Train running part axle box bearing fault monitoring method and system

Abstract

This invention belongs to the field of rail transit technology, specifically disclosing a method and system for monitoring faults in axle box bearings of a train running gear. The method includes the following steps: installing vibration and temperature sensors on multiple axle box bearings of the train running gear to collect temperature and vibration data in real time; extracting features from the collected data and performing collaborative processing to obtain corresponding collaborative feature values; performing feature-level fusion to preliminarily diagnose the bearing's health status; calculating the corresponding fault probability using the collaborative feature values ​​to perform a secondary diagnosis of the bearing's health status; and determining whether to issue an early warning based on the results of the preliminary and secondary diagnoses. By employing this technical solution and using a collaborative and fusion strategy, potential faults can be detected earlier and timely warnings issued, ensuring the continuous and stable operation of rail trains and providing strong protection for train operation safety.

Description

Technical Field

[0001] This invention belongs to the field of rail transit technology and relates to a method and system for monitoring the failure of axle box bearings in the running gear of a train. Background Technology

[0002] The safe operation of railcars depends on the stability of their running gear, and bearings, as a crucial component of the running gear, directly affect the train's operational efficiency and safety. During long-term operation, bearings can experience problems such as plastic deformation, fatigue fracture, pitting, and spalling due to alternating loads and complex environmental factors. Ignoring these faults can lead not only to unplanned downtime and damage to train equipment but also potentially to safety accidents. Therefore, real-time monitoring of the health status and high-precision fault diagnosis of train running gear bearings are crucial for ensuring safe train operation.

[0003] Due to the complex operating environment of trains, external interference (such as changes in road conditions) can easily affect the data acquired by sensors, leading to an increase in false alarms, which poses a challenge to fault detection. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for monitoring the failure of axle box bearings in the running gear of a train, which can detect potential failures earlier and provide timely warnings, ensuring the continuous and stable operation of rail trains and providing strong protection for the safety of train operation.

[0005] To achieve the above objectives, the basic solution of the present invention is: a method for monitoring the fault of axle box bearings in a train running gear, comprising the following steps:

[0006] Temperature sensors are installed on multiple axle box bearings in the train running gear to collect temperature and vibration data in real time.

[0007] The features of the collected temperature and vibration data are extracted and processed collaboratively to obtain the corresponding collaborative feature values.

[0008] By fusing collaborative feature values ​​at the feature level, a preliminary diagnosis of the bearing's health status can be made.

[0009] Using a fusion diagnostic method based on DS evidence theory, the corresponding failure probability is calculated through collaborative eigenvalues ​​to perform a secondary diagnosis of the bearing's health status.

[0010] Based on the results of the preliminary and secondary diagnoses, if the preliminary diagnosis indicates a bearing fault and the secondary diagnosis also indicates a faulty bearing, a warning message will be issued; if the preliminary diagnosis indicates a bearing fault but the secondary diagnosis indicates a normal bearing, no warning message will be issued.

[0011] The working principle and beneficial effects of this basic solution are as follows: This technical solution overcomes the shortcomings of a single sensor being susceptible to external interference by coordinating the vibration and temperature characteristic data of multiple acquisition nodes, thereby achieving more accurate bearing fault diagnosis.

[0012] This invention employs a diagnostic strategy combining feature-level fusion and decision-level fusion to avoid misdiagnosis and significantly improve diagnostic reliability and accuracy. It also provides real-time monitoring of bearing status, offering early warnings to ensure safe train operation and facilitate efficient maintenance.

[0013] Furthermore, the method for extracting features from the collected vibration data and performing collaborative processing to obtain the corresponding collaborative feature values ​​is as follows:

[0014] Four time-frequency domain cooperative features were extracted from the vibration data: cooperative RMS value, cooperative peak-to-peak value, cooperative kurtosis value, and cooperative root mean square frequency.

[0015] Collaborative RMS value X RMS for:

[0016]

[0017] Wherein, RMSa represents the RMS value of the vibration of the measured axle box, RMS i denoted as RMS value of the i-th axle box, where RMSa is not included in all RMSi values, and N represents the total number of axle boxes.

[0018] Collaborative peak-to-peak value X P for:

[0019]

[0020] Where Pa represents the peak-to-peak value of the vibration of the measured axle box, P i Pi represents the peak-to-peak value of the vibration of the i-th axle box, excluding Pa in all Pi values;

[0021] Cooperative kurtosis value X K for:

[0022]

[0023] Where Ka represents the vibration kurtosis of the measured axle box, K i Ki represents the vibration kurtosis of the i-th axle box, excluding Ka;

[0024] Cooperative root mean square frequency X RMSF for:

[0025]

[0026] Where RMSFa represents the root mean square frequency of the tested axle box, RMSFi represents the root mean square frequency of the i-th axle box, and all RMSFi do not include RMSFa.

[0027] RMS value measures the energy of a signal and reflects the health status of the bearing during operation; peak-to-peak value helps capture large instantaneous fluctuations caused by internal bearing damage; kurtosis value identifies spikes and provides timely warnings of minor impacts or early damage; and root mean square frequency extracts frequency domain features, revealing frequency information of specific fault modes. Feature-level fusion of multiple features helps to comprehensively monitor bearing condition, thereby improving the accuracy and sensitivity of fault diagnosis.

[0028] Furthermore, the method for extracting features from the collected temperature data and performing collaborative processing to obtain the corresponding collaborative feature values ​​is as follows:

[0029] Two corresponding collaborative features are extracted from the temperature data: collaborative absolute temperature X. T and the rate of temperature rise X VT =:

[0030]

[0031] Where Ta represents the temperature of the measured axle box, T i Let represent the temperature of the i-th axle box, excluding Ta in all Ti values, and N represent the total number of axle boxes; VTa represents the temperature rise rate of the measured axle box, and VT... i This represents the temperature rise rate of the i-th axle box, excluding VTa from all VTi values.

[0032] By collaboratively processing the feature information from multiple acquisition nodes, corresponding collaborative feature values ​​are obtained, which are beneficial for subsequent use.

[0033] Furthermore, the collaborative feature values ​​are fused at the feature level to preliminarily diagnose the health status of the bearing. The specific steps are as follows:

[0034] Principal component analysis (PCA) was performed on the extracted collaborative features, and the obtained collaborative feature quantities were standardized.

[0035]

[0036] in, ξ is the standardized feature quantity. j E(ξ) is a characteristic quantity. j S(ξ) represents the mean of the characteristic quantities. j ) represents the standard deviation of the characteristic quantity;

[0037] The sample matrix X is composed of all standardized features. n×m ,for:

[0038]

[0039] Each row represents an observation sample, and each column represents a collaborative feature.

[0040] Through sample matrix X n×m The covariance matrix M is calculated as follows:

[0041] M = cov(X) n×m );

[0042] The eigenvalues ​​λ of the covariance matrix M i The values ​​are arranged in descending order, and their corresponding eigenvectors are used to form the feature matrix K:

[0043]

[0044] Calculate the contribution rate G of the characteristic quantity i :

[0045]

[0046] The top j features with a cumulative contribution rate of 85% are selected as the principal component P. j ,for:

[0047]

[0048] The principal component P j Input the trained support vector machine (SVM) to make a preliminary judgment on whether the bearing has a fault.

[0049] By fusing collaborative feature values ​​at the feature level, a preliminary diagnosis of the bearing's health status can be made.

[0050] Furthermore, a fusion diagnostic method based on DS evidence theory is used to calculate the corresponding failure probability through collaborative eigenvalues, thereby performing a secondary diagnosis of the bearing's health status. The specific method is as follows:

[0051] The identification framework H for bearing fault diagnosis is defined as follows:

[0052] H = {h1, h2}

[0053] Where h1 represents a fault and h2 represents a fault-free condition;

[0054] Based on the collaborative feature values ​​of temperature and vibration data, the probability values ​​of fault presence, fault absence, and fault uncertainty are calculated for each collaborative feature value.

[0055] The fault probability value, fault-free probability value and fault uncertainty probability value of each cooperative feature value are normalized to obtain the basic probability function of each cooperative feature value.

[0056] Based on the basic probability functions of each collaborative feature value, decision fusion is performed to obtain the final basic probability function of bearing failure; among which, the basic probability function includes failure, no failure, and failure uncertainty, and the existence of failure is determined according to the probability magnitude.

[0057] To avoid misdiagnosis, the feature information from multiple acquisition nodes is first processed collaboratively to calculate the fault probability function of the corresponding collaborative feature values. Then, a fusion diagnosis method based on DS evidence theory is used to make a more accurate diagnosis of the bearing's operating status.

[0058] Furthermore, based on the collaborative feature values ​​of temperature and vibration data, the steps for calculating the probability of fault presence, probability of no fault, and probability of fault uncertainty corresponding to each collaborative feature value are as follows:

[0059] Through collaborative RMS value X RMS Calculate the probability m of failure. RMS The probability of no failure n RMS Fault uncertainty probability p RMS :

[0060]

[0061] Among them, X RMS α represents the co-operative RMS value, and α, β, and σ represent the correction coefficients for each formula.

[0062] Through the co-kurtosis value X K Calculate the probability m of failure. K The probability of no failure n K Fault uncertainty probability p K :

[0063]

[0064] Among them, X K Indicates the co-kurtosis value;

[0065] Through coordinated absolute temperature X T Calculate the probability m of failure. T The probability of no failure n T Fault uncertainty probability p T :

[0066]

[0067] Among them, X T Indicates the combined absolute temperature value;

[0068] Through coordinated temperature rise rate X VT = Calculate the probability m of a fault VT The probability of no failure nVT Fault uncertainty probability p VT :

[0069]

[0070] Among them, X VT This indicates the rate of coordinated temperature rise.

[0071] Determine the probability of each failure to enable accurate diagnosis later.

[0072] Furthermore, the method for normalizing the fault probability, fault-free probability, and fault uncertainty probability of each cooperative feature value to obtain the basic probability function of each cooperative feature value is as follows:

[0073] Through collaborative RMS value X RMS Calculate the basic probability function m for faults 11 (h1), the basic probability function for fault-free operation m 11 (h2) Fault uncertainty basic probability function m 11 (H):

[0074]

[0075] Through the co-kurtosis value X K Calculate the basic probability function m for faults 12 (h1), the basic probability function for fault-free operation m 12 (h2) Fault uncertainty basic probability function m 12 (H):

[0076]

[0077] Through coordinated absolute temperature X T Calculate the basic probability function m for faults 21 (h1), the basic probability function for fault-free operation m 21 (h2) Fault uncertainty basic probability function m 21 (H):

[0078] Through coordinated temperature rise rate X VT = Calculate the basic probability function m for faults 22 (h1), the basic probability function for fault-free operation m 22 (h2) Fault uncertainty basic probability function m 22 (H):

[0079]

[0080] The probabilities of each collaborative feature value are processed to facilitate data computation.

[0081] Furthermore, based on the basic probability functions of each collaborative feature value, decision fusion is performed, with the following specific steps:

[0082] The formula for synthesizing evidence in the DS theory is as follows:

[0083]

[0084] In the DS evidence theory, C is the fused target set, m(C) is the probability function of the target set, A and B represent the sets of two different pieces of evidence supporting C, m1 and m2 are two basic probability functions under the same identification framework, and K is the degree of trust conflict describing the conflict between the two, which is:

[0085] The synthesis formula can be written as:

[0086]

[0087] The basic probability function m for the co-RMS value 11 The fundamental probability function m of co-kurtosis 12 The fundamental probability function m1 of the vibration is obtained by fusion, that is:

[0088]

[0089] The fundamental probability function m of cooperative absolute temperature 21 The fundamental probability function m of the coordinated temperature rise rate 22 The basic probability function m2 for temperature is obtained by fusion:

[0090]

[0091] The fundamental probability function m1 for vibration and the fundamental probability function m2 for temperature are fused to obtain the final fundamental probability function m for bearing failure:

[0092]

[0093] By fusing the various probability data, errors in calculating the probability of failure can be reduced, thereby improving the accuracy of diagnosis.

[0094] The present invention also provides a fault monitoring system for axle box bearings of a train running gear, including a data acquisition module and a processing module. The data acquisition module is used to acquire temperature and vibration data in real time, and the output end of the data acquisition module is connected to the input end of the processing module.

[0095] The processing module executes the method described in this invention to monitor the faults of the axle box bearings of the train running gear.

[0096] This system utilizes data acquisition and processing modules to monitor axle box bearing failures in train running gear. It can detect potential faults earlier and provide timely warnings, ensuring the continuous and stable operation of rail trains and providing strong protection for train operation safety.

[0097] Furthermore, the data acquisition module includes a vibration temperature sensor and a pre-processor;

[0098] The vibration and temperature sensor can be detachably installed on the axle box of the running gear of the rail train. The vibration and temperature sensor is used to obtain the temperature and vibration information of the bearing of the axle box of the running gear during the operation of the rail train.

[0099] The preprocessor is connected to the output of the vibration temperature sensor and is used to convert the collected vibration temperature information into digital signals and transmit them to the processing module through a mesh network.

[0100] The data acquisition module has a simple structure and uses a wireless mesh network to transmit data, reducing wiring requirements and simplifying installation and maintenance. Attached Figure Description

[0101] Figure 1 This is a flowchart illustrating the method for monitoring axle box bearing failures in the train running gear according to the present invention.

[0102] Figure 2 This is a flowchart illustrating the preliminary diagnosis of the bearing health status in the train running gear axle box bearing fault monitoring method of the present invention.

[0103] Figure 3 This is a flowchart illustrating the secondary diagnosis of the bearing health status in the train running gear axle box bearing fault monitoring method of the present invention.

[0104] Figure 4 This is a schematic diagram of the installation structure of the vibration and temperature sensor in the train running gear axle box bearing fault monitoring system of the present invention. Detailed Implementation

[0105] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0106] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0107] In the description of this invention, unless otherwise specified and limited, it should be noted that the terms "installation", "connection" and "linking" should be interpreted broadly. For example, they can refer to mechanical or electrical connections, or internal connections between two components. They can be direct connections or indirect connections through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0108] This invention discloses a method for monitoring faults in axle box bearings of train running gear. It collaboratively calculates vibration and temperature characteristics from multiple nodes, comprehensively analyzing multi-dimensional information to improve the accuracy and reliability of fault diagnosis. This collaborative and integrated strategy enables earlier detection of potential faults and timely warnings, ensuring the continuous and stable operation of rail trains and providing strong protection for train operation safety.

[0109] like Figure 1 As shown, the method for monitoring axle box bearing failures in train running gear includes the following steps:

[0110] Step (1): Install vibration and temperature sensors on multiple axle box bearings of the train running gear to collect temperature and vibration data in real time;

[0111] Step (2) Extract the features of the collected temperature and vibration data and perform collaborative processing to obtain the corresponding collaborative feature values;

[0112] By fusing collaborative feature values ​​at the feature level, a preliminary diagnosis of the bearing's health status can be made.

[0113] Step (3): Using the fusion diagnostic method based on DS evidence theory, the corresponding failure probability is calculated through collaborative eigenvalues ​​to perform a secondary diagnosis of the bearing's health status.

[0114] Step (4): Based on the results of the preliminary diagnosis and the secondary diagnosis, if the preliminary diagnosis indicates a bearing failure and the secondary diagnosis also indicates a bearing failure, an early warning message will be issued, and the corresponding bearing box will be inspected and maintained in a timely manner; if the preliminary diagnosis indicates a bearing failure and the secondary diagnosis indicates a bearing is in normal condition, no early warning message will be issued.

[0115] In the condition monitoring of axle box bearings in the running gear of railcars, data from a single vibration sensor is often susceptible to interference from environmental factors such as changes in road conditions, leading to misjudgments of abnormal system states. Relying solely on the characteristics of a single sensor, such as RMS values, peak-to-peak values, or kurtosis values, may result in false alarms due to environmental changes.

[0116] By coordinating data from multiple sensors, features such as coordinating RMS value, coordinating peak-to-peak value, coordinating kurtosis value, and coordinating root mean square frequency can be extracted, effectively reducing the impact of noise and errors, and capturing the true vibration status of the system more stably and reliably. This improves the detection accuracy of abnormal conditions of axle box bearings and reduces the possibility of false alarms.

[0117] In a preferred embodiment of the present invention, the method for extracting features from the collected vibration data and performing collaborative processing to obtain corresponding collaborative feature values ​​is as follows:

[0118] Four time-frequency domain cooperative features were extracted from the vibration data: cooperative RMS value, cooperative peak-to-peak value, cooperative kurtosis value, and cooperative root mean square frequency.

[0119] Collaborative RMS value X RMS for:

[0120]

[0121] Wherein, RMSa represents the RMS value of the vibration of the measured axle box, RMS i denoted as RMS value of the i-th axle box, where RMSa is not included in all RMSi values, and N represents the total number of axle boxes.

[0122] Collaborative peak-to-peak value X P for:

[0123]

[0124] Where Pa represents the peak-to-peak value of the vibration of the measured axle box, P i Pi represents the peak-to-peak value of the vibration of the i-th axle box, excluding Pa in all Pi values;

[0125] Cooperative kurtosis value X K for:

[0126]

[0127] Where Ka represents the vibration kurtosis of the measured axle box, K i Ki represents the vibration kurtosis of the i-th axle box, excluding Ka;

[0128] Cooperative root mean square frequency X RMSF for:

[0129]

[0130] Where RMSFa represents the root mean square frequency of the tested axle box, RMSFi represents the root mean square frequency of the i-th axle box, and all RMSFi do not include RMSFa.

[0131] In temperature monitoring of axle box bearings in the running gear of railcars, the temperature value and temperature rise rate of a single sensor may be affected by changes in the ambient temperature of the vehicle's operating environment, making it difficult to accurately reflect the actual temperature state of the bearing. Relying solely on a single sensor may lead to misjudgments due to changes in ambient temperature.

[0132] By collaboratively processing temperature data from multiple sensor nodes and comprehensively calculating the collaborative characteristics of temperature and temperature rise rate, interference from environmental factors can be effectively reduced, the true changes in bearing temperature can be captured more accurately, and false alarms caused by environmental fluctuations can be avoided.

[0133] In a preferred embodiment of the present invention, the method for extracting features from the collected temperature data and performing collaborative processing to obtain corresponding collaborative feature values ​​is as follows:

[0134] Two corresponding collaborative features are extracted from the temperature data: collaborative absolute temperature X. T and the rate of temperature rise X VT =:

[0135]

[0136] Where Ta represents the temperature of the measured axle box, T i Let represent the temperature of the i-th axle box, excluding Ta in all Ti values, and N represent the total number of axle boxes; VTa represents the temperature rise rate of the measured axle box, and VT... i This represents the temperature rise rate of the i-th axle box, excluding VTa from all VTi values.

[0137] In a preferred embodiment of the present invention, such as Figure 2 As shown, feature-level fusion of collaborative feature values ​​is performed to preliminarily diagnose the health status of the bearing. The specific steps are as follows:

[0138] Principal component analysis (PCA) was performed on the extracted collaborative features, and the obtained collaborative feature quantities were standardized.

[0139]

[0140] in, ξ is the standardized feature quantity. j E(ξ) is a characteristic quantity. j S(ξ) represents the mean of the characteristic quantities. j ) represents the standard deviation of the characteristic quantity;

[0141] The sample matrix X is composed of all standardized features. n×m ,for:

[0142]

[0143] Each row represents an observation sample, and each column represents a collaborative feature.

[0144] Through sample matrix X n×m The covariance matrix M is calculated as follows:

[0145] M = cov(X) n×m );

[0146] The eigenvalues ​​λ of the covariance matrix M i The values ​​are arranged in descending order, and their corresponding eigenvectors are used to form the feature matrix K:

[0147]

[0148] Calculate the contribution rate G of the characteristic quantity i :

[0149]

[0150] The top j features with a cumulative contribution rate of 85% are selected as the principal component P. j ,for:

[0151]

[0152] The principal component P j Input the trained support vector machine (SVM) to make a preliminary judgment on whether the bearing has a fault.

[0153] In a preferred embodiment of the present invention, such as Figure 3 As shown, a fusion diagnostic method based on DS evidence theory is used to calculate the corresponding failure probability through collaborative eigenvalues, and to perform a secondary diagnosis of the bearing's health status. The specific method is as follows:

[0154] The identification framework H for bearing fault diagnosis is determined. Since it is only necessary to determine whether a fault has occurred, the identification framework is as follows:

[0155] H = {h1, h2}

[0156] Where h1 represents a fault and h2 represents a fault-free condition;

[0157] We need to calculate the basic probability function m(h1) for faulty conditions and the basic probability function m(h2) for fault-free conditions. Meanwhile, under the DS evidence theory, the basic probability function for uncertain fault conditions is represented by m(H).

[0158] While single sensors are highly sensitive to fault detection, they are also susceptible to external interference, such as changes in road conditions and ambient temperature. These disturbances can lead to errors in fault probability calculations, increasing the risk of false alarms. Therefore, collaborative processing of multi-sensor signals can effectively suppress the effects of noise and external interference, thereby improving diagnostic accuracy and system reliability.

[0159] Based on the collaborative feature values ​​of temperature and vibration data, the probability values ​​of fault presence, fault absence, and fault uncertainty are calculated for each collaborative feature value.

[0160] The fault probability value, fault-free probability value and fault uncertainty probability value of each cooperative feature value are normalized to obtain the basic probability function of each cooperative feature value.

[0161] Based on the basic probability functions of each collaborative feature value, decision fusion is performed to obtain the final basic probability function of bearing failure; among which, the basic probability function includes failure, no failure, and failure uncertainty, and the existence of failure is determined according to the probability magnitude.

[0162] In a preferred embodiment of the present invention, the steps for calculating the probability of failure, the probability of no failure, and the probability of failure uncertainty corresponding to each collaborative feature value based on the collaborative feature values ​​of temperature and vibration data are as follows:

[0163] Through collaborative RMS value X RMS Calculate the probability m of failure. RMS The probability of no failure n RMS Fault uncertainty probability p RMS :

[0164]

[0165] Among them, X RMS α represents the co-operative RMS value, and α, β, and σ represent the correction coefficients for each formula.

[0166] Through the co-kurtosis value X K Calculate the probability m of failure. K The probability of no failure n K Fault uncertainty probability p K :

[0167]

[0168] Among them, X K Indicates the co-kurtosis value;

[0169] Through coordinated absolute temperature X T Calculate the probability m of failure. T The probability of no failure n TFault uncertainty probability p T :

[0170]

[0171]

[0172] Among them, X T Indicates the combined absolute temperature value;

[0173] Through coordinated temperature rise rate X VT = Calculate the probability m of a fault VT The probability of no failure n VT Fault uncertainty probability p VT :

[0174]

[0175] Among them, X VT This indicates the rate of coordinated temperature rise.

[0176] In a preferred embodiment of the present invention, the method for normalizing the fault probability value, fault-free probability value, and fault uncertainty probability value of each cooperative feature value to obtain the basic probability function of each cooperative feature value is as follows:

[0177] Through collaborative RMS value X RMS Calculate the basic probability function m for faults 11 (h1), the basic probability function for fault-free operation m 11 (h2) Fault uncertainty basic probability function m 11 (H):

[0178]

[0179] Through the co-kurtosis value X K Calculate the basic probability function m for faults 12 (h1), the basic probability function for fault-free operation m 12 (h2) Fault uncertainty basic probability function m 12 (H):

[0180]

[0181] Through coordinated absolute temperature X T Calculate the basic probability function m for faults 21 (h1), the basic probability function of fault-free operation m 21 (h2) Fault uncertainty basic probability function m 21 (H):

[0182]

[0183] Through coordinated temperature rise rate X VT = Calculate the basic probability function m for faults 22 (h1), the basic probability function for fault-free operation m 22 (h2), Fault uncertainty basic probability function m 22 (H):

[0184]

[0185] In a preferred embodiment of the present invention, decision fusion is performed based on the basic probability function of each collaborative feature value, and the specific steps are as follows:

[0186] The formula for synthesizing evidence in the DS theory is as follows:

[0187]

[0188] In the DS evidence theory, C is the fused target set, m(C) is the probability function of the target set, A and B represent the sets of two different pieces of evidence supporting C, m1 and m2 are two basic probability functions under the same identification framework, and K is the degree of trust conflict describing the conflict between the two, which is:

[0189] The synthesis formula can be written as:

[0190]

[0191] The basic probability function m for the co-RMS value 11 The fundamental probability function m of co-kurtosis 12 The fundamental probability function m1 of the vibration is obtained by fusion, that is:

[0192]

[0193] The fundamental probability function m of cooperative absolute temperature 21 The fundamental probability function m of the coordinated temperature rise rate 22 The basic probability function m2 for temperature is obtained by fusion:

[0194]

[0195] The fundamental probability function m1 for vibration and the fundamental probability function m2 for temperature are fused to obtain the final fundamental probability function m for bearing failure:

[0196]

[0197] This invention also provides a fault monitoring system for axle box bearings of a train running gear, including a data acquisition module and a processing module. The data acquisition module is used to acquire temperature and vibration data in real time, and the output terminal of the data acquisition module is electrically connected to the input terminal of the processing module. The processing module executes the method described in this invention to monitor faults in the axle box bearings of the train running gear.

[0198] In a preferred embodiment of the present invention, the data acquisition module includes a vibration temperature sensor (i.e., a vibration temperature sensing module) and a preprocessor, such as... Figure 4 As shown, the vibration and temperature sensor can be detachably installed on the axle box of the running gear of the rail train. The vibration and temperature sensor is used to obtain the temperature and vibration information of the bearings of the axle box of the running gear during the operation of the rail train.

[0199] The pre-processor is located on the running frame at the bottom of each carriage of the train. The pre-processor is electrically connected to the output of the vibration and temperature sensor. It is used to convert the collected vibration and temperature information into digital signals and transmit them to the processing module (i.e., the on-board host) through the mesh network. The on-board host is responsible for storing the received data and processing the data to realize the fault diagnosis of the bearings of the running gear of the rail train.

[0200] The data acquisition module has a simple structure and uses a wireless mesh network to transmit data, reducing wiring requirements and simplifying installation and maintenance.

[0201] This invention overcomes the shortcomings of single sensors being susceptible to external interference by coordinating vibration and temperature characteristic data from multiple acquisition nodes, thus achieving more accurate bearing fault diagnosis. It adopts a diagnostic strategy that combines feature-level fusion and decision-level fusion to avoid misdiagnosis and significantly improve the reliability and accuracy of diagnosis.

[0202] Meanwhile, this invention utilizes a wireless mesh network to transmit data, reducing wiring requirements and simplifying installation and maintenance. This invention not only effectively reduces noise interference but also provides real-time monitoring of bearing status, offering early warnings and ensuring safe train operation and efficient maintenance.

[0203] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0204] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for monitoring faults in axle box bearings of a train running gear, characterized in that, Includes the following steps: Temperature sensors are installed on multiple axle box bearings in the train running gear to collect temperature and vibration data in real time. The features of the collected temperature and vibration data are extracted and processed collaboratively to obtain the corresponding collaborative feature values. By fusing collaborative feature values ​​at the feature level, a preliminary diagnosis of the bearing's health status can be made. Using a fusion diagnostic method based on DS evidence theory, the corresponding failure probability is calculated through collaborative eigenvalues ​​to perform a secondary diagnosis of the bearing's health status. Based on the results of the preliminary diagnosis and the secondary diagnosis, if the preliminary diagnosis indicates a bearing failure and the secondary diagnosis also indicates a bearing failure, a warning message will be issued; if the preliminary diagnosis indicates a bearing failure but the secondary diagnosis indicates a normal bearing condition, no warning message will be issued. A fusion diagnostic method based on DS evidence theory is used to calculate the corresponding failure probability through collaborative eigenvalues, and then perform a secondary diagnosis of the bearing's health status. The specific method is as follows: Based on the collaborative feature values ​​of temperature and vibration data, the probability values ​​of fault presence, fault absence, and fault uncertainty are calculated for each collaborative feature value. The fault probability value, fault-free probability value and fault uncertainty probability value of each cooperative feature value are normalized to obtain the basic probability function of each cooperative feature value. Based on the basic probability functions of each collaborative feature value, decision fusion is performed to obtain the final basic probability function of bearing failure; among which, the basic probability function includes failure, no failure, and failure uncertainty, and the existence of failure is determined according to the probability magnitude. The steps for calculating the probability of fault presence, probability of no fault, and probability of fault uncertainty corresponding to each collaborative feature value based on temperature and vibration data are as follows: Through collaborative RMS value X RMS Calculate the probability m of failure. RMS The probability of no failure n RMS Fault uncertainty probability p RMS : Among them, X RMS α represents the co-operative RMS value, and α, β, and σ represent the correction coefficients for each formula. Through the co-kurtosis value X K Calculate the probability m of failure. K The probability of no failure n K Fault uncertainty probability p K : Among them, X K Indicates the co-kurtosis value; Through coordinated absolute temperature X T Calculate the probability m of failure. T The probability of no failure n T Fault uncertainty probability p T : Among them, X T Indicates the combined absolute temperature value; Through coordinated temperature rise rate X VT = Calculate the probability m of a fault VT The probability of no failure n VT Fault uncertainty probability p VT : Among them, X VT This indicates the rate of coordinated temperature rise.

2. The method for monitoring axle box bearing failures in train running gear as described in claim 1, characterized in that, The method for extracting features from the collected vibration data and performing collaborative processing to obtain the corresponding collaborative feature values ​​is as follows: Four time-frequency domain cooperative features were extracted from the vibration data: cooperative RMS value, cooperative peak-to-peak value, cooperative kurtosis value, and cooperative root mean square frequency. Collaborative RMS value X RMS for: Among them, RMS a This represents the RMS value of the vibration of the measured axle box. i Represents the RMS value of the vibration of the i-th axle box, and all RMS values. i Excluding RMS a N represents the total number of axle boxes excluding the axle box being measured; Synergistic peak-to-peak value X P for: Among them, P a P represents the peak-to-peak value of the vibration of the measured axle box. i This represents the peak-to-peak value of the vibration of the i-th axle box, and all P... i Pa is not included; Cooperative kurtosis value X K for: Among them, K a K represents the vibration kurtosis of the measured axle box. i Represents the vibration kurtosis of the i-th axle box, and all K i Excluding K a ; Cooperative root mean square frequency X RMSF for: Among them, RMSF a The root mean square frequency (RMSF) of the measured axle box is given. i Represents the root mean square frequency (RMSF) of the i-th axle box, and all RMSFs i Excluding RMSF a .

3. The method for monitoring train running gear axle box bearing failure as described in claim 1, characterized in that, The method for extracting features from the collected temperature data and performing collaborative processing to obtain the corresponding collaborative feature values ​​is as follows: Two corresponding collaborative features are extracted from the temperature data: collaborative absolute temperature X. T and the rate of temperature rise X VT =: Among them, T a The temperature of the axle box being measured is represented by T. i Represents the temperature of the i-th axle box, and all T i Ta is not included; N represents the total number of axle boxes; VT a VT represents the rate of temperature rise of the measured axle box. i Represents the temperature rise rate of the i-th axle box, and all VT i VT is not included a .

4. The method for monitoring axle box bearing failures in train running gear as described in claim 1, characterized in that, The collaborative feature values ​​are fused at the feature level to preliminarily diagnose the health status of the bearing. The specific steps are as follows: Principal component analysis (PCA) was performed on the extracted collaborative features, and the obtained collaborative feature quantities were standardized. in, ξ is the standardized feature quantity. j E(ξ) is a characteristic quantity. j S(ξ) represents the mean of the characteristic quantities. j ) represents the standard deviation of the characteristic quantity; The sample matrix X is composed of all standardized features. n×m ,for: Each row represents an observation sample, and each column represents a collaborative feature. Through sample matrix X n×m The covariance matrix M is calculated as follows: M=cov(X n×m ); The eigenvalues ​​λ of the covariance matrix M i The values ​​are arranged in descending order, and their corresponding eigenvectors are used to form the feature matrix K: Calculate the contribution rate G of the characteristic quantity i : The top j features with a cumulative contribution rate of 85% are selected as the principal component P. j ,for: The principal component P j Input the trained support vector machine (SVM) to make a preliminary judgment on whether the bearing has a fault.

5. The method for monitoring axle box bearing failures in train running gear as described in claim 1, characterized in that, The method for normalizing the fault probability, fault-free probability, and fault uncertainty probability of each cooperative feature value to obtain the basic probability function of each cooperative feature value is as follows: Through collaborative RMS value X RMS Calculate the basic probability function m for faults 11 (h1), the basic probability function for fault-free operation m 11 (h2), Fault uncertainty basic probability function m 11 (H): Through the co-kurtosis value X K Calculate the basic probability function m for faults 12 (h1), the basic probability function for fault-free operation m 12 (h2) Fault uncertainty basic probability function m 12 (H): Through coordinated absolute temperature X T Calculate the basic probability function m for faults 21 (h1), the basic probability function of fault-free operation m 21 (h2) Fault uncertainty basic probability function m 21 (H): Through coordinated temperature rise rate X VT = Calculate the basic probability function m for faults 22 (h1), the basic probability function for fault-free operation m 22 (h2), Fault uncertainty basic probability function m 22 (H):

6. The method for monitoring train running gear axle box bearing failure as described in claim 1, characterized in that, Based on the basic probability functions of each collaborative feature value, decision fusion is performed, and the specific steps are as follows: The formula for synthesizing evidence in the DS theory is as follows: In the DS evidence theory, C is the fused target set, m(C) is the probability function of the target set, A and B represent the sets of two different pieces of evidence supporting C, m1 and m2 are two basic probability functions under the same identification framework, and K is the degree of trust conflict describing the conflict between the two, which is: The synthesis formula can be written as: The basic probability function m for the co-RMS value 11 The fundamental probability function m of co-kurtosis 12 The fundamental probability function m1 of the vibration is obtained by fusion, that is: The fundamental probability function m of cooperative absolute temperature 21 The fundamental probability function m of the coordinated temperature rise rate 22 The basic probability function m2 for temperature is obtained by fusion: The fundamental probability function m1 for vibration and the fundamental probability function m2 for temperature are fused to obtain the final fundamental probability function m for bearing failure:

7. A fault monitoring system for axle box bearings of a train running gear, characterized in that, It includes a data acquisition module and a processing module. The data acquisition module is used to acquire temperature and vibration data in real time, and the output end of the data acquisition module is connected to the input end of the processing module. The processing module executes the method described in any one of claims 1-6 to perform fault monitoring of the axle box bearings of the train running gear.

8. The train running gear axle box bearing fault monitoring system as described in claim 7, characterized in that, The data acquisition module includes a vibration temperature sensor and a pre-processor; The vibration and temperature sensor can be detachably installed on the axle box of the running gear of the rail train. The vibration and temperature sensor is used to obtain the temperature and vibration information of the bearing of the axle box of the running gear during the operation of the rail train. The preprocessor is connected to the output of the vibration temperature sensor and is used to convert the collected vibration temperature information into digital signals and transmit them to the processing module through a mesh network.

Images