Method, system and equipment for monitoring defects of trackside non-electric train wheels

Abstract

The invention provides a method, a system and equipment for monitoring trackside non-electric train wheel defects. The method comprises the following steps: filtering original wavelength data to obtain original smooth wavelength data; extracting an original vehicle passing waveform according to the original smooth wavelength data; carrying out normalization processing on the original vehicle passing waveform; and realizing wheel defect monitoring according to a corresponding relation between the reference waveform and the original vehicle passing waveform after normalization processing. Through detailed description of an automatic identification process of a defect waveform, rapid and accurate wheel tread defect detection is realized; on-line wheel health state monitoring can be achieved, train operation is not affected, and train transportation safety and efficiency are guaranteed. The accuracy of train wheel data extraction and the accuracy of wheel defect detection are guaranteed through an algorithm.

Description

Technical Field

[0001] The invention belongs to the field of train control, and in particular relates to a method, system and equipment for monitoring wheel defects of a non-electrical train on the trackside. Background Art

[0002] The healthy state of the contact tread between the wheel and the rail is related to the safety and comfort of rail transit driving. Long-term operation of the train will cause damage to the wheel tread.

[0003] At present, the detection of train wheel tread in China mainly relies on manual mechanical measurement, which is inefficient and has large errors. In the prior art, the online monitoring system based on fiber grating sensing of train wheel tread status fixes the fiber grating sensor under the rail and arranges an array composed of multiple fiber grating sensors to ensure that the detection range is larger than the circumference of the wheel tread. The wheel tread defects are identified by analyzing the waveform of the wheel passing through the monitoring area. However, the automatic identification process of the defect waveform is not described in detail. Summary of the invention

[0004] In view of the above problems, the present invention proposes a method for monitoring wheel defects of a train without electricity on the trackside, the method comprising:

[0005] Filtering the original wavelength data to obtain original smoothed wavelength data;

[0006] Extracting the original vehicle passing waveform according to the original smoothed wavelength data;

[0007] Normalizing the original vehicle passing waveform;

[0008] Wheel defect monitoring is achieved based on the correspondence between the reference waveform and the normalized original vehicle passing waveform.

[0009] Further, extracting the original vehicle passing waveform according to the original smoothed wavelength data includes:

[0010] determining the amount of waveform change between adjacent troughs in the raw smoothed wavelength data;

[0011] The original wavelength data of a single axle is extracted according to the relationship between the waveform change amount and the first threshold.

[0012] Further, extracting the original vehicle passing waveform according to the original smoothed wavelength data includes:

[0013] Determine the initial array element, the initial peak value and the initial trough value according to the collected original wavelength and the filtered initial wavelength;

[0014] Determine the initial state and the second state in sequence, alternately iterate the peak value and the trough value of the two states according to the waveform change, and update the array elements;

[0015] When there is a wavelength value greater than the trough value after the alternating iteration, the state is adjusted to the initial state, and the original wavelength data of a single axle is extracted according to the relationship between the waveform change amount in the updated array element and the first threshold.

[0016] Furthermore, constructing a reference waveform includes:

[0017] Determine the length relationship between the original reference waveform and the normalized original vehicle passing waveform;

[0018] When the original reference waveform length does not meet the normalized original vehicle passing waveform length of a predetermined value multiple, the original reference waveform is interpolated until the reference waveform is greater than the normalized original vehicle passing waveform length of a predetermined value multiple.

[0019] Furthermore, the wheel defect monitoring is realized according to the correspondence between the reference waveform and the normalized original vehicle passing waveform, including:

[0020] Determine the output matrix array of the reference waveform array length and the original vehicle passing waveform array length after normalization according to the longest common subsequence algorithm;

[0021] Determine the output result according to the ratio of the output matrix array to the normalized original vehicle passing waveform array length;

[0022] Wheel defect monitoring is achieved based on comparison of the output result with the second threshold value.

[0023] Furthermore, the matrix array for determining the reference waveform array length and the normalized original vehicle passing waveform array length according to the longest common subsequence algorithm includes:

[0024] Determine the initial matrix according to the reference waveform array length and the normalized original vehicle passing waveform array length;

[0025] Determining a third threshold of waveform tolerance according to waveform fluctuations;

[0026] The initial matrix is ​​modified according to the relationship between the difference between the reference waveform and the original vehicle passing waveform after normalization and the third threshold value;

[0027] Correct the current value of the initial matrix;

[0028] Output the corrected matrix array.

[0029] The present invention also provides a trackside non-electric train wheel defect monitoring system, the system comprising:

[0030] A filtering unit, an extraction unit, a normalization unit and a monitoring unit, wherein:

[0031] A filtering unit, used for filtering the original wavelength data to obtain original smoothed wavelength data;

[0032] An extraction unit, which is in communication with the filtering unit and is used to extract the original vehicle passing waveform according to the original smoothed wavelength data;

[0033] A normalization unit, connected to the extraction unit for normalization processing of the original vehicle passing waveform;

[0034] The monitoring unit is connected to the normalization unit for communication and is used to realize wheel defect monitoring according to the corresponding relationship between the reference waveform and the original vehicle passing waveform after normalization processing.

[0035] Further, the extraction unit is used to extract the original vehicle passing waveform according to the original smoothed wavelength data, including:

[0036] determining the amount of waveform change between adjacent troughs in the raw smoothed wavelength data;

[0037] The original wavelength data of a single axle is extracted according to the relationship between the waveform change amount and the first threshold.

[0038] Further, the extraction unit is used to extract the original vehicle passing waveform according to the original smoothed wavelength data, including:

[0039] Determine the initial array element, the initial peak value and the initial trough value according to the collected original wavelength and the filtered initial wavelength;

[0040] Determine the initial state and the second state in sequence, alternately iterate the peak value and the trough value of the two states according to the waveform change, and update the array elements;

[0041] When there is a wavelength value greater than the trough value after the alternating iteration, the state is adjusted to the initial state, and the original wavelength data of a single axle is extracted according to the relationship between the waveform change amount in the updated array element and the first threshold.

[0042] Furthermore, the system further comprises a construction unit, wherein the construction unit is used to construct a reference waveform including:

[0043] Determine the length relationship between the original reference waveform and the normalized original vehicle passing waveform;

[0044] When the original reference waveform length does not meet the normalized original vehicle passing waveform length of a predetermined value multiple, the original reference waveform is interpolated until the reference waveform is greater than the normalized original vehicle passing waveform length of a predetermined value multiple.

[0045] Furthermore, the monitoring unit is used to implement wheel defect monitoring according to the corresponding relationship between the reference waveform and the normalized original vehicle passing waveform, including:

[0046] Determine the output matrix array of the reference waveform array length and the original vehicle passing waveform array length after normalization according to the longest common subsequence algorithm;

[0047] Determine the output result according to the ratio of the output matrix array to the normalized original vehicle passing waveform array length;

[0048] Wheel defect monitoring is achieved based on comparison of the output result with the second threshold value.

[0049] The present invention also provides an electronic device, comprising at least one processor and at least one memory, wherein the memory is data-connected to the processor, wherein:

[0050] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the trackside non-powered train wheel defect monitoring method of the present invention.

[0051] The method, system and device for monitoring wheel defects of non-electrical trackside trains of the present invention can realize fast and accurate wheel tread defect detection by describing the automatic recognition process of defect waveform in detail; it can realize online wheel health status monitoring without affecting train operation, and ensure train transportation safety and efficiency. The accuracy of train wheel data extraction and wheel defect detection is guaranteed by the algorithm.

[0052] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0054] Figure 1 A schematic flow chart of a method for monitoring wheel defects of a train without power supply at the trackside in an embodiment of the present invention is shown;

[0055] Figure 2 A schematic diagram of the layout of the fiber Bragg grating sensor in an embodiment of the present invention is shown;

[0056] Figure 3 A schematic diagram showing comparison waveforms before and after filtering in an embodiment of the present invention is shown;

[0057] Figure 4 A schematic diagram of extracting original wavelength data of a single axle in an embodiment of the present invention is shown;

[0058] Figure 5 A schematic diagram of a normalized processing waveform change in an embodiment of the present invention is shown;

[0059] Figure 6 A schematic diagram of a reference waveform in an embodiment of the present invention is shown;

[0060] Figure 7 A schematic diagram of implementing wheel defect monitoring in an embodiment of the present invention is shown;

[0061] Figure 8 A schematic structural diagram of a train wheel defect monitoring system without power supply at the trackside in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] The health status of the tread where the wheel contacts the rail is related to the safety and comfort of rail transit. Long-term operation of trains can cause damage to the wheelset tread. Currently, the tread detection of train wheels in China mainly relies on manual mechanical measurement, which is inefficient. In the existing technology, the waveform of the wheel passing through the monitoring area is analyzed to identify the wheel tread defects, but the automatic identification process of the defect waveform is not specifically described.

[0064] The present invention provides a method for monitoring wheel defects of a train without power supply at the trackside. Figure 1 The figure shows a schematic flow chart of a method for monitoring wheel defects of a train without power supply at the trackside in an embodiment of the present invention. Figure 1 The method includes filtering the original wavelength data to obtain the original smoothed wavelength data; extracting the original vehicle passing waveform according to the original smoothed wavelength data; normalizing the original vehicle passing waveform; and realizing wheel defect monitoring according to the correspondence between the reference waveform and the normalized original vehicle passing waveform.

[0065] In the embodiment of the present invention, the specific implementation of the method for monitoring wheel defects of a train without electricity on the trackside is described:

[0066] In the embodiment of the present invention, by arranging fiber grating sensors beside the track, Figure 2 FIG. 1 shows a schematic diagram of the layout of the fiber Bragg grating sensor in an embodiment of the present invention. Figure 2 In the present invention, a fiber grating sensor 1 is arranged at the bottom of the rail through fasteners, and a strain detection grating 2 is fixed inside the fiber grating sensor 1. When a train passes by, the strain detection grating 2 arranged inside the fiber grating sensor 1 collects the strain signal when the train passes by to monitor the health status of the wheel tread.

[0067] In the embodiment of the present invention, there is no specific limitation on the models of the fiber grating sensor 1 and the strain detection grating 2, and any type that can realize the collection of train strain signals is within the protection scope of the present invention.

[0068] Specifically, in an embodiment of the present invention, in order to accurately extract the passing waveform of a single wheel and avoid interference of waveform jitter on the passing waveform selection, a low-pass smoothing filter is performed on the original wavelength ω to obtain the filtered wavelength λ. In an embodiment of the present invention, the comparison waveforms before and after the low-pass smoothing filter are described.

[0069] Figure 3 FIG. 4 shows a schematic diagram of waveform comparison before and after filtering in an embodiment of the present invention. Figure 3 In the figure, the original waveform is represented by a dotted line, the filtered waveform is represented by a solid line, the abscissa represents the sampling point, and the ordinate represents the wavelength change. In the embodiment of the present invention, the sampling points between 3670 and 3900 are selected for illustration. The original waveform has four fluctuation sections in the sampling interval. The sampling points of the first fluctuation section are between 3670 and 3740, and the waveform fluctuation range is between -15 and 55 pm; the sampling points of the second fluctuation section are between 3740 and 3790, and the waveform fluctuation range is between -15 and 44 pm; the sampling points of the third fluctuation section are between 3790 and 3840, and the waveform fluctuation range is between 8 and 75 pm. The fluctuation of about 8 values ​​of the left and right sampling points around the peak of the third fluctuation section exceeds 67 pm, and the fluctuation is relatively large; the sampling points of the fourth fluctuation section are between 3840 and 3880, and the waveform fluctuation range is between -15 and 58 pm. After low-pass smoothing filtering, the filtered waveform also represents four fluctuation segments, and smoothing filtering is mainly performed on the third segment with larger fluctuations. After smoothing filtering of the first original waveform segment, the waveform fluctuation range is between -15-50pm, and the waveform of the sampling point at 3690 is smoothed; after smoothing filtering of the second original waveform segment, the waveform fluctuation range is between -15-40pm; after smoothing filtering of the third original waveform segment, the waveform fluctuation range is between -15-50pm, and the waveform of the sampling point at 3790 is smoothed; after smoothing filtering of the fourth original waveform segment, the waveform fluctuation range is between -15-53pm, and the waveform of the sampling point at 3890 is smoothed.

[0070] Specifically, in an embodiment of the present invention, extracting the original vehicle passing waveform based on the original smoothed wavelength data includes: determining the waveform change between adjacent troughs in the original smoothed wavelength data; and extracting the original wavelength data of a single axle based on the relationship between the waveform change and a first threshold.

[0071] In the embodiment of the present invention, based on the filtered waveform, according to the positions of the two troughs of the phase relationship, and adding the waveform change amount as a condition, the original waveform of a single axle is extracted. The original waveform will have irregular oscillations, which may cause inaccurate waveform extraction of a single axle. By extracting the waveform between the two troughs of the filtered waveform and adding the restriction condition of the waveform change amount, some interference waveforms can be prevented from being misjudged as vehicle passing waveforms, thereby improving the accuracy of axle judgment. The following is a detailed description of the process of extracting the original vehicle passing waveform:

[0072] In an embodiment of the present invention, extracting the original vehicle passing waveform based on the original smoothed wavelength data specifically includes: determining the initial array element, the initial peak value and the initial trough value based on the collected original wavelength and the filtered initial wavelength; determining the initial state and the second state in sequence, alternately iterating the peak value and trough value of the two states based on the waveform change amount, and updating the array elements; when there is a wavelength value greater than the trough value after the alternating iteration, adjusting the state to the initial state, and extracting the original wavelength data of a single axle based on the relationship between the waveform change amount in the updated array element and the first threshold.

[0073] In the embodiment of the present invention, combined with Figure 4 The process of extracting the original vehicle passing waveform is described in detail. Figure 4 A schematic diagram of extracting the original wavelength data of a single axle in an embodiment of the present invention is shown:

[0074] The original wavelength collected is ω 0 , the initial wavelength after filtering is λ 0 , and set λ low =λ high =λ 0 ,ω raw =ω 0 , at this timeω raw The array has only one elementω 0 , set the initial state state = 1;

[0075] When state=1, the new filtered wavelength value λ 1 Less than or equal to λ low When λ 1 Assign to λ low , and the array ω is raw The value is updated to the new original wavelength ω 1 , at this timeω rawThe array has only one elementω 1 ,correspond Figure 4 The time period t0-t1 in;

[0076] When state=1, the new filtered wavelength value λ 2 Greater than λ low When , the new original wavelength value is stored in the array ω raw , at this time the array ω raw The element increases, when the wavelength value λ 2 Greater than λ high When λ 2 Assign to λ high , and update the descending threshold th down =0.5*(λ high -λ low ),correspond Figure 4 The time period t1-t2 in;

[0077] When state=1, when λ high With the new wavelength value λ 3 The difference is greater than the falling threshold th down (λ high -λ 3 >th down ), then 3 Assign to λ low , and change the state to 2, corresponding to the time point t3 in Figure 4;

[0078] When state = 2, the new original wavelength value is stored in the array ω raw , at this time the array ω raw The elements of are increased. When the new filtered wavelength value λ 4 Less than or equal to λ low When λ 4 Assign to λ low ,correspond Figure 4 The t3-t4 time period;

[0079] When the new wavelength value is greater than λ low When the state is set to 1, the array ω is judged raw The difference between the maximum and minimum values ​​in the array, when the difference is greater than 40, in the embodiment of the present invention, the difference represents the limiting condition of the waveform change when the axle original waveform is extracted, that is, the waveform change is greater than 40, and the waveform is considered to be a valid vehicle passing waveform, and the next step is performed. This value needs to be set according to different sensor sensitivities and vehicle weights, that is, the data recorded in this array is considered to be the original waveform data of a single wheel passing, and the next step is judged.

[0080] In the embodiment of the present invention, the process of normalizing the original vehicle passing waveform is also included. Figure 5 FIG. 4 shows a schematic diagram of the normalized waveform change in an embodiment of the present invention. Specifically, the original waveform data of a single wheel passing recorded during the extraction process is normalized to obtain ω nor In the embodiment of the present invention, the wavelength range of the normalized waveform is unified to 0-100.

[0081] In the embodiment of the present invention, wheel defect monitoring is achieved by setting a reference waveform and by using the correspondence between the reference waveform and the original vehicle passing waveform after normalization processing. Figure 6 Schematic diagram of reference waveform in an embodiment of the present invention is shown. Figure 6 In the construction of reference waveform, the ideal reference waveform ω is constructed based on multiple normal vehicle passing waveforms. ref , as shown in the figure below, in order to ensure that the algorithm can adapt to different vehicle speed conditions, the reference waveform and the passing waveform of a single wheel need to be analyzed before judging the wheel tread state. nor Perform length judgment. When the length of the reference waveform does not meet three times the length of the waveform to be measured, the reference waveform needs to be interpolated. The interpolation parameters need to ensure that the length of the reference waveform is greater than three times the length of the waveform to be measured.

[0082] In an embodiment of the present invention, wheel defect monitoring is implemented based on the correspondence between a reference waveform and an original vehicle passing waveform after normalization, including: determining an output matrix array of a reference waveform array length and an original vehicle passing waveform array length after normalization according to a longest common subsequence algorithm; determining an output result based on a ratio relationship between the output matrix array and the original vehicle passing waveform array length after normalization; and implementing wheel defect monitoring based on comparison of the output result with a second threshold.

[0083] In the embodiment of the present invention, based on the extracted vehicle passing waveform ω nor And the reference waveform ω ref , the longest common subsequence method is used to determine the matrix array of the reference waveform array length and the original vehicle passing waveform array length after normalization. The algorithm has low computational complexity and can effectively improve the accuracy of detection by setting a dynamic threshold. In the embodiment of the present invention, the monitoring method is also specifically described:

[0084] In the embodiment of the present invention, it is agreed that ref The length of the array is n, ω nor The array length is m. Set the matrix L to n+1 rows, m+1 columns, and the initial value to 0. The algorithm is as follows:

[0085] 1for i=2:1:n+1

[0086] 2for j=2:1:m+1

[0087] 3ifω<20

[0088] nor

[0089] 4thre=20

[0090] 5 else

[0091] 6thre=5

[0092] 7 end if

[0093] 8 if abs(ω ref (i-1)-ω nor (j-1)) <thre

[0094] 9 L(i, j) = L(i+1, j+1) + 1

[0095] 10 else

[0096] 11 L(i, j) = L(i+1, j+1)

[0097] 12 end if

[0098] 13 if L(i-1, j)>=L(i, j)

[0099] 14 L(i, j) = L(i-1, j)

[0100] 15 end if

[0101] 16 if L(i, j-1)>=L(i, j)

[0102] 17 L(i, j) = L(i, j-1)

[0103] 18 end if

[0104] 19end for

[0105] 20end for

[0106] In the embodiment of the present invention, the 3rd to 7th lines of the algorithm are used for setting the judgment threshold. Since the vehicle passing waveform usually fluctuates greatly at the bottom of the waveform, the bottom of the waveform, i.e., ω nor <20, the judgment threshold is set larger, the tolerance is high, and misjudgment is avoided. nor When >=20, the judgment threshold is set smaller to avoid missed judgments;

[0107] Behaviors 8-12 of the algorithm are the judgment part, that is, when the difference between the reference waveform and the passing vehicle waveform is less than the threshold, the value of the corresponding matrix L will be +1;

[0108] Lines 13-18 of the algorithm prevent the current value of the matrix from being smaller than the values ​​above and to the left.

[0109] After determining the final matrix, take the L(n+1,m+1) value divided by m and multiply by 100 to get the output result D=100*L(n+1,m+1) / m; in the embodiment of the present invention, when D is less than 100, it is judged that the passing vehicle waveform is abnormal and the wheel has tread damage. When D is equal to 100, the wheel is normal. Figure 7 FIG. 1 shows a schematic diagram of implementing wheel defect monitoring in an embodiment of the present invention. Figure 7 After comparison, the first undulating section of the original waveform is a normal wheel, with an X value of 3742 and a Y value of 100; the second undulating section of the original waveform is a normal wheel, with an X value of 3790 and a Y value of 100; the third undulating section of the original waveform is a damaged wheel, with an X value of 3842 and a Y value of 92.16; the fourth undulating section of the original waveform is a normal wheel, with an X value of 3883 and a Y value of 100.

[0110] After the judgment is completed, the wavelength λ 5 To λ low With λ high , and clear the array ω raw .

[0111] In an embodiment of the present invention, a train wheel defect monitoring system without power supply at the trackside is also provided. Figure 8 The schematic diagram of the structure of the train wheel defect monitoring system without power supply at the trackside in the embodiment of the present invention is shown. Figure 8 The system includes a filtering unit, an extraction unit, a normalization unit and a monitoring unit, wherein the filtering unit is used to filter the original wavelength data to obtain the original smoothed wavelength data; the extraction unit is communicated with the filtering unit and is used to extract the original vehicle passing waveform according to the original smoothed wavelength data; the normalization unit is communicated with the extraction unit and is used to normalize the original vehicle passing waveform; the monitoring unit is communicated with the normalization unit and is used to realize wheel defect monitoring according to the correspondence between the reference waveform and the normalized original vehicle passing waveform.

[0112] Specifically, in an embodiment of the present invention, the extraction unit is used to extract the original vehicle passing waveform based on the original smoothed wavelength data, including: determining the waveform change between adjacent troughs in the original smoothed wavelength data; and extracting the original wavelength data of a single axle based on the relationship between the waveform change and the first threshold.

[0113] In an embodiment of the present invention, a specific description is also given of a process in which an extraction unit is used to extract an original vehicle passing waveform based on original smoothed wavelength data. The extraction unit is used to extract an original vehicle passing waveform based on original smoothed wavelength data, including: determining an initial array element, an initial peak value, and an initial trough value based on the collected original wavelength and the filtered initial wavelength; determining an initial state and a second state in sequence, alternately iterating the peak value and trough value of the two states based on the waveform change, and updating the array elements; when there is a wavelength value greater than the trough value after alternating iterations, adjusting the state to the initial state, and extracting the original wavelength data of a single axle based on the relationship between the waveform change in the updated array element and the first threshold.

[0114] Specifically, the trackside non-powered train wheel defect monitoring system of the present invention further includes a construction unit, and the construction unit is used to construct a reference waveform including:

[0115] Determine the length relationship between the original reference waveform and the normalized original vehicle passing waveform;

[0116] When the original reference waveform length does not meet the normalized original vehicle passing waveform length of a predetermined value multiple, the original reference waveform is interpolated until the reference waveform is greater than the normalized original vehicle passing waveform length of a predetermined value multiple.

[0117] In an embodiment of the present invention, a monitoring unit is used to implement wheel defect monitoring based on the correspondence between a reference waveform and an original vehicle passing waveform after normalization processing, including: determining an output matrix array of a reference waveform array length and an original vehicle passing waveform array length after normalization processing according to a longest common subsequence algorithm; determining an output result based on a ratio relationship between the output matrix array and the original vehicle passing waveform array length after normalization processing; and implementing wheel defect monitoring based on a comparison of the output result with a second threshold value.

[0118] In an embodiment of the present invention, an electronic device is further provided, the electronic device comprising at least one processor and at least one memory, the memory being data-connected to the processor, wherein:

[0119] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for monitoring wheel defects of trains without power supply at the trackside described in an embodiment of the present invention.

[0120] The method, system and device for monitoring wheel defects of non-electrical trackside trains of the present invention can realize fast and accurate wheel tread defect detection by describing the automatic recognition process of defect waveform in detail; it can realize online wheel health status monitoring without affecting train operation, and ensure train transportation safety and efficiency. The accuracy of train wheel data extraction and wheel defect detection is guaranteed by the algorithm.

[0121] Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent substitutions for some of the technical features therein; and these modifications or substitutions do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring wheel defects of a train without electricity on the trackside, characterized in that: The method comprises: Filtering the original wavelength data to obtain original smoothed wavelength data; Extracting the original vehicle passing waveform according to the original smoothed wavelength data; Normalizing the original vehicle passing waveform; Wheel defect monitoring is achieved based on the correspondence between the reference waveform and the normalized original vehicle passing waveform.

2. The method for monitoring wheel defects of a trackside non-powered train according to claim 1, characterized in that: Extracting the original vehicle passing waveform based on the original smoothed wavelength data includes: determining the amount of waveform change between adjacent troughs in the raw smoothed wavelength data; The original wavelength data of a single axle is extracted according to the relationship between the waveform change amount and the first threshold.

3. The method for monitoring wheel defects of a train without power supply at the trackside according to claim 2, characterized in that: Extracting the original vehicle passing waveform based on the original smoothed wavelength data includes: Determine the initial array element, the initial peak value and the initial trough value according to the collected original wavelength and the filtered initial wavelength; Determine the initial state and the second state in sequence, alternately iterate the peak value and the trough value of the two states according to the waveform change, and update the array elements; When there is a wavelength value greater than the trough value after the alternating iteration, the state is adjusted to the initial state, and the original wavelength data of a single axle is extracted according to the relationship between the waveform change amount in the updated array element and the first threshold.

4. The method for monitoring wheel defects of a trackside non-powered train according to claim 1, characterized in that: Building a reference waveform includes: Determine the length relationship between the original reference waveform and the normalized original vehicle passing waveform; When the original reference waveform length does not meet the normalized original vehicle passing waveform length of a predetermined value multiple, the original reference waveform is interpolated until the reference waveform is greater than the normalized original vehicle passing waveform length of a predetermined value multiple.

5. The method for monitoring wheel defects of a trackside non-powered train according to any one of claims 1 to 4, characterized in that: The wheel defect monitoring is realized according to the corresponding relationship between the reference waveform and the original vehicle passing waveform after normalization, including: Determine the output matrix array of the reference waveform array length and the original vehicle passing waveform array length after normalization according to the longest common subsequence algorithm; Determine the output result according to the ratio of the output matrix array to the normalized original vehicle passing waveform array length; Wheel defect monitoring is achieved based on comparison of the output result with the second threshold value.

6. The method for monitoring wheel defects of a train without power supply at the trackside according to claim 5, characterized in that: The matrix array for determining the reference waveform array length and the normalized original vehicle passing waveform array length according to the longest common subsequence algorithm includes: Determine the initial matrix according to the reference waveform array length and the normalized original vehicle passing waveform array length; Determining a third threshold of waveform tolerance according to waveform fluctuations; The initial matrix is ​​modified according to the relationship between the difference between the reference waveform and the original vehicle passing waveform after normalization and the third threshold value; Correct the current value of the initial matrix; Output the corrected matrix array.

7. A trackside non-electric train wheel defect monitoring system, characterized in that: The system comprises: A filtering unit, an extraction unit, a normalization unit and a monitoring unit, wherein: A filtering unit, used for filtering the original wavelength data to obtain original smoothed wavelength data; An extraction unit, which is in communication with the filtering unit and is used to extract the original vehicle passing waveform according to the original smoothed wavelength data; A normalization unit, connected to the extraction unit for normalization processing of the original vehicle passing waveform; The monitoring unit is connected to the normalization unit for communication and is used to realize wheel defect monitoring according to the corresponding relationship between the reference waveform and the original vehicle passing waveform after normalization processing.

8. The trackside non-powered train wheel defect monitoring system according to claim 7, characterized in that: The extraction unit is used to extract the original vehicle passing waveform according to the original smoothed wavelength data, including: determining the amount of waveform change between adjacent troughs in the raw smoothed wavelength data; The original wavelength data of a single axle is extracted according to the relationship between the waveform change amount and the first threshold.

9. The trackside non-powered train wheel defect monitoring system according to claim 8, characterized in that: The extraction unit is used to extract the original vehicle passing waveform according to the original smoothed wavelength data, including: Determine the initial array element, the initial peak value and the initial trough value according to the collected original wavelength and the filtered initial wavelength; Determine the initial state and the second state in sequence, alternately iterate the peak value and the trough value of the two states according to the waveform change, and update the array elements; When there is a wavelength value greater than the trough value after the alternating iteration, the state is adjusted to the initial state, and the original wavelength data of a single axle is extracted according to the relationship between the waveform change amount in the updated array element and the first threshold.

10. The trackside non-powered train wheel defect monitoring system according to claim 7, characterized in that: The system further comprises a construction unit, wherein the construction unit is used to construct a reference waveform comprising: Determine the length relationship between the original reference waveform and the normalized original vehicle passing waveform; When the original reference waveform length does not meet the normalized original vehicle passing waveform length of a predetermined value multiple, the original reference waveform is interpolated until the reference waveform is greater than the normalized original vehicle passing waveform length of a predetermined value multiple.

11. The trackside non-powered train wheel defect monitoring system according to any one of claims 7 to 10, characterized in that: The monitoring unit is used to realize wheel defect monitoring according to the corresponding relationship between the reference waveform and the original vehicle passing waveform after normalization, including: Determine the output matrix array of the reference waveform array length and the original vehicle passing waveform array length after normalization according to the longest common subsequence algorithm; Determine the output result according to the ratio of the output matrix array to the normalized original vehicle passing waveform array length; Wheel defect monitoring is achieved based on comparison of the output result with the second threshold value.

12. An electronic device, comprising at least one processor and at least one memory, wherein the memory is data-connected to the processor, wherein: The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 6.

Images