Train bearing temperature monitoring method and system based on false data elimination

Abstract

The present invention discloses a train bearing temperature monitoring method and system based on pseudo data elimination, including a data acquisition terminal and a remote monitoring unit; the acquisition terminal is used to collect and send the temperature data of the train bearing; the remote monitoring unit is used to analyze and process the temperature data, obtain the results after analysis, and perform early warning processing or alarm processing according to the results after analysis. The present invention eliminates or identifies abnormal temperature data information by judging and analyzing the bearing temperature data within a certain continuous time period, and uses the temperature data to warn or alarm the bearing temperature, thereby eliminating the influence of abnormal temperature data on the monitoring results, improving the reliability and accuracy of bearing temperature monitoring, and ensuring the safe operation of rail transit.

Description

Technical Field

[0001] The present invention relates to the field of rail transit monitoring, and in particular to a train bearing temperature monitoring method and system based on pseudo data elimination. Background Art

[0002] With the rapid development of my country's transportation industry and the accelerated pace of modernization, rail transit has become the main means of transportation in cities, but it is also accompanied by safety risks. Bearings are one of the key parts of trains. During the operation of trains, vibration and impact may cause damage to bearings, affecting the safety performance of trains. Therefore, bearing monitoring has become an important part of ensuring the safe operation of trains.

[0003] At present, bearing monitoring mainly focuses on oil samples, noise, vibration and temperature. Among them, bearing monitoring based on bearing temperature information is more widely used due to its good effect. However, when measuring the bearing temperature of a train, it is easily affected by the external environment, train operating conditions and train running speed, etc., and it is easy for sensors to make errors or the temperature measured by the sensor to make errors, resulting in abnormalities in the measured temperature data, thereby causing false warnings and alarms.

[0004] Therefore, a train bearing temperature monitoring method and system based on pseudo data elimination is needed to solve the above problems. Summary of the invention

[0005] In view of this, the purpose of the present invention is to overcome the defects in the prior art and provide a train bearing temperature monitoring method and system based on pseudo data elimination, which can eliminate the influence of abnormal temperature data on the monitoring results and improve the reliability and accuracy of bearing temperature monitoring.

[0006] The train bearing temperature monitoring method based on pseudo data elimination of the present invention comprises the following steps:

[0007] S1. Collect the temperature T of the target bearing at time n n ;

[0008] S2. Determine the temperature T n Is it normal? If so, delete the oldest data in the data set and set the temperature T n Insert into the data set; if not, proceed to the next step;

[0009] S3. Collect the temperature T of the target bearing at time n+1 n+1 , determine the temperature T n+1 Is it normal? If so, then the temperature T n Update and get the updated temperature T n ′ ; If not, calculate the temperature T n+1With temperature T n The difference Δt between them is determined as follows:

[0010] If the difference Δt is greater than the preset threshold ΔT or is a non-positive real number, the acquisition device is abnormal and an abnormal alarm is issued;

[0011] If the difference Δt is less than or equal to the preset threshold ΔT and is a non-negative real number, proceed to the next step;

[0012] S4. Collect the temperature T of the target bearing at time n+2 n+2 , determine the temperature T n+2 Is it normal? If so, the acquisition device is abnormal and an abnormal alarm is issued; if not, the temperature T is calculated. n+2 With temperature T n+1 The difference Δt1 between them is determined as follows:

[0013] If the difference Δt1 is greater than the preset threshold ΔT1 or is a non-positive real number, the acquisition device is abnormal and an abnormal alarm is issued;

[0014] If the difference Δt1 is less than or equal to the preset threshold ΔT1 and is a non-negative real number, it indicates that the target bearing temperature rises within a reasonable range, and the temperature T n is the effective value and according to the temperature T n Early warning of bearing temperature.

[0015] Further, the data set includes a number of normal temperature data of the train bearing before time n; wherein the train bearing includes the target bearing.

[0016] Furthermore, according to the temperature T n Early warning of bearing temperature includes the following steps:

[0017] S5. Set the temperature T n With the warning threshold T Y The following comparison is made:

[0018] If the temperature T n Greater than or equal to the warning threshold T Y , then the target bearing temperature warning is issued and the next step is entered;

[0019] If the temperature T n Less than the warning threshold T Y , then the target bearing temperature is normal;

[0020] S6. Set the temperature T n With the warning threshold T B The following comparison is made:

[0021] If the temperature Tn Greater than or equal to the warning threshold T B , then the target bearing temperature alarm is issued;

[0022] If the temperature T n Less than the warning threshold T B , then the target bearing temperature warning is issued.

[0023] Further, the preset threshold ΔT is determined according to the following formula:

[0024] ΔT=Δt R *Δε;

[0025] Among them, Δt R It is the maximum temperature difference per unit time when the bearing runs at full load in the actual environment; Δε is the time interval for temperature data collection.

[0026] Further, the preset threshold ΔT1 is determined according to the following formula:

[0027] ΔT1=Δt R *Δε;

[0028] Among them, Δt R It is the maximum temperature difference per unit time when the bearing runs at full load in the actual environment; Δε is the time interval for temperature data collection.

[0029] Further, the warning threshold T is determined according to the following formula: Y :

[0030] T Y =T R -α*T s ;

[0031] Among them, T R is k times the maximum operating temperature of the bearing material; T s is the ambient temperature of the bearing; α is the environmental factor variable.

[0032] Further, the warning threshold T is determined according to the following formula: B :

[0033] T B =T V -α*T s ;

[0034] Among them, T V is the maximum operating temperature corresponding to the bearing material; T s is the ambient temperature of the bearing; α is the environmental factor variable.

[0035] A train bearing temperature monitoring system based on pseudo data elimination includes a data acquisition terminal and a remote monitoring unit;

[0036] The acquisition terminal is used to collect and send temperature data of train bearings;

[0037] The remote monitoring unit is used to analyze and process the temperature data to obtain analysis results, and perform early warning processing or alarm processing according to the analysis results.

[0038] Furthermore, the acquisition terminal includes a control module, a communication module and a temperature sensor module;

[0039] The temperature sensing module is used to collect temperature data of train bearings; the control module sends the temperature data to the remote monitoring unit through the communication module.

[0040] Furthermore, the remote monitoring unit includes a cloud platform; the cloud platform includes a network server and an alarm service device.

[0041] The beneficial effects of the present invention are as follows: the present invention discloses a train bearing temperature monitoring method and system based on pseudo data elimination, which eliminates or identifies abnormal temperature data information by judging and analyzing the bearing temperature data within a certain continuous time period, and saves the normal temperature information into a data set, and continuously updates the data set, so that normal or accurate temperature data can be retained, and uses the normal temperature data to warn or alarm the bearing temperature, eliminating the influence of abnormal temperature data on the monitoring results, improving the reliability and accuracy of bearing temperature monitoring, and ensuring the safe operation of rail transit. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0043] Figure 1 It is a schematic diagram of the monitoring method flow of the present invention;

[0044] Figure 2 It is a schematic diagram of the arrangement of the acquisition terminal of the present invention;

[0045] Figure 3 It is a structural schematic diagram of the acquisition terminal of the present invention;

[0046] Figure 4 It is a schematic diagram of the monitoring system architecture principle of the present invention. DETAILED DESCRIPTION

[0047] The present invention is further described below in conjunction with the accompanying drawings, as shown in the drawings:

[0048] The train bearing temperature monitoring method based on pseudo data elimination of the present invention comprises the following steps:

[0049] S1. Collect the temperature T of the target bearing at time n n ;

[0050] S2. Determine the temperature T n Is it normal? If so, delete the oldest data in the data set and set the temperature T n Insert into the data set; if not, proceed to the next step;

[0051] S3. Collect the temperature T of the target bearing at time n+1 n+1 , determine the temperature T n+1 Is it normal? If so, then the temperature T n Update and get the updated temperature T n '; Through the above steps, the temperature T can be determined n Although the display is abnormal, it may not trigger an early warning or alarm. n The displayed value is corrected to eliminate the temperature T n The display abnormal state, the temperature T n To update, you can set T n =(T n-1 +T n+1 ) / 2.

[0052] If not, calculate the temperature T n+1 With temperature T n The difference Δt between them is determined as follows:

[0053] If the difference Δt is greater than the preset threshold ΔT or is a non-positive real number, it indicates that the target bearing temperature changes too fast and does not conform to the actual train bearing temperature change, and the acquisition device is abnormal, and an abnormal alarm is issued; wherein the acquisition device is a temperature sensor for collecting bearing temperature;

[0054] If the difference Δt is less than or equal to the preset threshold ΔT and is a non-negative real number, proceed to the next step;

[0055] S4. Collect the temperature T of the target bearing at time n+2 n+2 , determine the temperature T n+2 Is it normal? If so, it means that the target bearing temperature changes too fast and does not conform to the actual train bearing temperature change. The acquisition equipment is abnormal and an abnormal alarm is issued; if not, the temperature T is calculated. n+2 With temperature T n+1 The difference Δt1 between them is determined as follows:

[0056] If the difference Δt1 is greater than the preset threshold ΔT1 or is a non-positive real number, it indicates that the target bearing temperature changes too fast and does not conform to the actual train bearing temperature change, and the acquisition device is abnormal, and an abnormal alarm is issued for the acquisition device;

[0057] If the difference Δt1 is less than or equal to the preset threshold ΔT1 and is a non-negative real number, it indicates that the target bearing temperature rises within a reasonable range, and the temperature T n is the effective value and according to the temperature T n Early warning of bearing temperature.

[0058] It should be noted that the above monitoring method can be used for the temperature monitoring of each bearing of the train bearing. By monitoring the temperature of the target bearing, the temperature status information of the target bearing can be objectively, effectively and accurately obtained, and thus the temperature warning or alarm can be carried out in time, thereby ensuring the safe and reliable operation of the train.

[0059] In this embodiment, the data set includes a number of normal temperature data of train bearings before time n; wherein the train bearings include target bearings. For example, there are 12 acquisition terminals, each of which acquires temperature information of a bearing, and the 12 acquisition terminals respectively acquire 256 normal temperature data of the corresponding bearing before time n, thereby forming a 12*256 matrix and a data set.

[0060] In this embodiment, according to the temperature T n Early warning of bearing temperature includes the following steps:

[0061] S5. Set the temperature T n With the warning threshold T Y The following comparison is made:

[0062] If the temperature T n Greater than or equal to the warning threshold T Y , then the target bearing temperature warning is issued and the next step is entered;

[0063] If the temperature T n Less than the warning threshold T Y , then the target bearing temperature is normal;

[0064] S6. Set the temperature T n With the warning threshold T B The following comparison is made:

[0065] If the temperature T n Greater than or equal to the warning threshold T B , then the target bearing temperature alarm is issued;

[0066] If the temperature T n Less than the warning threshold T B, then the target bearing temperature warning is issued.

[0067] Among them, the early warning is mainly a reminder, indicating that the bearing temperature is slightly high and needs to be tracked and prevented; the alarm indicates that the bearing temperature is too high and a danger has occurred, which requires timely processing.

[0068] Through the above steps, the early warning or alarm of the target bearing temperature is further strengthened.

[0069] In this embodiment, the preset threshold ΔT is determined according to the following formula:

[0070] ΔT=Δt R *Δε;

[0071] Among them, Δt R It is the maximum temperature difference per unit time when the bearing runs at full load in the actual environment; Δε is the time interval for temperature data collection.

[0072] The preset threshold ΔT1 is determined according to the following formula:

[0073] ΔT1=Δt R *Δε;

[0074] Among them, Δt R is the maximum temperature difference per unit time when the bearing is running at full load in the actual environment; Δε is the time interval for temperature data collection. Among them, the maximum temperature difference refers to the maximum difference in bearing temperature change per unit time, and the unit time can be taken as per minute.

[0075] Because the time interval for collecting temperature data is fixed, the preset threshold ΔT and the preset threshold ΔT1 may be equal.

[0076] In this embodiment, the warning threshold T is determined according to the following formula: Y :

[0077] T Y =T R -α*T s ;

[0078] Among them, T R is k times the maximum operating temperature of the bearing material; T s is the ambient temperature of the bearing; α is the environmental factor variable. Among them, k can be taken as 0.8;

[0079] Determine the warning threshold T according to the following formula B :

[0080] T B =T V -α*T s ;

[0081] Among them, T V is the maximum operating temperature corresponding to the bearing material; T s is the ambient temperature of the bearing; α is the environmental factor variable. The environmental factor variable α is [0.1, 0.3].

[0082] By constructing a RRCF model (Robust Random Cut Forest), and using the RRCF model as a recognition model to determine whether the bearing temperature data is normal; by inputting the bearing temperature into the constructed RRCF model, the judgment result of whether the bearing temperature is normal is output. The RRCF model judges the temperature data based on: first calculate the average abnormal score of the temperature data, if the average abnormal score exceeds the set threshold, the temperature is considered to be abnormal data; generally, the average abnormal score is a value between 0-1, and the set threshold can be 0.5; in this embodiment, the temperature data refers to the temperature value collected by the acquisition terminal at time n, and the average abnormal score refers to the average of multiple abnormal values ​​obtained after the temperature data at time n is input into the model (the RRCF model includes multiple decision trees, and when outputting the abnormal score, each decision tree will output an abnormal score, and these abnormal scores are all obtained based on the temperature value at time n); the RRCF model adopts the existing technology, which will not be repeated here.

[0083] The present invention also relates to a train bearing temperature monitoring system based on pseudo data elimination, the system corresponds to the above-mentioned train bearing temperature monitoring method based on pseudo data elimination, and can be understood as a system for implementing the above-mentioned method, the system includes a data acquisition terminal and a remote monitoring unit;

[0084] The acquisition terminal is used to collect and send temperature data of train bearings;

[0085] The remote monitoring unit is used to analyze and process the temperature data to obtain analysis results, and perform early warning processing or alarm processing according to the analysis results.

[0086] In this embodiment, Figure 3 As shown, the acquisition terminal includes a control module, a communication module and a temperature sensor module;

[0087] The temperature sensing module is used to collect the temperature data of the train bearing; the control module sends the temperature data to the remote monitoring unit through the communication module. Among them, the control module adopts the existing STM32 single-chip microcomputer; the communication module adopts the NB-IOT communication module, and the temperature sensing module includes several temperature sensors; the output end of the temperature sensor module is connected to the input end of the STM32, and the output end of the STM32 single-chip microcomputer is connected to the NB-IOT communication module, and then the temperature data is sent to the cloud platform. The acquisition terminal uses USB to access the power supply, and the train supplies power to it. The power supply equipment and power supply method adopt the existing technology, which will not be repeated here.

[0088] Among them, NB-IOT communication supports cellular data connection of low-power devices in wide area networks, also known as low-power wide area networks (LPWAN). Compared with other IoT technologies, NB-IOT has the characteristics of wide coverage, large number of connected devices, low power consumption and low cost, and is widely used in the field of industrial process monitoring.

[0089] In this embodiment, the remote monitoring unit includes a cloud platform; the cloud platform includes a network server and an alarm service device. The cloud platform refers to a hardware-based service that can provide computing, network and storage capabilities. The cloud platform uses existing technologies and will not be described in detail here.

[0090] The working principle of the train bearing temperature monitoring system is as follows: Figure 2 and Figure 4 As shown, the on-board acquisition terminal is responsible for collecting the temperature data of the corresponding bearings in the carriage, and then packs the temperature data of each bearing of the train at the same time and the carriage serial number data, and sends the data to the cloud platform through the NB-IOT communication module. The cloud platform classifies the received data according to the train number and equipment number, processes and analyzes the temperature data, and sends warning or alarm information according to the analysis results. That is, the temperature sensor module of the on-board acquisition terminal collects the axle temperature data every 1 minute, the STM32 single-chip microcomputer reads the axle temperature data collected by the sensor module, and the STM32 single-chip microcomputer packs the carriage serial number and temperature data and sends them to the NB-IOT communication module; the NB-IoT communication module sends the temperature data packet to the cloud platform, and the cloud platform unpacks the data packet to obtain the train serial number information, the carriage number information and the corresponding temperature data, and then stores it, processes and analyzes the subway axle temperature data according to the above train bearing temperature monitoring method, and issues warning or alarm information according to the analysis results.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A train bearing temperature monitoring method based on pseudo data elimination, characterized in that: The steps include: S1. Collect the temperature T of the target bearing at time n n ; S2. Determine the temperature T n Is it normal? If so, delete the oldest data in the data set and set the temperature T n Insert into the data set; if not, proceed to the next step; S3. Collect the temperature T of the target bearing at time n+1 n+1 , determine the temperature T n+1 Is it normal? If so, then the temperature T n Update and get the updated temperature T n ′ ; If not, calculate the temperature T n+1 With temperature T n The difference Δt between them is determined as follows: If the difference Δt is greater than the preset threshold ΔT or is a non-positive real number, the acquisition device is abnormal and an abnormal alarm is issued; If the difference Δt is less than or equal to the preset threshold ΔT and is a non-negative real number, proceed to the next step; S4. Collect the temperature T of the target bearing at time n+2 n+2 , determine the temperature T n+2 Is it normal? If so, the acquisition device is abnormal and an abnormal alarm is issued; if not, the temperature T is calculated. n+2 With temperature T n+1 The difference Δt1 between them is determined as follows: If the difference Δt1 is greater than the preset threshold ΔT1 or is a non-positive real number, the acquisition device is abnormal and an abnormal alarm is issued; If the difference Δt1 is less than or equal to the preset threshold ΔT1 and is a non-negative real number, it indicates that the target bearing temperature rises within a reasonable range, and the temperature T n is the effective value and according to the temperature T n Early warning of bearing temperature.

2. The train bearing temperature monitoring method based on pseudo data elimination according to claim 1 is characterized in that: The data set includes a number of normal temperature data of a train bearing before time n; wherein the train bearing includes a target bearing.

3. The train bearing temperature monitoring method based on pseudo data elimination according to claim 1 is characterized in that: According to the temperature T n Early warning of bearing temperature includes the following steps: S5. Set the temperature T n With the warning threshold T Y The following comparison is made: If the temperature T n Greater than or equal to the warning threshold T Y , then the target bearing temperature warning is issued and the next step is entered; If the temperature T n Less than the warning threshold T Y , then the target bearing temperature is normal; S6. Set the temperature T n With the warning threshold T B The following comparison is made: If the temperature T n Greater than or equal to the warning threshold T B , then the target bearing temperature alarm is issued; If the temperature T n Less than the warning threshold T B , then the target bearing temperature warning is issued.

4. The train bearing temperature monitoring method based on pseudo data elimination according to claim 1 is characterized in that: The preset threshold ΔT is determined according to the following formula: ΔT=Δt R *No; Among them, Δt R It is the maximum temperature difference per unit time when the bearing runs at full load in the actual environment; Δε is the time interval for temperature data collection.

5. The train bearing temperature monitoring method based on pseudo data elimination according to claim 1 is characterized in that: The preset threshold ΔT1 is determined according to the following formula: ΔT1=Δt R *No; Among them, Δt R It is the maximum temperature difference per unit time when the bearing runs at full load in the actual environment; Δε is the time interval for temperature data collection.

6. The train bearing temperature monitoring method based on false data elimination according to claim 1 is characterized in that: Determine the warning threshold T according to the following formula Y : T Y =T R -α*T s ; Among them, T R is k times the maximum operating temperature of the bearing material; T s is the ambient temperature of the bearing; α is the environmental factor variable.

7. The train bearing temperature monitoring method based on false data elimination according to claim 1 is characterized in that: Determine the warning threshold T according to the following formula B : T B =T V -α*T s ; Among them, T V is the maximum operating temperature corresponding to the bearing material; T s is the ambient temperature of the bearing; α is the environmental factor variable.

8. A train bearing temperature monitoring system based on pseudo data elimination using the train bearing temperature monitoring method according to any one of claims 1 to 7, characterized in that: Including data acquisition terminal and remote monitoring unit; The acquisition terminal is used to collect and send temperature data of train bearings; The remote monitoring unit is used to analyze and process the temperature data to obtain analysis results, and perform early warning processing or alarm processing according to the analysis results.

9. The train bearing temperature monitoring system based on pseudo data elimination according to claim 8 is characterized in that: The acquisition terminal includes a control module, a communication module and a temperature sensor module; The temperature sensing module is used to collect temperature data of train bearings; the control module sends the temperature data to the remote monitoring unit through the communication module.

10. The train bearing temperature monitoring system based on pseudo data elimination according to claim 8, characterized in that: The remote monitoring unit includes a cloud platform; the cloud platform includes a network server and an alarm service device.

Images