Track multi-source data alignment method, device and track system

Abstract

The present invention provides a track multi-source data alignment method, device and track system, wherein, since the characteristic points in the time domain signal of the reference monitoring device are matched with the event that the predetermined part of the train passes through the reference monitoring device, the precise reference time point when the predetermined part of the train passes through the reference monitoring device can be quickly and accurately obtained, so that the accurate time point when the predetermined part of the train passes through each general monitoring device can be deduced according to the reference time point and the known spacing between each monitoring device, and the corresponding time shift amount can be obtained, thereby realizing automatic time domain alignment of multi-source monitoring data, without the need for manual operation to perform time alignment processing, not only the alignment accuracy is higher, but also for the actual track monitoring scene with a large number of sensors, the time domain alignment of a large amount of data can be automatically realized, so that timely and efficient track monitoring big data analysis can be realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of track monitoring, and in particular relates to a track multi-source data alignment method, device and track system. Background Art

[0002] In the operation of rail transit, it is very important to monitor the track to ensure the stable and safe operation of the train. This monitoring is usually achieved by installing sensors and data collectors on the track to collect data, and further processing and analyzing the collected data. With the development of technology, in order to better monitor the operation of specific tracks and trains, many different types of sensors are now being installed on the track. For example, for floating plate track with elastic vibration reduction structure, a variety of sensors are installed to monitor fastener load, rail load, floating plate vibration acceleration, elastic element working status, etc.

[0003] In actual applications, due to various reasons, such as differences in the installation positions of sensors and the time differences of the self-systems of different data collectors, the data collected by different collectors have time deviations, which brings great difficulties and troubles to the subsequent data processing and analysis. At present, data alignment is mainly done manually, which is not only labor-intensive and resource-intensive, but also has limited alignment accuracy and low efficiency, and cannot meet the needs of track monitoring big data analysis.

[0004] Although there are some data alignment methods in the prior art, these methods are not applicable to the data alignment of multi-source sensors in track monitoring. For example, a commonly used control method is to connect multiple sensors to the same acquisition controller, and control multiple sensors to collect data simultaneously through the acquisition controller, so as to ensure the synchronization of data collection time. However, many sensors used in track monitoring are passive acquisition methods, such as load sensors. When the train passes over the rails above the load sensors, the load sensors collect the corresponding load data, so time synchronization cannot be achieved in this way. Another commonly used method is to send a time synchronization signal to all collectors at a predetermined interval to keep their system time as synchronized as possible, but in track monitoring, the reason for time asynchrony is not only the system time difference of each collector, but also other reasons such as the sensor installation position. This method cannot eliminate the time difference caused by other reasons. Moreover, the number of sensors used in track monitoring is very large. For example, hundreds of load detection sensors and dozens of distance sensors will be arranged in a track section with the length of a floating plate, and the corresponding number of collectors will also be very large. It is difficult to send a time synchronization signal to all collectors at the same time. Summary of the invention

[0005] The present invention is made to solve the above problems, and aims to provide a data alignment method that can efficiently achieve the time alignment of data of multi-source sensors for track monitoring, so as to more conveniently achieve big data analysis, as well as a device using the method, and a track system including the device and multiple monitoring devices. The present invention adopts the following technical solutions:

[0006] The present invention provides a method for aligning multi-source data of a track, which is used for time domain alignment of multi-source monitoring data of a track. The method has the following technical features, and comprises the following steps: step S1, respectively performing corresponding preprocessing on original signals collected by a plurality of monitoring devices to obtain corresponding time domain signals, wherein the plurality of monitoring devices include a general monitoring device and a reference monitoring device arranged at a predetermined position of the track; step S2, searching for predetermined feature points and their corresponding time points in the time domain signals of the reference monitoring device; step S3, obtaining the time point when a predetermined part of the train passes through the reference monitoring device as a reference time point according to the time difference between the plurality of feature points and the train structure; step S4, obtaining the time shift amount of the time domain signal of each general monitoring device according to the reference time point and the predetermined distance between the reference monitoring device and each general monitoring device; and step S5, respectively performing time shifting on each of the time domain signals according to the corresponding time shift amount, so as to obtain the aligned time domain signals of the plurality of monitoring devices.

[0007] The track multi-source data alignment method provided by the present invention may also have such a technical feature, wherein the reference monitoring device is a load monitoring device comprising a plurality of load sensors, which is arranged under the rails of the track and is used to collect load data; in step S2, a peak and its corresponding time point are searched in the time domain signal of the load monitoring device, and in step S3, the time point when the bogie passes through the load monitoring device is obtained as the reference time point based on the time difference between the plurality of peaks and the position distribution of the plurality of bogies in the train.

[0008] The track multi-source data alignment method provided by the present invention may also have such a technical feature, wherein, in step S2, all peak data in the time domain signal of the benchmark monitoring device are found through a peak detection algorithm, including the load value and time point of the peak, and step S3 includes the following sub-steps: step S3-1, based on the time point of the peak, calculating the time difference between two adjacent peaks to obtain a time difference array for each group of the peaks; step S3-2, based on the time difference array, calculating the cross-correlation coefficients between multiple groups of the peaks through a cross-correlation algorithm to obtain a cross-correlation coefficient array; step S3-3, finding the maximum value position in the cross-correlation coefficient array, and obtaining the time point corresponding to the maximum value position as the benchmark time point.

[0009] The track multi-source data alignment method provided by the present invention may also have such a technical feature, wherein step S4 includes the following sub-steps: step S4-1, calculating the time difference between a group of wave peaks generated by the bogie passing through the reference monitoring device as the peak time difference; step S4-2, calculating the train speed of the train based on the wheelbase of the train and the peak time difference; step S4-3, calculating the passing time of the train passing the predetermined distance based on the predetermined distance between the reference monitoring device and the general monitoring device and the train speed; step S4-4, calculating the time shift based on the passing time and the reference time point.

[0010] The track multi-source data alignment method provided by the present invention may also have such a technical feature, wherein the track includes a lower foundation and a plurality of floating track bed plates arranged in sequence, the floating track bed plates are arranged on the lower foundation through a plurality of vibration isolators, and the reference monitoring device is the load monitoring device arranged in the middle of the floating track bed plates.

[0011] The track multi-source data alignment method provided by the present invention may also have such technical features, wherein the multiple monitoring devices include a load monitoring device, which is arranged under the steel rails of the track, and includes multiple load sensors. The collected original data is original load data. For the original load data, step S1 includes the following sub-steps: smoothing the original data to obtain smoothed data; cleaning the smoothed data, including removing outliers, to obtain cleaned data; performing data zeroing processing on the cleaned data to obtain the time domain data.

[0012] The track multi-source data alignment method provided by the present invention may also have such technical features, wherein the multiple monitoring devices include an elastic element monitoring device and a rail displacement monitoring device, the elastic element monitoring device is arranged in the vibration isolator of the track, and it includes multiple ranging sensors, and the collected original data is original vibration isolator distance data, the rail displacement monitoring device includes a rail displacement ranging sensor arranged under the rail and facing the bottom surface of the rail, and the collected original data is original rail displacement distance data. For the original vibration isolator distance data and the original rail displacement distance data, in step S1, before filtering, it also includes: performing conversion calculation on the original signal to convert it from the original signal of electrical signal data to the original signal of digital signal data.

[0013] The track multi-source data alignment method provided by the present invention may also have such a technical feature, wherein the multiple monitoring devices include a vibration acceleration monitoring device, which includes a plurality of vibration acceleration sensors, which are respectively arranged at any one or more of the predetermined positions on the side of the rail of the track, on the roadbed plate, and above the lower foundation, and the original data collected by them is the original vibration acceleration data.

[0014] The present invention provides a track multi-source data alignment device, which is used for time domain alignment of multi-source monitoring data of the track. The device has the following technical features, including: a signal data acquisition unit, which respectively acquires the original signals collected by multiple monitoring devices arranged in the track, wherein the multiple monitoring devices include a reference monitoring device and a general monitoring device arranged at a predetermined position of the track; a data preprocessing unit, which is used to perform corresponding preprocessing on the original signals of the multiple monitoring devices respectively to obtain corresponding time domain signals; a feature point acquisition unit, which is used to search for predetermined feature points and their corresponding time points in the time domain signals corresponding to the reference monitoring devices; a reference time acquisition unit, which is used to obtain the time point when a specific part of the train passes through the reference monitoring device as the reference time point based on the time difference between the multiple feature points and the train structure; a shift amount acquisition unit, which is used to obtain the time shift amount of the time domain signal of each general monitoring device based on the reference time point and the predetermined distance between the reference monitoring device and each general monitoring device; and a time alignment unit, which is used to perform time shift on each of the time domain signals according to the time shift amount, so as to obtain the aligned time domain signals of the multiple monitoring devices.

[0015] The present invention provides a track system, which has the following technical features: a track; a plurality of monitoring devices, which are arranged in the track and are used to collect multi-source monitoring data of the track; and a track multi-source monitoring data alignment device, which is used to perform time domain alignment on the multi-source data, wherein the track multi-source data alignment device is the above-mentioned track multi-source data alignment device.

[0016] The present invention has the following functions and effects.

[0017] According to the method, device and track system for aligning multi-source data of a track provided by the present invention, the method comprises the steps of preprocessing the original signals of multiple sources, searching for predetermined characteristic points in the time domain signals of the reference monitoring device, finding the reference time point when the predetermined part of the train passes through the reference monitoring device, obtaining the time shift amount based on the reference time point and the predetermined device spacing, and translating the time domain signals of the general monitoring device according to the time shift amount to achieve the time domain alignment of the multi-source data. Among them, since the characteristic points in the time domain signals of the reference monitoring device are matched with the event that the predetermined part of the train passes through the reference monitoring device, the precise reference time point when the predetermined part of the train passes through the reference monitoring device can be obtained quickly and accurately, so that the accurate time point when the predetermined part of the train passes through each general monitoring device can be calculated according to the reference time point and the known spacing between each monitoring device, and the corresponding time shift amount can be obtained, thereby achieving the automatic time domain alignment of the multi-source monitoring data, without the need for manual operation for time alignment processing, not only the alignment accuracy is higher, but also for the track monitoring scene with a large number of sensors, the time domain alignment of a large amount of data can be automatically achieved, so that timely and efficient big data analysis can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flow chart of a method for aligning track multi-source data in an embodiment of the present invention;

[0019] Figure 2 is a schematic top view of a track system in an embodiment of the present invention;

[0020] Figure 3 is a cross-sectional schematic diagram of a track system in an embodiment of the present invention;

[0021] Figure 4 is a schematic diagram of the structure of a train in an embodiment of the present invention;

[0022] Figure 5 is a schematic structural diagram of an elastic element monitoring device according to an embodiment of the present invention;

[0023] Figure 6 is a structural diagram of a fixed tooling and a distance measuring sensor in an embodiment of the present invention;

[0024] Figure 7 is an installation position diagram of a load monitoring device in an embodiment of the present invention;

[0025] Figure 8 is a structural exploded view of a load monitoring device in an embodiment of the present invention;

[0026] Fig. 9 is a flow chart of step S1 in an embodiment of the present invention;

[0027] Fig.10is an example diagram of a load time history curve of a train passing through a plate in an embodiment of the present invention;

[0028] Fig.11 is an example diagram of a load time history curve of a train passing through a plate end in an embodiment of the present invention;

[0029] Fig.12 is a flow chart of step S3 in an embodiment of the present invention;

[0030] Fig.13 Schematic diagram of the corresponding relationship between the peak in the load-mileage curve and the bogie in an embodiment of the present invention;

[0031] Fig.14 is a flow chart of step S4 in an embodiment of the present invention;

[0032] Fig.15 is an example diagram of a plurality of aligned load time history curves in an embodiment of the present invention;

[0033] Fig.16 It is a logic block diagram of a track multi-source data alignment device in an embodiment of the present invention.

[0034] Reference numerals:

[0035] Track system 1000; elastic element monitoring device 100; fixing tool 10; base plate 11; flat plate portion 111; clearance notch 1112; fitting protrusion 112; fixing protrusion 113; mounting hole 1131; first sensor bracket 121; first fixing plate portion 1211; first mounting plate portion 1212; second sensor bracket 122; second fixing plate portion 1221; second mounting plate portion 1222; positioning assembly 13; positioning member 131; support member 14; support plate 15; first distance measuring sensor 21; second distance measuring sensor 22; power supply unit 30; data acquisition instrument 40; load monitoring device 200; loading pad 211; sensor fixing plate 212; load sensor 213; cable 214; load transfer column 215; base plate 216; fixing hole 2161; spike mounting hole 2162; fixing plate fastening screw 217; base fixing screw 218; spike 221; vibration acceleration monitoring device 300; rail vibration acceleration sensor 310; roadbed vibration acceleration sensor 320; foundation vibration acceleration sensor 330; rail displacement monitoring device 400; track multi-source data alignment device 600; signal data acquisition unit 610; data preprocessing unit 620; feature point acquisition unit 630; reference time acquisition unit 640; translation acquisition unit 650; time domain alignment unit 660; alignment device communication unit 670; alignment device control unit 680; train 800; carriage 810; bogie 820; wheelset 822; wheel 8221; axle 8222; track 900; lower foundation 910; floating track plate 920; vibration isolator 930; sleeve 931; elastic element 932; unit support plate 933; short sleeper 940; rail 950; rail pad 960; track fastener 970. DETAILED DESCRIPTION

[0036] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the track multi-source data alignment method, device and track system of the present invention are specifically described below in combination with embodiments and drawings.

[0037] Figure 1 4 is a flow chart of the track multi-source data alignment method in this embodiment.

[0038] like Figure 1 As shown, the method includes the following steps.

[0039] Step S1, pre-processing the original signals collected by various monitoring devices respectively to obtain corresponding time domain signals respectively, wherein the various monitoring devices include a reference monitoring device and a general monitoring device set at a predetermined position on the track.

[0040] Step S2, searching for predetermined feature points and their corresponding time points in the time domain signal corresponding to the reference monitoring device.

[0041] Step S3, obtaining the time point when a specific part of the train passes through the reference monitoring device as the reference time point according to the time difference between the multiple feature points and the train structure.

[0042] Step S4, obtaining the time shift of the time domain signal of each general monitoring device according to the reference time point and the predetermined distance between the reference monitoring device and each general monitoring device.

[0043] Step S5, performing time shift on each time domain signal according to the time shift amount, so as to obtain the aligned time domain signals corresponding to the various monitoring devices.

[0044] In order to better illustrate the present method, the algorithm flow of the present solution will be specifically described below by way of example in combination with an actual track structure.

[0045] Figure 2 is a top view of the track system in this embodiment, Figure 3 is a cross-sectional view of the track system in this embodiment.

[0046] like Figure 2 and Figure 3 As shown, taking a floating plate track arranged in a circular tunnel as an example, the track system 1000 includes a track 900, an elastic element monitoring device 100, a load monitoring device 200, a vibration acceleration monitoring device 300, and a rail displacement monitoring device 400.

[0047] The track 900 includes a lower foundation 910 , a plurality of floating track plates 920 , a plurality of vibration isolators 930 , a plurality of short sleepers 940 , and two steel rails 950 .

[0048] The lower foundation 910 is arranged at the bottom of the circular tunnel, and its upper surface is roughly flat. The floating track plate 920 is roughly in the shape of a rectangular plate as a whole, and is a precast concrete plate. Multiple floating track plates 920 are connected end to end and arranged in sequence along the tunnel. The short sleepers 940 are concrete short sleepers, which are arranged on the upper surface of the floating track plate 920 to form two rows, and the multiple short sleepers 940 in each row are arranged at roughly equal intervals. The rails 950 are I-shaped rails, and the two rails 950 are respectively placed on the two rows of short sleepers 940 and fixed by rail fasteners arranged on the short sleepers 940.

[0049] A plurality of vibration isolators 930 are embedded in the floating track plate 920. The vibration isolators 930 include a sleeve 931, an elastic element 932 and other mounting components pre-buried in the floating track plate 920. The elastic element 932 is arranged at the bottom of the sleeve 931, and the lower end of the elastic element 932 extends out of the sleeve 931 and abuts against the lower foundation 910. The vibration isolators 930 are distributed on both sides of the width direction of the floating track plate 920 to form two rows, and each vibration isolator 930 is located between two adjacent short sleepers 940 and is closer to one side edge of the floating track plate 920 in the width direction relative to the short sleepers 940. The floating track plate 920 is arranged on the lower foundation 910 through a plurality of vibration isolators 930, and there is a certain gap between the bottom surface of the floating track plate 920 and the upper surface of the lower foundation 910.

[0050] Optionally, an elastic pad for vibration isolation may be provided between the rail 950 and the short sleeper 940 .

[0051] In this embodiment, the size of the floating track plate 920 is 4710×2600×320mm (length×width×thickness). There are eight short sleepers 940 in a row on each floating track plate 920, and the distance between two adjacent short sleepers 940 is 595mm, which is the distance between two adjacent sets of track fasteners. There are four vibration isolators 930 in a row on each floating track plate 920, two of which are respectively arranged near the two side plate ends (two ends in the length direction) of the floating track plate 920, and the distance between two adjacent vibration isolators 930 is 1190mm. The elastic element 932 in the vibration isolator 930 is a rubber spring or includes a rubber spring.

[0052] Figure 4 is a schematic diagram of the structure of the train in this embodiment. In addition, Figure 3 The structure diagram of the train is also shown in FIG.

[0053] like Figure 3 and Figure 4 As shown, the train 800 includes multiple carriages 810 and multiple bogies 820. Multiple bogies 820 are arranged at the bottom of each carriage 810, and the bogies 820 include a support frame, a wheel pair 822, a spring device and a shock absorber, etc. The wheel pair 822 includes a wheel 8221 and an axle 8222.

[0054] In this embodiment, the train 800 has six carriages 810, and each carriage 810 has two bogies 820 at the bottom. The total length of the train is 118.8m, the vehicle fixed distance (the distance between two bogies at the bottom of a carriage) is 12.6m, and the wheelbase (the distance between two wheel pairs in a bogie, also called fixed wheelbase) is 2300mm.

[0055] Figure 5 is a schematic diagram of the structure of the elastic element monitoring device in this embodiment, Figure 6is a structural diagram of the fixed tooling and the distance measuring sensor in this embodiment.

[0056] like Figure 5 As shown, the vibration isolator 930 also includes a unit support plate 933, which is a metal flat plate, embedded in the sleeve 931 and abutted against a circle of steps near the bottom of the sleeve 931, and the upper end of the elastic unit 932 abuts against the unit support plate 933. The edge of the unit support plate 933 has a plurality of arc-shaped notches. The diameter of the elastic element 932 is smaller than the inner diameter of the sleeve 931, and is arranged approximately coaxially with the sleeve 931.

[0057] The elastic element monitoring device 100 is disposed in the vibration isolator 930 and is used to detect the dynamic deformation of the elastic element 932 (rubber spring). Figure 5 and Figure 6 As shown, the elastic element monitoring device 100 includes a fixing device 10 , a plurality of first distance measuring sensors 21 , a plurality of second distance measuring sensors 22 , corresponding power supply units 30 and a data acquisition device 40 .

[0058] The fixing tool 10 includes a base plate 11 , a plurality of first sensor brackets 121 , a plurality of second sensor brackets 122 , and a positioning assembly 13 .

[0059] The base plate 11 includes a disc-shaped flat plate portion 111, a plurality of engaging protrusions 112 formed on the edge of the flat plate portion 111, and a plurality of fixing protrusions 113. The edge of the flat plate portion 111 also has a plurality of clearance notches 1112.

[0060] The first sensor bracket 121 is L-shaped as a whole, and has a first fixing plate portion 1211 and a first mounting plate portion 1212 which are perpendicular to each other. The first fixing plate portion 1211 is fixedly mounted on the base plate 11, and the first mounting plate portion 1212 is vertically located above the base plate 11 and next to a clearance notch 1112. The first distance measuring sensor 21 is mounted on the outer end of the first mounting plate portion 1212, with its detection end facing downward, and can detect the distance from the sensor to the unit support plate 933 through the clearance notch 1112.

[0061] The structure of the second sensor bracket 122 is substantially the same as that of the first sensor bracket 121, and it has a second fixing plate portion 1221 and a second mounting plate portion 1222 which are perpendicular to each other. The second fixing plate portion 1221 is fixedly mounted on the base plate 11, and the second mounting plate portion 1222 vertically extends from the outside of the base plate 11 to the bottom of the base plate 11. The second distance measuring sensor 22 is mounted on the outer end of the second mounting plate portion 1222, with its detection end facing downward, and can detect the distance from the sensor to the lower foundation through the corresponding arc-shaped notch on the unit support plate 933.

[0062] In this embodiment, the first distance measuring sensor 21 and the second distance measuring sensor 22 are both laser distance measuring sensors.

[0063] The positioning assembly 13 includes a plurality of thin rod-shaped positioning members 131, which are respectively inserted into a plurality of corresponding mounting holes on the base plate 11, and the mounting positions of the positioning members 131 in the corresponding mounting holes are adjustable. When installing the base plate 11, the plurality of positioning members 131 can be used to keep the surface direction of the base plate 11 perpendicular to the axial direction of the sleeve, and the leveling of the base plate 11 can be further achieved by adjusting the positioning members 131.

[0064] The support member 14 is in the shape of a circular cylinder, and has external thread protrusions at both ends. The support plate 15 is in the shape of a circular plate, and is used to support the power supply unit and the data acquisition unit. One end of the support member 14 is installed in the mounting hole in the middle of the flat plate portion 111, and the other end is installed in the mounting hole in the middle of the support plate 15, so that the support plate 15 is installed above the flat plate portion 111, and the support plate 15 is parallel to the flat plate portion 111.

[0065] like Figure 2 As shown, in this embodiment, the eight vibration isolators 930 in a floating ballast plate 920 are all provided with elastic element monitoring devices 100. Optionally, the two vibration isolators 930 at the plate ends of an adjacent floating ballast plate 920 are also provided with elastic element monitoring devices 100.

[0066] Figure 7 2 is an installation position diagram of the load monitoring device in this embodiment.

[0067] like Figure 7 As shown, the overall shape of the load monitoring device 200 is consistent with the track iron pad, and it is installed in the track structure instead of the track iron pad, specifically between the short sleeper 940 and the rail 950. When the rail pad 960 is provided, it is located between the short sleeper 940 and the rail pad 960. The rail fastener 970 is fixed to the load monitoring device 200.

[0068] Figure 8 is a structural exploded view of the load monitoring device in this embodiment.

[0069] like Figure 8 As shown, the load monitoring device 200 includes a loading pad 211, a sensor fixing plate 212, a plurality of load sensors 213 and their cables 214, a plurality of load transfer columns 215, a base plate 216, a plurality of fixing plate fastening screws 217, a plurality of base fixing screws 218, and also includes a corresponding data acquisition instrument (not shown in the figure).

[0070] Among them, the lower surface of the loading pad 211 and the upper surface of the sensor fixing plate 212 have correspondingly arranged multiple sensor fixing grooves and wiring grooves. After being assembled, corresponding accommodating cavities and wiring channels are formed. Multiple load sensors 213 are respectively embedded in each accommodating cavity, and a load transfer column 215 is arranged at the bottom. Its cables 214 are gathered together through the wiring channel and led out from one side of the entire device. According to the distribution of the accommodating cavities, multiple load sensors 213 form two rows, which are respectively located below the two sides of the rail 950.

[0071] The base plate 216 has a plurality of fixing holes 2161 and a pair of spike mounting holes 2162 . The loading pad 211 , the sensor fixing plate 212 , and the base plate 216 are stacked and assembled and fixed on the short rail sleeper 940 through a plurality of spikes 221 .

[0072] The load sensor 213 may be a pressure sensor or a piezoelectric sensor. In this embodiment, it is a pressure sensor. There are six load sensors 213 in one device, forming two rows, with three in each row.

[0073] like Figure 2 As shown, in this embodiment, a load monitoring device 200 is provided at each short sleeper 940 on a floating track plate 920 .

[0074] like Figure 2 and Figure 3 As shown, there are two vibration acceleration monitoring devices 300, which are respectively arranged in the plate (middle in the length direction) of the floating track plate 920, and are respectively located on both sides (both sides in the width direction) of the floating track plate 920. The vibration acceleration monitoring device 300 includes a rail vibration acceleration sensor 310, a roadbed vibration acceleration sensor 320, and a foundation (tunnel wall) vibration acceleration sensor 330, which are respectively used to collect vibration acceleration data at predetermined positions above the rail 950, the floating track plate 920, and the lower foundation 910, and also includes a corresponding data acquisition instrument (not shown in the figure).

[0075] The rail vibration acceleration sensor 310 is installed at the outer side of the rail 950 .

[0076] The track bed vibration acceleration sensor 320 is installed on the upper surface of the floating track plate 920 and is located at the edge in the width direction thereof.

[0077] The foundation vibration acceleration sensor 330 is installed on the tunnel wall above the lower foundation 910.

[0078] like Figure 2 and Figure 3As shown, there are two rail displacement monitoring devices 400, which are also respectively arranged in the middle of the floating track plate 920 and are relatively close to both sides of the floating track plate 920. The rail displacement monitoring device 400 includes a rail displacement ranging sensor and a corresponding data acquisition instrument. The rail displacement ranging sensor is arranged on the upper surface of the floating track plate 920 and is located directly below the rail 950. Its detection end faces upward to the bottom surface of the rail 950, and is used to monitor the displacement of the rail in the middle of the plate.

[0079] The method of the present invention will be described in detail below in conjunction with the above-mentioned track system 1000.

[0080] Step S1, pre-processing the original signals collected by various monitoring devices respectively to obtain corresponding time domain signals respectively, wherein the various monitoring devices include a reference monitoring device and a general monitoring device set at a predetermined position on the track.

[0081] In this embodiment, the load monitoring device 200 arranged in the middle part (middle part in the length direction) of the floating track plate 920 is used as a reference monitoring device, and the load monitoring device 200 and other monitoring devices (elastic element monitoring device 100, vibration acceleration monitoring device 300, rail displacement monitoring device 400) arranged at other positions are used as general monitoring devices.

[0082] Fig. 9 is a flow chart of step S1 in this embodiment.

[0083] like Fig. 9 As shown, the preprocessing of the original signals collected by various monitoring devices is different.

[0084] The load monitoring device 200 collects the track load signal data. For the track load signal data, step S1 includes the following sub-steps.

[0085] Step SA1, downloading load signal data from a data acquisition instrument corresponding to the load monitoring device.

[0086] Step SA2: preview the load signal data.

[0087] Step SA3, smoothing the load signal data, wherein a low-pass filter or a moving average smoothing algorithm may be used to smooth the data to reduce noise.

[0088] Step SA4, performing data cleaning on the smoothed load signal data, including removing abnormal values ​​therein.

[0089] Step SA5, reset the cleaned load signal data to zero, wherein the load signal data is a time domain signal (time history curve), whose two axes are time and load value, respectively. By translating the time history curve as a whole along the axis of the load value, the load value of the time history curve is made to be 0 when the train does not pass through the load monitoring device, thereby eliminating the influence of the train passing on the installation status of the load monitoring device.

[0090] For the elastic element distance signal data collected by the elastic element monitoring device 100 , step S1 includes the following sub-steps.

[0091] Step SB1, downloading elastic element distance signal data, which is electrical signal data, from a data acquisition instrument corresponding to the elastic element monitoring device.

[0092] Step SB2, performing conversion calculation on the elastic element distance signal data of the electrical signal data to convert it into digital signal data.

[0093] Step SB3, previewing the elastic element distance signal data of the digital signal data.

[0094] Step SB4, smoothing the elastic element distance signal data.

[0095] Step SB5, cleaning the smoothed elastic element distance signal data, including removing abnormal values ​​therein.

[0096] Step SB6, returning the distance signal data of the cleaned elastic element to zero.

[0097] The specific operations of steps SB3 to SB6 are substantially the same as those of the above-mentioned steps SA2 to SA5, and thus will not be described again. The corresponding steps of the following other monitoring devices are also the same.

[0098] For the rail displacement distance signal data collected by the rail displacement monitoring device 400, step S1 includes the following sub-steps.

[0099] Step SC1, downloading rail displacement distance signal data, which is electrical signal data, from a data acquisition instrument corresponding to the rail displacement monitoring device.

[0100] Step SC2, converting and calculating the rail displacement distance signal data of the electrical signal data to convert it into digital signal data.

[0101] Step SC3: previewing the rail displacement distance signal data of the digital signal data.

[0102] Step SC4, smoothing the rail displacement distance signal data.

[0103] Step SC5, cleaning the smoothed rail displacement distance signal data, including removing abnormal values ​​therein.

[0104] Step SC6, returning the displacement distance signal data of the cleaned rail to zero.

[0105] For the vibration acceleration signal data collected by the vibration acceleration monitoring device 300, step S1 includes the following sub-steps.

[0106] Step SD1, downloading vibration acceleration signal data from a data acquisition instrument corresponding to the vibration acceleration monitoring device.

[0107] Step SD2, previewing the vibration acceleration signal data.

[0108] Step S2, searching for predetermined feature points and their corresponding time points in the time domain signal corresponding to the reference monitoring device.

[0109] In this embodiment, the reference monitoring device is a load monitoring device 200 arranged in the middle of the floating track plate 920. The characteristic point to be searched is the peak in the load signal data. All peak data in the time domain signal (load signal data) are found through the peak detection algorithm. The peak data includes the coordinates of the peak (i.e., the time point) and the value of the peak (i.e., the load value).

[0110] Fig.10 is an example diagram of the load time history curve collected by the load monitoring device in the plate in this embodiment, Fig.11 It is an example diagram of the load time history curve collected by the load monitoring device at the plate end in this embodiment.

[0111] like Fig.10 As shown, the train passes through the load monitoring device 200 in the middle of the plate. The load time curve collected by the load monitoring device 200 contains multiple peaks, wherein each peak corresponds to a wheel 8221 of the train 800, and two peaks with a relatively close distance correspond to a bogie 820 of the train 800. It can be seen that there is no obvious abnormality in the time curve, and the correspondence between the peak and the train structure is clear and definite. This is because the load monitoring device 200 (pad device) in the middle of the plate is surrounded by other pad devices to bear the load together, so the measured data is relatively stable.

[0112] like Fig.11 As shown in FIG. 1 , the load monitoring device 200 at the end of the train passes through the board. The load time curve collected by the load monitoring device 200 also contains multiple peaks, but the curve is obviously smaller than that of the load monitoring device 200. Fig.10 The curve is more complicated. In addition to the peaks corresponding to the two wheel pairs of the bogie, there are many other relatively small peaks, which makes the relationship between the train structure and the time curve not as good as Fig.10This is because the load monitoring device 200 at the end of the board only has other pad devices to bear the load together on one side close to the middle of the board, but not on the other side, so the stability is relatively worse. On the other hand, when the train passes the end of the board, the board end will be warped by the force, that is, the position close to the middle of the board sinks and the end of the board is lifted, so there will be abnormalities in the collected data.

[0113] Therefore, in this embodiment, the load monitoring device 200 in the middle of the plate is used as the reference monitoring device, and the focus is on determining the corresponding relationship between the time curve measured by the load monitoring device 200 at the plate end and the train structure.

[0114] Step S3, obtaining the time point when a specific part of the train passes through the reference monitoring device as the reference time point according to the time difference between the multiple feature points and the train structure.

[0115] Fig.12 is a flow chart of step S3 in this embodiment.

[0116] like Fig.12 As shown, step S3 specifically includes the following sub-steps:

[0117] Step S3-1, based on the time points of the peaks found, the time difference between two adjacent peaks is calculated to obtain a peak time difference array of each group of peaks.

[0118] Among them, the peak time difference array = [peak 2 coordinate - peak 1 coordinate, peak 3 coordinate - peak 2 coordinate, peak 4 coordinate - peak 3 coordinate, ...].

[0119] Step S3-2, based on the peak time difference array, calculate the cross-correlation coefficients between multiple groups of peaks through a cross-correlation algorithm to obtain a cross-correlation coefficient array.

[0120] Among them, firstly, a cross-correlation coefficient array is created to store the cross-correlation calculation results, and then the cross-correlation coefficients between multiple groups of peaks are calculated, wherein, for each possible lag value (the time lag of each group of peaks relative to the first group of peaks), the product sum of the overlapping parts of the group of peaks and the first group of peaks is calculated as the lag value, and the result of the lag value is stored in the cross-correlation coefficient array as the corresponding cross-correlation coefficient.

[0121] Step S3-3, searching for the maximum value position in the cross-correlation coefficient array, and obtaining the time point corresponding to the maximum value position as the reference time point, so as to determine the best alignment point between the signal data collected by different monitoring devices.

[0122] Fig.13 Schematic diagram of the corresponding relationship between the wave peak in the load time history curve and the bogie in this embodiment.

[0123] like Fig.13 As shown, through step S3, by cross-correlation calculation and finding the maximum value position, the peak value and its coordinates corresponding to the front wheel in the bogie 820 (i.e., the time point when the front wheel passes the load monitoring device 200 in the middle of the board) are found in the load signal data in the middle of the board. Similarly, in the load signal data at the end of the board, the peak value and its time point corresponding to the front wheel in the bogie 820 can also be found, so that the data collected by the load monitoring device 200 in the middle of the board and the data collected by the load monitoring device 200 at the end of the board can be aligned in the time domain using these two time points.

[0124] Cross-correlation calculation and finding the maximum position are very necessary here. Fig.11 As shown, the first peak circled by the dotted circle on the left is an abnormal value. In the time domain signal, such abnormal peaks may be caused by abnormal vibration, warping of the track plate and other factors. If the first peak in the time domain signals of the two monitoring devices is used to perform data time domain alignment, the wheel time point in one time domain signal may be aligned with the abnormal value time point in the other time domain signal, and accurate alignment cannot be achieved.

[0125] Moreover, it can be seen intuitively that the time history curve of the middle of the board is very different from that of the board end. Therefore, if the peak matching algorithm is used, the data of the middle of the board cannot be used to assist the correspondence between the peak value in the time history curve of the board end and the wheel position. However, although the time history curve of the board end is complex, there are also multiple groups of peaks with equal intervals corresponding to each bogie. Therefore, this can be used to accurately identify the peak corresponding to the wheel through the cross-correlation algorithm.

[0126] Step S4, obtaining the time shift of the time domain signal of each general monitoring device according to the reference time point and the predetermined distance between the reference monitoring device and each general monitoring device.

[0127] Fig.14 It is a flow chart of step S4 in this embodiment.

[0128] like Fig.14 As shown, step S4 specifically includes the following sub-steps:

[0129] Step S4-1, calculating the time difference between a group of wave peaks generated by the train bogie passing through the reference monitoring device as the peak time difference.

[0130] Step S4-2, calculating the train speed of the train according to the wheelbase and peak time difference of the train.

[0131] Step S4-3, calculating the time it takes for the train to pass the predetermined distance based on the predetermined distance between the reference monitoring device and the general monitoring device and the train speed.

[0132] The installation position of each monitoring device is fixed, so the distance from the reference monitoring device to each general monitoring device is known based on parameters such as the track size and the designed installation position. For example, the distance from the reference monitoring device (the load monitoring device 200 in the middle of the plate) to another load monitoring device 200 is one or more sleeper spacings, which is n×595mm in this embodiment; for another example, the distance from the reference monitoring device to an elastic element monitoring device 100 is 0.5 sleeper spacing plus one or more sleeper spacings, which is (n+0.5)×595mm in this embodiment. The vibration acceleration monitoring device 300 and the rail displacement monitoring device 400 are both arranged in the middle of the plate, so the distance from the reference monitoring device to them is 0.5 sleeper spacing.

[0133] Step S4-4, calculate the time shift of the time domain signal of the general monitoring device relative to the time domain signal of the reference monitoring device according to the elapsed time and the reference time point, that is:

[0134] Time shift = distance between general monitoring device and reference monitoring device / train speed + offset of reference time point relative to coordinate origin.

[0135] Step S5, time-shifting the time domain signals of each general monitoring device according to the corresponding time shift amount, so as to obtain the aligned time domain signals corresponding to the various monitoring devices.

[0136] That is, the time domain signal of the general monitoring device is shifted as a whole along the time axis according to the time shift amount, so as to obtain an aligned time domain signal.

[0137] Fig.15 is an example diagram of multiple aligned load time history curves in this embodiment, which shows the time domain signals of 18 load monitoring devices 200 after alignment.

[0138] like Fig.15 As shown, in actual track scenarios, the number of monitoring devices will be very large. In a track section used for monitoring, there are 18 load monitoring devices 200 alone, and each load monitoring device 200 includes 6 load sensors. If manual data time domain alignment is required, the workload will be very large. However, the above method can realize automatic time domain alignment of data from a large number of monitoring devices.

[0139] Fig.16 It is a logic block diagram of the track multi-source data alignment device in this embodiment.

[0140] like Fig.16As shown, a device corresponding to the above method can be formed, and the track multi-source data alignment device 600 includes a signal data acquisition unit 610, a data preprocessing unit 620, a feature point acquisition unit 630, a reference time acquisition unit 640, a translation acquisition unit 650, a time domain alignment processing unit 660, an alignment device communication unit 670 and an alignment device control unit 680.

[0141] The signal data acquisition unit 610 is used to acquire the corresponding signal data collected by each monitoring device from each data acquisition instrument.

[0142] The data preprocessing unit 620 preprocesses the acquired signal data using the method of step S1 to obtain a time domain signal corresponding to each monitoring device.

[0143] The feature point acquisition unit 630 uses the method of step S2 to search for predetermined feature points and their corresponding time points in the time domain signal corresponding to the reference monitoring device.

[0144] The reference time acquisition unit 640 obtains the reference time point using the method in step S3 above.

[0145] The shift amount acquisition unit 650 uses the method of step S4 above to obtain the time shift amount of the time domain signal of each general monitoring device based on the reference point and the predetermined monitoring device spacing.

[0146] The time domain alignment processing unit 660 uses the method of step S5 to perform time domain shift on each time domain signal according to the corresponding time shift amount, thereby obtaining the aligned time domain signal of the monitoring device.

[0147] The alignment device communication unit 660 is used to communicate with other devices, for example, the aligned time domain signal can be sent to other computing and analysis devices for further computing and analysis. Optionally, the track multi-source data alignment device 600 may further include a computing and analysis unit for further computing and analysis based on the aligned track multi-source monitoring data. Optionally, the track multi-source data alignment device 600 does not directly connect to each data acquisition instrument and obtain the signal data collected by it, but receives signal data from other devices through the alignment device communication unit 660.

[0148] The alignment device control unit 680 is used to control the operation of the above-mentioned functional units.

[0149] In this embodiment, the track multi-source data alignment device 600 is a laptop computer equipped with a corresponding computer program. The track system 1000 also includes the track multi-source data alignment device 600.

[0150] After achieving time domain alignment of multi-source track monitoring data, the multi-source monitoring data can be applied to realize automatic detection of track defects, optimal design of track structure and train structure, etc. Some application methods are exemplified below.

[0151] For example, the load monitoring device 200 includes two rows of load sensors 213, which are respectively arranged below the two sides of the rail bottom of the rail. Therefore, the load data measured by the two rows of load sensors 213 can be extracted according to the sensor number as the load data of the inner and outer sides of the rail, and compared to analyze the load changes of the inner and outer sides of the rail when the train passes through the load monitoring device 200, so as to timely discover problems in the track structure, such as excessive load on one side of the rail or excessive change, which may correspond to problems such as fastener loosening and failure. In addition, the track structure can also be optimized based on the comparison results. For example, when a rail pad is arranged under the rail, a corresponding variable stiffness rail pad can be designed according to the load difference between the inner and outer sides of the rail and the difference in load change when the train passes.

[0152] For example, the elastic deformation potential energy of each elastic element in a floating plate can be calculated based on the distance data measured by the elastic element monitoring device 100, and the vibration energy of the floating plate can be further calculated. Then, the vibration energy of the floating plate at different times can be compared and analyzed in combination with data such as time point, train speed, and train passenger weight, and the floating plate can be optimized accordingly based on the analysis results.

[0153] For another example, the above-mentioned multiple monitoring devices can also be set up in the curved section of the track. The load data of the inner and outer rails of the curved section of the track can be extracted according to the sensor number, and compared to analyze the load difference and load change between the inner and outer rails when the train passes through the curved section of the track, so as to timely discover problems in the track structure and optimize the design of the track structure. For example, when a rail pad is provided under the rail, rail pads with different stiffness can be designed for the inner and outer rails according to the load difference between the inner and outer rails and the difference in load change when the train passes.

[0154] This embodiment has the following functions and effects.

[0155] According to the track multi-source data alignment method and device provided by this embodiment, the method includes the steps of preprocessing the multi-source original signal, searching for the predetermined feature point in the time domain signal of the reference monitoring device, finding the reference time point when the predetermined part of the train passes through the reference monitoring device, obtaining the time shift based on the reference time point and the predetermined device spacing, and translating the time domain signal of the general monitoring device according to the time shift to achieve multi-source data time domain alignment. Among them, since the feature point in the time domain signal of the reference monitoring device is matched with the event that the predetermined part of the train passes through the reference monitoring device, the precise reference time point when the predetermined part of the train passes through the reference monitoring device can be obtained quickly and accurately, so that the accurate time point when the predetermined part of the train passes through each general monitoring device can be calculated based on the reference time point and the known spacing between each monitoring device, and the corresponding time shift can be obtained, thereby achieving automatic time domain alignment of multi-source monitoring data, without manual operation for time alignment processing, not only the alignment accuracy is higher, but also for the track monitoring scene with a large number of sensors, the time domain alignment of a large amount of data can be automatically achieved, so that timely and efficient big data analysis can be achieved.

[0156] In the embodiment, different preprocessing is performed on the raw data collected by various monitoring devices, including data smoothing, elimination of abnormal data, and data zeroing. After preprocessing, the noise in the raw data can be reduced, and the impact of passing trains on the installation status of each monitoring device can be eliminated, which is conducive to obtaining more accurate alignment results later.

[0157] Furthermore, the load monitoring device installed in the middle of the floating track slab is used as a benchmark monitoring device. Since there are multiple pad devices in the middle of the track slab that bear the load together and are relatively stable, there are basically no abnormal peaks in the time curve collected by the load monitoring device in the middle of the slab, and there are few other abnormal values. The correspondence between multiple peaks and the train bogie is clear and definite. Therefore, it is used as the benchmark monitoring device, and the search for characteristic points (peaks) is more convenient and accurate. Correspondingly, the subsequently calculated train speed and time shift are also more accurate, and the multi-source monitoring data can be more accurately aligned in the time domain.

[0158] Furthermore, the optimal alignment point between signals can be effectively determined through the maximum correlation algorithm (including calculating the cross-correlation results and finding the maximum value position). Due to the structural characteristics of the floating track plate, the time-history curves collected by the load monitoring device in the middle of the plate and the load monitoring device at the end of the plate are very different. It is impossible to use the data in the middle of the plate to assist the correspondence and confirmation of the wheel position in the data at the end of the plate through algorithms such as peak matching. However, by using the characteristic that the bogie will produce two peaks with equal time intervals when passing through the load monitoring device, the maximum correlation algorithm can well determine the optimal alignment point between the signals with very different differences, and then calculate the vehicle speed and other data through the peak data in the middle of the plate, which can assist in confirming the wheel position in the data at the end of the plate.

[0159] Furthermore, since various monitoring devices are fixedly installed in the track structure, the known device spacing, the calculated train speed, and the reference time point when the train passes the reference monitoring device can be used to obtain accurate time shifts with very little calculation complexity, which can be used to perform automated time domain alignment on large-scale multi-source monitoring data.

[0160] In the embodiment, the track is a floating plate track, and the multi-source monitoring data includes track load data, elastic unit related distance data, rail / roadbed plate / tunnel wall vibration acceleration data, and rail displacement distance data. After data alignment processing, these data can be used to analyze the wheel-rail contact, rail deformation, elastic unit installation and deformation, overall vibration reduction effect, etc., which is conducive to keeping the floating plate track in a more ideal state.

[0161] The above embodiments are only used to illustrate the specific implementation of the present invention, and the present invention is not limited to the description scope of the above embodiments. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only to illustrate the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

[0162] For example, in the above embodiment, the monitoring device includes an elastic element monitoring device, a load monitoring device, a vibration acceleration monitoring device, and a rail displacement monitoring device. In an alternative solution, the monitoring device may include only one or more of these types, and the specific structure and installation position of each monitoring device may also be different. For example, for a non-floating ballast plate track, there may be no elastic element monitoring device; for a track using a vibration isolation pad, the elastic element monitoring device may also use different detection elements and different structures.

[0163] In the above embodiment, the load monitoring device in the middle of the plate is used as the reference monitoring device, and the wave crest corresponding to the wheel is found from its time domain signal as the characteristic point for time domain alignment. In an alternative solution, other monitoring devices can also be used as reference monitoring devices, as long as there are relatively easy-to-identify waveform features in their time domain signals that can clearly correspond to specific structures on the train. The characteristic points can also use other waveform features, such as troughs, specific shape curve segments, etc.

[0164] In the above embodiment, the track multi-source data alignment device is a laptop computer. In an alternative solution, it may also be other computing and analysis equipment, and may also be a remote device.

Claims

1. A method for aligning multi-source data of a track, used for aligning multi-source monitoring data of a track in the time domain, characterized in that: The following steps are involved: Step S1, pre-processing the original signals collected by various monitoring devices respectively to obtain corresponding time domain signals, wherein the various monitoring devices include a general monitoring device and a reference monitoring device set at a predetermined position on the track; Step S2, searching for predetermined characteristic points and their corresponding time points in the time domain signal of the reference monitoring device; Step S3, obtaining the time point when the predetermined part of the train passes through the reference monitoring device as the reference time point according to the time difference between the plurality of characteristic points and the train structure; Step S4, obtaining the time shift of the time domain signal of each general monitoring device relative to the time domain signal of the reference monitoring device according to the reference time point and the predetermined distance between the reference monitoring device and each general monitoring device; and Step S5, performing time domain shift on each of the time domain signals according to the corresponding time shift amount, so as to obtain aligned time domain signals of the multiple monitoring devices. The reference monitoring device is a load monitoring device including a plurality of load sensors, which is arranged under the rails of the track and is used to collect load data. In step S2, all peak data in the time domain signal of the reference monitoring device are found through a peak detection algorithm, including the peak load value and the corresponding time point. In step S3, the time point at which the bogie passes through the load monitoring device is obtained as the reference time point according to the time difference between the plurality of wave peaks and the position distribution of the plurality of bogies in the train. Step S3 includes the following sub-steps: Step S3-1, based on the time points of the peaks, calculating the time difference between two adjacent peaks to obtain a peak time difference array of each group of peaks; Step S3-2, based on the peak time difference array, calculating the cross-correlation coefficients between the plurality of groups of the peaks by a cross-correlation algorithm to obtain a cross-correlation coefficient array; Step S3-3, searching for the maximum value position in the cross-correlation coefficient array, and obtaining the time point corresponding to the maximum value position as the reference time point.

2. The track multi-source data alignment method according to claim 1, Features: Wherein, step S4 includes the following sub-steps: Step S4-1, calculating the time difference between a group of wave peaks generated when the bogie passes through the reference monitoring device as the peak time difference; Step S4-2, calculating the train speed of the train according to the wheelbase of the train and the peak time difference; Step S4-3, calculating the time it takes for the train to pass the predetermined distance based on the predetermined distance between the reference monitoring device and the general monitoring device and the train speed; Step S4-4, calculating the time shift according to the elapsed time and the reference time point.

3. The track multi-source data alignment method according to claim 1, characterized in that: in, The track includes a lower foundation and a plurality of floating ballast plates arranged in sequence, wherein the floating ballast plates are arranged on the lower foundation through a plurality of vibration isolators. The reference monitoring device is the load monitoring device arranged in the middle of the floating track bed slab.

4. The track multi-source data alignment method according to claim 1, Features: in, The various monitoring devices include a load monitoring device, which is arranged under the rails of the track and includes a plurality of load sensors. The collected original signals are original load data. For the original load data, step S1 includes the following sub-steps: Performing data smoothing on the original signal to obtain smoothed data; Cleaning the smoothed data, including removing outliers, to obtain cleaned data; The cleaned data is subjected to data zeroing processing to obtain the time domain signal.

5. The track multi-source data alignment method according to claim 4, characterized in that: in, The various monitoring devices include elastic element monitoring devices and rail displacement monitoring devices, The elastic element monitoring device is arranged in the vibration isolator of the track, and comprises a plurality of distance measuring sensors. The original signal collected is the original vibration isolator distance data. The rail displacement monitoring device comprises a rail displacement distance measuring sensor arranged below the rail and facing the bottom surface of the rail, and the original signal collected by the sensor is the original rail displacement distance data. For the original vibration isolator distance data and the original rail displacement distance data, in step S1, before filtering, the following is further included: The original signal is converted by calculation to convert it from the original signal of electrical signal data to the original signal of digital signal data.

6. The track multi-source data alignment method according to claim 4, characterized in that: in, The various monitoring devices include a vibration acceleration monitoring device, which includes a plurality of vibration acceleration sensors, which are respectively arranged at any one or more predetermined positions on the side of the rail of the track, on the roadbed plate, and above the lower foundation. The original signal collected by the device is the original vibration acceleration data.

7. A track multi-source data alignment device, used for performing time domain alignment on track multi-source monitoring data, characterized in that: include: A signal data acquisition unit acquires original signals collected by multiple monitoring devices arranged in the track, wherein the multiple monitoring devices include a reference monitoring device and a general monitoring device arranged at a predetermined position of the track; A data preprocessing unit, used for preprocessing the original signals of the plurality of monitoring devices respectively to obtain corresponding time domain signals; A feature point acquisition unit, used to search for predetermined feature points and their corresponding time points in the time domain signal corresponding to the reference monitoring device; A reference time acquisition unit, used for obtaining a time point when a specific part of the train passes through the reference monitoring device as a reference time point according to the time difference between the plurality of feature points and the train structure; a shift amount acquiring unit, configured to obtain a time shift amount of a time domain signal of each of the general monitoring devices according to the reference time point and a predetermined distance between the reference monitoring device and each of the general monitoring devices; and A time domain alignment unit is used to perform time domain shift on each of the time domain signals according to the time shift amount, so as to obtain aligned time domain signals of the multiple monitoring devices. The reference monitoring device is a load monitoring device including a plurality of load sensors, which is arranged under the rails of the track and is used to collect load data. The characteristic point acquisition unit finds all the peak data in the time domain signal of the reference monitoring device through a peak detection algorithm, including the peak load value and the corresponding time point. The reference time acquisition unit obtains the time point when the bogie passes through the load monitoring device as the reference time point based on the time difference between the multiple peaks and the position distribution of the multiple bogies in the train, specifically including: based on the time point of the peak, calculating the time difference between two adjacent peaks to obtain a peak time difference array for each group of the peaks; based on the peak time difference array, calculating the cross-correlation coefficients between the multiple groups of peaks by a cross-correlation algorithm to obtain a cross-correlation coefficient array; searching for the maximum value position in the cross-correlation coefficient array, and obtaining the time point corresponding to the maximum value position as the reference time point.

8. A track system, characterized in that: include: track; A plurality of monitoring devices are arranged in the track and used for collecting multi-source monitoring data of the track; as well as A track multi-source data alignment device is used to perform time domain alignment on the multi-source monitoring data. Wherein, the track multi-source data alignment device is the track multi-source data alignment device described in claim 7.

Images