as the fundamental frequency _Ff . The inventors obtained the fundamental frequency of 3.1 Hz from the graph of FIG.

The inventors obtained the wavenumber ν of the fundamental frequency _Ff included in the transit period ts using the following equation (33). That is, the wavenumber ν is derived as the product of ts and _Ff .

In this case, ν=1.567×3.1=4.8577. Here, the number N of railcars of the moving railroad train is four. The inventors have found a feature that the wavenumber ν of the fundamental frequency _Ff included in the transit period ts is higher than N by about one. This feature is hereinafter referred to as the first feature. Therefore, the inventors assume that the number N of railroad vehicles included in the railroad train is obtained as a value obtained by subtracting 1 from the wavenumber ν of the fundamental frequency _Ff included in the transit period ts and rounded to an integer, It was found that it can be obtained using the following formula (34). A round function is a function that returns a rounded value of an argument. Also, the method of rounding to an integer is not limited to rounding off, and other methods such as rounding down or rounding up may be used in accordance with the characteristics of the bridge to be observed.

Also, the inventors obtained the fundamental period _Tf from the fundamental frequency _Ff using the following equation (35).

Then, the inventors took a moving average of the amount of deflection T _std (t) at the fundamental period T _f to apply a low-pass filtering process to T _std (t) for attenuating frequency components equal to or higher than the fundamental frequency. The low-pass filtering process may be another FIR filtering process that attenuates frequency components equal to or higher than the fundamental frequency. Let T _std (t) subjected to low-pass filtering be T _{std_lp} (t)=T _{std_lp} (kΔT). Here, k is a variable that indicates the order of observation when the amount of deflection is periodically observed at an observation point. That is, if the data period (time resolution) of the deflection amount observation is ΔT, then t=kΔT.
As shown in Equation (36) below, the moving average interval _kmf adjusted to the time resolution of the data is obtained from the fundamental period _Tf and ΔT.

Using _{kmf, T std_lp (} t) is obtained by the following equation (37).

The inventors subtracted T _{std_lp} (t) from the amount of deflection T _std (t), thereby performing high-pass filter processing for attenuating frequency components lower than the fundamental frequency to T _std (t). The high-pass filtering process may be any other FIR filtering process that attenuates frequency components below the fundamental frequency. Let T _std (t) subjected to high-pass filtering be T _{std_hp} (t). Specifically, the inventors obtained T _{std_hp (t) by subtracting T std_lp (} t) from T _std (t) as shown in the following equation (38).

The calculated T _{std_hp} (t) is superimposed on T _std (t) and shown in FIG. In the graph of FIG. 12, the horizontal axis indicates time (t=kΔT), and the vertical axis indicates the amount of deflection. The solid line graph in FIG. 12 indicates T std — _hp (k), and the dotted line graph indicates T _std (t).
From the graph of FIG. 12, the number of positive peaks of T _{std_hp} (t) in transit period t _s (period from time t _i to time t _o ) is six. Here, the positive peak is the upward convex peak of the bridge among the peaks of T _{std_hp} (t). Also, the number of negative peaks of T _{std_hp} (t) in transit period t _s is five. Here, the negative peak is a downward convex peak of the bridge among the peaks of T _{std_hp} (t). From this, the inventors found that the number of positive peaks of T _{std_hp} (t) in the transit period t _s (6) is two more than the number of rolling stock N (4) included in the railway train, and the negative A feature was found that the number of peaks (5) was one more than N (4). This feature is hereinafter referred to as a second feature.

The inventors verified whether the first feature and the second feature hold while changing the observation information and the environment information to various values. As a result, the inventors found that the first feature and the second feature are satisfied when L _c /2 < _LB < 3L _c /2.
Based on the first feature and the second feature, the derivation system 10 of the present embodiment calculates the number of railroad vehicles organized in the railroad train 6 from the displacement (deflection) of the bridge observed via the sensor device 2. Perform processing to derive

(1-4) Details of Elements Details of each of the measurement device 1, the sensor device 2, and the server device 3 of the derivation system 10 will now be described with reference to FIG.
Incidentally, in this embodiment, the environmental information for each of the railway train 6 and the unit bridge girder is already known. That is, the bridge length L _B of the unit bridge girder, the distance L _x from the entry end of the unit bridge girder to the observation point, the vehicle length L _c of each railroad vehicle of the railroad train 6, the number of axles a _r of each railroad vehicle of the railroad train 6 , L _a of each railcar of the railroad train 6 are known. Specifically, L _B =25 [m], L _x =12.5 m, L _c (1) to L _c (N) each =25 [m], a _r (1) to a _r (N) each = 4, L _a (a _w (1, 1)) to L _a (a _w (N, 1)) respectively = 2.5 [m], L _a (a _w (1, 2)) to L _a ( a _w (N, 2)) each = 2.5 [m], L _a (a _w (1, 3)) to L _a (a _w (N, 3)) each = 15 [m], L _a ( a _w (1, 4)) to L _a (a _w (N, 4))=2.5 [m] respectively.
In addition, in the present embodiment, the derivation system 10 uses observation information (the number N of railway vehicles organized into the railway train 6, the time t i when the railway train 6 enters the unit bridge girder, the time t _i when the railway train 6 exits the unit bridge girder, The time t _o and the period t _s during which the railway train 6 passes through the unit bridge girder are derived based on the data measured by the measuring device 1 . The number N of railway vehicles organized into the railway train 6 is sixteen.

A measuring device 1 measures deflection at an observation point via a sensor device 2 . Although the measuring device 1 is installed on the abutment 8b in this embodiment, it may be installed at another position. The measuring device 1 includes a control section 100 , a storage section 110 and a communication section 120 . The control unit 100 includes a processor such as a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and the like. The control unit 100 realizes each function of the measuring device 1 by developing various programs recorded in a ROM or the like into a RAM and executing them via the CPU. The storage unit 110 stores various programs, measured deflection data, and the like. The communication unit 120 includes a circuit used for wired or wireless communication with an external device.

The sensor device 2 detects acceleration as a predetermined physical quantity at the observation point. The sensor device 2 includes a control section 200 , an acceleration sensor 210 , a storage section 220 and a communication section 230 . The control unit 200 includes a processor such as a CPU, a ROM, a RAM, and the like. The control unit 200 realizes each function of the sensor device 2 by loading various programs recorded in ROM or the like into RAM and executing them via the CPU.

The acceleration sensor 210 is an acceleration sensor such as a crystal acceleration sensor, a MEMS acceleration sensor, or the like, which can detect acceleration occurring in each of three axial directions orthogonal to each other. In this embodiment, the acceleration sensor 210 is arranged so that one axis is parallel to the vertical direction in order to detect acceleration in the vertical direction more accurately. However, the installation location of the sensor device 2 in the upper structure 7 may be tilted. Even if one of the three detection axes of the acceleration sensor 210 is not aligned in the vertical direction, the measuring device 1 combines the accelerations of the three axes to detect acceleration in the vertical direction.

The control unit 200 of the sensor device 2 periodically detects vertical acceleration at an observation point on the bridge 5 via the acceleration sensor 210 and transmits data of the detected acceleration to the measuring device 1 . The control unit 100 of the measurement device 1 measures the vertical deflection of the bridge 5 at the observation point at the acceleration detection time based on the acceleration data transmitted from the sensor device 2 . In this embodiment, the control unit 100 obtains the vertical deflection of the bridge 5 at the observation point by integrating the acceleration indicated by the data transmitted from the sensor device 2 twice with respect to time. Then, the control unit 100 transmits the measured deflection data to the server device 3 . Note that, in the present embodiment, the sensor device 2 detects acceleration at a predetermined cycle ΔT. Therefore, the measuring device 1 measures the time-series data of the deflection of the ΔT period. That is, the time-series data to be measured is data of discrete values of displacement measured in the period of ΔT, and each discrete value is data associated with the measurement time.

The server device 3 derives the number of railcars included in the railroad train 6 based on the deflection at the observation point measured by the measuring device 1 . The server device 3 is an example of a derivation device. The server device 3 includes a control section 300 , a storage section 310 and a communication section 320 . The control unit 300 includes a processor such as a CPU, a ROM, a RAM, and the like. The control unit 300 implements the functions of an acquisition unit 301, a time derivation unit 302, and a number derivation unit 303 by loading various programs recorded in a ROM or the like into a RAM and executing them via the CPU. The storage unit 310 stores various programs, data of detected deflection, and the like. The communication unit 320 includes a circuit used for wired or wireless communication with an external device.

The acquisition unit 301 has a function of acquiring time-series data of deflection occurring at an observation point as a response when the railroad train 6 moves on each bridge in the bridge 5 . The control unit 300 acquires the time-series data of the deflection occurring at the observation point from the measurement device 1 using the function of the acquisition unit 301 . Hereinafter, the time-series data of the deflection acquired by the function of the acquisition unit 301 is u(t). FIG. 14 shows an example of u(t). The horizontal axis of the graph in FIG. 14 indicates time, and the vertical axis indicates the amount of deflection.

The time derivation unit 302 has a function of deriving the entry time and exit time of the railroad train 6 with respect to the unit bridge girder based on the time-series data acquired by the acquisition unit 301 . The control unit 300 uses the function of the time derivation unit 302 to perform FFT on u(t). FIG. 15 shows the result of performing FFT on u(t) shown in FIG. The horizontal axis of the graph in FIG. 15 indicates frequency, and the vertical axis indicates the intensity of the corresponding frequency component. Then, the control section 300 detects a peak from the FFT result. The control unit 300 identifies the peak corresponding to the minimum frequency among the detected peaks, excluding the peak of the side lobe caused by the window function used in the FFT. The control unit 300 derives the frequency corresponding to the specified peak as the fundamental frequency _Ff of u(t). In the example of FIG. 15, the derived fundamental frequency F _f is 3.01 Hz.

The control unit 300 applies a low-pass filter to u(t) to attenuate components of the fundamental frequency _Ff or higher as follows. First, based on the acquired fundamental frequency Ff, the control unit 300 derives the period _Tf by deriving the reciprocal of _Ff in the same manner as in Equation (35 ₎ . Control unit 300 derives interval _kmf using equation (36) based on derived _Tf and ΔT, which is the predetermined period. The control unit 300 applies a low-pass filter to u(t) by taking a moving average of each value of u(t) in the derived interval _kmf . Let u(t) subjected to low-pass filtering be u _lp (t)=u _lp (kΔT). Here, k is a variable that indicates the order of observation when the amount of deflection is periodically observed at an observation point. The control unit 300 derives u _lp (t) using the following equation (39) based on the derived interval _kmf . u _lp (t) is data of multiple discrete values, similar to u(t), which is data of multiple discrete values.

This low-pass filter may be an FIR filter that attenuates components of the fundamental frequency _Ff or higher.
Then, from u _lp (t), the control unit 300 identifies two continuous data that sandwich a predetermined threshold value C _L regarding the amount of deflection. Here, two consecutive data of u _lp (t) sandwiching C _L means a range sandwiched between values of two continuously measured displacement data included in u _lp (t), that is, , indicates that _CL is included in the range of the smaller value or more and the larger value or less of these displacement data.
This threshold C _L is the value of the deflection that occurs in the bridge when a railroad train enters the bridge. is the value of the deflection of the observation point of the bridge in case Also, this threshold value _CL may be any other value as long as it is possible to detect the entry of a railroad train into a bridge. It may be the amount of deflection. Also, the threshold value _CL may be a value such as a predetermined ratio (eg, 10%, 1%, etc.) of the maximum value of the deflection amount at the observation point of the bridge when the railroad train passes over the bridge. Also, the threshold C _L may be set to any data value included in u _lp (t).

FIG. 16 shows u _lp (t) and the threshold C _L . The horizontal axis of the graph in FIG. 16 indicates time (t=kΔT), and the vertical axis indicates the amount of deflection. The solid line graph in FIG. 16 indicates u _lp (t), and the dotted line graph indicates u(t). u _lp (k) and the threshold C _L intersect in the portion surrounded by the dotted line circle in FIG. 16 . Also, FIG. 17 shows an enlarged view of the crossing point of _{u lp} (t) and CL (left dotted circle portion in the graph of FIG. 16). The horizontal axis of the graph in FIG. 17 indicates time, and the vertical axis indicates the amount of deflection. Each black dot in FIG. 17 indicates discrete value data included in u _lp (t). In the example of FIG. 17, data k−1 and data k included in u _lp (t) sandwich the threshold C _L .
Control unit 300 identifies the later one of the two times corresponding to the two consecutive data sandwiching the identified _CL . In the example of FIG. 17, control unit 300 identifies time kΔT corresponding to data k.
In the example of FIG. 16, the control unit 300 also specifies two data _circled by dotted lines on the right side of FIG. Identify the later of the two times.

Then, the control unit 300 derives the earliest of the specified times as the entry time t _i of the railroad train 6 to the unit bridge girder. Further, the control unit 300 derives the later one of the specified times as the leaving time _{t 0} of the railroad train 6 from the unit bridge girder. In the example of FIG. 16, the control unit 300 derives the entry time t _i =7.2 [s] and the exit time t _o =12.795 [s]. Thus, in the present embodiment, the control unit 300 derives the times associated with any data included in _{u lp} (t) as the entering time t _i and the leaving time t 0 .

Thus, in the present embodiment, the control unit 300 selects the later of the two times corresponding to the two consecutive data sandwiching CL included in _{u lp} (t) as the entry time t _i , the entry time t i , Derived as time t _o . However, the control unit 300 may derive other times as the entry time _{t i} and the entry time t 0 . For example, the control unit 300 identifies, from u _lp (t), two consecutive data sandwiching a predetermined threshold C _L regarding the deflection amount, and after one of the times corresponding to the identified two data, The time included in the period before the other time may be derived as the approach time _{t i} and the approach time t 0 . In the example of FIG. 17, the control unit 300 controls the time after the time (k-1)ΔT corresponding to the data k-1 and before the time kΔT corresponding to the data k (for example, the time (k-1)ΔT, such as the time corresponding to the intersection of u _lp (t) and C _L ) may be derived as the approach time t _i . Further, the control unit 300 obtains a curve obtained by interpolating between each data included in u _lp (t), and obtains the times corresponding to the intersections of the obtained curve and C _L as t _i and t _o . good.

Also, it is conceivable that one of two consecutive data sandwiching C _L contained in u _lp (t) is equal to C _L . For example, in the example of FIG. 17, it is possible that the value of data _k is equal to CL. In that case, the control unit 300 selects one of the two sets of the data equal to _CL and the data immediately before, and the data equal to _CL and the data immediately after, It may be selected as two consecutive data sandwiching _CL . In the example of FIG. 17, when data k is equal to _CL , the control unit 300 generates two consecutive data sandwiching _CL , a set of data k−1 and data k, and a set of data k and data k+1. , and select either one of the two sets. The control unit 300 may derive the time within the period between the two times corresponding to the two data included in the selected set as t _i or t _o .

In the present embodiment, the control unit 300 derives the times associated with any data included in _{u lp} (t) as the entering time t _i and the leaving time t 0 . As a result, the control unit 300 can easily acquire the data of u _lp (t) corresponding to each measurement time of the ΔT interval including the entry time t _i and the exit time t 0 by referring to _{u lp} (t). and can be used. On the other hand, when the control unit 300 derives times that are not associated with any data included in _{u lp} (t) as the entry time t _i and the exit time t 0 , t _i , t ₀ are included. The data of u _lp (t) corresponding to each measurement time of the ΔT interval is obtained by resampling from the original u _lp (t) or the like, which increases the labor of processing.

The control unit 300 uses u _lp (t) in which the vibration component of the fundamental frequency or higher is attenuated to derive the entry time and the exit time, thereby reducing the influence of the vibration component of the fundamental frequency or higher and improving accuracy. Entry time and exit time can be derived.
However, the control unit 300 may not derive u _lp (t). In that case, the control unit 300 may derive t _i and t _o using u(t) instead of u _lp (t), for example.

Based on u(t), which is the time-series data obtained by the function of the obtaining unit 301, and the entry time t _i and the exit time t _o obtained by the function of the time derivation unit 302, the number derivation unit 303 This is a function of deriving the number of railway vehicles included in the railway train 6 .
The control unit 300 uses the function of the number derivation unit 303 to derive a candidate value for the number of railcars included in the railroad train 6 as a first candidate value based on the first feature. Based on t _i and t _o , the control unit 300 uses Equation (1) to derive the passage period t _s during which the railroad train 6 passes through the unit bridge girder. Then, based on the derived t _s and the fundamental frequency F _f derived based on u(t), the control unit 300 uses Equation (33) to calculate the fundamental frequency F _f included in the transit period t _s to derive the wavenumber ν of Based on the derived ν, the control unit 300 uses Equation (34) to derive the number N of railcars included in the railroad train 6 as the first candidate value. Here, since t _i =7.2 [s], t _o =12.795 [s], and F _f =3.01 [Hz], the control unit 300 sets the first candidate value to round (( 12.795−7.2)×3.01−1)=round(15.84)=16.

Further, the control unit 300 derives candidate values for the number of railway vehicles included in the railway train 6 as the second candidate value and the third candidate value based on the second feature. By subtracting u _lp (t) from u(t), the control unit 300 applies high-pass filtering to u(t) to attenuate components of frequencies lower than the fundamental frequency, and u Derive u _hp (t), which is (t).
Further, the control section 300 may derive u _hp (t) by multiplying u(t) by a high-pass filter, which is an FIR filter that attenuates frequency components lower than the fundamental frequency.
Then, the control unit 300 identifies the number of positive peaks from the data in the period from t _i to t _o in u _hp (t). Control unit 300 derives a value obtained by subtracting 2 from the number of identified positive peaks as a second candidate value. Also, the control unit 300 identifies the number of negative peaks from the data in the period from t _i to t _o in u _hp (t). The control unit 300 derives a value obtained by subtracting 1 from the specified number of negative peaks as a third candidate value. FIG. 18 shows u _hp (t) obtained by subjecting u(t) shown in FIG. 14 to high-pass filtering. The horizontal axis in FIG. 18 indicates time, and the vertical axis indicates the amount of deflection. The solid line graph in FIG. 18 indicates u _hp (t), and the dotted line graph indicates u(t). In the example of FIG. 18, the number of positive peaks of u _hp (t) is 18 and the number of negative peaks is 17 in time period t _s . Therefore, control section 300 derives 18-2=16 as the second candidate value. Also, the control unit 300 derives 17−1=16 as the third candidate value.

The control unit 300 compares the first candidate value, the second candidate value, and the third candidate value, and determines the final estimated value of N based on the comparison result. When the first candidate value, the second candidate value, and the third candidate value are equal, the control unit 300 determines these equal values as the estimated value of N.
In addition, if the first candidate value, the second candidate value, and the third candidate value are not all equal, and the second candidate value and the third candidate value are equal, the control unit 300 Either the second candidate value or the third candidate value is determined as the estimate of N. Since the number of positive peaks and the number of negative peaks of u _hp (t) in transit period t _s are correlated, if the second candidate value and the third candidate value are equal, the second candidate value and the third candidate value are considered to be more reliable than the first candidate value.
Further, if the first candidate value, the second candidate value, and the third candidate value are not all equal, and the second candidate value and the third candidate value are not equal, Of the second candidate value and the third candidate value, the value closer to (F _f ×t _s −1) is determined as the estimated value of N. In this way, when the second candidate value and the third candidate value are not equal due to the influence of noise or the like, the control unit 300 sets (F _f ×t _s −1) to a positive value, and the second candidate value and the third candidate value, whichever is closer to (F _f ×t _s −1) is determined as the estimated value of N. However, the control unit 300 may determine the first candidate value as the estimated value of N when the second candidate value and the third candidate value are not equal.
In this way, the control unit 300 determines the final estimated value of N based on the comparison result of the first candidate value, the second candidate value, and the third candidate value, thereby improving accuracy. , N can be derived. Here, since the first candidate value=second candidate value=third candidate value=16, the control unit 300 determines the final value of N to be 16. Here, since the number N of railroad vehicles of the railroad train 6 is 16, it was confirmed that the controller 300 was able to derive the number N of railroad vehicles of the railroad train 6 .

As described above, according to the configuration of the present embodiment, the derivation system 10 uses the time-series data of the displacement (deflection) of the observation point on the bridge 5 based on the first feature and the second feature. The number of railcars included in the train 6 can be derived. Also, the derivation system 10 obtains the first candidate value using the equations (33) and (34). The derivation system 10 also performed high-pass filtering on u(t) and used the number of peaks in u _hp (t) to obtain the second candidate value and the third candidate value. In this way, the derivation system 10 reduces the amount of calculation for each of the first candidate value, the second candidate value, and the third candidate value compared to the case of obtaining the number of railroad vehicles included in the railroad train 6 by the reverse analysis method. , can be obtained at low load. In this way, the derivation system 10 can obtain the number of railway vehicles that are organized in the railway train 6 with a lower load.

(2) Derivation process:
The process of deriving the number of railroad cars of the railroad train 6 executed by the server device 3 will be described with reference to FIG. 19 . The server device 3 starts the processing of FIG. 19 in response to the transmission of the displacement data at the observation point from the measuring device 1, but starts the processing of FIG. 19 at an arbitrary timing such as a designated timing. may
In S<b>100 , the control unit 300 acquires the time-series data u(t) of the deflection occurring at the observation point from the measurement device 1 using the function of the acquisition unit 301 . S100 is an example of an acquisition step.

In S105, the control unit 300 performs FFT on u(t) acquired in S100 by the function of the time derivation unit 302. FIG. The control unit 300 detects peaks from the FFT result. The control unit 300 identifies the peak corresponding to the minimum frequency among the detected peaks, excluding the peak of the side lobe caused by the window function used in the FFT. The control unit 300 derives the frequency corresponding to the specified peak as the fundamental frequency _Ff of u(t).

In S110, the control unit 300 uses the function of the time deriving unit 302 to derive the period _Tf based on the fundamental frequency _Ff by deriving the reciprocal of _Ff in the same manner as in Equation (35). Control unit 300 derives interval _kmf using equation (36) based on derived _Tf and ΔT, which is the predetermined period. The control unit 300 applies a low-pass filter to u(t) by taking a moving average of each value of u(t) in the derived interval _kmf . The control unit 300 derives u _lp (t) using Equation (39) based on the derived interval _kmf .

Then, the control unit 300 obtains the time at which u _lp (t) intersects with the predetermined threshold C _L regarding the deflection amount. The control unit 300 derives the earlier one of the obtained times as the entry time _ti of the railroad train 6 to the unit bridge girder. In addition, the control unit 300 derives the later one of the obtained times as the leaving time _{t 0} of the railroad train 6 from the unit bridge girder. S110 is an example of the time derivation step.

In S115, the control unit 300 uses the function of the number deriving unit 303 to derive the passage period ts during which the railway train 6 passes through the unit bridge girder using Equation (1) based on _{t i} and t _o . . Then, based on the derived t _s and the fundamental frequency F _f derived based on u(t), the control unit 300 uses Equation (33) to calculate the fundamental frequency F _f included in the transit period t _s to derive the wavenumber ν of Based on the derived ν, the control unit 300 uses Equation (34) to derive the number N of railcars included in the railroad train 6 as the first candidate value.

In S120, the control unit 300 uses the function of the number derivation unit 303 to perform a high-pass filtering process that attenuates components of frequencies below the fundamental frequency by subtracting u _lp (t) derived in S110 from u(t). Apply u(t) to derive u _hp (t), which is the high-pass filtered u(t).
Then, the control unit 300 identifies the number of positive peaks from the data in the period from t _i to t _o (period t _s ) in u _hp (t). Control unit 300 derives a value obtained by subtracting 2 from the number of identified positive peaks as a second candidate value. Also, the control unit 300 identifies the number of negative peaks from the data of the period t _s in u _hp (t). The control unit 300 derives a value obtained by subtracting 2 from the specified number of negative peaks as a third candidate value.

In S125, the control unit 300 uses the function of the number derivation unit 303 to determine whether or not the first candidate value, the second candidate value, and the third candidate value are all equal. If the control unit 300 determines that the first candidate value, the second candidate value, and the third candidate value are all equal, the process proceeds to S130. If the control unit 300 determines that the first candidate value, the second candidate value, and the third candidate value are not all equal, the process proceeds to S135.

In S130, the control unit 300 uses the function of the number deriving unit 303 to set the first candidate value (=second candidate value, third candidate value) as the final estimated value of the number N of railcars of the railroad train 6. decide.
In S<b>135 , the control unit 300 determines whether or not the second candidate value and the third candidate value are equal using the function of the number derivation unit 303 . If the control unit 300 determines that the second candidate value and the third candidate value are equal, the process proceeds to S145. If the control unit 300 determines that the second candidate value and the third candidate value are not equal, the process proceeds to S140.

In S140, the control unit 300 uses the function of the number derivation unit 303 to determine the value closer to (Ff×ts−1) from among the second candidate value and the third candidate value as the estimated value of N.
In S<b>145 , the control unit 300 uses the function of the number derivation unit 303 to determine the second candidate value (=third candidate value) as the final estimated value of the number N of railcars of the railroad train 6 .
S115 to S145 are an example of the number derivation step.

(3) Other embodiments:
The above embodiment is an example for carrying out the present invention, and various other embodiments can be adopted. The method of deriving the number of railcars of a railroad train from the displacement at the observation point as in the above embodiment can be implemented as an invention of a program or an invention of a method.

Furthermore, a configuration may be adopted in which the functions of the server device 3 are realized by a plurality of devices. Each function of the server device 3 may be distributed and implemented in a plurality of devices. Also, each function of the server device 3 may be implemented in another device. For example, the functions of the acquisition unit 301 , the time derivation unit 302 , and the number derivation unit 303 may be implemented in the measuring device 1 . A configuration or the like in which the server device 3 is distributed among a plurality of devices may also be used. Furthermore, the above-described embodiment is an example, and an embodiment in which some configurations are omitted or other configurations are added may be adopted.

In the above-described embodiment, the derivation system 10 derives the number of railroad vehicles included in the railroad train 6 in which railroad vehicles that are one or more mobile bodies are organized. However, the derivation system 10 may derive the number of moving bodies included in another organized moving body. For example, the derivation system 10 may derive the number of trolleys contained in a trolley train in which one or more trolleys are connected. Derivation system 10 may also derive the number of vehicles contained in a trailer with one or more vehicles attached.

Further, in the above-described embodiment, the derivation system 10 derives the number of moving bodies included in the train of moving bodies moving on the bridge 5 . However, the derivation system 10 may derive the number of moving bodies included in a train of moving bodies that moves on a structure other than a bridge, such as a concrete foundation that supports a railroad track.

Also, in the above-described embodiment, the number of sensor devices 2 included in the derivation system 10 is two, but the number may be one or three or more.

In the above-described embodiment, the control unit 300 acquires displacement (deflection) data measured from acceleration detected via the acceleration sensor 210 as the time-series data u(t). However, the control unit 300 uses, as u(t), bridge displacement data derived from physical quantities detected via sensors such as impact sensors, pressure sensors, strain gauges, image measuring devices, load cells, and displacement gauges. may be obtained. For example, the control unit 300 detects the displacement of the observation point by periodically photographing a predetermined object placed at the observation point of the bridge 5 via the image measuring device, and acquires the data of the detected displacement. You may Also, the control unit 300 may acquire data of a physical quantity different from the displacement of the bridge as u(t). For example, the control unit 300 may acquire the acceleration, velocity, stress, etc. at the observation point of the bridge 5 as u(t). Also, the control unit 300 acquires, as u(t), the number of pixels corresponding to the amount of displacement of the predetermined object placed at the observation point of the bridge 5 in the image captured by the image measuring device. good. Also, the control unit 300 may acquire data of a plurality of types of physical quantities (for example, displacement, stress, etc.) generated at the observation point as a response to the movement of the railway train 6 on the unit bridge girder.

Further, in the above-described embodiment, the derivation system 10 obtains the first candidate value, the second candidate value, and the third candidate value, respectively, and obtains the obtained first candidate value, the second candidate value, and the third candidate value. A final estimate of N was obtained based on the comparison results of . However, the derivation system 10 may derive any one of the first candidate value, the second candidate value, and the third candidate value as the final estimate of N. In that case, the other two candidate values may not be derived.

Further, in the above-described embodiment, the control unit 300 removes side lobes caused by the influence of the window function used in the FFT from the FFT result for the time-series data u(t) acquired by the function of the acquisition unit 301. , the peak corresponding to the lowest frequency was identified, and the identified peak was obtained as the fundamental frequency _Ff . However, the control unit 300 may obtain the fundamental frequency _Ff considering the influence of noise generated in the FFT result on u(t). For example, the control unit 300, from the FFT result for u (t), excludes side lobes caused by the influence of the window function used in the FFT, and identifies a peak above a predetermined threshold corresponding to the lowest frequency, The identified peak may be determined as the fundamental frequency _Ff .

The time-series data may be data acquired at a data rate that is at least twice the frequency of vibrations that are assumed to occur in structures due to movement of the moving body.

Furthermore, the present invention is also applicable as a computer-executed program or method. Moreover, the above programs and methods may be implemented as a single device or may be implemented using components provided in a plurality of devices, and include various modes. In addition, it can be changed as appropriate, such as a part of which is software and a part of which is hardware. Furthermore, the invention is established as a program recording medium. Of course, the recording medium for the program may be a magnetic recording medium, a semiconductor memory, or the like, and any recording medium that will be developed in the future can be considered in exactly the same way.

DESCRIPTION OF SYMBOLS 1... Measuring device, 2... Sensor device, 3... Server device, 4... Communication network, 5... Bridge, 6... Railroad train, 7... Superstructure, 7a... Bridge floor, 7b... Bearing, 7c... Rail, 7d... Sleeper , 7e... ballast, F... floor slab, G... main girder, 8... lower structure, 8a... bridge pier, 8b... abutment, 10... derivation system, 100... control unit, 110... storage unit, 120... communication unit, 200... Control unit 210 Acceleration sensor 220 Storage unit 230 Communication unit 300 Control unit 301 Acquisition unit 302 Time derivation unit 303 Number derivation unit 310 Storage unit 320 Communication unit

Claims (17)

an acquisition step of acquiring time-series data including physical quantities occurring at a predetermined observation point in a structure in response to movement of an organized mobile body in which one or more mobile bodies are organized;
a time derivation step of deriving an entry time and an exit time from the structure of the mobile organization based on the time-series data;
a number derivation step of deriving the number of the moving bodies included in the organized moving body based on the time-series data, the entering time, and the leaving time;
Derivation methods, including In the time deriving step, the time series data corresponds to two consecutive data sandwiching a predetermined threshold, which are included in the time series data subjected to low-pass filter processing for attenuating vibration components with frequencies equal to or higher than the fundamental frequency of the time series data. The entry time and the exit time are obtained by performing, for each of the entry time and the exit time, a time within a period after one of the two times and before the other time. 2. The derivation method according to claim 1, wherein the time is derived. In the number derivation step, a value obtained by subtracting 1 from a wave number that is a product of a period from the entry time to the exit time in the time series data and a fundamental frequency of the time series data and rounding to an integer is calculated as the 3. The derivation method according to claim 1, wherein the derivation is performed as the number of moving bodies. In the step of deriving the number of positive peaks within the period from the entry time to the exit time in the time series data that has been subjected to a high-pass filter process that attenuates frequency components lower than the fundamental frequency of the time series data 3. The derivation method according to claim 1, wherein a value obtained by subtracting 2 from the number is derived as the number of said moving bodies. In the step of deriving the number, negative peaks within the period from the entry time to the exit time in the time series data subjected to high-pass filtering to attenuate vibration components with frequencies lower than the fundamental frequency of the time series data 3. The derivation method according to claim 1 or 2, wherein a value obtained by subtracting 1 from the number of is derived as the number of said moving bodies. In the number derivation step, a first candidate value obtained by subtracting 1 from a wave number that is a product of the period from the entry time to the exit time in the time series data and the fundamental frequency of the time series data and rounding to an integer. and the number of positive value peaks within the period from the entry time to the exit time in the time series data that has been subjected to a high-pass filter process that attenuates the vibration component with a frequency lower than the fundamental frequency of the time series data A second candidate value from which 2 is subtracted, and a second candidate value from which 1 is subtracted from the number of negative value peaks within the period from the entry time to the exit time in the time-series data subjected to the high-pass filtering process. 3. The derivation method according to claim 1 or 2, wherein the number of the moving bodies included in the organized moving body is derived based on the result of comparison between the three candidate values. In the number derivation step, if the first candidate value and the second candidate value are not equal and the second candidate value and the third candidate value are equal, the second candidate value is used as the organization movement 7. The derivation method according to claim 6, wherein the number is determined as the number of said moving bodies included in a body. In the step of deriving the number, if the second candidate value and the third candidate value are not equal, one of the second candidate value and the third candidate value closer to the value obtained by subtracting 1 from the wave number is selected. 7. The derivation method according to claim 6, wherein the derivation method is derived as the number of the moving bodies included in the organized moving body. The derivation method according to any one of claims 1 to 8, wherein the model of the deflection of the structure is a formula based on the structure of the structure. 10. The derivation method according to any one of claims 1 to 9, wherein the structure is a simple beam with both ends supported. 11. The derivation method according to any one of claims 1 to 10, wherein said structure is a bridge. 12. The derivation method according to any one of claims 1 to 11, wherein the moving body is a railroad vehicle that moves the structure via wheels. The time-series data is based on data detected through at least one of an acceleration sensor, an impact sensor, a pressure sensor, a strain gauge, an image measuring device, a load cell, and a displacement gauge. 13. A derivation method according to any one of claims 1 to 12, which is data. 14. The derivation method according to any one of claims 1 to 13, wherein BWIM (Bridge Weigh in Motion) is applicable to the structure. an acquisition unit that acquires time-series data including physical quantities that occur at a predetermined observation point in a structure in response to movement of an organized mobile body in which one or more mobile bodies are organized;
a time derivation unit for deriving an entry time and an exit time to the structure of the mobile organization based on the time-series data;
a number deriving unit that derives the number of the moving bodies included in the organized moving body based on the time-series data, the entering time, and the leaving time;
A derivation device comprising a A derivation system comprising a derivation device and a sensor,
The derivation device is
A physical quantity generated at a predetermined observation point in a structure in response to movement of an organized mobile body in which one or more mobile bodies are organized, the physical quantity being measured via the sensor. , an acquisition unit that acquires time-series data;
a time derivation unit for deriving an entry time and an exit time to the structure of the mobile organization based on the time-series data;
a number deriving unit that derives the number of the moving bodies included in the organized moving body based on the time-series data, the entering time, and the leaving time;
A derivation system comprising to the computer,
an acquisition step of acquiring time-series data including physical quantities occurring at a predetermined observation point in a structure in response to movement of a structure by an organized mobile body in which one or more mobile bodies are organized;
a time derivation step of deriving an entry time and an exit time to the structure of the mobile organization based on the time-series data;
A number derivation step of deriving the number of the moving bodies included in the organized moving body based on the time-series data, the entering time, and the leaving time;
program to run the

Images