CN114660378A - Multi-source detection parameter-based contact network comprehensive diagnosis method - Google Patents
Multi-source detection parameter-based contact network comprehensive diagnosis method Download PDFInfo
- Publication number
- CN114660378A CN114660378A CN202210187673.7A CN202210187673A CN114660378A CN 114660378 A CN114660378 A CN 114660378A CN 202210187673 A CN202210187673 A CN 202210187673A CN 114660378 A CN114660378 A CN 114660378A
- Authority
- CN
- China
- Prior art keywords
- data
- detection data
- detection
- value
- standard
- 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
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses a contact net comprehensive diagnosis method based on multi-source detection parameters, which comprises the following steps: s1, automatically aligning the detection data; s2, standard state generation is carried out on the detection data after automatic alignment; and S3, carrying out potential defect analysis on the detection data in the standard state, wherein the potential defect analysis comprises time domain analysis and frequency domain analysis, and obtaining the potential defect of the overhead line system. The comprehensive diagnosis method for the contact network based on the multi-source detection parameters mainly utilizes multi-dimensional detection data for multiple times to comprehensively diagnose and evaluate the contact network. Multi-dimensional inspection data includes, but is not limited to, pull-out values, lead-height, bow net contact force, hard spots.
Description
Technical Field
The invention relates to the field of contact network detection, in particular to a comprehensive diagnosis method for a contact network based on multi-source detection parameters.
Background
Along with the rapid development of urban rail transit, the contact net detection system is more accurate and comprehensive, and can comprehensively monitor the contact net state from multiple aspects. At present, the utilization of the detection data mainly only stays in the judgment of whether the detection data is over-limited, the over-limit threshold value is generally set to be larger, and the data which is not over-limited is rarely concerned. The existing utilization methods for detection data include: 1. establishing a reliability model of the whole overhead line system by using a fault tree analysis method, an event number analysis method and a credibility theory; 2. analyzing the detection data of the contact network based on Hilbert-Huang transform, and evaluating the whole operation state of the contact network; 3. evaluating the health state of the contact net by utilizing a fuzzy comprehensive evaluation method, a gray cluster, a cloud model and the like; 4. and evaluating the health state of the contact network by using a set pair analysis method and an evidence theory, and fusing all evaluation indexes in the evaluation index system of the health state of the contact network layer by layer. In the above methods, method 1 only utilizes the defect data to perform fault analysis, and does not diagnose the non-overrun data; the method 2 is to analyze the whole line data and cannot judge the specific position; the method 3 analyzes the uncertainty of the processing information and the insufficiency of the information fusion, directly performs weighted fusion on each evaluation index of the contact network state, and is easy to misjudge the contact network state; the method 4 can accurately complete the evaluation of the health state of the contact network, but cannot accurately position the specific defect position, and only evaluates the health state of the whole line.
Disclosure of Invention
Aiming at the defects in the prior art, the comprehensive diagnosis method for the contact network based on the multi-source detection parameters solves the problem that the detection data of the contact network is incomplete.
In order to achieve the purpose of the invention, the invention adopts the technical scheme that: a contact network comprehensive diagnosis method based on multi-source detection parameters comprises the following steps:
s1, automatically aligning the detection data;
s2, standard state generation is carried out on the detection data after automatic alignment;
and S3, carrying out potential defect analysis on the detection data in the standard state, wherein the potential defect analysis comprises time domain analysis and frequency domain analysis, and obtaining the potential defect of the overhead line system.
Further: the specific steps of step S1 are:
s11, standard data establishment and detection data preprocessing;
s12, positioning the anchor section area through data matching and calculation;
and S13, performing corresponding difference value sampling operation on the data according to the matching result, and aligning the detected data waveform with the design curve.
Further: the standard data establishment comprises the following steps:
and establishing standard data by using positioning points and anchor section design values of the contact network and corresponding mileage information through a linear difference method.
Further: the detection data preprocessing comprises the following steps:
and (3) sequentially carrying out difference with the next data from the first detection data, when the level difference value is greater than a set threshold value T, dividing the previous data and the next data into two point sets, and finally removing the points of which the number of the point sets is less than a threshold value M to filter the abnormal jump points.
Further: the step S12 specifically includes:
s121, dividing the detection data and the standard data into a plurality of parts under the condition that the pull-out value is suddenly changed;
s122, matching each anchor segment of the detection data with an anchor segment of the standard data by using the correlation coefficient, and selecting the longest ordered subsequence from the matching result as an optimal solution;
s123, calculating and matching the stretching scale of the optimal solution relative to the standard data in pairs to obtain the stretching scale smaller than the threshold value LmaxAs a result of (1).
Further, the method comprises the following steps: the step S2 includes:
and counting the measurement standard range corresponding to each detection point according to the confidence interval according to the measured value of each detection point to generate a detection data standard state area.
Further: the time domain analysis in step S3 includes trend diagnosis and/or analysis of variance.
Further: the trend diagnosis is specifically as follows:
calculating the maximum difference diff between each acquisition point of the current detection data and the area data in the standard stateiIf the difference value is in the standard state area, the difference value is 0, and the maximum difference value sum sigma diff between each positioning point is obtained according to the spaniCalculating the change rate delta m of the detection data compared with the standard state area by using the sum of the maximum difference value and the central value of the standard state area;
and (4) synchronously analyzing the change rate of the detection data of different item points for many times, and if the change trends of the detection data for many times are the same, gradually exceeding the range of the standard state interval by the detection data, and regarding the detection data as the subhealth state of the contact network.
Further: the analysis of variance comprises:
s311, respectively calculating the height leading pulling value variance values of the whole line;
s312, setting the lead weight value as a pull-out value weight of 2 times; adding the variance and the assignment of the lead height pull-out value respectively to obtain a variance sum based on the lead height pull-out value;
and S313, calculating the variance of the detection data at different moments by the multiple detection data, and when the variance of the pull-out value is continuously increased for multiple times, the contact network has hidden danger. .
Further: the frequency domain analysis in step S3 includes:
s321, the detection data are subjected to averaging, and the overall average value of each detection data is subtracted to enable the overall average value of the data to be 0;
and S322, selecting Hilbert-yellowing to realize time-frequency conversion of detection data, calculating instantaneous energy of the signal according to time-frequency information of the detection data, setting an energy threshold, and judging a potential defect area if the instantaneous energy exceeds the energy threshold.
The invention has the beneficial effects that:
1. the comprehensive diagnosis method for the contact network based on the multi-source detection parameters mainly utilizes multi-dimensional detection data for multiple times to comprehensively diagnose and evaluate the contact network. Multi-dimensional inspection data includes, but is not limited to, pull-out values, lead-height, bow net contact force, hard points.
2. The invention automatically realizes the alignment of multiple detection data and avoids the failure of subsequent analysis caused by mileage error of the detection data.
3. According to the invention, a specific analysis method is selected according to the self characteristics of the detection points at different sources, and the characteristics of the catenary are amplified so as to effectively diagnose the detection data of the catenary.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a diagram illustrating standard pull-out value data and actual detection pull-out value data according to an embodiment of the present invention;
FIG. 3 is a diagram illustrating alignment results of a detected data waveform and a design value curve according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating variance versus time according to an embodiment of the present invention;
FIG. 5 is a graphical representation of line pressure data in an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and it will be apparent to those skilled in the art that various changes may be made without departing from the spirit and scope of the invention as defined and defined in the appended claims, and all matters produced by the invention using the inventive concept are protected.
As shown in fig. 1, a comprehensive diagnosis method for a catenary based on multi-source detection parameters includes the following steps:
s1, automatically aligning the detection data;
the position information of the detection data is usually from an on-vehicle speed sensor, a GPS, or the like, but an accumulated mileage error due to factors such as a reduction in the wheel diameter due to wheel wear due to relative sliding of the wheel rail is unavoidable. The mileage error of the detection data can directly influence the subsequent state evaluation and analysis of the contact network, the accurate positioning of the defects and the cause judgment of the defects. Fig. 2 represents standard pull-out value data and actual detection pull-out value data, which produce a significant mileage shift due to a gradual accumulated mileage error.
S11, standard data establishment and detection data preprocessing;
collecting design values of positioning points and anchor sections for erecting the contact network, realizing the establishment of all-line standard data by using a linear difference method according to the sampling interval of 0.25m point by using the design data of the positioning points and the anchor section joints and corresponding mileage information, and making the sequence of the routine as (x)1,x2,…xm) The pull-out value sequence is (y)1,y2,…ym)；
In the above formula, yiIs the ith pull-out value, y0Design value for pull out value at fix point, ymIs the m-th pull-out value, xmIs the mth mileage, x0For mileage information at a location point, xiIs the ith mileage;
the method comprises the steps that detection data of the contact network are generally influenced by data transmission, electromagnetic interference and the like, more abnormal points are possibly generated, the abnormal points need to be eliminated before subsequent alignment operation is carried out, abnormal jumping points are filtered in a data aggregation mode, the difference between the first detection data and the next detection data is sequentially carried out, when the difference value is larger than a set threshold value T, the previous detection data and the next detection data are divided into two point sets, the analogy is carried out, and finally the points with the number smaller than the threshold value M are eliminated.
S12, positioning the anchor section area;
the main characteristic of the catenary pull-out value curve in the anchor section area is pull-out value mutation, and the detection data and the standard data are respectively divided into a plurality of parts under the condition of pull-out value mutation, namely standard data smd1,smd2…smdnDetecting data Amd1,Amd2…Amdm；
And respectively matching each anchor segment of the detection data with the anchor segment of the standard data by using a correlation coefficient formula, wherein the correlation coefficient formula is as follows:
in the above formula, r is a correlation coefficient, xiAnd xi' is the ith standard data and detection data, and m is the number of the standard data and the detection data;
recording the matching pairs with the correlation coefficient exceeding the threshold value T, selecting the longest optimal sequence matching pair, recording the positions corresponding to the matched detection data according to the sequence of the standard data, selecting the longest ordered subsequence from the matching result as the optimal solution, calculating the stretching scale delta L of the matched optimal solution compared with the standard data in pairs, and deleting the stretching scale which is larger than the threshold value LmaxKeeping the stretching scale smaller than the threshold LmaxThe matching result of (1).
And S13, performing corresponding difference sampling operation on the data according to the matching result, and aligning the detected data waveform with the design curve, wherein the alignment result is shown in FIG. 3.
S2, standard state generation is carried out on the detection data after automatic alignment;
after the detection data are aligned, the mileage error of the detection data is eliminated, and the state judgment of the detection parameters cannot singly depend on the comparison between the design value and the measured value due to the existence of certain system errors of the detection equipment. The invention provides a method for generating a standard state of detection data. The method relies on multiple detection data, after the detection data are aligned, the measured value of each detection point is counted, and the measurement standard range corresponding to each point is counted according to 95% confidence interval
Wherein
Is mean value, sigma is variance, n is data number, and is generated according to measurement standard rangeAnd detecting the standard state range of the data.
And S3, carrying out potential defect analysis on the detection data in the standard state, wherein the potential defect analysis comprises time domain analysis and frequency domain analysis, and obtaining the potential defect of the overhead line system.
The static parameters (lead-high pull-out values) of the contact network are relatively stable and the influence of external factors is small, so that the relatively direct time domain analysis of the static parameters of the contact network is mainly selected for comprehensive diagnosis, and the time domain analysis comprises change trend diagnosis and variance analysis.
The trend diagnosis is specifically as follows:
calculating the maximum difference diff between each acquisition point of the current detection data and the area data in the standard stateiIf the difference value is in the middle of the standard state, the difference value is 0, and the maximum difference value sum sigma diff between each positioning point is obtained according to the spaniTaking the sum of the maximum difference value and the central value of the corresponding standard state area as the change rate delta m of the current detection data compared with the standard state area;
the calculation formula is as follows:
in the above formula, [ X ]imin,Ximax]Is a standard state area range;
the change rate can correctly reflect the trend change condition of the detection data, and the standard state area can effectively filter the change caused by normal fluctuation of the detection itself. And (4) synchronously analyzing the change rate of the detection data of different item points for many times, and if the change trends of the detection data for many times are the same, gradually exceeding the range of the standard state interval by the detection data, and regarding the detection data as the subhealth state of the contact network.
The overall variance of the detection data is an important element for evaluating whether the overhead line system is smooth, and the forward and backward comparison diagnosis of the overhead line system can be realized according to the detection data of the line for multiple times. Because the dynamic detection parameters have certain uncertainty, the invention selects the static detection parameters to lead the height and pull out value to carry out comprehensive evaluation on the contact network.
The analysis of variance specifically comprises: respectively solving the height-leading pulling value variance values of the whole line; according to weighting consideration of static quality calculation of a contact network in a CQI (channel quality indicator) contact network static detection evaluation method, setting a lead height weight value as a pull-out value weight of 2 times, wherein the lead height weight is 0.67 after normalization, and the pull-out value weight is 0.33; and respectively adding the variance and the assignment of the lead-high pull-out value to obtain the variance sum based on the lead-high pull-out value, and calculating the variance of the detection data at different moments by using the detection data for multiple times, wherein as shown in fig. 4, when the variance sum of the lead-high pull-out values is increased gradually for multiple times continuously, hidden dangers exist in the contact network at the moment.
The dynamic parameters (pressure hard points and the like) of the contact network are greatly influenced by the outside, and different pantograph and different vehicles can cause the detection data to have large difference, so that a frequency domain analysis method is selected, and corresponding frequency components are extracted from the detection data to carry out fault identification and judgment.
The frequency domain analysis specifically comprises:
the detection data are subjected to averaging, and the overall average value of each detection data is subtracted to enable the overall average value of the data to be 0;
and (3) selecting Hilbert-yellowing to realize time-frequency conversion of detection data, calculating instantaneous energy of a signal according to time-frequency information of the detection data, setting an energy threshold, and judging a potential defect area if the instantaneous energy exceeds the energy threshold.
Taking a certain line pressure data as an example, it can be seen in fig. 5 that the upper detection data obviously fluctuates most at the same time when the instantaneous energy exceeds the threshold line. For the dynamic detection parameters, the instantaneous energy of the dynamic detection parameters can sensitively reflect the fluctuation condition of the dynamic detection parameters, the smoothness of the contact net at each position is reflected, and the influence from different detection devices can be effectively avoided.
Claims (10)
1. A contact network comprehensive diagnosis method based on multi-source detection parameters is characterized by comprising the following steps:
s1, automatically aligning the detection data;
s2, standard state generation is carried out on the detection data after automatic alignment;
and S3, carrying out potential defect analysis on the detection data in the standard state, wherein the potential defect analysis comprises time domain analysis and frequency domain analysis, and obtaining the potential defect of the overhead line system.
2. The multi-source detection parameter-based catenary comprehensive diagnosis method according to claim 1, wherein the specific steps of step S1 are as follows:
s11, standard data establishment and detection data preprocessing;
s12, positioning the anchor section area through data matching and calculation;
and S13, performing corresponding difference value sampling operation on the data according to the matching result, and aligning the detected data waveform with the design curve.
3. The method for comprehensively diagnosing the overhead line system based on the multi-source detection parameters of claim 2, wherein the standard data is established and comprises the following steps:
and establishing standard data by using positioning points and anchor section design values of the contact network and corresponding mileage information through a linear difference method.
4. The method for comprehensively diagnosing the overhead line system based on the multi-source detection parameters of claim 2, wherein the detection data preprocessing comprises the following steps:
and (3) sequentially carrying out difference with the next data from the first detection data, when the level difference value is greater than a set threshold value T, dividing the previous data and the next data into two point sets, and finally removing the points of which the number of the point sets is less than a threshold value M to filter the abnormal jump points.
5. The multi-source detection parameter contact network comprehensive diagnosis method according to claim 2, wherein the step S12 specifically comprises:
s121, dividing the detection data and the standard data into a plurality of parts under the condition that the pull-out value is suddenly changed;
s122, matching each anchor segment of the detection data with an anchor segment of the standard data by using the correlation coefficient, and selecting the longest ordered subsequence from the matching result as an optimal solution;
s123, calculating and matching the stretching scale of the optimal solution relative to the standard data in pairs to obtain the stretching scale smaller than the threshold value LmaxThe result of (1).
6. The method for comprehensively diagnosing the overhead line system based on the multi-source detection parameters of claim 1, wherein the step S2 comprises:
and counting the measurement standard range corresponding to each detection point according to the confidence interval according to the measured value of each detection point to generate a detection data standard state area.
7. The method for comprehensively diagnosing the overhead line system based on the multi-source detection parameters of claim 1, wherein the time domain analysis in the step S3 includes trend analysis and/or variance analysis.
8. The method for comprehensively diagnosing the overhead line system based on the multi-source detection parameters of claim 7, wherein the trend of change diagnosis is specifically as follows:
calculating the maximum difference diff between each acquisition point of the current detection data and the standard state area dataiIf the locating point is in the standard state area, the locating point is 0, and the maximum difference sum sigma diff between every two locating points is obtained according to the strideiCalculating the change rate delta m of the detection data compared with the standard state area by using the sum of the maximum difference value and the central value of the standard state area;
and (4) synchronously analyzing the change rate of the detection data of different item points for many times, and if the change trends of the detection data for many times are the same, gradually exceeding the range of the standard state interval by the detection data, and regarding the detection data as the subhealth state of the contact network.
9. The method for comprehensively diagnosing the overhead line system based on the multi-source detection parameters of claim 8, wherein the analysis of variance comprises:
s311, respectively calculating the height leading pulling value variance values of the whole line;
s312, setting the lead weight value as a pull-out value weight of 2 times; adding the variance and the assignment of the lead height pull-out value respectively to obtain a variance sum based on the lead height pull-out value;
and S313, calculating the variance of the detection data at different moments by using the detection data for multiple times, and when the variance sum of the continuous multiple-time derivative pull-out values gradually increases, the contact network has hidden danger at the moment.
10. The multi-source detection parameter-based catenary comprehensive diagnosis method according to claim 1, wherein the frequency domain analysis in step S3 comprises:
s321, the detection data are subjected to averaging, and the overall average value of each detection data is subtracted to enable the overall average value of the data to be 0;
and S322, selecting Hilbert-yellowing to realize time-frequency conversion of detection data, calculating instantaneous energy of the signal according to time-frequency information of the detection data, setting an energy threshold, and judging a potential defect area if the instantaneous energy exceeds the energy threshold.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210187673.7A CN114660378A (en) | 2022-02-28 | 2022-02-28 | Multi-source detection parameter-based contact network comprehensive diagnosis method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210187673.7A CN114660378A (en) | 2022-02-28 | 2022-02-28 | Multi-source detection parameter-based contact network comprehensive diagnosis method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114660378A true CN114660378A (en) | 2022-06-24 |
Family
ID=82027456
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210187673.7A Pending CN114660378A (en) | 2022-02-28 | 2022-02-28 | Multi-source detection parameter-based contact network comprehensive diagnosis method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114660378A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114996318A (en) * | 2022-07-12 | 2022-09-02 | 成都唐源电气股份有限公司 | Automatic judgment method and system for processing mode of abnormal value of detection data |
| CN115081911A (en) * | 2022-07-04 | 2022-09-20 | 成都唐源电气股份有限公司 | Contact network potential risk identification method and device based on dynamic detection data |
-
2022
- 2022-02-28 CN CN202210187673.7A patent/CN114660378A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115081911A (en) * | 2022-07-04 | 2022-09-20 | 成都唐源电气股份有限公司 | Contact network potential risk identification method and device based on dynamic detection data |
| CN114996318A (en) * | 2022-07-12 | 2022-09-02 | 成都唐源电气股份有限公司 | Automatic judgment method and system for processing mode of abnormal value of detection data |
| CN114996318B (en) * | 2022-07-12 | 2022-11-04 | 成都唐源电气股份有限公司 | Automatic judgment method and system for processing mode of abnormal value of detection data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114660378A (en) | Multi-source detection parameter-based contact network comprehensive diagnosis method | |
| CN111562108A (en) | Rolling bearing intelligent fault diagnosis method based on CNN and FCMC | |
| CN111776006B (en) | Pre-alarming method and device for train axle temperature | |
| CN113657221B (en) | Power plant equipment state monitoring method based on intelligent sensing technology | |
| CN111460392A (en) | Magnetic suspension train and suspension system fault detection method and system thereof | |
| CN116662111B (en) | Intelligent network management control system for train | |
| CN113805018A (en) | Intelligent identification method for partial discharge fault type of 10kV cable of power distribution network | |
| CN117390378B (en) | Intelligent management method and system for dual-carbon platform data | |
| CN114422379B (en) | Analysis method for multi-platform equipment wireless networking | |
| CN115081331A (en) | Wind turbine generator running state abnormity detection method based on state parameter reconstruction error | |
| CN114739671A (en) | Bearing fault diagnosis method based on improved generalized S transformation | |
| CN112329272B (en) | Residual life fitting prediction method integrating photovoltaic module degradation and life data | |
| CN115656764B (en) | Illumination quality detection method for light-emitting diode | |
| CN116257651B (en) | Intelligent monitoring system for abnormal sound of through channel cab apron | |
| CN117630797A (en) | Ammeter health state detection method, system and storage medium based on working current | |
| CN112033656A (en) | Mechanical system fault detection method based on broadband spectrum processing | |
| CN113672859A (en) | Switch point machine fault acoustic diagnosis system | |
| Zhao et al. | CPDM: An efficient crowdsensing-based pothole detection and measurement system design | |
| CN113758627A (en) | Transient flow event detection method for water supply pipe network | |
| CN118100446B (en) | Intelligent detection method for node running state of power system | |
| US20240271990A1 (en) | Vibration measurement error determination method and vibration error discernment system using the same | |
| CN115158399B (en) | Time sequence signal abnormality detection method and system | |
| CN118091489B (en) | Method for detecting state of glass insulator of power transmission line | |
| CN117057621B (en) | Quantitative analysis method for investment cost and reliability improvement level of power generation system | |
| CN114001850A (en) | Pantograph pressure detection method and system |
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 |