CN114684217B - Rail transit health monitoring system and method - Google Patents
Rail transit health monitoring system and method Download PDFInfo
- Publication number
- CN114684217B CN114684217B CN202210259899.3A CN202210259899A CN114684217B CN 114684217 B CN114684217 B CN 114684217B CN 202210259899 A CN202210259899 A CN 202210259899A CN 114684217 B CN114684217 B CN 114684217B
- Authority
- CN
- China
- Prior art keywords
- data
- neural network
- layer
- health
- vibration
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000036541 health Effects 0.000 title claims abstract description 102
- 238000000034 method Methods 0.000 title claims abstract description 55
- 238000012544 monitoring process Methods 0.000 title claims abstract description 53
- 238000013528 artificial neural network Methods 0.000 claims abstract description 85
- 230000004044 response Effects 0.000 claims abstract description 28
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 239000000835 fiber Substances 0.000 claims description 58
- 239000000523 sample Substances 0.000 claims description 52
- 230000003287 optical effect Effects 0.000 claims description 43
- 238000012549 training Methods 0.000 claims description 43
- 230000006870 function Effects 0.000 claims description 42
- 238000011176 pooling Methods 0.000 claims description 26
- 238000010606 normalization Methods 0.000 claims description 19
- 230000008569 process Effects 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 13
- 238000012423 maintenance Methods 0.000 claims description 11
- 238000002310 reflectometry Methods 0.000 claims description 9
- 238000004458 analytical method Methods 0.000 claims description 8
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 6
- 238000003491 array Methods 0.000 claims description 6
- 230000007613 environmental effect Effects 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 6
- 230000009467 reduction Effects 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 4
- 239000000853 adhesive Substances 0.000 claims description 3
- 230000001070 adhesive effect Effects 0.000 claims description 3
- 239000004568 cement Substances 0.000 claims description 3
- 239000003795 chemical substances by application Substances 0.000 claims description 3
- 238000013527 convolutional neural network Methods 0.000 description 18
- 239000013307 optical fiber Substances 0.000 description 17
- 238000001514 detection method Methods 0.000 description 12
- 238000010586 diagram Methods 0.000 description 10
- 238000003860 storage Methods 0.000 description 10
- 238000004590 computer program Methods 0.000 description 7
- 230000004913 activation Effects 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 5
- 238000012806 monitoring device Methods 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 210000002569 neuron Anatomy 0.000 description 4
- 230000035945 sensitivity Effects 0.000 description 4
- 230000002085 persistent effect Effects 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 239000006185 dispersion Substances 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000005284 excitation Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 239000000758 substrate Substances 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 238000002759 z-score normalization Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000009545 invasion Effects 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 230000008376 long-term health Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000003801 milling Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000004642 transportation engineering Methods 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61K—AUXILIARY EQUIPMENT SPECIALLY ADAPTED FOR RAILWAYS, NOT OTHERWISE PROVIDED FOR
- B61K9/00—Railway vehicle profile gauges; Detecting or indicating overheating of components; Apparatus on locomotives or cars to indicate bad track sections; General design of track recording vehicles
- B61K9/08—Measuring installations for surveying permanent way
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
- B61L15/00—Indicators provided on the vehicle or train for signalling purposes
- B61L15/0054—Train integrity supervision, e.g. end-of-train [EOT] devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Mechanical Engineering (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to a rail transit health monitoring system and a method, wherein the method comprises the following steps: acquiring response data; preprocessing according to the response data, and determining vibration data to be detected; and inputting the vibration data to be tested into a first neural network and/or a second neural network which are completely trained, and determining the predicted track structure health category and the predicted vehicle body health category. According to the method, after vibration data to be measured are input into the related neural network, after the result is directly predicted, related personnel can refer to the predicted result to know the health condition and the running speed of the subway in time so as to reduce accidents.
Description
Technical Field
The invention relates to the technical field of traffic safety monitoring, in particular to a rail transit health monitoring system and method.
Background
In recent years, urban rail transportation systems are rapidly developed, and the healthy service of the urban rail transportation engineering structure serving as an important component of major structural engineering and urban traffic pulse-setting is important for the normal operation of cities. However, the track traffic engineering structure is gradually damaged along with operation, and especially under the influence of complex load conditions such as temperature, train, basic deformation, external invasion and the like, diseases can occur at any time, and the driving safety is influenced. The damage of the track traffic engineering structure shows frequent and sudden occurrence, so that the construction of a full-time global safety monitoring system for the track traffic engineering structure is needed to comprehensively and timely master the health conditions of vehicles and the track engineering structure, timely early warning and alarming potential accident hazards of the potential damage and the sudden accident, and protecting navigation for track traffic safety operation. At present, most of urban rail transit engineering structure state detection and maintenance in China are in a regular maintenance or post-fault repair mode, and the operation, maintenance and maintenance cost is high, the efficiency is low, and the real-time monitoring performance cannot be guaranteed. Therefore, the technology and the management level of the safety monitoring technology of the rail traffic engineering infrastructure are improved by utilizing new technology and new methods. Traditional electric detection system receives electromagnetic interference seriously in electrified circuit section, and the signal can't be transmitted in a long distance. The optical fiber sensor has the advantages of good long-term stability, long-distance signal transmission, easiness in networking, electromagnetic interference resistance and the like, gradually replaces an electric sensor with long-term stability which is difficult to guarantee, and is widely used in long-term health monitoring in the fields of bridges, tunnels, airports, railways and the like.
In the prior art, the invention patent 201110009091.1 'an on-line monitoring system for the tread state of a train wheel' discloses a train wheel health state system based on a plurality of fiber gratings. However, the reliability of the system detection is not high, false alarm is easy to be caused, for example, when a sensing device is in a problem, strain signals are abnormal, and false judgment is easy to be caused; the invention patent 202010666848.3 'a high-speed railway rail safe operation detection method based on optical fiber distributed vibration monitoring' adopts a traditional optical fiber distributed vibration detection method, and the traditional distributed optical fiber sensing technology has the defects that the scattering coefficient in optical fibers is low, the signal-to-noise ratio of a system is usually not high, and the spatial resolution and the detection sensitivity of a distributed sensing system are further affected. To solve this problem, researchers often use a multiple averaging algorithm or a method of artificially increasing the scattering coefficient. Averaging thousands of times can improve the signal-to-noise ratio but can result in an increase in system response time. The improvement of the back scattering intensity by ultraviolet exposure or femtosecond laser processing and other methods can effectively improve the signal to noise ratio, but the uniformity and timeliness of the process are limited. Therefore, how to effectively monitor the rail transit by using the fiber bragg grating sensor is a problem to be solved.
Disclosure of Invention
In view of the foregoing, it is necessary to provide a system and a method for monitoring the health of rail transit, which are used for solving the problem that it is difficult to efficiently utilize the fiber grating sensor to predict the condition of rail transit in the prior art.
In order to solve the above technical problems, the present invention provides a rail transit health monitoring system, including: vibration sensing network, data processing center and intelligent fortune dimension service platform of communication connection in proper order, wherein:
the vibration sensing network comprises a fiber grating vibration sensing optical cable arranged on a track traffic engineering structure;
the data processing center is used for processing monitoring signals of the fiber grating vibration sensing optical cable, the fiber grating strain sensing optical cable and the fiber grating array temperature sensing optical cable to determine response data;
the intelligent operation and maintenance service platform is used for performing intelligent analysis according to the response data and monitoring the track structure and the vehicle based on an intelligent analysis result.
Further, the fiber bragg grating arrays in the vibration sensing network are all made of fiber bragg grating array sensing probes, the fiber bragg grating array sensing probes comprise preset reflectivity grating arrays, and fiber bragg grating sensing points with preset scale are multiplexed on a single fiber.
Further, the fiber gratings on each fiber grating array sensing probe are distributed at equal intervals, the fiber gratings with equal intervals are used as nodes, and the adjacent nodes form vibration sensing units.
Further, vibration sensing optical cable in the vibration sensing network is closely attached to the tested structure, wherein:
if the tested structure belongs to the existing line, directly adhering the structure adhesive to the tested structure, or placing the vibration sensing optical cable after slotting, and then curing the vibration sensing optical cable and the tested structure by using cement or curing agent;
if the detected structure belongs to the newly repaired line, directly pouring the vibration sensing optical cable into the concrete.
The invention also provides a rail transit health monitoring method which is applied to the intelligent operation and maintenance service platform in the rail transit health monitoring system, and the method comprises the following steps:
acquiring response data;
preprocessing according to the response data, and determining vibration data to be detected;
and inputting the vibration data to be tested into a first neural network and/or a second neural network which are completely trained, and determining the predicted track structure health category and the predicted vehicle body health category.
Further, the response data includes vibration data of the pickup in the ith section between a first moment when the train approaches and a second moment when the train leaves, the preprocessing is performed according to the response data, and the determining of the data to be detected includes:
Adding environmental noise signals to the front end and the rear end of the vibration data to determine filling data;
and carrying out data normalization processing on the filling data to determine the data to be tested.
Further, the training process of the first neural network and/or the second neural network comprises:
outputting a predicted track structure health class when a sample set is input to the first neural network, wherein the sample set includes sample vibration data and a corresponding actual track structure health class;
outputting a predicted vehicle body health class when a sample set is input to the second neural network, wherein the sample set includes the sample vibration data and a corresponding actual vehicle body health class;
determining a first loss function according to the error between the predicted track structure health category and the actual track structure health category, and training a constructed first neural network according to the first loss function until convergence;
and determining a second loss function according to the error between the predicted vehicle body health category and the actual vehicle body health category, and training the constructed second neural network to be converged according to the second loss function.
Further, the first neural network sequentially includes: input layer, convolution layer, max-pooling layer, flame layer, full connection layer, and output layer, wherein:
the input layer is used for inputting the sample set;
the convolution layer is used for carrying out convolution operation on the sample set and determining first convolution data;
the maximum pooling layer comprises a pooling layer and a Batch Normalization layer, wherein the pooling layer is used for carrying out feature extraction and dimension reduction on the first convolution data, and the Batch Normalization layer is used for carrying out standardization on the dimension reduced first convolution data to determine standardized data;
the flat layer is used for expanding the standardized data into a one-dimensional vector;
the full connection layer is used for reducing the dimension of the one-dimensional vector and classifying the vector through a softmax function to determine the predicted vehicle body health category.
Further, the second neural network sequentially includes: BP neural network and one-dimensional neural network, wherein:
the BP neural network is used for training a speed training set and outputting a predicted vehicle speed, wherein the speed training set comprises sample vibration data and corresponding speed labels;
The one-dimensional neural network is used for training the sample set through the input layer, the convolution layer, the maximum pooling layer, the flat layer, the full connection layer and the output layer in sequence, and determining the predicted vehicle body health type.
Further, in the training process of the first neural network and/or the second neural network, an Adam optimizer is adopted to adaptively adjust a learning rate, and the first loss function and the second loss function are multiple cross entropy functions.
Compared with the prior art, the invention has the beneficial effects that: firstly, response data are effectively acquired; furthermore, relevant effective data are extracted through preprocessing the response data, vibration data to be detected are formed, and accuracy of the vibration data to be detected is guaranteed; finally, the vibration data to be detected is effectively identified and judged through the first neural network and/or the second neural network, the characteristic information is extracted, the track structure health type and the vehicle body health type are predicted, the rapid and efficient judgment of various traffic conditions based on the vibration information is ensured, and the vibration data is judged. In summary, the invention realizes high-sensitivity acquisition of response data based on the vibration sensing network, obtains various traffic running information only based on the vibration data to be detected after the vibration data to be detected is input into the related neural network, ensures the high efficiency and rapidity of an algorithm, and enables related personnel to know the health condition and running speed of the subway in time by referring to the prediction result after directly predicting the result so as to reduce accidents and ensure the timeliness of feedback.
Drawings
FIG. 1 is a schematic flow chart of an embodiment of a rail transit security monitoring system according to the present invention;
FIG. 2 is a schematic flow chart of an embodiment of a method for monitoring rail transit health according to the present invention;
fig. 3 is a flowchart of an embodiment of step S202 in fig. 2 according to the present invention;
FIG. 4 is a flow chart of an embodiment of a training process of the first neural network and/or the second neural network according to the present invention;
FIG. 5 is a schematic diagram illustrating a first embodiment of a neural network according to the present invention;
fig. 6 is a schematic structural diagram of an embodiment of a BP neural network according to the present invention;
fig. 7 is a schematic structural diagram of an embodiment of a rail transit health monitoring device provided by the present invention;
fig. 8 is a schematic structural diagram of an embodiment of an electronic device according to the present invention.
Detailed Description
Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form a part hereof, and together with the description serve to explain the principles of the invention, and are not intended to limit the scope of the invention.
In the description of the present invention, the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. Furthermore, the meaning of "a plurality of" means at least two, such as two, three, etc., unless specifically defined otherwise.
Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the invention. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly understand that the described embodiments may be combined with other embodiments.
The invention provides a rail transit health monitoring system and a rail transit health monitoring method, which are characterized in that the rail transit health and the vehicle body health are comprehensively judged by collecting persistent data sample information and classifying by utilizing a neural network, so that a new thought is provided for further improving the accuracy and the high efficiency of monitoring the rail transit health state.
Before the description of the embodiments, the related words are interpreted:
optical fiber sensor: is a sensor that converts the state of a measured object into a measurable optical signal. The working principle of the optical fiber sensor is that the light beam incident by the light source is sent into the modulator through the optical fiber, and the interaction with the external measured parameters in the modulator causes the optical properties of the light such as the intensity, wavelength, frequency, phase, polarization state and the like of the light to change into the modulated optical signal, and then the modulated optical signal is sent into the photoelectric device through the optical fiber and the measured parameters are obtained after the optical signal is sent into the demodulator. In the whole process, the light beam is led in through the optical fiber and then is emitted out after passing through the modulator, wherein the optical fiber firstly transmits the light beam and secondly plays the role of the optical modulator;
Optical fiber array: an array is formed by using a V-Groove (V-Groove) substrate to mount a bundle of optical fibers or an optical fiber ribbon on the substrate at a prescribed interval;
an optical cable: in order to meet the performance specifications of optical, mechanical or environmental, it is a communications cable assembly that utilizes one or more optical fibers disposed in a covering sheath as a transmission medium and that can be used alone or in groups.
Based on the description of the technical terms, in the prior art, the problem that the reliability of monitoring the health state of the wheels by the optical fiber sensor is low and false alarm is easy to occur is solved, and the system response time is easy to occur due to low signal to noise ratio of the system, so that the invention aims to provide a rapid and efficient rail transit health monitoring method.
Specific embodiments are described in detail below:
the embodiment of the invention provides a track traffic safety monitoring system, and referring to fig. 1, fig. 1 is a schematic flow diagram of an embodiment of the track traffic safety monitoring system provided by the invention, including: vibration sensing network 1, data processing center 2 and intelligent fortune dimension service platform 3 of communication connection in proper order, wherein:
the vibration sensing network 1 comprises a fiber grating vibration sensing optical cable 11 which is arranged on a track traffic engineering structure;
The data processing center 2 is configured to process monitoring signals of the fiber grating vibration sensing optical cable, the fiber grating strain sensing optical cable and the fiber grating array temperature sensing optical cable, and determine sensing data;
the intelligent operation and maintenance service platform 3 is used for performing intelligent analysis according to the sensing data and monitoring the engineering structure and train state of the rail transit based on the intelligent analysis result.
In the embodiment of the invention, the vibration sensing network is arranged, and the fiber bragg grating is utilized to vibrate the sensing optical cable, so that the accuracy of monitoring vibration information is ensured, and comprehensive monitoring is realized; the data processing center is arranged to convert the monitoring information of the plurality of optical cables into sensing data; by arranging the intelligent operation and maintenance service platform, corresponding intelligent analysis is carried out on the sensing data, and the engineering structure and the train state of the rail transit are effectively fed back from the result of the intelligent analysis, so that the rapid monitoring on the real-time state of the train and the running condition of the engineering structure is ensured, and corresponding alarm processing is timely sent out.
As a preferred embodiment, the fiber grating arrays in the vibration sensing network are all made of fiber grating array sensing probes, the fiber grating array sensing probes comprise preset reflectivity grating arrays, and fiber grating sensing points with preset scale are multiplexed on a single fiber.
In the embodiment of the invention, the multiplexing capacity of the fiber bragg grating is ensured by utilizing the fiber bragg grating array sensing probe, a rail transit line of tens of kilometers can be covered only by laying a sensing network consisting of a plurality of sensing optical cables, intelligent sensing of a whole domain without blind areas is realized, and the comprehensive acquisition of monitoring signals is facilitated.
The preset reflectivity grating array is preferably an extremely weak reflectivity grating array, the fiber gratings on each fiber probe are distributed at equal intervals, the reflection bandwidth is 2-3 nm, the reflectivity consistency is good, the reflectivity range is-30-50 dB according to the multiplexing capacity of the fiber gratings on the single fiber, and the multiplexing capacity of the single fiber is larger when the reflectivity is lower.
The fiber bragg grating sensing point with the preset scale is preferably a fiber bragg grating sensing point with the ultra-large scale, so that various monitoring signals can be conveniently collected.
As a preferred embodiment, the fiber gratings on each fiber grating array sensing probe are distributed at equal intervals, the fiber gratings with equal intervals are used as nodes, and the adjacent nodes form a vibration sensing unit.
In the embodiment of the invention, the adjacent nodes are utilized to form the high-sensitivity interference type vibration sensing unit, so that the vibration pickup can sense tiny vibration abnormal information, and the sensitivity of system detection is ensured.
As a more specific embodiment, the high-sensitivity interference type vibration sensing unit is formed by taking the equidistant fiber bragg grating as a node and taking the adjacent nodes as nodes. The node distance of the fiber grating array is adjustable.
Specifically, the node distance l is set to be 3 meters or more. Taking a node spacing of 5 meters as an example, the minimum phase difference delta phi which can be perceived by each vibration pickup is less than 10mrad according to the formula:
wherein: delta phi-optical signal phase difference between adjacent fiber bragg grating nodes; lambda-optical signal wavelength; delta l-variation of grating length between adjacent grating nodes;
the minimum strain quantity which can be perceived by each vibration pickup unit is less than 50n epsilon, so that tiny vibration abnormal information caused by degradation or damage of a vehicle and a track under the coupling excitation of a wheel track can be perceived, and the detection capability of the system is obviously improved.
In the embodiment of the invention, the adjacent nodes are utilized to form the high-sensitivity interference type vibration sensing unit, so that the vibration pickup can sense tiny vibration abnormal information, and the sensitivity of system detection is ensured.
As an preferable embodiment, the vibration sensing optical cable in the vibration sensing network is closely attached to the structure to be measured, wherein:
if the tested structure belongs to the existing line, directly adhering the structure adhesive to the tested structure, or placing the vibration sensing optical cable after slotting, and then curing the vibration sensing optical cable and the tested structure by using cement or curing agent;
If the detected structure belongs to the newly repaired line, directly pouring the vibration sensing optical cable into the concrete.
In the embodiment of the invention, the structure to be tested and the vibration sensing optical cable are efficiently cured by the arrangement methods under different conditions.
As a preferred embodiment, the data processing center comprises a demodulation meter, an acquisition data network platform, wherein:
the demodulation instrument is used for converting the monitoring signals of the fiber grating vibration sensing optical cable, the fiber grating strain sensing optical cable and the fiber grating array temperature sensing optical cable into digital signals;
the data acquisition network platform is used for acquiring the digital signals in real time, converting the digital signals into the sensing data and uploading the sensing data to the intelligent operation and maintenance service platform.
In the embodiment of the invention, a computer software and hardware and a network platform are adopted to have an information acquisition function, the fiber bragg grating array sensing optical cable is connected into the fiber bragg grating array sensing demodulation instrument, the optical signals sensed by the sensing optical cable are converted into digital signals through the demodulation instrument, and the monitoring data of each sensing unit in the sensing network are acquired in real time by utilizing the computer software and hardware and the network platform, so that the detection of the whole domain of the track traffic engineering structure is realized.
As a preferred embodiment, referring to fig. 2, fig. 2 is a schematic flow chart of an embodiment of a rail transit health monitoring method provided by the present invention, including steps S201 to S203, wherein:
in step S201, response data is acquired;
in step S202, preprocessing is performed according to the response data, and vibration data to be measured is determined;
in step S203, the vibration data to be tested is input to the first neural network and/or the second neural network with complete training, and the predicted track structure health class and the predicted vehicle body health class are determined.
In the embodiment of the invention, firstly, response data is effectively acquired; furthermore, relevant effective data are extracted through preprocessing the response data, vibration data to be detected are formed, and accuracy of the vibration data to be detected is guaranteed; finally, the vibration data to be detected is effectively identified and judged through the first neural network and/or the second neural network, the characteristic information is extracted, the track structure health type and the vehicle body health type are predicted, the rapid and efficient judgment of various traffic conditions based on the vibration information is ensured, and the vibration data is judged.
As a more specific embodiment, the ultra-weak reflectivity fiber bragg grating vibration sensing optical cable is laid along the track structure in a full length, and the track line is divided into N sections by fiber bragg grating nodes:
In the above formula, L is the track line length, and L is the fiber grating pitch.
The invention can collect signals of N vibration pickers in all intervals in real time, the sampling frequency is more than or equal to 1000Hz, but the invention utilizes wheel-rail coupling excitation data, and most of real-time collection is useless data, thereby increasing the storage capacity and wasting system resources. To facilitate data analysis and data management, it is necessary to reduce invalid data. In general, one persistent data sample includes the following information: interval mark i (i is more than or equal to 1 and less than or equal to N); a time t0 when the train approaches the section i; the moment t1 when the train leaves the section i; vibration data continuously collected by the interval i vibration pickup in the time period of t0 and t 1.
In general, the time when the pickup of the selectable section i-k detects that a car passes is denoted as t0, and the time when the pickup of the section i+k detects that a car passes is denoted as t1, wherein:
in the above, L c Is the train length.
As a preferred embodiment, referring to fig. 3, fig. 3 is a flowchart of an embodiment of step S202 in fig. 2 provided by the present invention, where step S202 includes steps S301 to S302, and the step S includes:
in step S301, adding environmental noise signals to the front and rear ends of the vibration data, and determining filling data;
In step S302, the filling data is subjected to data normalization processing, and the data to be measured is determined.
In the embodiment of the invention, the accuracy of the vibration data to be detected is ensured by utilizing data filling and data normalization processing so as to further ensure the accuracy of the identification result.
As a preferred embodiment, as seen in connection with fig. 4, fig. 4 is a schematic flow chart of an embodiment of a training process of the first neural network and/or the second neural network provided by the present invention, including steps S401 to S404, wherein:
in step S401, when a sample set is input to the first neural network, outputting a predicted track structure health class, wherein the sample set includes sample vibration data and a corresponding actual track structure health class;
in step S402, outputting a predicted vehicle body health class when a sample set is input to the second neural network, wherein the sample set includes the sample vibration data and a corresponding actual vehicle body health class;
in step S403, determining a first loss function according to an error between the predicted track structure health class and the actual track structure health class, and training the constructed first neural network to converge according to the first loss function;
In step S404, a second loss function is determined according to the error between the predicted vehicle body health class and the actual vehicle body health class, and the constructed second neural network is trained to converge according to the second loss function.
In the embodiment of the invention, the first neural network and the second neural network are effectively trained by using the sample set, so that the effective identification of the track structure health category and the vehicle body health category is completed.
As a preferred embodiment, referring to fig. 5, fig. 5 is a schematic structural diagram of an embodiment of a first neural network provided by the present invention, where the first neural network sequentially includes: input layer, convolution layer, max-pooling layer, flame layer, full connection layer, and output layer, wherein:
the input layer is used for inputting the sample set;
the convolution layer is used for carrying out convolution operation on the sample set and determining first convolution data;
the maximum pooling layer comprises a pooling layer and a Batch Normalization layer, wherein the pooling layer is used for carrying out feature extraction and dimension reduction on the first convolution data, and the Batch Normalization layer is used for carrying out standardization on the dimension reduced first convolution data to determine standardized data;
The flat layer is used for expanding the standardized data into a one-dimensional vector;
the full connection layer is used for reducing the dimension of the one-dimensional vector, classifying the vector through a softmax function and determining the health type of the predicted orbit structure.
In the embodiment of the invention, a first neural network is constructed to realize feature extraction of a sample set, and a corresponding relation with an actual track structure health class is determined to complete network convergence so as to effectively predict the vehicle body health class.
The following describes the track structure health assessment algorithm flow in a specific embodiment as follows:
firstly, subway signal acquisition: storing the acquired data into a database;
secondly, because the subway speeds are inconsistent, the subway speeds are slower when the subway enters and exits, and the running speed is faster in the midway running process, so that the obtained signals are inconsistent in length, data filling is required to be carried out on the signals with shorter duration, and environmental noise is added at the front end and the rear end of the signals, so that the lengths of the signals are consistent and are 729 data points;
thirdly, data normalization: because the obtained signals are inconsistent in amplitude and uneven in dispersion, training and convergence of the neural network are not facilitated, and therefore normalization preprocessing is necessary for data. The normalization mode commonly used at present comprises two methods of min-max normalization and Z-score normalization. The embodiment of the invention adopts Z-score standardization, and the specific formula is as follows:
Where μ is the mean of all sample data and σ is the standard deviation of all sample data. The processed data conforms to the standard normal distribution, namely the mean value is 0, and the standard deviation is 1.
Fourthly, training and predicting the neural network: the obtained data are divided into a training set and a testing set according to the proportion of 9:1. Convolutional neural networks are mainly used for detection of rail health. The following six categories are mainly classified: (1) rail health; (2) rail break; (3) orbital wave milling; (4) loosening the fastener; (5) breaking the spring strip; (6) offline foundation lesion. Convolutional neural networks tend to exhibit better results for the problem of predicting results from big data. The convolutional neural network can be classified into a one-dimensional convolutional neural network, a two-dimensional convolutional neural network, and the like according to the object of use. The two-dimensional convolutional neural network is mainly used for identifying and classifying images. For text data, a one-dimensional convolutional neural network is more suitable. Therefore, the embodiment of the invention adopts a one-dimensional convolutional neural network, the training set is used for training the neural network, and the model is continuously optimized until the good performance is achieved through the adjustment of a loss function and the counter propagation principle; the test set is used for evaluating the generalization capability and orbit prediction evaluation capability of the model;
The device mainly comprises an input layer, a convolution layer (Conv 1D), a Pooling layer (Pooling), a full connection layer (FC) and an output layer as shown in FIG. 5. The subway vibration signal is used as an input tensor and is sent into a neural network for training. First, the one-dimensional convolution layer is passed, the convolution kernel size is set to 9, the number of convolution kernels is set to 32, and after the data tensor passes through the convolution layer, the data tensor enters the pooling layer. And the pooling layer is used for carrying out feature extraction and dimension reduction treatment on the original data, so that the dimension of the data can be greatly reduced, and the training and convergence of the model are accelerated. The pooling layer has two kinds of average pooling and maximum pooling, and in practice, the maximum pooling layer can often obtain better effect, so the embodiment of the invention is intended to adopt the maximum pooling layer. To prevent over-fitting of model training, each pooling layer is followed by a layer Batch Normalization, which again normalizes the data, accelerates model convergence and prevents over-fitting. After three-wheel rolling and pooling operations, the two-dimensional tensor is unfolded into a one-dimensional vector through the flat operation, the dimension is reduced through the full-connection layer, finally, the two-dimensional tensor is divided into six types through a softmax function, and a classification result is output and used for evaluating the health condition of the track, wherein the specific parameter settings are shown in the following table 1:
TABLE 1
In order to train the neural network, the training process is compiled first, and there are three important parameters, namely, an optimizer, a loss function, and an evaluation index. The optimizer to be adopted by the invention is an Adam optimizer, and the optimization method is adopted because the learning rate can be adaptively adjusted according to different parameters, and the optimization method is more suitable for large-scale data and parameter scenes. The loss function adopts a multi-element cross entropy function, because the problem is a multi-classification problem, and the method is more suitable for the multi-element cross entropy function. The specific formula is as follows:
in the above formula, y is a real label,is a predictive label；
As for the evaluation index, the accuracy was selected as an evaluation index for the classification performance of the model. Accuracy refers to the ratio of the number of correctly classified samples to the total number;
fifthly, outputting a result, and displaying by an upper computer: after the neural network predicts the result, the result is displayed on an upper computer interface, and related personnel can refer to the predicted result to know the health condition of the track in time and process the track in time so as to reduce accidents.
As a preferred embodiment, referring to fig. 6, fig. 6 is a schematic structural diagram of an embodiment of a BP neural network provided by the present invention, where the second neural network sequentially includes: BP neural network and one-dimensional neural network, wherein:
The BP neural network is used for training a speed training set and outputting a predicted vehicle speed, wherein the speed training set comprises sample vibration data and corresponding speed labels;
the one-dimensional neural network is used for training the sample set through the input layer, the convolution layer, the maximum pooling layer, the flat layer, the full connection layer and the output layer in sequence, and determining the predicted vehicle body health type.
In the embodiment of the invention, a second neural network is constructed to realize feature extraction of a sample set, determine the corresponding relation with the actual vehicle speed and the actual vehicle body health class, and complete network convergence so as to effectively predict the vehicle body health class.
As a preferred embodiment, in the training process of the first neural network and/or the second neural network, an Adam optimizer is adopted to adaptively adjust a learning rate, and the first loss function and the second loss function are multiple cross entropy functions.
In the embodiment of the invention, the rapidity and the effectiveness of the training process are ensured by utilizing an Adam optimizer and a multi-element cross entropy function.
The following describes the vehicle health assessment algorithm flow in a specific embodiment as follows:
step 1, subway signal acquisition: storing the acquired data into a database to enable the data to be persistent;
Step 2, data filling: because the subway speeds are inconsistent, when the subway is in an outbound and in an inbound, the subway speed is slower, the running speed is faster in the midway running process, and the obtained signals are inconsistent in length, so that data filling is required to be carried out on the signals with shorter duration, and environmental noise is added at the front end and the rear end of the signals, so that the lengths of the signals are consistent and are 729 data points;
step 3, data normalization: because the obtained signals are inconsistent in amplitude and uneven in dispersion, training and convergence of the neural network are not facilitated, and therefore normalization preprocessing is necessary for data. The normalization mode commonly used at present comprises two methods of min-max normalization and Z-score normalization. The embodiment of the invention adopts Z-score standardization, and the specific formula is as follows:
where μ is the mean of all sample data and σ is the standard deviation of all sample data. The processed data conforms to the standard normal distribution, namely the mean value is 0, and the standard deviation is 1.
Step 4, neural network prediction: the module is divided into two parts, one part is a BP neural network and the other part is a one-dimensional convolutional neural network. The BP neural network structure is shown in fig. 6 and is used for detecting the subway speed. One-dimensional convolutional neural networks are shown in fig. 5 for classification of vehicle states. For subway speed detection, the method is a typical scalar regression problem, a model is built by utilizing the nonlinear learning capability of the BP neural network, training is carried out by using a training set sample of a manually marked speed label in advance, and then the training set sample is used for speed prediction. The model consists of an input layer, a hidden layer and an output layer. The input layer is the subway vibration signal extracted before, the hidden layer is 32 neurons, each neuron is followed by a Relu activation function, and finally the output layer (no activation function) of one neuron is used for outputting the predicted subway speed:
Wherein the output of the neurons of each hidden layer is:
f(x)＝σ(∑wx+b)
wherein w is a weight, b is a deviation, σ is an activation function, here a Relu activation function, and the specific formula is:
f(x)＝max(0,x)
the activation function is adopted to enhance the nonlinearity of the data, so that the training of the model is facilitated;
the loss function is selected as a mean square error function (MSE, mean squared error), and the evaluation index is selected as a mean absolute error (MAE, mean absolute error) with the following formulas:
the optimizer adopts an Adam optimizer, and after 150 times of training, the model verification set converges to an average absolute error of 0.7, so that the model verification set can be more accurately used for predicting the subway speed;
the one-dimensional convolutional neural network is mainly used for detecting the health condition of the subway body. The main classifications are the following four classes:
(1) The vehicle is healthy; (2) wheel flat defects; (3) wheel eccentricity; (4) loosening the wheels. Convolutional neural networks tend to exhibit better results for the problem of predicting results from big data. The convolutional neural network can be classified into a one-dimensional convolutional neural network, a two-dimensional convolutional neural network, and the like according to the object of use. The two-dimensional convolutional neural network is mainly used for identifying and classifying images. For text data, a one-dimensional convolutional neural network is more suitable;
The structure of the one-dimensional convolutional neural network is shown in fig. 5, and the compiling of the structure and training process and the selection of the loss function are consistent with those of the first neural network, but the classification labels are different, the dimension reduction is performed through the full-connection layer, and finally the classification results are classified into four types through a softmax function, so that the classification results are output and used for evaluating the health condition of the subway, and are not repeated here;
step 5, outputting a result, and displaying by an upper computer: after the neural network predicts the result, the result is displayed on an upper computer interface, and related personnel can refer to the predicted result to know the health condition and the running speed of the subway in time so as to reduce accidents.
The embodiment of the invention further provides a rail transit health monitoring device, and as seen in conjunction with fig. 7, fig. 7 is a schematic structural diagram of an embodiment of the rail transit health monitoring device provided by the invention, and the rail transit health monitoring device 700 includes:
an acquisition unit 701 for acquiring response data;
a processing unit 702, configured to perform preprocessing according to the response data, and determine vibration data to be measured;
and the prediction unit 703 is configured to input the vibration data to be tested to a first neural network and/or a second neural network with complete training, and determine a predicted track structure health class and a predicted vehicle body health class.
For a more specific implementation manner of each unit of the rail transit health monitoring device, reference may be made to the description of the above rail transit health monitoring method, and similar advantageous effects will be provided, which will not be repeated herein.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when being executed by a processor, implements the rail transit health monitoring method as described above.
In general, the computer instructions for carrying out the methods of the present invention may be carried in any combination of one or more computer-readable storage media. The non-transitory computer-readable storage medium may include any computer-readable medium, except the signal itself in temporary propagation.
The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of an embodiment of the present invention, a computer-readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
Computer program code for carrying out operations of the present invention may be written in one or more programming languages, or combinations thereof, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" language or similar programming languages, and in particular, the Python language suitable for neural network computing and TensorFlow, pyTorch-based platform frameworks may be used. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).
The embodiment of the present invention further provides an electronic device, and as shown in fig. 8, fig. 8 is a schematic structural diagram of an embodiment of the electronic device provided by the present invention, where the electronic device 800 includes a processor 801, a memory 802, and a computer program stored in the memory 802 and capable of running on the processor 801, and when the processor 801 executes the program, the method for monitoring rail transit health as described above is implemented.
As a preferred embodiment, the electronic device 800 further comprises a display 803 for displaying the processor 801 to perform the rail transit health monitoring method as described above.
By way of example, a computer program may be partitioned into one or more modules/units that are stored in memory 802 and executed by processor 801 to perform the present invention. One or more of the modules/units may be a series of computer program instruction segments capable of performing particular functions to describe the execution of the computer program in the electronic device 800. For example, the computer program may be divided into the acquisition unit 701, the processing unit 702, and the prediction unit 703 in the above embodiments, and specific functions of the respective units are described above, which are not described herein in detail.
The electronic device 800 may be a desktop computer, notebook, palm top computer, or smart phone device with an adjustable camera module.
The processor 801 may be an integrated circuit chip with signal processing capabilities. The processor 801 may be a general-purpose processor including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; but also digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The disclosed methods, steps, and logic blocks in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The Memory 802 may be, but is not limited to, a random access Memory (Random Access Memory, RAM), a Read Only Memory (ROM), a programmable Read Only Memory (Programmable Read-Only Memory, PROM), an erasable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), an electrically erasable Read Only Memory (Electric Erasable Programmable Read-Only Memory, EEPROM), etc. The memory 802 is configured to store a program, and the processor 801 executes the program after receiving an execution instruction, where the method for defining a flow disclosed in any of the foregoing embodiments of the present invention may be applied to the processor 801 or implemented by the processor 801.
The display 803 may be an LCD display screen or an LED display screen. Such as a display screen on a cell phone.
It is to be appreciated that the configuration shown in fig. 8 is merely a schematic diagram of one configuration of the electronic device 800, and that the electronic device 800 may include more or fewer components than shown in fig. 8. The components shown in fig. 8 may be implemented in hardware, software, or a combination thereof.
The computer readable storage medium and the electronic device according to the embodiments of the present invention may be implemented with reference to the content of the specific description of implementing the method for monitoring the health of the rail transit according to the present invention, and have similar beneficial effects as the method for monitoring the health of the rail transit according to the present invention, which are not described herein.
The invention discloses a rail transit health monitoring system and method, firstly, response data are effectively acquired; furthermore, relevant effective data are extracted through preprocessing the response data, vibration data to be detected are formed, and accuracy of the vibration data to be detected is guaranteed; finally, the vibration data to be detected is effectively identified and judged through the first neural network and/or the second neural network, the characteristic information is extracted, the track structure health type and the vehicle body health type are predicted, the rapid and efficient judgment of various traffic conditions based on the vibration information is ensured, and the vibration data is judged.
According to the technical scheme, response data are acquired with high sensitivity based on the vibration sensing network, after the vibration data to be measured are input into the related neural network, various traffic running information can be obtained only based on the vibration data to be measured, the high efficiency and the rapidity of an algorithm are guaranteed, after the result is directly predicted, related personnel can refer to the predicted result, the health condition and the running speed of the subway can be known in time, accidents are reduced, and the timeliness of feedback is guaranteed.
The present invention is not limited to the above-mentioned embodiments, and any changes or substitutions that can be easily understood by those skilled in the art within the technical scope of the present invention are intended to be included in the scope of the present invention.
Claims (6)
1. The rail transit health monitoring method is characterized by being applied to an intelligent operation and maintenance service platform in a rail transit health monitoring system, wherein the rail transit health monitoring system comprises the following components: vibration sensing network, data processing center and intelligent fortune dimension service platform of communication connection in proper order, wherein:
the vibration sensing network comprises a fiber grating vibration sensing optical cable arranged on a track traffic engineering structure;
the data processing center is used for processing the monitoring signals of the fiber bragg grating vibration sensing optical cable and determining response data;
the intelligent operation and maintenance service platform is used for performing intelligent analysis according to the response data and monitoring the track structure and the vehicle based on an intelligent analysis result;
the method comprises the following steps:
acquiring response data of the vibration sensing network;
preprocessing according to the response data, and determining vibration data to be detected;
inputting the vibration data to be tested into a first neural network and a second neural network which are completely trained, and determining the predicted track structure health category and the predicted vehicle body health category;
the training process of the first neural network and the second neural network comprises:
Outputting a predicted track structure health class when a sample set is input to the first neural network, wherein the sample set includes sample vibration data and a corresponding actual track structure health class;
outputting a predicted vehicle body health class when a sample set is input to the second neural network, wherein the sample set includes the sample vibration data and a corresponding actual vehicle body health class;
determining a first loss function according to the error between the predicted track structure health category and the actual track structure health category, and training a constructed first neural network according to the first loss function until convergence;
determining a second loss function according to the error between the predicted vehicle body health category and the actual vehicle body health category, and training a constructed second neural network to be converged according to the second loss function;
the first neural network comprises, in order: input layer, convolution layer, max-pooling layer, flame layer, full connection layer, and output layer, wherein:
the input layer is used for inputting the sample set;
the convolution layer is used for carrying out convolution operation on the sample set and determining first convolution data;
The maximum pooling layer comprises a pooling layer and a Batch Normalization layer, wherein the pooling layer is used for carrying out feature extraction and dimension reduction on the first convolution data, and the Batch Normalization layer is used for carrying out standardization on the dimension reduced first convolution data to determine standardized data;
the flat layer is used for expanding the standardized data into a one-dimensional vector;
the full connection layer is used for reducing the dimension of the one-dimensional vector, classifying the vector through a softmax function and determining the health category of the predicted orbit structure;
the second neural network comprises, in order: BP neural network and one-dimensional neural network, wherein:
the BP neural network is used for training a speed training set and outputting a predicted vehicle speed, wherein the speed training set comprises sample vibration data and corresponding speed labels;
the one-dimensional neural network is used for training the sample set through the input layer, the convolution layer, the maximum pooling layer, the flat layer, the full connection layer and the output layer in sequence, and determining the predicted vehicle body health type.
2. The method for monitoring the health of the rail transit according to claim 1, wherein the fiber bragg grating arrays in the vibration sensing network are all made of fiber bragg grating array sensing probes, the fiber bragg grating array sensing probes comprise a preset reflectivity grating array, and fiber bragg grating sensing points with preset scales are multiplexed on a single fiber.
3. The method for monitoring the health of the rail transit according to claim 1, wherein the fiber gratings on each fiber grating array sensing probe are distributed at equal intervals, the fiber gratings with equal intervals are used as nodes, and the adjacent nodes form vibration sensing units.
4. The method for monitoring the health of rail transit according to claim 1, wherein the vibration sensing optical cable in the vibration sensing network is closely attached to the structure to be tested, wherein:
if the tested structure belongs to the existing line, directly adhering the structure adhesive to the tested structure, or placing the vibration sensing optical cable after slotting, and then curing the vibration sensing optical cable and the tested structure by using cement or curing agent;
if the detected structure belongs to the newly repaired line, directly pouring the vibration sensing optical cable into the concrete.
5. The method of claim 1, wherein the response data includes vibration data of the pickup in the ith section between a first moment when the train is traveling in and a second moment when the train is traveling out, the preprocessing is performed according to the response data, and determining the data to be detected includes:
adding environmental noise signals to the front end and the rear end of the vibration data to determine filling data;
And carrying out data normalization processing on the filling data to determine the data to be tested.
6. The rail transit health monitoring method of claim 1, wherein a learning rate is adaptively adjusted using an Adam optimizer during training of the first and second neural networks, and the first and second loss functions are multiple cross entropy functions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210259899.3A CN114684217B (en) | 2022-03-16 | 2022-03-16 | Rail transit health monitoring system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210259899.3A CN114684217B (en) | 2022-03-16 | 2022-03-16 | Rail transit health monitoring system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114684217A CN114684217A (en) | 2022-07-01 |
| CN114684217B true CN114684217B (en) | 2024-03-01 |
Family
ID=82138179
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210259899.3A Active CN114684217B (en) | 2022-03-16 | 2022-03-16 | Rail transit health monitoring system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114684217B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115420476B (en) * | 2022-11-03 | 2023-03-24 | 高勘(广州)技术有限公司 | Optical cable-based railway health monitoring method, device, equipment and storage medium |
| CN116223763B (en) * | 2022-12-05 | 2024-03-26 | 湖北国际物流机场有限公司 | Method for monitoring health state of runway surface of airport runway landing zone |
| CN115808468A (en) * | 2022-12-08 | 2023-03-17 | 中铁第四勘察设计院集团有限公司 | Ballastless track disease rapid identification method based on vibration waveform |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106503642A (en) * | 2016-10-18 | 2017-03-15 | 长园长通新材料股份有限公司 | A kind of model of vibration method for building up for being applied to optical fiber sensing system |
| CN108647786A (en) * | 2018-07-10 | 2018-10-12 | 电子科技大学 | The rotating machinery on-line fault monitoring method of neural network is fought based on depth convolution |
| CN109446187A (en) * | 2018-10-16 | 2019-03-08 | 浙江大学 | Complex equipment health status monitoring method based on attention mechanism and neural network |
| CN109815892A (en) * | 2019-01-22 | 2019-05-28 | 武汉理工大学 | The signal recognition method of distributed fiber grating sensing network based on CNN |
| CN109829577A (en) * | 2019-01-17 | 2019-05-31 | 北京交通大学 | Track train operating status prediction technique based on deep neural network structural model |
| CN109829538A (en) * | 2019-02-28 | 2019-05-31 | 苏州热工研究院有限公司 | A kind of equipment health Evaluation method and apparatus based on deep neural network |
| CN110175434A (en) * | 2019-06-19 | 2019-08-27 | 西南交通大学 | A kind of rail fastener system injury detection method based on convolutional neural networks |
| CN110502991A (en) * | 2019-07-18 | 2019-11-26 | 武汉理工大学 | Internal combustion engine health monitor method and system based on random convolutional neural networks structure |
| CN110751073A (en) * | 2019-10-12 | 2020-02-04 | 武汉理工大学 | Pipeline early damage mode identification method based on distributed optical fiber sensing and deep learning |
| CN111071300A (en) * | 2020-02-12 | 2020-04-28 | 太原理工大学 | High-speed train rail transit fault safety monitoring and early warning system and signal processing method |
| CN111242021A (en) * | 2020-01-10 | 2020-06-05 | 电子科技大学 | Distributed optical fiber vibration signal feature extraction and identification method |
| CN112580153A (en) * | 2020-12-29 | 2021-03-30 | 成都运达科技股份有限公司 | Health state management system and method for vehicle running gear monitoring component |
| CN113465924A (en) * | 2021-06-21 | 2021-10-01 | 武汉理工大学 | Bearing fault diagnosis method and system based on improved BP neural network |
| CN113469060A (en) * | 2021-07-02 | 2021-10-01 | 浙大城市学院 | Multi-sensor fusion convolution neural network aeroengine bearing fault diagnosis method |
| CN113639646A (en) * | 2021-05-24 | 2021-11-12 | 武汉理工大学 | Full-time global online monitoring device and method for deformation of large engineering structure |
| CN113761649A (en) * | 2021-08-05 | 2021-12-07 | 江苏大学 | Intelligent automobile tire eccentric wear prediction method based on one-dimensional convolutional neural network |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108932480B (en) * | 2018-06-08 | 2022-03-15 | 电子科技大学 | Distributed optical fiber sensing signal feature learning and classifying method based on 1D-CNN |
| US11783452B2 (en) * | 2020-04-07 | 2023-10-10 | Nec Corporation | Traffic monitoring using distributed fiber optic sensing |
-
2022
- 2022-03-16 CN CN202210259899.3A patent/CN114684217B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106503642A (en) * | 2016-10-18 | 2017-03-15 | 长园长通新材料股份有限公司 | A kind of model of vibration method for building up for being applied to optical fiber sensing system |
| CN108647786A (en) * | 2018-07-10 | 2018-10-12 | 电子科技大学 | The rotating machinery on-line fault monitoring method of neural network is fought based on depth convolution |
| CN109446187A (en) * | 2018-10-16 | 2019-03-08 | 浙江大学 | Complex equipment health status monitoring method based on attention mechanism and neural network |
| CN109829577A (en) * | 2019-01-17 | 2019-05-31 | 北京交通大学 | Track train operating status prediction technique based on deep neural network structural model |
| CN109815892A (en) * | 2019-01-22 | 2019-05-28 | 武汉理工大学 | The signal recognition method of distributed fiber grating sensing network based on CNN |
| CN109829538A (en) * | 2019-02-28 | 2019-05-31 | 苏州热工研究院有限公司 | A kind of equipment health Evaluation method and apparatus based on deep neural network |
| CN110175434A (en) * | 2019-06-19 | 2019-08-27 | 西南交通大学 | A kind of rail fastener system injury detection method based on convolutional neural networks |
| CN110502991A (en) * | 2019-07-18 | 2019-11-26 | 武汉理工大学 | Internal combustion engine health monitor method and system based on random convolutional neural networks structure |
| CN110751073A (en) * | 2019-10-12 | 2020-02-04 | 武汉理工大学 | Pipeline early damage mode identification method based on distributed optical fiber sensing and deep learning |
| CN111242021A (en) * | 2020-01-10 | 2020-06-05 | 电子科技大学 | Distributed optical fiber vibration signal feature extraction and identification method |
| CN111071300A (en) * | 2020-02-12 | 2020-04-28 | 太原理工大学 | High-speed train rail transit fault safety monitoring and early warning system and signal processing method |
| CN112580153A (en) * | 2020-12-29 | 2021-03-30 | 成都运达科技股份有限公司 | Health state management system and method for vehicle running gear monitoring component |
| CN113639646A (en) * | 2021-05-24 | 2021-11-12 | 武汉理工大学 | Full-time global online monitoring device and method for deformation of large engineering structure |
| CN113465924A (en) * | 2021-06-21 | 2021-10-01 | 武汉理工大学 | Bearing fault diagnosis method and system based on improved BP neural network |
| CN113469060A (en) * | 2021-07-02 | 2021-10-01 | 浙大城市学院 | Multi-sensor fusion convolution neural network aeroengine bearing fault diagnosis method |
| CN113761649A (en) * | 2021-08-05 | 2021-12-07 | 江苏大学 | Intelligent automobile tire eccentric wear prediction method based on one-dimensional convolutional neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114684217A (en) | 2022-07-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114684217B (en) | Rail transit health monitoring system and method | |
| Peng et al. | Assessing the impact of reduced visibility on traffic crash risk using microscopic data and surrogate safety measures | |
| Chen et al. | Damage detection of highway bridges based on long-gauge strain response under stochastic traffic flow | |
| CN114575927A (en) | Rail transit safety monitoring system and method | |
| CN110926523A (en) | High-speed railway bridge safety perception and early warning system under complicated abominable condition | |
| CN201819697U (en) | Bridge safety monitoring and warning system | |
| CN111060270B (en) | Movable rapid monitoring and intelligent evaluation method for urban viaduct | |
| CN109472885A (en) | Tire safety management method, device, tire safety management equipment and automobile | |
| Liu et al. | Automatic interpretation of strain distributions measured from distributed fiber optic sensors for crack monitoring | |
| Lechner et al. | A wavelet-based bridge weigh-in-motion system | |
| Wang et al. | Research on application of deep convolutional network in high-speed railway track inspection based on distributed fiber acoustic sensing | |
| Ambros et al. | A feasibility study for developing a transferable accident prediction model for Czech regions | |
| Liao et al. | The moving load identification method on asphalt roads based on the BP neural network and FBG sensor monitoring | |
| EP3951344A1 (en) | Methods and systems for damage evaluation of structural assets | |
| Liu et al. | Turning telecommunication fiber‐optic cables into distributed acoustic sensors for vibration‐based bridge health monitoring | |
| Rahman et al. | Remote condition monitoring of rail tracks using distributed acoustic sensing (DAS): a deep CNN-LSTM-SW based model | |
| Al-Nasar et al. | A systematic review of structural materials health monitoring system for girder-type bridges | |
| Ganji et al. | Asphalt pavement macrotexture monitoring in cracked surfaces by using an acoustical low-cost continuous method | |
| CN107784261B (en) | Road surface state detection method and device | |
| CN106646670A (en) | Distributed micrometeorological monitoring method for transmission lines | |
| Xiong et al. | Structural health evaluation of arch bridge by field test and optimized BPNN algorithm | |
| US20230024104A1 (en) | Identification of false transformer humming using machine learning | |
| CN208036293U (en) | Rail of subway real-time state monitoring device | |
| Dell’Acqua et al. | Validation procedure for predictive functions of driver behaviour on two-lane rural roads | |
| Lu | A statistical based damage detection approach for highway bridge structural health monitoring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |