Train operation management system, train operation management device, and train operation management method

Abstract

To automatically adjust a predicted constant used for predicting the operation of a train.SOLUTION: A train operation management system includes an operation predicting function unit which generates a predicted timetable based on a day timetable indicating an operation schedule of a train on the day and a predicted constant for defining a condition when predicting the operation of the train, outputs the predicted timetable as data used for confirming and / or changing the day timetable, and accumulates the predicted timetable, and a predicted constant adjusting function unit for adjusting the predicted constant by comparing the accumulated predicted timetables with an actual timetable showing the results of operation of the train.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a train operation management system, a train operation management device, and a train operation management method.

Background Art

[0002] Conventionally, in order to perform train operation prediction, there is a technique described in Japanese Patent Application Laid-Open No. 2008-222004 (Patent Document 1). This publication states that "a train operation management system includes an operation management device having a prediction simulation function for performing operation prediction based on signal equipment status, operation status, and a first planned schedule, and an operation arrangement proposal function for making an operation arrangement proposal based on the operation prediction result by the prediction simulation function, and having a second planned schedule that is changed based on the input content input to the operation arrangement device, and the operation management device performs operation management based on the second planned schedule. In this system, the input content is directly incorporated into the prediction simulation function and the operation arrangement proposal function through the change of the first planned schedule."

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In operation prediction, various values such as the arrival and departure times of trains, the stopping time, and the time required between stations are the targets of prediction. These values are affected by factors such as day of the week, weather, and temperature. Therefore, prediction constants that define the operation prediction conditions are set for each combination of various factors, and the prediction constants that match the current situation are used. When the operation prediction deviates from the actual situation, the dispatcher or the like manually adjusts the prediction constants. Adjusting prediction constants places a heavy burden on the dispatcher, and setting inappropriate prediction constants can have a significant impact on the operation prediction. Therefore, an object of the present invention is to automatically adjust the prediction constants used for train operation prediction.

Means for Solving the Problem

[0005] To achieve the above object, one of the representative train operation management systems and train operation management devices of the present invention generates a predicted timetable based on the timetable for the day showing the train operation plan for the day and the prediction constants which are constants defining the conditions when predicting the operation of the train, outputs the predicted timetable as data to be used for the confirmation and / or modification of the timetable for the day, and includes an operation prediction function unit for accumulating the predicted timetable, and a prediction constant adjustment function unit for comparing the accumulated predicted timetable with the actual timetable showing the result of the train operation and adjusting the prediction constants. Also, one of the representative train operation management methods of the present invention includes steps in which a train operation management device generates a predicted timetable based on the timetable for the day showing the train operation plan for the day and the prediction constants which are constants defining the conditions when predicting the operation of the train, outputs the predicted timetable as data to be used for the confirmation and / or modification of the timetable for the day and accumulates the predicted timetable, and compares the accumulated predicted timetable with the actual timetable showing the result of the train operation and adjusts the prediction constants.

Advantages of the Invention

[0006] According to the present invention, the prediction constants used for train operation prediction can be automatically adjusted. Problems, configurations, and effects other than those described above will be clarified by the description of the following embodiments.

Brief Description of the Drawings

[0007]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Mode for Carrying Out the Invention

[0008] Hereinafter, embodiments of the present invention will be described with reference to the drawings.

Embodiment

[0009] FIG. 1 is a block diagram showing the configuration of a train operation management system 100 according to an embodiment of the present invention, the flow of processing and data. As shown in FIG. 1, the train operation management system 100 includes a central management device 1, a central control device 2, a command terminal 3, and a monitor 4, and manages the operation of the train 103.

[0010] The central management device 1 is a device that manages various information related to the operation of the train. The central control device 2 is a device that controls the operation of the train. The central management device 1 and the central control device 2 may be integrated devices. The command terminal 3 is a device corresponding to the train operation management device in the claims, and outputs various information to the dispatcher and receives operation inputs from the dispatcher. The monitor 4 is used for the display output to the dispatcher.

[0011] The detailed configuration of the command terminal 3 will be described. The command terminal 3 has an operation data storage unit 50, an operation prediction function unit 60, and a prediction constant adjustment function unit 40. The operation data storage unit 50 is a storage device such as an HDD (Hard Disk Drive). The operation data storage unit 50 stores the daily execution schedule 10, the daily changed schedule 11, the actual schedule 12, and the train tracking information 13. In addition, information managed and generated by the command terminal includes a predicted schedule 30, difference data 101, prediction constant data 102, and the like.

[0012] The operation prediction function unit 60 is a function unit that performs operation prediction, and the prediction constant adjustment function unit 40 is a function unit that adjusts prediction constants. The operation prediction function unit 60 and the prediction constant adjustment function unit 40 are realized, for example, by a CPU (Central Processing Unit) executing a predetermined program.

[0013] The same-day execution schedule 10 and the same-day change schedule 11 are same-day schedules showing the train operation plan for the same day. The same-day execution schedule 10 shows the train operation plan for the same day formulated in advance before the start of operation on the same day. The same-day change schedule 11 shows the train operation plan for the same day changed according to the operation status on the same day. The same-day change schedule 11 includes, for example, the operation plan of special trains. The actual schedule 12 shows how the trains were actually operated. The actual schedule 12 is determined after the operation on the same day ends. The train tracking information 13 is information that collects in real time the current position of each train and the arrival and departure times at each station. The actual schedule 12 is created based on the train tracking information 13, for example, by the central management device 1.

[0014] The command terminal 3 acquires the same-day execution schedule 10, the same-day change schedule 11, and the actual schedule 12 from the central management device 1 and stores them in the operation data storage unit 50, and acquires the train tracking information 13 from the central control device 2 and stores it in the operation data storage unit 50.

[0015] The operation prediction function unit 60 generates a predicted schedule 30 using the data stored in the operation management data storage unit 50, outputs the predicted schedule 30 as data to be used for checking and changing the same-day schedule, and accumulates the predicted schedule 30. The functions of the operation prediction function unit 60 include a prediction calculation condition definition function 20, a prediction condition adjustment function 21, a prediction time calculation function 22, a prediction unreasonableness detection function 23, and a trial prediction function 24.

[0016] The prediction calculation condition definition function 20 defines the prediction calculation conditions used for calculating the prediction time of the prediction time node. As the prediction time nodes, for example, arrival and departure nodes are set for each station on the line. The prediction time of the prediction time node is the prediction result of the arrival and departure times of the train. Also, the time from the prediction time of the arrival node to the prediction time of the departure node at a certain station is the stop time at that station. And the time from the prediction time of the departure node at a certain station to the prediction time of the next node is the required time between stations.

[0017] The prediction calculation condition definition function 20 defines the constraint conditions in the calculation of arrival / departure times, stop times, required times, etc. as prediction calculation conditions. For example, "For trains arriving on the same track, how many seconds or more should there be between the departure time of the previous train and the arrival time of the next train?", "The restrictions on the arrival / departure times of each train required for connection waiting", etc. are included in the prediction calculation conditions.

[0018] Here, arrival / departure times, stop times, required times, etc. are affected by factors such as day of the week, weather, temperature, etc. Therefore, the prediction calculation condition definition function 20 refers to the prediction constant data 102 in which prediction constants that define the operation prediction conditions for each combination of various factors are set, and defines the prediction calculation conditions using the prediction constants that match the current situation.

[0019] The master data of the prediction constant data 102 is held by the central management device 1, and the command terminal 3 holds the prediction constant data 102 with the same content as the central management device 1. When there is a change in the master data of the central management device 1, the change is reflected in the prediction constant data 102 of the command terminal 3.

[0020] The prediction condition adjustment function 21 adjusts and additionally defines the prediction calculation conditions based on the operator's input. For example, when the operator recognizes a situation that affects the operation due to platform congestion or the occurrence of a critically ill patient, it becomes possible to accurately obtain the prediction diagram by adjusting and additionally defining the prediction calculation conditions.

[0021] The prediction time calculation function 22 calculates the prediction time of each prediction time node using the prediction calculation conditions defined by the prediction calculation condition definition function 20 and the prediction condition adjustment function 21 as inter-node constraints. Also, the prediction time calculation function 22 can calculate the prediction time using the train tracking information 13 as well.

[0022] The prediction unreasonableness detection function 23 detects unreasonableness in which the magnitude relationship of the prediction times among a plurality of prediction time nodes circulates. When unreasonableness is detected, that is, when there is a contradiction between the prediction calculation conditions, the prediction calculation conditions are changed to resolve the contradiction.

[0023] The trial prediction function 24 changes the train diagram using prediction calculation conditions without unreasonableness and generates a trial train diagram. The trial train diagram reflects the train operation arrangement content before execution confirmation (for example, change of destination). The trial prediction function 24 calculates the prediction time based on the trial train diagram and generates a prediction train diagram 30. The prediction train diagram 30 predicts the operation of trains after the current time. The trial prediction function 24 displays and outputs the prediction train diagram 30 to the monitor 4 and stores the prediction train diagram 30 in a predetermined storage device. The dispatcher can confirm the influence before the execution confirmation of the train operation arrangement based on the prediction train diagram 30 output to the monitor 4.

[0024] The prediction constant adjustment function unit 40 compares the accumulated prediction train diagram 30 with the actual train diagram 12 indicating the result of the train operation and adjusts the prediction constants. The prediction constant adjustment function unit 40 obtains the difference between the time indicated by the prediction train diagram 30 and the time indicated by the actual train diagram 12 as difference data 101. When the difference data 101 is equal to or greater than the prediction constant correction threshold, the prediction constant adjustment function unit 40 corrects the prediction constant so that the difference becomes less than the prediction constant correction threshold. At this time, when the time indicated by the actual train diagram 12 is equal to or greater than the operation abnormality determination threshold, it is excluded from the correction target. This is because when a large disruption occurs in the train diagram due to an operation abnormality and it is reflected in the prediction constant, an inappropriate change is added to the value of the prediction constant. Further, instead of directly comparing the time values shown in the performance diagram 12 with the operation abnormality determination threshold value, when the difference between the value of a predetermined node indicated by the performance diagram 12 and the value of the previous node linked to the said node is equal to or greater than the operation abnormality determination value, it may be excluded from the objects to be corrected. For example, when there is a significant delay in the arrival time which is the value of the arrival node of a certain station, if there is also a significant delay in the departure time of the departure node of the previous station which is the previous node, the delay is considered to be due to operation abnormality before the previous station. Therefore, the difference of the previous node is evaluated to determine whether adjustment of the prediction constant is necessary. Similarly, even if there is a significant delay in the departure time which is the value of the departure node of a certain station, if the arrival time of the arrival node at the said station which is the previous node is similarly delayed, the departure delay is not an operation abnormality but due to the arrival delay. Therefore, there is no need to exclude it from the objects to be adjusted for the prediction constant.

[0025] The prediction constant adjustment function unit 40 adjusts the prediction constant by changing the master data of the prediction constant data managed by the central management device 1. When the master data of the prediction constant data is changed, the central management device 1 distributes the changed prediction constant data to all command terminals 3 that use the said prediction constant data. The command terminal 3 holds the updated prediction constant data. Further, the command terminal 3 displays and outputs a prediction constant change monitor screen on the monitor 4 to notify the operator of the update of the prediction constant data.

[0026] Figure 2 is an explanatory diagram of the prediction constant data. As shown in Figure 2, the prediction constant data includes the minimum operation minutes, minimum stop minutes, departure - departure continuous interval, platform conflict interval, tolerance interval, etc. The minimum operation minutes sets the minimum stop minutes for items such as the section, start station, end station, train type, passing / stopping at the start station, passing / stopping at the end station, and vehicle type. The minimum stop minutes sets the minimum stop minutes for items such as the station, vehicle type, number of cars, and train type. The departure - departure continuous interval sets the departure - departure continuous interval for items such as the station, departure line, direction, passing / stopping of the preceding train, and passing / stopping of the following train. The platform conflict time interval sets the platform conflict time interval for the items of station, platform, relative direction between trains, passing / stopping of the preceding train, passing / stopping of the following train, number of cars of the preceding train, and number of cars of the following train. The tolerance time interval sets the tolerance time interval for the items of station, preceding train route, following train route, passing / stopping of the preceding train, passing / stopping of the following train, number of cars of the preceding train, and number of cars of the following train.

[0027] The prediction constant data shown in FIG. 2 is further set for each day of the week, weather, and temperature. Therefore, for example, by specifying elements such as day of the week "Monday", predicted weather "rain", and predicted maximum temperature "5 degrees to 10 degrees", appropriate prediction constants can be specified under these conditions.

[0028] FIG. 3 is a flowchart showing an example of the processing procedure of the prediction constant adjustment function unit 40 of the train operation management system in the first embodiment. The operations based on this flowchart are as follows, and the operation subject of each step is the command terminal 3. Also, this process is executed for each node on the train schedule.

[0029] Step 200: The command terminal 3 starts the process of automatically adjusting the prediction constant and proceeds to step 201. Step 201: The command terminal 3 determines whether the magnitude of the difference between the predicted train schedule 30 and the actual train schedule 12 is equal to or greater than the prediction constant correction threshold (for example, 3 seconds), that is, if it exceeds the range where the adjustment of the prediction constant is unnecessary and can be ignored (YES), it proceeds to step 202; if it is less than 3 seconds (NO), it proceeds to step 205.

[0030] Step 202: The command terminal 3 determines whether the difference between the operation actual value of the previous link node and this node is less than the operation anomaly determination value. If the difference is less than the operation anomaly determination value (YES), it proceeds to step 203. If the difference is equal to or greater than the operation anomaly determination value (NO), since the data is not suitable for use in adjusting the prediction constant, it proceeds to step 205. Step 203: The command terminal 3 proceeds to step 204 if the target value is not the value of the special train schedule (YES), that is, if the data is not unsuitable for use in adjusting the prediction constant; it proceeds to step 205 if it is special train schedule data (NO).

[0031] Step 204: The command terminal 3 adjusts the prediction constant so that the predicted value is within the prediction constant correction threshold (for example, within 3 seconds). Then, it proceeds to step S205. Step 205: End the process of automatically adjusting the prediction constant. Note that the threshold of 3 seconds in steps 201 and 204 is just an example.

Example

[0032] In Example 1, the automatic adjustment of the prediction constant based on the difference between the actual schedule 12 and the predicted schedule 30 was explained. However, it may be configured such that the conductor can confirm the history of the execution of the automatic adjustment, that is, when, where, and what differences occurred resulting in the automatic adjustment.

[0033] Figure 4 is a block diagram showing the configuration of a system that enables confirmation of the history of automatic adjustment of the prediction constant, as well as the flow of processing and data. The block diagram shown in Figure 4 adds prediction constant adjustment history data 104 to Figure 1. The prediction constant adjustment history data 104 records the content when the prediction constant adjustment is performed by the prediction constant adjustment function unit 40, and is output to the prediction constant change monitor screen of the monitor 4 so that the conductor can confirm the content.

[0034] Figure 5 is a flowchart showing an example of the processing procedure of the prediction constant adjustment function unit 40 of a train operation management system that enables confirmation of the history of automatic adjustment of the prediction constant. The flowchart shown in Figure 5 adds the recording and output of the prediction constant adjustment history in step 206 and the output of an abnormality warning in step 207 after step 204 of Figure 3. Step 206: Record the prediction constant adjustment content in step 204 as the prediction constant adjustment history data 104 and output it to the monitor 4. Step 207: Output an alarm to the central management device 1 indicating that a difference has occurred between the actual diagram and the predicted diagram leading to the adjustment of the prediction constant in Step 204, record it as prediction constant adjustment history data 104, and output it to the alarm dialog of the monitor 4.

[0035] As described above, the disclosed train operation management system 100 generates a predicted diagram 30 based on the daily diagram showing the train operation plan for the day and the prediction constant which is a constant defining the conditions when predicting the operation of the train, outputs the predicted diagram 30 as data to be used for the confirmation and / or change of the daily diagram, accumulates the predicted diagram 30, and includes an operation prediction function unit 60 for comparing the accumulated predicted diagram 30 with the actual diagram 12 showing the result of the train operation, and a prediction constant adjustment function unit 40 for adjusting the prediction constant. With such a configuration and operation, the prediction constant used for train operation prediction can be automatically adjusted.

[0036] Further, when the difference between the time indicated by the predicted diagram 30 and the time indicated by the actual diagram 12 is equal to or greater than the prediction constant correction threshold, the prediction constant adjustment function unit 40 corrects the prediction constant so that the difference becomes less than the prediction constant correction threshold. Therefore, the accuracy of the predicted diagram 30 can be efficiently improved.

[0037] Also, when the time indicated by the actual diagram 12 is equal to or greater than the operation abnormality determination threshold, it is excluded from the object of the correction. Alternatively, when the difference between the value of a predetermined node indicated by the actual diagram 12 and the value of the previous node linked to the node is equal to or greater than the operation abnormality determination value, it is excluded from the object of the correction. By adopting such an operation, it is possible to prevent a decrease in prediction accuracy caused by reflecting an operation abnormality in the prediction constant.

[0038] Further, a plurality of prediction constants are provided according to at least one of the elements such as day of the week, weather, and temperature. The operation prediction function unit 60 generates the predicted train schedule using the prediction constant that conforms to the situation of the day among the plurality of prediction constants, and the prediction constant adjustment function unit 40 adjusts the prediction constant that conforms to the situation of the day among the plurality of prediction constants. Therefore, it is possible to improve the accuracy of prediction according to requirements such as the day of the week, weather, and temperature.

[0039] Further, the daily train schedule includes a daily execution train schedule formulated in advance and a daily modified train schedule modified according to the operation state of the day. Therefore, it is possible to evaluate the differences from the pre-formulated train schedule and the train schedule modified on the day, and adjust the prediction constants.

[0040] Further, a command terminal 3 that receives operations by a dispatcher includes the operation prediction function unit 60 and the prediction constant adjustment function unit 40. The command terminal 3 acquires the daily train schedule, the actual train schedule 12, and the prediction constants from the central management device 1, and the prediction constant adjustment function unit 40 adjusts the prediction constants managed by the central management device 1. In such a configuration, the adjustment of the prediction constants by the command terminal 3 can be applied to the entire operation system.

[0041] Further, the prediction constant adjustment function unit 40 evaluates the difference between the predicted train schedule 30 and the actual train schedule 12, and outputs an alarm notifying the dispatcher of the necessity to adjust the prediction constants according to the result of the evaluation. In such a configuration, information regarding the adjustment of the prediction constants can be appropriately notified to the dispatcher as necessary.

[0042] Note that the present invention is not limited to the above embodiments, and includes various modifications. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, not only the deletion of such a configuration is possible, but also the replacement and addition of configurations are possible. For example, in the above-described embodiment, a configuration in which an apparatus for performing anomaly detection from data and an apparatus for updating a model are integrated is illustrated. However, a configuration in which anomaly detection and model update are performed by separate apparatuses may also be used.

Explanation of Signs

[0043] 1: Central management apparatus, 2: Central control apparatus, 3: Command terminal, 4: Monitor, 10: Daily execution diagram, 11: Daily change diagram, 12: Performance diagram, 13: Train tracking information, 20: Prediction calculation condition definition function, 21: Prediction condition adjustment function, 22: Prediction time calculation function, 23: Prediction unreasonableness detection function, 24: Trial prediction function, 30: Prediction diagram, 40: Prediction constant adjustment function unit, 50: Operation data storage unit, 60: Operation prediction function unit, 100: Train operation management system, 101: Difference data, 102: Prediction constant data, 103: Train, 104: Prediction constant adjustment history data

Claims

1. A train operation prediction function unit that generates a predicted train schedule based on the current-day train schedule indicating the train operation plan for the day and prediction constants that define the conditions for predicting the train operation, outputs the predicted train schedule as data to be used for confirmation and / or modification of the current-day train schedule, and accumulates the predicted train schedule, A prediction constant adjustment function unit that compares the accumulated predicted train schedule with the actual train schedule indicating the result of the train operation and adjusts the prediction constants, Comprising, The prediction constant adjustment function unit corrects the prediction constants so that the difference between the time indicated by the predicted train schedule and the time indicated by the actual train schedule is less than the prediction constant correction threshold when the difference is equal to or greater than the prediction constant correction threshold. A train operation management system characterized by this.

2. The train operation management system according to claim 1, characterized in that when the time indicated by the actual train schedule is equal to or greater than the operation abnormality determination threshold, it is excluded from the object of the correction.

3. The train operation management system according to claim 1, characterized in that when the difference between the value of a predetermined node indicated by the actual train schedule and the value of the previous node linked to the node is equal to or greater than the operation abnormality determination value, it is excluded from the object of the correction.

4. A plurality of prediction constants are provided according to at least one of the elements of day of the week, weather, and temperature, The train operation prediction function unit generates the predicted train schedule using the prediction constant that matches the current situation among the plurality of prediction constants, The prediction constant adjustment function unit adjusts the prediction constant that matches the current situation among the plurality of prediction constants, The train operation management system according to claim 1, characterized by this.

5. The train operation management system according to claim 1, characterized in that the current-day train schedule includes a previously planned current-day execution train schedule and a current-day modified train schedule modified according to the current operation state.

6. A command terminal that receives an operation by a dispatcher includes the train operation prediction function unit and the prediction constant adjustment function unit, The command terminal acquires the current-day train schedule, the actual train schedule, and the prediction constants from a central management device, The prediction constant adjustment function unit adjusts the prediction constants managed by the central management device, The train operation management system according to claim 1, characterized by this.

7. The train operation management system according to claim 1, wherein the prediction constant adjustment function unit evaluates the difference between the predicted train schedule and the actual train schedule, and outputs an alarm notifying the dispatcher of the necessity of adjusting the prediction constant according to the result of the evaluation.

8. An operation prediction function unit that generates a predicted train schedule based on the daily train schedule indicating the train operation plan for the day and the prediction constant that is a constant defining the conditions when predicting the train operation, outputs the predicted train schedule as data for confirmation and / or modification of the daily train schedule, and accumulates the predicted train schedule. A prediction constant adjustment function unit that compares the accumulated predicted train schedule with the actual train schedule indicating the result of the train operation, and adjusts the prediction constant. A train operation management device, characterized by comprising the above.

9. A train operation management device includes steps of: generating a predicted train schedule based on the daily train schedule indicating the train operation plan for the day and the prediction constant that is a constant defining the conditions when predicting the train operation; outputting the predicted train schedule as data for confirmation and / or modification of the daily train schedule, and accumulating the predicted train schedule; comparing the accumulated predicted train schedule with the actual train schedule indicating the result of the train operation, and adjusting the prediction constant. The method for managing train operation, characterized in that the step of adjusting the prediction constant corrects the prediction constant so that the difference between the time indicated by the predicted train schedule and the time indicated by the actual train schedule is less than the prediction constant correction threshold when the difference is equal to or greater than the prediction constant correction threshold. ​

Images