CN106649745A - Locomotive application comprehensive analysis method - Google Patents
Locomotive application comprehensive analysis method Download PDFInfo
- Publication number
- CN106649745A CN106649745A CN201611218458.XA CN201611218458A CN106649745A CN 106649745 A CN106649745 A CN 106649745A CN 201611218458 A CN201611218458 A CN 201611218458A CN 106649745 A CN106649745 A CN 106649745A
- Authority
- CN
- China
- Prior art keywords
- data
- lkj
- time
- file
- recording
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/283—Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP
Abstract
A locomotive application comprehensive analysis method comprises the steps of obtaining data from a data source, wherein the data of the data source comprises LKJ operating data, 6A data, recording pen files, line end data, TCMS data, Locotrol data and detection equipment data; building databases corresponding to time of the data for the obtained data of the data source, then, adopting a timer shaft of LKJ as a reference timer shaft, and performing one-to-one corresponding hooking on the LKJ timer shaft and all the data in the databases according to time after the time in the databases corresponding to the data of the data source corresponds to the LKJ timer shaft; obtaining occurrence time of safety item points in LKJ operation, and calling data from any one or more databases while a certain database is called for checking a data object during certain item point occurrence on a display interface of an industrial personal computer. LKJ time is adopted as the shaft, key data of the data source is displayed during item point occurrence, and linkage analysis can be performed conveniently.
Description
Technical field
The present invention relates to locomotive field, and in particular to a kind of locomotive operation comprehensive analysis method.
Background technology
The important component part that engineering application overall analysis system is managed as locomotive operation, is to ensure locomotive operation safety
And engineering reorganizes and outfit the locomotive quality of maintenance, in prior art, when being analyzed to locomotive information, it is impossible to which all locomotive informations are entered
The comprehensive flexible linkage analysis of row so that in existing analysis, exist analysis information island, decision problem not comprehensively, solve to ask
Inscribe a halfway difficult problem, it is impossible to provide comprehensive information support for engineering safety analysis, it is impossible to provide effectively guarantor for motorcycle safety
Barrier.
The content of the invention
The technical problem to be solved in the present invention is to provide one kind service on buses or trains process data chain of locomotive is all integrated, kept away
Exempt from the information island of existing system, provide, to solve the problems, such as prior art.
To solve above-mentioned technical problem, the present invention is employed the following technical solutions：
A kind of locomotive operation comprehensive analysis method, including：
Data are obtained from data source：The data source data includes LKJ service datas, 6A data, recording pen file, row mantissa
According to, TCMS data, Locotrol data, testing equipment data；
Data processing is carried out to the data that data source is obtained：
LKJ service datas, 6A data, recording pen file, row mantissa evidence, TCMS data, Locotrol data, detection to obtaining
Device data, sets up respectively each data data base corresponding with the time, then using the time shafts of LKJ as fiducial time axle, will
LKJ service datas, 6A data, recording pen file, row mantissa evidence, TCMS data, Locotrol data, testing equipment data correspondence
Each data base in time and LKJ time shafts carry out it is corresponding after, all data in each data base and LKJ time shafts are entered
The upper one-to-one mounting of row time；
Data after to process are analyzed：
According to LKJ log datas, the operating each safe corner time of origins of LKJ are obtained so that show boundary in industrial computer
When on face by data object when calling certain number storehouse and checking on LKJ time shafts that certain corner occurs, while by calling it
It any one or multiple data bases data so as to it any one or multiple data bases data industrial computer display circle
Synchronously shown on face, or broadcast in audio-frequency module.
When data after to process are analyzed, while driver's gesture recognition module is set in industrial computer, to 6A numbers
According to video in the operating gesture of driver be identified, recognition result and other data are carried out into linkage display, judge driver's handss
Whether gesture is correct.
The LKJ service datas are dumped in expansion module from IC-card, and the 6A data are dumped to from 6A data storage cards
In expansion module, the recording pen file is dumped in expansion module from recording pen, and the encrypting module in expansion box runs to LKJ
After data, 6A data, recording pen file are encrypted, being transferred to industrial computer carries out Video processing and linkage analysis.
Data in the recording pen document data bank are the data after being calibrated with LKJ fiducial time axles, described right
Data in recording pen file by being arranged on industrial computer in automatic correcting time module carry out the method for automatic correcting time and be：
To LKJ alert datas, read aloud by standard pronunciation, the standard pronunciation read aloud is trained into generation alarm model data
Storehouse；
Recording file to the recording pen in the locomotive of acquisition, carries out respectively single pass audio-frequency noise and removes and twin-channel sound
Frequency noise remove；
Recording file to removing noise, extracts all sound bites therein and sound bite corresponding time；
By the interval time between adjacent sound bite is in the range of setting interval threshold and sound bite duration is in setting duration
Sound bite in threshold range is extracted, used as efficient voice fragment；
Efficient voice fragment is matched with the data in model database, the warning message in efficient voice fragment is obtained,
The warning classification of warning message is obtained simultaneously；
The warning message that will be obtained, is contrasted with the time belonging to the warning message of LKJ, obtains the delay of recording file,
According to the delay, recording file is calibrated.
The recording file of described pair of removal noise, when extracting all sound bites therein, is extracted by way of MFCC
Signal frequency feature therein.
The training pattern sets up process：The alarm voice signal read aloud standard pronunciation, is extracted using MFCC methods and is believed
Number frequecy characteristic, according to the frequecy characteristic for extracting, training pattern is set up by Markov algorithm.
It is described to be to the method that recording file is calibrated：
Arrange with the LKJ standard time as X-axis coordinate and the plane coordinate system of Y-axis coordinate, obtain warning message in select to
Few two warning messages, and at least one time point is obtained in each warning message, and obtain the time point in recording pen
Time and the time in the LKJ standard time, the time-sloped of multiple time points is calculated in above-mentioned plane coordinate system, according to
When the slope carries out simple computation school to the recording file of the recording pen record.
It is described set interval threshold scope as LKJ to identical warning message report twice when midfeather shortest time
Scope and maximum duration between.
It is described to set the duration of most long warning message and most short warning message in the warning message that duration threshold value is reported as LKJ
Duration between scope.
The mounting is referred to sets up contact by each data at the same time on a timeline, when in LKJ files row
When certain data is selected in table, can be either multiple or whole with certain by the data of LKJ listed files by arranging optionies
Data in data base are linked, according to the display packing of the display interface of the prior industrial computer for arranging, by LKJ service datas
Table data and other data carry out piecemeal and show.
Beneficial effects of the present invention：With the LKJ times as axle, when corner occurs integrated display and inquiry 6A video informations,
The critical datas such as recording pen information, and supporting driver's gesture identification function, to driver safety traveling provide sound, regarding, figure etc. majority
According to the comprehensive analysis of various dimensions, it is intended to realize LKJ, 6A video, recording pen file, TCMS, Locotrol, 6A system, ground inspection
The linkage analysis of the information such as measurement equipment, solve the information island in terms of analytical data, realize the multidimensional with intelligent data as guiding
Degree comprehensive analysis.
Description of the drawings
Fig. 1 is the system construction drawing of the present invention.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and detailed description.
A kind of locomotive operation comprehensive analysis evaluation method that the present invention is provided, comprises the following steps：
Data are obtained firstly the need of from data source, the data of the data source are the storage device during service on buses or trains of locomotive on car
The LKJ service datas of record, 6A data, recording pen file, row mantissa evidence, TCMS data, Locotrol data, testing equipment number
According to etc. data.Above-mentioned data are to be obtained according to prior art.
After obtaining data source data, the data obtained to data source are needed to carry out data processing, first the LKJ fortune to obtaining
Row data, 6A data, recording pen file, row mantissa evidence, TCMS data, Locotrol data, testing equipment data, set up respectively
Data base, the data base is and data data base corresponding with the time, then using the time shafts of LKJ as fiducial time axle, will
LKJ service datas, 6A data, recording pen file, row mantissa evidence, TCMS data, Locotrol data, testing equipment data correspondence
Each data base in time and LKJ time shafts carry out it is corresponding after, all data in each data base and LKJ time shafts are entered
The upper one-to-one mounting of row time；The meaning of mounting is：Each data at the same time on a timeline are set up into connection
System, when certain data is selected in LKJ listed files, can by arrange optionies, by the data of LKJ listed files and certain
Data in either multiple or all database are linked, according to the display side of the display interface of the prior industrial computer for arranging
Method, carries out the table data of LKJ service datas and other data piecemeal and shows.I.e. in the background will be all in each data base
Data be associated with the time on LKJ time shafts it is corresponding, when certain corner is clicked in operation pages, on interface show with
Data or the data in partial database on LKJ time shafts in all data bases of the time synchronized that the corner occurs.
I.e. in said process, first industrial computer display screen display LKJ log data table data, list
Each corner is shown in data, the event that certain corner occurs is shown if desired, then clicked in the LKJ listed files of display screen
The corner is selected, due to the data mounting that same time point in the corner time of origin and other data bases occurs so that in work
During data object when calling several storehouses to check on LKJ time shafts that the corner occurs on control machine display interface, can simultaneously by calling
The data of other one or more data bases so as to which its data synchronously shown on the display interface of industrial computer, if
Video and audio file, then can be broadcast by audio-frequency module.
It is video format except 6A data displays when above-mentioned data are shown, recording file is shown as audio format,
Other data can be arranged according to industrial computer inside and be shown as multi-form, such as curve form or tabular form.
After being shown on the display screen of industrial computer, can be by the analysis to these linkage data, automatically to data
Exception is identified and is formed files passe, it is also possible to is analyzed by artificial observation, is manually entered analysis result file.
As shown in the system of Fig. 1, industrial computer can be analyzed to above-mentioned to each database data.LKJ service datas are entered
When row shows analysis, can be analyzed by the data to whole record, the linkage point of search terms point list can be also carried out to corner
Analysis, while the form of recording curve can also be shown as.And when being analyzed to TCMS data, failure can be carried out to TCMS data
Data diagnosis are analyzed, and can also carry out statistical analysiss.During the document data analysis of 6A data, 6A data can be carried out accident analysis and
Generate 6A tracing analysiss.And 6A video files are used to be shown.
When 6A video datas after to process are analyzed, while driver's gesture identification mould can also be arranged in industrial computer
Block, is identified to the operating gesture of driver in the video of 6A data, and recognition result and other data are carried out into linkage display, sentences
Whether the operating gesture of driver is correct under disconnected current state.
In above-mentioned data source, LKJ service datas are dumped in expansion module from IC-card, and 6A data are from 6A data storage cards
In being dumped to expansion module, the recording pen file is dumped in expansion module from recording pen, the encrypting module pair in expansion box
After LKJ service datas, 6A data, recording pen file are encrypted, being transferred to industrial computer carries out Video processing and linkage analysis.
Data in recording pen document data bank are the data after being calibrated with LKJ fiducial time axles, and described pair is recorded
Data in file by being arranged on industrial computer in automatic correcting time module carry out automatic correcting time method it is as described below.
It is identified by the warning message of the LKJ recorded to recording file and time comparison, school is carried out to recording file
When, to realize the linkage analysis of audio frequency and video.
Because recording file can record the information such as vehicle whistle, driver's dialogue, automobile mechanical vibration, and in LKJ generally not
These information are recorded, and these information noises are larger, information identification has complexity, it is impossible to carried according to these information times of carrying out
Take, when then carrying out the school of recording file according to the standard time axle of LKJ.
And LKJ can carry out voice broadcast to various warning messages, each recording file can be carried out to these warning messages
Record, meanwhile, the report of the warning message recorded in recording file is unique, and school is carried out to recording according to these voices
When, accuracy is high, and complexity is low.
The method of the present invention carries out database training firstly the need of the warning message reported to LKJ.I.e. in noiselessness condition
Under, the warning message related to LKJ is repeatedly read aloud by machine standard pronunciation, by mel cepstrum frequency（MFCC）Carry
The feature in these standard pronunciations is taken, then by hidden markov model（HMM）The standard pronunciation read aloud is trained to into warning mould
Type data base.
Because recording file has single-channel and two-channel different recording files, therefore, for different recording texts
Part carries out noise removal and speech enhan-cement using the different drying methods that goes.Wherein, single-channel voice document is by optimum improvement
Logarithm spectral amplitude estimation（OMLSA）Existing algorithm carry out dry and speech enhan-cement, double-channel file based on human ear using being sheltered
The spectrum of effect cuts algorithm carries out dry and speech enhan-cement.
To removing noise and carrying out the recording file of speech enhan-cement, all sound bites therein and the sound bite are extracted
Corresponding former and later two times.When extracting all sound bites therein, signal frequency therein is extracted by way of MFCC
Feature, according to the frequecy characteristic of signal, acquisition is considered as the fragment of voice recording.
Because LKJ warning messages are when reporting, the report of warning message has unique rule, such as each warning message
Report twice, the midfeather very short time（Such as 1 second）, and the length of warning message is generally shorter, and recording pen is recorded
Sound except warning message, also driver's sound of speaking or other sound, these sound irregularities, or cannot enter
Row correct time determines, therefore the LKJ warning messages in recording file are identified than being identified more to other information
Plus easily, be calibrated according to these warning messages, during school faster, it is more accurate.
Concrete operation method is：Interval time when All Alerts information reports twice is calculated first, according to interval time
The maximum and minimum value of threshold value is manually set, threshold range is formed, the threshold range is the scope for setting interval threshold；And it is same
, timing, record most long report time and most short report time, root are carried out to the massage voice reading time of each warning message
The scope of duration threshold value is set according to these times.Due to the presence of external interference factor, above range can be adjusted as needed.
For the sound bite for extracting, by the interval time between adjacent sound bite in setting interval threshold model
Enclose sound bite of the interior and sound bite duration in setting duration threshold range to extract, as efficient voice fragment.
To efficient voice snippet extraction feature, the spy of each audio file in the feature and training pattern extracted
Levy, matched according to existing phonetic feature matching process, obtain whether efficient voice fragment is warning according to matching result
The result of information, is not warning message, then give up the efficient voice fragment, if warning message, is then obtained according to training pattern
Take the warning classification of the efficient voice fragment.
The warning message that will be obtained, is contrasted with the time belonging to the warning message of LKJ, obtains prolonging for recording file
Late, according to the delay, recording file is calibrated.
Even if same manufacturer production, also due to the problem of crystal oscillator causes the time delay of each recording pen different, therefore,
Recording file school is carried out constantly, for the file of different recording pens record, it should be calibrated respectively.
When the recording file to some recording pen record is calibrated, first from the warning extracted in recording file
Several warning messages are selected in information, time of some of each warning message point in recording pen is obtained and in LKJ marks
Time between punctual, it is in the coordinate system of standard time in X-axis and Y-axis, two times of calculating correspond respectively to two-dimentional seat
X-axis in mark system and time-sloped during Y-axis, simple computation school is carried out according to the slope to the recording file of recording pen record
When.
Above-described is only the preferred embodiment of the present invention, it is noted that for a person skilled in the art,
Under the premise of without departing from general idea of the present invention, some changes and improvements can also be made, these should also be considered as the present invention's
Protection domain.
Claims (10)
1. a kind of locomotive operation comprehensive analysis method, it is characterised in that include：
Data are obtained from data source：The data source data includes LKJ service datas, 6A data, recording pen file, row mantissa
According to, TCMS data, Locotrol data, testing equipment data；
Data processing is carried out to the data that data source is obtained：
LKJ service datas, 6A data, recording pen file, row mantissa evidence, TCMS data, Locotrol data, detection to obtaining
Device data, sets up respectively each data data base corresponding with the time, then using the time shafts of LKJ as fiducial time axle, will
LKJ service datas, 6A data, recording pen file, row mantissa evidence, TCMS data, Locotrol data, testing equipment data correspondence
Each data base in time and LKJ time shafts carry out it is corresponding after, all data in each data base and LKJ time shafts are entered
The upper one-to-one mounting of row time；
Data after to process are analyzed：
According to LKJ log datas, the operating each safe corner time of origins of LKJ are obtained so that show boundary in industrial computer
When on face by data object when calling certain number storehouse and checking on LKJ time shafts that certain corner occurs, while by calling it
It any one or multiple data bases data so as to it any one or multiple data bases data industrial computer display circle
Synchronously shown on face, or broadcast in audio-frequency module.
2. a kind of locomotive operation comprehensive analysis method according to claim 1, it is characterised in that：Data after to process
When being analyzed, while arranging driver's gesture recognition module, the operating gesture to driver in the video of 6A data in industrial computer
It is identified, recognition result and other data is carried out into linkage display, judges whether driver's gesture is correct.
3. a kind of locomotive operation comprehensive analysis method according to claim 1, it is characterised in that：The LKJ service datas
It is dumped in expansion module from IC-card, the 6A data are dumped in expansion module from 6A data storage cards, the recording pen text
Part is dumped in expansion module from recording pen, and the encrypting module in expansion box is to LKJ service datas, 6A data, recording pen file
After being encrypted, being transferred to industrial computer carries out Video processing and linkage analysis.
4. a kind of locomotive operation comprehensive analysis method according to claim 1, it is characterised in that：The recording pen number of files
It is the data after being calibrated with LKJ fiducial time axles according to the data in storehouse, the data in the file to recording pen are by setting
Put the automatic correcting time module in industrial computer and carry out the method for automatic correcting time and be：
To LKJ alert datas, read aloud by standard pronunciation, the standard pronunciation read aloud is trained into generation alarm model data
Storehouse；
Recording file to the recording pen in the locomotive of acquisition, carries out respectively single pass audio-frequency noise and removes and twin-channel sound
Frequency noise remove；
Recording file to removing noise, extracts all sound bites therein and sound bite corresponding time；
By the interval time between adjacent sound bite is in the range of setting interval threshold and sound bite duration is in setting duration
Sound bite in threshold range is extracted, used as efficient voice fragment；
Efficient voice fragment is matched with the data in model database, the warning message in efficient voice fragment is obtained,
The warning classification of warning message is obtained simultaneously；
The warning message that will be obtained, is contrasted with the time belonging to the warning message of LKJ, obtains the delay of recording file,
According to the delay, recording file is calibrated.
5. a kind of locomotive operation comprehensive analysis method according to claim 4, it is characterised in that：Described pair removes noise
Recording file, when extracting all sound bites therein, extracts signal frequency feature therein by way of MFCC.
6. a kind of locomotive operation comprehensive analysis method according to claim 4, it is characterised in that：The training pattern is set up
Process is：The alarm voice signal read aloud standard pronunciation, extracts the frequecy characteristic of signal, according to what is extracted using MFCC methods
Frequecy characteristic, by Markov algorithm training pattern is set up.
7. a kind of locomotive operation comprehensive analysis method according to claim 4, it is characterised in that：It is described that recording file is entered
Method during row school is：
Arrange with the LKJ standard time as X-axis coordinate and the plane coordinate system of Y-axis coordinate, obtain warning message in select to
Few two warning messages, and at least one time point is obtained in each warning message, and obtain the time point in recording pen
Time and the time in the LKJ standard time, the time-sloped of multiple time points is calculated in above-mentioned plane coordinate system, according to
When the slope carries out simple computation school to the recording file of the recording pen record.
8. a kind of locomotive operation comprehensive analysis method according to claim 4, it is characterised in that：The setting interval threshold
Scope is scope when LKJ reports twice to identical warning message between the shortest time of midfeather and maximum duration.
9. a kind of locomotive operation comprehensive analysis method according to claim 4, it is characterised in that：The setting duration threshold value
Scope in the warning message reported for LKJ between the duration of most long warning message and the duration of most short warning message.
10. a kind of locomotive operation comprehensive analysis method according to claim 1, it is characterised in that：It is described mounting refer to by
Each data at the same time on a timeline set up contact, when certain data is selected in LKJ listed files, can lead to
Setting optionies are crossed, the data of LKJ listed files are linked with the data in certain either multiple or all database,
According to the display packing of the display interface of the prior industrial computer for arranging, the table data of LKJ service datas and other data are entered
Row piecemeal shows.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611218458.XA CN106649745A (en) | 2016-12-26 | 2016-12-26 | Locomotive application comprehensive analysis method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611218458.XA CN106649745A (en) | 2016-12-26 | 2016-12-26 | Locomotive application comprehensive analysis method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106649745A true CN106649745A (en) | 2017-05-10 |
Family
ID=58827355
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611218458.XA Pending CN106649745A (en) | 2016-12-26 | 2016-12-26 | Locomotive application comprehensive analysis method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106649745A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109698965A (en) * | 2017-10-20 | 2019-04-30 | 株洲中车时代电气股份有限公司 | A kind of audio video synchronization playback method, device, equipment and computer storage medium |
| CN112254977A (en) * | 2020-09-03 | 2021-01-22 | 北汽福田汽车股份有限公司 | Data processing method and device based on automatic emergency braking system |
| CN112769933A (en) * | 2021-01-05 | 2021-05-07 | 株洲中车时代电气股份有限公司 | Vehicle-mounted multi-system data fusion analysis system of rail locomotive |
| CN113256153A (en) * | 2021-06-16 | 2021-08-13 | 北京铁道工程机电技术研究所股份有限公司 | Evaluation method and device, storage medium and electronic equipment |
| CN114565989A (en) * | 2022-02-28 | 2022-05-31 | 西安开天铁路电气股份有限公司 | Locomotive 5G dump device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104766605A (en) * | 2015-04-23 | 2015-07-08 | 郑州畅想高科股份有限公司 | Time synchronizing system and method for haulage motor recording device and LKJ |
| CN105528406A (en) * | 2015-12-07 | 2016-04-27 | 河南思维信息技术有限公司 | Locomotive vehicle-mounted file linkage analysis method |
-
2016
- 2016-12-26 CN CN201611218458.XA patent/CN106649745A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104766605A (en) * | 2015-04-23 | 2015-07-08 | 郑州畅想高科股份有限公司 | Time synchronizing system and method for haulage motor recording device and LKJ |
| CN105528406A (en) * | 2015-12-07 | 2016-04-27 | 河南思维信息技术有限公司 | Locomotive vehicle-mounted file linkage analysis method |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109698965A (en) * | 2017-10-20 | 2019-04-30 | 株洲中车时代电气股份有限公司 | A kind of audio video synchronization playback method, device, equipment and computer storage medium |
| CN109698965B (en) * | 2017-10-20 | 2021-07-30 | 株洲中车时代电气股份有限公司 | Video synchronous playing method, device, equipment and computer storage medium |
| CN112254977A (en) * | 2020-09-03 | 2021-01-22 | 北汽福田汽车股份有限公司 | Data processing method and device based on automatic emergency braking system |
| CN112769933A (en) * | 2021-01-05 | 2021-05-07 | 株洲中车时代电气股份有限公司 | Vehicle-mounted multi-system data fusion analysis system of rail locomotive |
| CN113256153A (en) * | 2021-06-16 | 2021-08-13 | 北京铁道工程机电技术研究所股份有限公司 | Evaluation method and device, storage medium and electronic equipment |
| CN113256153B (en) * | 2021-06-16 | 2024-03-08 | 北京铁道工程机电技术研究所股份有限公司 | Evaluation method and device, storage medium and electronic equipment |
| CN114565989A (en) * | 2022-02-28 | 2022-05-31 | 西安开天铁路电气股份有限公司 | Locomotive 5G dump device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106649745A (en) | Locomotive application comprehensive analysis method | |
| Koizumi et al. | ToyADMOS: A dataset of miniature-machine operating sounds for anomalous sound detection | |
| CN110334241B (en) | Quality inspection method, device and equipment for customer service record and computer readable storage medium | |
| Jelil et al. | Spoof Detection Using Source, Instantaneous Frequency and Cepstral Features. | |
| JP6953464B2 (en) | Information push method and equipment | |
| CN101149928B (en) | Sound signal processing method, sound signal processing apparatus and computer program | |
| EP2817800B1 (en) | Modified mel filter bank structure using spectral characteristics for sound analysis | |
| JP2007256153A (en) | System for detecting railway vehicle truck abnormality | |
| US6876964B1 (en) | Apparatus for detecting fatigue and doze by voice, and recording medium | |
| JP2012242214A (en) | Strange noise inspection method and strange noise inspection device | |
| CN113705412B (en) | Track state detection method and device for multi-source data fusion based on deep learning | |
| CN105913059A (en) | Vehicle VIN code automatic identifying system and control method therefor | |
| CN104464755A (en) | Voice evaluation method and device | |
| Gong et al. | Vocalsound: A dataset for improving human vocal sounds recognition | |
| CN110176251B (en) | Automatic acoustic data labeling method and device | |
| CN104572617A (en) | Oral test answer deviation detection method and device | |
| CN113095204B (en) | Double-recording data quality inspection method, device and system | |
| Nayyar et al. | Damage detection based on output-only measurements using cepstrum analysis and a baseline-free frequency response function curvature method | |
| CN109817224A (en) | A kind of voice sensitive word monitor system and method | |
| CN106601241A (en) | Automatic time correcting method for recording file | |
| CN109920447A (en) | Recording fraud detection method based on sef-adapting filter Amplitude & Phase feature extraction | |
| CN107539038B (en) | Vehicle tire pressure state monitoring method and device | |
| CN116741200A (en) | Locomotive fan fault detection method and device | |
| Zambon et al. | Real-time urban traffic noise maps: the influence of Anomalous Noise Events in Milan Pilot area of DYNAMAP | |
| CN110956072B (en) | Driving skill training method based on big data analysis |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170510 |
|
| RJ01 | Rejection of invention patent application after publication |