CN109870683B - Rail abnormal fastener detection method based on radar signal time-frequency characteristic analysis - Google Patents
Rail abnormal fastener detection method based on radar signal time-frequency characteristic analysis Download PDFInfo
- Publication number
- CN109870683B CN109870683B CN201910168959.9A CN201910168959A CN109870683B CN 109870683 B CN109870683 B CN 109870683B CN 201910168959 A CN201910168959 A CN 201910168959A CN 109870683 B CN109870683 B CN 109870683B
- Authority
- CN
- China
- Prior art keywords
- time
- frequency
- fastener
- rail
- database
- 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
Images
Landscapes
- Train Traffic Observation, Control, And Security (AREA)
Abstract
The invention provides a method for detecting an abnormal rail fastener based on radar signal time-frequency characteristic analysis, which is used for rail maintenance. According to the method, the unmanned aerial vehicle carrying the 24GHz continuous wave radar acquires data, the unmanned aerial vehicle patrols and examines along the rail, transmits a 24GHz continuous wave signal when reaching the overhead position of the rail fastener, and acquires echo data before, during and after passing of the train. The method comprises the steps of firstly carrying out radar monitoring on a normal fastener, carrying out short-time Fourier transform on echo signals, storing obtained time-frequency spectrograms into a database in a classified mode, then comparing the time-frequency spectrograms of the fastener to be detected with the time-frequency spectrograms of corresponding classes in the database one by one, and judging whether the rail fastener is abnormal or not according to the difference of the time-frequency spectrograms. The method is simple to operate, high in detection efficiency and accuracy, and capable of efficiently and accurately judging whether the rail fasteners are abnormal or not.
Description
Technical Field
The invention relates to the technical field of railway maintenance and radar signal processing, in particular to an abnormal fastener detection method based on radar signal time-frequency characteristic analysis.
Background
In recent years, railways are rapidly developed, a high-speed train brings more convenience to people to go out, and the people's going-out efficiency is improved. The railway construction plays an extraordinary role in the economic construction of China, the distance between cities is shortened, the problem of insufficient transportation capacity is solved, the railway construction is an important power for promoting the sustainable and rapid development of economy, and therefore the development of railways has an important strategic position for the nation. While enjoying the benefits of railways, we should pay more attention to the safety of railway operation. In addition to the train itself, the safety problem of the railway line is not negligible. In the process of routing inspection, whether the rail is intact or not, whether foreign matters exist on the line or not and the like are all necessary detection items. The fastener is a middle part for connecting the rail and the sleeper, is used as a key part for fixing the rail, ensures that the rail is in a normal state, and has a vital significance in ensuring the safety of a line. At present, the mode of manual inspection is mainly adopted in China, the detection is carried out by naked eyes, the time and the labor are wasted, and the omission ratio is high. The automatic detection technology for the abnormal fastener is an optical image processing technology, which can detect with high efficiency, but is easily affected by resolution, light and environmental factors, so that the technology has certain limitation. How to carry out high-efficient detection to unusual fastener is a problem that needs to solve urgently.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a method for detecting an abnormal rail fastener based on radar signal time-frequency characteristic analysis, which aims to solve the problem of automatic detection of the abnormal rail fastener.
According to the method for detecting the rail abnormal fasteners based on the radar signal time-frequency characteristic analysis, the realized hardware comprises an unmanned aerial vehicle and a remote computer; the unmanned aerial vehicle is provided with a 24GHz continuous wave radar and a data acquisition module; the unmanned aerial vehicle patrols and examines along the rail, and transmits a 24GHz continuous wave signal when reaching the overhead position of the rail fastener, and the data acquisition module acquires radar echo data; the computer is provided with a database and a data processing module. The detection method comprises the following steps 1-3.
And 3, detecting abnormal fasteners. The data processing module compares the transmitted time-frequency spectrogram X with time-frequency spectrograms in subcategories of corresponding frequency ranges in corresponding situation categories in the database one by one, and calculates the difference coefficient between the time-frequency spectrogram X and the time-frequency spectrogram in the corresponding subcategories; and when the difference coefficient is larger than a preset threshold value, judging as an abnormal fastener, otherwise, judging as a normal fastener, and storing the time-frequency spectrogram of the fastener into a corresponding category of the database.
The difference coefficient is defined as:
compared with the prior art, the invention has the advantages and positive effects that: (1) according to the method, the normal fastener is subjected to radar detection according to the three conditions of the train passing through the front part, the middle part and the rear part, corresponding time-frequency spectrograms are obtained, and the detection of the fastener is more accurate and comprehensive due to data acquisition of the three conditions; (2) because the speed of the train is different and the frequency of the echo signal is different, the method divides the subclasses according to the frequency range of the weak signal, so that the database is established, the data acquisition mode is simple, and the normal data is easy to acquire; in addition, during detection, the detection is quicker and the data calculation amount is less only by comparing the frequency range of the weak signal of the fastener to be detected with the subclasses in the corresponding subclasses in the database; (3) the method has the advantages of simple operation, all-weather, high detection efficiency and accuracy and the like, and can efficiently and accurately judge whether the rail fastener is abnormal, such as loosening and the like.
Drawings
FIG. 1 is a schematic view of a rail normal clip;
FIG. 2 is a schematic view of a rail anomalous fastener;
FIG. 3 is a schematic flow chart of a method for detecting an abnormal rail fastening according to the present invention;
FIG. 4 is a time-frequency spectrum of a normal fastener signal before a train passes;
FIG. 5 is a time-frequency spectrum of a normal fastener signal as a train passes;
FIG. 6 is a time-frequency spectrum of a normal fastener signal after a train has passed;
FIG. 7 is a signal time-frequency spectrum of an abnormal fastener pulled by a person;
FIG. 8 is a time-frequency spectrum of an abnormal fastener signal when a train passes.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the present invention will be further described in detail with reference to the accompanying drawings and embodiments.
Rail clips are important parts of the track, also known as intermediate connecting parts, for connecting rails and sleepers. The rail track has the functions of effectively ensuring the reliable connection between the rail and the sleeper for a long time, preventing the longitudinal movement between the rail and the sleeper, ensuring the rail gauge to be normal, fully playing the buffering and damping performance under the action of the power of a train and slowing down the accumulation of the residual deformation of the track. A rail normal clip as shown in fig. 1, the rail normal clip is slightly vibrated by vibration of a rail when a train passes by. When a rail clip is in an abnormal state, such as when the clip is loose, as shown in fig. 2, in the process of train passing, in addition to vibration following the rail, the vibration itself generates vibration of a larger magnitude, and the vibration causes radial velocity relative to radar, which is the basis of the present invention for detecting abnormal clips based on radar signals.
The hardware realized by the invention comprises an unmanned aerial vehicle and a remote computer. 24GHz continuous wave radar and a data acquisition module are mounted on the unmanned aerial vehicle. And building a database and a data processing module on the remote computer. The flow of the method for detecting the rail abnormal fastener provided by the invention is shown in fig. 3, and the overall working mode is divided into a training mode and a detection mode. And transmitting the two-dimensional time-frequency spectrogram of the echo signal into different modules according to different working modes. When the fastener is in a training mode, the two-dimensional time-frequency spectrogram is transmitted into a database, and only the time-frequency spectrogram of a normal fastener is stored in the database; when the fastener is in the detection mode, the acquired time-frequency spectrogram of the fastener is transmitted into the data processing module. And establishing a normal fastener database in a training mode, and detecting the fastener in real time in a detection mode.
First, an implementation of data acquisition of the present invention is explained. Unmanned aerial vehicle follows the train and patrols and examines along the rail, and 24GHz continuous wave signal is launched to the fastener of radar below when unmanned aerial vehicle reachd the overhead position of rail fastener, gathers echo signal data. Unmanned aerial vehicle's position is in the oblique top of fastener, data acquisition under three kinds of circumstances: before a train passes, when a train passes, and after a train passes. Unmanned aerial vehicle generally stops 1 ~ 15 seconds of time in fastener oblique sky.
Secondly, the basis of the radar of the invention for detecting the vibrating target is the Doppler effect. Based on the doppler effect, when an object has a radial velocity relative to the radar, the reflected signal generates a doppler shift, and the frequency shift has the following relationship with the velocity:
where λ is the wavelength of the radar, v is the radial velocity of the target, and f is the doppler shift produced by the target. The abnormal fastener can generate radial velocity relative to the radar when vibrating, so that the detection of the abnormal fastener can be realized by measuring the Doppler frequency shift of echo signals. Since the abnormal fastener will vibrate irregularly, the radial velocity relative to the radar is time-varying, and the doppler shift is also time-varying.
In order to show the vibration state of the abnormal fastener, the invention carries out time-frequency analysis on the fastener echo signal. Common time-frequency analysis methods include short-time fourier transform, Wigner-Ville distribution, wavelet transform, and the like. The short-time Fourier transform has the advantages of simple calculation, no cross terms and the like, so the invention adopts the short-time Fourier transform to display the time-frequency spectrum characteristic of the radar echo signal. The formula for the short-time fourier transform can be expressed as follows:
where s (τ) represents the radar echo signal, w (τ -t) represents a window function, superscript denotes the conjugate, t denotes time, f denotes the frequency of the echo signal, τ denotes the time delay, and j is an imaginary unit.
The resolving power of the short-time Fourier transform is completely dependent on the time window radius Delta of the window functiontSum frequency window radius Δw. However, the values of the time window radius and the frequency window radius of the window function are not independent, and according to the inaccuracy measuring principle, the following relationship exists:
it can be seen that for short-time fourier transforms, the time and frequency resolution are always contradictory. In the time-frequency spectrum of the signal, high time resolution and frequency resolution cannot be obtained at the same time, one of which is narrowed and the other of which is necessarily widened. This means that only time resolution can be sacrificed in exchange for higher frequency resolution or conversely that a decrease in frequency resolution is exchanged for an increase in time resolution. When the window function is a gaussian function, the equal sign is established, and the time and frequency resolution are compromised. Therefore, the window function of the present invention employs a gaussian window function.
The following describes implementation steps of the rail abnormal fastening detection method of the present invention, and as shown in fig. 3, the method includes the following steps 1 to 3.
Unmanned aerial vehicle follows the train and patrols and examines along the rail, respectively before the train process, when passing through and under three kinds of circumstances after passing through, the radar carries out 24GHz continuous wave signal transmission to normal fastener, gathers the echo signal of normal fastener under these three kinds of circumstances. The data acquisition module carries out short-time Fourier transform with the echo signal of the normal fastener who gathers, obtains echo signal's two-dimensional time-frequency spectrogram to with time-frequency spectrogram incoming database storage. In the training stage, what data acquisition module gathered is the echo signal of normal fastener. The database divides the time-frequency spectrogram into three types for storage according to the three conditions of the train before, when and after passing.
Fig. 4, 5 and 6 are radar return signals of a normal fastener before, during and after a train pass, respectively. The train in the invention is just before and after passing, namely the train is just before and just after passing the fastener. The first half of fig. 4 is before the train passes and the second half is when the train is passing; the first half of fig. 6 is that the train is passing and the second half is that the train has left, so in the second half of fig. 4 and the first half of fig. 6 is a reflection component that exists between the train body and the gap between the two cars. It has been found that the rails, stones and fasteners can still vibrate before and for a short period of time after the train has passed. When a train passes by, the instantaneous frequency of the echo signal can be divided into two parts, one part is a strong echo signal and the other part is a weak echo signal. Strong echo signals are produced by reflections from the train body. When a train passes by, the speed component of the train body relative to the radar and the train body shake generated in the process of train moving can generate Doppler frequency shift. Because of the gap between the two cars, the reflection component of the train body exhibits a periodic regularity. Weak echo signals are generated by rails, stones, fasteners, etc. and the frequency of vibration of these reflectors is significantly greater than that of the vibrations generated by the train body. But the frequency intensity is less than the signal reflected by the train car body. At this point, the clip is firmly secured to the rail and therefore vibrates in the same manner as the rail.
The frequency ranges of both the strong echo signal and the weak echo signal vary with the speed of the train. The method of the invention therefore also sets a plurality of subcategories in three cases, depending on the different frequency ranges of the signal. Each subcategory represents a frequency range, and embodiments of the present invention subdivide the plurality of subcategories according to the frequency range of the weak echo signal. For example, for echo data under the condition that a train passes by, the frequency range of weak echo signals in the acquired time-frequency spectrum is 0-1000Hz, the spectrum is stored into subcategories of 0-1000Hz under the condition that the train passes by, and if the subcategories of the frequency range do not exist, the subcategories of 0-1000Hz are established. As shown in fig. 5, if the frequency range of the collected weak echo signal is 500-. The frequency ranges of the different subcategories may partially overlap.
In the real-time detection process, the data of the fastener to be detected are compared with the data of corresponding subclasses in three classes in the database one by one according to the frequency range of the weak echo signal, so that the detection is carried out.
And 3, comparing the transmitted time-frequency spectrogram with the time-frequency spectrograms in the sub-categories under the corresponding situation category in the database one by the data processing module, calculating the difference coefficient of the time-frequency spectrograms in the X and the corresponding sub-categories, judging as an abnormal fastener when the difference coefficient is larger than a preset threshold, and otherwise, judging as a normal fastener. When the fastener is a normal fastener, the time-frequency spectrogram of the fastener is stored in the subcategory of the corresponding situation, and the database data is updated.
When the clip becomes loose, the clip itself will vibrate to a greater extent than it will vibrate following the rail. FIG. 7 is a time-frequency spectrum of an echo signal of an abnormal fastener pulled by a person. The abnormal fastener can generate irregular vibration along with the pulling of a person, and the highest frequency of the vibration of the fastener can reach 100Hz due to the low pulling frequency of the person. When the train is pressed, the vibration frequency of the abnormal fastener is higher, and the frequency intensity generated by the fastener is lower than that generated by the train body.
FIG. 8 is a time-frequency spectrum of an abnormal fastener as a train passes. For safety reasons, all fasteners which are acquired by the invention and passed by the train are normal fasteners, so that the echo signals of the abnormal fasteners are reasonable artificially synthesized signals. Because the vibration of the abnormal fastener is generated by pulling by people, the vibration frequency of the fastener is lower, the highest frequency is 100Hz, and the frequency is coincident with the frequency of the train body. When a train passes through the abnormal fastener, the fastener can continuously vibrate, and the vibration can also occur when the joint of the two carriages passes through. Therefore, comparing fig. 5 and fig. 8, in the gap between two cars, the intensity of the frequency component generated by the train is small, and the intensity of the frequency component generated by the abnormal fastener is strong, so that the two time-frequency spectrograms are obviously different, and the fastener can be determined to be the abnormal fastener. When the train passes through the abnormal fastener, the vibration frequency of the fastener is higher, so the time-frequency spectrogram difference of the normal fastener and the abnormal fastener is more obvious, and the detection accuracy is higher.
When the fastener echo signal is in the detection mode, the acquired fastener echo signal time-frequency spectrogram is transmitted into the data processing module. And comparing the newly acquired signal time-frequency spectrogram with the time-frequency spectrograms of the corresponding classes in the database one by one, and calculating the difference coefficient between the time-frequency spectrogram and the time-frequency spectrogram in the database. The difference coefficient is defined as the ratio of the number of the time-frequency spectrograms which have obvious difference with the time-frequency spectrogram of the fastener to be detected in the corresponding category of the database to the total number of the time-frequency spectrograms of the category in the database, namely:
the time-frequency spectrogram of the normal fastener has smaller difference with the time-frequency spectrogram of the corresponding category in the database, so the difference coefficient is smaller; the anomalous fasteners will also vibrate to a greater extent, and therefore a greater coefficient of variation, in addition to following the rail vibrations. If the difference coefficient is larger than the threshold value, the fastener is judged to be an abnormal fastener; if the difference coefficient is smaller than the threshold value, the fastener is judged to be a normal fastener, the time-frequency spectrogram is transmitted into a database, and the database data is updated. The threshold value can be set to any value from 0 to 1, and the threshold value is selected to be 0.5 in the embodiment of the invention, namely when the time-frequency spectrogram of the fastener to be tested is obviously different from more than half of the time-frequency spectrogram of the corresponding category in the database, the fastener is judged to be an abnormal fastener. The lower the threshold setting, the higher the sensitivity of the detection.
When the two-dimensional time-frequency spectrograms are compared, the two-dimensional time-frequency spectrogram of the fastener to be detected and the pixel value of the time-frequency spectrogram in the database are subtracted to obtain the difference value of the two-dimensional time-frequency spectrograms, and if the difference value is larger than the total pixel value which is alpha times, the difference between the two-dimensional time-frequency spectrogram of the fastener to be detected and the time-frequency spectrogram of the corresponding category of the database is considered to be obvious. Wherein, alpha is a threshold value, and the value interval is (0, 1).
It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Claims (6)
1. A rail abnormal fastener detection method based on radar signal time-frequency characteristic analysis is characterized in that realized hardware comprises an unmanned aerial vehicle and a remote computer; the unmanned aerial vehicle is provided with a 24GHz continuous wave radar and a data acquisition module; the unmanned aerial vehicle patrols and examines along the rail, when arriving at the overhead position of the rail fastener, transmits 24GHz continuous wave signals to the fastener, and the data acquisition module acquires echo data; the computer is provided with a database and a data processing module; the detection method detects the abnormal fastener by measuring the Doppler frequency shift of the echo signal, and comprises the following steps:
step 1, establishing a normal fastener database, comprising: the radar transmits 24GHz continuous wave signals to the normal rail fasteners and collects echo signals of the normal fasteners under three conditions of before, during and after passing of a train; carrying out short-time Fourier transform on the collected radar echo signals to obtain a two-dimensional time-frequency spectrogram of the signals, and storing the time-frequency spectrogram into a database; the database divides the time-frequency spectrogram into three types for storage according to three conditions of the train before, when and after passing, and sub-types are set according to the frequency range of the signal under each type, and each sub-type represents a frequency range; storing the time-frequency spectrogram into a subcategory corresponding to the frequency range under the corresponding situation category;
step 2, starting a detection mode, transmitting a 24GHz continuous wave signal to the fastener by a radar, carrying out short-time Fourier transform on an acquired echo signal to obtain a two-dimensional time-frequency spectrogram X of the signal, and transmitting the time-frequency spectrogram X and a corresponding acquisition condition into a data processing module; the acquisition condition refers to the acquisition before, during or after the train passes;
step 3, the data processing module compares the transmitted time-frequency spectrogram X with the time-frequency spectrograms in the sub-categories of the corresponding frequency ranges in the corresponding situation categories in the database one by one, and calculates the difference coefficient between the time-frequency spectrogram X and the time-frequency spectrogram in the corresponding sub-categories; when the difference coefficient is larger than a preset threshold value, judging as an abnormal fastener, otherwise, judging as a normal fastener, and storing a time-frequency spectrogram of the fastener into a corresponding category of a database;
the difference coefficient is defined as:
2. the method according to claim 1, wherein the unmanned aerial vehicle is located obliquely above the fastener and stays for 1-15 seconds.
3. The method according to claim 1, wherein in step 1, the window function used for the short-time fourier transform is a gaussian window function.
4. The method according to claim 1, wherein in step 1, the subcategories are set according to the frequency range of the weak echo signal.
5. The method according to claim 1, wherein in the step 3, when comparing the two time-frequency spectrograms, the difference between the time-frequency spectrogram of the fastener to be tested and the pixel value of the time-frequency spectrogram in the database is obtained by subtracting the pixel values of the two time-frequency spectrograms, and if the difference is greater than a times of the total pixel value of the time-frequency spectrogram in the database, the difference between the time-frequency spectrogram of the fastener to be tested and the time-frequency spectrogram of the database is considered to be obvious, wherein α is a set threshold, and the value range is (0, 1).
6. The method according to any one of claims 1 to 4, wherein the threshold value preset in step 3 is 0.5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910168959.9A CN109870683B (en) | 2019-03-06 | 2019-03-06 | Rail abnormal fastener detection method based on radar signal time-frequency characteristic analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910168959.9A CN109870683B (en) | 2019-03-06 | 2019-03-06 | Rail abnormal fastener detection method based on radar signal time-frequency characteristic analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109870683A CN109870683A (en) | 2019-06-11 |
| CN109870683B true CN109870683B (en) | 2020-09-04 |
Family
ID=66919876
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910168959.9A Active CN109870683B (en) | 2019-03-06 | 2019-03-06 | Rail abnormal fastener detection method based on radar signal time-frequency characteristic analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109870683B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115436935A (en) * | 2022-08-21 | 2022-12-06 | 唯识感知科技(成都)有限公司 | Vehicle-mounted Doppler radar detection system, array and method |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000230208A (en) * | 1999-02-12 | 2000-08-22 | Kobe Steel Ltd | Slackness inspecting device for rail fastening device |
| JP2007147412A (en) * | 2005-11-27 | 2007-06-14 | Teruya:Kk | Real-time inspection system of loose rail fastening using dc battery-less rfid tag with sensor input functions |
| CN105501248A (en) * | 2016-02-16 | 2016-04-20 | 株洲时代电子技术有限公司 | Railway line inspection system |
| CN106815552A (en) * | 2016-12-09 | 2017-06-09 | 云南航天工程物探检测股份有限公司 | Data signal post-processing approach based on time frequency analysis |
| CN107403139A (en) * | 2017-07-01 | 2017-11-28 | 南京理工大学 | A kind of municipal rail train wheel flat fault detection method |
| CN107748862A (en) * | 2017-09-21 | 2018-03-02 | 清华大学 | A kind of unmanned plane sorting technique and device based on dual-frequency radar signal time-frequency distributions |
| CN108058721A (en) * | 2018-01-26 | 2018-05-22 | 北京市劳动保护科学研究所 | A kind of rail fastener loose condition detection method and system |
| CN207946369U (en) * | 2017-10-13 | 2018-10-09 | 成都精工华耀科技有限公司 | A kind of railway rail clip abnormal detector based on pointolite array linear array imaging |
| CN109409225A (en) * | 2018-09-21 | 2019-03-01 | 清华大学 | Unmanned plane classification method and device based on the fusion of radar multipath signal time-frequency characteristics |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9463817B2 (en) * | 2015-02-16 | 2016-10-11 | Electro-Motive Diesel, Inc. | Automatic disabling of unpowered locked wheel fault detection for slipped traction motor pinion |
-
2019
- 2019-03-06 CN CN201910168959.9A patent/CN109870683B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000230208A (en) * | 1999-02-12 | 2000-08-22 | Kobe Steel Ltd | Slackness inspecting device for rail fastening device |
| JP2007147412A (en) * | 2005-11-27 | 2007-06-14 | Teruya:Kk | Real-time inspection system of loose rail fastening using dc battery-less rfid tag with sensor input functions |
| CN105501248A (en) * | 2016-02-16 | 2016-04-20 | 株洲时代电子技术有限公司 | Railway line inspection system |
| CN106815552A (en) * | 2016-12-09 | 2017-06-09 | 云南航天工程物探检测股份有限公司 | Data signal post-processing approach based on time frequency analysis |
| CN107403139A (en) * | 2017-07-01 | 2017-11-28 | 南京理工大学 | A kind of municipal rail train wheel flat fault detection method |
| CN107748862A (en) * | 2017-09-21 | 2018-03-02 | 清华大学 | A kind of unmanned plane sorting technique and device based on dual-frequency radar signal time-frequency distributions |
| CN207946369U (en) * | 2017-10-13 | 2018-10-09 | 成都精工华耀科技有限公司 | A kind of railway rail clip abnormal detector based on pointolite array linear array imaging |
| CN108058721A (en) * | 2018-01-26 | 2018-05-22 | 北京市劳动保护科学研究所 | A kind of rail fastener loose condition detection method and system |
| CN109409225A (en) * | 2018-09-21 | 2019-03-01 | 清华大学 | Unmanned plane classification method and device based on the fusion of radar multipath signal time-frequency characteristics |
Non-Patent Citations (2)
| Title |
|---|
| "Broken railway fastener detection based on Adaboost algorithm";Yiqi Xia等;《2010 international conference on optoelectronics and image processing》;20101231;第313-316页 * |
| "基于SIFT特征的铁路扣件状态检测算法";赵珊珊等;《传感器与微系统》;20181231;第37卷(第11期);第148-154页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109870683A (en) | 2019-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1812907B (en) | Rail and train monitoring system and method | |
| CN101900708B (en) | Vibration and audio signal-based high-speed train track defect detecting method | |
| US5402346A (en) | System for the calculation of at least one vehicular traffic check parameter | |
| US9121934B2 (en) | Radar system and detection method | |
| CN103076393B (en) | Knocking scan type bridge damage detecting system | |
| CN104881995B (en) | A kind of trackside dualbeam microwave radar traffic flow detecting device and method | |
| DE112009005228B4 (en) | Ultrasonic detection device | |
| KR970702499A (en) | Traffic monitoring method for the automatic detection of vehicle-related incidents | |
| CN101876705A (en) | Frequency domain vehicle detecting method based on single-frequency continuous wave radar | |
| JP6624601B2 (en) | Radar apparatus and target object detection method | |
| CN107917957B (en) | Damage detection method for plate-shaped structure | |
| CN105738898A (en) | Multilane radar speed measurement method and device based on combined distance, angle and speed measurement | |
| CN103680138A (en) | A method for monitoring vehicle road conditions | |
| CN109870683B (en) | Rail abnormal fastener detection method based on radar signal time-frequency characteristic analysis | |
| US12106664B2 (en) | Dynamic road traffic noise mapping using distributed fiber optic sensing (DFOS) over telecom network | |
| CN105423924A (en) | Noise barrier on-line measurement method and system | |
| CN100485732C (en) | Date collecting method for measuring automobile external noise | |
| CN114964330A (en) | Fault monitoring system and method based on optical fiber sensing and multi-parameter fusion | |
| CN104471438B (en) | For producing and assessing ultrasonic signal, in particular for determination vehicle and the apparatus and method of the distance of barrier | |
| IWATA et al. | Improved algorithms of seismic parameters estimation and noise discrimination in earthquake early warning | |
| CN202814931U (en) | Self-adaption ultrasonic rail fault detection device based on frequency spectrum cognition | |
| CN110514861B (en) | High-speed rail running speed estimation method utilizing amplitude spectrum autocorrelation | |
| WO2004102056A3 (en) | Systems and methods for non-destructively testing conductive members employing electromagnetic back scattering | |
| KR100378956B1 (en) | Acoustical Vehicle Detection System | |
| CN103150905A (en) | Method for installing wave frequency detector at roadside to detect traffic flow |
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 |