Railway foundation settlement management early warning method and system

Abstract

The present application is suitable for the field of railway foundation management, and provides a railway foundation settlement management early warning method and system, the method comprising: installing a vehicle-mounted displacement sensor at the target position of a running vehicle; deploying a continuously distributed optical fiber sensing monitor group and a temperature sensor inside the railway line; establishing a space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network, fusing the real-time collected vehicle body longitudinal displacement data with the optical fiber strain data of the corresponding mileage section, and calculating the longitudinal change degree of the foundation in the monitoring section; setting a multi-level threshold trigger and a recheck strategy; fusing multi-source data, generating and sending a ladder intelligent early warning notice based on the multi-level threshold trigger and the recheck strategy; the present application constructs a "vehicle-line-foundation" three-in-one intelligent monitoring system, significantly improving the space-time resolution and accuracy of railway foundation settlement monitoring. The multi-level threshold and recheck mechanism reduces the false alarm and missed alarm rate, and the ladder type early warning realizes risk grading control.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of railway foundation management, and particularly relates to a railway foundation settlement management and early warning method and system. BACKGROUND

[0002] Railway foundation settlement monitoring is a key technology to ensure the safe operation of trains. The existing monitoring means mainly rely on fixed-point sensors or distributed optical fiber sensing systems. However, traditional point monitoring has the problem of insufficient spatial resolution and difficulty in capturing local sudden settlement. Although pure distributed optical fiber monitoring can achieve continuous measurement, it is limited by the long distance characteristics of railway lines, and there is a time lag effect on the instantaneous settlement response under dynamic vehicle load.

[0003] In addition, the existing technology lacks coupling analysis of vehicle dynamic displacement and foundation static settlement, and cannot accurately identify the abnormal longitudinal displacement of the vehicle body caused by uneven foundation settlement during vehicle driving, which is easy to cause missed detection or false alarm, and is difficult to meet the real-time requirement of millimeter-level settlement monitoring for high-speed rail operation. In view of the problems of dynamic load response lag and insufficient multi-source data fusion in the prior art, it is urgent to develop a more mature railway foundation settlement management and early warning method and system. SUMMARY

[0004] The purpose of the present application is to provide a railway foundation settlement management and early warning method and system, which aims to solve the problems raised in the background art.

[0005] The present application is realized in this way. On the one hand, a railway foundation settlement management and early warning method, the method comprising:

[0006] Installing a vehicle-mounted displacement sensor at the targeted position of the driving vehicle to collect real-time vehicle body longitudinal displacement data during vehicle driving, the targeted position including the axle box of the bogie, the central section of the vehicle body frame and the vehicle body directly above the rail contact surface;

[0007] Deploying a continuous distributed optical fiber sensing monitor group and a temperature sensor inside the railway line to collect real-time continuous strain distribution data and temperature data of the railway foundation, the continuous distributed optical fiber sensing monitor group comprising an optical fiber sensing network laid along the railway line;

[0008] Establishing a space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network, fusing the real-time vehicle body longitudinal displacement data and the corresponding optical fiber strain data of the mileage section, and calculating the longitudinal variation degree of the foundation in the monitoring interval;

[0009] Setting a multi-level threshold trigger and review strategy;

[0010] Fusing multi-source data, generating and sending a ladder intelligent early warning notice based on the multi-level threshold trigger and review strategy.

[0011] As a further scheme of the present application, the space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network fuses the real-time collected vehicle body longitudinal displacement data with the optical fiber strain data of the corresponding mileage section, and calculates the ground longitudinal variation degree in the monitoring interval, which specifically includes:

[0012] A space-time mapping relationship is established based on the vehicle-mounted GPS time stamp of the running vehicle and the route mileage marker;

[0013] The vehicle-mounted displacement data is denoised by a Kalman filtering algorithm, and the optical fiber strain data is simultaneously temperature-compensated;

[0014] The compensated optical fiber strain value is calculated as follows:

[0015] ;

[0016] In the formula, is the original optical fiber strain measurement value without temperature compensation, is the optical fiber thermal expansion coefficient, is the real-time temperature change amount.

[0017] As a further scheme of the present application, the space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network fuses the real-time collected vehicle body longitudinal displacement data with the optical fiber strain data of the corresponding mileage section, and calculates the ground longitudinal variation degree in the monitoring interval, which specifically includes:

[0018] Based on a dynamic time warping algorithm, the vehicle-mounted longitudinal displacement curve is phase-aligned with the optical fiber strain curve of the corresponding mileage section;

[0019] Based on a difference algorithm, the ground longitudinal deformation gradient value in the monitoring interval is calculated;

[0020] The calculation process of the ground longitudinal deformation gradient value is as follows:

[0021] ;

[0022] In the formula, is the longitudinal deformation gradient of the monitoring surface, is the longitudinal displacement amount of the vehicle-mounted sensor at the corresponding section, is the section spacing, is the vertical strain conversion displacement calculated by the optical fiber sensing monitor group.

[0023] As a further scheme of the present application, the multi-level threshold triggering and review strategy specifically comprises:

[0024] The preset two-level longitudinal deformation degree threshold values are , the primary deformation degree threshold value is , and the emergency deformation degree threshold value is

[0025] When the ground longitudinal deformation gradient value of three consecutive sampling periods exceeds the primary deformation degree threshold value , a regional-level review instruction is triggered;

[0026] Based on the regional-level review instruction, the optical fiber static water level sensor group arranged in the monitoring section is controlled to measure the pressure change of each reference point in real time through the liquid level communication pipe;

[0027] The relative settlement amount of each reference point is calculated ;

[0028] The calculation process of the relative settlement amount is as follows:

[0029] ;

[0030] In the formula, is the pressure change value, is the density of the communication liquid, 9.81 ;

[0031] When the ground longitudinal deformation gradient value exceeds the emergency deformation degree threshold value , a full-line level real-time scanning is triggered.

[0032] As a further scheme of the present application, the multi-level threshold triggering and review strategy specifically comprises:

[0033] The fusion warning model including the back propagation neural network is constructed, and the input parameters of the fusion warning model include the vehicle-mounted longitudinal displacement amount, the optical fiber strain value, the liquid level pressure change rate, the environment temperature and humidity, and the train running speed;

[0034] When the relative settlement amount of any reference point exceeds the target threshold value, a ladder classification warning signal is generated, and the ballastless section threshold value is set, and the ballasted section threshold value ;

[0035] In the ballastless section, the ladder classification warning signal specifically includes:

[0036] Yellow warning: when When the longitudinal displacement of the vehicle body exceeds the first threshold value, the artificial inspection instruction is generated and sent;

[0037] Orange warning: when the longitudinal displacement of the vehicle body exceeds the second threshold value, the speed reduction instruction is generated and sent;

[0038] Orange warning: when the longitudinal displacement of the vehicle body exceeds the third threshold value, the emergency stop instruction is generated and sent.

[0039] As a further scheme of the present application, in another aspect, a railway foundation settlement management early warning system, the system comprises:

[0040] The first real-time acquisition module is used for installing a vehicle-mounted displacement sensor at a targeted position of a running vehicle, and collecting vehicle body longitudinal displacement data in a real-time manner during vehicle running;

[0041] The targeted position includes a bogie axle box, a central section of a vehicle body underframe, and a vehicle body directly above a rail contact surface;

[0042] The second real-time acquisition module is used for deploying a continuous distributed optical fiber sensing monitor group and a temperature sensor inside a railway line, and collecting continuous strain distribution data and temperature data of a railway foundation in a real-time manner;

[0043] The continuous distributed optical fiber sensing monitor group comprises an optical fiber sensing network laid along the railway line;

[0044] The space-time synchronization calibration model module is used for establishing a space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network;

[0045] The calculation module is used for fusing the real-time collected vehicle body longitudinal displacement data and the optical fiber strain data of the corresponding mileage section, and calculating the longitudinal variation degree of the foundation in the monitoring interval;

[0046] The control module is used for setting a multi-level threshold trigger and a recheck strategy;

[0047] The generation and sending module is used for fusing multi-source data, generating and sending a step-by-step intelligent early warning notice based on the multi-level threshold trigger and the recheck strategy.

[0048] As a further scheme of the present application, the space-time synchronization calibration model module specifically comprises:

[0049] The mapping unit is used for establishing a space-time mapping relationship based on a vehicle-mounted GPS timestamp of a running vehicle and a line mileage mark;

[0050] The synchronization compensation unit is used for carrying out denoising processing on the vehicle-mounted displacement data through a Kalman filtering algorithm, and synchronously carrying out temperature compensation on the optical fiber strain data.

[0051] As a further scheme of the present application, the calculation module comprises: ​​

[0052] a phase alignment unit, configured to perform phase alignment between the vehicle-mounted longitudinal displacement curve and the fiber strain curve of the corresponding mileage section based on a dynamic time warping algorithm;

[0053] a first calculation unit, configured to calculate the ground longitudinal deformation gradient value in the monitoring section based on a difference algorithm .

[0054] As a further scheme of the present application, the management and control module specifically comprises:

[0055] a two-level longitudinal change degree threshold unit, configured to preset two-level longitudinal change degree thresholds;

[0056] a re-inspection unit, configured to trigger a regional-level re-inspection instruction when the ground longitudinal deformation gradient value in three consecutive sampling periods exceeds the primary change degree threshold .

[0057] a real-time measurement unit, configured to control the fiber static level sensor group arranged in the monitoring section to measure the pressure change of each reference point in real time through the liquid level communication pipe based on the regional-level re-inspection instruction;

[0058] a second calculation unit, configured to calculate the relative settlement of each reference point .

[0059] a whole-line-level real-time scanning unit, configured to trigger whole-line-level real-time scanning when the ground longitudinal deformation gradient value exceeds the emergency change degree threshold .

[0060] As a further scheme of the present application, the generation and sending module specifically comprises:

[0061] a fusion early warning model unit, configured to construct a fusion early warning model comprising a back propagation neural network;

[0062] The input parameters of the fusion early warning model include the vehicle-mounted longitudinal displacement, the fiber strain value, the liquid level pressure change rate, the environmental temperature and humidity, and the train running speed;

[0063] a generation unit, configured to generate a ladder classification early warning signal when the relative settlement of any reference point exceeds the target threshold.

[0064] The application provides a railway foundation settlement management early warning method and system, the method and system of the application construct a "vehicle-line-foundation" three-in-one intelligent monitoring system, significantly improving the spatial and temporal resolution and accuracy of railway foundation settlement monitoring. Multi-level threshold and recheck mechanism reduce the false alarm and missed alarm rate, and the step-by-step early warning realizes risk grading control, effectively guaranteeing train operation safety. The intelligent early warning system links railway dispatching and operation and maintenance, improving emergency response efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0065] Figure 1 It is a main flowchart of a railway foundation settlement management early warning method.

[0066] Figure 2 It is a flowchart of the first embodiment of a railway foundation settlement management early warning method, in which a space-time synchronization calibration model of a vehicle-mounted displacement sensor and a fiber sensing network is established, and real-time collected vehicle body longitudinal displacement data and corresponding fiber strain data of a mileage section are fused and processed to calculate the longitudinal change degree of a monitoring interval.

[0067] Figure 3 It is a flowchart of the second embodiment of a railway foundation settlement management early warning method, in which a space-time synchronization calibration model of a vehicle-mounted displacement sensor and a fiber sensing network is established, and real-time collected vehicle body longitudinal displacement data and corresponding fiber strain data of a mileage section are fused and processed to calculate the longitudinal change degree of a monitoring interval.

[0068] Figure 4 It is a flowchart of a railway foundation settlement management early warning method, in which a multi-level threshold triggering and recheck strategy is set.

[0069] Figure 5 It is a flowchart of a railway foundation settlement management early warning method, in which a multi-level threshold triggering and recheck strategy is set.

[0070] Figure 6 It is a main structure diagram of a railway foundation settlement management early warning system.

[0071] Figure 7 It is a structure block diagram of a space-time synchronization calibration model module in a railway foundation settlement management early warning system.

[0072] Figure 8 It is a structure block diagram of a calculation module in a railway foundation settlement management early warning system.

[0073] Figure 9 It is a structure block diagram of a management and control module in a railway foundation settlement management early warning system.

[0074] Figure 10 It is a structure block diagram of a generation and sending module in a railway foundation settlement management early warning system. Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0076] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0077] The present invention provides a railway foundation settlement management and early warning method and system, which solves the technical problems in the background art.

[0078] like Figure 1 The diagram shown is a main flowchart of a railway foundation settlement management and early warning method according to an embodiment of the present invention. The railway foundation settlement management and early warning method includes:

[0079] Step S100: Install an on-board displacement sensor at the target position of the vehicle to collect longitudinal displacement data of the vehicle body in real time during the vehicle's movement.

[0080] The target locations include the bogie axle box, the central section of the car body underframe, and the car body directly above the rail contact surface;

[0081] Step S200: Deploy a continuous distributed fiber optic sensor monitoring group and temperature sensor within the railway line to collect continuous strain distribution data and temperature data of the railway foundation in real time.

[0082] The continuous distributed optical fiber sensor monitoring group includes an optical fiber sensor network laid along the railway line.

[0083] Step S300: Establish a spatiotemporal synchronization calibration model for the vehicle displacement sensor and the fiber optic sensing network, fuse the real-time collected longitudinal displacement data of the vehicle body with the fiber optic strain data of the corresponding mileage segment, and calculate the degree of longitudinal change of the foundation within the monitoring interval.

[0084] Step S400: Set multi-level threshold triggering and re-inspection strategies;

[0085] Step S500: Integrate multi-source data, generate and send tiered intelligent early warning notifications based on multi-level threshold triggering and re-inspection strategies;

[0086] In application, the application constructs a "vehicle-line-foundation" three-in-one intelligent monitoring system. At the data acquisition level, high-precision vehicle-mounted displacement sensors are accurately installed on the bogie axle box, the central section of the vehicle body underframe and the vehicle body directly above the rail contact surface, covering a MEMS inertial measurement unit with a sampling frequency of ≥100 Hz and a Beidou differential positioning module with a centimeter-level accuracy, to capture vehicle body longitudinal displacement, pitch angle and acceleration data in real time at a high frequency interval of ≤50 ms, accurately reflecting the dynamic response of the vehicle during operation. Meanwhile, two single-mode optical fibers are laid parallel at a depth of 0.3-0.5 m below the roadbed surface on both sides of the railway line, forming a distributed optical fiber sensing network with a spacing of 50-100 m. With the help of BOTDA technology, continuous strain data acquisition with a spatial resolution of ≤1 m is realized, and the interference of temperature on optical fiber strain measurement is eliminated. At the data processing and analysis stage, a space-time synchronization calibration model is established based on the vehicle-mounted GPS timestamp and the line mileage marker. The Kalman filter algorithm is used to denoise the vehicle-mounted displacement data, and the dynamic time warping algorithm is used to align the phase of the vehicle-mounted longitudinal displacement curve and the optical fiber strain curve. The longitudinal variation degree of the foundation in the monitoring interval is calculated, the vehicle-ground data is deeply integrated, two levels of longitudinal variation degree thresholds and corresponding regional and full-line level recheck strategies are set, and the relative settlement is calculated based on the connected pipe liquid column balance principle through the optical fiber static level sensor group. A multi-source data fusion early warning model is constructed based on the BP neural network, and multiple parameters such as vehicle-mounted longitudinal displacement, optical fiber strain value and liquid level pressure change rate are coupled and analyzed. According to different safety thresholds of ballastless and ballasted tracks, three-level ladder warning signals of yellow, orange and red are generated, and the railway dispatching and operation and maintenance system is linked to realize intelligent management of the whole process from monitoring to disposal.

[0087] As shown in Figure 2 , as a preferred embodiment of the application, the space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network fuses and processes the real-time collected vehicle body longitudinal displacement data and the optical fiber strain data corresponding to the mileage section, and calculates the longitudinal variation degree of the foundation in the monitoring interval, which specifically includes:

[0088] Step S301: establishing a space-time mapping relationship based on the vehicle-mounted GPS timestamp of the running vehicle and the line mileage marker;

[0089] Step S302: denoising the vehicle-mounted displacement data by the Kalman filter algorithm and simultaneously temperature compensating the optical fiber strain data;

[0090] The calculation process of the compensated optical fiber strain value is as follows:

[0091] ;

[0092] In the formula, a raw optical fiber strain measurement value without temperature compensation, a fiber thermal expansion coefficient, a real-time temperature change amount;

[0093] In application, the embodiment establishes a space-time mapping relationship between the vehicle-mounted GPS time stamp and the line mileage mark. During the driving process, the vehicle-mounted GPS can accurately record the position of the vehicle at different times, and the line mileage mark clearly indicates the mileage information of each position on the railway line. By combining the two, the data collected on the vehicle can be accurately corresponded to the specific position and time point of the railway line. Then, the Kalman filtering algorithm is used to denoise the vehicle displacement data. Since the vehicle displacement sensor will be affected by factors such as vibration and electromagnetic interference during the driving process when collecting data, the data will have noise. The Kalman filtering algorithm can optimally estimate the data based on the historical information and current measurement value, effectively remove noise interference, and make the vehicle displacement data more accurate and reliable. At the same time, the fiber strain data is temperature compensated. The fiber strain data will be affected by changes in environmental temperature, and temperature changes will cause the fiber to expand and contract, thereby affecting the strain measurement results. After the space-time mapping, denoising and temperature compensation processing, the data provides data support for subsequent phase alignment of the vehicle-mounted longitudinal displacement curve and the fiber strain curve using the dynamic time warping algorithm, and calculation of the ground longitudinal variation degree in the monitoring interval.

[0094] As Figure 3 shown, as a preferred embodiment of the present application, the establishment of the space-time synchronous calibration model of the vehicle-mounted displacement sensor and the fiber sensing network fuses the real-time collected vehicle body longitudinal displacement data and the fiber strain data corresponding to the mileage section, and calculates the ground longitudinal variation degree in the monitoring interval, which specifically includes:

[0095] Step S311: based on the dynamic time warping algorithm, the vehicle-mounted longitudinal displacement curve is phase-aligned with the fiber strain curve corresponding to the mileage section;

[0096] Step S312: based on the difference algorithm, the ground longitudinal deformation gradient value in the monitoring interval is calculated .

[0097] The calculation process of the ground longitudinal deformation gradient value is as follows:

[0098] ;

[0099] In the formula, is the longitudinal deformation gradient of the first monitoring surface, is the longitudinal displacement amount of the vehicle-mounted sensor at the corresponding section, is the section spacing, Convert the vertical strain of the fiber sensor monitor group into displacement;

[0100] In the application of the embodiment, through the cooperative application of the dynamic time warping algorithm and the difference algorithm, deep fusion analysis of the vehicle-mounted dynamic response data and the ground static strain data is realized, and a high-precision longitudinal deformation gradient calculation model is constructed. In the time and space synchronization calibration link, due to the differences in the sampling frequency and data transmission delay of the vehicle-mounted displacement sensor and the distributed fiber sensing network, and the change of the train running speed will cause the phase deviation of the vehicle-mounted data time series and the line mileage mark, the dynamic time warping algorithm aligns the phase of the vehicle-mounted longitudinal displacement curve and the fiber strain curve of the corresponding mileage section by matching the dynamic bending characteristics of the time series. Specifically, the algorithm takes the minimization of the cumulative sum of the Euclidean distance between two curve points as the target, solves the time-space asynchronous problem caused by factors such as vehicle acceleration, deceleration and vibration by constructing a cost matrix and finding the optimal path, and ensures that the vehicle-mounted displacement data and the fiber strain data of the same monitoring section correspond in time and space dimensions. The ground longitudinal deformation gradient value , which can reflect the deformation gradient intensity in the monitoring section.

[0101] As shown in Figure 4 , as a preferred embodiment of the present application, the multi-level threshold triggering and review strategy specifically includes:

[0102] Step S401: preset two-level longitudinal change degree thresholds;

[0103] The primary change degree threshold is , and the emergency change degree threshold is ;

[0104] Step S402: when the ground longitudinal deformation gradient values of the continuous 3 sampling periods exceed the primary change degree threshold , trigger the regional review instruction;

[0105] Step S403: based on the regional review instruction, control the fiber static water level sensor group deployed in the monitoring section to measure the pressure change of each reference point in real time through the liquid level communication pipe;

[0106] Step S404: calculate the relative settlement amount of each reference point;

[0107] The calculation process of the relative settlement amount is as follows:

[0108] ;

[0109] In the formula, is the pressure change value, For the liquid density, Take 9.81 ;

[0110] Step S405: When the ground longitudinal deformation gradient value exceeds the emergency change degree threshold , a full-line level real-time scanning is triggered;

[0111] It should be understood that when the ground longitudinal deformation gradient value calculated by the system breaks through the primary change degree threshold for 3 consecutive sampling periods, a regional level re-inspection instruction is triggered. The instruction activates the optical fiber static level sensor group deployed in the 50-200 meter monitoring section, and the sensors are distributed in the bridge-subgrade transition section, soft soil foundation and other key areas at an interval of 5-10 meters, and the pressure change data of each reference point is collected in real time through the liquid level communication pipe. Based on the principle of liquid column balance of the communication pipe, the relative settlement of each reference point is calculated, and if the ground longitudinal deformation gradient value breaks through the emergency change degree threshold , a full-line level real-time scanning program is immediately started. The program will mobilize the distributed optical fiber sensing network along the railway, the vehicle-mounted monitoring equipment and the fixed monitoring points along the line to conduct synchronous detection of the whole line. The distributed optical fiber sensing network continuously monitors the ground strain with a spatial resolution of 1m, the vehicle-mounted sensor synchronously collects dynamic displacement data during driving, and the fixed monitoring points along the line real-time return environmental temperature and humidity, underground water level and other parameters.

[0112] As shown in Figure 5 , as a preferred embodiment of the present application, the fusion of multi-source data, based on multi-level threshold triggering and re-inspection strategy, generates and sends a ladder intelligent early warning notice, which specifically includes:

[0113] Step S501: build a fusion early warning model containing a back propagation neural network;

[0114] The input parameters of the fusion early warning model include: vehicle-mounted longitudinal displacement, optical fiber strain value, liquid level pressure change rate, environmental temperature and humidity, and train running speed;

[0115] Step S502: when the relative settlement of any reference point exceeds the target threshold, a ladder classification early warning signal is generated;

[0116] And set the ballast section threshold , the ballast section threshold ;

[0117] In the ballast section, the ladder classification early warning signal specifically includes:

[0118] Yellow warning: when , generate and send a manual inspection instruction;

[0119] Orange warning: when , generate and send speed reduction instructions;

[0120] Orange warning: when , generate and send emergency stop instructions;

[0121] In the application of the embodiment, a back propagation neural network is used to build a fusion early warning model. The input parameters of the model are multiple, including the vehicle longitudinal displacement, which can reflect the dynamic changes of the vehicle body in the longitudinal direction during train running; the fiber strain value, which can capture the deformation information of the foundation; the hydraulic pressure change rate, which is helpful to monitor the fluctuation of the internal pressure of the foundation; the environmental temperature and humidity, because environmental factors can have a potential impact on the stability of the foundation; the train running speed, the force borne by the railway foundation is different at different speeds; in the settlement monitoring and threshold setting, the relative settlement of each reference point is accurately calculated through specific technical means. And according to the different track structure characteristics, the differentiated settlement thresholds are set, the threshold of ballastless section , the threshold of ballasted section . The tolerance difference of different track types to settlement is fully considered; when the yellow warning state is triggered, the artificial inspection instruction is generated and sent. This is in the case of preliminary abnormality of the settlement amount, but it is not serious. Through artificial close-range inspection, the actual condition of the foundation is further determined; when the orange warning state is triggered, the speed reduction instruction is generated and sent. It means that the settlement has reached a certain degree, which may affect the safe and smooth running of the train. By reducing the train speed, the force on the foundation is reduced, and time is also gained for further disposal; when the red warning state is triggered, the emergency stop instruction is generated and sent immediately. It indicates that the foundation settlement has reached a state of serious threat to train safety. Emergency stop can maximize the prevention of accidents and ensure the safety of personnel life and equipment property.

[0122] As shown in Figure 6 , as another preferred embodiment of the present application, on the other hand, a railway foundation settlement management early warning system, the system comprises:

[0123] The first real-time acquisition module 100 is used to install a vehicle-mounted displacement sensor at a targeted position of a running vehicle, and to acquire vehicle body longitudinal displacement data in real time during vehicle running;

[0124] The targeted position includes the axle box of the bogie, the central section of the vehicle body frame, and the vehicle body directly above the rail contact surface;

[0125] The second real-time acquisition module 200 is used to deploy a continuous distributed optical fiber sensing monitor group and a temperature sensor inside the railway line, and to acquire continuous strain distribution data and temperature data of the railway foundation in real time;

[0126] The continuous distributed optical fiber sensing monitor group comprises an optical fiber sensing network laid along a railway line;

[0127] The space-time synchronization calibration model module 300 is configured to establish a space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network.

[0128] The computing module 400 is configured to fuse the real-time collected vehicle body longitudinal displacement data and the optical fiber strain data corresponding to the mileage section, and calculate the ground longitudinal variation degree in the monitoring section.

[0129] The control module 500 is configured to set a multi-level threshold triggering and review strategy.

[0130] The generating and sending module 600 is configured to fuse multi-source data, generate and send a ladder intelligent early warning notice based on the multi-level threshold triggering and review strategy.

[0131] In application, the vehicle-mounted displacement sensor is installed at the targeted position of the running vehicle, the first real-time collection module 100 collects the vehicle body longitudinal displacement data in the running process of the vehicle in real time, the continuous distributed optical fiber sensing monitor group and the temperature sensor are deployed inside the railway line, the second real-time collection module 200 collects the continuous strain distribution data and the temperature data of the railway ground in real time, the space-time synchronization calibration model module 300 establishes the space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network, the computing module 400 fuses the real-time collected vehicle body longitudinal displacement data and the optical fiber strain data corresponding to the mileage section, and calculates the ground longitudinal variation degree in the monitoring section, the control module 500 sets the multi-level threshold triggering and review strategy, the generating and sending module 600 generates and sends the ladder intelligent early warning notice based on the multi-level threshold triggering and review strategy.

[0132] As shown in Figure 7 , as another preferred embodiment of the present application, the space-time synchronization calibration model module 300 specifically comprises:

[0133] The mapping unit 301 is configured to establish a space-time mapping relationship based on the vehicle-mounted GPS time stamp of the running vehicle and the line mileage mark.

[0134] The synchronization compensation unit 302 is configured to perform denoising processing on the vehicle-mounted displacement data through a Kalman filtering algorithm, and synchronously perform temperature compensation on the optical fiber strain data.

[0135] In application, the mapping unit 301 establishes a space-time mapping relationship based on the vehicle-mounted GPS time stamp of the running vehicle and the line mileage mark, and the synchronization compensation unit 302 performs denoising processing on the vehicle-mounted displacement data through a Kalman filtering algorithm, and synchronously performs temperature compensation on the optical fiber strain data.

[0136] As shown in Figure 8As another preferred embodiment of the present application, the computing module 400 specifically comprises:

[0137] The phase alignment unit 401 is configured to perform phase alignment between the vehicle-mounted longitudinal displacement curve and the fiber strain curve of the corresponding mileage segment based on the dynamic time warping algorithm.

[0138] The first computing unit 402 is configured to calculate the ground longitudinal deformation gradient value in the monitoring interval based on the difference algorithm. .

[0139] In the application of the present embodiment, the phase alignment unit 401 performs phase alignment between the vehicle-mounted longitudinal displacement curve and the fiber strain curve of the corresponding mileage segment based on the dynamic time warping algorithm, and the first computing unit 402 calculates the ground longitudinal deformation gradient value in the monitoring interval based on the difference algorithm. .

[0140] As shown in Figure 9 , as another preferred embodiment of the present application, the management and control module 500 specifically comprises:

[0141] The two-level longitudinal change degree threshold unit 501 is configured to preset two-level longitudinal change degree thresholds.

[0142] The re-inspection unit 502 is configured to trigger a regional-level re-inspection instruction when the ground longitudinal deformation gradient value in the continuous three sampling periods exceeds the primary change degree threshold .

[0143] The real-time measurement unit 503 is configured to control the fiber static level sensor group deployed in the monitoring section to measure the pressure change of each reference point in real time through the liquid level communication pipe based on the regional-level re-inspection instruction.

[0144] The second computing unit 504 is configured to calculate the relative settlement amount of each reference point .

[0145] The full-line-level real-time scanning unit 505 is configured to trigger full-line-level real-time scanning when the ground longitudinal deformation gradient value exceeds the emergency change degree threshold .

[0146] In the application of the present embodiment, the two-level longitudinal change degree threshold unit 501 presets two-level longitudinal change degree thresholds, and when the ground longitudinal deformation gradient value in the continuous three sampling periods exceeds the primary change degree threshold When the relative settlement amount of any reference point exceeds the target threshold value, the generating unit 602 generates a step-by-step early warning signal. When the local longitudinal deformation gradient value Exceeds the emergency change degree threshold value The full-line level real-time scanning unit 505 triggers full-line level real-time scanning.

[0147] As shown in Figure 10 As another preferred embodiment of the present application, the generating and sending module 600 specifically includes:

[0148] The fusion early warning model unit 601 is configured to construct a fusion early warning model containing a back propagation neural network.

[0149] The fusion early warning model input parameters include: vehicle-mounted longitudinal displacement, fiber strain value, liquid level pressure change rate, environmental temperature and humidity, and train running speed.

[0150] The generating unit 602 is configured to generate a step-by-step early warning signal when the relative settlement amount of any reference point exceeds the target threshold value.

[0151] In the application of the present embodiment, the fusion early warning model unit 601 constructs a fusion early warning model containing a back propagation neural network, and the generating unit 602 generates a step-by-step early warning signal when the relative settlement amount of any reference point exceeds the target threshold value.

[0152] ​​The railway foundation settlement management early warning method and system provided in the above embodiments of the present application construct a "vehicle-line-foundation" three-in-one intelligent monitoring system. At the data acquisition level, high-precision vehicle-mounted displacement sensors are accurately installed on the bogie axle box, the central section of the vehicle body underframe, and the vehicle body directly above the rail contact surface of the running vehicle, covering a MEMS inertial measurement unit with a sampling frequency of 100 Hz or more and a Beidou differential positioning module with a centimeter-level accuracy, to capture vehicle body longitudinal displacement, pitch angle, and acceleration data at a high frequency interval of 50 ms or less in real time, accurately reflecting the dynamic response of the vehicle during operation. Meanwhile, two single-mode optical fibers are laid parallel at a depth of 0.3-0.5 m below the roadbed surface on both sides of the railway line, forming a distributed optical fiber sensing network with a spacing of 50-100 m. With the help of BOTDA technology, continuous strain data acquisition with a spatial resolution of 1 m or less is achieved, and the interference of temperature on optical fiber strain measurement is eliminated. At the data processing and analysis stage, a space-time synchronization calibration model is established based on the vehicle-mounted GPS timestamp and the line mileage marker, and Kalman filtering algorithm is used to denoise the vehicle-mounted displacement data. The dynamic time warping algorithm is used to align the phase of the vehicle-mounted longitudinal displacement curve and the optical fiber strain curve, calculate the longitudinal variation degree of the foundation in the monitoring interval, realize the deep fusion of vehicle-ground data, set two levels of longitudinal variation degree threshold and corresponding regional and full-line level recheck strategies, calculate the relative settlement based on the connected pipe liquid column balance principle through the optical fiber static level sensor group; a multi-source data fusion early warning model is constructed based on the BP neural network, and multiple parameters such as vehicle-mounted longitudinal displacement, optical fiber strain, and liquid level pressure change rate are coupled and analyzed. According to the different safety thresholds of ballastless and ballasted tracks, yellow, orange, and red three-level step-by-step early warning signals are generated, and the railway dispatching and operation and maintenance system is linked to realize intelligent management of the whole process from monitoring to disposal. The method and system construct a "vehicle-line-foundation" three-in-one intelligent monitoring system, significantly improving the spatial and temporal resolution and accuracy of railway foundation settlement monitoring. The multi-level threshold and recheck mechanism reduces the false alarm and missed alarm rate, the step-by-step early warning realizes risk grading control, and effectively guarantees the safety of train operation. The intelligent early warning system links the railway dispatching and operation and maintenance, and improves the emergency response efficiency.

[0153] In order to enable the above-mentioned method and system to run smoothly, the system can include more or fewer components than described above, or combine certain components, or different components, such as input and output devices, network access devices, buses, processors, and memories.

[0154] The processor can be a central processing unit, and can also be other general purpose processors, digital signal processors, application specific integrated circuits, programmable logic devices, discrete gates or transistor logic components, discrete hardware components, or the like. The general purpose processor can be a microprocessor or can also be any conventional processor. The processor is a control center of the system, and connects various parts by using various interfaces and lines.

[0155] Any combination of the technical features in the above-described embodiments can be made, and in order to make the description simple, all possible combinations of the technical features in the above-described embodiments are not described, however, as long as the combination of the technical features does not exist, it should be considered that it is within the scope of the description.

[0156] The above-described embodiments only express several implementation manners of the present application, the description is relatively specific and detailed, but it should not be understood as a limitation on the patent scope of the present application. It should be pointed out that, for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

[0157] The above-described only the preferred embodiments of the present application, and does not limit the present application, any modification, equivalent replacement and improvement within the spirit and principle of the present application, etc., should be included in the protection scope of the present application.

Claims

1. A railway subgrade settlement management early warning method, characterized in that, The method comprises: installing a vehicle-mounted displacement sensor at a target position of the running vehicle to collect vehicle body longitudinal displacement data in real time during vehicle running, the target position including a bogie axle box, a central section of a vehicle body underframe, and a vehicle body directly above a rail contact surface; deploying a continuous distributed optical fiber sensing monitor group and a temperature sensor inside a railway line to collect continuous strain distribution data and temperature data of a railway foundation in real time, the continuous distributed optical fiber sensing monitor group including an optical fiber sensing network laid along the railway line; establishing a space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network, fusing the collected vehicle body longitudinal displacement data with optical fiber strain data of the corresponding mileage section, and calculating the longitudinal variation degree of the foundation in the monitoring section; setting a multi-level threshold triggering and rechecking strategy; fusing multi-source data, generating and sending a ladder intelligent early warning notice based on the multi-level threshold triggering and rechecking strategy; The method further comprises: aligning the vehicle-mounted longitudinal displacement curve with the optical fiber strain curve of the corresponding mileage section based on a dynamic time warping algorithm; Based on the difference algorithm, the gradient value of the longitudinal deformation of the ground in the monitoring interval is calculated ; the ground longitudinal deformation gradient value The calculation process is as follows: ； wherein, is the first is the longitudinal deformation gradient of the monitoring surface, is the longitudinal displacement of the on-board sensor at the corresponding section, is the section spacing, is the vertical strain conversion displacement calculated by the fiber-optic sensor monitor group; The multi-level threshold triggering and rechecking strategy comprises: The preset two-stage longitudinal change degree threshold values are a primary change degree threshold value , and an emergency change degree threshold value . When the ground longitudinal deformation gradient value of 3 consecutive sampling periods exceeds the primary change degree threshold , the regional level review instruction is triggered; controlling a group of optical fiber hydrostatic level sensors deployed in the monitoring section based on a regional rechecking instruction to measure the pressure change of each reference point in real time through a liquid level communication pipe; calculating the relative settlement of each reference point ; The relative sedimentation amount The calculation process is as follows: ； wherein is the pressure change value, is the communicating liquid density, Take 9.81 ; local longitudinal deformation gradient value exceeds an emergency change degree threshold triggering a full-line level real-time scan.

2. The railway subgrade settlement management pre-warning method according to claim 1, characterized in that, The method further comprises: establishing a space-time mapping relationship based on a vehicle-mounted GPS timestamp of the running vehicle and a line mileage marker; performing denoising processing on the vehicle-mounted displacement data and temperature compensation on the optical fiber strain data simultaneously through a Kalman filtering algorithm; Compensated fiber strain value The calculation process is as follows: ； In the formula, is the original optical fiber strain measurement value without temperature compensation, is the optical fiber thermal expansion coefficient, is the real-time temperature change amount.

3. The railway subgrade settlement management and early warning method according to claim 1, characterized in that, The method further comprises: constructing a fusion early warning model containing a back propagation neural network, and the input parameters of the fusion early warning model include vehicle-mounted longitudinal displacement, optical fiber strain value, liquid level pressure change rate, environmental temperature and humidity, and train running speed; When the relative settlement amount of any reference point A stepwise escalation warning signal is generated when the target threshold is exceeded, and a ballastless section threshold is set , ballast section threshold ; In a ballastless section, the ladder classification early warning signal specifically includes: Yellow alert: when the artificial patrol instruction is generated and sent; Orange alert: when a speed reduction instruction is generated and sent; Red alert: when an emergency stop command is generated and sent.

4. A railway subgrade settlement management early warning system, characterized by, The system comprises: a first real-time collection module for installing a vehicle-mounted displacement sensor at a target position of the running vehicle to collect vehicle body longitudinal displacement data in real time during vehicle running; the target position including a bogie axle box, a central section of a vehicle body underframe, and a vehicle body directly above a rail contact surface; a second real-time collection module for deploying a continuous distributed optical fiber sensing monitor group and a temperature sensor inside a railway line to collect continuous strain distribution data and temperature data of a railway foundation in real time; the continuous distributed optical fiber sensing monitor group including an optical fiber sensing network laid along the railway line; The space-time synchronization calibration model module is configured to establish a space-time synchronization calibration model of the vehicle-mounted displacement sensor and the optical fiber sensing network. The calculation module is configured to fuse the real-time collected vehicle body longitudinal displacement data and the optical fiber strain data corresponding to the mileage section, and calculate the ground longitudinal variation degree in the monitoring section. The control module is configured to set a multi-level threshold trigger and a recheck strategy. The generation and sending module is configured to fuse multi-source data, generate and send a ladder intelligent early warning notice based on the multi-level threshold trigger and the recheck strategy.

5. The railway subgrade settlement management pre-warning system according to claim 4, characterized in that, The space-time synchronization calibration model module specifically includes: A mapping unit is configured to establish a space-time mapping relationship based on the vehicle-mounted GPS time stamp of the traveling vehicle and the route mileage mark. A synchronization compensation unit is configured to perform denoising processing on the vehicle-mounted displacement data and temperature compensation on the optical fiber strain data by using a Kalman filtering algorithm.

6. The railway subgrade settlement management pre-warning system according to claim 4, characterized in that, The calculation module includes: A phase alignment unit is configured to perform phase alignment on the vehicle-mounted longitudinal displacement curve and the optical fiber strain curve corresponding to the mileage section based on a dynamic time warping algorithm. The first computing unit is configured to calculate the gradient value of the ground longitudinal deformation in the monitoring interval based on a difference algorithm. .

7. The railway subgrade settlement management pre-warning system according to claim 4, characterized in that, The control module specifically includes: A two-level longitudinal variation degree threshold unit is configured to preset two-level longitudinal variation degree thresholds. The retest unit is used to measure the longitudinal deformation gradient value of the foundation over three consecutive sampling periods. Exceeding the primary change threshold At that time, a regional-level re-inspection command is triggered; A real-time measurement unit is configured to control the optical fiber static leveling sensor group deployed in the monitoring section to perform real-time measurement on the pressure variation of each reference point through the liquid level communication pipe based on the regional recheck instruction. a second computing unit for computing the relative settlement of each reference point ; Line-wide real-time scanning unit for local longitudinal deformation gradient values Exceeding an emergency change degree threshold Triggering line-wide real-time scanning.

8. The railway subgrade settlement management pre-warning system according to claim 4, characterized in that, The generation and sending module specifically includes: A fusion early warning model unit is configured to construct a fusion early warning model including a back propagation neural network. The fusion early warning model input parameters include: the vehicle-mounted longitudinal displacement amount, the optical fiber strain value, the liquid level pressure variation rate, the environmental temperature and humidity, and the train running speed. The generating unit is configured to generate a relative settlement amount of any reference point When the target threshold is exceeded, a stepwise hierarchical early warning signal is generated.

Images