Intelligent monitoring system for high-speed rail roadbed
By combining millimeter-wave radar, Beidou RTK measuring instrument and ultrasonic meteorological station, the intelligent monitoring system of high-speed railway subgrade settlement monitoring is solved all-weather, high-precision and data utilization problems of high-speed railway subgrade settlement monitoring, real-time monitoring and timing prediction are achieved, and high-speed railways are improved Operational security and data utilization.
Patent Information
- Application Number
- CN202420824077.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-19
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2034-04-19
AI Technical Summary
The existing high-speed railway subgrade settlement monitoring technology cannot achieve all-weather, high-precision, and intelligent monitoring, and is limited in application in harsh environments, and the data processing and storage security is insufficient, so it is impossible to make full use of monitoring data for timing prediction to improve operational security.
Using millimeter wave radar combined with Beidou high-precision RTK measuring instrument, ultrasonic integrated meteorological station and cloud data platform, a high-speed railway subgrade intelligent monitoring system is built to achieve high-precision, all-weather and real-time monitoring, and establish a large model for timing prediction through deep learning algorithms. .
It realizes high-precision, all-weather and real-time monitoring of subsidence of high-speed railways, reduces the economic cost of the monitoring system, is suitable for complex terrain and harsh environments, improves the storage security and utilization rate of data, and improves the safety of high-speed railway operations and reliability.
Smart Images

Figure CN222049014U_ABST
Abstract
Description
Technical Field
[0001] The utility model relates to the technical field of high-speed railways, and in particular to an intelligent monitoring system for high-speed railway subgrades. Background Art
[0002] As an important part of the national economy, high-speed railways are of great significance for fast, high-precision and real-time monitoring of roadbed settlement and deformation. Although traditional detection methods such as total stations, laser rangefinders and liquid settlement systems can be used for observation during the construction phase, they cannot achieve all-weather, fully automatic real-time monitoring due to the need for manual operation and low efficiency. The safe and efficient operation of high-speed railways plays an important role in various fields such as my country's economy, politics, and culture. In order to ensure the safety of high-speed railway operation, scientific researchers have conducted a lot of research on all-weather, high-precision and intelligent monitoring of high-speed railway settlement:
[0003] The patent "An Adaptive Scanning Roadbed Settlement Remote Monitoring System and Method", publication number CN101709968A, uses an optical system to achieve settlement monitoring. Its main disadvantages are: the laser is greatly affected by rain, fog, dust and other environmental factors, and the device contains mechanical moving parts, so the reliability is not high, and it is difficult to achieve long-term, stable, and all-weather operation.
[0004] Patent "A multi-point laser displacement monitoring device, system and method", publication number CN115289981A. The utility model uses a laser rangefinder, a galvanometer motor and a servo control module to monitor multiple objects at the same time. The main disadvantage is that since laser ranging is greatly affected by environmental factors, especially atmospheric humidity, the device is limited in application scenarios and is more suitable for indoor model tests and other environmental conditions. It is limited in application in areas with large changes in environmental conditions.
[0005] Patent "A real-time monitoring system for high-speed railway subgrade settlement", publication number CN103884317A. Microwave radar is used in this patent to monitor settlement. Compared with optical system monitoring and liquid method monitoring, its main advantages are that microwave radar can realize high-precision continuous monitoring of multiple measuring points under unmanned conditions, and has good application potential. The disadvantage is that there is no subsequent processing of the monitoring data, and the storage security and full utilization of the data cannot be guaranteed. In addition, for the safe operation of high-speed rail, monitoring is only passive, and full use should be made of data and artificial intelligence to build large models to achieve time series prediction, thereby improving the safety of high-speed rail operations. Utility Model Content
[0006] The purpose of the utility model is to provide a high-speed railway subgrade intelligent monitoring system to solve the problems existing in the above-mentioned prior art.
[0007] In order to achieve the above object, the utility model provides a high-speed railway subgrade intelligent monitoring system, which adopts the following technical solutions:
[0008] A high-speed railway subgrade intelligent monitoring system, comprising:
[0009] Millimeter wave radar transceiver;
[0010] A corner reflector, wherein the corner reflector is connected to the millimeter wave radar transceiver signal;
[0011] Settling plate;
[0012] A metal rod, wherein the metal rod is arranged on the settling plate, and the corner reflector is arranged on the metal rod;
[0013] An integrated weather station, wherein a signal output terminal of the integrated weather station is connected to a signal input terminal of the millimeter wave radar transceiver;
[0014] A signal base station, wherein the signal output terminal of the millimeter wave radar transceiver is connected to the signal input terminal of the signal base station;
[0015] Cloud data platform, the signal output end of the signal base station is connected to the signal input end of the cloud data platform.
[0016] Furthermore, the integrated meteorological station includes a supporting body, which is fixedly supported by a supporting connector, and is provided with a temperature sensor module, a sound pressure sensor module, an atmospheric humidity measurement module, a full-angle wind direction sensor module, an ultrasonic wind speed sensor module and a communication module. The signal output ends of the temperature sensor module, the sound pressure sensor module, the atmospheric humidity measurement module, the full-angle wind direction sensor module and the ultrasonic wind speed sensor module are all connected to the signal input end of the communication module, and the signal output end of the communication module is connected to the signal input end of the millimeter wave radar transceiver.
[0017] Furthermore, the millimeter-wave radar transceiver includes a chip, a modulator, a power amplifier, a transmitting module, a receiving module and a communication interface. The signal output end of the chip is connected to the modulator, the power amplifier and the transmitting module in sequence, the signal output end of the receiving module is connected to the signal input end of the chip, and the signal output end of the chip is connected to the signal input end of the communication interface.
[0018] Furthermore, the chip is a MMIC chip.
[0019] Furthermore, the settling plate is connected to three of the metal rods, and each of the metal rods is provided with one of the corner reflectors.
[0020] Furthermore, the settlement plates are provided in plurality, which are respectively arranged at the top of the foundation, the middle of the roadbed, the side of the track, and the midpoint of the section.
[0021] Furthermore, the millimeter wave radar is arranged at the upper end of the equal-diameter annular concrete pole.
[0022] Furthermore, the equal-diameter annular concrete rods are arranged beside the railway section embankment whose settlement is to be monitored.
[0023] Furthermore, the height of the equal-diameter annular concrete pole is 6 to 8 m.
[0024] Furthermore, it also includes a concrete base, and a permanent control point is set at the upper end of the concrete base. The setting of the permanent control point shall refer to the relevant provisions of the "General Specification for Engineering Surveying" GB 55018-2021 for setting.
[0025] The beneficial effects of the utility model are as follows: the utility model combines millimeter wave radar, settlement plate, corner reflector, Beidou high-precision RTK measuring instrument, ultrasonic integrated all-in-one weather station and network signal base station to achieve high-precision, all-weather, real-time monitoring of high-speed railway subgrade settlement. Beidou RTK high-precision measuring instrument is used to encrypt control points and build permanent control points with millimeter-level accuracy. In the absence of other data, the monitoring data can be corrected through these encrypted control points to reduce the economic cost of the monitoring system. The key high-speed railway line monitoring section uses an ultrasonic integrated all-in-one weather station to measure multiple data, such as rainfall, temperature, wind speed, humidity, wind direction, noise and atmospheric pressure, for data correction and analysis. The utility model is a non-contact subgrade settlement monitoring system with a comprehensive environmental compensation algorithm that can be read remotely and automatically. It is particularly suitable for subgrade settlement monitoring in complex terrain and harsh environmental conditions. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] In order to more clearly illustrate the specific implementation of the utility model or the technical solution in the prior art, the following is a brief introduction to the drawings required for the specific implementation or the prior art description. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, each element or part is not necessarily drawn according to the actual scale.
[0027] Figure 1 A structural schematic diagram of a high-speed railway subgrade intelligent monitoring system according to an embodiment of the utility model is shown.
[0028] Figure 2 A structural diagram of an integrated meteorological station of a high-speed railway subgrade intelligent monitoring system according to an embodiment of the utility model is shown.
[0029] Figure 3A structural diagram of a millimeter-wave radar transceiver of a high-speed railway subgrade intelligent monitoring system according to an embodiment of the utility model is shown.
[0030] Figure 4a A recommended measurement point arrangement diagram of a high-speed railway subgrade intelligent monitoring system according to an embodiment of the utility model is shown, taking a standard subgrade cross section of a single-track high-speed railway as an example.
[0031] Figure 4b The following diagram shows the recommended measurement point arrangement of a high-speed railway subgrade intelligent monitoring system according to an embodiment of the utility model, taking the standard subgrade cross section of a double-track high-speed railway as an example.
[0032] Figure 5a An overall schematic diagram of a vertical monitoring arrangement of a high-speed railway subgrade intelligent monitoring system according to an embodiment of the utility model is shown.
[0033] Figure 5b A schematic diagram of equal-diameter annular concrete poles in a vertical monitoring arrangement of a high-speed railway subgrade intelligent monitoring system according to an embodiment of the utility model is shown.
[0034] Figure 5c A schematic diagram of a settlement plate in a vertical monitoring arrangement of a high-speed railway subgrade intelligent monitoring system according to an embodiment of the utility model is shown.
[0035] Figure 6 A cloud data platform workflow diagram of a high-speed railway subgrade intelligent monitoring system according to an embodiment of the utility model is shown.
[0036] Reference numerals:
[0037] 101. Millimeter wave radar transceiver; 102. Corner reflector; 103. Settlement plate; 104. Metal rod; 105. High-precision permanent control point; 106. Integrated meteorological station; 1061. Support body; 1062. Support connector; 1063. Temperature sensor module; 1064. Sound pressure sensor module; 1065. Atmospheric humidity measurement module; 1066. Full-angle wind direction sensor module; 1067. Ultrasonic wind speed sensor module; 1068. Communication module; 107. Signal base station; 108. Cloud data platform; 201. Equal-diameter annular concrete pole; 202. Roadbed; 301. MMIC chip; 302. Modulator; 303. Power amplifier; 304. Transmitter module; 305. Receiving module; 306. Communication interface; 307. Host computer. DETAILED DESCRIPTION
[0038] The following describes the implementation of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific implementations, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.
[0039] In the description of the present invention, unless otherwise specified, "multiple" means two or more; the terms "upper", "lower", "left", "right", "inner", "outer", "front end", "rear end", "head", "tail" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", "third" and the like are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.
[0040] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0041] The specific implementation of the utility model is further described in detail below in conjunction with the drawings and embodiments.
[0042] The embodiment of the utility model provides a high-speed railway subgrade intelligent monitoring system, which is composed of a benchmark encrypted control point system built by the Beidou Huace RTK measuring instrument, a millimeter-wave radar real-time monitoring system, and a cloud data analysis platform. It strives to create a new generation of all-weather, high-precision, high-anti-interference, and high-performance smart cloud monitoring platform. Through the deep learning algorithm of artificial intelligence, a large model of high-speed railway settlement is established to achieve time series prediction, thereby providing more advanced, scientific, and reliable data services for the safe operation of high-speed railways.
[0043] The wavelength of millimeter waves is between centimeter waves and light waves, which makes millimeter waves have the advantages of both microwave guidance and photoelectric guidance. At the same time, it has strong resistance to environmental interference and can monitor settlement in all weather conditions. Therefore, the utility model uses a millimeter-wave radar system to measure tiny displacements, and can obtain high-precision displacement measurement results in combination with a microwave corner reflector fixed on the settlement plate. At the same time, RTK measuring instruments are used along the high-speed railway to encrypt control points, build permanent control points, and then obtain standard data through millimeter-wave radars and reflectors, so that other data do not need to be corrected by a series of sensor devices such as temperature sensors, humidity sensors, and wind speed sensors, thereby automatically monitoring the settlement of the high-speed railway subgrade. In addition, cloud-based data calculation can provide high-quality training data for subsequent data large model training, thereby improving model training efficiency.
[0044] The subsidence monitoring and early warning system consists of a subsidence monitoring system and a cloud data analysis center. The main function of the subsidence monitoring system is to provide high-precision monitoring and measurement data, while the cloud analysis platform performs data cleaning, analysis, and deep learning LSTM large model training, thereby realizing the construction of a large model based on monitoring data, providing more scientific and reliable technical support for the subsequent intelligent analysis of high-speed rail operations.
[0045] The roadbed settlement monitoring system is mainly composed of millimeter wave radar transceiver, microwave corner reflector, settlement plate with metal rod (detailed structural parts see the subsequent construction design drawing), Beidou RTK measuring instrument, temperature sensor, humidity sensor, wind force and direction sensor, etc. First, during the construction of high-speed railway, settlement plates are pre-buried at the control position of the roadbed and roadbed. Three metal rods are welded on each settlement plate. The upper end of the metal rod is exposed to the structural surface and connected to the microwave corner reflector. The metal rod is covered with a protective layer to prevent deformation caused by direct contact with the roadbed structure and cause monitoring failure; the number of pre-buried settlement plates can be flexibly adjusted according to the specific conditions of the route section to be detected. It is recommended to pre-buried 5 to 10 settlement plates for double-line and 2 to 5 settlement plates for single-line high-speed railway. The recommended measurement point layout diagram is given in the standard roadbed cross-section diagram in the attached figure of the specification for reference for other monitoring projects. Afterwards, a 6-8 meter high pole is built along the high-speed rail monitoring section to carry millimeter wave radar and ultrasonic integrated weather station. The utility model adopts Beidou RTK high-precision measuring instrument to encrypt high-precision control points, and the encrypted points can be built into permanent control points with millimeter-level accuracy. It can be used to correct monitoring data when there is a lack of temperature, humidity, wind speed and other measurement data, thereby reducing the economic cost of the monitoring system. In the key control section of the high-speed rail line monitoring, ultrasonic integrated weather station can be used to measure rainfall, temperature, wind speed, humidity, wind direction, noise, atmospheric pressure and other data for high-precision data correction and analysis. The microwave corner reflector connected to the upper end of the metal pole should be adjusted to point to the microwave radar receiving antenna in the direction of its strongest reflection, thereby ensuring the original accuracy of the monitoring data.
[0046] The millimeter wave radar transceiver mainly consists of DDS (Direct Digital Synthesizer), modulator, power amplifier, mixer, low-frequency amplifier, signal processor and communication interface circuit. The function of DDS is to generate a low-frequency continuous wave (CW) or frequency modulated continuous wave (FMCW) circuit. It generates a high-stability and high-precision FMCW frequency signal through digital control. Then the modulator modulates the baseband signal generated by DDS with the RF carrier by linear frequency modulation (LFM) and continuous wave linear modulation (CWL) to generate the RF signal required by the millimeter wave radar. Among them, the power amplifier is used to enhance the amplitude of the RF signal to ensure that the signal has a sufficient power level for transmission. In millimeter wave radar, due to the high operating frequency, a power amplifier array is used to provide sufficient output power. The RF signal generated by the modulator is mixed with the received echo signal in the mixer to obtain a lower frequency signal for subsequent signal processing. The low-frequency amplifier is used to amplify the low-frequency signal output by the mixer, which contains information such as the distance, speed and angle of the target object, and needs to be amplified so that the signal processor can process the information. The signal processor is responsible for processing and analyzing the received echo signal. It extracts the characteristic information of the target object, such as distance, speed, spatial angle and other series of information through time domain and frequency domain analysis algorithms. The communication interface circuit of the millimeter wave radar is used to transmit the target object information processed by the signal processor to other devices or systems. Serial communication interfaces such as UART, SPI, I2C are often used to transmit data to the host computer or other control system for further processing and display. In the utility model, the communication interface will work together with the network signal base station to upload the preliminary analyzed characteristic information of the target object to the cloud data center for subsequent analysis, monitoring and training.
[0047] The cloud data analysis center is of great significance in the monitoring system. It has the important functions of secondary analysis of monitoring data, monitoring and early warning, and emergency coordination. At the same time, it uses monitoring data such as temperature, humidity, wind force, wind speed, atmospheric pressure, rainfall and sedimentation data to establish a real-time database and a deep learning model dedicated to high-speed rail operation monitoring. With the development of the timeline, the increase in the amount of training data will completely speed up the establishment of the data model. Ultimately, the high-speed rail real-time monitoring model will guide subsequent high-speed rail monitoring, maintenance and other aspects, thereby realizing the strategic concept of intelligent high-speed rail operation and maintenance.
[0048] It should be noted that the big data analysis methods applied by the cloud data analysis center mentioned above all use existing analysis methods. The purpose of the present invention is to obtain sufficient and accurate data through structural design.
[0049] like Figure 1As shown, the monitoring system includes a millimeter wave radar transceiver 101, a corner reflector 102, a settlement plate 103, a metal rod 104, a high-precision permanent control point 105, an integrated meteorological station 106, a signal base station 107, and a cloud data platform 108.
[0050] The above-mentioned millimeter wave radar transceiver 101 uses a frequency modulated continuous wave FMCW transceiver, whose operating frequency range is 1GHz-100GHz and wavelength is 1-10mm. The corner reflector 102 can be selected in three specifications: 55mm, 77.6mm, and 98.5mm. The specific selection needs to be flexibly selected according to the actual measurement point layout and the monitoring range of the millimeter wave radar.
[0051] The high-precision permanent control point 105 should be obtained by encrypting the control point using a high-precision RTK surveying instrument, and finally needs to be built into a permanent control point. The utility model uses the Beidou Huace RTK surveying instrument with millimeter-level accuracy. Please refer to the relevant provisions of the General Specification for Engineering Surveying GB 55018-2021 for the setting of the high-precision permanent control point 105.
[0052] The main parameters of the integrated weather station 106 are as follows: DC power supply 10-30V; power consumption 0.12W; wind speed range: 0~40m / s; wind direction: 0~360° accuracy; working temperature -40~80°℃; working humidity 0~95%RH; wind resistance 75m / s.
[0053] like Figure 2 As shown, the integrated meteorological station 106 includes a supporting body 1061, which is fixedly supported by a supporting connector 1062. A temperature sensor module 1063, a sound pressure sensor module 1064, an atmospheric humidity measurement module 1065, a full-angle wind direction sensor module 1066, an ultrasonic wind speed sensor module 1067 and a communication module 1068 are arranged on the supporting body 1061. The signal output ends of the temperature sensor module 1063, the sound pressure sensor module 1064, the atmospheric humidity measurement module 1065, the full-angle wind direction sensor module 1066 and the ultrasonic wind speed sensor module 1067 are all connected to the signal input end of the communication module 1068, and the signal output end of the communication module 1068 is connected to the signal input end of the millimeter wave radar transceiver 101.
[0054] Millimeter wave radar transceiver working process Figure 3As shown, first, the modulator 302 generates a millimeter wave FMCW signal of a specific modulation frequency under the control of the MMIC chip 301, and then the power amplifier 303 strengthens the strength of the FMCW wave, and then the transmitting module 304 sends the millimeter wave signal through the antenna to form a transmitting signal. At the same time, the receiving module 305 of the transceiver receives the millimeter wave signal reflected from the target through the antenna to form a receiving signal. The received signal is then processed by the MMIC chip 301 for low noise amplification, mixing, etc., and converted into an intermediate frequency or baseband signal. Finally, the signal processing module performs a series of operations such as ADC sampling, filtering, and FFT on the intermediate frequency or baseband signal to obtain a range-Doppler matrix (RDM) (each range-Doppler cell corresponds to the strength of a target). In addition, the signal processing module performs CFAR detection, DOA estimation and other algorithms on the RDM to obtain information such as displacement, velocity, azimuth, etc. of the target, and then transmits the object feature information to the host computer 307 through the communication interface 306, which is a cloud data center in the utility model.
[0055] In this embodiment, the data feedback in the millimeter-wave radar transceiver is based on a multi-threaded method, which can simultaneously analyze and process all the reflected signals on three monitoring sections within a length of 150 meters, thereby improving the processing performance of the reflected signals.
[0056] The main function of the signal processing module is to digitally process the intermediate frequency or baseband signal output by the receiving module, and extract information such as the distance, speed, and azimuth of the target. Among them, ADC sampling is the process of converting analog signals into digital signals, that is, converting continuous waveform signals into discrete numerical sequences. In ADC sampling, it is necessary to follow the sampling theorem, that is, the sampling frequency must be greater than or equal to twice the highest frequency of the signal, otherwise it will cause aliasing, that is, the original analog signal cannot be restored from the digital signal. The performance indicators of ADC sampling include sampling rate, resolution, dynamic range, distortion, etc.; filtering is the process of removing or attenuating certain frequency components in the digital signal, that is, selectively passing or blocking the signal according to different frequency characteristics. The function of filtering is to remove noise, interference or useless information in the digital signal, and improve the signal-to-noise ratio and resolution. In the utility model, filtering is mainly low-pass filtering, and in terms of implementation, it has both finite impulse response (FIR) filtering and infinite impulse response (IIR) filtering, thereby improving operability. FFT is the abbreviation of Fast Fourier Transform. It is an algorithm that transforms digital signals from the time domain to the frequency domain, that is, decomposing digital signals into the superposition of different frequency components. FFT can effectively reduce computational complexity and time consumption and improve computational efficiency. The role of FFT is to analyze the frequency characteristics in digital signals, such as the range-Doppler matrix (RDM), that is, each range-Doppler cell corresponds to the intensity of a target.
[0057] The arrangement of measurement points for settlement monitoring mainly depends on the millimeter wave radar power, monitoring budget and on-site monitoring construction space. Figure 4a and Figure 4b Taking the standard cross section of single-track and double-track high-speed railway as examples, the recommended measurement point layout diagram is given. The specific field layout implementation can be modified in a targeted manner in combination with Figure 4. This embodiment takes the standard high-speed railway subgrade cross section as an example to give a recommended settlement observation point layout scheme. Figure 4a and Figure 4b As shown, when measuring points are arranged on the spot, three metal rods 103 are connected to each settlement plate 104. The measurements obtained by the three corner reflectors 102 are grouped together, which can eliminate the interference of difference data to a certain extent. The measurement data of all corner reflectors on each monitoring section constitute a section monitoring information, so as to comprehensively characterize the overall and local settlement of the monitoring section, which is of great significance for later data analysis and even large model training. Specifically, the data within the group on each settlement plate of this part is more superior to the traditional single-point data in terms of anti-interference ability for failure data and abnormal data. The data structure design of small group data + large cross-section group data provides better training data for the subsequent model training stage. From the perspective of the maintenance of the high-speed railway subgrade in the later stage, it is crucial to determine the source of the settlement and deformation of the high-speed railway subgrade and make corresponding disposal decisions. The settlement plates 104 in the utility model are respectively arranged at the top of the foundation, the middle of the roadbed, the side of the track, and the midpoint of the section. In the later data analysis stage, the settlement displacement value of each measuring point on the roadbed cross section can be obtained, so as to determine whether the settlement of the high-speed railway subgrade comes from the roadbed part or the foundation part. After that, the maintenance plan for the next level section can be formulated according to the corresponding disposal process.
[0058] In the data structure of small group data + large group data of cross section, the data of each settlement plate is regarded as "small group data" when storing data, and the data of all settlement plates on a cross section is "large group data", and this data structure should be based on "cross section number-settlement plate number-plate serial number", which can be expressed as follows as an example only:
[0059] {“b1-01”: Sedimentation value 1,
[0060] "b1-02": Sedimentation value 2,
[0061] "b1-03": Settlement value 3} This is an example of the data structure of plate No. 1 on a certain section;
[0062] {“#1-b1”:{{“b1-01”: Sedimentation value 1,
[0063] "b1-02": Sedimentation value 2,
[0064] "b1-03": Sedimentation value 3}},
[0065] "#1-b2":{{"b2-01":sedimentation value 1,
[0066] "b2-02": Sedimentation value 2,
[0067] "b2-03": Sedimentation value 3}},
[0068] …
[0069] }
[0070] The above data can be stored in a dictionary type, and the storage is implemented based on a key-value pair. Therefore, the advantage of the data structure of small group data + large cross-sectional group data obtained by using the utility model is that the structured data is more conducive to the later analysis of the data and has higher storage security.
[0071] For the deployment of the on-site monitoring system, please refer to the attached Figure 5a-5c To plan the overall layout. First, install the commonly used equal-diameter annular concrete poles 201 at a certain distance next to the railway section roadbed where the settlement needs to be monitored. The height should be 6 to 8 meters to ensure that it is far enough from the ground. The distance between these equal-diameter annular concrete poles 201 and the rails and roadbed 202 should be kept as far away from the roadbed settlement as possible. The millimeter-wave radar transceiver 101 and the integrated meteorological station 106 are fixed on the top of the equal-diameter annular concrete poles 201, and the antenna surface of the radar is spherical. In order to enhance the signal of the settlement plate 103 and facilitate radar detection, the strongest reflection direction of the corner reflector 102 should point to the microwave radar antenna.
[0072] Exemplarily, the concrete pole is a standard structure of the railway system, which is easy to manufacture and install, and is beneficial to cost control and subsequent maintenance.
[0073] In the establishment and training phase of the cloud LSTM algorithm model, the cloud data platform 108 first needs to process the data sent back by the signal base station 107, mainly including measured settlement, temperature, humidity, atmospheric pressure, wind speed, wind direction and other data. After that, the data is cleaned and preprocessed to remove outliers and process missing values, and the features are normalized and stored in the database. At the same time, the preprocessed data is divided into a training set, a validation set and a test set, usually divided according to the time series to ensure that the model has good generalization ability.
[0074] The cloud data platform 108 may select an existing data processing platform. For example, the server may select HPE ProLiant DL380 Gen10, the disk array may select HPE (HPE Smart Array), and the storage area network (SAN) may select Fibre Channel (FC).
[0075] Existing data analysis software can be configured in the cloud data platform 108, such as an LSTM algorithm model to analyze the collected data.
[0076] For example, Figure 6 As shown in the figure, the model is built using Python language, and the environment is built based on third-party libraries such as PyTorch, NumPy, Matplotlib, and Seaborn for deep learning. The input sequence is used as the input of the model, and the temporal dependencies in the sequence data are captured through multiple LSTM units. The complexity and representation ability of the model can be flexibly controlled by adjusting the number of layers and the number of units in each layer. The input layer and output layer of the model as well as the dimensions of the input and output are preliminarily formulated with reference to the existing model, and whether to add other neural network layers is determined according to the needs of the actual task. In the model training phase, the LSTM model is trained using the training set, and the model parameters are gradually learned and optimized through the back propagation algorithm and optimizer to minimize the difference between the predicted output and the actual observed value. Through the correlation between relevant factors in the training data (such as precipitation value, temperature, wind speed, wind direction, and atmospheric humidity), the model can learn the complex nonlinear relationship between them and future precipitation values. The predicted output result is obtained by passing the input sequence into the model. Then, the loss between the predicted results and the measured data is calculated, and the parameters of the model are updated using the back propagation algorithm. The weights and biases of the model are optimized based on optimization algorithms such as Adam and SGD. The above training steps are iterated until the model training effect reaches the preset loss convergence condition. In the model evaluation stage, the performance of the trained model is evaluated using the validation set. Evaluation indicators such as root mean square error or mean absolute error are calculated between the predicted results and the measured data. According to the performance of the validation set, hyperparameters can be adjusted, such as adding network layers and adjusting the learning rate. Training and evaluation are repeated until the expected model performance is obtained. In the model application stage, the performance of the tuned LSTM model is finally evaluated using the test set, and the performance of the model is verified on real data. The trained model is used for prediction, and the new sedimentation data and related environmental variable factors are input into the model, and the prediction results are obtained through forward propagation. The difference between the predicted results and the actual situation is analyzed, and the model is improved or optimized to improve the accuracy and performance of the prediction. Based on the trained model, the real-time received data is predicted and monitored. By comparing the measured value with the predicted value, an alarm signal is issued in time to notify the high-speed rail operation and maintenance department to take countermeasures. At the same time, the prediction results can be visualized to provide users with real-time monitoring and decision support.
[0077] Finally, it will enter the normal operation stage. In this stage, the cloud data center will continuously optimize and iterate the cloud data analysis architecture based on feedback from actual applications and new data, including improving model performance, optimizing computing resource utilization, and improving system reliability, to provide high-quality high-speed rail settlement monitoring and forecasting services.
[0078] It should be noted that the cloud data platform 108 in the present invention mainly plays the role of data storage and analysis and processing. Data storage depends on the memory, and the analysis and processing function depends on the processor. The analysis and processing method is based on existing known algorithms. The machine learning algorithms listed above are only examples. The present invention can also use other algorithms to analyze data, such as unsupervised learning algorithm models, etc.
[0079] It should be noted that the principle, structure and method of the utility model can also be applied to the settlement monitoring of different structures such as railways, highways, bridges, etc., as well as cloud data analysis and large model construction. The difference lies in the working environment of various target structures, the type and size of loads, and the settlement limit values. According to the specific situation, targeted modifications can be made to the monitoring layout, monitoring accuracy, and the detailed technology stack used in the training model, deep learning model, etc. Based on the basic principles and scope of the utility model, technicians can make corresponding changes and improvements without violating the principles, and these changes and improvements belong to the protection scope of the present invention.
[0080] The above implementation modes are only used to illustrate the present invention, but not to limit the present invention. Ordinary technicians in the relevant technical field can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all equivalent technical solutions also belong to the scope of the present invention. The scope of patent protection of the present invention should be limited by the claims.
Claims
1. A high-speed railway subgrade intelligent monitoring system, characterized in that: include: Millimeter wave radar transceiver; A corner reflector, wherein the corner reflector is connected to the millimeter wave radar transceiver signal; Settling plate; A metal rod, wherein the metal rod is arranged on the settling plate, and the corner reflector is arranged on the metal rod; An integrated weather station, wherein a signal output terminal of the integrated weather station is connected to a signal input terminal of the millimeter wave radar transceiver; A signal base station, wherein the signal output terminal of the millimeter wave radar transceiver is connected to the signal input terminal of the signal base station; Cloud data platform, the signal output end of the signal base station is connected to the signal input end of the cloud data platform.
2. The high-speed railway subgrade intelligent monitoring system according to claim 1 is characterized in that: The integrated meteorological station includes a supporting body, which is fixedly supported by a supporting connector. A temperature sensor module, a sound pressure sensor module, an atmospheric humidity measurement module, a full-angle wind direction sensor module, an ultrasonic wind speed sensor module and a communication module are arranged on the supporting body. The signal output ends of the temperature sensor module, the sound pressure sensor module, the atmospheric humidity measurement module, the full-angle wind direction sensor module and the ultrasonic wind speed sensor module are all connected to the signal input end of the communication module, and the signal output end of the communication module is connected to the signal input end of the millimeter wave radar transceiver.
3. The high-speed railway subgrade intelligent monitoring system according to claim 1 is characterized in that: The millimeter-wave radar transceiver includes a chip, a modulator, a power amplifier, a transmitting module, a receiving module and a communication interface. The signal output end of the chip is connected to the modulator, the power amplifier and the transmitting module in sequence, the signal output end of the receiving module is connected to the signal input end of the chip, and the signal output end of the chip is connected to the signal input end of the communication interface.
4. The high-speed railway subgrade intelligent monitoring system according to claim 3 is characterized in that: The chip is an MMIC chip.
5. The high-speed railway subgrade intelligent monitoring system according to claim 1 is characterized in that: The three metal rods are connected to the settling plate, and each metal rod is provided with a corner reflector.
6. The high-speed railway subgrade intelligent monitoring system according to claim 1 is characterized in that: The settlement plates are arranged in plurality and are respectively arranged at the top of the foundation, the middle of the roadbed, the side of the track, and the midpoint of the section.
7. The high-speed railway subgrade intelligent monitoring system according to claim 1 is characterized in that: The millimeter wave radar is arranged at the upper end of an equal-diameter annular concrete rod.
8. The high-speed railway subgrade intelligent monitoring system according to claim 7 is characterized in that: The equal-diameter annular concrete rods are arranged beside the railway section embankment whose settlement is to be monitored.
9. The high-speed railway subgrade intelligent monitoring system according to claim 7 is characterized in that: The height of the equal-diameter annular concrete pole is 6 to 8 meters.
10. The high-speed railway subgrade intelligent monitoring system according to claim 1 is characterized in that: The utility model also comprises a concrete base, and a permanent control point is arranged at the upper end of the concrete base.
