Deterioration detection device, deterioration detection system, deterioration detection method, weight measurement device, weight measurement method, and program

Abstract

The present invention continuously detects deterioration in a bridge. This deterioration detection device comprises: an acquisition unit for collecting time series data pertaining to a parameter that represents a travel-direction site on a subject portion of a bridge where a sensor is provided; an extraction unit for extracting, from the time series data, specific portion data from a time when a specific vehicle passes through a measured segment of the bridge on the basis of an assessment result indicating whether the specific vehicle has passed by, the assessment being made by a neural network that outputs the assessment result upon receiving input of the time series data; an amplitude calculation unit for calculating, on the basis of the specific portion data, an amplitude value pertaining to the amount of travel-direction expansion and contraction of the bridge at the time when the specific vehicle passes by; a deterioration assessment unit for assessing that the bridge has deteriorated when the amplitude value is greater than a preset criterion value; and a re-training command unit for determining whether to re-train the neural network on the basis of pre-established determination criteria, and outputting a re-training command that commands re-training of the neural network on the basis of the result of the determination as to whether to re-train the neural network.

Description

Technical Field

[0001] The present invention relates to a deterioration detection device, a deterioration detection system, a deterioration detection method, a weight measurement device, a weight measurement method, and a program.

Background Art

[0002] Conventionally, bridges have been visually inspected about once every five years. In order to further enhance the safety of bridges, it is desirable to, for example, constantly monitor bridges, predict the future state of bridges based on the results of the monitoring, and manage bridges in a planned manner.

[0003] Patent Document 1 describes a technique for measuring the strain of a floor slab when a vehicle passes over a bridge using a strain gauge and detecting the characteristics of the passing vehicle based on the measured strain. Patent Document 2 describes a technique for calculating the axle ratio of a vehicle from the waveform of the strain measured by a strain gauge and identifying the axle distance, vehicle speed, and vehicle type of the vehicle by comparing the calculated axle ratio with the axle ratio registered in a database. Patent Document 3 describes a technique for calculating a provisional axle load value of a vehicle based on the strain of longitudinal ribs and transverse ribs when a vehicle passes, and correcting the provisional axle load value with the vehicle weight value calculated based on the strain of the transverse ribs. Patent Document 4 describes a technique for detecting the strain of a floor slab when a vehicle passes over a bridge and calculating the weight of the vehicle based on the detected strain. Non-Patent Document 1 describes a technique for installing an acceleration sensor under the rear wheel spring of a route bus to calculate the deflection characteristics of a bridge.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Patent Document 3

Patent Document 4

Non-Patent Document

[0005]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] By the way, as a bridge deteriorates, the strain when a vehicle passes increases. However, it is difficult to easily and continuously measure to what extent the bridge is strained due to the influence of deterioration without relying on personnel.

[0007] In addition, in order to monitor a bridge, it is conceivable to use a model that outputs an output value according to a plurality of input values. However, the detection of bridge deterioration requires long-term monitoring over several years. For this reason, a deviation occurs between the characteristics of the bridge and the characteristics of the model during long-term operation, and it becomes impossible to accurately detect the deterioration of the bridge.

[0008] The present invention has been made in view of the above, and an object thereof is to provide a deterioration detection device, a deterioration detection system, a deterioration detection method, a weight measurement device, a weight measurement method, and a program that can accurately detect bridge deterioration over a long period of time.

Means for Solving the Problems

[0009] In order to solve the above-described problems and achieve the object, a deterioration detection device according to the present invention includes an acquisition unit that collects time-series data of a parameter representing displacement in the traveling direction of a target portion of the bridge where the sensor is provided from a sensor provided on the bridge, and a neural network that inputs the time-series data and outputs a determination result indicating whether or not a specific vehicle has passed. Based on the determination result, an extraction unit that extracts specific portion data when the specific vehicle passes through a measurement section of the bridge from the time-series data, an amplitude calculation unit that calculates an amplitude value of an expansion and contraction amount in the traveling direction of the bridge when the specific vehicle passes based on the specific portion data, a deterioration determination unit that determines that the bridge has deteriorated when the amplitude value becomes larger than a preset reference value, and a re-learning instruction unit that determines whether or not to re-learn the neural network based on a predetermined determination criterion and outputs a re-learning instruction for instructing re-learning of the neural network based on a determination result of whether or not to re-learn the neural network.

Effect of the Invention

[0010] According to the present invention, deterioration of a bridge can be accurately detected over a long period of time.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments will be described with reference to the drawings.

[0013] (First Embodiment) FIG. 1 is a diagram showing a deterioration detection system 10 according to the first embodiment. The deterioration detection system 10 outputs alarm information when the bridge deteriorates.

[0014] The deterioration detection system 10 includes a sensor 20, a transmission device 22, and a deterioration detection device 30.

[0015] The sensor 20 is provided at a predetermined target portion on the bridge. The sensor 20 detects a parameter representing the displacement in the traveling direction at the target portion where the sensor 20 of the bridge is provided. In the present embodiment, the sensor 20 measures the amount of expansion and contraction in the traveling direction at the target portion of the bridge. The amount of expansion and contraction is, for example, a change in distance from several nanometers to about several hundred nanometers between two points at a distance of about several tens of centimeters.

[0016] Note that as long as the sensor 20 can detect a parameter representing the displacement in the traveling direction, instead of the amount of expansion and contraction in the traveling direction, it may detect other physical quantities. For example, the sensor 20 may be a strain gauge that detects the strain in the traveling direction at the target portion of the bridge. Also, for example, the sensor 20 may be a vibration meter that detects the magnitude of the natural vibration frequency in the traveling direction at the target portion of the bridge or the magnitude of the natural vibration frequency in the vertical direction at the target portion of the bridge.

[0017] The sensor 20 continuously detects a parameter representing the displacement in the traveling direction at the target portion of the bridge at predetermined time intervals. For example, the sensor 20 detects the parameter every several milliseconds. The sensor 20 continuously detects the parameter at predetermined time intervals, for example, during the period when the power is turned on. The sensor 20 may continuously detect the parameter at predetermined time intervals constantly, for example, 24 hours a day.

[0018] The transmission device 22 transmits the parameter detected by the sensor 20 to the deterioration detection device 30 via a network. The network may be wired, wireless, or a mixture of wired and wireless. The network is, for example, a LAN (Local Area Network), a VPN (Virtual Private Network), or a WAN (Wide Area Network) to which the LAN is connected via a router. Also, the network may include the Internet or a telephone communication line, etc.

[0019] The deterioration detection device 30 receives time-series data of parameters representing the displacement in the traveling direction of the target portion of the bridge, which is transmitted from the transmission device 22 via the network. The deterioration detection device 30 determines whether the bridge has deteriorated based on the received time-series data of the parameters. Then, when the deterioration detection device 30 determines that the bridge has deteriorated, it outputs alarm information to, for example, an administrator or an information processing device held by the administrator.

[0020] The deterioration detection device 30 is a computer such as a server device connectable to the network. The deterioration detection device 30 may be a single computer, or may be configured by a plurality of computers such as a cloud system.

[0021] Note that the deterioration detection system 10 may be configured not to include the transmission device 22. In this case, the deterioration detection device 30 directly acquires parameters representing the displacement in the traveling direction of the target portion of the bridge from the sensor 20. Also, in this case, the deterioration detection device 30 may be provided near the sensor 20, that is, near the bridge.

[0022] FIG. 2 is a diagram showing the arrangement of the sensor 20 when the bridge is viewed from the side. FIG. 3 is a diagram showing the arrangement of the sensor 20 when the bridge is viewed from above. The sensor 20 is attached, for example, on the end side rather than the center in the traveling direction of the bridge. The sensor 20 is attached, for example, on the lower surface of the bridge near the abutment. Thereby, even after the bridge is completed, the operator can easily attach the sensor 20 to the bridge. Note that the sensor 20 may be attached at any position in the traveling direction of the bridge. For example, although it may be difficult for the operator to attach the sensor 20, the sensor 20 may be attached to the central portion in the traveling direction of the bridge.

[0023] FIG. 4 is a diagram showing a part of the sensor 20 and the main girder 54 of the bridge. The sensor 20 measures the change amount of the distance between the first point 62 and the second point 64 on the lower surface 56 of the main girder 54 of the bridge as the expansion and contraction amount in the traveling direction.

[0024] The first point 62 and the second point 64 are at the same position in the width direction of the bridge and at different positions in the traveling direction. The distance between the first point 62 and the second point 64 is, for example, about several tens of centimeters. In the example of FIG. 4, the distance between the first point 62 and the second point 64 is 35 centimeters. The sensor 20 measures the amount of change in the distance in the traveling direction between the first point 62 and the second point 64 in units of, for example, several nanometers to several hundreds of nanometers.

[0025] The sensor 20 includes a first member 66, a second member 68, and a displacement detection device 70.

[0026] The first member 66 is a cantilever having a support portion 66a and a beam portion 66b. One end of the support portion 66a is a fixed end 66c fixed to the first point 62. The support portion 66a extends downward from the first point 62 by a predetermined distance in a direction perpendicular to the lower surface 56 of the main girder 54. The beam portion 66b extends from the end of the support portion 66a opposite to the fixed end 66c by a predetermined distance toward the second point 64 in the traveling direction. The end of the beam portion 66b on the side not connected to the support portion 66a is a free end 66d not connected to any member. In the present embodiment, the free end 66d of the first member 66 is disposed near the approximate center of the line connecting the first point 62 and the second point 64.

[0027] The second member 68 is a cantilever having a support portion 68a and a beam portion 68b. One end of the support portion 68a is a fixed end 68c fixed to the second point 64. The support portion 68a extends downward from the second point 64 by a predetermined distance in a direction perpendicular to the lower surface 56 of the main girder 54. The beam portion 68b extends from the end of the support portion 68a opposite to the fixed end 68c by a predetermined distance toward the first point 62 in the traveling direction. The end of the beam portion 68b on the side not connected to the support portion 68a is a free end 68d not connected to any member. In the present embodiment, the free end 68d of the second member 68 is disposed near the approximate center of the line connecting the first point 62 and the second point 64.

[0028] Here, the free end 66d of the first member 66 and the free end 68d of the second member 68 are arranged at positions overlapping in the traveling direction without mechanical interference. As a result, the free end 66d of the first member 66 and the free end 68d of the second member 68 are arranged at positions facing each other in a direction perpendicular to the lower surface 56 of the main girder 54. When the distance between the first point 62 and the second point 64 changes, the relative positions of the free end 66d of the first member 66 and the free end 68d of the second member 68 shift in the traveling direction.

[0029] Note that the sensor 20 shown in FIG. 4 was configured such that both the first member 66 and the second member 68 were cantilever beams. However, the second member 68 may be a cantilever beam, and the first member 66 may not be a cantilever beam. In this case, the second member 68 is arranged at a position overlapping in the traveling direction with at least a part of the first member 66 without mechanical interference at the free end 68d. Even in such a configuration, when the distance between the first point 62 and the second point 64 changes, the relative positions of the first member 66 and the free end 68d of the second member 68 shift in the traveling direction.

[0030] The displacement detection device 70 is provided at a portion where the free end 66d of the first member 66 and the free end 68d of the second member 68 face each other. The displacement detection device 70 detects the displacement of the relative positions of the free end 66d of the first member 66 and the free end 68d of the second member 68. Then, the displacement detection device 70 outputs the detected displacement as the amount of expansion and contraction between two points in the traveling direction of the bridge.

[0031] FIG. 5 is a diagram showing the displacement detection device 70 together with the first member 66 and the second member 68. The displacement detection device 70 includes an optical element 72 and a detector 74.

[0032] The optical element 72 is attached to one of the free end 66d of the first member 66 or the free end 68d of the second member 68. The detector 74 is attached to the other of the free end 66d of the first member 66 or the free end 68d of the second member 68 where the optical element 72 is not attached.

[0033] The optical element 72 is an optical member whose reflected light amount or transmitted light amount changes according to the irradiation position of light with respect to the traveling direction. For example, the optical element 72 is a mirror with a plurality of light absorbers applied to its surface at predetermined intervals in the traveling direction. Also, the optical element 72 may be a diffraction grating in which a plurality of optical slits are formed at predetermined intervals in the traveling direction.

[0034] The detector 74 includes a half mirror 76, a light emitting unit 78, a light receiving unit 80, and a detection circuit 82. The half mirror 76 reflects a part of the irradiated light and transmits the other part.

[0035] The light emitting unit 78 irradiates light to the optical element 72 via the half mirror 76. The light receiving unit 80 receives the light reflected by the optical element 72 via the half mirror 76. The detection circuit 82 outputs, as a displacement amount, a signal representing the displacement of the relative position between the first member 66 and the second member 68 based on the change in the light amount of the light detected by the light receiving unit 80.

[0036] The position of the light irradiated to the optical element 72 is shifted in the traveling direction according to the shift in the position in the traveling direction of the relative position between the first member 66 and the second member 68. Since a plurality of light absorbers or a plurality of optical slits arranged in the traveling direction are formed in the optical element 72, the reflected light amount of the optical element 72 increases or decreases according to the shift in the traveling direction at the light irradiation position. Specifically, when the position of the light irradiated to the optical element 72 is shifted by the interval between the plurality of light absorbers or the plurality of optical slits arranged, the increase and decrease of the light amount complete one cycle. Therefore, for example, the detection circuit 82 can obtain the amount of change in the relative position between the first member 66 and the second member 68 by counting the increase and decrease of the signal output from the light receiving unit 80.

[0037] Further, the displacement detection device 70 may include two optical elements 72 that are shifted from each other by 1 / 4 cycle with respect to the pitch of the light absorption material or the optical slit, and two light emitting units 78 and two light receiving units 80 corresponding to the two optical elements 72. Thereby, the two light receiving units 80 can output two periodic signals that are shifted in phase by 1 / 4 cycle with respect to the change in the relative position between the first member 66 and the second member 68. Therefore, for example, the detection circuit 82 can detect the direction of change in the relative position between the first member 66 and the second member 68, and the amount of change in the relative position between the first member 66 and the second member 68 at intervals shorter than the period of the stripe, based on the values of the two signals.

[0038] Also, in the example of FIG. 5, the optical element 72 is configured to reflect light. Instead of this, the optical element 72 may be configured to transmit light. In this case, the amount of transmitted light changes according to the irradiation position of the light with respect to the traveling direction. For example, the optical element 72 may be glass or plastic or the like having a plurality of light absorption materials applied to the surface at a predetermined interval in the traveling direction. In such a case, the light receiving unit 80 receives the light transmitted through the optical element 72.

[0039] Further, the detector 74 may be configured not to include the half mirror 76. Here, it is assumed that the detector 74 is provided on the first member 66. Also, it is assumed that the optical element 72 is provided on the second member 68. And let the position on the first member 66 that faces the center in the width direction of the optical element 72 be P. In such a case, the light emitting unit 78 is arranged at a position on the first member 66 that is shifted by a predetermined distance in the width direction from P. Also, the light receiving unit 80 is arranged at a position on the first member 66 that is shifted by a predetermined distance in the width direction on the side opposite to the light emitting unit 78 from P. The light emitting unit 78 emits light in a direction toward the center in the width direction of the optical element 72. The optical element 72 reflects the incident light in the direction of the light receiving unit 80 when the light from the light emitting unit 78 is incident at a predetermined angle. And the light receiving unit 80 receives the light reflected by the optical element 72. Such a detector 74 can have the same function as the configuration shown in FIG. 5.

[0040] The displacement detection device 70 with such a configuration can be attached to the lower surface 56 of the main girder 54 in a bridge. For example, the displacement detection device 70 can be externally attached without embedding a telescopic member in the bridge like a strain gauge. Thereby, the displacement detection device 70 can be attached later to a completed bridge. Also, the displacement detection device 70 can be attached without degrading the strength of the bridge. Further, the displacement detection device 70 can be easily maintained even after attachment.

[0041] Also, the displacement detection device 70 with such a configuration detects the change in the distance between two points using a cantilever beam by an optical sensor. Thereby, the displacement detection device 70 can accurately detect very small telescopic movements in a bridge using simple components with a low cost.

[0042] FIG. 6 is a diagram showing the functional configuration of the deterioration detection device 30. The deterioration detection device 30 includes an acquisition unit 112, a time-series data storage unit 114, a cutout unit 116, an extraction unit 118, an amplitude calculation unit 120, a deterioration determination unit 122, an alarm output unit 124, a collection unit 132, a re-learning data storage unit 134, a re-learning instruction unit 136, and a re-learning unit 138.

[0043] The acquisition unit 112 collects time-series data of parameters representing the displacement in the traveling direction at the target portion where the sensor 20 of the bridge is provided from the sensor 20 provided on the bridge. In the present embodiment, the acquisition unit 112 acquires the time-series data of the parameters via a network. Also, in the present embodiment, the parameter is the amount of telescopic movement in the traveling direction at the target portion. The time-series data of the parameters has the detected time associated with the parameters.

[0044] The time-series data storage unit 114 stores the time-series data of the parameters collected by the acquisition unit 112.

[0045] The cut-out unit 116 divides the time-series data of the parameters stored in the time-series data storage unit 114 in units of a predetermined time length, and cuts out partial data. Then, the cut-out unit 116 sequentially supplies the cut-out partial data to the extraction unit 118 one by one. The time length of the partial data is at least longer than the time from the point when the change in the amount of expansion and contraction in the traveling direction starts to the point when the change ends by a specific vehicle passing through the measurement section of the bridge.

[0046] The partial data is data of a predetermined number of samples. More specifically, the partial data is data of the number of samples to be input to the first neural network used in the extraction unit 118. In the present embodiment, the cut-out unit 116 cuts out a part of two adjacent partial data in the time direction so as to overlap. That is, each partial data overlaps with the latter half part of the immediately preceding partial data in terms of time and the first half part of the immediately following partial data in terms of time. For example, the first half 1 / 2 of the data in each partial data may be the same as the latter half 1 / 2 of the data in the immediately preceding partial data. Also, the latter half 1 / 2 of the data in each partial data may be the same as the first half 1 / 2 of the data in the immediately following partial data. Thereby, when a specific vehicle passes through the bridge, the cut-out unit 116 can include all of the change from the point when the change in the amount of expansion and contraction in the traveling direction of the bridge starts to the point when the change ends in any of the plurality of partial data.

[0047] Based on the determination result of the first neural network, the extraction unit 118 extracts specific partial data, which is partial data when a specific vehicle passes over a bridge, from the time-series data of the parameters stored in the time-series data storage unit 114. The first neural network inputs the target partial data and determines whether a specific vehicle has passed through the measurement section in the target partial data. The first neural network has previously acquired the time-series data obtained when a specific vehicle passes over a bridge and is trained using the previously acquired time-series data as teacher data. In the present embodiment, the first neural network is a convolutional neural network (CNN). By using the convolutional neural network, the extraction unit 118 can accurately detect that a specific vehicle has passed over a bridge even if the data obtained when the specific vehicle passes over the bridge is included in any time portion of the partial data. Then, in response to obtaining a determination result from the first neural network that a specific vehicle has passed over a bridge, the extraction unit 118 outputs the partial data input to the first neural network as specific partial data.

[0048] The specific vehicle is, for example, a vehicle that regularly passes over a bridge and passes over the bridge at approximately the same speed and approximately the same weight each time. For example, the specific vehicle is a route bus. The route bus travels according to a predetermined timetable every day. Therefore, the route bus is scheduled to pass over the bridge at a predetermined time every day. Also, the specific vehicle may be, for example, a garbage collection vehicle or the like. The garbage collection vehicle has a predetermined route and time of travel. Therefore, the garbage collection vehicle is scheduled to pass over the bridge at a predetermined time on the garbage collection day.

[0049] Each time the extraction unit 118 extracts specific partial data, the amplitude calculation unit 120 calculates an amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when the specific vehicle passes, based on the extracted specific partial data. For example, when the parameter represents the amount of expansion and contraction of the target portion, the amplitude calculation unit 120 calculates the difference between the maximum value and the minimum value in the extracted specific partial data as the amplitude value.

[0050] When the parameter is not the amount of expansion and contraction of the target part, the amplitude calculation unit 120 may output, as the amplitude value, a value correlated with the amplitude value of the amount of expansion and contraction of the target part. For example, when the parameter is the magnitude of the natural frequency, the amplitude calculation unit 120 may calculate, as the amplitude value, the difference between the maximum value and the minimum value in the extracted specific part data.

[0051] Each time the amplitude calculation unit 120 calculates an amplitude value, the deterioration determination unit 122 compares the calculated amplitude value with a preset reference value. Then, when the amplitude value becomes larger than the reference value, the deterioration determination unit 122 determines that the bridge has deteriorated.

[0052] The deterioration determination unit 122 may calculate the moving average value of the amplitude values of a predetermined number of samples in the immediate vicinity, and when the moving average value becomes larger than the reference value, determine that the bridge has deteriorated. Further, the deterioration determination unit 122 may perform a predetermined filtering process such as noise removal processing on the amplitude values of a plurality of sample numbers, and when the result of the filtering process becomes larger than the reference value, determine that the bridge has deteriorated. Note that a plurality of different reference values may be preset in the deterioration determination unit 122. Then, each time the amplitude value becomes larger than each reference value, the deterioration determination unit 122 may determine that the bridge has deteriorated.

[0053] When it is determined that the bridge has deteriorated, the alarm output unit 124 outputs alarm information indicating that the bridge has deteriorated. Note that when a plurality of different reference values are preset in the deterioration determination unit 122, the alarm output unit 124 may acquire alarm information including level information indicating the magnitude of the reference value each time the amplitude value becomes larger than each reference value. Thereby, the alarm output unit 124 can notify the administrator or the like of the level of the deterioration degree of the bridge.

[0054] Each time the extraction unit 118 extracts specific part data, the collection unit 132 collects the extracted specific part data. The re-learning data storage unit 134 stores the specific part data collected by the collection unit 132.

[0055] The relearning instruction unit 136 determines whether to relearn the first neural network based on a predetermined criterion, and outputs a relearning instruction for instructing the relearning of the first neural network based on the determination result of whether to relearn the first neural network. For example, every time the amplitude calculation unit 120 calculates an amplitude value, the relearning instruction unit 136 determines whether to relearn the first neural network based on a predetermined criterion.

[0056] When a relearning instruction is output from the relearning instruction unit 136, the relearning unit 138 relearns the first neural network using the specific partial data collected by the collection unit 132 as teacher data. That is, the relearning unit 138 relearns the first neural network that has already been learned. For example, the relearning unit 138 may perform relearning by the error backpropagation method or the like with the network parameters such as the weights and biases set immediately before relearning as the initial values, or may change the network parameters set immediately before relearning to random values or predetermined values and then perform relearning by the error backpropagation method or the like.

[0057] The relearning unit 138 relearns the first neural network using the specific partial data collected after the last relearning as teacher data. When relearning has not yet been performed since the start of the deterioration determination, the relearning unit 138 may relearn the first neural network using the specific partial data collected after the start as teacher data. Thereby, the relearning unit 138 can relearn the first neural network so as to perform appropriate determination processing according to the state of the most recent bridge. Note that the relearning unit 138 may relearn the first neural network using the time-series data collected in advance as teacher data in addition to the specific partial data collected by the collection unit 132.

[0058] FIG. 7 is a diagram showing an example of time-series data of the amount of expansion and contraction in the traveling direction when a route bus passes over a bridge. When a vehicle passes over a bridge, the bridge changes in the expanding direction with respect to the traveling direction, expands to the maximum value, then changes in the contracting direction, contracts to the minimum value, and then returns to its original state.

[0059] When vehicles of the same type pass over a bridge at approximately the same weight and approximately the same speed, the waveforms of the amount of expansion and contraction of the target part of the bridge are approximately the same if the degree of deterioration of the bridge is the same. For example, when vehicles of the same type pass over a bridge at approximately the same weight and approximately the same speed, the amplitude value (the difference between the maximum value and the minimum value) and the change time (the period from the start time of the change to the end time of the change) in the waveform of the amount of expansion and contraction are approximately the same if the degree of deterioration of the bridge is the same. Note that FIG. 7 is an example in the case where the sensor 20 is provided at the end of the bridge on the vehicle entry side. The sensor 20 may be provided at the end of the bridge on the vehicle exit side. In this case, the bridge changes in the opposite way to that in FIG. 7. That is, in this case, the bridge changes in the contracting direction with respect to the traveling direction, contracts to the minimum value, then changes in the expanding direction, expands to the maximum value, and then returns to its original state.

[0060] FIG. 8 is a diagram showing the time change of the amplitude value and the error range of the amplitude value.

[0061] Bridges deteriorate over time. When the deterioration of a bridge progresses, even when vehicles of the same type pass over the bridge at approximately the same weight and approximately the same speed, the amplitude value of the amount of expansion and contraction becomes larger. That is, the amplitude value of the amount of expansion and contraction in the target part of the bridge when vehicles of the same type pass over the bridge at approximately the same weight and approximately the same speed becomes larger as time passes.

[0062] Even when vehicles of the same type pass over a bridge at approximately the same weight and at approximately the same speed, errors can occur in the amplitude values of the expansion and contraction amounts due to the surrounding environment, measurement conditions, etc. Also, when the specific vehicle is a route bus, the number of passengers and the passing speed also vary from day to day. The amplitude values of the expansion and contraction amounts calculated when a specific vehicle passes over a bridge also include errors due to weight and passing speed.

[0063] However, referring to the experimental results of Non-Patent Document 1, for example, as shown in FIG. 8, it is predicted that the amount of change in the amplitude value due to the aging deterioration of the bridge is sufficiently larger than the error distribution range of the amplitude value calculated when the route bus passes over the bridge. Therefore, when a specific vehicle such as a route bus passes over a bridge and the amplitude value of the expansion and contraction amount of the target portion becomes larger than a preset reference value, if the reference value is set sufficiently larger than the error distribution range of the amplitude value at the start of measurement, it can be said that the bridge has deteriorated compared to the start of measurement.

[0064] Therefore, when the amplitude value of the expansion and contraction amount in the traveling direction of the bridge when a specific vehicle passes is larger than the reference value, the deterioration detection device 30 can determine that the bridge has deteriorated.

[0065] FIG. 9 is a diagram showing the time change of the amplitude value and the threshold for re-learning the first neural network.

[0066] When the data input to the machine learning model changes over time, there is a problem that the accuracy of the neural network deteriorates when it is operated for a long period. This is considered to be due to the deviation between the situation when the neural network was learned and the situation after a long period has passed. Therefore, when the prediction accuracy of the neural network deteriorates, it is re-learned. However, it is difficult to determine to what extent the prediction accuracy has deteriorated during operation.

[0067] When the bridge deteriorates, the amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when a vehicle passes increases over time. Therefore, it is considered that the accuracy of the first neural network into which the amount of expansion and contraction in the traveling direction of the bridge is input deteriorates according to the change in the amplitude value of this amount of expansion and contraction. Thus, the relearning instruction unit 136 outputs a relearning instruction, for example, when the amplitude value of the amount of expansion and contraction in the traveling direction of the bridge becomes larger than a threshold value.

[0068] Thereby, the deterioration detection device 30 can appropriately determine the timing of relearning the first neural network without determining how much the prediction accuracy has deteriorated. Therefore, the deterioration detection device 30 can accurately detect the deterioration of the bridge over a long period.

[0069] For example, a single or a plurality of different threshold values are set in the relearning instruction unit 136. The single or plurality of threshold values are values between the initial value, which is the amplitude value at the start of the determination of deterioration, and the reference value for determining that the bridge has deteriorated. In this case, the relearning instruction unit 136 outputs a relearning instruction every time the amplitude value calculated by the amplitude calculation unit 120 becomes larger than each of the single or plurality of threshold values. Further, the relearning instruction unit 136 may calculate the moving average value of the amplitude values of a predetermined number of most recent samples, and output a relearning instruction every time the moving average value becomes larger than each of the single or plurality of threshold values.

[0070] Also, when a plurality of different reference values are set in advance and the deterioration determination unit 122 determines that the bridge has deteriorated every time the amplitude value becomes larger than each of the reference values, the relearning instruction unit 136 may have each of the plurality of different reference values match the threshold value. Further, in this case, one or more threshold values may be set between the reference values.

[0071] Further, for example, when the calculated amplitude value changes in a preset pattern, the relearning instruction unit 136 may determine to relearn the neural network and output a relearning instruction. For example, when the amplitude value calculated by the amplitude calculation unit 120 changes beyond a preset range since the last relearning, the relearning instruction unit 136 outputs a relearning instruction. Note that when relearning has not been performed yet since the start of the deterioration determination, the relearning instruction unit 136 outputs a relearning instruction when the amplitude value calculated by the amplitude calculation unit 120 changes beyond a preset range since the start.

[0072] For example, when the amplitude value calculated by the amplitude calculation unit 120 changes by a preset ratio since the start or the last relearning, the relearning instruction unit 136 may output a relearning instruction. In this case, the relearning instruction unit 136 may output a relearning instruction when the amplitude value changes by a predetermined ratio or more in the increasing direction, or may also output a relearning instruction when the amplitude value changes by a predetermined ratio or more in the decreasing direction.

[0073] Note that the relearning instruction unit 136 may output a relearning instruction when a predetermined time has elapsed since the last relearning. Note that when relearning has not been performed yet since the start of the deterioration determination, the relearning instruction unit 136 outputs a relearning instruction when a predetermined time has elapsed since the start.

[0074] FIG. 10 is a diagram showing a functional configuration of the deterioration detection device 30 according to a modified example.

[0075] The deterioration detection device 30 may further include any one or a plurality of a passing time acquisition unit 142, a passing detection information acquisition unit 144, a scheduled time estimation unit 146, a date acquisition unit 148, a passing speed acquisition unit 152, and a weight acquisition unit 154.

[0076] When the passage time acquisition unit 142 determines that the specific vehicle is scheduled to cross the bridge at the same first time every day, it acquires information indicating the first time, which is the passage time of the specific vehicle, from an external device or the like. For example, when the specific vehicle is a route bus, the unit may acquire the timetable information of the route bus and calculate the first time based on the acquired timetable information. When the passage time acquisition unit 142 acquires the first time, the extraction unit 118 extracts specific partial data in the time series data of the parameter based on the acquired first time. For example, the extraction unit 118 cuts out a predetermined time range before and after the first time in the time series data of the parameter and extracts the specific partial data from the cut-out range. As a result, the extraction unit 118 only needs to execute the extraction process on the partial data in a part of the time period in the time series data of the parameter, so that the specific partial data can be accurately extracted with a small processing amount.

[0077] The passage detection information acquisition unit 144 acquires a detection signal detected by a passage detection device that detects that the specific vehicle is crossing the bridge. For example, the passage detection device acquires image data from a camera that images the vehicle crossing the bridge, analyzes the acquired image data, and determines whether the specific vehicle is crossing the bridge. When the passage detection device determines that the specific vehicle is crossing the bridge, it gives a detection signal indicating that the specific vehicle is crossing the bridge to the deterioration detection device 30. Further, for example, the passage detection device may be a receiving device that receives identification information for identifying the specific vehicle from a wireless communication device provided in the specific vehicle by a wireless signal. In this case, the passage detection device is provided near the bridge, and when it receives the identification information from the specific vehicle, it gives a detection signal indicating that the specific vehicle is crossing the bridge to the deterioration detection device 30.

[0078] When the passage detection information acquisition unit 144 receives detection information indicating that a specific vehicle is passing over a bridge from the passage detection device, it gives time information indicating the time when the detection information was received to the extraction unit 118. The extraction unit 118 extracts specific partial data in the time-series data of the parameters based on the received time information. For example, the extraction unit 118 cuts out a predetermined time range before and after the time indicated by the received time information from the time-series data of the parameters, and extracts the specific partial data from the cut-out range. Thereby, since the extraction unit 118 only needs to execute the extraction process on the partial data for a part of the time zone in the time-series data of the parameters, it can extract the specific partial data with high accuracy with a small processing amount.

[0079] The scheduled time estimation unit 146 acquires passage information notifying that a specific vehicle has passed a predetermined first position. For example, when the specific vehicle is a route bus, the first position is a bus stop immediately before the bridge or a specific bus stop on the bus route. When the specific vehicle is a route bus, the scheduled time estimation unit 146 receives passage information indicating that it has passed the first position. The passage information is transmitted, for example, from a transmitter provided at the bus stop, a transmitter provided on the route bus, or a management device that manages the operation information of the route bus.

[0080] When the scheduled time estimation unit 146 receives the passage information, it estimates the scheduled time for the specific vehicle to pass over the bridge based on the passage information and the predicted travel time of the specific vehicle from the first position to the bridge. For example, the scheduled time estimation unit 146 calculates, as the scheduled time, the time obtained by adding the predicted travel time to the time when the passage information was received.

[0081] Then, the scheduled time estimation unit 146 provides the estimated scheduled time to the extraction unit 118. Based on the received scheduled time, the extraction unit 118 extracts specific partial data in the time-series data of the parameters. For example, the extraction unit 118 cuts out a predetermined time range before and after the time indicated by the scheduled time from the time-series data of the parameters, and extracts the specific partial data from the cut-out range. Thereby, since the extraction unit 118 only needs to execute the extraction process on the partial data in a part of the time zone in the time-series data of the parameters, it can extract the specific partial data with high accuracy with a small amount of processing.

[0082] The date acquisition unit 148 acquires date information indicating a preset date. The date acquisition unit 148 may acquire the day of the week as the date. In this case, the specific vehicle is a vehicle scheduled to pass over the bridge at the same time every day. The date acquisition unit 148 provides the acquired date information to the amplitude calculation unit 120.

[0083] When the date information is acquired, the amplitude calculation unit 120 calculates an amplitude value based on the specific partial data extracted from the time-series data on the date indicated by the acquired date information. For example, the amplitude calculation unit 120 calculates the amplitude value based on the specific partial data extracted from the time-series data on a weekday date, and does not calculate the amplitude value based on the specific partial data extracted from the time-series data on a holiday date. For example, when the specific vehicle is a route bus, the number of passengers is significantly different between weekdays and holidays, and as a result, the weight may be significantly different. Also, when the specific vehicle is a route bus, the traffic congestion level is different between weekdays and holidays, and the passing speed when passing over the bridge may be significantly different. Therefore, when the specific vehicle is a route bus, the amplitude value of the expansion and contraction amount may have significantly different waveform characteristics between weekdays and holidays.

[0084] Therefore, by calculating the amplitude value based on the specific partial data extracted from the time-series data at a preset date, the amplitude calculation unit 120 can output the amplitude value of the amount of expansion and contraction when vehicles of the same type pass at approximately the same weight and approximately the same speed. Thereby, the deterioration detection device 30 can accurately determine whether the bridge has deteriorated or not.

[0085] The time zone acquisition unit 150 acquires time zone information indicating a preset time zone within a day. The time zone acquisition unit 150 may acquire information such as from 5:00 am to 7:00 am as the time zone. The time zone acquisition unit 150 provides the acquired time zone information to the amplitude calculation unit 120.

[0086] When acquiring the time zone information, the amplitude calculation unit 120 calculates the amplitude value based on the specific partial data extracted from the time-series data in the time zone indicated by the acquired time zone information. For example, the amplitude calculation unit 120 calculates the amplitude value based on the specific partial data extracted from the time-series data in the designated time zone, and does not calculate the amplitude value based on the specific partial data extracted from the time-series data outside the designated time zone. For example, the traffic congestion level varies depending on the time zone within a day, and there may be a significant difference in the passing speed of a specific vehicle passing over the bridge. Therefore, the amplitude value of the amount of expansion and contraction may significantly differ in waveform characteristics depending on the time zone even within a day.

[0087] Therefore, by calculating the amplitude value based on the specific partial data extracted from the time-series data in the preset time zone, the amplitude calculation unit 120 can output the amplitude value of the amount of expansion and contraction when vehicles of the same type pass at approximately the same weight and approximately the same speed. Thereby, the deterioration detection device 30 can accurately determine whether the bridge has deteriorated or not.

[0088] The passing speed acquisition unit 152 acquires the speed when a specific vehicle passes over a bridge. For example, the passing speed acquisition unit 152 may acquire log data measured by a speedometer provided in the specific vehicle. Then, the passing speed acquisition unit 152 may acquire the speed at the time when the specific vehicle passes over the bridge from the log data. The passing speed acquisition unit 152 provides speed information indicating the acquired speed to the amplitude calculation unit 120.

[0089] When acquiring the speed information, the amplitude calculation unit 120 calculates an amplitude value based on specific partial data extracted from time-series data in which the speed when the specific vehicle passes over the bridge is within a preset range. Thereby, the amplitude calculation unit 120 can output the amplitude value of the expansion and contraction amount when vehicles of the same type pass at substantially the same speed. Therefore, the deterioration detection device 30 can more accurately determine whether the bridge has deteriorated.

[0090] The weight acquisition unit 154 acquires the weight when a specific vehicle passes over a bridge. For example, a garbage collection vehicle generally includes a weighing scale for measuring its own weight. The weight acquisition unit 154 may acquire log data measured by a weighing scale provided in a specific vehicle such as a garbage collection vehicle. Then, the weight acquisition unit 154 may acquire the weight at the time when the specific vehicle passes over the bridge from the log data. Also, when the specific vehicle is a route bus, the weight acquisition unit 154 may, for example, acquire the number of passengers at a bus stop immediately before the bridge or a specific bus stop on the bus route, and estimate the weight based on the acquired number of passengers. Then, the weight acquisition unit 154 provides weight information indicating the acquired weight to the amplitude calculation unit 120.

[0091] When acquiring the weight information, the amplitude calculation unit 120 calculates an amplitude value based on specific partial data extracted from time-series data in which the weight when the specific vehicle passes over the bridge is within a preset range. Thereby, the amplitude calculation unit 120 can output the amplitude value of the expansion and contraction amount when vehicles of the same type pass at substantially the same weight. Therefore, the deterioration detection device 30 can more accurately determine whether the bridge has deteriorated.

[0092] FIG. 11 is a flowchart showing the flow of processing of the deterioration detection device 30. As an example, the deterioration detection device 30 executes processing in the flow as shown in FIG. 11.

[0093] First, in S11, the deterioration detection device 30 determines whether or not a route bus, which is a specific vehicle, has passed a bus stop (or a specific bus stop) immediately before a bridge on the bus route. If it has not passed the immediately preceding bus stop (No in S11), the deterioration detection device 30 waits in S11 for the processing. If it has passed the immediately preceding bus stop (Yes in S11), the deterioration detection device 30 advances the processing to S12.

[0094] In S12, the deterioration detection device 30 estimates the scheduled time to pass the bridge. Subsequently, in S13, the deterioration detection device 30 determines whether or not it has reached the time a predetermined time before the scheduled time. If it has not reached the time a predetermined time before the scheduled time (No in S13), the deterioration detection device 30 waits in S13 for the processing. If it has reached the time a predetermined time before the scheduled time (Yes in S13), the deterioration detection device 30 advances the processing to S14.

[0095] In S14, the deterioration detection device 30 turns on the power of the sensor 20. For example, the deterioration detection device 30 gives an instruction signal to the sensor 20 by wireless communication or the like to turn on the power.

[0096] Subsequently, in S15, the deterioration detection device 30 receives and stores the time-series data of the parameters detected by the sensor 20. For example, the deterioration detection device 30 receives and stores the time-series data of the amount of expansion and contraction of the target portion of the bridge.

[0097] Subsequently, in S16, the deterioration detection device 30 determines whether or not it has reached the time a predetermined time after the scheduled time. If it has not reached the time a predetermined time after the scheduled time (No in S16), the deterioration detection device 30 waits in S16 for the processing. If it has reached the time a predetermined time after the scheduled time (Yes in S16), the deterioration detection device 30 advances the processing to S17.

[0098] In S17, the deterioration detection device 30 turns off the power supply of the sensor 20. For example, the deterioration detection device 30 gives an instruction signal to the sensor 20 through wireless communication or the like to turn off the power supply.

[0099] Subsequently, in S18, the deterioration detection device 30 extracts specific partial data, which is partial data when the route bus, which is a specific vehicle, passes over the bridge, from the time-series data of the stored parameters. Subsequently, in S19, the deterioration detection device 30 calculates the amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when the route bus, which is a specific vehicle, passes through, from the extracted specific partial data. For example, when the parameter represents the amount of expansion and contraction of the target part, the deterioration detection device 30 calculates the difference between the maximum value and the minimum value in the extracted specific partial data as the amplitude value. Subsequently, in S20, the deterioration detection device 30 stores the calculated amplitude value.

[0100] Subsequently, in S21, the deterioration detection device 30 determines whether the calculated amplitude value is greater than the reference value. If the amplitude value is not greater than the reference value (No in S21), the deterioration detection device 30 returns the process to S11 and repeats the process from S11. If the amplitude value is greater than the reference value (Yes in S21), the deterioration detection device 30 proceeds to S22.

[0101] In S22, the deterioration detection device 30 outputs alarm information indicating that the bridge has deteriorated to, for example, an administrator or an information processing device held by the administrator. When S22 ends, the deterioration detection device 30 ends this flow. Note that the deterioration detection device 30 may increase the reference value by a predetermined amount after S22 and return the process to S11.

[0102] As described above, the deterioration detection system 10 according to the first embodiment calculates the amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when a specific vehicle such as a road bus passes over the bridge, based on the specific part data at that time. When the calculated amplitude value becomes larger than a preset reference value, it is determined that the bridge has deteriorated, and alarm information indicating that the bridge has deteriorated is output. Thus, according to the deterioration detection system 10 according to the first embodiment, the deterioration of the bridge can be easily and continuously detected over a long period. Therefore, according to the deterioration detection system 10 according to the first embodiment, the state of the bridge can be constantly monitored at low cost, and the bridge can be managed in a planned manner.

[0103] Furthermore, since the deterioration detection system 10 according to the first embodiment retrains the first neural network used for detecting a specific vehicle at an appropriate timing, the deterioration of the bridge can be accurately detected over a long period.

[0104] (Second Embodiment) Next, the weight measurement system 210 according to the second embodiment will be described. In the description of the second embodiment, elements having substantially the same configuration as those in the first embodiment are denoted by the same reference numerals, and detailed description thereof will be omitted except for the differences.

[0105] FIG. 12 is a diagram showing the weight measurement system 210 according to the second embodiment. The weight measurement system 210 accurately measures the weight of a vehicle passing over the bridge while correcting the relationship information according to the deterioration of the bridge.

[0106] The weight measurement system 210 includes a sensor 20, a transmission device 22, and a weight measurement device 230.

[0107] The weight measurement device 230 receives time-series data of parameters representing the displacement in the traveling direction of the target portion of the bridge, which is transmitted from the transmission device 22 via the network. Based on the received time-series data of the parameters, the weight measurement device 230 calculates a vehicle amplitude value representing the amplitude value of the expansion and contraction amount in the traveling direction of the bridge when the vehicle passes. Then, the weight measurement device 230 calculates the weight of the vehicle based on the relationship information representing the correspondence between the amplitude value and the weight and the calculated vehicle amplitude value. Further, when it is determined that the bridge has deteriorated, the weight measurement device 230 corrects the relationship information representing the correspondence between the amplitude value and the weight.

[0108] The weight measurement device 230 is a computer such as a server device connectable to the network. The weight measurement device 230 may be a single computer or may be configured by a plurality of computers such as a cloud system.

[0109] Note that the weight measurement system 210 may be configured not to include the transmission device 22. In this case, the weight measurement device 230 directly acquires parameters representing the displacement in the traveling direction of the target portion of the bridge from the sensor 20. Also, in this case, the weight measurement device 230 may be provided near the sensor 20, that is, near the bridge.

[0110] FIG. 13 is a diagram showing the functional configuration of the weight measurement device 230. The weight measurement device 230 includes an acquisition unit 112, a time-series data storage unit 114, a cutout unit 116, a vehicle extraction unit 242, a vehicle amplitude calculation unit 244, a relationship information storage unit 246, a weight calculation unit 248, an extraction unit 118, an amplitude calculation unit 120, a correction unit 250, a vehicle collection unit 262, a vehicle re-learning data storage unit 264, a collection unit 132, a re-learning data storage unit 134, a re-learning instruction unit 136, a re-learning unit 138, and a vehicle re-learning unit 266.

[0111] The acquisition unit 112, the time-series data storage unit 114, and the cutout unit 116 have the same functions and configurations as those in the first embodiment.

[0112] Based on the determination result by the vehicle determination neural network, the vehicle extraction unit 242 extracts vehicle partial data, which is partial data when the vehicle passes over the bridge, from the time-series data of the parameters stored in the time-series data storage unit 114. Here, the vehicle is not limited to a specific vehicle such as a route bus, and can be any type of vehicle.

[0113] The vehicle determination neural network inputs the target partial data and determines whether the vehicle has passed over the bridge for the target partial data. The vehicle determination neural network has previously acquired the time-series data obtained when the vehicle passes over the bridge, and has been learned using the previously acquired time-series data as teacher data. In the present embodiment, the vehicle determination neural network is a convolutional neural network (CNN). By using the convolutional neural network, the vehicle extraction unit 242 can accurately detect that the vehicle has passed over the bridge even if the waveform data obtained when the vehicle passes over the bridge is included in any time portion of the partial data. Then, in response to obtaining a determination result that the vehicle has passed over the bridge by the vehicle determination neural network, the vehicle extraction unit 242 outputs the partial data input to the vehicle determination neural network as vehicle partial data. Note that the vehicle determination neural network may be a neural network having the same configuration as the first neural network. However, in this case, the parameters included in the vehicle determination neural network have different values from the parameters included in the first neural network as a result of learning.

[0114] Each time the vehicle extraction unit 242 extracts vehicle partial data, the vehicle amplitude calculation unit 244 calculates a vehicle amplitude value representing the amplitude value of the expansion and contraction amount in the traveling direction of the bridge when the vehicle passes, based on the extracted vehicle partial data. For example, when the parameter represents the expansion and contraction amount of the target portion, the vehicle amplitude calculation unit 244 calculates the difference between the maximum value and the minimum value in the extracted vehicle partial data as the vehicle amplitude value.

[0115] When the parameter is not the amount of expansion and contraction of the target part, the vehicle amplitude calculation unit 244 may output, as the vehicle amplitude value, a value correlated with the amplitude value of the amount of expansion and contraction of the target part. For example, when the parameter is the magnitude of the natural frequency, the vehicle amplitude calculation unit 244 may calculate, as the vehicle amplitude value, the difference between the maximum value and the minimum value in the extracted vehicle part data.

[0116] The relationship information storage unit 246 stores relationship information representing the correspondence between the amplitude value and the weight of the vehicle.

[0117] The amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when the vehicle passes over the bridge is correlated with the weight of the vehicle. More specifically, the amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when the vehicle passes over the bridge increases according to the weight of the vehicle. Therefore, an administrator or the like generates in advance relationship information representing the correspondence between the amplitude value and the weight of the vehicle passing over the bridge, and stores it in the relationship information storage unit 246.

[0118] The administrator or the like may generate the relationship information, for example, by running test vehicles of various weights on the bridge and measuring the correspondence between the weight of the vehicle and the amplitude value. Also, the administrator or the like may generate the relationship information by further using the simulation results. Also, the administrator or the like may generate the relationship information of the bridge by using the relationship information applied to other bridges.

[0119] The relationship information storage unit 246 may store, as the relationship information, a table associating each of a plurality of amplitude values with the corresponding weight. Also, the relationship information storage unit 246 may store, as the relationship information, a function or arithmetic expression or the like that outputs the weight by inputting the amplitude value.

[0120] The vehicle weight calculation unit 248 calculates the weight of the vehicle every time the vehicle amplitude calculation unit 244 calculates the vehicle amplitude value. More specifically, the weight calculation unit 248 calculates the weight of the vehicle based on the relationship information representing the correspondence between the amplitude value and the weight stored in the relationship information storage unit 246 and the calculated vehicle amplitude value. The weight calculation unit 248 outputs the calculated weight to, for example, an information processing device or a server held by an administrator. In this case, the weight calculation unit 248 may also output the time when the vehicle passed over the bridge and the like. As a result, an administrator or the like can associate the vehicle that has passed over the bridge with the weight.

[0121] The extraction unit 118 has the same functions and configuration as in the first embodiment, and extracts specific partial data, which is partial data when a specific vehicle passes over a bridge, from the time-series data of the parameters stored in the time-series data storage unit 114 based on the determination result by the first neural network.

[0122] The amplitude calculation unit 120 has the same functions and configuration as in the first embodiment, and calculates a specific vehicle amplitude value representing the amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when the specific vehicle passes, based on the specific partial data every time the extraction unit 118 extracts the specific partial data.

[0123] The correction unit 250 compares the calculated specific vehicle amplitude value with a preset reference value every time the amplitude calculation unit 120 calculates the specific vehicle amplitude value. Then, when the specific vehicle amplitude value is greater than the reference value, the correction unit 250 corrects the correspondence relationship stored in the relationship information storage unit 246.

[0124] The correction unit 250 may calculate the moving average value of the specific vehicle amplitude values of a predetermined number of most recent samples, and correct the relationship information when the moving average value is greater than the reference value. Further, the correction unit 250 may perform a predetermined filtering process such as noise removal processing on the specific vehicle amplitude values of a plurality of sample numbers, and correct the relationship information when the result of the filtering process is greater than the reference value.

[0125] The vehicle collection unit 262 collects the extracted vehicle partial data every time the vehicle extraction unit 242 extracts vehicle partial data. The vehicle relearning data storage unit 264 stores the vehicle partial data collected by the vehicle collection unit 262.

[0126] The collection unit 132 has the same functions and configuration as in the first embodiment, and collects the extracted specific partial data every time the extraction unit 118 extracts specific partial data. The relearning data storage unit 134 has the same functions and configuration as in the first embodiment, and stores the specific partial data collected by the collection unit 132.

[0127] The relearning instruction unit 136 determines whether to relearn the first neural network and the vehicle determination neural network based on the same predetermined judgment criteria as in the first embodiment, and outputs a relearning instruction based on the determination result of whether to relearn the first neural network and the vehicle determination neural network. The relearning instruction unit 136 supplies the relearning instruction to the vehicle relearning unit 266 and the relearning unit 138.

[0128] When a relearning instruction is output from the relearning instruction unit 136, the vehicle relearning unit 266 relearns the vehicle determination neural network using the vehicle partial data collected by the vehicle collection unit 262 as teacher data. That is, the vehicle relearning unit 266 relearns the already learned vehicle determination neural network. For example, the vehicle relearning unit 266 may perform relearning by the error backpropagation method or the like with the network parameters such as the weights and biases set immediately before relearning as the initial values, or may change the network parameters set immediately before relearning to random values or predetermined values and then perform relearning by the error backpropagation method or the like.

[0129] The vehicle relearning unit 266 relearns the vehicle determination neural network using the vehicle partial data collected after the last relearning as teacher data. When relearning has not yet been performed since the start of deterioration determination, the vehicle relearning unit 266 may relearn the vehicle determination neural network using the vehicle partial data collected after the start as teacher data. Thereby, the vehicle relearning unit 266 can relearn the vehicle determination neural network so as to perform appropriate determination processing according to the state of the most recent bridge. Note that the vehicle relearning unit 266 may relearn the vehicle determination neural network using, as teacher data, time-series data collected in advance in addition to the vehicle partial data collected by the vehicle collection unit 262.

[0130] The relearning unit 138 has the same configuration as that in the first embodiment, and when a relearning instruction is output from the relearning instruction unit 136, it relearns the first neural network using the specific partial data collected by the collection unit 132 as teacher data.

[0131] Note that the weight measurement device 230 may further include any one or more of the passage time acquisition unit 142, passage detection information acquisition unit 144, scheduled time estimation unit 146, date acquisition unit 148, time zone acquisition unit 150, passage speed acquisition unit 152, and weight acquisition unit 154 shown in FIG. 10.

[0132] FIG. 14 is a diagram showing the relationship between the weight of the test vehicle and the amplitude value of the expansion and contraction amount of the bridge. In FIG. 14, the graph with black circles represents the relationship when the deterioration of the bridge is small, and the graph with white squares represents the relationship when the deterioration of the bridge becomes larger than a predetermined reference value.

[0133] As the deterioration of the bridge increases, the amplitude value of the expansion and contraction amount in the traveling direction of the bridge when a vehicle passes increases. Therefore, as shown in FIG. 14, when the deterioration of the bridge becomes larger than a reference value, for example, and when the deterioration is smaller than the reference value, the relationship between the weight of the vehicle and the amplitude value of the expansion and contraction amount in the traveling direction of the bridge when the vehicle passes is different. Thus, for example, the administrator generates first relationship information when the deterioration of the bridge is below the reference value and second relationship information when the deterioration of the bridge is larger than the reference value, and stores them in the relationship information storage unit 246.

[0134] Then, when the specific vehicle amplitude value is below the reference value, the correction unit 250 causes the first relationship information to be output from the relationship information storage unit 246 to the weight calculation unit 248. Also, when the specific vehicle amplitude value is larger than the reference value, the correction unit 250 causes the second relationship information to be output from the relationship information storage unit 246 to the weight calculation unit 248. Thereby, the correction unit 250 can accurately measure the weight of the vehicle regardless of the degree of deterioration of the bridge.

[0135] Note that a plurality of different reference values may be set in advance for the correction unit 250. And each time the specific vehicle amplitude value becomes larger than each reference value, the correction unit 250 may adjust the correction amount of the relationship information according to the magnitude of the reference value. Thereby, the correction unit 250 can appropriately correct the relationship information according to the level of deterioration of the bridge.

[0136] FIG. 15 is a flowchart showing the flow of the correction process of the relationship information by the weight measuring device 230. As an example, the weight measuring device 230 executes the process in the flow as shown in FIG. 15.

[0137] Note that the processes from S11 to S21 in FIG. 15 execute the same processes as the deterioration detection device 30 in FIG. 11. However, in the weight measurement system 210, since the weight of the vehicle passing over the bridge is measured, the power supply of the sensor 20 is always on. Therefore, when the sensor 20 used for weight measurement and the sensor 20 used when a specific vehicle passes are the same, the weight measurement device 230 does not have to execute the processes of S13, S14, S16, and S17. However, in this case, the weight measurement device 230, in S18, cuts out a range from a predetermined time before the scheduled time to a predetermined time after the scheduled time from the time-series data of the parameters, and extracts specific partial data from the cut-out range.

[0138] In S21, the weight measurement device 230 determines whether the calculated amplitude value is greater than the reference value. If the amplitude value is greater than the reference value (Yes in S21), the weight measurement device 230 advances the process to S31.

[0139] In S31, the weight measurement device 230 corrects the correspondence relationship stored in the relationship information storage unit 246. When S31 ends, the weight measurement device 230 ends this flow. Note that the weight measurement device 230 may increase the reference value by a predetermined amount after S31 and return the process to S11.

[0140] The weight measurement system 210 according to the second embodiment as described above calculates a vehicle amplitude value representing the amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when a vehicle passes, and based on the relationship information representing the correspondence relationship between the amplitude value and the weight of the vehicle, and the calculated vehicle amplitude value, calculates the weight of the vehicle. Therefore, the weight measurement system 210 can accurately measure the weight of the vehicle passing over the bridge.

[0141] Furthermore, when the specific vehicle amplitude value representing the amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when a specific vehicle passes is greater than the reference value, the weight measurement system 210 corrects the relationship information. Therefore, the weight measurement system 210 can correct the relationship information according to the deterioration of the bridge.

[0142] As described above, according to the weight measurement system 210 according to the second embodiment, it is possible to accurately measure the weight of a vehicle passing over a bridge while correcting the relationship information as the bridge deteriorates.

[0143] (Hardware Configuration of Deterioration Detection Device 30 and Weight Measurement Device 230) FIG. 16 is a diagram showing the hardware configurations of the deterioration detection device 30 and the weight measurement device 230. The deterioration detection device 30 and the weight measurement device 230 are realized, for example, with a hardware configuration similar to that of a general computer. The deterioration detection device 30 and the weight measurement device 230 include a CPU (Central Processing Unit) 301, an operation device 302, a display device 303, a ROM (Read Only Memory) 304, a RAM (Random Access Memory) 305, a storage device 306, a communication device 307, and a bus 309. Each part is connected by the bus 309.

[0144] The CPU 301 executes various processes in cooperation with various programs stored in advance in the ROM 304 or the storage device 306 using a predetermined area of the RAM 305 as a work area, and comprehensively controls the operations of each part constituting the deterioration detection device 30 or the weight measurement device 230. Further, the CPU 301 operates the operation device 302, the display device 303, the communication device 307, etc. in cooperation with programs stored in advance in the ROM 304 or the storage device 306.

[0145] The operation device 302 is an input device such as a touch panel, a mouse, or a keyboard. It receives information input by the user as an operation input as an instruction signal and outputs the instruction signal to the CPU 301.

[0146] The display device 303 is a display unit such as an LCD (Liquid Crystal Display). The display device 303 displays various information based on the display signal from the CPU 301.

[0147] The ROM 304 stores, in a non-rewritable manner, programs and various setting information used for controlling the deterioration detection device 30 or the weight measurement device 230. The RAM 305 is a volatile storage medium such as an SDRAM (Synchronous Dynamic Random Access Memory). The RAM 305 functions as a working area for the CPU 301.

[0148] The storage device 306 is a rewritable recording device such as a semiconductor storage medium like a flash memory, or a magnetic or optical recordable storage medium. The storage device 306 stores programs used for controlling the deterioration detection device 30 or the weight measurement device 230.

[0149] The communication device 307 transmits and receives data with other devices. Also, the communication device 307 may transmit and receive data with a server or the like via a network.

[0150] The programs executed by the deterioration detection device 30 and the weight measurement device 230 are stored, for example, on a computer connected to a network such as the Internet and provided by being downloaded via the network. Also, the programs executed by the deterioration detection device 30 and the weight measurement device 230 may be provided by being pre-embedded in a portable storage medium or the like.

[0151] The program executed by the deterioration detection device 30 has a module configuration including a collection module, a cut-out module, an extraction module, an amplitude calculation module, a deterioration determination module, an alarm output module, a collection module, a relearning instruction module, and a relearning module. The CPU 301 reads such a program from a storage medium or the like and loads the respective modules into the RAM 305. Then, by executing such a program, the CPU 301 functions as an acquisition unit 112, a cut-out unit 116, an extraction unit 118, an amplitude calculation unit 120, a deterioration determination unit 122, an alarm output unit 124, a collection unit 132, a relearning instruction unit 136, and a relearning unit 138. Note that part or all of the acquisition unit 112, extraction unit 118, amplitude calculation unit 120, deterioration determination unit 122, alarm output unit 124, collection unit 132, relearning instruction unit 136, and relearning unit 138 may be configured by hardware.

[0152] The program executed by the weight measurement device 230 has a module configuration including a collection module, a cut-out module, a vehicle extraction module, a vehicle amplitude calculation module, a weight calculation module, an extraction module, an amplitude calculation module, a correction module, a vehicle collection module, a collection module, a relearning instruction module, a vehicle relearning module, and a relearning module. The CPU 301 reads such a program from a storage medium or the like and loads the respective modules into the RAM 305. Then, by executing such a program, the CPU 301 functions as an acquisition unit 112, a cut-out unit 116, a vehicle extraction unit 242, a vehicle amplitude calculation unit 244, a weight calculation unit 248, an extraction unit 118, an amplitude calculation unit 120, a correction unit 250, a vehicle collection unit 262, a collection unit 132, a relearning instruction unit 136, a vehicle relearning unit 266, and a relearning unit 138. Note that part or all of the acquisition unit 112, vehicle extraction unit 242, vehicle amplitude calculation unit 244, weight calculation unit 248, extraction unit 118, amplitude calculation unit 120, and correction unit 250 may be configured by hardware.

[0153] As described above, the embodiments of the present invention have been explained. However, these embodiments are presented as examples and are not intended to limit the scope of the invention. The embodiments can be modified in various ways.

Explanation of Signs

[0154] 10 Deterioration detection system 20 Sensor 22 Transmitter 30 Deterioration detection device 54 Main girder 56 Bottom surface 62 First point 64 Second point 66 First member 68 Second member 70 Displacement detection device 72 Optical element 74 Detector 76 Half mirror 78 Light emitting part 80 Light receiving part 82 Detection circuit 112 Acquisition part 114 Time series data storage part 118 Extraction part 120 Amplitude calculation part 122 Deterioration determination part 124 Alarm output part 132 Collection part 134 Relearning data storage part 136 Relearning instruction part 138 Relearning part 142 Passage time acquisition part 144 Passage detection information acquisition part 146 Scheduled time estimation part 148 Date acquisition part 150 Time zone acquisition part 152 Passage speed acquisition part 154 Weight acquisition part 210 Weight measurement system 230 Weight measurement device 242 Vehicle extraction part 244 Vehicle amplitude calculation part 246 Relationship information storage part 248 Weight calculation unit 250 Correction unit 262 Vehicle collection unit 264 Vehicle relearning data storage unit 266 Vehicle relearning unit

Claims

1. An acquisition unit that collects time-series data of a parameter representing displacement in the traveling direction in a target portion of the bridge where the sensor is provided, from the sensor provided on the bridge; An extraction unit that extracts specific partial data when the specific vehicle passes through the measurement section of the bridge, from the time-series data, based on the determination result indicating whether the specific vehicle has passed through, output by a neural network that inputs the time-series data; An amplitude calculation unit that calculates an amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when the specific vehicle passes through, based on the specific partial data; A deterioration determination unit that determines that the bridge has deteriorated when the amplitude value becomes larger than a preset reference value; A relearning instruction unit that determines whether to relearn the neural network based on a predetermined determination criterion, and outputs a relearning instruction for instructing relearning of the neural network based on the determination result of whether to relearn the neural network; A deterioration detection device comprising the above.

2. An alarm output unit that outputs alarm information indicating that the bridge has deteriorated, when it is determined that the bridge has deteriorated. The deterioration detection device according to claim 1, further comprising the above.

3. One or more threshold values are set between an initial value, which is the amplitude value at the start of the determination of the deterioration of the bridge, and the reference value, in the relearning instruction unit; The relearning instruction unit outputs the relearning instruction every time the amplitude value becomes larger than each of the one or more threshold values. The deterioration detection device according to claim 1 or 2.

4. The relearning instruction unit outputs the relearning instruction when the amplitude value changes in a preset pattern. The deterioration detection device according to claim 1 or 2.

5. When a predetermined time has elapsed since the start of the deterioration determination or the immediately preceding relearning, the relearning instruction unit outputs the relearning instruction. The deterioration detection device according to claim 1 or 2.

6. A collection unit that collects the specific part data; A relearning unit that, when the relearning instruction is output, relearns the neural network using the collected specific part data as teacher data; The deterioration detection device according to any one of claims 1 to 5, further comprising:

7. The relearning unit relearns the neural network using the specific part data collected later since the start or the immediately preceding relearning as the teacher data. The deterioration detection device according to claim 6.

8. The neural network is a convolutional neural network. The deterioration detection device according to any one of claims 1 to 7.

9. A sensor provided in a target part of a bridge, for detecting a parameter representing a displacement in the traveling direction of the target part of the bridge; An acquisition unit that collects time-series data of the parameter from the sensor; Based on the determination result by a neural network that inputs the time-series data and outputs a determination result indicating whether or not a specific vehicle has passed, specific part data when the specific vehicle passes through the measurement section of the bridge is extracted from the time-series data; An amplitude calculation unit that calculates an amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when the specific vehicle passes, based on the specific part data; A deterioration determination unit that determines that the bridge has deteriorated when the amplitude value becomes larger than a preset reference value; Determine whether to retrain the neural network based on a predetermined judgment criterion, and output a retraining instruction for instructing retraining of the neural network based on the judgment result of whether to retrain the neural network. A deterioration detection system comprising.

10. Collect time-series data of a parameter representing the displacement in the traveling direction of the target portion of the bridge where the sensor is provided from a sensor provided on the bridge. Based on the determination result by the neural network that inputs the time-series data and outputs a determination result indicating whether a specific vehicle has passed, extract specific portion data when the specific vehicle passes through the measurement section of the bridge from the time-series data. Based on the specific portion data, calculate an amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when the specific vehicle passes. When the amplitude value becomes larger than a preset reference value, determine that the bridge has deteriorated. Determine whether to retrain the neural network based on a predetermined judgment criterion, and output a retraining instruction for instructing retraining of the neural network based on the judgment result of whether to retrain the neural network. A deterioration detection method.

11. A program for causing an information processing apparatus to function as a deterioration detection apparatus, The information processing apparatus, An acquisition unit that collects time-series data of a parameter representing the displacement in the traveling direction of the target portion of the bridge where the sensor is provided from a sensor provided on the bridge. An extraction unit that extracts specific portion data when the specific vehicle passes through the measurement section of the bridge from the time-series data based on the determination result by the neural network that inputs the time-series data and outputs a determination result indicating whether a specific vehicle has passed. An amplitude calculation unit that calculates an amplitude value of the amount of expansion and contraction in the traveling direction of the bridge when the specific vehicle passes based on the specific portion data. When the amplitude value becomes larger than a preset reference value, a deterioration determination unit that determines that the bridge has deteriorated, and a relearning instruction unit that determines whether to relearn the neural network based on a predetermined determination criterion and outputs a relearning instruction for instructing the relearning of the neural network based on the determination result of whether to relearn the neural network, and a program for causing it to function.

12. an acquisition unit that collects time-series data of a parameter representing the displacement in the traveling direction of a target portion of the bridge where the sensor of the bridge is provided from the sensor provided on the bridge, based on the determination result by a neural network for vehicle determination that inputs the time-series data and outputs a determination result indicating whether a vehicle has passed, vehicle partial data, which is partial data when the vehicle passes through the measurement section of the bridge, is extracted from the time-series data by a vehicle extraction unit, a vehicle amplitude calculation unit that calculates a vehicle amplitude value representing the amplitude value of the expansion and contraction amount of the bridge in the traveling direction when the vehicle passes based on the vehicle partial data, a weight calculation unit that calculates the weight of the vehicle based on relationship information representing the correspondence between the amplitude value and the weight of the vehicle and the calculated vehicle amplitude value, based on the determination result by a neural network that inputs the time-series data and outputs a determination result indicating whether a specific vehicle has passed, specific partial data, which is partial data when the specific vehicle passes through the measurement section of the bridge, is extracted from the time-series data by an extraction unit, an amplitude calculation unit that calculates a specific vehicle amplitude value representing the amplitude value of the expansion and contraction amount of the bridge in the traveling direction when the specific vehicle passes based on the specific partial data, a correction unit that corrects the relationship information when the specific vehicle amplitude value becomes larger than a preset reference value, Based on a predetermined criterion, it is determined whether to retrain the neural network for vehicle determination and the neural network, and a retraining instruction unit outputs a retraining instruction for instructing retraining of the neural network for vehicle determination and the neural network based on the determination result of whether to retrain the neural network for vehicle determination and the neural network. A weighing device comprising the same.

13. Collect time-series data of a parameter representing the displacement in the traveling direction of the target portion of the bridge where the sensor is provided from a sensor provided on the bridge. Based on the determination result by the neural network for vehicle determination that inputs the time-series data and outputs a determination result indicating whether a vehicle has passed, vehicle partial data, which is partial data when the vehicle passes through the measurement section of the bridge, is extracted from the time-series data. Based on the vehicle partial data, a vehicle amplitude value representing the amplitude value of the amount of expansion and contraction of the bridge in the traveling direction when the vehicle passes is calculated. Based on the relationship information representing the correspondence between the amplitude value and the weight of the vehicle and the calculated vehicle amplitude value, the weight of the vehicle is calculated. Based on the determination result by the neural network that inputs the time-series data and outputs a determination result indicating whether a specific vehicle has passed, an extraction unit extracts specific partial data, which is partial data when the specific vehicle passes through the measurement section of the bridge, from the time-series data. Based on the specific partial data, a specific vehicle amplitude value representing the amplitude value of the amount of expansion and contraction of the bridge in the traveling direction when the specific vehicle passes is calculated. When the specific vehicle amplitude value becomes larger than a preset reference value, the relationship information is corrected. Determine whether to retrain the vehicle determination neural network and the neural network based on a predetermined criterion, and output a retraining instruction to instruct retraining of the vehicle determination neural network and the neural network based on the determination result of whether to retrain the vehicle determination neural network and the neural network. Weight measurement method.

14. A program for causing an information processing device to function as a weight measuring device, the information processing device, an acquisition unit that collects time-series data of a parameter representing a displacement in the traveling direction of a target portion of the bridge where the sensor is provided from a sensor provided on the bridge; a vehicle extraction unit that extracts vehicle partial data, which is partial data when the vehicle passes through the measurement section of the bridge, from the time-series data based on the determination result by a vehicle determination neural network that inputs the time-series data and outputs a determination result indicating whether a vehicle has passed; a vehicle amplitude calculation unit that calculates a vehicle amplitude value representing an amplitude value of the amount of expansion and contraction of the bridge in the traveling direction when the vehicle passes, based on the vehicle partial data; a weight calculation unit that calculates the weight of the vehicle based on relationship information representing the correspondence between the amplitude value and the weight of the vehicle and the calculated vehicle amplitude value; an extraction unit that extracts specific partial data, which is partial data when the specific vehicle passes through the measurement section of the bridge, from the time-series data based on the determination result by a neural network that inputs the time-series data and outputs a determination result indicating whether the specific vehicle has passed; an amplitude calculation unit that calculates a specific vehicle amplitude value representing an amplitude value of the amount of expansion and contraction of the bridge in the traveling direction when the specific vehicle passes, based on the specific partial data; a correction unit that corrects the relationship information when the specific vehicle amplitude value becomes larger than a preset reference value; Based on a predetermined criterion, determine whether to relearn the neural network for vehicle determination and the neural network, and output a relearning instruction for instructing the relearning of the neural network for vehicle determination and the neural network based on the determination result of whether to relearn the neural network for vehicle determination and the neural network; A program that causes it to function.

Images