CN110044616B - Fault diagnosis method and system for bearing and gear of railway locomotive - Google Patents
Fault diagnosis method and system for bearing and gear of railway locomotive Download PDFInfo
- Publication number
- CN110044616B CN110044616B CN201910256518.4A CN201910256518A CN110044616B CN 110044616 B CN110044616 B CN 110044616B CN 201910256518 A CN201910256518 A CN 201910256518A CN 110044616 B CN110044616 B CN 110044616B
- Authority
- CN
- China
- Prior art keywords
- data
- fault
- module
- diagnosis
- real
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/028—Acoustic or vibration analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M17/00—Testing of vehicles
- G01M17/08—Railway vehicles
Abstract
The invention discloses a fault diagnosis method and a fault diagnosis system for a bearing and a gear of a railway locomotive, wherein the method comprises the following steps: arranging a vibration acceleration sensor, a rotating speed measuring sensor and a GPS sensor at preset positions of a locomotive bearing and a gear to acquire real-time data; screening the real-time data and dividing the real-time data into data segments with preset interval duration, and calculating a time domain index array and a frequency domain matrix in real time; and judging the fault degree and calculating the residual life based on the time domain index array through self-learning training, and judging the fault type and position based on the frequency domain matrix through self-learning training. The system comprises a power supply module, a data acquisition module, a cache preprocessing module, a calculation module, a diagnosis module, a storage transmission module, an airborne alarm module and an input/output interface. The method and the system are realized in a vehicle-mounted hardware unit and do not depend on external computing equipment and terminals, so that the stability and the reliability of fault diagnosis are ensured.
Description
Technical Field
The invention belongs to the technical field of fault diagnosis and signal processing analysis, and particularly relates to a fault diagnosis method and a fault diagnosis system for a bearing and a gear of a railway locomotive.
Background
The bearing and the gear are one of the core components of a traction motor and an axle box of the railway locomotive, and directly influence the running safety of the locomotive. The conventional fault diagnosis system for the bearing and the gear of the railway locomotive generally alarms according to simple characteristic indexes, so that the fault diagnosis rate is low, the conditions of missing alarm and false alarm are serious, the maintenance personnel at the locomotive service section cannot accurately master the real fault state of the bearing and the gear, and great potential safety hazard is caused; in addition, in the conventional railway locomotive monitoring and diagnosing system, the failure diagnosis modes of the bearing and the gear are as follows: the airborne hardware is only responsible for data acquisition, data analysis and fault diagnosis are carried out by excessively depending on a software system of a computing terminal, and due to the fact that the data transmission quantity is large, much data cannot be analyzed and selectively deleted, the fault degree and the fault type of a locomotive bearing and a gear cannot be accurately found at the first time, and the diagnosis result has hysteresis; in addition, the existing railway locomotive monitoring and diagnosing system is simple in function, excessive external equipment is used for fault data processing and fault alarming, the related equipment is multiple, wiring is long, cost is high, stability is poor, and urgent requirements of the current locomotive service section on accurate real-time diagnosis of bearings and gears cannot be met.
The above information disclosed in this background section is only for enhancement of understanding of the background of the invention and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a fault diagnosis method and a fault diagnosis system for a bearing and a gear of a railway locomotive, which can realize the calculation of fault characteristic indexes and the calculation of frequency spectrums in onboard hardware, realize the real-time fault diagnosis in the hardware and then display the fault diagnosis result at a local terminal and a remote terminal.
The invention aims to realize the purpose through the following technical scheme, and the method for diagnosing the faults of the bearing and the gear of the railway locomotive comprises the following steps of:
in the first step, a vibration acceleration sensor, a rotating speed measuring sensor and a GPS sensor are arranged at preset positions of a locomotive bearing and a gear to acquire real-time data;
in the second step, screening the real-time data, dividing the real-time data into data segments with preset interval duration, and caching the data segments;
in the third step, the data segment is filtered, and a time domain index array and a frequency domain matrix are generated based on the filtered data segment, wherein the time domain index array at least comprises a first index, a second index and a third index, the frequency domain matrix comprises an FFT, an envelope spectrum, a cepstrum and an energy spectrum, and the first index, the second index and the third index are respectively:
wherein F represents the filtered vibration signal, N is the total number of points in F, m and i are both positive integers, F' represents the derivative of F,
denotes the average value of F ', F' denotes the second derivative of F,
denotes the mean value of F', Δ E i ＝F 2 i+1 -F 2 i ，
Is the average value of Δ E;
and in the fourth step, judging the fault degree and calculating the residual life through self-learning training based on the time domain index array, and judging the fault type and position through self-learning training based on the frequency domain matrix.
In the second step, when the speed of the locomotive reaches a preset percentage of the rated running speed and is not in the acceleration and deceleration process, the locomotive is judged to be in the normal running process, the real-time data of the normal running process is resampled at equal angles and is divided into data segments with preset interval duration, and the data segments are cached.
In the second step, when the change of the speed of the locomotive is smaller than the preset speed amplitude, the locomotive is judged to be in a stable running process, when the change of the rotating speed and the vibration is smaller than a first preset range, the real-time data of the normal running process is sampled and divided into data segments with preset interval duration, and the data segments are cached.
In the method, in the second step, when the locomotive does not turn and is not accelerated or decelerated and the rotating speed is a rated speed, the locomotive is determined to reach a preset position based on a GPS sensor, real-time data from the preset position is sampled and divided into data sections with preset interval duration, and the data sections are cached.
In the method, in the second step, when the real-time data from the preset position is sampled and the change of the rotating speed and the vibration is larger than a second preset value, the speed of the locomotive is adjusted to enable the change of the rotating speed and the vibration to be smaller than a second preset range, the real-time data is continuously sampled and divided into data sections with preset interval duration, and the data sections are cached.
In the method, the time interval is 1-5 seconds, in the third step, the time domain index array further comprises a mean value, a standard deviation, a variance, a skewness index, a kurtosis index, a peak value, a mean square amplitude, an average amplitude, a square root amplitude, a waveform index and a peak value index, and in the fourth step, self-learning training is performed through a convolutional neural network and a knowledge vector machine.
According to another aspect of the present invention, a diagnostic system for implementing the method for diagnosing the bearing and gear failure of the railway locomotive comprises,
the data acquisition module is arranged at a preset position of a locomotive bearing and a gear to acquire real-time data;
a cache preprocessing module, which is connected with the data acquisition module and comprises,
a pre-processing unit that screens the real-time data and divides it into data segments of a predetermined interval duration,
a cache unit connected to the preprocessing unit, the cache unit configured to cache the data segment;
a computing module, connected to the cache preprocessing module, comprising,
a time domain index calculation unit generating a time domain index array based on the filtered data segments,
a spectrum calculation unit which generates a frequency domain matrix based on the filtered data segment;
and the diagnosis module is connected with the calculation module, judges the fault degree and calculates the residual life based on the time domain index array self-learning training, and judges the fault type and position based on the frequency domain matrix through self-learning training.
In the diagnostic system, the system further comprises,
the power module comprises a vibration energy collecting unit, a rechargeable battery unit and an external power supply, and if the external power supply is connected, the power supply is directly used for supplying power and charging the rechargeable battery; if the power supply is not connected, the rechargeable battery supplies power to the system according to the stored electric quantity, and meanwhile, the vibration energy collecting unit acquires vibration energy to convert the vibration energy into electric energy and charges the rechargeable battery;
the storage transmission module is connected with the diagnosis module and is configured to store and transmit fault information, the transmission mode comprises a wired mode and a wireless mode, and the fault information comprises fault degree, residual service life, fault type and position;
the alarm module comprises an alarm module and a control module,
an alarm unit that issues an alarm based on the degree of the failure,
a display unit that displays failure information;
an input-output interface, comprising,
a sensor interface for reading data of the vibration acceleration sensor, the rotating speed measuring sensor and the GPS sensor,
a data interface for exchanging data with an external storage device,
and the output interface transmits the airborne data to the remote equipment in a wired mode.
In the diagnosis system, the data acquisition module comprises a vibration acceleration sensor, a rotating speed measurement sensor and a GPS sensor; the storage transmission module comprises a storage module for storing fault information and a transmission module for transmitting the fault information; the alarm unit comprises a buzzer and an LED lamp; the display unit includes a liquid crystal display. When no fault exists, only the time domain index array in the calculation module is stored and sent, and the rest data and the frequency spectrum matrix can be discarded, so that the data volume is greatly reduced. When a fault occurs, the storage module stores the data segment and the fault information, the transmission module transmits the complete data of the data segment and the corresponding fault information, the remote terminal can analyze the data segment in detail, the alarm unit sends out an alarm, and the liquid crystal display screen displays the fault information.
In the diagnosis system, the diagnosis module is in wireless connection with the mobile terminal, and the mobile terminal comprises a computer, a mobile phone, a bracelet, a large screen and a cloud server.
Compared with the prior art, the invention has the following advantages:
the invention is integrated in the hardware unit, thus being capable of carrying out real-time early warning and diagnosis on the fault in real time; an external storage unit, a button group, an LED and a liquid crystal display screen are introduced to realize parameter setting and result display of fault diagnosis, and fault early warning and diagnosis can be carried out without depending on any external equipment. By adopting a cache and selective mechanism, the computing unit can conveniently and rapidly compute the real-time characteristic indexes, only the computed characteristic indexes are stored and transmitted when no fault exists, and the original data and the frequency spectrum can be discarded, so that the data volume is greatly reduced. The method has three modes: the system comprises an active local detection mode, a steady detection mode and a global detection mode, and can be switched randomly, wherein the mode switching function is realized by a key set, the active local detection mode only acquires vibration data of a steady rotating speed and a straight track part, the data quality is best, and the diagnosis result has the most referential property; the steady detection mode only adopts the data of the steady section of the rotating speed, the data volume is large, the diagnosis results can be verified mutually, and the reliability of diagnosis can be improved; the global detection mode ensures monitoring throughout the run, emphasizes the integrity of time domain index monitoring, but is time consuming. By adopting three modes, the comprehensive monitoring on fault diagnosis can be ensured, and the operation safety can be ensured to the maximum extent.
Drawings
Various other advantages and benefits of the present invention will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. It is obvious that the drawings described below are only some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort. Also, like parts are designated by like reference numerals throughout the drawings.
In the drawings:
FIG. 1 is a schematic representation of the steps of a method of diagnosing a bearing and gear failure of a railroad locomotive according to one embodiment of the present invention;
FIG. 2 is a schematic illustration of implementation of rail locomotive traction motor drive end bearing vibration acceleration and velocity in accordance with one embodiment of the present invention;
FIG. 3 is a schematic diagram of a diagnostic system implementing a method of diagnosing a bearing and gear failure of a railroad locomotive according to one embodiment of the present invention;
FIG. 4 is a schematic illustration of data transmission and connection for implementing a method for diagnosing a bearing and gear failure of a railroad locomotive according to one embodiment of the present invention;
FIG. 5 is a schematic diagram of a cache preprocessing module of a diagnostic system for implementing a method for diagnosing a bearing and gear failure of a railroad locomotive according to one embodiment of the present invention;
FIG. 6 is a flow chart of the operation of the memory transfer module of the diagnostic system implementing the method for diagnosing railroad car bearings and gear faults in accordance with one embodiment of the present invention.
The invention is further explained below with reference to the figures and examples.
Detailed Description
Specific embodiments of the present invention will be described in more detail below with reference to fig. 1 to 6. While specific embodiments of the invention are shown in the drawings, it should be understood that the invention may be embodied in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.
It should be noted that certain terms are used throughout the description and claims to refer to particular components. As one skilled in the art will appreciate, various names may be used to refer to a component. This specification and claims do not intend to distinguish between components that differ in name but not function. In the following description and in the claims, the terms "include" and "comprise" are used in an open-ended fashion, and thus should be interpreted to mean "include, but not limited to. The description which follows is a preferred embodiment of the invention, but is made for the purpose of illustrating the general principles of the invention and not for the purpose of limiting the scope of the invention. The scope of the invention is to be determined by the claims appended hereto.
For the purpose of facilitating understanding of the embodiments of the present invention, the following description will be made by taking specific embodiments as examples with reference to the accompanying drawings, and the drawings are not to be construed as limiting the embodiments of the present invention.
For better understanding, fig. 1 is a schematic diagram illustrating the steps of a method for diagnosing a bearing and gear fault of a railway locomotive according to an embodiment of the present invention, and as shown in fig. 1, the method for diagnosing the bearing and gear fault of the railway locomotive comprises the following steps:
in the first step S1, a vibration acceleration sensor, a rotating speed measuring sensor and a GPS sensor are arranged at preset positions of a locomotive bearing and a gear to acquire real-time data;
in a second step S2, screening the real-time data, dividing the real-time data into data segments with preset interval duration, and caching the data segments;
in a third step S3, the data segment is filtered, and a time domain index array and a frequency domain matrix are generated based on the filtered data segment, where the time domain index array at least includes a first index, a second index, and a third index, and the frequency domain matrix includes an FFT, an envelope spectrum, a cepstrum, and an energy spectrum, where the first index, the second index, and the third index are respectively:
wherein F represents the filtered vibration signal, N is the total number of points in F, m and i are both positive integers, F' represents the derivative of F,
denotes the average value of F ', F' denotes the second derivative of F,
denotes the mean value of F', Δ E i ＝F 2 i+1 -F 2 i ，
Is the average value of Δ E;
and in a fourth step S4, judging the fault degree and calculating the residual life through self-learning training based on the time domain index array, and judging the fault type and position through self-learning training based on the frequency domain matrix.
In one embodiment of the method, in the second step S2, when the speed of the locomotive reaches a predetermined percentage of the rated operating speed and is not in the acceleration and deceleration process, it is determined that the locomotive is in the normal operating process, the real-time data of the normal operating process is resampled at equal angles and divided into data segments with predetermined interval duration, and the data segments are cached.
In one embodiment of the method, in the second step S2, when the change of the locomotive speed is smaller than the predetermined speed amplitude, it is determined that the locomotive is in a steady operation process, and when the change of the rotating speed and the vibration is smaller than the first predetermined range, the real-time data of the normal operation process is sampled and divided into data segments with predetermined interval duration, and the data segments are cached.
In one embodiment of the method, in a second step S2, when the locomotive is not turning and is not accelerating or decelerating, and the rotation speed is a rated speed, it is determined that the locomotive reaches a predetermined position based on the GPS sensor, real-time data from the predetermined position is sampled and divided into data segments of a predetermined interval duration, and the data segments are buffered.
In one embodiment of the method, in the second step S2, when the real-time data from the predetermined position is sampled and the variation of the rotation speed and the vibration is greater than the second predetermined value, the speed of the locomotive is adjusted so that the variation of the rotation speed and the vibration is smaller than the second predetermined range, the real-time data is continuously sampled and divided into data segments with predetermined interval duration, and the data segments are cached.
In an embodiment of the method, the time interval is 1 to 5 seconds, in a third step S3, the time domain index array further includes a mean value, a standard deviation, a variance, a skewness index, a kurtosis index, a peak-to-peak value, a peak-to-mean value, a mean amplitude, a square root amplitude, a waveform index, and a peak index, and in a fourth step S4, self-learning training is performed via a convolutional neural network and a knowledge vector machine.
To further understand the present invention, in one embodiment, the diagnostic objects are bearings and gears of a railway locomotive, and a vibration acceleration sensor, a rotational speed measurement sensor, and a GPS sensor are disposed at respective positions of the bearings and gears of the locomotive. After the system is powered on, the equipment is turned on, real-time data acquired by the sensor is acquired through the sensor interface, and the original data is preprocessed through the cache and preprocessing module; transmitting the preprocessed effective data into a calculation module, and calculating a time domain index and a frequency spectrum; inputting the calculation result into a diagnosis module, and comparing the difference between the current time domain index and the frequency spectrum with the difference in a preset parameter library to obtain a diagnosis result of the fault type and the fault degree; and according to whether the fault occurs or not, carrying out corresponding early warning display in the onboard alarm module, storing the diagnosis result and corresponding data in a local storage unit, and sending the diagnosis result and the corresponding data to a remote terminal in a wired or wireless mode.
The minimum system representation of the invention can realize data acquisition, fault diagnosis and fault display without depending on a remote system. The remote system shows that when the system is connected with a wire or a wireless network, the system can further perform real-time early warning display of faults at terminals such as an enterprise control terminal, a notebook computer, a mobile phone, a bracelet, a large screen and the like.
In one embodiment, as shown in the schematic diagram of vibration acceleration and velocity in fig. 2, the geographical environment of the locomotive running on the railway line is complex, and the altitude terrain and velocity fluctuation is large according to the GPS track of the railway locomotive running line and the vibration acceleration and velocity during the running process. The influence of the change of the operation condition on the vibration is far stronger than the influence of the weak fault of the bearing on the vibration, and if the vibration data of the whole line is adopted to carry out fault diagnosis on the traction motor bearing, the false alarm and the false alarm are very easily caused. Therefore, in the embodiment, the local active detection mode is adopted for fault diagnosis of the traction motor bearing, and the GPS can be obtained by analysis 1 And GPS 2 The road section between the GPS and the GPS is a road section with stable altitude, straight track and stable rotating speed 1 And GPS 2 The corresponding internal identification values are 831 and 923. Its corresponding vibration is as in the GPS of FIG. 2 1 And GPS 2 Between the two vertical lines shown, the speed was now steady at 68Km/h. Under this condition, the peak alarm threshold value predetermined according to past experience is: slight fault [15g-18 g), medium fault [18g-22 g), and serious fault greater than or equal to 22g. The method comprises the following steps:
and the external power supply supplies power to start the equipment switch.
And inputting values of a starting position 831 and an ending position 923 through a man-machine interaction panel of the system. The location parameters 831 and 923 indicate that the system analyzes only data collected between the sections 831 to 923 as valid diagnostic data, while the data of the remaining sections need not be analyzed. Meanwhile, three threshold parameters of 18g, 20g and 25g of slight faults, 20g of moderate faults and severe faults of the traction motor bearing at the road section and the rotating speed are imported into the system through storage equipment and used for judging the fault degree.
After the train is started, the system acquires real-time vibration data, rotating speed data and GPS data acquired by the sensor through the sensor interface.
The original data are segmented and screened in real time through the cache and preprocessing module, 1 second vibration data, GPS data and rotating speed data are cached every time, whether the GPS position of the train is located between preset GPS positions [831, 923] in the current 1 second moment operation stage is judged, and if the GPS position is located between the preset GPS positions, the collected vibration data, GPS data and rotating speed data are transmitted to the calculation module and the diagnosis module.
In the calculation module, the present example calculates the frequency domain indicators mean, standard deviation, variance, skewness indicator, kurtosis indicator, peak-to-peak, mean square amplitude, average amplitude, square root amplitude, waveform indicator, peak indicator, and also calculates the frequency spectrum, envelope spectrum, and cepstrum in real time. In this example, the peak value index is used as the failure degree determination index.
And analyzing the data in real time through a diagnosis module, comparing the current peak value index with the threshold values in the parameter library, and finding that the vibration peak value of the current road section is between a slight fault threshold value 15g and a moderate fault threshold value 18g, so that the system judges the current road section to be a slight fault. Meanwhile, the frequency spectrum is analyzed according to an intelligent algorithm, and the fault of the retainer is diagnosed; finally, the residual service life of the bearing with the fault is deduced to be 2601 hours by comparing the evolution rule of the past fault.
Because the bearing has a fault, GPS data, rotating speed data, vibration data, time domain indexes and frequency spectrums are all stored in an onboard storage device.
The alarm system drives the buzzer to give an alarm, lights the first LED which represents the slight fault of the retainer in the LED lamp array, and simultaneously displays the 'slight fault of the retainer, and the estimated operation time 2601 hour' on the liquid crystal display screen.
Through wireless network, remote cell-phone APP receives current traction motor's real-time running state index in real time to receive current warning suggestion: "bearing failure is a minor foreign failure, and the current bearing is expected to still be usable for 2601 hours".
As shown in fig. 3, a diagnostic system for implementing the method for diagnosing a bearing and gear fault of a railway locomotive includes,
the data acquisition module is arranged at a preset position of a locomotive bearing and a gear to acquire real-time data;
a cache preprocessing module, which is connected with the data acquisition module and comprises,
a pre-processing unit that screens the real-time data and divides it into data segments of a predetermined interval duration,
a cache unit connected to the preprocessing unit, the cache unit configured to cache the data segment;
a computing module, connected to the cache preprocessing module, comprising,
a time domain index calculation unit generating a time domain index array based on the filtered data segments,
a spectrum calculation unit which generates a frequency domain matrix based on the filtered data segment;
and the diagnosis module is connected with the calculation module, judges the fault degree and calculates the residual life based on the time domain index array self-learning training, and judges the fault type and position based on the frequency domain matrix through self-learning training.
In one embodiment of the diagnostic system, the system further comprises,
the power module comprises a vibration energy collecting unit, a rechargeable battery unit and an external power supply, and if the external power supply is connected, the power supply is directly used for supplying power and charging the rechargeable battery; if the power supply is not connected, the rechargeable battery supplies power to the system according to the stored electric quantity, and meanwhile, the vibration energy collecting unit acquires vibration energy to convert the vibration energy into electric energy and charges the rechargeable battery;
the storage transmission module is connected with the diagnosis module and is configured to store and transmit fault information, the transmission mode comprises a wired mode and a wireless mode, and the fault information comprises fault degree, residual service life, fault type and position;
the alarm module comprises a plurality of alarm modules and a plurality of alarm modules,
an alarm unit that issues an alarm based on the degree of the failure,
a display unit that displays failure information;
an input-output interface, comprising,
a sensor interface for reading the data of the vibration acceleration sensor, the rotating speed measuring sensor and the GPS sensor,
a data interface for exchanging data with an external storage device,
and the output interface transmits the airborne data to the remote equipment in a wired mode.
In one embodiment of the diagnostic system, the data acquisition module comprises a vibration acceleration sensor, a rotating speed measurement sensor and a GPS sensor; the storage transmission module comprises a storage module for storing fault information and a transmission module for transmitting the fault information; the alarm unit comprises a buzzer and an LED lamp; the display unit includes a liquid crystal display. When a fault occurs, the storage module stores the data segment and the fault information, the transmission module transmits the data segment and the fault information, the alarm unit gives an alarm, and the liquid crystal display screen displays the fault information.
In an embodiment of the diagnosis system, as shown in fig. 4, the diagnosis module is wirelessly connected to a mobile terminal, and the mobile terminal includes a computer, a mobile phone, a bracelet, a large screen, and a cloud server.
To further understand the present invention, in one embodiment, the diagnostic system includes a power module, a data acquisition module, a cache preprocessing module, a calculation module, a diagnostic module, a storage transmission module, an onboard alarm module, a sensor interface, a data interface, and an output interface.
In one embodiment, the monitored objects are bearings and gears of a railway locomotive, and a vibration acceleration sensor, a rotating speed measuring sensor and a GPS sensor are arranged at corresponding positions of the bearings and the gears of the locomotive. After the system is powered on, the equipment is turned on, real-time data acquired by the sensor is acquired through the sensor interface, and the original data is preprocessed through the cache and preprocessing module; transmitting the preprocessed effective data into a calculation module, and calculating a time domain index and a frequency spectrum; inputting the calculation result into a diagnosis module, and comparing the difference between the current time domain index and the frequency spectrum with the difference in a preset parameter library to obtain a diagnosis result of the fault type and the fault degree; and according to whether the fault occurs or not, carrying out corresponding early warning display in the onboard alarm module, storing the diagnosis result and corresponding data in a local storage unit, and sending the diagnosis result and the corresponding data to a remote terminal in a wired or wireless mode.
In one embodiment, the power module comprises a vibration energy harvesting unit, a rechargeable battery unit and an external power supply, and the three units jointly supply power to the system. The power supply sequence is that if the external power supply is connected, the power supply is directly used for supplying power and the rechargeable battery is charged; if the power supply is not connected, the rechargeable battery supplies power to the system according to the stored electric quantity, meanwhile, the vibration energy collecting module starts to work continuously, vibration energy is obtained from the outside, the vibration energy is converted into electric energy, and the rechargeable battery is charged.
In one embodiment, the data acquisition module mainly acquires key operation parameters of locomotive operation, and in one embodiment, the data acquisition module comprises vibration acceleration acquisition, rotating speed acquisition and GPS acquisition.
In one embodiment, as shown in fig. 5, the cache preprocessing module preprocesses data according to three types of determination criteria, divides the collected continuous data into data segments with a duration of 1-5 seconds, screens valid data therein, discards invalid data, reduces data size, and facilitates real-time calculation by the subsequent calculation module. The running route of the railway locomotive is long, the running working conditions are complex, but the railway track is fixed, and the difference of the running working conditions at different times is not very large. Through past experience, the whole process of locomotive operation can be divided into three types, the first type is a common operation process, and the judgment conditions are as follows: the speed of the locomotive reaches more than 30% of the rated running speed and is not in the process of acceleration and deceleration. When the condition is met, the train is in a normal running process; the second type is a process of straight track and stable rotation speed, and the judgment conditions are as follows: the train does not turn, does not accelerate or decelerate, and the rotating speed is the rated speed, and is in the optimal running state, at this stage, the system reminds the train conductor to carry out the stability adjustment of the running speed of the train, or can be accessed to a micro-program controller (MCU) platform to carry out the stable control of the train speed. The road section meeting the condition can be determined through the past operation experience, and the data collected on the road section is most valuable for fault diagnosis of the bearing and the gear; the third type is a smooth running process, and the judgment condition is H3: the speed of rotation is smooth but the railway track is not necessarily straight. The path length meeting the judgment condition is long, so that the data is large, and the method has important value for fault diagnosis. For the three road sections, the corresponding cache preprocessing module has three modes: an active local detection mode (corresponding to H2), a stationary detection mode corresponding to H3, and a global detection mode corresponding to H1, wherein the selection of the modes is set by a function key in the panel.
The calculation module calculates the data characteristic indexes and frequency spectrums output by the cache preprocessing module in real time, and in order to diagnose the faults of the bearing and the gear more effectively, the corresponding state indexes are calculated after filtering is carried out in a frequency bandwidth of 20Hz around the frequency where the fault is most likely to occur to the gear and the bearing, and the calculation module mainly comprises a time domain index calculation unit and a frequency spectrum calculation unit.
The time domain index calculation unit includes: mean, standard deviation, variance, skewness index, kurtosis index, peak-to-peak value, mean-square amplitude, average amplitude, square root amplitude, waveform index, peak index. In addition, the following three status indicators are included:
state index 1:
wherein F represents the filtered vibration signal, N is the total number of points in F, m and i are both positive integers F' representing the derivative of F,
the average value of F' is shown.
The state index 2:
wherein F' represents the second derivative of F,
represents the average value of F'.
The state index 3:
wherein, delta E i ＝F 2 i+1 -F 2 i ，
Is the average value of Δ E.
The spectrum calculating unit comprises real-time calculation of various spectrums, including FFT, envelope spectrum, cepstrum and energy spectrum.
And the diagnosis module automatically diagnoses the fault by adopting an intelligent learning method according to the calculation results of the time domain index calculation module and the frequency spectrum calculation module and in combination with the parameters of a preset parameter library, and predicts the residual life of the fault component. The fault degree is judged mainly according to the comparison between a specific time domain index and a set threshold value, and the fault type and the fault position are judged according to whether the frequency spectrum graph contains the frequency multiplication of the frequency of the fault component. The setting of the fault degree judging threshold value in the diagnosis module is realized through a data panel, and the set parameters in the external data storage device can be directly read into the system through a data interface. The invention adopts two modes of a convolutional neural network and a knowledge vector machine to train and perfect the diagnosis module, the accuracy rate is continuously improved along with the gradual accumulation of data, and the diagnosis module is ensured to be used better and more accurate.
The overall calculation results of the calculation module and the diagnosis module are stored according to the format shown in the table 1.
TABLE 1 data Format of the results of the calculations
And n is the number of different road sections and time periods divided by the cache preprocessing module.
As shown in fig. 6, the storage and transmission module is responsible for storing and sending the calculation result in table 1, the transmission modes include a wired transmission mode and a wireless transmission mode, and the onboard alarm module or the remote terminal displays the result in real time through the buzzer, the LED lamp array, the LCD bar screen or the rest of the mobile terminals according to the diagnosis result in table 1.
According to different diagnosis results, the system has different fault processing procedures, if the diagnosis result is that a fault exists, all data in the table 1 are stored in a storage unit, and are sent to a terminal in a wired or wireless mode, and are displayed and alarmed on an onboard device and a remote terminal. And if the fault does not exist, only storing and sending the index array in the table 1, discarding the rest data, and displaying no alarm on the onboard equipment and the remote terminal.
Although the embodiments of the present invention have been described above with reference to the accompanying drawings, the present invention is not limited to the above-described embodiments and application fields, and the above-described embodiments are illustrative, instructive, and not restrictive. Those skilled in the art, having the benefit of this disclosure, may effect numerous modifications thereto without departing from the scope of the invention as defined by the appended claims.
Claims (1)
1. A diagnostic system for railway locomotive bearing and gear faults is characterized in that:
the system is used for executing the following railway locomotive bearing and gear fault diagnosis method, and the method comprises the following steps:
in a first step (S1), a vibration acceleration sensor, a rotation speed measuring sensor and a GPS sensor are arranged at preset positions of a locomotive bearing and a gear to acquire real-time data;
in the second step (S2), when the speed of the locomotive reaches more than the preset percentage of the rated operation speed and is not in the acceleration and deceleration process, the locomotive is judged to be in the normal operation process, when the real-time data from the preset position is sampled, when the change of the rotating speed and the vibration is more than a second preset value, the speed of the locomotive is adjusted to ensure that the change of the rotating speed and the vibration is less than a second preset range, the real-time data is continuously sampled and is divided into data sections with preset interval duration, the data sections are cached, the preset interval duration is 1 second, when the change of the speed of the locomotive is less than a preset speed amplitude value, the locomotive is judged to be in the stable operation process, when the change of the rotating speed and the vibration is less than the first preset range, the real-time data in the normal operation process is sampled and is divided into the data sections with the preset interval duration, and the data sections are cached;
in a third step (S3), the data segment is filtered, and a time domain index array and a frequency domain matrix are generated based on the filtered data segment, where the time domain index array at least includes a first index, a second index, and a third index, and the frequency domain matrix includes a cepstrum and an energy spectrum, where the first index, the second index, and the third index are respectively:
wherein F represents the filtered vibration signal, N is the total number of points in F, m and i are both positive integers, F' represents the derivative of F,
denotes the average value of F ', F' denotes the second derivative of F,
denotes the average value of F', delta E i ＝F 2 i+1 -F 2 i ，
Is the average value of Delta E, and the time domain index array further comprises the average value, standard deviation, variance, skewness index and kurtosis indexPeak-to-peak value, mean square amplitude, average amplitude, square root amplitude, waveform index and peak index;
in the fourth step (S4), fault degree and residual life are judged through self-learning training based on the time domain index array, and fault types and positions are judged through self-learning training based on the frequency domain matrix;
in addition, the method completes the calculation of the fault characteristic index and the calculation of the spectrogram in the airborne hardware, realizes the real-time fault diagnosis in the airborne hardware, and displays the fault diagnosis result at the local terminal and the remote terminal, thereby not depending on the calculation capability of the external terminal equipment;
in the second step (S2), the real-time data in the normal operation process is resampled at equal angles and divided into data segments with preset interval duration, and the data segments are cached;
in the second step (S2), when the locomotive does not turn and is not accelerated or decelerated and the rotating speed is a rated speed, determining that the locomotive reaches a preset position based on a GPS sensor, sampling real-time data from the preset position, dividing the real-time data into data sections with preset interval duration, and caching the data sections;
in the third step (S3), in the fourth step (S4), self-learning training is carried out through a convolutional neural network and a knowledge vector machine;
wherein the system comprises:
the data acquisition module is arranged at a preset position of a locomotive bearing and a gear to acquire real-time data;
a cache preprocessing module, which is connected with the data acquisition module and comprises,
a pre-processing unit that screens the real-time data and divides it into data segments of a predetermined interval duration,
a cache unit connected to the preprocessing unit, the cache unit configured to cache the data segment;
a computing module, connected to the cache preprocessing module, comprising,
a time domain index calculation unit generating a time domain index array based on the filtered data segments,
a spectrum calculation unit which generates a frequency domain matrix based on the filtered data segment;
the diagnosis module is connected with the calculation module, judges the fault degree and calculates the residual life based on the time domain index array self-learning training, and judges the fault type and position based on the frequency domain matrix through the self-learning training;
the diagnosis system is arranged in the airborne hardware, realizes real-time fault diagnosis in the airborne hardware, and displays fault diagnosis results at local and remote terminals, so that the diagnosis system does not depend on the computing capacity of external terminal equipment;
wherein the content of the first and second substances,
the system further comprises a control unit for controlling the operation of the motor,
the power module comprises a vibration energy collecting unit, a rechargeable battery unit and an external power supply, and if the external power supply is connected, the power supply is directly used for supplying power and charging the rechargeable battery; if the power supply is not connected, the rechargeable battery supplies power to the system according to the stored electric quantity, and meanwhile, the vibration energy collecting unit acquires vibration energy to convert the vibration energy into electric energy and charges the rechargeable battery;
the storage transmission module is connected with the diagnosis module and is configured to store and transmit fault information, the transmission mode of the storage transmission module comprises a wired mode and a wireless mode, and the fault information comprises fault degree, residual life, fault type and position;
the alarm module comprises an alarm module and a control module,
an alarm unit that issues an alarm based on the degree of the failure,
a display unit that displays failure information, and the display unit includes local and remote terminals that perform display of failure diagnosis results;
an input-output interface, comprising,
a sensor interface for reading the data of the vibration acceleration sensor, the rotating speed measuring sensor and the GPS sensor,
a data interface for exchanging data with an external storage device,
the output interface transmits the airborne data to the remote equipment in a wired mode;
the data acquisition module comprises a vibration acceleration sensor, a rotating speed measurement sensor and a GPS sensor; the storage transmission module comprises a storage module for storing fault information and a transmission module for transmitting the fault information; the alarm unit comprises a buzzer and an LED lamp; the display unit comprises a liquid crystal display screen, when a fault occurs, the storage module stores the data segment and the fault information, the transmission module transmits the data segment and the fault information, the alarm unit gives an alarm, and the liquid crystal display screen displays the fault information;
the diagnosis module is wirelessly connected with the mobile terminal;
and also,
the system is integrated in a hardware unit, such that: carrying out real-time early warning and diagnosis on the fault in real time; an external storage unit, a button group, an LED and a liquid crystal display screen are introduced to realize parameter setting and result display of fault diagnosis, and fault early warning and diagnosis can be carried out without depending on any external equipment;
the system adopts a cache and selective mechanism, so that a computing unit can conveniently and rapidly compute real-time characteristic indexes, only the computed characteristic indexes are stored and sent when no fault exists, and original data and frequency spectrum can be discarded, thereby greatly reducing the data volume;
the cache preprocessing module preprocesses data according to three types of judgment criteria, divides the acquired continuous data into data segments with the time length of 1-5 seconds, screens effective data in the data segments, discards invalid data, reduces the data volume and facilitates real-time calculation of a subsequent calculation module; the running route of the railway locomotive is long, the running working conditions are complex, but the railway track is fixed, the difference of the running working conditions at different times is not very large, and the whole process of the locomotive running can be divided into three types through the past experience:
the first type is a common operation process, and the judgment condition is H1: the speed of the locomotive reaches more than 30 percent of the rated running speed and is not in the process of acceleration and deceleration; when the condition is met, the train is in a normal running process;
the second type is the process of straight track and stable rotating speed operation, and the judgment conditions are H2: the train does not turn, does not accelerate or decelerate, and the rotating speed is a rated speed, and is in an optimal running state, at this stage, the system reminds the train conductor to carry out the stability adjustment of the running speed of the train, or can be accessed to a Microprogrammed Control Unit (MCU) platform to carry out the stable control of the train speed; the road section meeting the condition can be determined through the past operation experience, and the data collected on the road section is most valuable for fault diagnosis of the bearing and the gear;
the third type is a smooth running process, and the judgment condition is H3: the rotating speed is smooth, but the railway track is not necessarily straight; the road section meeting the judgment condition is long, so that the data is more, and the method also has important value for fault diagnosis; for the three operation processes, the corresponding cache preprocessing module has three modes: the active local detection mode corresponds to H2, the stable detection mode corresponds to H3 and the global detection mode corresponds to H1, wherein the mode is selected through a function key in a panel for setting;
this enables the system to have three modes: the system comprises an active local detection mode, a steady detection mode and a global detection mode, and can be switched randomly, wherein the mode switching function is realized by a key set, the active local detection mode only acquires vibration data of a steady rotating speed and a straight track part, the data quality is best, and the diagnosis result has the most referential property; the steady detection mode only adopts the data of the steady section of the rotating speed, the data volume is large, and the diagnosis results can be mutually verified so as to improve the reliability of diagnosis; the global detection mode ensures the monitoring of the whole operation process and emphasizes the integrity of time domain index monitoring, but consumes time; three modes are adopted to ensure the omnibearing monitoring of fault diagnosis and ensure the operation safety to the maximum extent;
wherein the content of the first and second substances,
when the monitored objects are bearings and gears of a railway locomotive, arranging a vibration acceleration sensor, a rotating speed measuring sensor and a GPS sensor at corresponding positions of the bearings and the gear parts of the locomotive; after the system is powered on, the equipment is turned on, real-time data acquired by the sensor is acquired through the sensor interface, and the original data is preprocessed through the cache and preprocessing module; transmitting the preprocessed effective data into a calculation module, and calculating a time domain index and a frequency spectrum; inputting the calculation result into a diagnosis module, and comparing the difference between the current time domain index and the frequency spectrum with the difference in a preset parameter library to obtain a diagnosis result of the fault type and the fault degree; according to whether the fault occurs or not, carrying out corresponding early warning display in the onboard alarm module, storing the diagnosis result and corresponding data in a local storage unit, and sending the diagnosis result and the corresponding data to a remote terminal in a wired or wireless mode;
in the calculation module, the mean value, the standard deviation, the variance, the skewness index, the kurtosis index, the peak-peak value, the mean square amplitude, the average amplitude, the square root amplitude, the waveform index and the peak index of the frequency domain indexes are calculated, in addition, the frequency spectrum, the envelope spectrum and the cepstrum are also calculated in real time, wherein the peak index is used as the fault degree judgment index;
when the calculation module calculates the data characteristic indexes and the frequency spectrums output by the cache preprocessing module in real time, in order to diagnose the faults of the bearing and the gear more effectively, the corresponding state indexes are calculated after filtering is carried out in the frequency bandwidth of 20Hz around the frequency where the fault is most likely to occur on the gear and the bearing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910256518.4A CN110044616B (en) | 2019-04-01 | 2019-04-01 | Fault diagnosis method and system for bearing and gear of railway locomotive |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910256518.4A CN110044616B (en) | 2019-04-01 | 2019-04-01 | Fault diagnosis method and system for bearing and gear of railway locomotive |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110044616A CN110044616A (en) | 2019-07-23 |
| CN110044616B true CN110044616B (en) | 2023-01-03 |
Family
ID=67275740
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910256518.4A Active CN110044616B (en) | 2019-04-01 | 2019-04-01 | Fault diagnosis method and system for bearing and gear of railway locomotive |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110044616B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113473510B (en) * | 2020-03-30 | 2024-05-03 | 大唐移动通信设备有限公司 | Data extraction method and device for positioning transmission link faults and electronic equipment |
| CN112232405B (en) * | 2020-10-13 | 2022-09-02 | 中车青岛四方机车车辆股份有限公司 | Fault prediction, monitoring and diagnosis method of gearbox and corresponding device |
| CN113065097B (en) * | 2021-03-18 | 2022-01-04 | 湖南南方宇航高精传动有限公司 | Method for calculating load coefficient of load-sharing test of planetary gear box |
| CN115033609B (en) * | 2022-08-15 | 2022-11-18 | 江苏苏港智能装备产业创新中心有限公司 | Method, device and equipment for monitoring and analyzing vibration of hoisting mechanism under unsteady state load and storage medium |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101274636A (en) * | 2007-03-26 | 2008-10-01 | 林贵生 | Vehicle-mounted intelligent supervising and early warning device for running status of track traffic vehicle |
| CN103809586A (en) * | 2014-02-21 | 2014-05-21 | 北京六捷科技有限公司 | Fault analyzing and diagnosing system and method of railway dispatching orders |
| CN104535323B (en) * | 2015-01-12 | 2017-03-08 | 石家庄铁道大学 | A kind of train wheel Method for Bearing Fault Diagnosis based on angular domain time-domain and frequency-domain |
| CN106153335A (en) * | 2015-02-10 | 2016-11-23 | 中国科学院声学研究所 | A kind of train bearing acoustics online system failure diagnosis and method |
| CN108318257A (en) * | 2018-01-30 | 2018-07-24 | 中道恒通（北京）科技有限公司 | Locomotive vibration monitors system |
| CN108583624B (en) * | 2018-04-12 | 2020-11-24 | 中车青岛四方机车车辆股份有限公司 | Train running state visualization method and device |
| CN109186964B (en) * | 2018-07-28 | 2019-12-17 | 华中科技大学 | rotary machine fault diagnosis method based on angle resampling and ROC-SVM |
-
2019
- 2019-04-01 CN CN201910256518.4A patent/CN110044616B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110044616A (en) | 2019-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110044616B (en) | Fault diagnosis method and system for bearing and gear of railway locomotive | |
| CN109144014B (en) | System and method for detecting operation condition of industrial equipment | |
| CN107884214B (en) | A kind of train EEF bogie unit failure multi-parameter comprehensive decision-making technique and device | |
| CN103700160A (en) | Motor vehicle onboard terminal based on microsensor and driving behavior judgment method | |
| CN112061988B (en) | Crane distributed network monitoring and single machine monitoring management system and method | |
| CN107449508A (en) | Automobile vibration fault detection system and detection data analysing method | |
| CN110469462A (en) | A kind of Wind turbines intelligent condition monitoring system based on multi-template | |
| CN105976630A (en) | Vehicle speed monitoring method and device | |
| CN111509847A (en) | Intelligent detection system and method for power grid unit state | |
| CN112327807A (en) | Automobile fault diagnosis and monitoring system and method based on IVI host | |
| CN110095179A (en) | A kind of detection of motor oscillating and diagnostic method | |
| CN110901695A (en) | Train stability detection system and method | |
| CN110456694A (en) | A kind of transport intelligent monitor system of wind power generating set and component | |
| CN109932178A (en) | A kind of multiple strand chain drive failure diagnostic system and method | |
| CN102506016A (en) | Wind power station equipment status inspection device | |
| WO2020060853A1 (en) | Using vehicle electrical system monitored values | |
| CN116880454A (en) | Intelligent diagnosis system and method for vehicle faults | |
| CN113442728B (en) | Comprehensive driving mileage management system of electric vehicle | |
| CN116338500B (en) | Rail transit vehicle storage battery operation monitoring system | |
| CN108389282A (en) | A kind of vehicle mounted failure self diagnosis early warning cloud system | |
| KR20220121932A (en) | System for monitoring battery of vehicle | |
| CN105988460A (en) | Dynamic track detection method, apparatus, and system for vehicle | |
| CN111791919A (en) | Method and system for monitoring state of top cover | |
| CN114757366B (en) | Fault prediction method and system for vehicle | |
| CN211642196U (en) | Train stationarity detecting system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |