JP2013107417A - Maintenance system of rolling stock - Google Patents

Abstract

PROBLEM TO BE SOLVED: To achieve more secure state base maintenance and prevent an adverse effect of an unnecessary emergency stop command or the like onto a train operation by absolutely preventing an erroneous determination of abnormality sign due to an adverse effect of maintenance.SOLUTION: The maintenance system of rolling stock includes an abnormality sign determination device that collects and analyzes sensor data or control command values of a rolling stock and determines an abnormality sign in the rolling stock. A maintenance determination device completely determines that the rolling stock is under maintenance on the basis of position data of the rolling stock or the arrival of a predetermined planned maintenance time. When it is determined that the rolling stock is under maintenance, an abnormality sign determination suppression device prohibits determination by the abnormality sign determination device or changes a threshold value to a value that causes abnormality sign detection to become difficult, thereby absolutely preventing an erroneous determination of abnormality sign due to an adverse effect of maintenance.

Description

  The present invention relates to a railway vehicle maintenance system that performs abnormality prediction based on an estimated vehicle state.

  In order to improve the maintenance efficiency of railway vehicles, the introduction of a state-based maintenance system that estimates vehicle conditions from vehicle data (sensor values, control command values), predicts abnormalities based on the estimated vehicle conditions, and performs maintenance inspections. Progressing. Patent Document 1 below proposes an abnormality sign detection that automatically detects an abnormality sign by combining a plurality of vehicle data in this state-based maintenance system.

  In addition, although related to the maintenance of automobiles, the following Patent Document 2 states that the diagnosis conditions for performing abnormality diagnosis are made to be stricter conditions during maintenance work that affects the operation of in-vehicle devices. The discovery of a new failure factor or a potential failure factor at an early stage, and the following Patent Document 3 indicate that an abnormality diagnosis report accompanying parts replacement work or the like is prevented.

JP 2006-160153 A JP 2010-269635 A International publication WO2009 / 038028

However, when the abnormality sign determination is simply performed based on the collected data as in Patent Document 1, vehicle conditions and various detection parameters are caused by lubrication, adjustment, disassembly maintenance, parts replacement, etc. performed during maintenance such as maintenance inspection. In some cases, there is a possibility that an erroneous determination is made that there is a sign of abnormality even though it is normal when viewed from the vehicle state after maintenance.
Patent Documents 2 and 3 disclose that the cause of a failure is discovered early in connection with the maintenance work of an automobile and that an erroneous abnormality diagnosis report accompanying a parts replacement work is prevented. It cannot be applied to a system for detecting abnormal signs.

  Therefore, in order to solve the above-described problems, the present invention focuses on the fact that maintenance is performed at a depot in a railway vehicle, and that a regular maintenance plan is formulated. A maintenance determination device that reliably determines whether or not the maintenance is in progress. In this maintenance determination device, when it is determined that maintenance is being performed, determination by the abnormal sign determination device is prohibited, or the threshold value is set to a value that is difficult to detect. change.

More specifically, the following technical measures were taken in the railway vehicle maintenance system of the present invention. That is,
(1) In a railway vehicle maintenance system provided with an abnormality sign determination device that collects and analyzes railway vehicle sensor data or control command values from the railway vehicle and determines an abnormality sign of the railway vehicle, the position of the railway vehicle A maintenance determination device for determining that the railway vehicle is under maintenance based on the arrival of data or a predetermined maintenance plan time, and when the maintenance determination device determines that maintenance is being performed, the abnormality An abnormality sign determination suppression means for prohibiting the determination by the sign determination device or changing the threshold value to a value that is difficult to detect the abnormality sign is provided.

(2) In the railway vehicle maintenance system, when the maintenance includes parts replacement / adjustment, and the amount of change in the determination index of the abnormality sign determination device is greater than a predetermined value after the maintenance, the abnormality sign determination The suppression means re-learns the normal model, and the threshold for performing the abnormal sign is reset based on the normal model.

(3) In the above railway vehicle maintenance system, is the maintenance determination device performing maintenance based on position information and time information transmitted from the railway vehicle and a predetermined diagram or maintenance plan? Judged whether or not.

(4) In the above railway vehicle maintenance system, the start of maintenance execution is determined when the railway vehicle is in the depot at the scheduled maintenance time.

(5) In the above railway vehicle maintenance system, the start of maintenance execution is determined when data download for performing an abnormality determination sign is started from the railway vehicle in the vehicle base.

(6) In the above railway vehicle maintenance system, the end of maintenance execution is determined when the speed of the railway vehicle is greater than a predetermined value or when the railway vehicle is separated from the depot by a predetermined distance.

(7) In the above railway vehicle maintenance system, a display device that displays that the maintenance determination device determines that maintenance is being performed is provided.

  According to the present invention, based on the position data of a railway vehicle or the arrival of a predetermined maintenance schedule time, it is determined reliably that the railway vehicle is under maintenance, and when it is determined that maintenance is being performed, Judgment by the judgment device is prohibited, or the threshold value is changed to a value that makes it difficult to detect abnormal signs, so it is possible to reliably prevent erroneous predictions of abnormal signs due to the effects of maintenance and to make state-based maintenance more reliable. In addition, it is possible to prevent an unnecessary emergency stop command from affecting the railway operation.

The figure which shows an example of the abnormality sign determination apparatus. (A) is a relationship of the diagnostic index at the time of abnormality sign determination, (b) is a figure which shows the change of the diagnostic index during a maintenance. The figure which shows an example of the diagnostic block diagram of this invention. The figure which shows an example of the diagnostic flowchart of this invention. The figure which shows an example of the flowchart of a maintenance determination apparatus. FIG. 3 is a diagram illustrating an example of a time chart according to the first embodiment. The figure which shows another example of the flowchart of a maintenance determination apparatus. FIG. 6 is a diagram illustrating an example of a diagnosis time chart according to the second embodiment. FIG. 10 is a diagram illustrating an example of a diagnosis flowchart according to the third embodiment. The figure which shows an example of an abnormality sign determination process. The figure which shows the time chart at the time of normal model reset. The figure which shows an example of the system configuration | structure of this invention.

  Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows an example of an abnormality determination device for a door. In general, cleaning, lubrication, adjustment, obstacle sensor check, disassembly and maintenance, replacement of consumable parts, etc. are determined for each fixed period of time, and the dates and times for such maintenance inspections are set. A maintenance plan is formulated for each vehicle.
For an abnormality determination device mounted on a vehicle, the relationship between the door operation time required for opening and closing in the normal state, the driving air pressure (door driving pressure) for driving the door, and the outside air temperature is normal using, for example, vector quantization clustering. Learn as a model.

When diagnosing an abnormal sign of a door, door data detection results such as door operation time, door drive pressure, outside temperature, etc. are downloaded from the vehicle to the ground side center where the abnormal sign determination device is installed and downloaded. The distance (Mahalanobis distance) in the vector space between the door data and the normal model is calculated.
Since this distance becomes larger as the current state deviates from the normal state, an abnormal sign is determined using this distance as a diagnostic index, and the determination result of the abnormal sign is displayed on the display device.
Specifically, this display device shows a trend of diagnostic indicators, and displays a warning when the diagnostic indicators exceed a predetermined value.

The problem of the abnormality sign determination apparatus will be described with reference to FIG.
As shown in FIG. 2A, the diagnostic index calculated from the abnormality sign determination apparatus described in FIG. 1 increases as the distance from the normal state increases, and a sign or abnormality can be determined based on the magnitude.
However, as shown in FIG. 2 (b), during vehicle maintenance, lubrication, adjustment, parts replacement, overhauling, etc. are performed, so that a situation different from that during vehicle operation, that is, parameters learned as a normal model are Therefore, during maintenance, for example, the correlation between the door operation time and the door driving pressure may change greatly.

  In such a case, as shown in FIG. 2 (b), the diagnostic index transmitted from the vehicle side enters the sign area, and the door device itself is normal, but the sign of an abnormality occurring in the railway vehicle. In order to prevent an accident from occurring, the ground-side center that has detected this error may cause an emergency stop command to be sent to the railway vehicle, which may affect vehicle operation. is there.

  Therefore, in the present invention, an erroneous diagnosis during maintenance is prevented by automatically notifying the abnormality determination device that the railway vehicle is under maintenance. In other words, when the present invention is applied, the determination result output from the abnormal sign determination device by the abnormal sign determination suppression means differs depending on whether or not the railway vehicle is under maintenance even for the same diagnostic index value. This makes it possible to reliably prevent such erroneous determination.

FIG. 3 shows an example of a diagnostic block diagram of the present invention. Sensor / control data and a time stamp are sent from the vehicle (301) to the abnormality sign determination device (304), and a diagnosis determination based on the abnormality sign is performed.
At this time, the maintenance determination device (303) compares the time and position information transmitted from the vehicle (301) with the diagram (vehicle operation information) or maintenance plan obtained from the ground database, and is attached to the sensor and control data. It is determined whether maintenance is executed at the time of the time stamp. An abnormality determination result of the abnormality sign determination device and a maintenance flag (maintenance execution flag) during maintenance execution by the maintenance determination device are sent to the display device (305), and a diagnostic index for abnormality determination and a maintenance execution flag are time-series. Is displayed.

  As shown in FIG. 12, the maintenance determination device (1204), the abnormality sign determination device (1203), and the display device (1206) are generally installed at the ground-side center. Alternatively, some or all of these devices may be mounted on the vehicle, and signals from these devices may be transmitted from the vehicle upper side to the ground side center.

Example 1
A first embodiment of the present invention will be described with reference to FIGS.
FIG. 4 is an example of a flowchart for prohibiting abnormality sign determination according to the present embodiment.
In step S401, maintenance determination processing is performed. This maintenance determination process will be described later with reference to FIG.

In step S402, it is determined whether or not maintenance is being performed. If it is determined that maintenance is being performed, the process proceeds to step S403, and processing for prohibiting abnormality determination is performed. During this process, as will be described in detail with reference to FIG. 6, no abnormality determination is performed even if the diagnostic index becomes larger than the threshold value.
On the other hand, if it is determined in step S402 that the maintenance is not in progress, the process proceeds to step S404, and the sign determination as described in FIG. 1 is performed. In this way, by prohibiting abnormality sign diagnosis during maintenance, it is possible to prevent erroneous diagnosis due to maintenance work and improve the accuracy of abnormality sign diagnosis.

FIG. 5 is an example of a maintenance determination process flowchart.
In step S501, it is determined whether or not the vehicle is in a vehicle base (depot). Specifically, the position of the vehicle is estimated based on the diagram and time, or the position of the vehicle is also measured using position information from a GPS device mounted on the vehicle, and it is determined whether or not the vehicle is in a depot.

  If it is determined in step S501 that there is a vehicle in the vehicle base, the process proceeds to step S502. Otherwise, the process proceeds to step S505. In step S502, it is determined whether or not the vehicle speed is equal to or lower than a predetermined value. If the vehicle speed is equal to or lower than the predetermined value, the process proceeds to step S503, and if not, the process proceeds to step S505.

In step S503, it is determined whether the time stamp of the sensor or control data transmitted from the vehicle is within the maintenance plan time obtained from the database. If within the time, the process proceeds to step S504, and the maintenance flag is being executed in step S504. Set the flag to 1. This flag is cleared to 0 (step S505) when the vehicle position is away from the depot (step S501) or the vehicle speed is equal to or higher than a predetermined speed (step S502).
That is, according to the present invention, since the vehicle is in the depot according to the maintenance plan, it is automatically and reliably determined that the maintenance is in progress, and erroneous diagnosis during the maintenance is prevented.

FIG. 6 is an example of a time chart when the flowcharts of FIGS. 4 and 5 are implemented.
When the speed transmitted from the vehicle is equal to or lower than the predetermined speed and the maintenance plan time flag obtained by extracting the maintenance plan time based on a predetermined diagram or maintenance plan is 1, the maintenance execution flag is set to 1. Set. As a result, while the maintenance is being executed and the flag is 1, even if the diagnostic index exceeds the sign determination threshold, the sign determination flag does not become 1, it is not determined as an abnormal sign, and the determination of an abnormal sign is suppressed. Is done. In the present embodiment, it is possible to prevent erroneous diagnosis due to maintenance work by prohibiting predictive judgment during execution of maintenance.

(Example 2)
There may be cases where maintenance is not performed as planned in advance due to trouble during operation. A maintenance execution determination method that is also useful in such a case will be described with reference to FIGS. That is, in this embodiment, when the vehicle enters the depot, the vehicle information such as various sensor information and control information accumulated in the traveling vehicle is downloaded to the computer in the depot for maintenance.

  FIG. 7 is another example of the maintenance determination flowchart. Steps S701 and S702 are the same as steps S501 and S502 described in FIG. If YES is determined in step S701 and step S702, the process proceeds to step S703. In step S703, it is determined whether downloading of vehicle data has started. If YES is determined in step S703, it is determined that maintenance is being performed, and the process proceeds to step S704 where the maintenance in progress flag is set to 1.

As described above, in this embodiment, the data stored in the vehicle is downloaded in a batch at the depot.
Although some data is downloaded at any time using the mobile wireless communication network during operation, it is necessary to download a large amount of data accumulated on the vehicle to perform predictive diagnosis. Returning to FIG. 2, a large amount of data is downloaded by WiFi wireless. In particular, in the event of a vehicle failure, downloading of vehicle information is an operation that must be performed in order to perform abnormality analysis, and therefore it can be determined that maintenance is being performed with this method.

  FIG. 8 is an example of a time chart in the case where the sign determination is prohibited using download as a trigger. The maintenance execution flag is set to 1 triggered by the download of the vehicle data, and it can be confirmed that no sign determination is made even if the diagnostic index becomes larger than the sign determination threshold.

(Example 3)
If the parameter correlation during normal operation changes significantly due to component replacement or adjustment during maintenance, it is necessary to reset (relearn) the normal model. An embodiment corresponding to such a case will be described with reference to FIGS.

  FIG. 9 shows a flowchart for resetting the abnormality sign determination. Steps S901 to S904 are the same as steps S401 to S404 in FIG. In the present embodiment, step S905 is newly added. In step S905, the process shown in FIG. 10 is performed. Retrain the normal model.

  FIG. 10 is an example of a flowchart of the abnormality sign determination reset process. In step S1001, it is determined whether the maintenance is completed based on whether the maintenance flag has changed from 1 to 0. If YES in step S1001, the process proceeds to step S1002. In step S1002, it is determined whether there is a part replacement / adjustment in the maintenance plan. Alternatively, if this process is performed in a batch after the maintenance is completed, the maintenance record information may be referred to confirm whether the parts have been replaced or adjusted.

If it is determined that parts replacement / adjustment has been performed, the process advances to step S1003. In step S1003, it is determined whether the diagnostic index has changed more than a predetermined value after parts replacement / adjustment.
If it is determined that the amount of change is larger than the predetermined value, the process proceeds to step S1004, the predictive diagnosis normal model relearning flag is set to 1, and the normal model is relearned.
As described above, in this embodiment, parts are exchanged and adjusted, and as a result, only when the diagnostic index changes greatly, the normal model is automatically reset to re-learn the normal model. The threshold value for judging the sign or abnormality based on the normal model is correctly reset depending on the value, the judgment of the sign of abnormality is suppressed, and erroneous diagnosis at the time of parts replacement / adjustment can be prevented.

  FIG. 11 is an example of a time chart when the normal model is reset. The replacement / adjustment flag indicates that the replacement / adjustment has been performed during maintenance. When the maintenance execution flag changes from 1 to 0 in this state and the maintenance is completed, the diagnosis index is a predictive judgment threshold. As a trigger, the normal model relearning flag is set to 1. Then, when the normal model is re-learned and the diagnostic index is correctly calculated, the normal model re-learning flag and the exchange / adjustment flag are reset to zero.

  As described above, according to the present invention, it is reliably determined that the railway vehicle is under maintenance, and when it is determined that the maintenance is being performed, the determination by the abnormality sign determination device is prohibited or the threshold value is set. Because it is changed to a value that makes it difficult to detect abnormal signs, it is possible to reliably prevent erroneous judgments of abnormal signs due to the effects of maintenance, to realize more reliable state-based maintenance, and to prevent the effects of unnecessary emergency stop commands etc. on railway operations It can be expected to be adopted as an enabling railway vehicle maintenance system.

DESCRIPTION OF SYMBOLS 301 Railway vehicle 302 Ground database 303 Maintenance determination apparatus 304 Abnormal sign determination apparatus 305 Display apparatus 1201 Railway vehicle 1202 Maintenance center (ground)
1203 Abnormality sign determination device 1204 Maintenance determination device 1205 Ground database 1206 Display device

Claims (7)

In a railway vehicle maintenance system comprising an abnormality sign determination device that collects and analyzes railway vehicle sensor data or control command values from the railway vehicle and determines an abnormality sign of the railway vehicle,
A maintenance determination device for determining that the railway vehicle is under maintenance based on the position data of the rail vehicle or the arrival of a predetermined maintenance plan time is determined by the maintenance determination device to determine that maintenance is being performed. In some cases, the maintenance system for a railway vehicle is provided with an abnormality sign determination suppressing means for prohibiting the determination by the abnormality sign determination device or changing a threshold value to a value that is difficult to detect the abnormality sign.   2. The abnormality sign determination suppressing means according to claim 1, wherein when the maintenance includes parts replacement / adjustment, and the amount of change in the determination index of the abnormality sign determination device is greater than a predetermined value after the maintenance, the abnormality sign determination suppression unit re-starts the normal model. A railway vehicle maintenance system characterized by learning and resetting a threshold for performing an abnormality sign based on the normal model.   3. The maintenance determination device according to claim 1, wherein the maintenance determination device determines whether or not maintenance is being performed based on position information and time information transmitted from the railcar and a predetermined diagram or maintenance plan. A maintenance system for railway vehicles.   4. The railway vehicle maintenance system according to claim 3, wherein the maintenance start of the railway vehicle is determined when the railway vehicle is in the depot at the planned maintenance time.   3. The railway vehicle maintenance system according to claim 1 or 2, wherein a maintenance execution start is determined when data download for performing an abnormality determination sign is started from the railway vehicle in a vehicle base.   6. The railway vehicle maintenance system according to claim 1, wherein the end of the maintenance is determined when the speed of the railway vehicle is greater than a predetermined value or when the railway vehicle is separated from the depot by a predetermined distance. .   7. The railway vehicle maintenance system according to claim 1, further comprising a display device that displays that the maintenance determination device determines that maintenance is being performed.

Images