A high-speed rail running part working condition load recording system based on voiceprint features

Abstract

The application relates to a high-speed rail running part working condition load recording system based on a voiceprint feature, relates to the field of high-speed rail running parts, and comprises a host computer and a plurality of monitoring terminals for detecting wheel hubs; one monitoring terminal is placed at a position opposite to one wheel hub, the monitoring terminal is used for collecting sound and vibration generated by the corresponding wheel hub in the driving process to output sound detection signals and vibration detection signals; the host computer is arranged on the high-speed rail, is connected with the plurality of monitoring terminals, is used for analyzing the sound detection signals and the vibration detection signals, judging whether abnormal sound and vibration are generated by the wheel hub in the driving process of the high-speed rail, storing generated abnormal detection data, and transmitting the abnormal detection data to a data platform when the high-speed rail stops. The application can reduce the cost of monitoring the high-speed rail running part.

Description

Technical Field

[0001] This application relates to the field of high-speed railway running gear, and in particular to a high-speed railway running gear operating condition load recording system based on voiceprint characteristics. Background Technology

[0002] With the rapid development of high-speed rail technology, high-speed trains are traveling at increasingly higher speeds. This means that any malfunction on a high-speed train traveling at such a high speed will inevitably seriously endanger railway safety and cause significant losses. Therefore, fault detection of high-speed trains is absolutely essential.

[0003] In related technologies, each running gear includes four hubs, and each hub is equipped with a monitoring device. The monitoring device is used to monitor the sound and vibration generated by the corresponding hub, and analyze the monitored sound and vibration to determine whether the hub has produced abnormal noises or vibrations, and thus determine whether the hub has malfunctioned.

[0004] Because there are many running gears on a high-speed train, and each running gear needs to be equipped with four of the aforementioned monitoring devices, the cost of monitoring all running gears on a high-speed train is relatively high. Summary of the Invention

[0005] To reduce the cost of monitoring the running gear of high-speed railways, this application proposes a high-speed railway running gear operating condition load recording system based on voiceprint characteristics.

[0006] The high-speed rail running gear load recording system based on voiceprint features provided in this application adopts the following technical solution:

[0007] A high-speed rail running gear condition load recording system based on acoustic signature features includes a main unit and multiple monitoring terminals for detecting wheel hubs;

[0008] A monitoring terminal is placed opposite to a wheel hub. The monitoring terminal is used to collect the sound and vibration generated by the corresponding wheel hub during driving, and to output sound detection signal and vibration detection signal.

[0009] The host is installed on the high-speed train and connected to multiple monitoring terminals. It is used to analyze the sound detection signal and vibration detection signal to determine whether the wheel hub produces abnormal noises and vibrations during the high-speed train's operation. It is used to store the generated abnormal detection data and transmit the abnormal detection data to the data platform when the high-speed train stops.

[0010] By adopting the above technical solution, the monitoring terminal can collect the sound and vibration generated by the wheel hub during operation, and the host can analyze the sound detection signal and vibration detection signal to determine whether the high-speed train is generating sound and vibration from the wheel hub during operation. Compared with related technologies, this application can realize the acquisition function and the analysis function through two separate devices, thereby reducing the cost of monitoring the running gear of the high-speed train.

[0011] In one possible implementation: the host includes a GPS module, a speed monitoring unit, an analysis and processing unit, and a storage unit;

[0012] The GPS module is used to locate the position of the high-speed train;

[0013] The speed monitoring unit is used to monitor the speed of the high-speed train in real time;

[0014] The storage unit is used to store multiple standard models, each standard model corresponding to a vehicle type;

[0015] The analysis and processing unit is connected to the storage unit, the speed monitoring unit, and multiple monitoring terminals. It is used to analyze the travel speed, sound detection signal, and vibration detection signal at the same time according to a standard model, to determine whether the high-speed train has abnormal noises and vibrations from the wheel hub during its operation, so as to generate abnormal detection data and transmit the abnormal detection data to the data platform when the high-speed train stops.

[0016] The storage unit is also used to store the anomaly detection data.

[0017] In one possible implementation: the analysis and processing unit is further configured as follows:

[0018] Acquire sound and vibration data for each wheel hub. Both the sound and vibration data carry a time identifier, a speed identifier, a location identifier, and a wheel hub number. The location identifier includes the longitude, latitude, and altitude of the location data.

[0019] The sound data is divided into a first static group and a first dynamic group based on the positioning identifier and the time identifier, and the vibration data is divided into a second static group and a second dynamic group.

[0020] Based on the standard model, the data of the first static group and the data of the second static group are analyzed to obtain the first analysis result for each wheel hub;

[0021] The static sound pattern and static vibration wave pattern of each wheel hub are determined according to the positioning mark and the speed mark. The static sound pattern is the sound pattern generated when the wheel hub is stationary, and the static vibration wave pattern is the vibration wave pattern generated when the wheel hub is stationary.

[0022] Based on the static acoustic pattern, the static vibration ripple, and the standard model, the data of the first dynamic group and the data of the second dynamic group are analyzed to obtain the second analysis result for each wheel hub.

[0023] Anomaly detection data is generated by combining the first analysis result and the second analysis result.

[0024] In one possible implementation: the analysis and processing unit is further configured as follows:

[0025] The step of analyzing the data of the first dynamic group and the data of the second dynamic group based on the static sound pattern, the static vibration ripple, and the standard model to obtain the second analysis result for each wheel hub includes:

[0026] Based on the static acoustic pattern, static vibration wave pattern, standard model, sound data, and vibration data of the same wheel hub, the environmental acoustic pattern and environmental vibration wave pattern of the high-speed rail at different locations on the railway are determined. The environmental acoustic pattern is the acoustic pattern caused by the environment to the wheel hub, and the environmental vibration wave pattern is the vibration wave pattern caused by the environment to the wheel hub.

[0027] Select a sound data or a vibration data from the first dynamic group or the second dynamic group;

[0028] Select environmental sound patterns or environmental vibration patterns based on the location identifiers and wheel hub number identifiers of the selected sound or vibration data.

[0029] The selected sound data or vibration data are analyzed based on the standard model, the static sound pattern or the static vibration wave pattern, and the environmental sound pattern or the environmental vibration wave pattern to obtain a judgment result. The judgment result is used to reflect whether the wheel hub produces abnormal noise or abnormal vibration at a certain speed.

[0030] The second analysis result is obtained by combining all the judgment results for the same wheel hub.

[0031] In one possible implementation: the analysis and processing unit is further configured as follows:

[0032] The determination of the environmental sound signature and environmental vibration wave at different locations on the railway based on the static sound signature, static vibration wave, standard model, sound data, and vibration data of the same wheel hub includes:

[0033] Select the sound and vibration data of the first N wheel hubs;

[0034] Based on the sound data, the static acoustic signature, and the standard model, determine the environmental acoustic signature reference value for each wheel hub at each location;

[0035] The mode value of the ambient sound reference value when each wheel hub is in the same position is taken as the ambient sound of that position.

[0036] Based on the vibration data, the static vibration ripples, and the standard model, determine the reference value of the environmental vibration ripples for each wheel hub at each location;

[0037] The mode value of the environmental vibration ripple reference value when all wheel hubs are located at the same position is taken as the environmental vibration ripple at that position.

[0038] In one possible implementation: the analysis and processing unit is further configured as follows:

[0039] The method for determining the environmental acoustic reference value of a wheel hub at a location based on the sound data, the static acoustic signature, and the standard model is as follows:

[0040] The static acoustic signature of the wheel hub and the acoustic signature corresponding to the speed identifier of the sound data in the standard model are filtered out from the sound data to obtain the environmental acoustic signature reference value.

[0041] The method for determining the reference value of environmental vibration ripple for a wheel hub at a location based on the vibration data, the static vibration ripple, and the standard model is as follows:

[0042] The static vibration ripples of the wheel hub and the vibration ripples corresponding to the velocity labels of the vibration data in the standard model are filtered out from the vibration data to obtain the environmental vibration ripple reference value.

[0043] In one possible implementation: the analysis and processing unit is further configured as follows:

[0044] The step of analyzing the selected sound data or vibration data based on the standard model, the static sound pattern or the static vibration ripple, and the environmental sound pattern or the environmental vibration ripple to obtain the judgment result includes:

[0045] The target soundprint is determined based on the selected sound data, the static soundprint of the corresponding wheel hub, and the environmental soundprint whose positioning identifier matches the positioning identifier of the sound data.

[0046] The voiceprint in the standard model that matches the speed identifier of the sound data is compared with the target voiceprint to obtain a judgment result;

[0047] The target vibration ripple is determined based on the selected vibration data, the static vibration ripple of the corresponding wheel hub, and the environmental vibration ripple whose positioning mark matches the positioning mark of the vibration data.

[0048] The vibration ripples in the standard model that match the velocity identifier of the vibration data are compared with the target vibration ripples to obtain the judgment result.

[0049] In one possible implementation: the analysis and processing unit is further configured as follows:

[0050] The step of determining the target soundprint based on the selected sound data, the static soundprint of the corresponding wheel hub, and the ambient soundprint whose positioning identifier matches the positioning identifier of the sound data includes:

[0051] The target soundprint is obtained by filtering out the static soundprint of the corresponding wheel hub and the environmental soundprint whose positioning identifier is consistent with the positioning identifier of the sound data from the selected sound data.

[0052] The step of determining the target vibration ripple based on the selected vibration data, the static vibration ripple of the corresponding wheel hub, and the environmental vibration ripple whose positioning marker matches the positioning marker of the vibration data includes:

[0053] The target vibration ripple is obtained by filtering out the static vibration ripples of the corresponding wheel hub and the environmental vibration ripples whose positioning marks are consistent with the positioning marks of the vibration data from the selected vibration data.

[0054] In one possible implementation: the analysis and processing unit is further configured as follows:

[0055] The step of dividing the sound data into a first static group and a first dynamic group based on the positioning identifier and the time identifier, and dividing the vibration data into a second static group and a second dynamic group, includes:

[0056] The data with the same altitude and consecutive time markers in the sound data are identified as the first static group, and the remaining data in the sound data are identified as the first dynamic group.

[0057] The vibration data with the same altitude and continuous time markers are identified as the second static group, and the remaining data in the vibration data are identified as the second dynamic group.

[0058] In one possible implementation: the analysis and processing unit is further configured as follows:

[0059] The process of determining the static sound signature and static vibration ripple of each wheel hub based on the positioning identifier and the speed identifier includes:

[0060] The voiceprints in the sound data whose location identifier is a specified location and whose speed identifier is zero are identified as static voiceprints.

[0061] The vibration ripples in the vibration data that are located at a specified position and have a velocity of zero are identified as static vibration ripples.

[0062] In summary, this application includes at least one of the following beneficial technical effects:

[0063] In this application, by setting up a host and multiple monitoring terminals, the two processes of collecting the sound and vibration generated by the wheel hub during operation and analyzing the sound detection signal and vibration detection signal can be carried out in two devices, thereby reducing the cost of monitoring the running gear of high-speed trains. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of a high-speed rail running gear load recording system based on voiceprint features, according to an embodiment of this application.

[0065] Figure 2 This is a schematic diagram of the data interaction between the high-speed rail running gear condition load recording system based on voiceprint features and the data platform according to an embodiment of this application.

[0066] Figure 3 This is a flowchart illustrating the analysis algorithm configured in the analysis processing unit of this application embodiment.

[0067] Explanation of reference numerals in the attached diagram: 1. Main unit; 11. Verification unit; 12. Speed ​​monitoring unit; 13. Storage unit; 14. Analysis and processing unit; 15. Mobile communication unit; 16. GPS module; 2. Monitoring terminal; 21. Sound monitoring unit; 22. Vibration monitoring unit; 23. Transmission unit; 24. Wireless communication unit; 25. Sound calibration unit; 26. Vibration calibration unit; 3. Signal repeater; 4. Data platform. Detailed Implementation

[0068] To make the purpose, technical solution, and advantages of this application clearer, the following description is provided in conjunction with the appendix. Figure 1-2 The present application will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.

[0069] This application discloses a high-speed rail running gear load recording system based on voiceprint features. (Refer to...) Figure 1 The high-speed rail running gear condition load recording system based on acoustic signature features includes a main unit 1 and multiple monitoring terminals 2. The multiple monitoring terminals 2 can collect the sound and vibration generated by the wheel hubs during high-speed rail operation. The main unit 1 can analyze the sound and vibration collected by the multiple monitoring terminals 2, allowing the sound and vibration collection process and the analysis process to be performed separately in two devices, thereby reducing the cost of monitoring the high-speed rail running gear.

[0070] Specifically, the monitoring terminal 2 is used to collect the sound and vibration generated by the wheel hub during driving, so as to output sound detection signals and vibration detection signals, and transmit sound detection signals and vibration detection signals. It includes a sound monitoring unit 21, a vibration monitoring unit 22, a transmission unit 23 and a wireless communication unit 24.

[0071] The sound monitoring unit 21 is used to collect the sound generated by the wheel hub during driving and output a sound detection signal. The sound monitoring unit 21 can be a device with audio acquisition function, such as a microphone or an audio acquisition device. In this embodiment, the sound monitoring unit 21 is a sound sensor.

[0072] The vibration monitoring unit 22 is used to collect the vibration generated by the wheel hub during driving and output a vibration detection signal. The vibration monitoring unit 22 is specifically a device with vibration collection function, such as a vibration sensor or a vibration acquisition device.

[0073] The transmission unit 23 is connected to the sound monitoring unit 21 and the vibration monitoring unit 22 respectively, and is used to receive and transmit sound detection signals and vibration detection signals. In this embodiment of the application, the transmission unit 23 is a microcontroller.

[0074] The wireless communication unit 24 is connected to the transmission unit 23 and is used to transmit sound detection signals and vibration detection signals wirelessly. Generally, the wireless communication unit 24 can be a WiFi unit, Bluetooth unit, or LoRa unit, etc., for short-range wireless communication.

[0075] It is understandable that, since the sound monitoring unit 21, vibration monitoring unit 22 and transmission unit 23 can collect the sound and vibration generated by the wheel hub during the operation and transmit the collected sound and vibration, multiple monitoring terminals 2 including the above-mentioned sound monitoring unit 21, vibration monitoring unit 22 and transmission unit 23 can be placed in the high-speed rail to collect data from all wheel hubs at the same time, so that the host 1 can uniformly analyze each collected sound detection signal and vibration detection signal.

[0076] Since the closer the monitoring terminal 2 is to the wheel hub, the clearer the sound and vibration collected by the monitoring terminal 2 are, in this embodiment of the application, each monitoring terminal 2 is placed in a position opposite to a wheel hub, that is, placed under the seat or on the floor inside the carriage opposite to the wheel hub.

[0077] It is worth noting that each monitoring terminal 2 can not only collect the sound and vibration generated by the corresponding wheel hub during high-speed train operation, but also collect the sound and vibration generated by adjacent wheel hubs during high-speed train operation, so as to verify the sound detection signal and vibration detection signal collected from adjacent wheel hubs, thereby improving the accuracy of sound and vibration collection.

[0078] To illustrate with an example, suppose a high-speed train has 100 wheel hubs, namely hub A0-1, hub A0-2, hub A1-1, hub A1-2, hub A2-1, hub A2-2, ..., hub A49-1, hub A49-2. Correspondingly, monitoring terminal 2 also has 100 terminals, namely monitoring terminal B0-1, monitoring terminal B0-2, monitoring terminal B1-1, monitoring terminal B1-2, monitoring terminal B2-1, monitoring terminal B2-2, ..., monitoring terminal B49-1, monitoring terminal B49-2. Among them, monitoring terminal B1-1 is used to collect the sound and vibration generated by wheel hubs A0-1 and A1-1; monitoring terminal B1-2 is used to collect the sound and vibration generated by wheel hubs A0-2 and A1-2; monitoring terminal B2-1 is used to collect the sound and vibration generated by wheel hubs A1-1 and A2-1; monitoring terminal B2-2 is used to collect the sound and vibration generated by wheel hubs A1-2 and A2-2; ..., monitoring terminal B49-1 is used to collect the sound and vibration generated by wheel hubs A48-1 and A49-1; monitoring terminal B49-2 is used to collect the sound and vibration generated by wheel hubs A48-2 and A49-2. "AX-1" refers to one side of the wheel hub, "AX-2" refers to the other side, "BX-1" refers to a monitoring terminal placed on one side of the wheel hub, and "BX-2" refers to a monitoring terminal placed on the other side of the wheel hub.

[0079] Specifically, the monitoring terminal 2 also includes a sound calibration unit 25 and a vibration calibration unit 26.

[0080] The sound calibration unit 25 is used to collect the sound generated by adjacent wheel hubs during driving, and output a sound calibration signal. The vibration calibration unit 26 is used to collect the vibration generated by adjacent wheel hubs during driving, and output a vibration calibration signal. Since both the sound calibration unit 25 and the sound monitoring unit 21 are used to collect the sound generated by the wheel hubs, and both the vibration calibration unit 26 and the vibration monitoring unit 22 are used to collect the vibration generated by the wheel hubs, the sound calibration unit 25 and the vibration calibration unit 26 can be configured with reference to the sound monitoring unit 21 and the vibration monitoring unit 22 described above.

[0081] The transmission unit 23 is also connected to the sound calibration unit 25 and the vibration calibration unit 26 respectively, for receiving and transmitting sound calibration signals and vibration calibration signals.

[0082] Reference Figure 1 and Figure 2Similarly, the host 1 is also equipped with a wireless communication unit 24, which enables the host 1 to connect with multiple monitoring terminals 2 for verifying the sound detection signal and the sound calibration signal, verifying the vibration detection signal and the vibration calibration signal, and analyzing the sound detection signal and the vibration detection signal to determine whether there are abnormal noises and vibrations in the wheel hub during operation. It is also used to store the generated abnormal detection data and transmit the abnormal detection data to the data platform 4 when the high-speed train stops.

[0083] It is understood that the main unit 1 is located inside the carriage. Considering that the main unit 1 and multiple monitoring terminals 2 are wirelessly connected, and that the signal strength weakens with distance, a signal repeater 3 is also installed inside the carriage to forward the sound detection signal, vibration detection signal, sound calibration signal, and vibration calibration signal output by the monitoring terminals 2 that are far away from the main unit 1. This ensures that the main unit 1 can receive the sound detection signal, vibration detection signal, sound calibration signal, and vibration calibration signal output by all monitoring terminals 2.

[0084] Furthermore, the host 1 includes a verification unit 11, a speed monitoring unit 12, a storage unit 13, and an analysis and processing unit 14.

[0085] The verification unit 11 is connected to the wireless communication unit 24 within the host 1. It is used to verify the sound detection signal based on the sound calibration signal and the vibration detection signal based on the vibration calibration signal, and to transmit both the sound and vibration detection signals. It is worth noting that in special circumstances, such as when a monitoring terminal 2 disconnects from the host 1, the corresponding sound calibration signal and vibration calibration signal can be used as backups. That is, the verification unit 11 directly transmits the sound calibration signal and vibration calibration signal for later analysis.

[0086] The speed monitoring unit 12 is used to monitor the speed of the high-speed train in real time. It can be a speed sensor, or other devices with speed measurement functions can be selected.

[0087] The analysis and processing unit 14 is connected to the verification unit 11 and the speed monitoring unit 12, respectively. It is used to analyze the driving speed, sound detection signal and vibration detection signal at the same time mark according to the standard model stored in the storage unit 13, and to determine whether there are abnormal noises and vibrations generated by the wheel hub during driving, so as to generate abnormal detection data. Among them, the driving speed, sound detection signal and vibration detection signal at the same time mark are driving speed, sound detection signal and vibration detection signal matched with the same time mark.

[0088] Specifically, storage unit 13 is used to store multiple standard models. These standard models are speed-sound / vibration models obtained after multiple tests for each vehicle model. Taking a certain vehicle model as an example, its standard model presents the normal sound fluctuation range and normal vibration range corresponding to each speed.

[0089] Reference Figure 3 Furthermore, the specific analysis process of the analysis and processing unit 14 is as follows:

[0090] Step S100: Obtain sound and vibration data for each wheel hub.

[0091] The sound data refers to the sound detection signal collected by monitoring terminal 2, and the vibration data refers to the vibration detection signal collected by monitoring terminal 2. Both the sound and vibration data carry time stamps, speed stamps, location stamps, and wheel hub numbers. The speed stamp reflects the travel speed corresponding to the sound or vibration data at different times; it is the speed detection signal collected by speed monitoring unit 12. The location stamp reflects the geographical location of the train at different times, and in this embodiment, it is represented by longitude, latitude, and altitude. The location stamp can be obtained from GPS module 16.

[0092] Step S200: Divide the sound data into a first static group and a first dynamic group according to the positioning identifier and the time identifier, and divide the vibration data into a second static group and a second dynamic group.

[0093] Understandably, high-speed rail lines are typically designed based on the terrain of the cities they pass through, resulting in uneven terrain and potentially including uphill, downhill, and level sections. The load on the wheel hubs differs between uphill and downhill sections. Therefore, wheel hub testing during train operation must consider the impact of different terrain sections. Conversely, compared to uphill and downhill sections, level sections have a smaller impact on the wheel hubs. The magnitude of sound and vibration data from level sections can be considered the same as the test results in the standard model to facilitate analysis of whether individual wheel hubs are producing abnormal noises or vibrations. The standard model is one of the standard models retrieved from storage unit 13 based on the train type.

[0094] Specifically, the first static group contains sound data from horizontal road sections, the first dynamic group contains sound data from uphill or downhill road sections, the second static group contains vibration data from horizontal road sections, and the second dynamic group contains vibration data from uphill or downhill road sections. The amount of data in the first static group is the same as that in the second static group, and the amount of data in the first dynamic group is the same as that in the second dynamic group.

[0095] In a specific embodiment, the method for dividing sound data into a first static group and a first dynamic group, and vibration data into a second static group and a second dynamic group, can be as follows:

[0096] From the sound data, select data with the same altitude and continuous time stamps as the first static group, and use the remaining data from the sound data as the first dynamic group. From the vibration data, select data with the same altitude and continuous time stamps as the second static group, and use the remaining data from the vibration data as the second dynamic group. It should be noted that there may be multiple horizontal road sections along the entire route; the sound and vibration data for multiple horizontal road sections can be obtained by filtering according to the above method.

[0097] Of course, in other embodiments, other methods can be used to divide the sound data into a first static group and a first dynamic group, and to divide the vibration data into a second static group and a second dynamic group.

[0098] Step S300: Analyze the data of the first static group and the data of the second static group according to the standard model to obtain the first analysis result for each wheel hub.

[0099] Since the magnitude of the sound data in the first static group is the same as that of the test results in the standard model, and the magnitude of the vibration data in the second static group is the same as that of the test results in the standard model, the data in the first static group and the data in the second static group can be directly compared with the test results in the standard model to determine whether there are abnormal noises or vibrations in each wheel hub on a horizontal road section.

[0100] Understandably, when determining whether there are abnormal noises or vibrations in each wheel hub, it is necessary to judge each wheel hub in turn. When judging each wheel hub, it is also necessary to judge whether the sound and vibration generated by the wheel hub in real time during the train's operation are abnormal.

[0101] Taking a wheel hub as an example, the specific method to determine whether the wheel hub produces abnormal noise or vibration is as follows:

[0102] You can select each sound or vibration data point in order from morning to night according to the time markers.

[0103] Assuming sound data is selected, the corresponding sound pattern needs to be found in the standard model based on the speed identifier of the sound data. In the standard model, each speed can correspond to a normal sound pattern range and a vibration ripple range, or the midpoint of the normal sound pattern range and the midpoint of the normal vibration ripple range. Then, the sound data is compared with the normal sound pattern range corresponding to the speed in the standard model. If the sound data falls within the normal sound pattern range, it indicates that the wheel hub is not producing any abnormal noise at that moment.

[0104] Furthermore, corresponding vibration data can be determined based on the time or location markers of the selected sound data to analyze the vibration generated by the wheel hub at that moment. The analysis method is the same as that used for analyzing sound data. Once both the sound and vibration data at a certain moment have been analyzed, it can be determined whether the wheel hub produced any abnormal noise or vibration at that moment.

[0105] In other embodiments, vibration data can be selected first, or sound data or vibration data can be selected according to other rules. Of course, each piece of sound and vibration data can also be analyzed separately, and the results can be integrated according to the wheel hub after all the sound and vibration data have been analyzed.

[0106] After analyzing the data from the first and second static groups, the state of each wheel hub in all horizontal road sections can be obtained, which is the first analysis result. Since the first and second static groups only contain partial data, it is necessary to integrate the first analysis result with the state of each wheel hub in uphill and downhill road sections.

[0107] Step S400: Determine the static sound pattern and static vibration ripple of each wheel hub based on the positioning mark and the speed mark.

[0108] Understandably, the relationships between speed and sound, and speed and vibration, in the standard model are derived from numerous simulation tests. However, for uphill and downhill sections, the sound and vibration data collected by monitoring terminal 2 are influenced by various factors. Therefore, directly analyzing the sound and vibration data using the test results from the standard model would lead to misjudgments. Thus, it is necessary to filter out the sound patterns and vibration ripples caused by various influencing factors in the sound and vibration data before comparing them with the standard model to obtain accurate analysis results.

[0109] First, it should be noted that during the train's operation, the wheel hub comes into contact with the rails, and is thus affected by the forces acting on it, causing the wheel hub itself to produce sound and vibration.

[0110] In this embodiment, the static sound pattern and static vibration wave of each wheel hub can be determined based on the location identifier and speed identifier. The static sound pattern is the sound produced by the wheel hub itself under the influence of force, and the static vibration wave is the vibration produced by the wheel hub itself under the influence of force. Specifically, the sound pattern with the location identifier at a specified position and the speed identifier at zero in the sound data is identified as the static sound pattern, and the vibration wave with the location identifier at a specified position and the speed identifier at zero in the vibration data is identified as the static vibration wave. It should be noted that the specified position is the station where the train stops. It is understood that since the wheel hub will wear during train operation, the static sound pattern and static vibration wave should also change in real time. For ease of analysis, the static sound pattern contained in the sound data and the static vibration wave contained in the vibration data of each wheel hub between two adjacent stations can be assumed to be the sound data and vibration data with a speed identifier of zero measured at the previous station. Similarly, the static sound pattern and static vibration wave also carry the location identifier and wheel hub number.

[0111] Of course, in other embodiments, other methods can be used to determine static acoustic patterns and static vibration ripples.

[0112] Step S500: Analyze the data of the first dynamic group and the second dynamic group based on the static sound pattern, the static vibration ripple, and the standard model to obtain the second analysis result for each wheel hub.

[0113] Optionally, step S500 includes the following steps: (steps S510 to S550)

[0114] Step S510: Determine the environmental sound and vibration waveforms of the high-speed train at different locations on the railway based on the static sound and vibration waveforms, standard model, sound data, and vibration data of the same wheel hub.

[0115] Among them, environmental sound pattern is the sound pattern caused by the environment to the wheel hub, and environmental vibration wave pattern is the vibration wave pattern caused by the environment to the wheel hub.

[0116] Before analyzing the data from the first and second dynamic groups, it is necessary to determine the impact of the environment on the sound and vibration data on uphill and downhill sections. It is worth noting that although the environmental impact on the sound and vibration data changes continuously during train operation, the effect on each wheel hub at the same location is the same.

[0117] Based on this, embodiments of this application provide a method for determining environmental sound signatures and environmental vibration ripples.

[0118] Understandably, since trains have wheel hubs on both sides, the impact of the same location on the left wheel hub will differ from that on the right wheel hub. Therefore, it is necessary to determine the environmental sound signature and environmental vibration ripples of the wheel hubs on different sides at different locations.

[0119] Specifically, the method for determining environmental voiceprints is as follows:

[0120] First, select the sound and vibration data of the first N wheel hubs.

[0121] Where N is an even number greater than 2 and less than the total number of wheel hubs. In some embodiments, for ease of calculation, N can be an even number greater than 10 and less than 20. Because it is necessary to determine the environmental sound signature and environmental vibration ripple of the wheel hubs on different sides, wheel hubs need to be selected from both sides simultaneously.

[0122] Then, based on the sound data, static sound signature, and standard model, the ambient sound signature reference value for each wheel hub at each location is determined.

[0123] Among them, the environmental soundprint reference value is the estimated environmental soundprint.

[0124] Specifically, taking a wheel hub as an example, its environmental sound pattern reference value at a certain location is obtained by filtering out the static sound pattern of the wheel hub from the sound data, as well as the sound pattern corresponding to the speed label of the sound data in the standard model. Since each sound data carries a wheel hub number and a location identifier, the corresponding sound data can be determined once the wheel hub and location are determined. Furthermore, the corresponding static sound pattern can also be determined based on the wheel hub number and location identifier. At the same time, the corresponding sound pattern can also be found in the standard model based on the speed label of the sound data. Assuming that the wheel hub does not have abnormal noise or vibration, the sound patterns corresponding to each speed in the standard model can be considered as the actual sound patterns generated by the wheel hub. Thus, by filtering out the static sound pattern from the sound data and the sound pattern corresponding to the speed label of the sound data in the standard model, the environmental sound pattern reference value can be obtained.

[0125] Finally, the mode value of the ambient sound reference values ​​when each wheel hub is in the same position is taken as the ambient sound at that position.

[0126] Because abnormal noises or vibrations may exist in each wheel hub, the ambient sound signature reference values ​​obtained for the same location may differ, indicating that some wheel hubs may be generating abnormal noises or vibrations. In this case, the most common value can be selected as the ambient sound signature. The ambient sound signature also carries a location identifier.

[0127] It should be noted that when determining the ambient sound signature of the left wheel hub at the same location, the sound data of the left wheel hub needs to be selected and analyzed; when determining the ambient sound signature of the right wheel hub at the same location, the sound data of the right wheel hub needs to be selected and analyzed.

[0128] By repeating the above method, you can obtain the ambient sound signatures of different wheel hubs at different locations.

[0129] Furthermore, the method for determining environmental vibration ripples is as follows:

[0130] First, based on vibration data, static vibration ripples, and a standard model, determine the environmental vibration ripple reference value for each wheel hub at each location. Then, take the mode value of the environmental vibration ripple reference values ​​for all wheel hubs at the same location as the environmental vibration ripple for that location.

[0131] Since the principle is the same as that of the method for determining environmental voiceprints mentioned above, it will not be repeated here.

[0132] Similarly, environmental vibration ripples also carry location markers.

[0133] Step S520: Select a sound data or a vibration data from the first dynamic group or the second dynamic group.

[0134] Step S530: Select the ambient sound pattern and ambient vibration pattern based on the location identifier and wheel hub number of the selected sound data or vibration data.

[0135] The wheel hub number can be used to determine whether the wheel hub is located on the left or right side, which makes it easier to select the corresponding side's ambient sound pattern or ambient vibration pattern based on the positioning mark.

[0136] Step S540: Analyze the selected sound data or vibration data based on the standard model, static sound pattern, static vibration waveform, and environmental sound pattern and environmental vibration waveform to obtain the judgment result.

[0137] The judgment result is used to reflect whether the wheel hub produces abnormal noise or vibration at a certain speed.

[0138] Specifically, the method for analyzing sound data is as follows:

[0139] First, the target soundprint is determined based on the selected sound data, the static soundprint of the corresponding wheel hub, and the environmental soundprint whose positioning identifier matches the positioning identifier of the sound data.

[0140] Among them, the target soundprint is the soundprint actually generated by the wheel hub.

[0141] Specifically, the target soundprint is obtained by filtering out the static soundprint of the corresponding wheel hub and the ambient soundprint whose location identifier matches the location identifier of the sound data from the selected sound data. Since each sound data carries a wheel hub number and a location identifier, the corresponding sound data can be determined once the wheel hub and location are determined. Furthermore, based on the wheel hub number and location identifier, the corresponding static soundprint and ambient soundprint can also be determined. Thus, by filtering out the static soundprint of the corresponding wheel hub and the ambient soundprint whose location identifier matches the location identifier of the sound data from the selected sound data, the target soundprint can be obtained. At this point, the magnitude of the obtained target soundprint is the same as that of the test results in the standard model, and can be compared with the test results in the standard model.

[0142] Then, the voiceprints that match the speed identifier of the sound data in the standard model are compared with the target voiceprint to obtain the judgment result.

[0143] The specific comparison method can refer to the method of comparing the data of the first static group and the second static group with the standard model mentioned above. That is, when the target soundprint is within the normal soundprint range, the judgment result is that the wheel hub does not produce abnormal noise at that moment. Conversely, when the target soundprint is not within the normal soundprint range, it indicates that the wheel hub produces abnormal noise, and the judgment result is that the wheel hub has abnormal noise. Further details will not be elaborated here.

[0144] Furthermore, the method for analyzing vibration data is as follows:

[0145] First, the target vibration wave is determined based on the selected vibration data, the corresponding static vibration wave of the wheel hub, and the environmental vibration wave whose positioning marker matches the positioning marker of the vibration data. Then, the vibration wave in the standard model that matches the velocity marker of the vibration data is compared with the target vibration wave to obtain the judgment result. The target vibration wave is the actual vibration wave generated by the wheel hub.

[0146] Since the principle is the same as that of the above-mentioned method for analyzing sound data, it will not be repeated here.

[0147] This completes the analysis of sound or vibration data for a specific wheel hub at a specific location. The same method can be repeated to analyze the data from the first dynamic group and the second dynamic group.

[0148] Step S550: Combine all the judgment results for the same wheel hub to obtain the second analysis result.

[0149] Step S600: Generate anomaly detection data by combining the first analysis result and the second analysis result.

[0150] By integrating all the judgment results and the initial analysis results for the same wheel hub, it can be determined whether the wheel hub has abnormal noise, and under what circumstances it will produce abnormal noise. This enables the analysis of the sound patterns and vibration waves generated by each wheel hub at every moment. After analysis, anomaly detection data is finally generated. The anomaly detection data contains data information related to the wheel hub that produces abnormal noise or vibration.

[0151] Of course, when integrating the initial analysis results and judgment results, integration can also be carried out according to the road segments divided by the stations.

[0152] Reference Figure 1 and Figure 2 The storage unit 13 is connected to the analysis and processing unit 14 and is also used to store anomaly detection data, so that the analysis and processing unit 14 can transmit the anomaly detection data to the data platform 4 when the high-speed train stops.

[0153] To facilitate the transmission of anomaly detection data by the analysis and processing unit 14, the analysis and processing unit 14 is also connected to a mobile communication unit 15, which can specifically use 4G for communication.

[0154] The data platform 4 is connected to the analysis and processing unit 14, which is used to receive anomaly detection data and generate anomaly detection reports based on the anomaly detection data.

[0155] Because the data platform 4 and the analysis and processing unit 14 are connected via 4G, there may be periods of no signal while the high-speed train is in motion, preventing timely transmission of anomaly detection data to the data platform 4. Therefore, the analysis and processing unit 14 only transmits the anomaly detection data to the data platform 4 when the high-speed train stops.

[0156] Specifically, the analysis and processing unit 14 is also connected to a GPS module 16. The GPS module 16 is used to determine the positional relationship between the high-speed train and each stop, and outputs a distance signal. The distance signal reflects the distance between the current position of the high-speed train and the next stop.

[0157] The analysis and processing unit 14 outputs anomaly detection data when the distance value reflected by the received distance signal is less than a preset value. Otherwise, it does not output anomaly detection data. It is understood that when the distance value reflected by the distance signal is less than the preset value, it indicates that the high-speed train has entered the station, so the anomaly detection data can be transmitted to the data platform 4.

[0158] Of course, this is just one way to trigger the analysis and processing unit 14 to output abnormal detection data, for reference only, and is not intended to impose any restrictions.

[0159] The implementation principle of the high-speed rail running gear load recording system based on acoustic signature features in this application is as follows: A monitoring terminal 2 is set up to collect and transmit the sound and vibration generated by the wheel hub during high-speed rail operation. A host computer 1 is set up to analyze all sound and vibration detection signals. This eliminates the need for each monitoring terminal 2 to have a separate module for analyzing sound and vibration detection signals, thus significantly reducing the manufacturing cost of the monitoring terminal 2.

[0160] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A high-speed rail running gear load recording system based on voiceprint features, characterized in that: It includes a main unit (1) and multiple monitoring terminals (2) for detecting wheel hubs; A monitoring terminal (2) is placed opposite to a wheel hub. The monitoring terminal (2) is used to collect the sound and vibration generated by the corresponding wheel hub during driving, so as to output sound detection signal and vibration detection signal. The host (1) is set on the high-speed rail and connected to multiple monitoring terminals (2) for analyzing the sound detection signal and vibration detection signal to determine whether the high-speed rail has abnormal noise and vibration generated by the wheel hub during operation. It is used to store the generated abnormal detection data and transmit the abnormal detection data to the data platform (4) when the high-speed rail stops. The host (1) includes a GPS module (16), a speed monitoring unit (12), an analysis and processing unit (14), and a storage unit (13); The GPS module (16) is used to locate the position of the high-speed train; The speed monitoring unit (12) is used to monitor the speed of the high-speed train in real time; The storage unit (13) is used to store multiple standard models, each standard model corresponding to a vehicle model; The analysis and processing unit (14) is connected to the storage unit (13), the speed monitoring unit (12) and multiple monitoring terminals (2), and is used to analyze the driving speed, sound detection signal and vibration detection signal at the same time according to a standard model, to determine whether the high-speed rail has abnormal noise and vibration in the wheel hub during the operation, so as to generate abnormal detection data, and transmit the abnormal detection data to the data platform (4) when the high-speed rail stops; The storage unit (13) is also used to store the anomaly detection data; The analysis and processing unit (14) is further configured as follows: Acquire sound and vibration data for each wheel hub. Both the sound and vibration data carry a time identifier, a speed identifier, a location identifier, and a wheel hub number. The location identifier includes longitude, latitude, and altitude. The sound data is divided into a first static group and a first dynamic group based on the positioning identifier and the time identifier, and the vibration data is divided into a second static group and a second dynamic group. Based on the standard model, the data of the first static group and the data of the second static group are analyzed to obtain the first analysis result for each wheel hub; The static sound pattern and static vibration wave pattern of each wheel hub are determined according to the positioning mark and the speed mark. The static sound pattern is the sound pattern generated when the wheel hub is stationary, and the static vibration wave pattern is the vibration wave pattern generated when the wheel hub is stationary. Based on the static acoustic pattern, the static vibration ripple, and the standard model, the data of the first dynamic group and the data of the second dynamic group are analyzed to obtain the second analysis result for each wheel hub. Anomaly detection data is generated by combining the first analysis result and the second analysis result.

2. The high-speed rail running gear condition load recording system based on voiceprint features according to claim 1, characterized in that: The analysis and processing unit (14) is further configured as follows: The step of analyzing the data of the first dynamic group and the data of the second dynamic group based on the static sound pattern, the static vibration ripple, and the standard model to obtain the second analysis result for each wheel hub includes: Based on the static acoustic pattern, static vibration wave pattern, standard model, sound data, and vibration data of the same wheel hub, the environmental acoustic pattern and environmental vibration wave pattern of the high-speed rail at different locations on the railway are determined. The environmental acoustic pattern is the acoustic pattern caused by the environment to the wheel hub, and the environmental vibration wave pattern is the vibration wave pattern caused by the environment to the wheel hub. Select a sound data or a vibration data from the first dynamic group or the second dynamic group; Select environmental sound patterns or environmental vibration patterns based on the location identifiers and wheel hub number identifiers of the selected sound or vibration data. The selected sound data or vibration data are analyzed based on the standard model, the static sound pattern or the static vibration wave pattern, and the environmental sound pattern or the environmental vibration wave pattern to obtain a judgment result. The judgment result is used to reflect whether the wheel hub produces abnormal noise or abnormal vibration at a certain speed. The second analysis result is obtained by combining all the judgment results for the same wheel hub.

3. The high-speed rail running gear condition load recording system based on voiceprint features according to claim 2, characterized in that: The analysis and processing unit (14) is further configured as follows: The determination of the environmental sound signature and environmental vibration wave at different locations on the railway based on the static sound signature, static vibration wave, standard model, sound data, and vibration data of the same wheel hub includes: Select the sound and vibration data of the first N wheel hubs; Based on the sound data, the static acoustic signature, and the standard model, determine the environmental acoustic signature reference value for each wheel hub at each location; The mode value of the ambient sound reference value when each wheel hub is in the same position is taken as the ambient sound of that position. Based on the vibration data, the static vibration ripples, and the standard model, determine the reference value of the environmental vibration ripples for each wheel hub at each location; The mode value of the environmental vibration ripple reference value when all wheel hubs are located at the same position is taken as the environmental vibration ripple at that position.

4. The high-speed rail running gear condition load recording system based on voiceprint features according to claim 3, characterized in that: The analysis and processing unit (14) is further configured as follows: The method for determining the environmental acoustic reference value of a wheel hub at a location based on the sound data, the static acoustic signature, and the standard model is as follows: The static acoustic signature of the wheel hub and the acoustic signature corresponding to the speed identifier of the sound data in the standard model are filtered out from the sound data to obtain the environmental acoustic signature reference value. The method for determining the reference value of environmental vibration ripple for a wheel hub at a location based on the vibration data, the static vibration ripple, and the standard model is as follows: The static vibration ripples of the wheel hub and the vibration ripples corresponding to the velocity labels of the vibration data in the standard model are filtered out from the vibration data to obtain the environmental vibration ripple reference value.

5. The high-speed rail running gear condition load recording system based on voiceprint features according to claim 2, characterized in that: The analysis and processing unit (14) is further configured as follows: The step of analyzing the selected sound data or vibration data based on the standard model, the static sound pattern or the static vibration ripple, and the environmental sound pattern or the environmental vibration ripple to obtain the judgment result includes: The target soundprint is determined based on the selected sound data, the static soundprint of the corresponding wheel hub, and the environmental soundprint whose positioning identifier matches the positioning identifier of the sound data. The voiceprint in the standard model that matches the speed identifier of the sound data is compared with the target voiceprint to obtain a judgment result; The target vibration ripple is determined based on the selected vibration data, the static vibration ripple of the corresponding wheel hub, and the environmental vibration ripple whose positioning mark matches the positioning mark of the vibration data. The vibration ripples in the standard model that match the velocity identifier of the vibration data are compared with the target vibration ripples to obtain the judgment result.

6. The high-speed rail running gear condition load recording system based on voiceprint features according to claim 5, characterized in that: The analysis and processing unit (14) is further configured as follows: The step of determining the target soundprint based on the selected sound data, the static soundprint of the corresponding wheel hub, and the ambient soundprint whose positioning identifier matches the positioning identifier of the sound data includes: The target soundprint is obtained by filtering out the static soundprint of the corresponding wheel hub and the environmental soundprint whose positioning identifier is consistent with the positioning identifier of the sound data from the selected sound data. The step of determining the target vibration ripple based on the selected vibration data, the static vibration ripple of the corresponding wheel hub, and the environmental vibration ripple whose positioning marker matches the positioning marker of the vibration data includes: The target vibration ripple is obtained by filtering out the static vibration ripples of the corresponding wheel hub and the environmental vibration ripples whose positioning marks are consistent with the positioning marks of the vibration data from the selected vibration data.

7. The high-speed rail running gear condition load recording system based on voiceprint features according to claim 1, characterized in that: The analysis and processing unit (14) is further configured as follows: The step of dividing the sound data into a first static group and a first dynamic group based on the positioning identifier and the time identifier, and dividing the vibration data into a second static group and a second dynamic group, includes: The data with the same altitude and consecutive time markers in the sound data are identified as the first static group, and the remaining data in the sound data are identified as the first dynamic group. The vibration data with the same altitude and continuous time markers are identified as the second static group, and the remaining data in the vibration data are identified as the second dynamic group.

8. The high-speed rail running gear load recording system based on voiceprint features according to claim 1, characterized in that: The analysis and processing unit (14) is further configured as follows: The step of determining the static sound signature and static vibration ripple of each wheel hub based on the positioning identifier and the speed identifier includes: Voiceprints in the sound data whose location identifier is a specified location and whose speed identifier is zero are identified as static voiceprints. The vibration ripples in the vibration data that are located at a specified position and have a velocity of zero are identified as static vibration ripples.

Images