A wheel tread polygon fault diagnosis method and system

Abstract

The application provides a wheel tread polygon fault diagnosis method and system, the method comprises the following steps: obtaining the vibration acceleration of the wheel axle box position; performing frequency spectrum analysis on the single sample data of the vibration acceleration to obtain a vibration frequency spectrum; extracting a polygon vibration feature of the vibration frequency spectrum; constructing a polygon fault quantitative model according to the polygon vibration feature; and realizing polygon fault diagnosis and outputting alarm information according to the polygon fault quantitative model, a preset threshold value and a preset rule. Compared with the static detection in the prior art, the application can realize fault diagnosis and alarm in a driving state; compared with dynamic detection, the application constructs a fault quantitative model through processing of multiple single sample data, realizes fault diagnosis through the fault quantitative model and related rules, and the overall scheme can effectively improve the recognition rate of the tread polygon fault, so that the polygon fault can be accurately diagnosed and timely alarmed. The system has the same beneficial effects.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis technology, and in particular to a method and system for diagnosing polygonal faults in wheel treads. Background Technology

[0002] Tread polygonal deformation is a common wheel defect caused by long-term wear between the wheel and the rail. This defect triggers a series of dynamic responses in the wheel-rail system, significantly impacting driving safety. Existing detection methods are mainly divided into static detection and dynamic detection.

[0003] Currently, static detection involves using sensors to directly measure the tread condition of a single wheel when it is stationary. This method is direct, but inefficient, time-consuming, and cannot be used for monitoring while the vehicle is in motion.

[0004] Dynamic detection collects data such as vehicle acceleration and displacement, and monitors the condition of wheel treads through feature extraction. It has a wide range of applications. However, the existing technology still has a low recognition rate for polygonal faults on the tread.

[0005] Therefore, providing a wheel tread polygon fault diagnosis method and system that can effectively improve the recognition rate of polygon faults, so as to accurately diagnose polygon faults and output graded alarm information, is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for diagnosing polygonal faults in wheel treads. This method is logically clear, safe, effective, reliable and easy to operate. It can effectively improve the recognition rate of polygonal faults in wheel treads, so as to accurately diagnose polygonal faults and output graded alarm information.

[0007] Based on the above objectives, the technical solution provided by the present invention is as follows:

[0008] A method for diagnosing polygonal faults in wheel treads includes the following steps:

[0009] Obtain the vibration acceleration at the position of the wheel axle box;

[0010] The vibration spectrum is obtained by performing spectral analysis on the single sample data of the vibration acceleration.

[0011] Extract the polygonal vibration features from the vibration spectrum;

[0012] A quantitative model of polygonal faults is constructed based on the polygonal vibration characteristics.

[0013] Based on the polygonal fault quantitative model, preset threshold values, and preset rules, polygonal fault diagnosis and alarm information output are achieved.

[0014] Preferably, the extraction of polygonal vibration features from the vibration spectrum includes the following steps:

[0015] Obtain the frequency corresponding to the maximum amplitude of the vibration spectrum in the vibration spectrum;

[0016] Determine whether the fault frequency meets the preset conditions;

[0017] If so, then obtain the polygon order and corresponding polygon features in the vibration spectrum.

[0018] Preferably, the preset conditions are as follows:

[0019] m = mod(f0 / f n )

[0020] When m = 0, the polygon order J i =f0 / f n ;

[0021] If m≠0, then the polygon order J i =0;

[0022] Where mod is the modulo function, f0 is the frequency corresponding to the maximum amplitude of the vibration spectrum, and f n The rotational frequency of the wheels.

[0023] Preferably, the step of constructing a polygonal fault quantitative model based on the polygonal vibration characteristics includes the following steps:

[0024] The polygon order magnitude is obtained based on the polygon order and the corresponding polygon features.

[0025] The polygon order magnitude is weighted to obtain the weighted polygon order magnitude;

[0026] The polygonal fault quantitative model is constructed based on the weighted polygonal order magnitude and the preset wheel speed.

[0027] Preferably, the formula for calculating single-sample data used to construct the polygonal fault quantitative model is as follows:

[0028]

[0029] in, , where Q is the fault value, R is the weighting coefficient, R is the amplitude of the polygon order, and N is the wheel speed.

[0030] Preferably, the step of realizing polygonal fault diagnosis and outputting alarm information based on the polygonal fault quantitative model, preset threshold value, and preset rules includes the following steps:

[0031] Construct an m-dimensional order statistical array according to the preset rules based on the polygon order;

[0032] Based on the polygonal fault quantitative model and the polygonal order, an m-dimensional fault value statistical array is constructed according to the preset rules.

[0033] The fault value is classified into alarm levels according to the preset threshold value;

[0034] According to preset rules, the continuity of the m-dimensional order statistical array and the continuity of the m-dimensional fault value statistical array are cyclically counted respectively, and polygonal fault diagnosis and alarm information are output according to the alarm classification.

[0035] The preset rule is specifically the first-in, first-out (FIFO) principle.

[0036] Preferably, the step of classifying the fault value into alarm levels based on the preset threshold value specifically involves:

[0037] The preset threshold value is Alarm = (Alarm1, Alarm2, Alarm3);

[0038] Define the fault quantity value When it is time to issue a warning, when When it indicates a Level 1 alarm, This indicates a level two alarm.

[0039] Preferably, the step of cyclically calculating the continuity of the m-dimensional order statistical array according to a preset rule includes the following steps:

[0040] Let the order error be ΔJ, the sample cycle factor be K, the counting factor be C, and the sample continuity factor be H;

[0041] Let K = 1, C = 1, K ≤ m;

[0042] Determine the order J of a polygon K Does J satisfy? K =J K-1 ±ΔJ;

[0043] If so, then C = C + 1, K = K + 1;

[0044] If not, then C = 1, K = K + 1;

[0045] When C = H, the continuity cutoff position P = K is obtained, and the above loop ends. The loop continues to perform the continuous statistics of the next single sample data in the m-dimensional order statistical array.

[0046] When K > m and C < H, the m-dimensional order statistical array does not meet the preset order continuity condition. Therefore, according to the first-in-first-out principle, the cycle continues to count the continuity of the next m-dimensional order statistical array.

[0047] Preferably, the step of cyclically statistically analyzing the continuity of the m-dimensional fault value statistical array according to preset rules, and realizing polygonal fault diagnosis and outputting alarm information according to the alarm classification, includes the following steps:

[0048] Define the alarm grading cycle factor as Y (Y≤3), the count factor as C′, the sample cycle factor as K′ (K′≤P), and the continuity cutoff position of the m-dimensional order statistical array as P;

[0049] Let C′ = 0, K′ = P - H + 1, Y = 3;

[0050] Determine the fault value Does it meet the requirements?

[0051] If so, then C′=C′+1, K′=K′+1;

[0052] If not, then C′=0, K′=K′+1;

[0053] When C′=H, the alarm information corresponding to the alarm classification cycle factor Y is output;

[0054] When C′≠H, Y=Y-1, then continue to perform the loop to count the continuity of the next single sample data in the m-dimensional fault value statistical array;

[0055] When Y = 1 and C′ < H, the m-dimensional fault value statistics array does not meet the preset fault value continuity condition. Then, according to the first-in-first-out principle, the cycle continues to count the continuity of the next m-dimensional fault value statistics array.

[0056] A wheel tread polygon fault diagnosis system includes:

[0057] The acquisition module is used to obtain the vibration acceleration at the position of the wheel axle box;

[0058] The spectrum analysis module is used to perform spectrum analysis on the single sample data of the vibration acceleration to obtain the vibration spectrum;

[0059] The feature extraction module is used to extract polygonal vibration features from the vibration spectrum;

[0060] The model building module is used to build a polygonal fault quantitative model based on the polygonal vibration characteristics.

[0061] The fault diagnosis and alarm module is used to perform polygon fault diagnosis and output alarm information based on the polygonal fault quantitative model, preset threshold value and preset rules.

[0062] The method for diagnosing polygonal faults in wheel treads provided by this invention obtains vibration acceleration by detecting the position of the wheel axle box; performs spectral analysis on the single sample data of the obtained vibration acceleration to obtain a vibration spectrum; extracts polygonal vibration features from the vibration spectrum; constructs a polygonal fault quantitative model based on the extracted polygonal vibration features; and realizes polygonal fault diagnosis and outputs alarm information through a polygonal fault quantitative module, preset threshold values, and preset rules.

[0063] This invention extracts the vibration characteristics of polygonal tread surfaces after spectral analysis of vibration acceleration data. Based on these characteristics, a quantitative model is established. This quantitative model, along with relevant rules, enables accurate diagnosis and alarm of polygonal tread surface faults. Compared to static detection in existing technologies, this invention can achieve fault diagnosis and alarm while the vehicle is in motion. Compared to dynamic detection, this invention processes multiple single-sample data to construct a quantitative fault model. Fault diagnosis is then achieved through this quantitative model and relevant rules. The overall solution effectively improves the recognition rate of polygonal tread surface faults, enabling accurate diagnosis and timely alarm.

[0064] The present invention also discloses a wheel tread polygon fault diagnosis system. Since the diagnosis system and the diagnosis method solve the same technical problem and belong to the same technical concept, they should have the same beneficial effects, and will not be described in detail here. Attached Figure Description

[0065] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0066] Figure 1 A flowchart of a method for diagnosing polygonal faults in wheel treads provided in an embodiment of the present invention;

[0067] Figure 2 A flowchart of step S3 provided in an embodiment of the present invention;

[0068] Figure 3 A flowchart of step S4 provided in an embodiment of the present invention;

[0069] Figure 4 A flowchart of step S5 provided in an embodiment of the present invention;

[0070] Figure 5 This is a schematic diagram of a polygonal fault diagnosis system for wheel tread provided in an embodiment of the present invention. Detailed Implementation

[0071] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0072] The embodiments of this invention are written in a progressive manner.

[0073] This invention provides a method and system for diagnosing polygonal faults in wheel treads. It primarily addresses the technical problems in existing technologies where static detection cannot diagnose faults while the vehicle is in motion, and dynamic detection has low accuracy in recognizing polygonal faults in wheel treads.

[0074] like Figure 1 As shown, a method for diagnosing polygonal faults in wheel treads includes the following steps:

[0075] S1. Obtain the vibration acceleration at the position of the wheel axle box;

[0076] S2. Perform spectral analysis on single-sample data of vibration acceleration to obtain the vibration spectrum;

[0077] S3. Extract polygonal vibration features from the vibration spectrum;

[0078] S4. Construct a polygonal fault quantitative model based on polygonal vibration characteristics;

[0079] S5. Based on the polygonal fault quantitative model, preset threshold values, and preset rules, realize polygonal fault diagnosis and output alarm information.

[0080] It should be noted that the above diagnostic method can be divided into two parts depending on the purpose. The first part is single sample identification, which mainly includes steps S1 to S4, and the purpose is to construct a polygonal fault quantitative model. The second part is fault diagnosis and alarm, which mainly includes step S5, and the purpose is to realize polygonal fault diagnosis and alarm of wheel tread through the polygonal fault quantitative model.

[0081] In step S1, vibration acceleration is collected in real time by a vibration sensor installed at the wheel axle box.

[0082] In step S2, spectral analysis is performed on the single sample data of vibration acceleration to obtain the vibration spectrum;

[0083] In this embodiment, the i-th single sample data V(n) is selected from the vibration acceleration; the wheel rotation frequency f is obtained through a preset sampling time period. n0 (f n0 >0), the vibration spectrum H(f) is obtained by FFT calculation of V(n);

[0084] It should be noted that the method of transforming a time-domain signal to the frequency domain for analysis is called spectrum analysis. The purpose of spectrum analysis is to decompose a complex time-history waveform into several individual harmonic components through Fourier transform in order to obtain the frequency structure of the signal and the information of each harmonic and phase.

[0085] In step S3, polygonal vibration features are extracted from the obtained vibration frequencies;

[0086] It should be noted that when the health status of critical components changes, the frequency components in the sample data signal spectrum will change accordingly. Therefore, the signal spectrum information can be accurately characterized by analyzing the frequency domain characteristics of the vibration signal, and then the health status of critical components under different operating conditions can be obtained.

[0087] In step S4, a polygonal fault quantitative model is constructed using polygonal vibration features;

[0088] In step S5, polygonal fault diagnosis is achieved through a polygonal fault quantitative model and preset rules, and alarm information is output through preset threshold values.

[0089] like Figure 2 As shown, preferably, step S3 includes the following steps:

[0090] A1. Obtain the frequency corresponding to the maximum amplitude of the vibration spectrum in the vibration spectrum;

[0091] A2. Determine whether the fault frequency meets the preset conditions;

[0092] A3. If so, obtain the polygon order and corresponding polygon features in the vibration spectrum.

[0093] Preferably, the preset conditions are as follows:

[0094] m = mod(f0 / f n )

[0095] When m = 0, the polygon order J i =f0 / f n ;

[0096] If m≠0, then the polygon order J i =0;

[0097] Where mod is the modulo function, f0 is the frequency corresponding to the maximum amplitude of the vibration spectrum, and f n The rotational frequency of the wheels.

[0098] In step A1, the maximum value (i.e. the maximum amplitude of the vibration spectrum) H(f0) in the spectrum H(f) and its corresponding frequency f0 (f0≠0) are obtained;

[0099] In steps A2 and B3, using the remainder formula described above, determine whether the frequency f0 is the wheel rotation frequency f. n When the remainder is 0, the polygonal order in the vibration spectrum is specifically the frequency f0 and the rotational frequency f. n The quotient value; when the remainder is not 0, the polygon order in the vibration spectrum is specifically 0; based on the obtained polygon order, the corresponding polygon frequency is extracted from the vibration spectrum as polygon feature.

[0100] like Figure 3 As shown, preferably, step S4 includes the following steps:

[0101] B1. Obtain the polygon order magnitude based on the polygon order and the corresponding polygon features;

[0102] B2. Weight the polygon order magnitude to obtain the weighted polygon order magnitude;

[0103] B3. Construct a polygonal fault quantitative model based on the weighted polygon order magnitude and the preset wheel speed.

[0104] Preferably, the formula for calculating single-sample data used to construct the polygonal fault quantitative model is as follows:

[0105]

[0106] in, , where Q is the fault value, R is the weighting coefficient, R is the amplitude of the polygon order, and N is the wheel speed.

[0107] In step B1, based on the obtained polygon order and the corresponding polygon frequency, the amplitude of the polygon order is obtained, that is, the time domain index corresponding to the frequency of the i-th polygon order.

[0108] In step B2, by using preset weighting coefficients, weights are added to the time-domain index corresponding to the i-th polygon order frequency to obtain the weighted polygon order amplitude.

[0109] In step B3, a quantitative model of component faults is generated based on the single-sample data calculation formula, the weighted polygon order amplitude, and the preset wheel speed (which can also be detected by sensors).

[0110] It should be noted that the formula for calculating decibels (dB) in fault values ​​is generally divided into two types: 1. 10log(M); 2. 20log(M); when characterizing the amplitude of a single variable (R in the above formula), the parameter is selected as 20, and the wheel speed in the above formula is a fixed value.

[0111] like Figure 4 As shown, preferably, step S5 includes the following steps:

[0112] C1. Construct an m-dimensional order statistical array according to preset rules based on the polygon order;

[0113] C2. Construct an m-dimensional fault quantity statistical array according to preset rules based on the polygonal fault quantitative model and polygonal order;

[0114] C3. Classify alarm levels for fault values ​​based on preset threshold values;

[0115] C4. According to preset rules, the continuity of the m-dimensional order statistical array and the continuity of the m-dimensional fault value statistical array are cyclically counted respectively, and according to the alarm level, polygonal fault diagnosis and alarm information output are realized.

[0116] The specific preset rule is: First-In-First-Out (FIFO) principle.

[0117] In step C1, based on the obtained polygon orders and following the first-in-first-out principle, an m-dimensional order statistical array J = (J0, J1, ..., J...) is constructed. i , ..., J m-1 ), J i This represents the order value of the i-th single sample.

[0118] In step C2, based on the polygon order and the polygon fault quantification model, an m-dimensional fault value statistical array is constructed according to the first-in-first-out principle. This represents the fault value of the i-th single sample.

[0119] In step C3, a threshold value is set for the i-th single-sample fault value. Alarm classification;

[0120] In step C4, first, the continuity of the m-dimensional order statistical array J is counted in a loop, and then the continuity of the m-dimensional fault quantity statistical array V is counted in a loop. dB Based on the first-in-first-out principle and corresponding conditions and alarm levels, polygonal fault diagnosis and alarm information output are achieved.

[0121] Preferably, step C3 specifically includes:

[0122] The preset threshold value is Alarm = (Alarm1, Alarm2, Alarm3);

[0123] Define the fault quantity value When it is time to issue a warning, when When it indicates a Level 1 alarm, This indicates a level two alarm.

[0124] In practical applications, fault values ​​are defined by setting a threshold matrix Alarm = (Alarm1, Alarm2, Alarm3). When the first threshold is exceeded, a warning is issued, indicating the fault value. Exceeding the second threshold indicates a Level 1 alarm, with the fault value... When the third threshold is exceeded, a level two alarm is triggered. The threshold matrix can be set according to actual needs, and the corresponding alarm level can be further set by setting the Xth threshold.

[0125] Preferably, step C4 involves cyclically calculating the continuity of the m-dimensional statistical array, which includes the following steps:

[0126] D1. Define the order error as Δ, the sample cycle factor as K, the counting factor as C, and the sample continuity factor as H;

[0127] D2. Let K = 1, C = 1, K ≤ m;

[0128] D3. Determine the order of a polygon J K Does J satisfy? K =J K-1 ±ΔJ;

[0129] D41. If so, then C = C + 1, K = K + 1;

[0130] D42. If not, then C = 1, K = K + 1;

[0131] D5. When C = H, the continuity cutoff position P = K is obtained, and the above loop ends. Then, the loop continues to perform continuous statistics on the continuity of the next single sample data in the m-dimensional order statistical array.

[0132] D6. When K > m and C < H, the m-dimensional order statistical array does not meet the preset order continuity condition. Therefore, the loop continues to calculate the continuity of the next m-dimensional order statistical array according to the first-in-first-out principle.

[0133] In step D1, the order error, sample cycle factor, counting factor, and sample continuity factor are defined respectively;

[0134] In step D2, the continuous cyclic statistics of the m-dimensional order statistical array are initialized, that is, the sample cyclic factor is set to 1 and the count factor is set to 1. The range of the sample cyclic factor does not exceed the established m-dimensional dimension value m.

[0135] In step D3, determine the order J of the polygon. K Does it satisfy formula J? K =J K-1 ±ΔJ, that is, from J=(J0, J1, ..., J i , ..., J m-1 Starting with J1 in J, each single sample data in J is judged separately;

[0136] In steps D41 and D42, if a single sample data satisfies the above formula, the counting factor is incremented by 1 and the cyclic sample factor is incremented by 1; if a single sample data does not satisfy the above formula, the counting factor is set to 1 (i.e., the counting factor is initialized) and the cyclic sample factor is incremented by 1.

[0137] In step D5, when the counting factor equals the sample continuity factor, it is determined that the m-dimensional order statistical array is continuous. The value of the sample cycle factor at this time is obtained, and this value is determined to be the continuity cutoff position P of the m-dimensional order statistical array. At the same time, the above cycle ends, and the cycle continues to count the continuity of the next single sample data of the m-dimensional order statistical array.

[0138] In step D6, when the sample cyclic factor is greater than the dimension value m and the counting factor is less than the sample continuous factor, that is, when the current m-dimensional order statistical array J = (J0, J1, ..., J...) is... i , ..., J m-1 If the above continuity condition is not met, then the loop continues to count the next m-dimensional order statistical array J = J1, ..., J according to the first-in-first-out principle. i , ..., J m-1 J m The continuity of ).

[0139] In this embodiment, the continuity of the m-dimensional (let m = 10)-dimensional statistical array is cyclically counted. The sample continuity factor H = 7 is set. That is, if 7 consecutive samples satisfy the above formula, then the 10-dimensional statistical array is considered to be continuous.

[0140] The current 10-dimensional order statistical array J is cyclically statistically analyzed according to steps E2 to E6, specifically as follows:

[0141] Let the counting factor be equal to 1 and the sample cycle factor be equal to 1. At this time, the value of the sample cycle factor is a positive integer between 1 and 10.

[0142] Determine if J1 satisfies J1=J0±ΔJ. If it does, the count factor is equal to 2 (if not, the count factor is equal to 1), and the cyclic sample factor is equal to 2. Continue to determine if J2 satisfies J2=J1±ΔJ. If it does, the count factor is equal to 3 (if not, the count factor is equal to 1), and the cyclic sample factor is equal to 3. Continue to determine if J3 satisfies J3=J2±ΔJ. If it does, the count factor is equal to 4 (if not, the count factor is equal to 1), and the sample cyclic factor is equal to 4. Repeat the above calculations multiple times. When the value of the count factor equals the value of the sample continuity factor, record the continuity cutoff position of the 10-dimensional order statistical array as P=K=8, that is, the 7 samples before the 8th single sample order value are continuous, and end the above loop. Initialize the count factor and the sample cyclic factor, that is, set the count factor to 1 and the sample cyclic factor to 1, and then perform the continuity judgment of J9.

[0143] When the sample cycle factor is greater than the dimension value and the count factor is less than the sample continuity factor, that is, when J 11 If the counting factor (or K takes a value greater than 10) is less than 7, then the current 10-dimensional order statistical array J = (J0, J1, ..., J...) is considered to be... i If the above continuity condition is not met, then the cyclical statistics of the next 10-dimensional order statistical array J = (J1, ..., J9) will continue according to the first-in-first-out principle. i ..., J9, J 10 The continuity of ).

[0144] It should be noted that the above embodiments are only examples illustrating the specific implementation of the continuity of the m-dimensional order statistical array in a loop. The values ​​and dimension values ​​defined in step D1 can be set according to actual needs.

[0145] Preferably, step C4 involves cyclically statistically analyzing the continuity of the m-dimensional fault value statistical array and, based on the alarm classification, realizing polygonal fault diagnosis and outputting alarm information, including the following steps:

[0146] E1. Define the alarm grading cycle factor as Y (Y≤3), the count factor as C′, the sample cycle factor as K′ (K′≤P), and the continuity cutoff position of the m-dimensional order statistical array as P;

[0147] E2. Let C′=0, K′=P-H+1, Y=3;

[0148] E3. Determine the fault value V dBk′ Does it satisfy V? dBk′ ≥Alarm(Y);

[0149] E41. If so, then C′=C′+1, K′=K′+1;

[0150] E42. If not, then C′=0, K′=Kv+1;

[0151] E5. When C′=H, the alarm information corresponding to the alarm classification cycle factor Y is output;

[0152] E6. When C′≠H, Y=Y-1, then continue to perform loop statistics on the continuity of the next single sample data in the m-dimensional fault value statistical array;

[0153] E7. When Y = 1 and C′ < H, the m-dimensional fault value statistics array does not meet the preset fault value continuity condition. Then, according to the first-in-first-out principle, the loop continues to count the continuity of the next m-dimensional fault value statistics array.

[0154] In step E1, the alarm grading cycle factor (the value range of Y corresponds to the grading number in step C3, i.e., Y≤3), the counting factor, and the sample cycle factor are defined respectively, and the continuity cutoff position P of the m-dimensional order statistical array is obtained; the value of the sample continuity factor H is the same as the set value in the continuity of the m-dimensional order statistical array of the cycle statistics.

[0155] Step E2: Initialize the continuous cyclic statistics of the m-dimensional fault value statistical array, that is, set the counting factor to 0, the sample cyclic factor to the difference between the value of the continuous cutoff position of the m-dimensional order statistical array and the sample continuous factor plus 1, and the alarm classification cyclic factor to 3.

[0156] In step E3, the fault quantity value is determined. Does it satisfy the formula? That is, from In Begin by examining V separately. dB Judge each single sample data in the data;

[0157] In steps E41 and E42, if a single sample data satisfies the above formula, the counting factor is incremented by 1 and the cyclic sample factor is incremented by 1; if a single sample data does not satisfy the above formula, the counting factor is set to 0 (i.e., the counting factor is initialized) and the cyclic sample factor is incremented by 1.

[0158] In step E5, when the counting factor equals the sample continuity factor, the value of the alarm classification cycle factor Y is obtained at this time, and the corresponding alarm information is output according to the definition in step C3.

[0159] In step E6, when the counting factor is not equal to the sample continuity factor, the alarm grading loop factor is decremented by 1, and the continuity of the next single sample data in the m-dimensional fault value statistics array continues.

[0160] In step E7, when the alarm classification cycle factor is equal to 1 and the counting factor is less than the sample continuous factor, that is, when the current m-dimensional fault value statistical array is... If the above continuity condition is not met, the next m-dimensional fault value statistical array will continue to be calculated according to the first-in-first-out principle. The continuity.

[0161] In this embodiment, the continuity of the m-dimensional fault value statistical array is cyclically counted (let m = 10), and the alarm classification cyclic factor Y is set to a maximum of 3 levels, that is, the alarm classification is three levels: early warning, first-level alarm and second-level alarm.

[0162] The current 10-dimensional fault value statistical array V dB Perform cyclical statistics from steps E2 to E7, specifically as follows:

[0163] Let the counting factor be 0, and the sample cycle factor K′ be equal to the difference between the continuity cutoff position P of the m-dimensional order statistical array and the sample continuity factor H plus 1, that is, K′=8-7+1=2, Y=3;

[0164] Determine fault values Does it satisfy the formula? If the condition is met, the count factor equals 1 (if not, the count factor equals 0), and the sample cycle factor equals 3; continue to determine the fault quantity value. Does it satisfy the formula? If the condition is met, the count factor equals 2 (if not, the count factor equals 0), and the sample cycle factor equals 4; continue to determine the fault value. Does it satisfy the formula? If the condition is met, the counting factor is equal to 3 (if not, the counting factor is equal to 0), and the sample cycle factor is equal to 5. The above rules are used to perform multiple cyclic calculations. When the value of the counting factor is equal to the value of the sample continuity factor, that is, C′=H=7, which means that the fault quantities of the 7 single samples are all greater than or equal to the third threshold, a level 2 alarm is output.

[0165] When C′≠H≠7, that is, when one of the seven single-sample fault quantities is less than the third threshold, the alarm grading cycle factor Y is reduced by 1, i.e., Y = 2; assuming The formula is not satisfied. Then continue the loop to check. Does it meet the requirements? That is, from Begin determining whether the continuity exceeds the second threshold;

[0166] When the alarm grading cycle factor Y = 1, and the value of the count factor is less than the value of the sample continuity factor, i.e., assuming Not satisfied That is, continue the loop to judge. Does it meet the requirements? At this point, the alarm grading cycle factor Y = 1, and the counting factor is less than 7, so the current 10-dimensional fault value statistical array is considered to be... If the above continuity condition is not met, the loop continues to count the next 10-dimensional statistical array according to the first-in-first-out principle. The continuity.

[0167] like Figure 5 As shown, a polygonal fault diagnosis system for wheel treads includes:

[0168] The acquisition module is used to obtain the vibration acceleration at the position of the wheel axle box;

[0169] The spectrum analysis module is used to perform spectrum analysis on single sample data of vibration acceleration to obtain the vibration spectrum;

[0170] The feature extraction module is used to extract polygonal vibration features from the vibration spectrum;

[0171] The model building module is used to construct a polygonal fault quantitative model based on polygonal vibration characteristics.

[0172] The fault diagnosis and alarm module is used to diagnose polygonal faults and output alarm information based on the polygonal fault quantitative model, preset threshold values ​​and preset rules.

[0173] In practical application, the acquisition module obtains the vibration acceleration at the wheel axle box position and sends the vibration acceleration to the spectrum analysis module; the spectrum analysis module performs spectrum analysis on the single sample data of vibration acceleration to obtain the vibration spectrum and sends the vibration spectrum to the feature extraction module; the feature extraction module extracts polygonal vibration features from the vibration spectrum and sends the polygonal vibration features to the model building module; the model building module constructs a polygonal fault quantitative model based on the polygonal vibration features and sends the polygonal fault quantitative model to the fault diagnosis and alarm module; the fault diagnosis and alarm module, based on the polygonal fault quantitative model, preset threshold values, and preset rules, realizes polygonal fault diagnosis and outputs alarm information.

[0174] In the embodiments provided in this application, it should be understood that the disclosed methods and systems can be implemented in other ways. The system embodiments described above are merely illustrative. For example, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple modules or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, and can be electrical, mechanical, or other forms.

[0175] Furthermore, in the various embodiments of the present invention, each functional module can be fully integrated into a processor, or each module can be a separate device, or two or more modules can be integrated into a device; each functional module in the various embodiments of the present invention can be implemented in hardware or in the form of hardware plus software functional units.

[0176] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by program instructions and related hardware. The aforementioned program instructions can be stored in a computer-readable storage medium. When the program instructions are executed, they perform the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0177] It should be understood that the use of terms such as "system," "device," "unit," and / or "module" in this application is merely one method of distinguishing different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0178] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "a," and / or "the" are not specifically singular and may include the plural. Generally, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements. An element defined by the phrase "comprising an..." does not exclude the presence of other identical elements in the process, method, product, or apparatus that includes the element.

[0179] In the description of the embodiments of this application, unless otherwise stated, " / " means "or", for example, A / B can mean A or B; "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the description of the embodiments of this application, "multiple" refers to two or more.

[0180] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0181] If a flowchart is used in this application, it is used to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0182] The foregoing has provided a detailed description of a method and system for diagnosing polygonal faults in wheel treads provided by the present invention. The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for diagnosing polygonal faults in wheel treads, characterized in that, Includes the following steps: Obtain the vibration acceleration at the position of the wheel axle box; The vibration spectrum is obtained by performing spectral analysis on the single sample data of the vibration acceleration. Extract the polygonal vibration features from the vibration spectrum; A quantitative model of polygonal faults is constructed based on the polygonal vibration characteristics. Based on the aforementioned polygonal fault quantitative model, preset threshold values, and preset rules, polygonal fault diagnosis and alarm information output are achieved, specifically including the following steps: Constructed according to the polygon order and the preset rules m Dimensional order statistical array; Based on the polygonal fault quantification model and the polygonal order, a system is constructed according to the preset rules. m Dimensional fault quantity statistics array; The fault value is classified into alarm levels according to the preset threshold value; According to the preset rules, the statistics are cyclically counted one by one. m The continuity of the dimensional order statistical array and the aforementioned m The continuity of the fault quantity statistics array is determined, and based on the alarm classification, polygonal fault diagnosis and alarm information output are realized; Specifically, the preset rule is: First-In-First-Out (FIFO) principle; The extraction of polygonal vibration features from the vibration spectrum includes the following steps: Obtain the frequency corresponding to the maximum amplitude of the vibration spectrum in the vibration spectrum; Determine whether the frequency meets the preset conditions; If so, then obtain the polygon order and corresponding polygon features in the vibration spectrum; The preset conditions are specifically as follows: ； when When, then the polygon order ; like When the order is 0, the polygon order is... ; in, It is a modulo function. The frequency corresponding to the maximum amplitude of the vibration spectrum in the vibration spectrum. The rotational frequency of the wheels.

2. The method for diagnosing polygonal wheel tread faults as described in claim 1, characterized in that, The step of constructing a polygonal fault quantitative model based on the polygonal vibration characteristics includes the following steps: The polygon order magnitude is obtained based on the polygon order and the corresponding polygon features. The polygon order magnitude is weighted to obtain the weighted polygon order magnitude; The polygonal fault quantitative model is constructed based on the weighted polygonal order magnitude and the preset wheel speed.

3. The method for diagnosing polygonal wheel tread faults as described in claim 2, characterized in that, The specific formula for calculating single-sample data used to construct the polygonal fault quantitative model is as follows: ； in, This is the fault value. These are weighting coefficients. The magnitude of the polygon order. This refers to the wheel rotation speed.

4. The method for diagnosing polygonal wheel tread faults as described in claim 3, characterized in that, The alarm classification of the fault value based on the preset threshold value is specifically as follows: The preset threshold value is ; Define the fault quantity value When it is time to issue a warning, when When it indicates a Level 1 alarm, when This indicates a level two alarm.

5. The method for diagnosing polygonal wheel tread faults as described in claim 4, characterized in that, The statistics are cyclically calculated according to preset rules. m The continuity of a dimensional array is determined by the following steps: Define the order error as The sample cycle factor is The counting factor is The sample continuity factor is ; make ; Determine the order of a polygon Does it meet the requirements? ; If so, then ; If not, then ; when When the continuity cutoff position is obtained, the continuity cutoff position is then determined. And end the above loop, and continue to execute the loop statistics. m The continuity of the next single sample data in a dimensional statistical array; when ,and When, then the stated m If the dimensional order statistical array does not meet the preset order continuity condition, then the loop continues to perform the next cyclic statistical operation according to the first-in-first-out principle. Continuity of a dimensional statistical array.

6. The method for diagnosing polygonal wheel tread faults as described in claim 5, characterized in that, The statistics are cyclically calculated according to preset rules. m The continuity of the fault value statistics array is determined, and based on the alarm classification, polygonal fault diagnosis and alarm information output are implemented, including the following steps: Define the alarm grading cycle factor as , The counting factor is The sample cycle factor is , , m The continuity cutoff position of the dimensional statistical array is ; make =0, , =3; Determine the fault value Does it meet the requirements? If so, then ， ; If not, then =0 ， ; when When this happens, the output will be the alarm grading cycle factor. Corresponding alarm information; when hour, Then continue executing the loop to count the above. m The continuity of the next single sample data in the fault quantity statistics array; when ,and At that time, the m If the fault value statistics array does not meet the preset fault value continuity condition, then the loop continues to count the next fault value according to the first-in-first-out principle. m The continuity of the fault quantity statistics array.

7. A polygonal fault diagnosis system for wheel treads, characterized in that, include: The acquisition module is used to obtain the vibration acceleration at the position of the wheel axle box; The spectrum analysis module is used to perform spectrum analysis on the single sample data of the vibration acceleration to obtain the vibration spectrum; The feature extraction module is used to extract polygonal vibration features from the vibration spectrum; The model building module is used to build a polygonal fault quantitative model based on the polygonal vibration characteristics. The fault diagnosis and alarm module is used to perform polygon fault diagnosis and output alarm information based on the polygonal fault quantitative model, preset threshold values, and preset rules. Specifically, it is used for: Constructed according to the polygon order and the preset rules m Dimensional order statistical array; Based on the polygonal fault quantification model and the polygonal order, a system is constructed according to the preset rules. m Dimensional fault quantity statistics array; The fault value is classified into alarm levels according to the preset threshold value; According to the preset rules, the statistics are cyclically counted one by one. m The continuity of the dimensional order statistical array and the aforementioned m The continuity of the fault quantity statistics array is determined, and based on the alarm classification, polygonal fault diagnosis and alarm information output are realized; Specifically, the preset rule is: First-In-First-Out (FIFO) principle; The feature extraction module is specifically used for: Obtain the frequency corresponding to the maximum amplitude of the vibration spectrum in the vibration spectrum; Determine whether the frequency meets the preset conditions; If so, then obtain the polygon order and corresponding polygon features in the vibration spectrum; The preset conditions are specifically as follows: ； when When, then the polygon order ; like When the order is 0, the polygon order is... ; in, It is a modulo function. The frequency corresponding to the maximum amplitude of the vibration spectrum in the vibration spectrum. The rotational frequency of the wheels.

Images