Railway vehicle fault diagnosis information acquisition method

Abstract

The invention discloses a rail vehicle fault diagnosis information acquisition method. An important information acquisition and storage strategy is adopted. The method comprises the following steps: setting an important information acquisition starting condition of a monitoring point and a single acquisition total duration T0; meanwhile, setting a time-division frequency-division acquisition scheme in the total time period T0; the time-phased frequency-division acquisition scheme comprises the steps that the total acquisition time T0 is divided into T1, T2,..., Ti,... and TN, the sampling frequency corresponding to each time period is F1, F2,..., Fi,... and FN, and T0 is equal to T1 + T2 +... + Ti +... + TN and F1gt; f2gt; gt,... gt; figt; gt,... gt; i is greater than or equal to 1 and less than or equal to N, and N is greater than or equal to 2, namely, the acquisition frequencies in all time periods are different and are sequentially reduced. According to the invention, all effective information of railway vehicle daily fault diagnosis can be collected, invalid information is little or even no information exists, and the problems that the number of monitoring points for railway vehicle fault diagnosis is large, the information collection amount is large, and the calculation processing amount is large can be solved.

Description

Technical Field

[0001] This invention relates to the field of electrical fault diagnosis technology for rail vehicles, and specifically to a method for collecting fault diagnosis information for rail vehicles. Background Technology

[0002] In rail transit systems such as high-speed rail, subways, and light rail, the reliability of the electrical circuits of rail vehicles is directly related to vehicle operation safety, especially public safety. The control circuits of rail vehicles drive different electrical circuits according to various control commands (such as switches, buttons, and driver controllers) and certain logical relationships, ultimately realizing the various functions required for train operation and their fault protection. Different electrical circuits in the vehicle system are controlled and electrically connected through their core control components such as relays and contactors. These components can degrade under long-term electrical and mechanical stress, leading to control failures in the vehicle system.

[0003] Currently, control failure monitoring in rail vehicle systems is primarily achieved by monitoring, recording, and analyzing abnormal electrical parameters occurring in electrical circuits. Because rail vehicles require monitoring a large number of electrical components such as relays, contactors, and switches, traditional on-off digital monitoring methods, while relatively simple in terms of data acquisition, only generate a small amount of data. However, traditional rail vehicle fault monitoring schemes only detect faults after they occur, lacking process information and making it difficult to monitor the quantitative changes in electrical components, thus hindering early warning and maintenance to prevent future problems.

[0004] In recent years, fault early warning technology for rail vehicles has received increasing attention. However, this early warning analysis requires waveform recording at every monitoring point of the electrical components. Because instantaneous changes need to be captured, the information acquisition frequency generally needs a sampling rate of 200KHz or higher. For a rail vehicle with thousands of monitoring points, averaging 16 hours of monitoring per day, and each piece of information calculated as 10 bytes (including time, status, detection value, etc.), the daily data volume reaches TB (10^10 kilobytes). 12 The records will be extremely large (in bytes), but only a very small portion of this data is truly useful for fault early warning and diagnosis. Both data storage, retrieval and processing face great difficulties, so there is an urgent need for a more reasonable information collection method. Summary of the Invention

[0005] Currently, the collection of fault diagnosis information for rail vehicles typically uses a fixed frequency. A low collection frequency may miss crucial fault information, making source tracing and analysis difficult; a high frequency, however, may result in insufficient data volume and processing capacity. This situation makes system design difficult to balance; both high and low frequencies could lead to significant waste of storage and computing resources without contributing to fault diagnosis.

[0006] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a new method for collecting fault diagnosis information of rail vehicles.

[0007] To achieve the above objectives, the present invention designs a method for collecting fault diagnosis information of rail vehicles. The information collection method includes an important information collection and storage strategy, wherein the important information includes information on the presence of action, state change or abnormality at monitoring points. The important information collection and storage strategy includes: setting the start condition (or trigger condition) for important information collection at monitoring points and the total duration T0 of a single collection, that is, collecting data only when needed. Generally, when electrical components have switching actions or state changes, information such as voltage, current, temperature, or vibration is collected at least from the start time of the change to the end time of the change, thereby greatly reducing the amount of data collected; different monitoring points can define different start conditions and total duration T0 of a single collection; at the same time, a time-segmented and frequency-divided collection scheme is set within the total duration T0. The time-segmented and frequency-divisional acquisition scheme includes dividing the total acquisition duration T0 into T1, T2, ..., T... i ... T N The sampling frequencies corresponding to each time period are F1, F2, ..., F... i ..., F N And T0 = T1 + T2 + ... + T i +...+T N F1>F2>...>F i >...>F N Where 1≤i≤N, N≥2, that is, the sampling frequency is different in each time period and decreases sequentially.

[0008] Furthermore, the information acquisition start conditions include the first contact or disengagement time of the action switch, or the state change rate condition (state change per unit time); the total duration of a single acquisition is at least as long as the time from the start to the end of the action or state change of the measuring point, and can be slightly longer than this time, to ensure sufficient information acquisition and no omissions in state change analysis.

[0009] Furthermore, the state change rate condition judgment method includes continuously collecting and cyclically storing at least two cycles of information for each monitoring point at high frequency, calculating and analyzing the state change per unit time, and activating an important information collection and storage strategy when the change exceeds a threshold.

[0010] Furthermore, the method for implementing the first contact or disengagement time condition of the action switch includes setting linkage contacts at the monitoring point. The linkage contacts are distinguished by their sequence of action. The contact that acts first sends an important information collection and storage start signal, and the contact that acts later is the monitored contact.

[0011] Furthermore, the time-division and frequency-division acquisition scheme includes: T0=1s, N=3, and wherein: T1 time period: 0~100ms, system action start, large disturbance stage status data, monitoring the highest sampling frequency F1=250kHz, (1s / 250k=4us). It is not important whether the time period boundary points include the sampling frequency of this time period, and it is acceptable. The same applies below. T2 period: 100-200ms, state data during the middle of a large disturbance in the system, with the highest sampling frequency F2=200kHz during monitoring; T3 period: 200ms to 1s, which is the state data after a large disturbance in the system. The highest sampling frequency during monitoring is F3=50Hz.

[0012] Furthermore, the important information collection and storage strategy also includes: during the T1 or T2 period, if the same signal encounters a new start criterion, the sequential collection and storage will not be restarted; during the monitoring process in the T3 period, if the same signal encounters a new start criterion, the collection and storage will restart from T1 and proceed sequentially according to the predetermined strategy.

[0013] Furthermore, the time-segmented and frequency-divisional acquisition scheme includes: T0=1s, N=2, and wherein: T1 time period: 0~200ms, system action start, large disturbance stage status data, monitoring the highest sampling frequency F1=250kHz, (1s / 250k=4us). It is not important whether the time period boundary points include the sampling frequency of this time period, and it is acceptable. The same applies below. T2 period: 200ms to 1s, which is the mid-term state data after a large disturbance in the system. The highest sampling frequency during monitoring is F2=50Hz.

[0014] Furthermore, the important information collection and storage strategy also includes the following: when the same monitoring point encounters a restart signal (a new important information collection and storage start signal) during the important information collection and storage process: if the restart signal occurs within time period T1, then collection and storage continue according to the established strategy; if the restart signal occurs within time period T... N During the time period, the strategy starting from time period T1 will continue or re-collect and store data.

[0015] Furthermore, the information acquisition method also includes a general information acquisition and storage strategy, which includes a strategy for low-frequency continuous acquisition and storage of monitoring point information under normal conditions.

[0016] Furthermore, the information includes at least one of the following: voltage, current, vibration, temperature, etc. at the monitoring point.

[0017] The advantages and beneficial effects of this invention are as follows: This invention designs an important information acquisition and storage strategy, namely, a high-frequency acquisition and storage strategy is used during periods of drastic signal changes to preserve detailed information about the drastic signal changes; a high / medium-frequency acquisition and storage strategy is used during periods of moderate signal changes; a medium / low-frequency acquisition and storage strategy is used during periods of slow signal changes; and a low-frequency acquisition and storage strategy is used during periods of relatively constant signal changes. By combining the important information acquisition and storage strategy with the general information acquisition and storage strategy and rationally setting the information acquisition start conditions, this invention achieves the complete acquisition of all effective information for daily fault diagnosis of rail vehicles, with little or no invalid information, thus solving the problems of a large number of monitoring points, large amount of information acquisition, and large amount of computational processing required for fault diagnosis of rail vehicles.

[0018] For an independent monitoring point, the preliminary estimate is that the average number of actions or status changes per day is 100, and the cumulative storage time required is generally 100 seconds. Moreover, the proportion of high-frequency data in these 100 seconds of data is less than 1 / 5 of the traditional 1 second of data. Based on an estimate of 16 hours and 57,600 seconds per day, the amount of data per day for a single contact point is less than 1 / 2880 of the original high-frequency recording method. On the other hand, in reality, most monitoring points have far fewer than 100 actions or status changes per day. It can be seen that the data storage capacity of this invention is significantly reduced, which will also lead to a corresponding significant reduction in the workload of data retrieval and computation processing. Attached Figure Description

[0019] Figure 1 This is a flowchart of a rail vehicle fault information collection method according to the present invention; Figure 2 This is a waveform diagram of the voltage at the instant the coil of a relay is de-energized; Figure 3 This is a waveform diagram of the current at the instant the coil of a relay is energized; Figure 4 This is a waveform diagram of the current at the instant the coil of a relay is de-energized; Figure 5 This is a voltage waveform diagram of the normally open contact of a relay at the moment of energization; Figure 6 This is a voltage waveform diagram of the normally open contact of a relay at the moment of de-energization. Detailed Implementation

[0020] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings and examples. The following examples are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0021] Example 1: like Figure 1As shown, the present invention is a method for collecting fault diagnosis information of rail vehicles. The information collection method includes an important information collection and storage strategy, and the important information includes information on the presence of action, status change or abnormality at monitoring points. The important information collection and storage strategy includes: setting the start condition (or trigger condition) for important information collection at monitoring points and the total duration T0 of a single collection, that is, collecting data only when needed. Generally, when electrical components have switching actions or state changes, information such as voltage, current, temperature, or vibration is collected at least from the start time of the change to the end time of the change, thereby greatly reducing the amount of data collected; different monitoring points can define different start conditions and total duration T0 of a single collection; at the same time, a time-segmented and frequency-divided collection scheme is set within the total duration T0. The time-segmented and frequency-divisional acquisition scheme includes dividing the total acquisition duration T0 into T1, T2, ..., T... i ... T N The sampling frequencies corresponding to each time period are F1, F2, ..., F... i ..., F N And T0 = T1 + T2 + ... + T i +...+T N F1>F2>...>F i >...>F N Where 1≤i≤N, N≥2, that is, the sampling frequency is different in each time period and decreases sequentially.

[0022] The basic idea of ​​this invention is to analyze multiple high-frequency acquisitions of information from different types of monitoring points on rail vehicles to obtain the time-frequency characteristics (empirical information and rules) of related information changes when the state changes. Based on these time-frequency characteristics, acquisition periods are divided: high-frequency acquisition and storage strategies are used for periods of drastic signal change (each period having a relatively fixed duration) to preserve detailed information about these changes; high / medium-frequency acquisition and storage strategies are used for periods of moderate signal change; medium / low-frequency acquisition and storage strategies are used for periods of slow signal change; and low-frequency acquisition and storage strategies are used for periods of relatively constant signal. There are no clear boundaries for determining the specific high, medium, and low frequency values; they are generally determined based on the signal characteristics. For specific parameters, such as the high-frequency band of a contactor being close to 100kHz, the sampling frequency (information acquisition frequency) should be set to 200kHz or higher. For low-frequency signals, the sampling frequency can be selected within the range of 0.01Hz, 0.1Hz, 1Hz, 10Hz, or even lower frequencies, such as sampling once every 10 minutes or 30 minutes. Generally, different monitored information may have different time-frequency characteristics. However, for convenience, if the time-frequency characteristics are similar, all or most monitoring points can adopt the same time-segmented and frequency-segmented acquisition and storage scheme, that is, the total duration, time segmentation, and acquisition frequency of each time segment are consistent. This also facilitates the subsequent unified data processing and analysis process.

[0023] Based on the above data acquisition and storage strategy, when a monitoring point experiences an action or abnormal state, according to the transient time-frequency characteristics of the monitoring point and following the principle of facilitating fault source analysis, transient data should be recorded in detail at an appropriate frequency. Steady-state data, which is of little significance for fault analysis but is abundant, is discarded or recorded in small quantities. Furthermore, different data acquisition and storage strategies can be specified for each monitoring point. This effectively solves the problem of a large number of monitoring points and a large volume of monitoring data for rail vehicles. The fault diagnosis system can be configured with different data acquisition and storage mechanisms according to the actual scenario to maximize the utilization of storage space.

[0024] Preferably, the information acquisition start conditions include the first contact or disengagement time of the action switch (which can be implemented through hardware settings, such as linkage contacts, but the actions are sequential; the contact that acts first sends an important information acquisition and storage start signal, and the contact that acts later is the monitored contact; or the important information acquisition and storage start signal is given manually / semi-manually, or a centralized start signal is given according to a fixed process and cycle of control actions, and important information is acquired and stored at each monitoring point according to a predetermined process at a fixed time period), or the state change rate condition (state change per unit time). This embodiment uses the state change rate condition for analysis and judgment. The total duration of a single acquisition is at least not shorter than the time from the start to the end of the measurement point action or state change, and can be slightly longer than this time to ensure sufficient information acquisition and no omission of state change analysis.

[0025] Preferably, the state change rate condition judgment method described in this embodiment includes continuously collecting and cyclically storing at least two cycles of information for each monitoring point at high frequency (equivalent to caching; according to the first-in-first-out principle, expired information is not normally stored; in this embodiment, to more accurately determine the start time of state change, temporary sampling information for 100 cycles is continuously cached), and calculating and analyzing the state change per unit time. When the value exceeds a threshold, an important information collection and storage strategy is activated. The formally stored information may include the most recent continuously collected and cyclically stored information. The state change rate of the measured quantity can be calculated after two consecutive sampling cycles. If there are multiple cycles, the calculation of the change rate is more reliable. Related methods include the general difference method or the least squares method, which are mature technologies. In contrast, dynamic high-frequency acquisition is used. It's relatively easy to implement a cyclic storage system to determine whether important information collection and storage needs to be initiated. However, confirming whether the state has changed requires a targeted judgment method and threshold based on the time-frequency characteristics of the contact action. The specific threshold setting depends on the actual magnitude of the signal value at the monitoring point, its detection error, and the allowable variation. Different thresholds can be set for different monitoring points. In fact, using the threshold judgment method is simpler and more convenient, and there are no major drawbacks. Even if it's an interference signal (not caused by actual action or state change), recording it doesn't have a significant adverse impact, as it can be used to analyze interference characteristics and patterns. It only occupies some more storage space, but this extra space is not too large, and the total duration is limited. If the duration is too long, manual intervention and analysis can be initiated to find the cause and solve the problem.

[0026] Preferably, the information includes at least one of the following: voltage, current, vibration, temperature, etc. at the monitoring point.

[0027] Preferably, in this embodiment, for monitoring the coil voltage of AC and DC contactors, relays, various switches, solenoid valves, sensors, etc. in general rail vehicles, the time-segmented and frequency-division acquisition scheme includes: T0=1s, N=2, and wherein: T1 time period: 0~200ms, system action start, large disturbance stage status data, monitoring the highest sampling frequency F1=250kHz, (1s / 250k=4us). It is not important whether the time period boundary points include the sampling frequency of this time period, and it is acceptable. The same applies below. Time period T2 (200ms to 1s) represents the intermediate state data after a large system disturbance, with a maximum sampling frequency of F2 = 50Hz during monitoring. After the total duration ends, the system returns to normal and data collection and storage ceases. Alternatively, data can be continuously collected and stored at low frequencies, or continuously collected and stored at high frequencies in a loop. For specific methods, please refer to the preferred scheme in Example 1.

[0028] Preferably, the information acquisition method further includes a general information acquisition and storage strategy, which includes a strategy of continuously acquiring and storing monitoring point information at low frequency under normal conditions. In order to ensure the continuity and integrity of information storage, a low-frequency acquisition strategy can also be adopted simultaneously under normal conditions, that is, a general information acquisition and storage strategy, such as continuously acquiring and storing the information of the measurement points at sampling frequencies of 0.01Hz, 0.1Hz, 1Hz, 10Hz, etc. That is, under normal conditions, low-frequency acquisition and storage of all information (because there is no change or the change is very small, low frequency is sufficient to acquire signal characteristics), high-frequency acquisition of buffer window information (once there is a change, it can be known in time, and the information can also be effectively saved), and high-frequency / medium-frequency acquisition and storage of information for a specific duration when there is an anomaly or when it is needed.

[0029] Taking the relay coil voltage as an example, according to... Figure 2 The image shown is a magnified view of the voltage fluctuation at the instant the relay coil loses power. When the relay loses power, the coil voltage momentarily generates a reverse voltage of approximately -180V, which persists for about 20ms before gradually disappearing. The entire power-off process lasts approximately 40ms. A high-frequency sampling period of 200ms is sufficient to completely record the state of the relay coil at the instant it loses power.

[0030] Regarding the signal restart issue, if a restart signal is encountered during the collection and storage of important information, and if the total storage space is not tight and the restart probability is not high, from the perspective of the comprehensiveness of information preservation, the collection and storage can be restarted as soon as a new start criterion is encountered during the entire collection and recording process, and the previous storage is still preserved (the specific detection information is accompanied by time information).

[0031] Therefore, a more comprehensive rail vehicle fault information collection scheme, for each monitoring point, includes the following specific steps: S1. Set the start conditions for collecting important information at the monitoring point and the total duration of a single collection, T0. At the same time, set the time-segmented and frequency-segmented collection scheme within the total duration T0. S2. Low-frequency acquisition and storage of general information, synchronous continuous high-frequency acquisition and cyclic storage, and determination of whether the conditions for starting the acquisition of important information are met based on the high-frequency acquisition data. S3. When the conditions for starting the collection of important information are met, the collection and storage will be carried out at frequency F1 during time period T1 and at frequency F2 during time period T2, and so on, until the collection is completed according to the total duration T0. S4. During the data collection and storage process in step S3, if the same monitoring point encounters new important information collection start conditions, the data collection and storage will continue from time period T1 until the data collection is completed according to the new total duration T0. S5. Analyze and process the data collected and stored at each monitoring point according to the established method to determine whether a fault has occurred or whether data anomalies may cause a fault (i.e., fault warning), and thus prompt or alarm for handling. There are mature technologies in this area, and it is not a problem that needs to be solved by this invention.

[0032] Example 2: The difference from Embodiment 1 is that the information collection start condition criterion in this embodiment is the first contact or disengagement time of the action switch. The implementation method includes setting linkage contacts at the monitoring point. The action of the linkage contacts is distinguished by the order of action. The contact that acts first sends an important information collection and storage start signal, and the contact that acts later is the monitored contact. Generally, the contact that acts later only needs to delay the action by about 1ms relative to the contact that acts earlier at the normal action speed to ensure that the collection and storage record can contain the action start state information of the monitored contact.

[0033] Example 3: The difference from Example 1 is that this example monitors the contact voltage of AC and DC contactors, relays, and various switches in general rail vehicles. The time-segmented and frequency-division-based acquisition scheme includes: T0=1s, N=3, and wherein: T1 time period: 0~100ms, system action start, large disturbance stage status data, monitoring the highest sampling frequency F1=250kHz, (1s / 250k=4us). It is not important whether the time period boundary points include the sampling frequency of this time period, and it is acceptable. The same applies below. T2 period: 100-200ms, state data during the middle of a large disturbance in the system, with the highest sampling frequency F2=200kHz during monitoring; T3 period: 200ms to 1s, representing the state data after a large system disturbance, with a maximum sampling frequency of F3=50Hz during monitoring. After the total duration ends, the system returns to normal and no further data is collected or stored. Alternatively, data can be collected and stored continuously at low frequencies, or continuously collected at high frequencies and stored cyclically. For specific methods, please refer to the preferred scheme below.

[0034] Specifically, such as Figure 3 As shown, taking the current waveform during the energization of a relay as an example for analysis, the fluctuations are relatively violent in the T1 period (approximately 100-200ms), the fluctuations slow down in the T2 period (approximately 200-300ms, not fully shown in the figure), and there are still occasional fluctuations in the T3 period (after approximately 300ms, not shown in the figure), but the frequency is significantly reduced. The important information acquisition strategy adopted in this embodiment is reasonable. Other typical waveforms are shown below. Figure 4 This is a waveform diagram of the current at the instant the relay coil is de-energized. Figure 5 This is a voltage waveform diagram of the normally open contact of a relay at the moment of energization. Figure 6It is the voltage waveform diagram of the normally open contact of a certain relay at the moment of losing power. The violent fluctuations all occur within the first 200 ms, and most within the first 100 ms.

[0035] Preferably, specifically, in the auxiliary diagnosis of sudden faults (transient state), when the start condition is met (the current value rises or falls), the system starts collecting and storing according to the established strategy from T1; regarding the restart problem of signal collection, the important information collection and storage strategy further includes: within the T1 or T2 period, if the same signal encounters a new start criterion, it is collected and stored sequentially without restarting; during the monitoring process of the T3 period, if the same signal encounters a new start criterion, it starts from T1 again and is collected and stored sequentially according to the established strategy.

[0036] Preferably, in this embodiment, continuous high-frequency collection for the state criterion and cyclic storage are for 2-cycle information, which is the simplest state. If the signal monitoring values between adjacent two cycles change beyond the set threshold, the collection and storage of important information are started.

[0037] Embodiment 4: Regarding the restart problem, the difference from Embodiment 1 is that the important information collection and storage strategy in this embodiment further includes when the same monitoring point encounters a restart signal during the important information collection and storage process (a new important information collection and storage start signal): If the restart signal occurs within the T1 period, continue to collect and store according to the established strategy; if the restart signal occurs within the T N period, continue or restart the collection and storage according to the strategy starting from the T1 period. Here, essentially, the collection frequency strategy starts over, and the storage can be overwritten or continuously stored. Generally, this situation is not common, and continuous storage does not increase the obvious storage requirements; if N>2 and the restart signal occurs within the T i period, when 1<i<N, it needs to be analyzed according to specific circumstances. Generally, when T1+T{2}+...+T i <(T i+1 +...+T N ) / 2 and F i <F i+1 / 2, continue to collect and store according to the established strategy, otherwise start over according to the strategy starting from the T1 period; as an insurance measure, when a restart signal appears within the T2 and subsequent periods, continue or restart the collection and storage according to the strategy starting from the T1 period.

[0038] Embodiment 5: The difference from Embodiment 1 is that this embodiment only executes the important information collection and storage strategy and does not execute the general information collection and storage strategy. That is, information collection and storage are only started when the monitoring point has an action, state change or abnormality. Under normal circumstances, information is not collected, and the normal state information of the monitoring point can be obtained through empirical values or a small amount of real-time collection.

[0039] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, including different segment divisions, different embodiments and recombination of their local solutions, etc. These improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for collecting fault diagnosis information of rail vehicles, characterized in that, The information collection method includes an important information collection and storage strategy, which includes: setting the start conditions for important information collection at the monitoring point and the total duration T0 of a single collection; and setting a time-segmented and frequency-segmented collection scheme within the total duration T0. The time-segmented and frequency-divisional acquisition scheme includes dividing the total acquisition duration T0 into T1, T2, ..., T... i ... T N The sampling frequencies corresponding to each time period are F1, F2, ..., F... i ..., F N And T0 = T1 + T2 + ... + T i +...+T N F1>F2>...>F i >...>F N Where 1≤i≤N, N≥2, that is, the sampling frequency is different in each time period and decreases sequentially.

2. The method for collecting fault diagnosis information of rail vehicles according to claim 1, characterized in that, The information acquisition start conditions include the first contact or disengagement time of the action switch, or the state change rate condition; the total duration of a single acquisition is at least not less than the time from the start to the end of the action or state change of the measuring point.

3. The method for collecting fault diagnosis information of rail vehicles according to claim 2, characterized in that, The state change rate condition judgment method includes continuously collecting and cyclically storing at least two cycles of information for each monitoring point at high frequency, calculating and analyzing the state change per unit time, and activating an important information collection and storage strategy when the change exceeds a threshold.

4. The method for collecting fault diagnosis information of rail vehicles according to claim 2, characterized in that, The method for determining the initial contact or disengagement time of the action switch includes setting linkage contacts at the monitoring point. The linkage contacts are activated sequentially, with the first contact activating sending an important information collection and storage start signal, and the subsequent contact activating is the monitored contact.

5. The method for collecting fault diagnosis information of rail vehicles according to claim 1, characterized in that, The time-division and frequency-division acquisition scheme includes: T0=1s, N=3, and wherein: T1 time period: 0-100ms, F1=250kHz; T2 time period: 100-200ms, F2=200kHz; T3 time period: 200ms~1s, F3=50Hz.

6. The method for collecting fault diagnosis information of rail vehicles according to claim 5, characterized in that, The important information collection and storage strategy also includes: during the T1 or T2 period, if the same signal encounters a new start criterion, the sequential collection and storage will not be restarted; during the monitoring process in the T3 period, if the same signal encounters a new start criterion, the collection and storage will start again from T1 and proceed sequentially according to the established strategy.

7. The method for collecting fault diagnosis information of rail vehicles according to claim 1, characterized in that, The time-division and frequency-division acquisition scheme includes: T0=1s, N=2, and wherein: T1 time period: 0~200ms, F1=250kHz; T2 time period: 200ms~1s, F2=50Hz.

8. The method for collecting fault diagnosis information of rail vehicles according to claim 1, characterized in that, The important information collection and storage strategy also includes the following: when the same monitoring point encounters a restart signal during the important information collection and storage process: if the restart signal occurs within time period T1, collection and storage will continue according to the established strategy; if the restart signal occurs within time period T... N During the time period, the strategy starting from time period T1 will continue or re-collect and store data.

9. A method for collecting fault diagnosis information of rail vehicles according to claim 1, characterized in that, The information acquisition method also includes a general information acquisition and storage strategy, which includes a strategy for low-frequency continuous acquisition and storage of monitoring point information under normal conditions.

10. A method for collecting fault diagnosis information of rail vehicles according to claim 1, characterized in that, The information includes at least one of the following: voltage, current, vibration, and temperature information of the monitoring point.

Images