Search

CN-122009267-A - On-line monitoring device based on subway environmental noise inversion rail ripples grinds state

CN122009267ACN 122009267 ACN122009267 ACN 122009267ACN-122009267-A

Abstract

The invention discloses an on-line monitoring device for inverting a rail wave grinding state based on subway environmental noise, and relates to the technical field of rail transit. The device comprises a base and a support frame, wherein the base and the support frame are arranged beside a track, an acoustic signal acquisition unit and a working condition acquisition unit are arranged on the support frame, the base is provided with an electric cabinet, and an edge calculation processing unit, a self-calibration unit, a cloud platform data management unit and a power supply unit are integrated in the electric cabinet. According to the invention, a non-contact acoustic monitoring scheme is adopted, the signal-to-noise ratio of the wave mill characteristic is improved through a two-stage signal purification algorithm, the wave mill full-parameter high-precision inversion is realized by combining a deep learning model, the self-calibration and cloud platform operation and maintenance functions are matched, the all-weather online monitoring, intelligent early warning and full-line closed-loop operation and maintenance of the subway steel rail wave mill can be realized, the steel rail structure is not damaged, and the complex operation environment of the subway tunnel is adapted.

Inventors

  • LI MINGDI
  • LIU QINGJIE

Assignees

  • 华东交通大学

Dates

Publication Date
20260512
Application Date
20260319

Claims (9)

  1. 1. The on-line monitoring device for inverting the rail wave grinding state based on subway environmental noise is characterized by comprising an electric cabinet (1), a base (3), an acoustic signal acquisition unit, a working condition acquisition unit and a support frame (7), wherein the base (3) is arranged beside a track, the support frame (7) is arranged on the base, the acoustic signal acquisition unit and the working condition acquisition unit are arranged on the support frame (7), and the electric cabinet (1) is further arranged on the base (3); The sound signal acquisition unit is used for acquiring subway environment multichannel sound signals when a train passes through and completing amplification, filtering and analog-to-digital conversion of the signals; The working condition acquisition unit is used for acquiring running working condition data of a train passing through in real time and pre-storing line basic parameters of a monitoring section; an edge computing processing unit, a self-calibration unit, a cloud platform data management unit and a power supply unit are arranged in the electric cabinet (1); The edge calculation processing unit is respectively and electrically connected with the acoustic signal acquisition unit and the working condition acquisition unit and is used for purifying the acoustic signal and extracting the characteristics; the self-calibration unit is electrically connected with the edge calculation processing unit and is used for correcting monitoring errors caused by environmental changes and device drift; The cloud platform data management unit is in communication connection with the edge calculation processing unit through the wireless communication module and is used for completing storage, trend prediction, visual display and intelligent maintenance decision of full-line wave mill data; The power supply unit provides stable power supply for each unit of the device.
  2. 2. The on-line monitoring device based on subway environmental noise inversion rail wave grinding state of claim 1, wherein the acoustic signal acquisition unit is arranged on the inner side of the support frame (7) and comprises a microphone array, a pre-amplification module, a band-pass filter module and a high-speed AD acquisition module; The microphone array is a linear uniform array and is longitudinally arranged along the inner side of the supporting frame (7), the array comprises 8-16 microphone array elements, the array element distance is 5-10cm, the microphone array corresponds to half wavelength of the rail wave mill core characteristic frequency, the pre-amplification module is a multichannel low-noise amplifier and corresponds to the microphone array elements one by one, the passband range of the band-pass filter module is 200Hz-5kHz, the core frequency interval of subway rail wave mill fundamental waves and higher harmonics is covered, the sampling rate of the high-speed AD acquisition module is more than or equal to 20kHz, the resolution is more than or equal to 16 bits, and multichannel acoustic signals are synchronously acquired.
  3. 3. The on-line monitoring device for inverting the rail wave grinding state based on the subway environmental noise according to claim 1 is characterized in that the working condition acquisition unit comprises a laser Doppler speed measurement sensor (5), an axle weight identification module and a line parameter storage module; The laser Doppler speed measuring sensor (5) is arranged at the top of the supporting frame (7) and used for collecting the instantaneous speed of a train passing through the monitoring section in real time, the axle weight identification module is connected into the subway train TCMS system and used for obtaining the axle weight data of the train, and the line parameter storage module is used for pre-storing the rail model, curve radius, super-high value, fastener type and sleeper interval line basic parameters of the monitoring section.
  4. 4. The on-line monitoring device for inverting the rail wave grinding state based on the subway environmental noise according to claim 1 is characterized in that the edge computing processing unit adopts an ARM+FPGA heterogeneous processor architecture and comprises a signal preprocessing module, a wave grinding characteristic extraction module, a wave grinding state inversion module and an abnormality early warning module; The system comprises a signal preprocessing module, a wave grinding characteristic extraction module, a wave grinding state inversion module, an abnormality early warning module, a cloud platform data management unit and a cloud platform data management unit, wherein the signal preprocessing module is used for carrying out time delay alignment and gain calibration on multichannel acoustic signals, the wave grinding characteristic extraction module is used for carrying out time frequency analysis of the combination of wavelet transformation and Hilbert-Huang transformation on purified wave grinding characteristic noise signals and extracting wave grinding characteristic frequency, harmonic components and time frequency energy distribution characteristics, the wave grinding state inversion module is internally provided with a multi-parameter fusion inversion model which is finished through pre-training, and the abnormality early warning module is internally provided with a wave grinding safety threshold and a development rate threshold and synchronously uploads early warning information to the cloud platform data management unit.
  5. 5. The on-line monitoring device based on subway environmental noise inversion rail wave grinding state of claim 4, wherein the multi-parameter fusion inversion model is a CNN-LSTM deep learning inversion model, and the pre-training data set comprises a laboratory wheel-rail test bed full-working-condition wave grinding noise data set and a field line calibration actual measurement data set, and is adapted to rail and working conditions of different subway lines through migration learning.
  6. 6. The on-line monitoring device for inverting the rail wave grinding state based on the subway environmental noise according to claim 1 is characterized in that the self-calibration unit comprises a standard sound source, an environmental parameter sensor module (6) and a calibration control module; The system comprises a laser Doppler velocity measurement sensor (5), a standard sound source, an environment parameter sensor module (6) and a calibration control module, wherein the standard sound source is arranged at the side of the laser Doppler velocity measurement sensor (5) and can emit standard sine sound signals with fixed frequency and fixed sound pressure level, the environment parameter sensor module (6) is arranged at a proper position on a support frame (7) and comprises temperature, humidity and air pressure sensors, environmental parameters of monitoring points are collected in real time and used for correcting the influence of sound velocity change on beam forming time delay, and the calibration control module is used for automatically starting a calibration process in a daily skylight period.
  7. 7. The on-line monitoring device for inverting the rail wave grinding state based on the subway environmental noise according to claim 1 is characterized in that the cloud platform data management unit comprises a data storage module, a wave grinding development prediction module, a full-line visualization module and an intelligent maintenance decision module; The data storage module is used for receiving and storing the wave grinding state data, the original sound signal data, the working condition data and the early warning information uploaded by the edge calculation processing units of the monitoring points for a long time; the full-line visualization module is used for visually displaying the position, the severity and the development state of the wave mill of each monitoring point of the full line, and the intelligent maintenance decision module is used for generating a steel rail grinding maintenance scheme according to the real-time state and the prediction result of the wave mill.
  8. 8. The on-line monitoring device for inverting the rail wave grinding state based on the subway environmental noise is characterized in that the power supply unit comprises a subway direct current power supply module and a standby lithium battery pack, wherein the subway direct current power supply module is connected to a subway power supply grid, and the standby lithium battery pack is used for coping with a power grid outage scene.
  9. 9. The on-line monitoring device based on subway environment noise inversion rail ripples ground state of claim 1, wherein the protection screen (2) is covered on the top of the support frame (7) and is used for covering the calibration sound source module (4), the laser Doppler speed sensor (5) and the environment parameter sensor module (6).

Description

On-line monitoring device based on subway environmental noise inversion rail ripples grinds state Technical Field The invention relates to the technical field of rail transit, in particular to an on-line monitoring device for inverting a rail wave grinding state based on subway environmental noise. Background The rail wave mill is a periodical wave-shaped abrasion phenomenon of the top surface of the subway rail along the longitudinal direction, and is one of typical diseases of a subway rail system. The wave grinding can greatly aggravate wheel rail vibration and radiation noise, accelerate fatigue damage of rail and vehicle parts such as fasteners, sleepers, bearings and the like, and even influence train running stability and running safety when serious. The existing subway rail wave mill monitoring technology mainly has the following defects: 1. The off-line detection technology relies on the on-line operation of an artificial skylight period, adopts the rail wave mill measuring instrument to measure section by section, has extremely low operation efficiency and high labor intensity, cannot capture the dynamic development process of the wave mill in real time, and is easy to cause disease omission and early warning lag. 2. The contact type on-line monitoring technology needs to install acceleration and displacement sensors on rail web and rail bottom drilling, so that the structural strength of the steel rail is damaged, the sensors are easily damaged in severe environments of moist, dust and strong vibration of a subway tunnel, the maintenance cost is extremely high, meanwhile, the sensors are influenced by strong impact of wheel rails, the service life of the sensors is short, and the data stability is poor. 3. The existing non-contact acoustic monitoring technology only collects noise through a single microphone to carry out simple spectrum analysis, can not solve the interference problem of strong background noise (carriage noise, ventilation noise and tunnel wall reflection reverberation) in a subway tunnel, is difficult to accurately extract wave mill characteristic noise, can only qualitatively judge whether wave mill exists or not, can not quantitatively invert core parameters such as wave mill wavelength, wave depth, position and the like, has low inversion precision, and can not meet engineering operation and maintenance requirements. Therefore, the research and development of the rail corrugation monitoring device which is non-contact, high in anti-interference capability, high in monitoring precision and capable of running on line all the day becomes an urgent need in the field of subway rail operation and maintenance. Disclosure of Invention Aiming at the defects of the prior art, the invention solves the technical problems that the prior rail wave mill monitoring technology has the defects of damage to the rail structure, poor real-time performance of off-line detection, weak anti-interference capability of acoustic monitoring, incapability of quantitatively inverting the wave mill full parameters and difficulty in realizing full-line closed-loop operation and maintenance, and provides an on-line monitoring device for inverting the rail wave mill state based on subway environmental noise. In order to achieve the above purpose, the technical scheme adopted by the invention is as follows: The on-line monitoring device based on subway environment noise inversion steel rail wave grinding state comprises an electric cabinet, a base, an acoustic signal acquisition unit, a working condition acquisition unit and a support frame, wherein the base is arranged beside a rail and is provided with the support frame, the support frame is provided with the acoustic signal acquisition unit and the working condition acquisition unit, and the base is also provided with the electric cabinet; The sound signal acquisition unit is used for acquiring subway environment multichannel sound signals when a train passes through and completing amplification, filtering and analog-to-digital conversion of the signals; The working condition acquisition unit is used for acquiring running working condition data of a train passing through in real time and pre-storing line basic parameters of a monitoring section; an edge computing processing unit, a self-calibration unit, a cloud platform data management unit and a power supply unit are arranged in the electric cabinet; the edge calculation processing unit is respectively and electrically connected with the acoustic signal acquisition unit and the working condition acquisition unit, and is used for purifying and extracting characteristics of acoustic signals, inverting the rail wave grinding state by combining working condition data, and completing local abnormality early warning; The self-calibration unit is electrically connected with the edge calculation processing unit and is used for correcting monitoring errors caused by environmental changes and device drift and guara