JP2005041332A - Program and train diagram evaluation support device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide program and train diagram evaluation support device to evaluate whether or not a prepared train diagram is a train diagram difficult to be affected by disturbance, etc. or support the evaluation. <P>SOLUTION: A train diagram network forming part 120 forms a train diagram network by referring to train diagram data 740. A predictive diagram preparing part 140 prepares distribution of a generation time of passengers by a random number generated based on a generation rate of passengers for every OD defined in a passenger generation rate table 780. The predictive diagram preparing part 140 sets required getting on/off time calculated by a getting on/off time calculating part 142 as weight in accordance with the distribution of the generation time of passengers in relation of a stop time arc, and a critical path of the train diagram network is calculated to prepare predictive diagram data. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a program for evaluating a train diagram and a train diagram evaluation support apparatus.

  In railways, transportation services are provided according to the train schedule that defines the train operation plan. However, in the event of an accident or disaster, the train will be delayed due to an increase in passenger arrival and departure times during times of congestion, resulting in disruption of the schedule. There is a case. Conventionally, research has been conducted to provide a stable transportation service that is not easily affected by the disturbance that causes the train delay as described above.

On the other hand, as disclosed in Patent Document 1, a technique for creating a train operation plan that realizes an automatic train operation method that is resistant to disturbances, timetable changes, and the like is known. Specifically, a section in which the traveling time changes for each place is divided into several small sections having a certain traveling condition. For each small section, the train operation method and evaluation value for the small section are unique when the speed and time conditions for entering the small section and the speed and time conditions for exiting the small section are arbitrarily given. The method for determining the value is preset. Then, a combination of conditions that minimizes the evaluation value of the entire traveling section is selected from combinations of conditions such as speed and time at the time of crossing the boundary of each small section. A train operation plan for the entire travel section is determined using the train operation method.
JP-A-6-219282

  By the way, it is important to evaluate whether or not the train schedule created is less likely to be affected by the above-mentioned disturbances, etc., in order to provide stable transportation services. Currently, little research has been done for this purpose. That is, as in the technique disclosed in Patent Document 1, a technique for creating a train operation plan that is considered “strong” against disturbances and the like has been studied, but research on evaluation of “strength” has been made. The current situation is not.

  Therefore, an object of the present invention is to provide a program and a train diagram evaluation support device for evaluating whether or not the created train diagram is a train diagram that is not easily affected by disturbance or the like, or for supporting the evaluation. To do.

The first invention for solving the above problems is:
Computer
Train diagram input means for inputting a train diagram having at least arrival / departure time data of each train station (for example, manual input may be used, or reading from a storage medium may be performed),
A node representing each arrival event and departure event at each station of each train constituting the inputted train diagram, and an arc representing a predetermined temporal dependency between each arrival event and departure event at each station of each train And a network generation means for generating a train diagram network (for example, the train diagram network generation unit 120 shown in FIG. 2),
Setting means (for example, prediction diagram creation unit 140 shown in FIG. 2) for setting a new time dependency by changing the time dependency represented by the arc constituting the generated train diagram network,
Prediction diagram creation means for creating a prediction diagram, which is a virtual train diagram that satisfies the set time dependency, from a train diagram network in which the time dependency is varied (for example, prediction diagram creation shown in FIG. 2) Part 140),
It is a program to make it function as.

Moreover, the train diagram evaluation support device of the seventh invention is
Train schedule input means for inputting a train schedule having at least arrival / departure time data for each station of each train;
A node representing each arrival event and departure event at each station of each train constituting the inputted train diagram, and an arc representing a predetermined temporal dependency between each arrival event and departure event at each station of each train And a network generation means for generating a train diagram network based on
Setting means for setting a new time dependency by changing the time dependency represented by the arc constituting the generated train diagram network;
A prediction diagram creating means for creating a prediction diagram that is a virtual train diagram that satisfies the set time dependency, from a train diagram network in which the time dependency is variable;
It is characterized by having.

  According to the first or seventh aspect of the invention, a train diagram network representing the input train diagram is generated, and the time dependency represented by the arcs constituting the train diagram network is changed to create a new time dependency. By setting the relationship, a virtual prediction diagram that satisfies the newly set temporal dependency relationship can be created. According to this, it is possible to predict the disturbance of the train diagram when the time dependency represented by the arc constituting the train diagram network changes. Therefore, it is possible to support the evaluation of whether or not the input train diagram is a train diagram that is not easily affected by the disturbance that causes the train delay.

The second invention is the program of the first invention,
The setting means determines the variable amount of the time dependency according to a predetermined probability distribution, thereby causing the computer to function to set a plurality of types of time dependency;
The prediction diagram creating means causes the computer to function so as to create a prediction diagram for each of the set plural types of time dependency relationships.

  According to the second aspect of the invention, it is possible to create a prediction diagram for each of a plurality of types of temporal dependency relationships according to the variable amount of the temporal dependency relationship represented by the arc constituting the train diagram network. That is, it is possible to create a prediction diagram corresponding to a temporal dependency changed according to a predetermined probability distribution.

The third invention is the program of the second invention,
The computer is caused to function as distribution calculation means (for example, a prediction diamond creation unit 140 shown in FIG. 2) for calculating distribution of arrival times included in a plurality of prediction diamonds created by the prediction diamond creation means. .

  According to the third aspect of the present invention, it is possible to calculate the distribution of arrival times included in a plurality of prediction diagrams created by changing the temporal dependency relationship represented by the arcs constituting the train diagram network. Therefore, it is possible to calculate the distribution of arrival times at each station of each train that changes according to the temporal dependency changed according to the predetermined probability distribution.

The fourth invention is the program of the second or third invention,
An arrival delay time is calculated by comparing each of a plurality of prediction diamonds created by the prediction diamond creation means with the input arrival time data of the train diamond, and based on the calculated arrival delay time The computer is caused to function as evaluation value calculation means (for example, an evaluation value calculation unit 160 shown in FIG. 2) for calculating a predetermined evaluation value for the input train schedule.

The fifth invention is the program of the fourth invention,
The evaluation value calculation means includes (1) an overall average of the delay time at each station of each train, (2) an overall variance of the delay time at each station of each train, and (3) a delay time for the predetermined station of the predetermined train. At least one of the maximum value, (4) the probability that a delay has occurred at each station of each train, and (5) the probability that the delay time at each station of each train has exceeded a predetermined time is calculated as an evaluation value. The computer is made to function.

  According to the fourth or fifth invention, a difference between each of a plurality of predicted diagrams created by changing the time dependency represented by the arcs constituting the train diagram network and the arrival time data of the train diagram. On the basis of the arrival delay time, a predetermined evaluation value for the inputted train schedule can be calculated.

A sixth invention is a program according to any one of the second to fifth inventions,
The setting means is
An occurrence distribution generating means (for example, a prediction diagram creating unit 140 shown in FIG. 2) for generating a virtual occurrence distribution of passengers;
A boarding / alighting time calculating means (for example, a boarding / alighting time calculating unit 142 shown in FIG. 2) that calculates a boarding / alighting time for passenger boarding / alighting at each station of each train based on the generated passenger distribution,
Among the arcs constituting the train diagram network generated by the network generation means, the time dependency expressed by the arc representing the time dependency from the arrival event to the departure event by the same train is calculated at the time of the getting on / off. Incoming event time relationship setting means (for example, the prediction diagram creation unit 140 shown in FIG. 2) to be newly set based on minutes,
The computer is made to function so as to have

  According to the sixth aspect of the present invention, it is possible to calculate the boarding / alighting time for passenger boarding / exiting at each station of each train based on the virtual generation distribution of passengers. Then, among the arcs constituting the train diagram network, the time dependency from the arrival event to the departure event by the same train can be newly set based on the calculated time of getting on and off, and a prediction diagram can be created. Therefore, it is possible to create a prediction diagram corresponding to the passenger boarding time at each station of each train.

  ADVANTAGE OF THE INVENTION According to this invention, disorder of the train diagram when the time dependence which the arc which comprises a train diagram network represents changes can be estimated. Therefore, it is possible to evaluate whether or not the input train schedule is a train schedule that is not easily affected by disturbances or the like that cause train delay.

  Hereinafter, a program and a train schedule evaluation support apparatus to which the present invention is applied will be described in detail with reference to the drawings.

  The train schedule defines the train operation plan and is created taking into account the standard operation time between stations, the stop time at each station, the expected congestion of trains, or the connection with other trains. The An example of a train diagram is shown in FIG. In FIG. 1A, the operation of the train A and the train B is represented by “train lines” with the horizontal axis representing time and the vertical axis representing stations X, Y, and Z.

  FIG.1 (b) is a figure which shows an example of the train diagram network which expressed the train diagram shown to the figure (a) by the PERT network. The train diagram network shown in FIG. 1B includes a reference time node, an arrival event node that represents an arrival event at each station of each train, and an occurrence event node that represents an occurrence event from each station of each train (hereinafter referred to as an arrival event node). And the event node are appropriately referred to as “event nodes”) and arcs representing the time dependency of each node.

  The arc from the reference time node to each event node (hereinafter referred to as “reference time arc”) is weighted with the time at which the corresponding event occurs. In FIG. 1B, the reference time arc is indicated by a dotted arrow. For example, the departure time from the station Y of the train A is weighted to the reference time arc A10 from the reference time node to the departure event node N10 representing the departure from the station Y of the train A. Further, the arrival time of the train A at the station Z is weighted to the reference time arc A12 from the reference time node to the arrival event node N12 indicating arrival of the train A at the station Z.

  In addition, the arc connecting the event nodes is weighted by the time from the occurrence of the event represented by the event node on the arc start point side to the occurrence of the event represented by the event node on the end point side of the arc.

  Specifically, (1) an arc connecting event nodes indicating arrival and departure events at the same station of the same train (hereinafter referred to as “stop-time arc”), the train stops at the station. The stop time is weighted. This stop time is determined in consideration of the time taken for passengers and crew members to get on and off. In FIG. 1 (b), the stop time arc is indicated by a thick arrow. For example, in FIG. 1 (b), the stop time arc A30 from the arrival event node N14 to the departure event node N16 is weighted with the stop time at which the train B stops at the station Y.

  (2) An arc connecting event nodes representing arrival and departure events at different stations of the same train (hereinafter referred to as “running time arc”) is defined in advance as the running time between the corresponding stations. The standard operation time is weighted. In FIG.1 (b), the operation time arc is shown by the solid line arrow. For example, in FIG. 1B, the reference operation time between the stations Y and Z of the train B is weighted to the operation time arc A50 from the originating event node N16 to the arrival event node N18.

  Further, (3) arcs connecting event nodes representing arrival events and originating events in the same station of different trains (hereinafter referred to as “temporal arcs”) include, for example, time intervals between occurrence times of the corresponding events. (Time interval) is weighted. This time interval is determined in consideration of the time required for a train arriving from the same station to travel safely. In FIG.1 (b), a time-interval arc is shown with the dashed-dotted line arrow. For example, in FIG. 1B, the interval arc A70 from the departure event node N10 to the arrival event node N14 includes the time from the departure time from the station Y of the train A to the arrival time of the train B at the station Y. The time intervals are weighted.

  In the present embodiment, as described above, the train diagram is expressed by the PERT network in order to support the evaluation of whether or not the created train diagram is a train diagram that is not easily affected by disturbances that cause train delays. Train network is used. Specifically, the critical path of the train diagram network is calculated, the occurrence time of each arrival event at each station represented by each arrival event node (that is, the arrival time at each station of each train), and each occurrence event node A prediction diagram is created by determining the time of occurrence of an event occurring at each station of each train represented by (ie, the time of occurrence of each train at each station). Then, the weight set for each arc is changed, the critical path is calculated again, and the process of creating a prediction diagram is repeated a specified number of times (in this embodiment, 1000 times), thereby specifying the type of the specified number of times (hereinafter referred to as “the number of specified times”) , “Designated number of types”)). In addition, based on the obtained specified types of prediction diagrams, the arrival time distribution at each station of each train and the evaluation value based on the arrival delay time of each train at each station are output.

  In the following, for the sake of concise explanation, among the arcs constituting the train diagram network, the weights set for the arcs of the reference time arc, the operation time arc, and the time interval arc are fixed, and the stationary time arc is weighted. A case where the set stop time is variable will be described. Specifically, using a random number, a virtual generation distribution of passengers defining a departure station and a destination station is generated. Then, based on the generated passenger distribution, the required boarding time at each station of each train is calculated, and the calculated boarding time is set as a stop time, and a weight is set for the stop time arc. According to this, the prediction diagram which predicted the train operation according to the fluctuation | variation for the passenger's boarding / alighting time in each station can be created.

  Next, the functional configuration of the train schedule evaluation support apparatus 10 will be described. FIG. 2 is a block diagram illustrating an example of an internal configuration of the train schedule evaluation support apparatus 10. As shown in FIG. 2, the train schedule evaluation support device 10 includes a CPU 100, a ROM 200, an input device 300, a display device 400, a RAM 500, a storage device 600, and a storage medium 700 provided in the storage device 600. This is a device that can be realized by a general-purpose computer or the like.

  The CPU 100 executes processing based on a predetermined program in accordance with an input instruction, performs an instruction to each functional unit, data transfer, and the like, and controls the train diagram evaluation support device 10 in an integrated manner. Specifically, the CPU 100 reads various programs stored in the storage medium 700 included in the ROM 200 or the storage device 600 in accordance with an operation signal or the like input from the input device 300, expands the program in the RAM 500, and processes the program according to the program. Execute. Then, the processing result is held in the RAM 500, and a display signal for displaying the processing result is appropriately output to the display device 400. Further, part or all of the processing results stored in the RAM 500 are stored in the storage medium 700.

  Further, the CPU 100 executes a train schedule evaluation support process according to the train schedule evaluation support program 720. In this CPU 100, a train diagram network generation process is executed according to the train diagram network generation program 722, a function unit for generating a train diagram network is executed as a train diagram network generation unit 120, a prediction diagram creation program 724 is executed, A functional unit that creates a prediction diagram is called an evaluation value calculation unit 160 and a functional unit that executes an evaluation value calculation process according to the prediction diagram creation unit 140 and the evaluation value calculation program 726 and calculates an evaluation value of the prediction diamond. Moreover, in the prediction diagram creation part 140, the function part which performs the boarding / alighting time calculation process according to the boarding / alighting time calculation program 724a, and calculates the required boarding / alighting time in each station of each train is called the boarding / alighting time calculation part 142.

  The train diagram network generation unit 120 refers to the train diagram data 740 to generate a train diagram network. FIG. 3A shows an example of the data structure of train data stored in the train diagram data 740. As shown in FIG. 3 (a), the train data is associated with a train number, and diagram data for each station where the train with the train number arrives and departs is stored. Each diamond data defines the arrival and departure times of the train. FIG. 3B is a diagram illustrating an example of train diagram data 740, in which all train data constituting the train diagram is accumulated.

  FIG. 4 is an example in which the data content (train diagram) of the train diagram data 740 shown in FIG. In FIG. 4, the train data constituting the train diagram data 740 is represented by train stripes, with the horizontal axis representing time and the vertical axis representing stations S1 to S6. As shown in FIG. 4, train schedules are assigned train numbers T10 to T15 that operate from the first station S1 to the final station S6, and train numbers T20 to T25 that operate from the first station S3 to the final station S6. Consists of each assigned train.

  The train diagram network generation unit 120 generates a train diagram network composed of a reference time node, event nodes representing arrival and departure events at each station of each train constituting the train diagram, and an arc connecting the nodes. Generate.

  Then, the train diagram network generation unit 120 sets a weight for each of the arc of the reference time arc, the time interval arc, and the operation time arc for which a fixed weight is set. Specifically, the train diagram network generation unit 120 refers to the train diagram data 740, sets the occurrence time of the event corresponding to the reference time arc as a weight, and sets the event corresponding to the time interval arc. The time interval of occurrence time is set as a weight. Further, the train diagram network generation unit 120 refers to the reference operation time data 760 in which the reference operation time between stations by each train is defined, and uses the reference operation time corresponding to the operation time arc as a weight. Set.

  The prediction diagram creation unit 140 creates a passenger occurrence distribution pattern with reference to the passenger occurrence rate table 780. FIG. 5 shows an example of the passenger occurrence rate table 780. The passenger incidence rate table 780 shown in FIG. 5 defines the passenger incidence for each OD (combination of departure station and destination station; hereinafter referred to as “OD”) per door. That is, for each departure station, the rate of occurrence of passengers (the number of passengers generated per minute) with each stop station as the target station is defined. For example, the incidence of passengers getting on the train at the station S3 and getting off at the station S4 is “5.0 (person / minute)”.

  In addition, the incidence rate of passengers for each OD is determined, for example, by predicting passenger boarding / exiting status at each station. For example, it is assumed that the stations S1, S2, and S3 are the nearest stations in a residential area, the station S4 is a transfer station to another line, and the stations S5 and S6 are the nearest stations in the office area. As for station S1, S2 and S3, there are many passengers and few passengers, station S4 has many passengers and passengers, and stations S5 and S6 have few passengers and many passengers. It can be predicted, and based on this, the incidence of passengers per OD is determined.

  Specifically, the prediction diagram creation unit 140 generates a random number according to the Poisson distribution based on the passenger generation rate for each OD defined in the passenger generation rate table 780 described above, and the passenger is generated according to the generated random number. Create a distribution of virtual occurrence times. FIG. 6 is a diagram illustrating a distribution example of generation times of passengers who get on the train from the station S1 and get off at the station S2. In FIG. 6, the horizontal axis indicates the time of occurrence of passengers (t), and the time of occurrence of each passenger at the departure station (station S1) determined by random numbers is shown. The prediction diagram creation unit 140 creates a passenger generation time distribution for each OD and creates a passenger generation distribution pattern.

  The prediction diagram creation unit 140 creates 1000 passenger generation distribution patterns as described above and stores them in the RAM 500 as passenger generation distribution data 520. FIG. 7 shows an example of passenger generation distribution data 520. As shown in FIG. 7, the passenger generation distribution data 520 includes 1000 passenger generation distribution patterns 522 (522-1 to 522-n (n = 1000)) in which the distribution of generation times of passengers is defined for each OD. ) Is accumulated. Each passenger generation distribution pattern 522 holds passenger generation times at the departure station in a CSV (Comma Separated Value) format in chronological order for each OD.

  Then, the prediction diagram creation unit 140 sequentially selects the created passenger occurrence distribution patterns (hereinafter, the selected passenger occurrence distribution patterns are referred to as “target passenger occurrence distribution patterns” as appropriate), and the train schedule network constitutes a stop time. For the minute arc, a weight is set with the required boarding time calculated by the boarding time calculating unit 142 as the stopping time.

Here, the process of the boarding / alighting time calculation unit 142 will be described. The boarding / alighting time calculation unit 142 calculates the required boarding / alighting time d based on the following expression (1) according to the target passenger occurrence distribution pattern.
d = f (r _off ) c _off n _off + f (r _on ) c _on n _on (1)

In Equation (1), d is the required boarding time per door, n _off is the number of passengers getting off, n _on is the number of passengers, f (r _off ) is the boarding rate just before the train arrives, and f (r _on ) is getting off. occupancy immediately before but the ride begins to exit, alighting time required per If the c _off riding 0%, and riding time required per If the c _on riding 0% To do. Further, the boarding rate f (r _off ) immediately before the arrival of the train and the boarding rate f (r _on ) immediately before the getting- _off and starting the boarding are expressed by the following functional expression (2). In addition, this boarding rate is determined as what a passenger boards the train which arrives earliest after generating in a station.
f (r) = ar ² +1, a> 0 (2)

  The prediction diagram creation unit 140 calculates the critical path of the train diagram network after weighting the required on / off time calculated by the on / off time calculating unit 142 as described above, and calculates each arrival event node. The arrival time of the arrival event represented by, and the occurrence time of the occurrence event represented by each occurrence event node are determined, and the prediction diagram information 540 is updated.

  FIG. 8 is a diagram illustrating an example of the prediction diagram information 540. As shown in FIG. 8, 1000 prediction diagram data 542 (542-1 to 542-n (n = 1000)) are included in the prediction diagram information 540, similar to the train diagram data 740 described with reference to FIG. 3. It is stored in the data format. That is, the prediction diagram creation unit 140 selects a target passenger generation distribution pattern from the passenger generation distribution patterns stored in the passenger generation distribution data 520, and the boarding / alighting time calculation unit 142 calculates based on the selected target passenger generation distribution pattern. After the required boarding / alighting time is weighted to the stop time arc, the prediction diagram data 542 is created by calculating the critical path. Then, the prediction diagram creation unit 140 creates 1000 pieces of prediction diamond data 542-1 to 542-n by repeating the above-described processing.

  Then, the prediction diagram creation unit 140 refers to the prediction diagram information 540, for example, creates a arrival time distribution in which arrival times of each train are plotted for each station, and causes the display device 400 to display the arrival time distribution.

  The evaluation value calculation unit 160 calculates the arrival delay time at each station of each train and stores it in the RAM 500 as arrival delay time information 580. Here, the arrival delay time is the difference between the arrival time at each station of each train set in the prediction diagram data 542 and the arrival time at each station defined in the train diagram data 740. Yes, each of the prediction diagram data 542-1 to 542-n constituting the prediction diagram information 540 is calculated.

  FIG. 9 shows an example of arrival delay time information 580. As shown in FIG. 9, arrival delay time / minute information 580 stores arrival delay time / minute data 582-1 to 582-n corresponding to the prediction diagram data 542-1 to 542-n, respectively. That is, the arrival delay time data 582 stores the arrival delay time at each station of each train constituting each prediction diagram data 542.

  The evaluation value calculation unit 160 calculates and calculates evaluation values corresponding to the following evaluation items (1) to (5) based on the calculated arrival delay time at each station of each train as an evaluation scale of the train schedule. Each evaluation value is stored in the RAM 500 as evaluation value data 560 and is displayed on the display device 400.

(1) Overall average for arrival delay time The evaluation value calculation unit 160 calculates an overall average for arrival delay time at each station of each train. That is, the evaluation value calculation unit 160 calculates a total average of arrival delay times at each station of each train set in the aforementioned arrival delay time data 582-1 to 582-n.

(2) Total variance for arrival delay time The evaluation value calculation unit 160 calculates the total variance for arrival delay time at each station of each train. That is, the evaluation value calculation unit 160 obtains the deviation of arrival delay time at each station of each train set in the arrival delay time data 582-1 to 582-n based on the train diagram data 740. The total variance is calculated by calculating the arithmetic mean of the square of the deviation.

(3) Maximum value for arrival delay time The evaluation value calculation unit 160 calculates the maximum value for arrival delay time for a predetermined station of a predetermined train. That is, the evaluation value calculation unit 160 averages the arrival delay times at the same station (for example, the station S4) of the same train (for example, the train number T10) set in the arrival delay time / minute data 582-1 to 582-n. Each value is calculated and its maximum value is determined.

(4) Probability of occurrence of delay The evaluation value calculation unit 160 calculates the probability that each train is delayed at each station. That is, the evaluation value calculation unit 160 determines whether or not each train set in the arrival delay time / minute data 582-1 to 582-n is delayed at each station, and a delay occurs by obtaining a delay rate. Probability is calculated.

(5) Probability that the delay time is a predetermined time or more The evaluation value calculation unit 160 calculates the probability that the arrival delay time at each station of each train is a predetermined time (for example, 20 seconds). That is, the evaluation value calculation unit 160 calculates the probability that the arrival delay time at each station of each train set in the arrival delay time data 582-1 to 582-n is 20 seconds or more. The probability calculation method is the same as the probability that the delay in (4) has occurred.

  FIG. 10 shows an example of the evaluation value data 560. As shown in FIG. 10, evaluation value data 560 stores evaluation values calculated in association with the evaluation items (1) to (5) described above.

  The ROM 200 stores an IPL (Initial Program Loader) program executed by the CPU 100, various initial setting values related to the IPL program, and the like.

  The input device 300 includes a keyboard having a cursor key, a numeric keypad, various function keys, and a pointing device such as a mouse. The input device 300 outputs an operation signal such as a key pressing signal pressed on the keyboard or a mouse position signal to the CPU 100. To do.

  The display device 400 is configured by a CRT (Cathode Ray Tube), an LCD (Liquid Crystal Display), or the like, and displays display information corresponding to the display signal based on a display signal input from the CPU 100.

  The RAM 500 includes a memory area that temporarily holds various programs executed by the CPU 100 and data related to the execution of these programs. In particular, the RAM 500 stores and holds passenger generation distribution data 520, prediction diagram information 540, and evaluation value data 560 in order to realize the present embodiment.

  The storage device 600 includes a storage medium 700 in which programs, data, and the like are stored in advance, and the storage medium 700 is configured by a magnetic or optical storage medium, a semiconductor memory, or the like. The storage medium 700 is fixedly attached to the storage device 600 or detachably attached, and stores the various programs described above, data processed according to these programs, and the like. In particular, the storage medium 700 stores a train schedule evaluation support program 720, train schedule data 740, reference operation time data 760, and a passenger occurrence rate table 780. The train diagram evaluation support program 720 includes a train diagram network generation program 722, a prediction diagram creation program 724, and an evaluation value calculation program 726, and the prediction diagram creation program 724 includes a boarding / alighting time calculation program 724a.

  Next, the operation of the train schedule evaluation support apparatus 10 will be described. FIG. 11 is a processing flow of the CPU 100 related to the train schedule evaluation support process realized by executing the train schedule evaluation support program 720.

  As shown in FIG. 11, first, the train diagram network generation unit 120 executes a train diagram network generation process according to the train diagram network generation program 722 (step s10). Specifically, the train diagram network generation unit 120 refers to the train diagram data 740, generates a train diagram network, and sets a fixed time weight, a reference time arc, a time interval arc, and an operation time arc. A weight is set for each arc.

  Subsequently, the prediction diagram creation unit 140 executes a prediction diagram creation process according to the prediction diagram creation program 724. That is, first, the prediction diagram creation unit 140 creates a specified number of types of passenger occurrence distribution patterns using random numbers with reference to the passenger occurrence rate table 780 (step s20). At this time, the created several types of passenger generation distribution patterns are held in the passenger generation distribution data 520.

  Next, the prediction diagram creation unit 140 sets the required boarding / alighting time at each station of each train corresponding to the target passenger occurrence distribution pattern as a stop time, and sets a weight for the stop time arc (step s30). More specifically, at this time, the boarding / alighting time calculation unit 142 executes boarding / alighting time calculation processing according to the boarding / alighting time calculation program 724a, and the required boarding / alighting time at each station of each train based on the target passenger occurrence distribution pattern. Is calculated.

  Next, the prediction diagram creation unit 140 calculates the critical path, determines the occurrence time of the arrival event represented by each arrival event node, and the occurrence time of the occurrence event represented by each occurrence event node, and creates prediction diagram data. (Step s40). At this time, the created prediction diagram data is held in the prediction diagram information 540.

  Then, the prediction diagram creation unit 140 repeats the processes of steps s30 and s40 described above a specified number of times (1000 times) to create specified types of prediction diagram data. Then, the prediction diagram creation unit 140 creates the specified number of types of prediction diagram data (step s50: YES), then creates the arrival time distribution for each train station with reference to the prediction diagram information 540, and the display device 400 is displayed (step s60).

  FIG. 12 shows an example of distribution of arrival times of trains to the station S6. In FIG. 12, the arrival time distribution at the station S6 for each train number T13, T14, T15, T23, T24, and T25 is shown with the horizontal axis representing time (minutes) and the vertical axis representing probability (%). According to FIG. 12, in the prediction diagram created by calculating the required boarding / alighting time based on the distribution of the generation time of passengers determined using random numbers, and setting the weighting to the stop time arc, trains T12, T13, It can be seen that there is a delay in the arrival time at the station 6 at T14. At this time, the prediction diagram creation unit 140 may create a departure time distribution for each station of each train and display it on the display device 400. In addition, the prediction diagram creation unit 140 may display a change in the boarding rate of the train calculated by the getting-on / off time calculation unit 142 for each station on the display device 400 in a graph.

  Subsequently, the evaluation value calculation unit 160 executes an evaluation value calculation process according to the evaluation value calculation program 726. That is, first, the evaluation value calculation unit 160 calculates the arrival delay time at each station of each train for the specified number of types of prediction diagram data stored in the prediction diagram information 540 (step s70). At this time, arrival delay times calculated for each of the designated types of prediction diagram data are held in the arrival delay time information 580.

  Next, the evaluation value calculation unit 160, based on the calculated arrival delay time at each station of each train, (1) an overall average for the arrival delay time, (2) an overall variance for the arrival delay time, (3) arrival Evaluation values of the maximum value for the delay time, (4) the probability of occurrence of the delay, and (5) the probability of the delay time being a predetermined time or more are calculated and displayed on the display device 400 (step s80). Exit. At this time, the calculated evaluation values are held in the evaluation value data 560.

  As described above, according to the train diagram network evaluation support apparatus 10 to which the present invention is applied, a train diagram network representing an input train diagram can be generated. Then, for the stop time arc constituting the train diagram network, the necessary boarding time calculated based on the distribution of passenger occurrence times determined using random numbers is set as a weight, and the critical path of the train diagram network is set. Is calculated to create a prediction diagram, and the above-described processing is repeated to create a plurality of types of prediction diagrams. Moreover, the arrival time distribution which plotted the arrival time in each station of each train which comprises the prepared several kinds of prediction diamonds created can be created. Furthermore, the arrival delay time at each station of each train is calculated based on the specified number of types of prediction diagrams created, and the predetermined evaluation value is calculated based on the calculated arrival delay time at each station of each train. be able to. It can support the evaluation of whether or not the train schedule created is a train schedule that is not easily affected by fluctuations in boarding / exiting times that cause train delays.

  Therefore, because it is possible to create a prediction diagram according to the passenger's boarding time fluctuation at each station, whether the created train diagram is a train diagram that is not easily affected by disturbances that cause train delays. Can support the evaluation of whether or not.

  Here, the process of this embodiment is performed with respect to a plurality of train schedules, and the results of comparing and considering the obtained arrival time distributions and the respective evaluation values will be described. Specifically, in the train schedule plan 1 and the train schedule plan 1, the train schedule shown in FIG. 4 is a train schedule plan in which the departure times at all stations of the trains with the train numbers T20 to T25 are all shifted by one minute. 2. In train schedule plan 1, train schedules with all departure times at train stations T20 to T25 shifted one minute later are set as train schedule plan 3, and prediction schedules are created for train schedule plans 1 to 3, respectively. did. At this time, the necessary boarding / alighting time (stopping time) set as a weight for the stopping time arc was calculated using the same passenger generation distribution data.

  First, the tendency of each train schedule plan 1-3 will be described. Since the passengers generated at each station get on the train that arrived earliest after the occurrence, the boarding rate (congestion degree) of the train depends on the interval of the trains arriving at each station. That is, if the arrival interval of the train is wide, the number of passengers waiting for the train at the station increases, so the number of passengers who get on the train that has arrived increases, and if the arrival interval of the train is narrow, the number of passengers who get on the train decreases.

  As described above, since the station S4 is a transfer station to another line, the passenger generation rate is set high. Therefore, in the case of the train schedule plan 3, as compared with the train schedule plan 1 and the train schedule plan 2, there is a possibility that the number of passengers riding on the train of the train numbers T20 to T25 starting at the station S3 at the station S4 may increase. high. In other words, the trains of train numbers T10 to T15 already have passengers from stations S1, S2 and S3, so train schedule plan 1 and train schedule plan 2 are compared with train schedule plan 3. Thus, the boarding rate of the trains with the train numbers T10 to T15 increases. As a result, there is a high possibility that the required boarding / alighting time at each station after the station S3 increases and a train delay occurs. Therefore, in order not to cause a delay, it is important to reduce the congestion degree of the train and not to increase the necessary boarding time at each station of each train. It is expected that this is a train schedule that is less likely to cause delays than schedule plan 2 (not easily affected by disturbances, etc.).

  FIG. 13A shows an example of the arrival time distribution of trains to the station S6 in the train schedule plan 1, FIG. 13B shows an example of the arrival time distribution of trains to the station S6 in the train schedule plan 2, () Is a figure which shows an example of the arrival time distribution of the train to the station S6 in the train schedule plan 3, respectively. 13 (a) to (c), each train of train numbers T13, T14, T15, T23, T24, and T25 of each train schedule plan, with the horizontal axis representing time (minutes) and the vertical axis representing probability (%). The distribution of arrival times at station S6 is shown. According to FIG. 13, in accordance with the above-mentioned prediction, the arrival time of the train schedule plan 3 shown in (c) as compared with the train schedule plan 1 shown in (a) and the train schedule plan 2 shown in (b). It can be seen that this train train has a narrow distribution width and is unlikely to cause delays.

  FIG. 14 (a) shows the change in the train boarding rate at each station in the train schedule plan 1, (b) shows the change in the train boarding rate at each station in the train schedule plan 2, and (c) shows the train schedule plan. FIG. 3 is a diagram showing changes in the boarding ratio of trains at each station in FIG. 14 (a) to 14 (c), the horizontal axis represents the station and the vertical axis represents the occupancy rate (%), and the change in the occupancy rate at each station of each train number T10, T15, T25 of each train schedule plan. Each is shown. According to FIG. 14, in the train schedule plan 3 shown in (c), the boarding rate in each train is relaxed compared to the train schedule plan 1 shown in (a) and the train schedule plan 2 shown in (b). I understand that.

  FIG. 15 is a diagram illustrating an example of evaluation values obtained by executing the processing of the present embodiment on the train schedule plans 1 to 3. In FIG. 15, (1) Total average of arrival delay time (2) Total variance of arrival delay time (3) Maximum value of arrival delay time (4) Probability of occurrence of delay (5) Delay time is predetermined The evaluation value of each evaluation item of probability over time is shown for each train schedule plan 1-3. As shown in FIG. 15, the evaluation value of the train schedule plan 3 was smaller than the evaluation values of the train schedule plan 1 and the train schedule plan 2. From the above experimental results, it is considered that according to the train diagram of the train diagram plan 3, a more stable train operation is possible with less train delay due to an increase in the required boarding time.

  The preferred embodiments of the present invention have been described above, but the present invention is not limited to those described above, and can be changed as appropriate without departing from the spirit of the invention.

  For example, in the above-described embodiment, when calculating the necessary boarding / alighting time, the train boarding rate is determined as the one that gets on the train that arrives the earliest after the passenger arrives at the station. Good. In other words, for example, determining the train occupancy rate by taking into account the passenger's boarding and getting-off behavior toward the destination station, such as switching to a train with a different train type, such as normal, rapid, express, or limited express Good. Specifically, the prediction diagram creation unit 140 creates a distribution of passenger occurrence times at the departure station (station S1) for each train connection route, and stores the distribution in the RAM 500 as passenger occurrence distribution data. The boarding / alighting time calculation unit 142 calculates the necessary boarding / alighting time by referring to the passenger occurrence distribution data and sets it as a weight for the stop-time arc.

  In the above-described embodiment, the case where the stopping time (required boarding / alighting time) set as the stopping time arc among the arcs constituting the train diagram network has been described, but the weights set for other arcs May be variable. For example, assuming the variation of travel time between stations, the weight set for the operation time arc may be varied. Specifically, the prediction diagram data creation unit 140 determines, for example, a running time between stations by each train using a random number, and sets it as a weight to be set for the running time arc.

  Moreover, it is good also as preparing the passenger generation rate table for every predetermined time slot | zone, and creating the distribution of the generation time of the passenger for every OD by referring the passenger occurrence table according to the time slot | zone. According to this, it is possible to create a distribution of passenger generation times in consideration of the degree of congestion in the morning and evening commuting hours, and to calculate the required boarding / alighting time at each station of each train.

(A) is an example of a train diagram, (b) is a diagram showing an example of a train diagram network based on the train diagram shown in (a). The block diagram which shows an example of an internal structure of a train schedule evaluation assistance apparatus. The figure which shows an example of train schedule data. The figure which shows an example which trained the train diagram data shown in FIG. The figure which shows an example of a passenger incidence table. The figure which shows the example of distribution of the generation | occurrence | production time of a passenger. The figure which shows an example of passenger generation distribution data. The figure which shows an example of prediction diamond information. The figure which shows an example of arrival delay time information. The figure which shows an example of evaluation value data. The flowchart which shows an example of a train schedule evaluation assistance process. The figure which shows an example of the arrival time distribution of the train to station S6. The figure which shows an example of the arrival time distribution of the train to the station S6 in the train schedule plans 1-3. The figure which shows an example of the congestion degree of each train in the train schedule plans 1-3. The figure which shows an example of the evaluation value of the train schedule plans 1-3.

Explanation of symbols

10 Train schedule evaluation support device 100 CPU
120 Train Diagram Network Generation Unit 140 Prediction Diagram Creation Unit 142 Boarding / Exit Time Calculation Unit 160 Evaluation Value Calculation Unit 200 ROM
300 input device 400 display device 500 RAM
520 Passenger generation distribution data 540 Prediction diagram information 560 Evaluation value data 580 Arrival delay time / minute information 600 Storage device 700 Storage medium 720 Train diagram evaluation support program 722 Train diagram network generation program 724 Prediction diagram creation program
724a Entry / exit time calculation program 726 Evaluation value calculation program 740 Train diagram data 760 Standard operation time data 780 Passenger occurrence rate table

Claims (7)

Computer
Train diagram input means for inputting a train diagram having at least arrival / departure time data of each train station,
A node representing each arrival event and departure event at each station of each train constituting the inputted train diagram, and an arc representing a predetermined temporal dependency between each arrival event and departure event at each station of each train Based on the above, network generation means for generating a train diagram network,
Setting means for setting a new temporal dependency by changing the temporal dependency represented by the arc constituting the generated train diagram network,
A prediction diagram creating means for creating a prediction diagram that is a virtual train diagram that satisfies the set temporal dependency relationship from a train diagram network in which the temporal dependency relationship is variable;
Program to function as. The program according to claim 1,
The setting means determines the variable amount of the time dependency according to a predetermined probability distribution, thereby causing the computer to function to set a plurality of types of time dependency;
A program for causing the computer to function so that the prediction diagram creation means creates a prediction diagram for each of the set types of temporal dependency relationships. The program according to claim 2,
A program for causing the computer to function as distribution calculation means for calculating distribution of arrival times included in a plurality of prediction diamonds created by the prediction diamond creation means. The program according to claim 2 or 3,
An arrival delay time is calculated by comparing each of a plurality of prediction diamonds created by the prediction diamond creation means with the input arrival time data of the train diamond, and based on the calculated arrival delay time A program for causing the computer to function as evaluation value calculation means for calculating a predetermined evaluation value for an inputted train schedule. The program according to claim 4,
The evaluation value calculation means includes (1) an overall average of the delay time at each station of each train, (2) an overall variance of the delay time at each station of each train, and (3) a delay time for the predetermined station of the predetermined train. At least one of the maximum value, (4) the probability that a delay has occurred at each station of each train, and (5) the probability that the delay time at each station of each train has exceeded a predetermined time is calculated as an evaluation value. A program for causing the computer to function. A program according to any one of claims 2 to 5,
The setting means is
An occurrence distribution generating means for generating a virtual occurrence distribution of passengers;
A boarding / alighting time calculating means for calculating a boarding / alighting time for passengers getting on and off at each station of each train based on the generated distribution of passengers,
Among the arcs constituting the train diagram network generated by the network generation means, the time dependency expressed by the arc representing the time dependency from the arrival event to the departure event by the same train is calculated at the time of the getting on / off. Incoming event time relationship setting means to newly set based on minutes,
A program for causing the computer to function so as to have Train schedule input means for inputting a train schedule having at least arrival / departure time data for each station of each train;
A node representing each arrival event and departure event at each station of each train constituting the inputted train diagram, and an arc representing a predetermined temporal dependency between each arrival event and departure event at each station of each train And a network generation means for generating a train diagram network based on
Setting means for setting a new time dependency by changing the time dependency represented by the arc constituting the generated train diagram network;
A prediction diagram creating means for creating a prediction diagram that is a virtual train diagram that satisfies the set time dependency, from a train diagram network in which the time dependency is variable;
A train diagram evaluation support device comprising:

Images