Sensor diagnosis system and sensor diagnosis method

Abstract

To provide a technique that can prevent a performance degradation due to a position of a train or an external factor from being determined as an abnormality of a sensor and enables achievement of stable train operation.SOLUTION: A sensor diagnosis system for evaluating a performance of a sensor mounted on a train, includes: a database storing information on an existing facility that can be detected by the sensor according to a position of the train; a train control unit that detects the position of the train; and a sensor abnormality determination unit that determines that there is an abnormality in the sensor when a state of the existing facility detected by the sensor does not match information on the existing facility stored in the database.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a sensor diagnostic system and a sensor diagnostic method for detecting an abnormality in a sensor. [Background technology]

[0002] In recent years, research has been conducted into automating operation of existing track transportation systems due to concerns about labor shortages caused by the aging of drivers and the need to reduce operation costs. In track transportation systems, where transportation vehicles run on a track, if there is an obstacle on the track, the vehicle cannot steer to avoid it, so detecting obstacles on the track is important for improving the safety and operability of the track transportation system. Currently, drivers detect obstacles on the track and route by visual inspection.

[0003] On the other hand, unmanned operation requires a mechanism for automatically detecting obstacles on the route, and methods using sensors such as millimeter wave radar, LiDAR, and cameras are being researched. Obstacle detection depends on the performance of the sensors, and if some kind of abnormality occurs and the sensor is no longer able to perform as expected, there is a problem that it will no longer be able to detect obstacles. Therefore, in order to ensure the safety of train operation, it is necessary to be able to detect abnormalities in the sensors.

[0004] As a technology for detecting abnormalities in a sensor, for example, Patent Document 1 discloses a configuration in which a database is stored that records the installation positions and installation identifiers of ground-based objects, the installation identifier is identified from sensor information including information detected by the sensor, and the database is referred to. As a sensor's accuracy assurance performance, an accuracy assurance detection distance is calculated, and the calculated detection distance is used to determine sensor failure. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2021-069162 A Summary of the Invention [Problem to be solved by the invention]

[0006] In unmanned operation, if some abnormality occurs in the sensor and it is unable to detect obstacles, the safety of the train's operation cannot be guaranteed, and the train must be stopped. On the other hand, since the actual performance of the sensor can vary depending on the position of the train (ground installations) and surrounding external factors (weather, brightness), it is possible that the sensor's detection performance may decrease even if there is no abnormality in the sensor, and the decrease in detection performance may be judged as a sensor abnormality. If such a decrease in the sensor's detection performance due to the surrounding environment is judged as a sensor abnormality and the train is stopped, the train may stop frequently, and stable train operation may not be realized. In the technology described in Patent Document 1, if the accuracy guarantee detection distance fluctuates due to the influence of the surrounding environment, it may be judged as a sensor abnormality.

[0007] Therefore, it is necessary to build a sensor diagnostic system that, when a degradation in sensor performance is detected, can determine whether the degradation is due to the train's position or external factors, and whether the degradation is acceptable from the perspective of ensuring the safety of train operation, or whether the degradation is something that cannot be ignored from a safety standpoint, thereby enabling both safety and stable operation to be achieved.

[0008] An object of the present invention is to provide technology that prevents performance degradation due to train position or external factors from being judged as a sensor abnormality, thereby enabling stable train operation. [Means for solving the problem]

[0009] In order to solve the above problems, one representative sensor diagnosis system of the present invention is a sensor diagnosis system that evaluates the performance of a sensor installed on a train, and includes a database that stores information on existing equipment that can be detected by the sensor according to the position of the train, a train control unit that detects the position of the train, and a sensor abnormality determination unit that determines that there is an abnormality in the sensor if the state of the existing equipment detected by the sensor at the position of the train detected by the train control unit does not match the information of the existing equipment stored in the database. Effect of the Invention

[0010] As a result of the above, by creating a database of changes in sensor performance due to changes in train positions and external factors, it is possible to prevent performance degradation due to train positions or external factors from being judged as a sensor abnormality, making it possible to achieve stable train operations.

[0011] Problems, configurations and effects other than those described above will become apparent from the following description of the preferred embodiment of the invention. [Brief description of the drawings]

[0012] [Figure 1] FIG. 1 is a diagram illustrating an overview of a sensor diagnostic system according to a first embodiment of the present invention. [Diagram 2] 1 is a diagram showing a configuration of an on-board control device according to a first embodiment. [Diagram 3] FIG. 4 is a diagram showing an example of a data configuration of a detected object DB in the first embodiment. [Figure 4] FIG. 4 is a diagram illustrating a process flow of sensor abnormality detection in the first embodiment. [Diagram 5] FIG. 11 is a diagram showing an outline of a sensor diagnostic system according to a second embodiment of the present invention. [Figure 6] FIG. 13 is a diagram showing an example of a data configuration of a detected object DB according to the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Note that the present invention is not limited to the embodiment. In the description of the drawings, the same parts are denoted by the same reference numerals. EXAMPLES

[0014] Fig. 1 shows an overview of the sensor diagnostic system in the first embodiment of the present invention. Fig. 2 shows the configuration of an on-board control device 201 mounted on a train 101. The operation of the on-board control device 201 shown in Fig. 2 will be described with reference to Fig. 1. The explanation will be given taking the case where rails are targeted as an example of existing detectable equipment. "Existing equipment" refers to existing equipment installed along the railway line, and in addition to rails, possible examples include signals, tool boxes, etc.

[0015] The train 101 detects the rails using the sensor 203. In this embodiment, a camera is used as the sensor 203, and the rails are detected by image processing of the image captured by the sensor 203. The sensor control unit 204 detects the rails by image processing of the image input from the sensor 203, and can calculate the length of the detected rail as the rail detection distance 103. Assuming that the rails are installed on a flat surface, the rail position in three-dimensional space can be analyzed from the two-dimensional image, and the length of the rail shown in the two-dimensional image, that is, the rail detection distance, can be calculated geometrically. The detected object DB 102 is a database in which the length of the rail that can be detected at the position of the train 101, that is, the rail detection distance, is stored. Three databases, morning, afternoon, and night, are created in advance according to the time zone in which the train 101 runs. One of the three DBs is selected as the detected object DB 102 to be used according to the time zone in which the train 101 runs.

[0016] The sensor abnormality determination unit 202 receives the train position from the train control unit 205, refers to the detected object DB 102 at the received train position, and obtains the rail detectable distance at the current position. The sensor abnormality determination unit 202, which has obtained the rail detectable distance, compares the rail detectable distance with the rail detection distance 103 calculated by the sensor control unit 204, and can determine that a sensor abnormality has occurred if the rail detection distance 103 is shorter than the rail detectable distance by more than a predetermined value. However, at a train position where the rail shape is curved, when passing an oncoming train, the oncoming train may become an obstacle and the rail beyond the curve may not be visible, resulting in a shorter rail detection distance 103. Therefore, at a train position where the rail shape is curved, the presence or absence of an oncoming train is determined by image processing, and if there is an oncoming train, sensor diagnosis is not performed. This makes it possible to prevent erroneous sensor diagnosis due to an oncoming train. Similarly, it is assumed that the rail detection distance 103 may be shortened due to installations on the ground, but in such a case, the rail detection distance in the detected object DB 102 at the train position is also shortened, so this does not pose a problem.

[0017] An example of the data configuration of the detected object DB 102 is shown in Fig. 3. Since it is expected that the rail detectable distance will differ between the upbound and downbound directions, a detected object DB 102 is created for each upbound and downbound direction, and Fig. 3 shows an example for the downbound direction. "Position" refers to the distance in kilometers of the position of the train 101, and refers to the starting point where the rail detectable distance changes. "Detectable distance" is the rail detectable distance at that position. Therefore, for example, in Fig. 3, the rail detectable distance when the train is located 300m to 400m is 200m. The data for the detected object DB 102 can be created in advance, for example, by statistically processing the detection results of the sensor 203 when the train has run repeatedly in advance.

[0018] 4 shows a flow of processing for determining a sensor abnormality by the sensor abnormality determination unit 202. The sensor abnormality determination unit 202 periodically performs the flow shown in FIG.

[0019] Step 401: The sensor abnormality determination unit 202 acquires the rail detection distance 103 currently detected by the sensor 203 from the sensor control unit 204. At this time, the sensor control unit 204 also performs image processing to determine whether the rail shape is straight or curved, and to determine the presence or absence of an oncoming train, and the sensor abnormality determination unit 202 acquires each of these results from the sensor control unit 204.

[0020] Step 402: The sensor abnormality determination unit 202 acquires the current train position from the train control unit 205, and refers to the detected object DB 102 using the acquired train position to acquire the rail detectable distance at the current position.

[0021] Step 403: Based on the results of the determination in step 401 of whether the rail shape is straight or curved and the presence or absence of an oncoming train, the sensor anomaly determination unit 202 ends the process without performing sensor diagnosis if the rail shape is curved and there is an oncoming train. If the rail shape is not curved or there is no oncoming train, the process proceeds to step 404 to perform sensor diagnosis.

[0022] Step 404: The sensor abnormality determination unit 202 compares the rail detection distance 103 acquired in the step 401 with the rail detectable distance acquired in the step 402, and determines whether the rail detection distance 103 is shorter than the rail detectable distance by a predetermined value or more. If it is shorter, the process proceeds to step 405. If it is not shorter, it is determined that there is no abnormality in the sensor and the process ends.

[0023] Step 405: The sensor abnormality determination unit 202 determines that the sensor is abnormal and executes a predetermined abnormality process. The abnormality process may be, for example, braking output if the train is unable to run, or only an alarm output if the train can continue to run. The process may be determined in advance taking into account the importance of the sensor abnormality.

[0024] Through the above processing, by selecting the database to be referenced depending on the state of external factors, it becomes possible to determine sensor abnormalities taking external factors into account; by appropriately selecting the database, it becomes possible to prevent a deterioration in sensor performance due to the surrounding environment from being determined as a sensor abnormality, thereby enabling stable train operation to be achieved.

[0025] In this embodiment, three databases, one for morning, one for afternoon, and one for night, are created, and a database is selected by time. This allows the sensor to be diagnosed without referring to the illuminance around the train. On the other hand, a configuration may be used in which a database is selected according to the illuminance around the train, rather than the time. Specifically, the illuminance at the time each database was created is recorded in the three databases, and a light meter is connected to the on-board control device 201. The sensor abnormality determination unit 202 inputs the illuminance around the train from the light meter, compares the input current illuminance with the illuminance recorded in each database, and selects the database with the closest illuminance. By referring to the illuminance, the sensor can be correctly diagnosed even if the illuminance around the train changes due to factors other than the time, such as weather.

[0026] Furthermore, instead of using the rail detection distance in the database as is, the rail detection distance at the current position may be calculated by interpolating values ​​from multiple databases. Specifically, when using illuminance, a database with illuminance higher than the current illuminance and a database with illuminance lower than the current illuminance are referenced to obtain the rail detection distance at each current position. Then, the illuminance in each database and the current illuminance are used to interpolate the two obtained rail detection distances to calculate the rail detection distance at the current position under the current illuminance. Interpolation enables sensor anomaly detection with higher accuracy.

[0027] In addition, in this embodiment, the configuration has been described with a camera as the sensor, but sensors other than cameras can be diagnosed in the same way as long as they can detect rails. For example, infrared cameras, LiDAR, and millimeter wave radars can also be diagnosed with the configuration of this embodiment. Since it is expected that the rail detection distance 103 differs depending on the sensor, when diagnosing multiple sensors, the detected object DB 102 is created separately for each sensor.

[0028] When creating a detected object DB102 for a plurality of sensors, the detected object DB102 may be created and selected according to the characteristics of the sensor. As for an infrared camera, since the sensor performance changes with the change in the ambient temperature, a plurality of detected object DB102 with different ambient temperatures may be created, a thermometer may be connected to the on-board control device 201, and a database to be used may be selected according to the ambient temperature. This allows the infrared camera to be used as a sensor. As for a LiDAR, since the performance decreases due to rain, a plurality of detected object DB102 with different rainfall amounts may be created, a rain gauge may be connected to the on-board control device 201, and a database to be used may be selected according to the amount of rainfall. This allows the LiDAR to be used as a sensor. In addition, the image captured by the camera is affected not only by the illuminance but also by weather such as rain and snow. To take into account the effects of illuminance and weather, a detected object DB102 may be created for each combination of illuminance and weather. Then, a database to be used may be selected according to the illuminance and weather. The weather may be input by the driver.

[0029] In this embodiment, a configuration for diagnosing the sensor on rails has been shown. This makes it possible to diagnose the sensor even in non-electrified sections or sections with few stations. On the other hand, it is also possible to diagnose the sensor with a similar configuration even when the target is something other than rails. For example, it is possible to diagnose the sensor by using the number of poles supporting the overhead wires along the line that can be detected by the sensor or the length of the platform instead of the rail detection distance 103. The length of the platform can also be calculated by the same algorithm as the rail detection distance. This makes it possible to diagnose the sensor using the poles supporting the overhead wires along the line or the platform. EXAMPLES

[0030] Fig. 5 shows an overview of a sensor diagnosis system in a second embodiment of the present invention. In the first embodiment, the sensor was diagnosed using the rail length detected from a camera image, but in this embodiment, a configuration is shown in which the sensor is diagnosed using the rail-to-rail distance. By using the rail-to-rail distance in addition to the rail detection distance in the first embodiment for sensor diagnosis, the accuracy of the sensor diagnosis is improved.

[0031] Compared to the first embodiment, the data of the detected object DB 102, the sensor control unit 204, and the sensor abnormality determination unit 202 are changed.

[0032] An example of the data configuration of the detected object DB 102 in this embodiment is shown in Fig. 6. Compared to Fig. 3, "Rail distance error" has been added. This means the allowable error of the rail distance in the rail detection distance of the train position.

[0033] In this embodiment, the sensor control unit 204 can calculate the position of the rail detection distance, that is, the rail distance at the end of the rail detected by the sensor, in addition to the rail detection distance 103. The rail distance at the rail detection distance can be calculated using a pinhole camera model based on the relationship between the rail detection distance and the focal length of the lens. The sensor abnormality determination unit 202 in this embodiment acquires the rail detection distance 103 and the rail distance from the sensor control unit 204. The sensor abnormality determination unit 202 performs the same determination process as in the first embodiment for the rail detection distance 103 and diagnoses the sensor. If no sensor abnormality is detected in the rail detection distance 103, the sensor can be further diagnosed based on the rail distance. Since the true value of the rail distance is known as the gauge, it is determined whether the rail distance calculated by the sensor control unit 204 is within the error of the rail distance at the train position stored in the detected object DB 102, and if it is equal to or greater than the error stored in the detected object DB 102, it is determined that the sensor is abnormal.

[0034] By diagnosing the sensor using a known rail-to-rail distance, more accurate sensor diagnosis is possible.

[0035] The present invention is not limited to the above-mentioned embodiment, and various modifications are included. For example, the above-mentioned embodiment has been described in detail to explain the present invention in an easy-to-understand manner, and is not necessarily limited to those having all the configurations described. In addition, it is possible to add, delete, or replace a part of the configuration of each embodiment with another configuration. In addition, each of the above-mentioned configurations, functions, processing units, processing means, etc. may be realized in hardware by designing a part or all of them as an integrated circuit, for example. In addition, each of the above-mentioned configurations, functions, etc. may be realized in software by a processor interpreting and executing a program that realizes each function. Information such as a program, table, file, etc. that realizes each function can be placed in a memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0036] The following are some possible embodiments of the present invention, but the present invention is not limited to these.

[0037] (Aspect 1) A sensor diagnostic system for evaluating the performance of a sensor mounted on a train, comprising: a database storing information on existing facilities that can be detected by the sensor according to the position of the train; A train control unit that detects the position of the train; a sensor abnormality determination unit that determines that an abnormality exists in the sensor when a state of the existing equipment detected by the sensor at the position of the train detected by the train control unit does not match information on the existing equipment stored in the database; and A sensor diagnostic system comprising:

[0038] (Aspect 2) 2. A sensor diagnostic system according to claim 1, comprising: A sensor diagnosis system in which multiple databases are created depending on the state of external factors that affect the performance of the sensor, and the sensor abnormality determination unit determines which database to use depending on the state of the external factors around the train.

[0039] (Aspect 3) 3. A sensor diagnostic system according to claim 2, comprising: A sensor diagnosis system in which a plurality of the databases are created according to time, and the sensor abnormality determination unit determines which of the databases to use according to time.

[0040] (Aspect 4) 3. A sensor diagnostic system according to claim 2, comprising: A sensor diagnosis system in which a plurality of the databases are created according to illuminance, and the sensor abnormality determination unit determines which database to use according to the illuminance around the train.

[0041] (Aspect 5) A sensor diagnostic system according to any one of aspects 2 to 4, comprising: A sensor diagnosis system which creates a plurality of the databases according to temperature, and the sensor abnormality determination unit determines which database to use according to the temperature around the train.

[0042] (Aspect 6) A sensor diagnostic system according to any one of aspects 2 to 5, comprising: A sensor diagnosis system which creates a plurality of the databases according to the amount of rainfall, and the sensor abnormality determination unit determines which database to use according to the amount of rainfall around the train.

[0043] (Aspect 7) A sensor diagnostic system according to any one of claims 1 to 6, comprising: The existing equipment is a rail, and the sensor abnormality determination unit determines an abnormality in the sensor using the length of the rail detected by the sensor.

[0044] (Aspect 8) 8. A sensor diagnostic system according to claim 7, comprising: The existing equipment is a rail, and the sensor abnormality determination unit determines an abnormality in the sensor using the rail distance detected by the sensor.

[0045] (Aspect 9) A sensor diagnostic system according to any one of claims 1 to 6, comprising: The existing equipment is a pillar, and the sensor abnormality determination unit determines an abnormality in the sensor using the number of pillars detected by the sensor.

[0046] (Aspect 10) A sensor diagnostic system according to any one of claims 1 to 6, comprising: The existing facility is a platform, and the sensor abnormality determination unit determines an abnormality in the sensor using the length of the platform detected by the sensor.

[0047] (Aspect 11) A sensor diagnostic system according to any one of claims 1 to 10, comprising: The sensor abnormality determination unit interrupts diagnosis of the sensor when the rail shape is curved and there is an oncoming train.

[0048] (Aspect 12) A sensor diagnostic system according to any one of claims 2 to 6, comprising: A sensor diagnosis system in which the sensor abnormality determination unit selects a plurality of the databases, creates a state of the existing equipment to be used for determining an abnormality in the sensor from the states of the existing equipment stored in the plurality of selected databases by interpolating according to the states of external factors around the train, and determines an abnormality in the sensor using the interpolated states of the existing equipment.

[0049] (Aspect 13) A sensor diagnostic method for evaluating performance of a sensor mounted on a train, comprising the steps of: storing information of existing facilities detectable by the sensor in a database according to the position of the train; A train control unit detects a position of the train; a step of a sensor abnormality determination unit determining that an abnormality exists in the sensor when a state of the existing equipment detected by the sensor at the position of the train detected by the train control unit does not match information of the existing equipment stored in the database; A sensor diagnostic method comprising: [Explanation of symbols]

[0050] 101...Train 102…Detection object DB 103…Rail detection distance 201...On-board control device 202...sensor abnormality determination unit 203…Sensor 204...sensor control unit 205…Train control unit

Claims

1. A sensor diagnostic system for evaluating the performance of a sensor mounted on a train, comprising: a database storing information on existing facilities that can be detected by the sensor according to the position of the train; A train control unit that detects the position of the train; a sensor abnormality determination unit that determines that an abnormality exists in the sensor when a state of the existing equipment detected by the sensor at the position of the train detected by the train control unit does not match information on the existing equipment stored in the database; and A sensor diagnostic system comprising:

2. 2. The sensor diagnostic system of claim 1, A sensor diagnosis system in which multiple databases are created depending on the state of external factors that affect the performance of the sensor, and the sensor abnormality determination unit determines which database to use depending on the state of the external factors around the train.

3. 3. The sensor diagnostic system of claim 2, A sensor diagnosis system in which a plurality of the databases are created according to time, and the sensor abnormality determination unit determines which of the databases to use according to time.

4. 3. The sensor diagnostic system of claim 2, A sensor diagnosis system in which a plurality of the databases are created according to illuminance, and the sensor abnormality determination unit determines which database to use according to the illuminance around the train.

5. 3. The sensor diagnostic system of claim 2, A sensor diagnosis system which creates a plurality of the databases according to temperature, and the sensor abnormality determination unit determines which database to use according to the temperature around the train.

6. 3. The sensor diagnostic system of claim 2, A sensor diagnosis system which creates a plurality of the databases according to the amount of rainfall, and the sensor abnormality determination unit determines which database to use according to the amount of rainfall around the train.

7. The sensor diagnostic system according to any one of claims 1 to 6, The existing equipment is a rail, and the sensor abnormality determination unit determines an abnormality in the sensor using the length of the rail detected by the sensor.

8. 8. The sensor diagnostic system of claim 7, The existing equipment is a rail, and the sensor abnormality determination unit determines an abnormality in the sensor using the rail distance detected by the sensor.

9. The sensor diagnostic system according to any one of claims 1 to 6, The existing equipment is a pillar, and the sensor abnormality determination unit determines an abnormality in the sensor using the number of pillars detected by the sensor.

10. The sensor diagnostic system according to any one of claims 1 to 6, The existing facility is a platform, and the sensor abnormality determination unit determines an abnormality in the sensor using the length of the platform detected by the sensor.

11. 2. The sensor diagnostic system of claim 1, The sensor abnormality determination unit interrupts diagnosis of the sensor when the rail shape is curved and there is an oncoming train.

12. 3. The sensor diagnostic system of claim 2, A sensor diagnosis system in which the sensor abnormality determination unit selects a plurality of the databases, creates a state of the existing equipment to be used for determining an abnormality in the sensor from the states of the existing equipment stored in the plurality of selected databases by interpolating according to the states of external factors around the train, and determines an abnormality in the sensor using the interpolated states of the existing equipment.

13. A sensor diagnostic method for evaluating performance of a sensor mounted on a train, comprising the steps of: storing information of existing facilities detectable by the sensor in a database according to the position of the train; A train control unit detects a position of the train; a step of a sensor abnormality determination unit determining that an abnormality exists in the sensor when a state of the existing equipment detected by the sensor at the position of the train detected by the train control unit does not match information of the existing equipment stored in the database; A sensor diagnostic method comprising:

Images