GB2497706A

GB2497706A - Operation organization assistance system

Info

Publication number: GB2497706A
Application number: GB1306420.9A
Authority: GB
Inventors: Shuichiro Sakikawa; Mitsunori Okada; Saori Kurematsu
Original assignee: Hitachi Ltd
Current assignee: Hitachi Ltd
Priority date: 2010-10-13
Filing date: 2011-09-12
Publication date: 2013-06-19
Also published as: WO2012049929A1; JP2012081887A; GB201306420D0; JP5504127B2

Abstract

An operation organization task for recovering an operation plan when a train operation is disrupted due to bad weather, a car trouble, or the like is a very complicated planning task which requires multipurpose, big-picture determinations. To reduce a burden on a person who gives a command and manages the task and assist an upgrade of a technique, on the basis of a train operation track record obtained, a train operation for several hours in the future is simulated. In an operation organization assistance system which has a schedule data table which stores an operation plan, a data update unit which updates the operation plan, a track record schedule update unit which updates track record time with respect to the operation plan, a prediction calculation unit which performs a train operation simulation on the basis of the operation plan and the track record time, and a station device which obtains a traveling track record of a train, a correction data generation unit which generates correlation data between different line section entering time and exiting time of a train which extends over line sections at a line section boundary station, a prediction time correction unit which corrects a prediction calculation result of the line section boundary station, and a correction data table which stores data generated by the correction data generation unit are provided.

Description

DESCRIPTION

(Title of Invention]

OPERATION CONTROL ASSISTANCE SYSTEM

(Tec1itical Field]

(0001] The present invention relates to an operation control assistance system for performing simulations particularly of train operations in operation control work.

(Background Art]

[0002] Operation control work involving restoring an operation plait when train operations are disrupted due to bad weather, car trouble or the like is a very complicated scheduling task that requires not only assessing physical conditions such as is train speeds arid siding facilities at stations but also making multi-purpose, big-picture decisions taking into account an entire transport system typically covering turnover rates of the rolling stock and train crew as well as passenger services.

(0003] In the past, techniques were developed for simulating train operations several hours into the future based on the train operation track records obtained up to the present moment so that the burdens on operators engaged in operation control work maybe alleviated and their skills upgraded. Patent Literature 1 discloses a technique which, where trains are operated in a plurality of line sections under management, transmits train trouble information or other information about the current line section to the other line sections so that train operation plan information may be changed. And Patent Literature 2 discloses S a train operation management system covering a plurality of line sections.

(Citation List] [Patent Literature] (0004] (PTL1] Japanese Unexamined Patent Publication No. 2001-287651 (PTL 2] Japanese Unexamined Patent Publication No. 2002-29423

(Summary of Invention]

is (Technical Problem] (0005] However, although the above-cited Patent Literature 1 describes a structure f or merging the train trouble information over a plurality of line sections, there is no mention of how to calculate a future timetable of trains operated over a plurality of line sections. Also, although Patent Literature 2 discloses a basic structure of train operation management over a plurality of line sections, there is no mention of how to calculate a specific train operation timetable for a given line section.

[0006] That is, in the past, there was no attempt at exchanging operation prediction information between different line sections and predictiig train operations based on that information, with the result that no accurate prediction calculations were performed for the current line section. This has made it impossible toobtaina sufficient level of prediction accuracy and has required an extra time allowance for absorbing prediction errors.

[0007] The present invention provides a new technique which, in anadvanced railway line -sharing situation as inurban railway services in recent years, performs simulations of train operations several hours into the future at a level of accuracy necessary for operation control work in order to make more accurate train operation predictions than before.

[Solution to Problem] [0003] According to the present invention, there is provided an operation control assistance system including a database which has a schedule data table which stores an operat ion plan; a central processing unit havinga data update unit which updates the operation plan, a track record schedule update unit which updates cracK record times of trio operation pan, ana a prediction calculationunit which perf onus simulations of train operations based on the operation plan and the track record times; anda station device which obtains train traveling track records from different stations, the operation control assistance system performing trai.n operation predictions over S a plurality of line sections. The central processing unit includes a correction data generation unit which generates correlation data between a different line section entering time and a different line section exiting time of trains across a line section boundary station between line sections and a predicted time correction unit which corrects results of prediction calculations for the line section boundary station, and the database includes a correction data table which stores the correlation data generated by the correction data generation unit, whereby operation predictions regarding a plurality of line sections are complemented with one another.

[0009] Also in the ooeration control assistance system of this invention, the correction data generation unit may generate, regarding trains which come under control of a line section adjacent to a prediction calculation target line section and whici., from the adjacent line section, enter the prediction calculation target line section for a run therethrough before exiting to the adjacent line section, a train ID upon entry andatrainiDuponexit, apredictedenteringtimeof theentering train and a predicted exiting time of the exiting train obtained through prediction calculations over the prediction calculationtargetlinesection, andatimeallowance calculated as the difference between "the time required between entry and exit for a run with an operationally tolerable minimum running S time and stoppage time° and "the time required between entry and exit obtained through prediction calculations," as predicted time correction data for the line section boundary station.

(ooio] Also in the operation control assistance system of this invention, when the exiting train is included in those registered records in the correction data table which are targeted for prediction calculations, the prediction time correction unit may correct the predicted time of the exiting train at the boundary station by adding to the predicted time of the exiting train at the boundary station the value obtained by subtracting the time allowance registered in the record of the exiting train from a delay time calculated as the difference between the predicted entering time registered in the record of the exiting train in the correction data table on the one hand and the predicted entering time obtained through prediction calculations on the other hand.

(0011] Also in the operation control assistance system of this invention, if the predicted exiting time registered in the record of the exiting train is larger than the value of the corrected predicted time, then the corrected value may be regarded as the predicted exiting time.

[0012] Also in the operation control assistance systeni of this invention, four processes made up of aprocessof the prediction calculation unit, a process of the correction data unit, a process of the predicted time correction unit, and a reference update access to the correction data table may each be performed independently as parallel processing per line section.

[0013] Further according to the present invention, there is provided a control method for an operation control assistance system having a central processing unit including a data update i unit which updates an operation plan, a track record schedule update unit which updates track record times of the operation plan, a prediction calculation unit which performs simulations of train operations based on the operation plan and the track record times, a correction data qeneration unit which cxenerates correlation data between a different line section entering time and a different line section exiting time of trains across a line section boundary station between line sections, and a predicted time correction unit which corrects results of prediction calculations for the line section boundary station; a database having a schedule data table which stores the operation plan and a correction data table which stores the correlation data generated by the correction data generation unit; arid a station device which obtains train, traveling track records from diiferent stations. When the exiting train is s included in those registered records in the correction data table which are targeted for prediction calculations, the predicted time of the exiting train at tb.e boundary station is corrected b adding to tnc predictec time of the ex:tng train at the boundary station the value onLanea by subtractlng a time allowance registered in the record of the exiting train from a delay time calculated as the difference between the predicted entering time registered in the record of the exiting train in the correction data table on the one hand and the predicted entering time obtained through prediction is calculations on the other hand.

[003.4] Further with the control method for an operation control assistance system of this invention, if the predicted exiting time registered in the record of the exiting train is larger than the value of the corrected predicted time, then the corrected value may be regarded as the predicted exiting time.

1.0 015] Further with the control system for an operation control assistance system of this invention, four processes made up of a process of the prediction calculation unit, a process of

S

the correction data unit, a process of the predicted time correction unit, anda reference update access to the correction data table may each he performed independently as parallel processing per line section.

s (Advantageous Effects of Invention] [0016] The present invention provides an operation control assistance system±ncludingadatabasewhichstores an operation plan, a central processing unit having a data update unit and a prediction calculation unit which performs simulations of the operation plan and train operations, and a station device which obtains train traveling track records, the system thereby performing train operation predictions over a plurality of line sections. Further included are a correction data generation unit which generates correlation data between a different line section entering time and a different line section exiting time of trains across a line section boundary station between line sections, a predicted time correction unit which corrects results of prediction calculations for the line sectionboundary station, andacorrecticndata Lablewhichstores the correlation data generated by the correction data generation unit, so that operation predictions regarding a plurality of line sections are complemented with one another. Where trains are operated over a plurality of line sections, the operation predictions are calculated accurately and the results of the prediction calculations on these line sections are complemented with one another in order to ensure operational consistency at the line section boundary station.

(Brief Description of Drawings]

s (0017] (Fig. 1] Fig. 1 is a block diagram showing an operation control assistance system of the present invention.

(Fig. 2] Fig. 2 is a flowchart showing a process flow of the operation control assistance system.

(Fig. 3] Fig. 3 is a schedule diagram showing an operation plan stored in a schedule table of this invention.

(Fig. 4] Fig. 4 is a schedule diagram showing data generated by the correction data generation unit of this invention.

(Fig. 5] rig. 5 is an explanatory view showing a table structure of the correction data table.

[Fig. 6] Fig. 6 is a schedule diagram showing results of predicted times corrected by the predicted time correction unit of this invention.

[Fig. 7] Fig. 7 is another schedule diagram showing results of predicted times corrected by the predicted time correction unit of this invention.

(Description of Embodiments]

s (00181 An operation control assistance system as an embodiment of the present invention is explained below in reference to Figs. 1 through 7. The embodiment described hereunder involves applying the operation control assistance system to a railway operation system.

(Operation control assistance system] First, the configuration of the operation control assistance system embodying the present invention is explained.

(0019] is The operation control assistance system of this embodiment is configured as a wide-area distributed computing system including a central processing unit 1100 that executes a simulation program for operation prediction, a database 1.200 that stores diverse operation data, an input device 1300 that accepts input tromauser, andadisplaydevice 1400 that displays various GtJIs including schedule diagrams. Also included are a network device 1500 that functions as a communication path between the central processing unit 1100 and the other devices, a station device 2100 that acquires track record times constituting the actual operation times of trains, and a ii wide-area network device 2200 that functions as a communication path between the station devices 2100 within a line section.

Incidentally, the station device 2100 is installed per station and the wide-area network device 2200 is provided per line section.

[0020] The database 1200 includes a schedule data table 1210 that stores schedule data about the operation plan and track record times, and a correction data table 1220 that stores correction data.

[0021] The central processing unit 1100 includes a schedule data update unit 1110 that performs operation plan updates such as the most recent schedule changes due to car trouble or the like in the scheduled data table 1210 of the database 1200, a track record schedule update unit 1120 that updates the track record times in the schedule data table 1210, and a prediction calculation unit 1130 that performs prediction calculations per line section several hours into the future on the basis of the operation plan and track record times, [0022] The central processing unit 1100 further includes a correction data generation unit 1140 that generates linkage data between a different line section entering time and a different line section exiting time of trains operated over a plurality of line sections, the linkage data being used to correcttheresultsofpredictioncalculations foralinesection boundary station between line sections, and a predicted time correction unit 1150 that corrects the results of prediction calculations for line section boundary stations. The above-mentioned components of the central processing unit 1100 are configured and executed as a simulation program.

(0023] The input device 1300 is a device used for input such as a mouse and a keyboard, and the display device 1400 is a device used for display such as a display unit. The station device 2100 is installed at each station, acquires the track record times of trains arriving at and departing from the station in question, and transmits the acquired track record times to is the central processingunit llooviathe wide-areanetwork device 2200 and network device 1500. It should be noted that this embodiment is explained here as an operation control assistance system addressed to two line sections: A line section and B line section.

(schedule prediction process spanning multiple line sections] The steps making up a schedule prediction process are explained below in reference to Figs. 1 and 2. Fig. 2 is a flowchart showing a flowof atrainoperationscheduleprediction process performed by the operation control assistance system of this embodiment. The operation control assistance system of this embodiment performs the schedule prediction process independently for each line section and parallelly for the line sections involved, and the access to the correction data table 1220 is shared between the prediction processes. With this s embodiment, the schedule prediction process is carried out in parallel on the A and B line sections. Steps 82110 through 82150 constitute the schedule prediction process for the A line section, and steps $2210 through 82250 make up the schedule prediction process for the B line section.

(00241 The schedule prediction process for the A line section is started in step 82110. A starting trigger may be an event such as a schedule change or the timing of a predetermined constant cycle. Instep 82120, the prediction calculation unit 1130 performs prediction calculations on the A line section.

The prediction calculations involve calculating what effects a train delay obtained as a track record time will have on the schedule several hours into the future on the basis of the operation plan and track record times and by taking into account the train speeds, stoppage times, and inter-train distances defined by the operation plan and basic data.

(Operation plan) Fig. 3 is a schedule diagram showing an operation plan in a two-dimensional coordinate system with its vertical axis denoting line sections and its horizontal axis representing time T. Fig. 3 shows a schedule diagram regarding A and B line sections adjacent to each other with an X station located on the boundary therebetween. Line segments 3110, 3120 and 3130 denote respectively a OOla train, a 002a train and a 003a train S operating in the A line section, and line segments 3210, 3220 and 3230 represent respectively a OOlb train, a 002b train and a 003b train operating in the B line section.

(0025] The OOla train 3110 and the OOlb train 3210 are in a relation to each other spanning the different line sections. Although the rolling stock is the same, it operates as two trains across the X station as the boundary, each train having an operation plan regarding each line section. The 002a train 3120 and the 002b train 3220 are likewise in a relation to each other spanning is the different line sections, and so are the 003a train 3130 and the 003b train 3230.

(0026] A vertical line 3310 is a current-time line indicative of the present time. Any train foundon the left of the current-time line 3310 is shown to be either a train of which the track record has already been obtained or a train of which the track record has yet to be acquired because of a delay. On the assumption that no delay has occurred, the OOlb train 3210 straddling the current-time line 3310 is shown to have its track record obtained in the line section on the left of an intersection point between the current-time line 3310 and the line segment indicative of the train and to have its track record yet to be acquired in the line section on the right of the intersection point.

(0027J with this embodiment, it is assumed that the OOlb train 3210 has a delay of 5 minutes incurred at a way station and that the other trains have yet to run, i.e., have their track records yet to be obtained, With this assumption taken into account, the line segment indicative of the OOlb train 3210 turns into a line segment 3320 that is inclined more in the horizontal direction than the line segment of the operation plan. The other five trains have the same line segments as those of the operation plan at the present time.

[Schedule prediction process on the A line section] In Fig. 3, the OOla train 3110, 002a train 3120, and 003a train 3130 in the A line section have their track records yet to be obtained, i.e., have yet to incur delays. Thus the predicted times of the trains belonging to the A line section, to be calculated by the prediction calculation in step 82120 of Fig. 2, have the same values as those of the operation plan.

(0028] In the predicted time correction of step S213 0, the predicted time correction unit 1150 in Fig. 1 corrects the result of the prediction calculation performed in step 82120 by reference to the data stored in the correctiondata table 1220. The ensuing explanation assumes that no data is stored yet in the correction data table 1220 and that the process of step S2130 is not carried out.

(0029] In the correction data registration of step 32140, the correction data generation unit 1140 in Fig. 1 generates linkage data between a predicted entering time of a train entering the h line section from the adjacent B line section on the one hand, and a predicted exiting time of a train exiting from the A line section to the B line section on the other hand, and stores the generated data in the correction data table 1220.

(0030] The correctiondata has data items including the trainnumbers of entering and exiting trains, the predicted entering time is of the trainenteringthekline sectionandthepredictedexiting time of the train exiting to the B line section as calculated on the line section targeted for prediction calculation, and the time allowance obtained as the difference between "the time required between entry and exit for a run with an operationally tolerable minimum running time and stoppage time" and "the time required between entry and exit obtained through prediction calculations." (0031] Fig. 4 is a schedule diagram showing correction data generated in step $2140 of Fig. 2. With the A line section taken as the base point, a OOlb train 4210 and a 003b train 4230 are shown to be entering trains, a 002b train 4220 to be an exiting train, and a OOlb train 4210 and a 00Th train 4220 to be a pair running as the same rolling stock in the A line section and targeted S for preparation of correction data.

(0032] The predicted entering time of the OOlb train 4210 is set to be the same as the predicted time of a OOla train 4110 at anX station as calculated in the prediction calculation targeted for the A line section. in this example, the results of the prediction calculations targeted for the A line section are the same as those of the operation plan, so that the predicted entering time of the OOlb train 4210 is set to 9:00. The predicted exiting time of the 002b train 4220 is set to be the is same as the predicted time fora 002a train 4120 at the Xstation; the time is set to 10:00 in this example.

[0033] The time allowance between the entry of the OOlb train 4210 and the exit of the 002b train 4220 is obtained as the difference between "the time required between entry and exit for a run with an operationally tolerable minimum running time and stoppage time" and "the time required between entry and exit obtained through prediction calculations." With this embodiment, it is assumed that a time allowance of 2 minutes is gained when the stoppage time is minimized. A line segment 4410 indicates the correction data for the 002a train 4120 in effect when the stoppage time is minimized. The train 4210 has incurred a delay of 5 minutes at a way station and therefore takes on a line segment 4320.

S (correction data table] Fig. 5 shows a data structure of the correction data table 1220 indicated in Fig. 1. The table has the items including a line section 5110 indicating the line sections that entering and exiting trains enter and exit, an entering train 5120 indicating the train numbers of entering trains, a predicted entering time 5130 indicating the predicted entering times of entering trains, an exiting train 5140 indicating the train numbers of exiting trains, a predicted exiting time 5150 indicating the predicted exiting times of exiting trains, and a time allowance 5160 indicating time allowances between the entry of entering trains and the exit of exiting trains.

(0034] A record 5210 indicates the correction data about the OOlb train 4210 and 002b train 4220 shown in Fig. 4, indicating that the OOlb train 4210 and 002 train 4220 are of the same rolling stock, that the predicted entering time is 9:00, that the predicted exiting time is 10:00, and that the time allowance is 2 minutes.

(Schedule prediction process on the B line section] Upon completion of the schedule prediction process targeted for the A line section in step $2150 of Fig. 2, the schedule prediction process targeted for the B line section is carried out successively. Although the schedule prediction process on the A line section and that on the B line section can be S performed in parallel, the explanation of this embodiment assumes that the process on the B line section starts the moment the process on the A line section is finished.

[0035] Instep $2210, the schedule prediction process on the B line section is started. In step $2220, the prediction calculation targeted for the B line section is carried out.

(0036] Fig. 6 is a schedule diagram showing results of prediction calculations targeted for the B line section. A OOlb train 6210 has a track record of a five-minute delay incurred at a way station, so that a line segment 6330 indicates the result of the prediction calculation on the OOlb train 6210, with the predicted time at the X station calculated to be 9:05. A 002b train 6220 and a 003b train 6230 have their track records yet to be obtained, so that their predicted times are the same as those of the operation plan, with the line segments 6220 and 6230 remaining indicative of the prediction results.

(0037] In step $2230 of Fig. 2, the predicted time correction unit 1150 corrects the prediction result calculated in step $2220 by reference to the data stored in the correction data table 1220. What is targeted here f or correction among the data in the correction data table 1220 shown in Fig. 5 is the predicted exiting time of the exiting train 5140 at the boundary station.

(0038] Since the data of the record 5210 in Fig. 5 is stored in step $2140, what is targeted for correction in Fig. 6 by the predicted time correction unit 1150 is the predicted time, at the X station, of the 002b train running in the B line section.

The corrected time is obtained by adding to the predicted exiting time the value acquired by subtracting the time allowance from a delay time of the entering train. The delay time of the entering train is obtained as the difference between the time indicated in the predicted entering time 5130 in the correction is data table 1220 on the one hand, and the predicted time at the x station of the entering train acquired through schedule prediction on the other hand.

(0039] With this embodiment, the original line segment is corrected into line segments 6320 and 6330 due to the delayat away station.

Since the predicted time of the OOlb train 6210 as the entering train is 9:05 and the predicted entering time of the OQib train 6210 in the correction data table 1220 is 9;00, the delay time of the entering train is calculated to be 5 minutes. And since the time allowance 5160 in the correction data table 1220 is 2 minutes, the corrected predicted time at the X station of the 002b train 6220 is 10:03 obtained by adding to the predicted time 10:00 of the 002b train 6220 a three-minute time that is the difference between the delay time of the entering train S and the time allowance.

(0040] Also, a comparison is made between the corrected time and the predicted exiting time 5150 in the correction data table 1220. If the predicted exiting time 5150 is found larger than the corrected time, the value of the predicted exiting time 5150 is adopted as the newly corrected time. This is to simulate a maximum delay time, i.e., the largest possible time of delay as the delay time of the line section. Following the above processing, the corrected 002b train 6220 takes ona line segment 6340.

(0041] After the predicted exiting time is corrected in step $2230 of Fig. 2, control is returned to step $2220 because it is necessary to perform the prediction calculation based on the corrected value. With this embodiment, the 002b train 6220 in Fig. 6 is corrected in step S2230, so that control is returned to step $2220 to repeat the prediction calculation. In this prediction calculation, the 003b train 6230 as the returning 002b train 6220 in Fig. 6 takes on a line segment 6350 after being calculated to be delayed by 3 minutes due to the delay of the 002b train 6220.

(0042] Thereafter, the predicted time is again corrected in step $2230. If there is no train targeted for correction, control S is passed to step 82240 for registration of the corrected data.

With this embodiment, control is passed to step $2240 because there is no train targeted for correction in the predicted time correction to be performed for a second time in step 82230.

(0043] to In step 82240, data is stored into the correction data table 1220 on the basis of the result of the prediction calculation targeted f or the B line section. With this embodiment, there are two data records to be registered: a record 5220 in Fig. in which a OOla train 6110 in Fig. 6 is the exiting train, and a record 5230 in which a 002a train 6120 is the entering train and a 003a train 6130 is the exiting train.

(0044] In the data items of the record 5220 in the correction data table 1220, the entering train 5120 and the predicted entering time 5130 are left blank for the OOla train 6110 since there is no entering train in the schedule as the train to be returned.

The predicted time 9:00 of the 001b train 6210 at the X station is registered as the predicted exiting time 5150, and zero is registered as the time allowance.

(00451 In the record 5230 inwhich the 002a train 6120 is the entering trainandthe OO3atrainGI3O istheexitingtrain, thepredicted time of the 00Th train 6220 at the X station is registered as the predicted entering time 5130, and as the predicted time of the 003b train 6230 at the X station is registered as the predicted exiting ttrne 5150. Also with this embodiment, the time allowance is assumed to be zero for data registration.

Then in step 92250, a termination process is performed to terminate the prediction process targeted for the B line section.

[correction of the schedule prediction process on the A line section] At this stage, the deiayhasyet tobe ref lectedinthe schedule of the A line section, with the operation plan remaining unchanged. Thus upon completion of the prediction process on theBlinesectioninstepS22sO, thepredictionprocesstargeted for the A line section is again carried out on the basis of the above result. That is, the prediction process targeted for theAline section is started in step 52110, and thepredicted time calculation is performed in step 52120.

[00461 The processes in steps 92120, 92130, and 52140 are explained kelowinreference to Fig.?. Asintheabove-descrihedschedule prediction process targeted i-or the A line section, it is also assumed in step 52120 that a OOla train 7110, a 002a train 7120, and a 003a train 7130 running in the A line section have their track records yet to be acquired and that the results of the prediction calculations are the same as those of the operation plan.

(0047] In the predicted time correction process of step $2130, the predicted times of the OOla train 7110 and 003a train 7130 at the X station are corrected on the basis of the records 5220 and 5230 indicated in Fig. 5. Since the entering train 5120 is not set in the record 5220 shown in Fig. 5, the corrected time of the OOla train 7110 is set to be the larger of the two time values compared: the predicted time of the OOla train 7110 and the predicted exiting time 5150 thereof. With this embodiment, the predicted exiting time 5150 turns out to be the larger of the two values, so that the corrected predicted time of the OOta train 7110 at the X station is 9:05.

(00481 The predicted time at the X station of the 003a train 7130 is corrected through comparison of the predicted exiting time 5150 thereof in the record 5230 with the predicted time of the QOla train 7110, given that the delay of the 002a train 7120 as the entering train is 0 and that the time allowance of the record 5160 is also 0. In this example, the value of the predicted exiting time 5150 is the larger of the two values compared, so that the corrected predicted time of the 003a train 7130 at the X station is 11:03.

(00491 With the correction process of step 52130 completed as explained above, the prediction calculation process of step 52120 is again carried out to have the result of the correction s reflected in the result of the prediction for theA line section.

As a result of this, the 002a train 7120 in Pig. 7 is delayed and takes on a line segment 7350 as shown due to the delay of the OOla train 7110 to be returned. The calculations on the OOla train 7110 arid 003a train 7130 targeted for correction result in line segments 7330 and 7360, respectively.

[00501 In the manner described above, the operation plan spanning the A and B line sections is corrected through exchanges or mutual complementing of data therebetween, whichmakes accurate is operation schedule predictions possible.

(Reference Signs List] [0051] 1100 Central processing unit 1200 Database 1300 Input device 2400 Display device 1500 Network 2100 station device 2200 wide-area network 1110 schedule data update unit 1120 Track record schedule update unit 1130 prediction calculation unit 1140 Correction data generation unit 1150 Predicted time correction unit 1210 schedule data table 1220 Correction data table