JP5680433B2 - Program and system for determining number of trains - Google Patents

Description

The present invention relates to a train number determination system for determining the number of trains of a target train.

  In high-quality trains such as Shinkansen trains and limited express trains currently operating in Japan, multiple seat types (hereinafter referred to as “seat types”) such as reserved seats and non-reserved seats are set in the same train. Conventionally, the setting of the number of vehicles in one train (hereinafter referred to as “the number of trains”) and the seat type has been fixedly determined for each train system, based on past transportation trend results. For example, in a certain system, on the basis of driving with a 10-car train, in all the sections to be driven, 6 out of 10 cars are reserved seat vehicles and the remaining 4 cars are fixed seat vehicles. Was decided.

  Reserved seats are reserved on a first-come-first-served basis, and if the reserved seat is full and cannot be reserved, the passenger will purchase a ticket for a free seat. For railway operators, it is preferable that the number of reserved seats is full and the number of free seats is full, but there is a problem of reduced riding efficiency (road factor) due to an imbalance of congestion. For example, there are many phenomena in which a vacant seat is conspicuous in a designated seat even though the free seat is full, or conversely, a vacant seat is conspicuous in a free seat even though the designated seat is full. In view of this problem, Patent Document 1 discloses a mechanism for promoting the use of reserved seats and reducing vacant reserved seats.

JP 2008-40816 A

  By the way, due to the imbalance of congestion, a problem of abandonment of reservations has occurred in addition to a decrease in boarding efficiency. Since the boarding section varies depending on the passenger, the reserved seat may not be reserved by the passenger even though the reserved seat is partially vacant. When a drop in boarding efficiency or an abandonment of reservations occurs, there is a problem that railway operators can not obtain seat charges that can be obtained originally, and transportation revenues are reduced. Furthermore, when the demand for non-reserved seats exceeds the supply, there is a problem that the passenger seating efficiency of non-reserved seats exceeds 100% and passengers who cannot be seated are generated.

  The present invention has been made in view of the above-described problems, and the object of the present invention is to achieve the effects such as improvement in boarding efficiency and improvement in revenues of railway operators. The proposal is to propose a new method for creating a setting plan.

The first form is
Plan candidates including at least the number of trains of the target train to be planned for the number of trains, the number of seats supplied for each seat between the stations where the target train stops, and the seat fee for each seat between the stations (for example, FIG. Plan candidate generation step (for example, step A1 in FIG. 3) for generating the number of trained vehicle solution candidate data 530) of 4;
Using the seat type selection model that models the probability of the seat type selected by the passenger using the boarding time and the seat type charge, the boarding time between each station of the target train, and the plan candidate, the plan candidate A first seat type selection probability calculating step (for example, step A3 in FIG. 3) for calculating the seat type selection probability;
Estimating the number of seat type users between the stations of the target train and estimating the number of seat type users between the stations of the target train using the predetermined user demand between the stations of the target train and the seat type selection probability. Steps (eg, step A3 in FIG. 3);
A first evaluation step (e.g., step A5 in FIG. 3) for evaluating the plan candidate using the seat type supply seat number and the seat type user number;
And generating an arbitrary number of the plan candidates related to the target train in the candidate generation step, the first seat type selection probability calculation step, the first seat type user number estimation step, and the first This is a method for determining the number of trains for a train, in which the number of trains for the target train is determined based on the plan candidate having the best evaluation result.

As another form,
A plan for generating a plan candidate including at least the number of trains of the target train to be planned for the number of trains, the number of seats supplied for each seat between the stations where the target train stops, and the seat fee for each seat between the stations Candidate generation means;
Using the seat type selection model that models the probability of the seat type selected by the passenger using the boarding time and the seat type charge, the boarding time between each station of the target train, and the plan candidate, the plan candidate First seat type selection probability calculating means for calculating a seat type selection probability;
Estimating the number of seat type users between the stations of the target train and estimating the number of seat type users between the stations of the target train using the predetermined user demand between the stations of the target train and the seat type selection probability. Means,
First evaluation means for evaluating the plan candidate using the seat type supply seat number and the seat type user number;
The candidate generation means generates an arbitrary number of the plan candidates related to the target train, and for each plan candidate, the first seat type selection probability calculation means, the first seat type user number estimation means, A train number determination system that determines the number of trains of the target train based on the plan candidate that is executed by the first evaluation means and has the best evaluation result (for example, the number of trains and seat type setting in FIG. 13). The plan creation device 1) may be configured.

  According to the first embodiment, an arbitrary number of plans including at least the number of trains of the target train to be planned for the number of trains, the number of seats supplied for each seat between the stations, and the seat fee for each station. Candidates are generated. Then, each plan candidate is evaluated using the seat type supply seat number included in the plan candidate and the estimated number of seat type users between the stations, and based on the plan candidate having the best evaluation result. The number of trains for the target train is determined.

  Here, since the number of seat type users is estimated using the seat type selection probability calculated using the seat charge between stations included in the plan candidate, the seat type estimated for each plan candidate is estimated. Usage demand is different. Thereby, for example, it becomes possible to determine the number of trained vehicles so as to minimize the difference between the number of seats supplied for each seat type and the number of users.

As a second form, there is a method for determining the number of organized vehicles according to the first form,
The first evaluation step is a step of evaluating the plan candidate using a difference between the seat type supply seat number and the seat type user number between adjacent stop stations of the target train.
You may comprise the formation vehicle number determination method.

  According to this second mode, the plan candidate is evaluated using the difference between the number of seat-type supply seats and the number of seat-type users between adjacent stop stations of the target train. As a result, for example, it is possible to improve the boarding efficiency of the designated seat by setting the plan candidate that minimizes the difference between the number of seats supplied for each seat type and the number of users as the best evaluation result. It is also possible to reduce the number of passengers who can not be seated.

As a third form,
The number of seats by type of seating at each station where the target train stops, set based on the number of organized vehicles determined by executing the first or second form of the number of organized vehicles, and between the stations A seat type plan candidate generating step (for example, step A13 in FIG. 3) for generating a seat type plan candidate (for example, the seat type solution candidate data 550 in FIG. 11) including at least a seat type fee;
A second seat type selection probability calculating step for calculating a seat type selection probability for the seat type plan candidate using the seat type selection model, a boarding time between stations of the target train, and the seat type plan candidate. (For example, step A15 in FIG. 3);
Using the predetermined user demand between the stations of the target train and the seat type selection probability calculated in the second seat type selection probability calculation step, the seat type between the stations for the seat type plan candidate A second seat type user number estimation step (for example, step A15 in FIG. 3) for estimating the number of users;
1) Seat type utilization rate obtained from the seat type supply seat number included in the seat type plan candidate and the seat type user number estimated in the second seat type user number estimation step, 2) the second The seat type plan candidate using at least one of the number of seat type users estimated in the seat type user number estimation step and the charge revenue obtained from the seat type charge candidate included in the seat type plan candidate A second evaluation step (e.g., step A17 in FIG. 3) for evaluating
The number of seat type plan candidates related to the target train is generated in the seat type plan candidate generation step, the second seat type selection probability calculating step, and the second seat type user number estimation A seat type setting plan creation method for executing the step and the second evaluation step and creating a seat type setting plan for the target train based on the seat type plan candidate that has the best evaluation result. good.

  According to the third aspect, an arbitrary number of seat type plan candidates including at least the seat type supply seat number between the stations set based on the determined number of organized vehicles and the seat type fee between the stations is generated. The Each seat type plan candidate is evaluated using at least one of 1) seat type utilization rate and 2) toll revenue, and based on the seat type plan candidate with the best evaluation result, A seat type setting plan is created. As a result, for example, by setting a seat type plan candidate that maximizes the seat type utilization rate as the best evaluation result, the boarding efficiency of the seat type can be maximized. Further, by making the seat type plan candidate that maximizes the fee revenue the best evaluation result, it is possible to improve the revenue on the railroad operator side.

Explanatory drawing of the formation of the Shinkansen train and the number of seats. Explanatory drawing of the line section of the Shinkansen train. The flow chart of the number of formation vehicles / seat type setting plan creation processing. The data structural example of solution candidate data for number of formation vehicles. Explanatory drawing of a demand prediction calculation process. The data structural example of reserved seat selection probability candidate data. The data structural example of user demand OD data. Explanatory drawing of seat type demand calculation. Illustration of the number of reserved seats sold. Explanatory drawing of evaluation index value calculation. The data structural example of the solution candidate for seat classification. Explanatory drawing of evaluation index value calculation. The structural example of the number of formation vehicles and seat classification setting plan creation apparatus in 1st Example. The flow chart of the number of formation vehicles / seat type setting plan creation processing. FIG. 15 is a continuation of the flowchart of FIG. 14. The structural example of the Shinkansen operation system in 2nd Example. An example of a non-reserved seat express ticket. An example of a seat guidance screen. An example of a seat type display system. An example of a seat type individual guidance device.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the applicable embodiment of the present invention is not limited to this.

[principle]
First, the principle of creating a setting plan for the number of organized vehicles and seats in this embodiment will be described. In the present embodiment, a case will be described in which a plan for setting the number of vehicle formations / seat types for the “Shinkansen” which is a type of honor train is created. In addition, the Shinkansen vehicle will be described as a “Shinkansen vehicle”, and the operated train will be described as a “Shinkansen train”.

  FIG. 1 is a diagram illustrating the formation of a Shinkansen train and the number of seats in the present embodiment. As shown in FIG. 1, the Shinkansen vehicle in this embodiment is formed by M vehicles. The number of seats m of each vehicle is all equal, and therefore the total number of seats is “m × M”. In addition, the seat types of the Shinkansen train (hereinafter referred to as “seat types”) are two types, “designated seats” and “free seats”.

FIG. 2 is a diagram illustrating a configuration of a section of a Shinkansen train in the present embodiment. As shown in FIG. 2, the line section of the Shinkansen train in this embodiment is comprised of four stations, A station to D station.
The Shinkansen train runs from A station, which is the first station, to D station, which is the last station, and stops at each station on the way. Here, the section between A station and B station, which is between adjacent stop stations, is “first section”, the section between B station and C station is “second section”, and the section between C station and D station is “first section”. This is referred to as “3 sections”.

  And in this embodiment, the number of train formations of the Shinkansen train is set variably, and then the number of seats assigned to each seat type is set variably for each adjacent stop station section, Create a seat type setting plan. In this embodiment, OD (Origin to Destination), which is a combination of departure and arrival stations, is “A station → B station”, “A station → C station”, “A station → D station”, as shown in FIG. It is a combination of six departure / arrival stations of “B station → C station”, “B station → D station”, and “C station → D station”.

  FIG. 3 is a flowchart showing a flow of creating the number of organized vehicles / seat type setting plan in the present embodiment. FIG. 3 shows a case where a plan for setting the number of trains and seat types is created for all Shinkansen trains operating on a certain day.

  In general, first, the number of trains for each of the Shinkansen trains to be created is determined (steps A1 to A11). Next, for all these Shinkansen trains, the number of seats for each seat type is determined assuming that the Shinkansen train is organized with the determined number of trains set (steps A13 to A23). Here, if the number of trains on the Shinkansen train is determined, the total number of seats on the Shinkansen train is determined. And if the number of seats of the designated seat which is one seat type is determined, the number of free seats which is the other seat type is also determined. Therefore, with regard to the number of seats, a designated seat that is one of the seat types is determined.

  Specifically, in the determination of the number of trains, the process of Loop A for all the Shinkansen trains is executed. In this loop A, a plurality of (L) candidate trains for determining the number of trains (the train candidates for trains: plan candidates) are set for the target Shinkansen train (target train) (step A1).

  FIG. 4 is a diagram illustrating an example of the solution candidate data 530 for the number of trained vehicles. As shown in FIG. 4, the solution candidate data 530 for the number of trained vehicles includes a solution candidate No. that is an identification number. In association with 531, trained vehicle number candidate data 532, designated seat supply seat number candidate data 533, seat designation fee candidate data 534, and designated seat sales coefficient candidate data 535 are stored.

  The number of trained vehicle candidates data 532 is data that defines candidates for the number of trained vehicles. The number of trained vehicles is set at random, for example, so as to be equal to or less than the maximum number of trainable vehicles (for example, 15 cars).

  The reserved seat supply seat number candidate data 533 is data in which candidates for the supply seat number of the specified seat are determined in units of sections. In FIG. 4, the number of seats to be supplied for the designated seats in each of the first section to the third section is determined. The number of seats supplied in each section is set at random, for example, so as to be equal to or less than the maximum number of seats of the target Shinkansen train, which is determined from the number of trains set in the train number candidate data 532.

  The seat designation fee candidate data 534 is data in which seat fee candidates for the designated seat are determined in OD units. In FIG. 4, for each combination (OD) of departure / arrival stations between the A station and the D station, the seat fee for the designated seat is determined in a table format. The seat fee for the designated seat is set at random, for example, so as to be a value within a predetermined range (for example, 310 yen to 710 yen). Since the target Shinkansen trains run from station A (the first station) to station D (the last station), the seat price is set to “-(none)” for the OD in the opposite direction to the service direction. ing.

  The reserved seat sales coefficient candidate data 535 is data in which candidates for the reserved seat sales coefficient are determined in OD units. In FIG. 4, the sales coefficient of the reserved seat is defined in a table format for each combination (OD) of departure and arrival stations between A station and D station. Here, the sales coefficient is a value indicating the ratio of the number of sold seats actually sold to the number of demand seats of the designated seat. This sales coefficient is set at random, for example, so as to be a value within the range of “0.0 to 1.0”.

  Originally, for all ODs, it is desirable to sell the same number of seats as the demanded seats with the sales coefficient of the designated seat set to “1”. However, since the riding section varies depending on passengers, all ODs If the sales coefficient is set to “1”, seats may not be sold to all passengers depending on the section. Therefore, the sales coefficient of the designated seat is set in the range of “0.0 to 1.0”, and adjustment is made so that the seat allocation is flexibly performed. As for the sales coefficient, the OD in the direction opposite to the direction of operation of the target Shinkansen train is set to “-(none)”, similarly to the seat charge for the reserved seat.

  If a plurality of (L) trained vehicle number solution candidates are set for the target train, subsequently, a demand prediction calculation process is performed using the set trained vehicle number solution candidates (step A3).

  FIG. 5 is a diagram for explaining the flow of the demand prediction calculation process. According to FIG. 5, first, the designated seat selection probability candidate data 611 is calculated using the seat designation fee candidate data 534, the seat type selection model 522, and the other data 523 included in the solution candidates for the number of trains for each train. To do.

  The seat type selection model 522 is a predetermined mathematical model for calculating the seat type selection probability. The inventor of the present application conducted a trip survey targeting passengers to model the selection behavior of the seat type. As a result, in addition to factors such as boarding time and seat-type fees, passenger seat type selection behavior includes factors such as `` seat certainty '', which is a conscious factor assumed to exist behind selection behavior, It was found that the "degree of freedom" factor had an effect. Based on the results of this survey, a mathematical model was constructed to calculate the seat type selection probability using the seat type fee, boarding time, seating certainty, and seating freedom as variables.

According to the mathematical model, the seat type selection probability P _ij is calculated by the following equation (1).

In equation (1), the subscript “i” of the selection probability P _ij indicates the OD identification number, and in this embodiment, there are six ODs, so “i = 1, 2,..., 6”. Take one of the following values:
The subscript “j” indicates a seat type. In this embodiment, “j = 1” indicates a designated seat and “j = 2” indicates a free seat.

Further, “U _ij ” is calculated by the following equations (2) and (3).

In Expressions (2) and (3), “c _ij ” indicates the seat type fare / charge when the j-th seat type is used in the i-th OD, and “t _i ” indicates the i-th OD The boarding time is shown, “r _j ” shows the certainty of seating at the i-th OD, and “f _i ” shows the degree of freedom of seating at the i-th OD. Further, “α ₁ ” to “α ₄ ” are weights of the seat type fare / charge c, the boarding time t, the seating certainty r, and the seating freedom degree f, respectively.

That is, in the above equation (1), the seat charge of the designated seat included in the seat reservation fee candidate data 534, the fare, the boarding time t included in the other data 523, and the seating factor (the seating certainty r and The seating type selection probability _Pij is calculated by substituting the degree of freedom of seating f). Then, the selected selection probability P _ij is collected in units of OD, and the reserved seat selection probability candidate data 611 is generated.

  FIG. 6 shows an example of the reserved seat selection probability candidate data 611. As shown in FIG. 6, the reserved seat selection probability candidate data 611 is generated corresponding to each of the N trained vehicle number solution candidates, and the specified seat selection probability, which is the probability that the passenger selects the reserved seat, is expressed as OD. Data determined in units.

  Next, seat type demand calculation is performed using the user demand OD data 521 and the generated designated seat selection probability candidate data 611.

  FIG. 7 shows an example of user demand OD data 521. As shown in FIG. 7, the user demand OD data 521 is data in which the number of passengers used is determined in OD units for each Shinkansen train. The user demand OD data 521 may be data of the actual value of the user or data of the estimated value of the user.

  FIG. 8 is a diagram illustrating the flow of seat type demand calculation. As shown in FIG. 8, the user demand OD data 521 corresponding to the target train is multiplied by each of the L reserved seat selection probability candidate data 611i (i = 1, 2,..., L). Thus, L reserved seat demand candidate data 612i (i = 1, 2,..., L) is calculated.

  Specifically, the number of users included in the user demand OD data 521 and the reserved seat selection probability included in the reserved seat selection probability candidate data 611 are respectively multiplied by the OD unit, so that the specified seat of the corresponding OD is obtained. , The reserved seat candidate data 612 is calculated.

  Once the reserved seat demand candidate data 612 is calculated, subsequently, the reserved seat demand candidate data 612 and the reserved seat sales coefficient candidate data 535 included in the solution candidates for the number of trains for each train are used. Perform number calculations.

  FIG. 9 is a diagram illustrating the flow of calculating the number of reserved seat sales. As shown in FIG. 9, the designated seat sales candidate number data 613 is calculated by multiplying the designated seat demand candidate data 612 and the designated seat sales coefficient candidate data 535.

  Specifically, for each of the L number of trained vehicle solution candidates i (i = 1, 2,..., L), the corresponding designated seat demand candidate data 612i (i = 1, 2,..., L). L) and designated seat sales coefficient candidate data 535i (i = 1, 2,..., L), and by multiplying the number of designated seat demand seats and the designated seat sales coefficient in OD units, Calculate N reserved seat sales seat number candidate data 613i (i = 1, 2,..., L) corresponding to each of the set vehicle number solution candidates i (i = 1, 2,..., L). To do.

  Here, by calculating the difference between the user demand OD data 521 and the reserved seat sales seat number candidate data 613, the free seat demand candidate, which is data in which candidates for the demand seat number of the free seats are determined in OD units. Data 614 can be determined. It can be said that the number of demand seats for non-reserved seats indicates the number of users of non-reserved seats of the corresponding OD. This unreserved seat demand candidate data 614 can be used, for example, to calculate the expected free seat congestion rate, as will be described later in the second embodiment.

  Once the reserved seat sales seat number candidate data 613 is calculated, the constraint condition is checked for each of the specified seat sales seat number candidate data 613. The restriction condition is, for example, that the number of reserved seat sales seats included in the reserved seat sales seat number candidate data 613 does not exceed the total number of seats of the Shinkansen vehicle. Then, the solution candidates for the number of organized vehicles corresponding to the reserved seat sales seat number candidate data 613 that does not satisfy the constraint conditions are discarded. On the other hand, the solution candidate for the number of organized vehicles corresponding to the designated seat sales seat number candidate data 613 that satisfies the constraint condition is stored, and the demand prediction calculation process is terminated.

  Returning to FIG. 3, when the demand prediction calculation process is completed, the evaluation index value calculation process is subsequently performed (step A5). In the present embodiment, the vehicle composition solution candidate is evaluated using “the sum of squares of supply and demand residuals” representing the difference between the demand and supply for each seat type as the evaluation index value (first evaluation index value).

  FIG. 10 is a diagram for explaining a method of calculating the supply-demand residual sum of squares. As shown in FIG. 10, for each stored solution for the number of trained cars, the supply-demand residual sum of squares for each seat type is calculated from the corresponding designated seat sales seat number data 613 and designated seat supply seat number candidate data 533. Is calculated as an evaluation value.

“Supply and demand residual sum of squares” is calculated by the following equation (4).
In the above equation (4), “X _sj ” is the number of seats X _sj sold for the seat type j in the section s, and “Y _sj ” is the number of seats supplied for the seat type j in the section s. The seat type j is j = 1, 2, where “j = 1” indicates a designated seat and “j = 2” indicates a free seat. In the section s, s = 1 to 3, “s = 1” represents the first section, “s = 2” represents the second section, and “s = 3” represents the third section.

  In other words, for the reserved seat, based on the reserved seat sales seat number candidate data 613, the number of seats sold in the section unit is calculated as the demand seat number of the reserved seats in the section unit from the number of seats sold in the OD unit. . For the free seats, based on the free seat demand candidate data 614, the number of free seats in the section unit is calculated from the number of free seats in the OD unit. Next, the number of reserved seats for each section is calculated from the designated seat supply seat number candidate data 533 and the total number of seats. Next, for each section, calculate the “Supply and Demand residual square value”, which is the value obtained by squaring the difference between the number of demand seats and the number of seats supplied for the seat type (designated seats and unreserved seats). The “sum of the squares of residual supply and demand”, which is the sum of the “square of the difference”, is used as the “first evaluation index value” for the Shinkansen train.

  If the evaluation index value calculation process is performed, the solution candidate having the smallest calculated first evaluation index value is selected as the best solution from the stored solution candidates for the number of trained vehicles (step A7), and selected. The best solution is saved (step A9). Loop A is performed in this way.

  When the processing of Loop A for all the Shinkansen trains is completed, the number of trains for each Shinkansen train is determined by referring to the stored best solution of each Shinkansen train, the vehicle operation plan, and the like (step A11). .

  Once the number of trains for each Shinkansen train has been determined, a seat type setting plan for each Shinkansen train is created.

  In creating the seat type setting plan, the process of Loop B for each of all Shinkansen trains is executed. In the process of Loop B, first, for a target Shinkansen train (target train), a plurality of (N) seat type setting plan creation candidate candidates (seat type solution candidates: seat type plan candidates) are set. Perform (Step A13).

  FIG. 11 is a diagram showing an example of the seat type solution candidate data 550. As shown in FIG. 11, the data structure of the seat type solution candidate data 550 corresponds to the data structure obtained by excluding the trained vehicle number candidate data 532 from the trained vehicle number solution candidate data 530 (see FIG. 4). , The solution candidate No. which is the identification number. In association with 551, designated seat supply seat number candidate data 553, seat designated fee candidate data 554, and designated seat sales coefficient candidate data 555 are stored.

  In the designated seat supply seat number candidate data 553, the supply seat number of each section is set at random, for example, so as to be equal to or less than the maximum seat number determined by the determined number of trains of the target train.

  If a plurality (N) of seat type solution candidates are set for the target train, subsequently, a demand prediction calculation process using the set seat type solution candidates is performed (step A15). This demand prediction calculation process is the same process as the demand prediction calculation process (see FIG. 5) in the determination of the number of trained vehicles described above. That is, using the seat specification fee candidate data 554, the seat type selection model 522, and the other data 523 included in the seat type solution candidate data 550, the designated seat demand candidate data 612 corresponding to each seat type solution candidate. Is calculated. Next, the designated seat sales seat number candidate data 613 is calculated from the designated seat demand candidate data 612 and the designated seat sales coefficient candidate data 555. Then, among the calculated designated seat sales seat number candidate data 613, the seat type solution candidate corresponding to the designated seat sale seat number candidate data 613 that does not satisfy the restriction condition is discarded, and the designated seat sale seat number candidate satisfying the restriction condition is discarded. The seat type solution candidate corresponding to the data 613 is stored.

  If the demand prediction calculation process is performed, then the evaluation index value calculation process is performed (step A17). In this embodiment, seat type solution candidates are evaluated using the boarding efficiency (load factor) of the designated seat as the evaluation index value (second evaluation index value).

  FIG. 12 is a diagram illustrating a method for calculating the riding efficiency. As shown in FIG. 12, for each stored seat type solution candidate, the second evaluation of the riding efficiency for each section is performed from the corresponding designated seat supply seat number candidate data 553 and designated seat sales seat number candidate data 613. Calculated as an index value.

  That is, based on the reserved seat sales seat number candidate data 613, the number of reserved seats for each section is calculated from the number of reserved seats for each OD unit. Next, for each section, divide the number of seats sold for the designated seat by the number of seats supplied for the designated seat to calculate the boarding efficiency for the section, and calculate the average value of boarding efficiency for all sections as the second evaluation for the Shinkansen train. The index value.

  When the evaluation index value calculation processing is performed, the solution candidate having the maximum calculated second evaluation index value is selected as the best solution from the stored seat type solution candidates (step A19), and the selected best The solution is saved (step A21). Loop B is performed in this way.

  When the processing of Loop B for all the Shinkansen trains is completed, a seat type setting plan for all the Shinkansen trains operating on the operating day is created using the stored best solution of each Shinkansen train (step A23). ). That is, the seat type setting plan for the operation day is created by uniquely associating the number of seats supplied from the best solution for the designated seats between stations, the seat charge in OD units, and the sales coefficient with the Shinkansen train.

[First embodiment]
Next, a description will be given of an embodiment of the number-of-scheduled-vehicles / seat-type setting plan creation device 1 which is a device for creating the number-of-scheduled-vehicles / seat-type setting plan for the Shinkansen train according to the principle described above.

<Functional configuration>
FIG. 13 is a block diagram showing a functional configuration of the number of organized vehicles / seat type setting plan creation device 1. According to FIG. 13, the number of organized vehicles / seat type setting plan creation device 1 functionally includes a control unit 10, an input unit 20, a display unit 30, a communication unit 40, and a storage unit 50. Configured as a computer system.

  The control unit 10 is realized by an arithmetic device such as a CPU, for example, and based on programs and data stored in the storage unit 50, data input from the input unit 20, and the like, the number of trained vehicle / seat type setting plan creation device Instructions and data transfer to each part constituting 1 are performed, and a formation vehicle number / seat type setting plan creation process (see FIGS. 14 and 15) according to the number of formation vehicles / seat type setting plan creation program 510 is performed.

  The input unit 20 is an input device realized by, for example, a keyboard, a mouse, a touch panel, various switches, and the like, and outputs an input signal corresponding to an operation input to the control unit 10.

  The display unit 30 is a display device realized by, for example, an LCD (Liquid Crystal Display) or an ELD (Electronic Luminescent Display), and displays various screens based on display signals input from the control unit 10.

  The communication unit 40 is a communication device realized by, for example, a wireless communication module, a router, a modem, a TA, a wired communication cable jack, or a control circuit, and performs data communication with an external device.

  The storage unit 50 is a system program for realizing various functions for the control unit 10 to integrally control the number of organized vehicles / seat type setting plan creation device 1, and a program and data for realizing the present embodiment. And the like, and is used as a work area of the control unit 10 and temporarily stores calculation results executed by the control unit 10 according to various programs and an input signal from the input unit 20. In the present embodiment, the storage unit 50 stores a number of organized vehicles / seat type setting plan creation program 510 as a program, and as data, user demand OD data 521, seat type selection model data 522, Other data 523, solution candidate data for train number 530, seat type solution candidate data 550, designated seat selection probability candidate data 611, designated seat demand candidate data 612, designated seat sales seat number candidate data 613, , Unreserved seat demand candidate data 614, best solution data 621, trained vehicle number plan data 622, and seat type setting plan data 623 are stored.

<Process flow>
14 and 15 are flowcharts for explaining the number of trained vehicle / seat type setting plan creation processing. 14 and 15, the same steps as those in the set vehicle number / seat type setting plan creation process shown in FIG. 3 are denoted by the same step numbers, and the description will be simplified, and different steps will be mainly described.

  According to FIG. 14, first, the control unit 10 performs a process of Loop A for all the Shinkansen trains. In the loop A, the control unit 10 sets a plurality (L pieces) of solution candidates for the number of organized vehicles for the target Shinkansen train (target train) (step A1). Next, a demand prediction process using the set candidate solution for each set number of vehicles or predetermined user demand OD data 521 or the like is performed, and designated seat demand candidate data 612 corresponding to each set number of set vehicle number solution candidates. Alternatively, the number of reserved seat sales candidate data 613 and the unreserved seat demand candidate data 614 are calculated (step A3).

  Subsequently, an evaluation index value calculation process is performed, and a supply-demand residual sum of squares representing a difference between the demand and supply for each seat type is calculated as a first evaluation index value for each solution candidate for the number of trained vehicles (step A5). ). Then, a solution candidate having the smallest first evaluation index value is selected from the solution candidates for the number of trained vehicles, and stored as a provisional best solution (step A7).

  Subsequently, the control unit 10 performs solution candidate update processing using the solution candidate for the number of formation vehicles from which the provisional best solution is removed from the stored solution candidates for the number of formation vehicles (step B9). That is, for each of the trained vehicle number solution candidates excluding the provisional best solution, the trained vehicle number candidate data 532, the designated seat supply seat number candidate data 533, the seat designation fee candidate data 534 included in the solution candidate, and the designation Some values of the seat sales coefficient candidate data 535 are changed.

  The change of the solution candidate may be changed by, for example, determining the number of values to be changed or the amount of change at random, or the candidate number data 532 for the number of formation vehicles included in each solution candidate for the number of formation vehicles, the designated seat The combination of the supplied seat number candidate data 533, the seat designation fee candidate data 534, and the designated seat sales coefficient candidate data 535 may be exchanged, or a completely new solution candidate for the number of formation vehicles may be regenerated at random. good.

  When the solution candidate update process is completed, the control unit 10 performs a demand prediction calculation process using the updated solution candidates for the number of trained vehicles (step B11), and subsequently performs an evaluation index value calculation process. A first evaluation index value is calculated for each solution candidate for the number of trained vehicles (step B13).

  Then, the trained vehicle number solution candidate having the smallest first evaluation index value is selected as the best solution from the updated trained vehicle number solution candidates (step B15), and the first evaluation index value of the selected best solution is selected. And the stored first evaluation index value of the provisional best solution. As a result, if the first evaluation index value of the best solution selected this time is smaller than the stored first evaluation index value of the provisional best solution (step B17: YES), the best solution selected this time is set to a new one. Update / save as provisional best solution (step B19).

  Thereafter, it is determined whether an end condition is satisfied. Here, the “end condition” is: (1) The number of consecutive times that the first evaluation index value of the newly selected best solution is not smaller than the first evaluation index value of the best solution previously selected is a predetermined number of times (for example, 3000 times) or (2) the number of repetitions of steps B9 to B19 has reached a predetermined upper limit number (for example, 100,000 times).

  As a result, if it is determined that the end condition is not satisfied (step B21: NO), the process returns to step B9. On the other hand, if it is determined that the termination condition is satisfied (step B21: YES), loop A is terminated.

  When the processing of Loop A for all Shinkansen trains is completed, the control unit 10 determines each Shinkansen based on the best solution (provisional best solution) stored for each Shinkansen train, the vehicle operation plan, and the like. The number of trains is determined (step A11).

  As shown in FIG. 15, when the number of trains for each Shinkansen train is determined, the control unit 10 subsequently performs a process of Loop B for each of all the Shinkansen trains. In loop B, a plurality (N) of seat type solution candidates are set for the target Shinkansen train (target train) (step A13). Next, by performing a demand prediction process using the set solution candidates for each seat type and the user demand OD data 521 and the like determined in advance, the designated seat demand candidate data 612 corresponding to each solution type solution candidate, The designated seat sales seat number candidate data 613 and the unreserved seat demand candidate data 614 are calculated (step A15).

  Subsequently, an evaluation index value calculation process is performed, and the boarding efficiency of the designated seat is calculated as the second evaluation index value for each of the seat type solution candidates (step A17). Then, a solution candidate having the maximum calculated second evaluation index value is selected from the solution candidates for the seat type and stored as a provisional best solution (step B23).

  Subsequently, the control unit 10 performs a solution candidate update process using the seat type solution candidates excluding the provisional best solution from the stored seat type solution candidates (step B25). That is, for each seat type solution candidate excluding the provisional best solution, a part of the reserved seat supply seat number candidate data 553, the seat specified fee candidate data 554, and the specified seat sales coefficient candidate data 555 included in the solution candidate. Change the value of.

  When the solution candidate update process is completed, the control unit 10 performs a demand prediction calculation process using the updated seat type solution candidate (step B27), and subsequently performs an evaluation index value calculation process. The second evaluation index value of each type solution candidate is calculated (step B29).

  Then, the seat type solution candidate having the maximum second evaluation index value is selected as the best solution from the updated seat type solution candidates (step B31), and the second evaluation index value of the selected best solution is stored. The second evaluation index value of the provisional best solution being compared is compared. As a result, if the second evaluation index value of the selected best solution is greater than the stored second evaluation index value of the provisional best solution (step B33: YES), the currently selected best solution is used as a new provisional solution. Update / save as the best solution (step B35).

  Thereafter, it is determined whether an end condition is satisfied. Here, the “end condition” is: (1) The number of consecutive times that the second evaluation index value of the newly selected best solution is not greater than the second evaluation index value of the best solution selected last time is a predetermined number of times (for example, 3000). Or (2) the number of repetitions of steps B25 to B35 has reached a predetermined upper limit number (for example, 100,000 times).

  As a result, if it is determined that the end condition is not satisfied (step B37: NO), the process returns to step B25. On the other hand, if it is determined that the termination condition is satisfied (step B37: YES), the loop B is terminated.

  After completing the processing of Loop B for all Shinkansen trains, use the best solution stored for each Shinkansen train (provisional best solution) to create a seat type setting plan for all Shinkansen trains on the operating day. Create (step A23). Thereafter, the control unit 10 causes the display unit 30 to display the created train number / seat type setting plan created (step B37), and ends the train number / seat type setting plan creation process.

<Action and effect>
As described above, according to the first embodiment, the number of formation vehicles / seat type setting plan creation device 1 sets a plurality of solution candidates (solution candidates for the number of formation vehicles) for each Shinkansen train to be processed, The number of formation vehicles is determined using a solution candidate having the best evaluation result (best solution). Thereafter, similarly, a plurality of solution candidates (seat type solution candidates) are set for each of the Shinkansen trains, and a seat type setting plan is created using the solution candidate (best solution) with the best evaluation result.

  At this time, when determining the number of trained vehicles, each of the solution candidates is evaluated using the sum of squares of supply and demand residuals representing the difference between the demand and supply of the seat type as the first evaluation index value, The solution candidate with the smallest sum is selected as the best solution, and when creating the seat type setting plan, each solution candidate is evaluated using the riding efficiency of the designated seat as the second evaluation index value. By selecting the solution candidate with the largest value as the best solution, it is possible to create a seat type setting plan that matches the passenger demand. And the boarding efficiency of each train can be improved by operating the train based on the seat type setting plan created in this way.

[Second Embodiment]
Next, an embodiment of the Shinkansen operation system 100, which is a system including the number of organized vehicles / seat type setting plan creation device 1 of the first embodiment, will be described.

  FIG. 16 is a diagram illustrating an example of a system configuration of the Shinkansen operation system 100. According to FIG. 16, the Shinkansen operation system 100 is configured to include the number of organized vehicles / seat type setting plan creation system 3, a ticket sales system 5, a seat assignment notification system 7, and a plurality of Shinkansen vehicles 9. The

  The number of organized vehicles / seat type setting plan creation system 3 is a computer system that creates the number of organized vehicles / seat type setting plan and seat allocation plan for all Shinkansen trains. For example, the number of organized vehicles / seat type setting A plan creation device 1 and a seat allocation plan creation device 2 are provided. In addition, the number of organized vehicles / seat type setting plan creation system 3 sends the created number of organized vehicles / seat type setting plan and seat allocation plan to the ticket sales system 5 and seat allocation through a predetermined communication network N. To the notification system 7.

  The number of organized vehicles / seat type setting plan creation device 1 corresponds to the number of organized vehicles / seat type setting plan creation device 1 described in the first embodiment, and creates the number of organized vehicles / seat type setting plan for all Shinkansen trains. To do. Then, the number of organized vehicles / seat type setting plan creation device 1 provides the created number of organized vehicles / seat type setting plan to the seat allocation plan creation device 2.

  The seat allocation plan creation device 2 creates a seat allocation plan for each of the Shinkansen trains based on the number of organized vehicles / seat type setting plan acquired from the number of organized vehicles / seat type setting plan creation device 1. Specifically, the seat allocation plan creation device 2 is configured based on the seat type supply seat number candidate data, the seat type charge candidate data, and the seat type sales coefficient candidate data included in the number of organized vehicles / seat type setting plan. The number of seats sold by seat type between adjacent stations of the Shinkansen train is calculated, and seat allocation for the calculated number of seats sold by seat type is determined.

  The ticket sales system 5 is a ticketing window system that sells tickets such as tickets and limited express tickets. A ticket vending machine where passengers themselves purchase tickets by specifying trains, etc. It includes a system of general counters where station staff can meet and specify trains and issue tickets. The ticket sales system 5 includes, for example, a server 51 and a plurality of ticket vending machines 53, and further includes a ticket issuing system for station staff (not shown) as a general window system. The The ticket sales system 5 is connected to the number of trained vehicles / seat type setting plan creation system 3 and the seat assignment notification system 7 through a predetermined communication network N.

  The server 51 determines the number of seats and seats planned for each Shinkansen train based on the number of trains and seat types setting plan acquired from the number of trains and seat type setting plan creation system 3 and the seat allocation plan. Control is performed to display seat guidance including allocation on the display unit of each ticket machine 53. In addition, each ticket vending machine 53 is controlled to issue a ticket on which seat guidance is printed.

  Further, the server 51 uses the number of seats supplied and the number of seats to be sold for each Shinkansen train included in the seat type setting plan, and the expected congestion rate for each seat type (hereinafter referred to as “expected congestion rate”). Is calculated for each adjacent stop station. For example, if the seat type is a non-reserved seat, the number of seats sold and the number of seats supplied for each section between adjacent stations is calculated, and the ratio between the number of seats sold and the number of seats supplied is calculated. As a result, the expected congestion rate of the section is calculated. Then, the calculated expected congestion rate is displayed on the display unit of the ticket vending machine 53 together with the seat allocation plan.

  As described with reference to FIG. 9, the number of demand seats for the unreserved seats is obtained by calculating the difference between the user demand OD data 521 and the reserved seat sales seat number candidate data 613, as obtained from the unreserved seat demand candidate data 614. It can be obtained by converting into candidate data in section units. Here, since the free seat is allowed to stand, the number of free seat demand seats is made equal to the number of free seat sales. The number of seats supplied for free seats is the difference between the total number of seats and the number of reserved seats supplied.

  The ticket vending machine 53 is a ticket issuing device that issues tickets such as boarding tickets and limited express tickets under the control of the server 51. The ticket vending machine 53 includes a display unit integrated with a touch panel. And according to the purchase operation of the ticket of a Shinkansen train, the seat guide of the said Shinkansen train is displayed on a display part. In addition, a ticket is printed on which information on seat allocation and expected congestion rate in a boarding section where passengers wish to board is printed.

  FIG. 17 is a view showing a non-reserved seat limited express ticket 200 which is an example of a ticket issued by the ticket vending machine 53 of the ticket sales system 5. The non-reserved seat limited express ticket 200 is a limited express ticket necessary for the user to get on the unreserved seat on the Shinkansen train. According to FIG. 17, the passenger boarding section (A station → C station) is printed on the upper part of the non-reserved seat limited express ticket 200. Also, below this boarding section, the date (August 10), the train number (XXX), the departure time of the boarding station (17:34), The arrival time at the getting-off station (18:52) and the seat charge (520 yen) for the unreserved seat in the boarding section are printed.

  In addition, on the lower part of the non-reserved seat limited express ticket 200, a free seat guide is printed. This free seat guidance includes, for each OD included in the boarding section, a seat number to which a free seat is assigned in the OD and an expected congestion rate. For example, in FIG. 17, “A station → C station” which is the boarding section is “ABC seat of No. 4 car No. 7 to No. 10” is an unreserved seat, and an unoccupied seat in that section Is predicted to be “150%”.

  FIG. 18 is a diagram illustrating an example of a seat guidance screen W1 displayed on the display unit of the ticket vending machine 53. Note that the same seat guidance screen W1 is displayed not only at the ticket vending machine 53 but also at the ticketing system related to the general window system. Here, a case where “A station → D station” is selected as the boarding section will be described.

  According to FIG. 18, the seat guidance screen W1 displays the seat guidance map of the Shinkansen train that is color-coded by designated seats and unreserved seats for each of the adjacent stop stations from station A to station D. Seat assignments can differ for each Shinkansen train. For example, the seats assigned to reserved seats are colored orange, and the seats assigned free seats are colored blue, so that passengers or station staff can grasp the seat assignment at a glance, so that the seat type is color-coded. ing. For example, when a passenger or a station attendant wishes to reserve a reserved seat, he or she can make a reservation for the seat by touching the part of the seat to be reserved with a finger or a touch pen.

  In addition, the expected congestion rate of free seats is displayed below the seat guide map between adjacent stop stations. In the seat guidance screen W1 shown in FIG. 18, the expected congestion rate of the free seat “A station → B station” is “91%”, and the expected congestion rate of the free seat “B station → C station” is “105%”. ”And the fact that the expected congestion rate of“ C station → D station ”is“ 82% ”is displayed. By displaying the expected congestion rate for each section, the passenger can easily determine the seat type to be selected. Moreover, if it is a station clerk who operates a general window reception system, it is possible to guide a passenger about the prediction of the free seat congestion rate.

  The seat assignment notification system 7 is a communication system that obtains a seat assignment plan for each Shinkansen train from the number of trains and seat type setting plan creation system 3 and notifies the Shinkansen vehicle 9 that operates the Shinkansen train. The seat assignment notification system 7 includes, for example, a ground communication control device 71 communicably connected to each Shinkansen vehicle 9 and a vehicle operation DB (Data Base) 73 by a predetermined communication system.

  The vehicle operation DB 73 is a database in which vehicle operation data defining the correspondence between the Shinkansen vehicle 9 and the Shinkansen train operated by the vehicle is stored and stored.

  The seat assignment notification system 7 is communicatively connected to the number of organized vehicles / seat type setting plan creation system 3 and the ticket sales system 5 through a predetermined communication network N. The seat assignment notification system 7 is configured to allow data communication with each Shinkansen vehicle 9 through a predetermined communication system.

  The seat assignment notification system 7 receives the seat assignment plan for each Shinkansen train from the number of trained vehicles / seat type setting plan creation system 3. Then, the Shinkansen vehicle 9 that operates the Shinkansen train is selected with reference to the vehicle operation data stored and stored in the vehicle operation DB 73. Then, the ground communication control device 71 is controlled to transmit / notify the corresponding seat allocation plan to the selected Shinkansen vehicle 9.

  The bullet train 9 is an excellent train having a plurality of seat types (for example, free seats, reserved seats, and green seats). The Shinkansen vehicle 9 in the present embodiment includes a seat type display system 91, an individual seat type guidance device 93, and a control device 95 for each seat.

  FIG. 19 is a diagram illustrating an example of the seat type display system 91. As shown in FIG. 19, the seat type display system 91 transports (moves) the cabin electronic display board 310, which is arranged so as to be suspended from the ceiling of the cabin 300, and the cabin electronic display board 310. And a transport mechanism to be configured.

  Two rails 314 are extended on the ceiling portion of the passenger room 300 along the advancing / retreating direction (front-rear direction) of the train, and the guest room electronic display board is integrated with the movable part 312 fitted in the groove of the rail 314. 310 is installed. When the movable portion 312 moves in the rail 314, the cabin electronic display board 310 moves in the cabin 300 in the front-rear direction. A well-known transport mechanism can be applied as the transport mechanism of the cabin electronic display panel 310. For example, a spiral groove mechanism or the like can be applied by providing a spiral groove in the rail 314 or a shaft provided with a convex portion on the outer periphery and rotating the shaft. By this transport mechanism, the cabin electronic display panel 310 is transported to the boundary position between the free seat and the designated seat.

  On the guest room electronic display board 310, information on the seat type assigned to the seat in the guest room 300 is displayed. For example, a seat number to which a reserved seat and a free seat are assigned and an expected congestion rate of the free seat are displayed for each adjacent station. In FIG. 19, it is displayed that “No. 1 to No. 5” seats are designated seats and “No. 6 to No. 20” seats are non-reserved seats in the section “A station → B station”.

  The control device 95 determines the seat type of each seat in each guest room based on the seat allocation plan for the Shinkansen train notified from the seat allocation notification system 7. Then, the boundary position between the unreserved seat and the designated seat is calculated, the transport mechanism is controlled toward the position to move the cabin electronic display board 310, and the display according to the seat allocation plan is displayed on the cabin electronic display board 310. Control to perform.

  FIG. 20 is a diagram illustrating an example of the seat type individual guidance device 93. The seat type individual guidance device 93 is configured to include a seat electronic display board 322 at the upper back of the seat 320. The seat electronic display board 322 displays information on the seat type assigned to the seat immediately behind the seat 320, information on the expected congestion rate of the seat type, and the like.

  The control device 95 determines the seat type of each seat in each guest room based on the seat type allocation plan notified from the seat allocation notification system 7, and displays it on the individual seat type guidance device 93 corresponding to the seat. Display information is generated and transmitted to display the display information. For example, in FIG. 20, a message “The seat of the customer is a free seat at station A → C station and a reserved seat from station C” is displayed. Therefore, the passenger seated in the seat immediately behind the seat 320 can know that the current seat can be used up to the station C, but the passenger must move to another seat from the station C. .

[Modification]
Note that embodiments to which the present invention can be applied are not limited to the above-described embodiments, and it is needless to say that the embodiments can be appropriately changed without departing from the spirit of the present invention.

(A) Application system For example, in the above-described embodiment, the case of creating a set number of trains / seat type setting plan for a Shinkansen train, which is a kind of an honor train, has been described, but for an honor train other than the Shinkansen, The method of the present invention is applicable. For example, the creation of a seat type setting plan for reserved seats and unreserved seats on express trains and express trains can be performed in the same manner as in the above-described embodiment.

  Also, the seat types are not limited to the two types of reserved seats and free seats described above. For example, create a plan for setting the number of trains and type of seats that include green cars (for example, deluxe green cars and super green cars) that have already been adopted or will be adopted for various types of high-end trains. Is also possible.

(B) Types of evaluation index values Further, in the above-described embodiment, it has been described that the supply-demand residual sum of squares for each seat type is used as the first evaluation index value for evaluating the solution candidates for determining the number of trained vehicles. However, any value may be used as long as it represents a difference between demand and supply of a certain seat type. For example, the average value or sum of absolute values of the difference between the demand and supply for each seat type may be used, or the sum of cubes may be used.

  Moreover, although demonstrated as what uses the boarding efficiency of the whole train as a 2nd evaluation index value which evaluates the solution candidate for creating a seat classification setting plan, this 2nd evaluation index value is good also as the following values. It is preferable to determine which one is used according to the operation status / operation time of the train, the purpose of the railway operator, and the like.

  For example, the transportation revenue of the railway operator can be used as the second evaluation index value. Specifically, for each seat type, transportation revenue is calculated by multiplying the number of seats sold by the seat charge. And the value which calculated transportation revenue calculated about all the seats is made into the 2nd evaluation index value. In this case, the larger the transportation revenue, the better for the railway operator. Therefore, the solution candidate with the largest transportation revenue may be selected as the best solution.

  Further, the number of reserved seats may be used as the second evaluation index value. Taking reserved seats as an example, the difference between the number of reserved seats demanded seats and the number of reserved seats sold is the number of passengers who wanted to reserve reserved seats, that is, the number of reserved abandonments. Represents. Therefore, for the seat types that can be reserved, the number of reserved seats is calculated by subtracting the number of seats for sale from the number of demanded seats to obtain the second evaluation index value. In this case, the smaller the number of reservation abandons, the better for the railway operator and the passenger. Therefore, the solution candidate having the smallest number of reservation abductions may be selected as the best solution.

(C) Best solution selection method based on evaluation index value In addition, the best solution is selected using a combination of a plurality of types of evaluation index values instead of selecting the best solution using one type of evaluation index value. It is also possible. For example, regarding the seat type solution candidates, all of the above-described three types of second evaluation index values are calculated for each of these seat type solution candidates. Then, the solution candidates are comprehensively evaluated using the three types of second evaluation index values, and the solution candidate having the best evaluation result is selected as the best solution.

  Further, the best solution may be selected by giving priority to the second evaluation index value. For example, the priority order of the second evaluation index value is determined in advance, such as (1st) boarding efficiency, (2nd) transportation revenue, (3rd) reservation abandonment number. Then, the second evaluation index value having a higher priority is preferentially used to select the best solution. However, when the second evaluation index values of a plurality of solution candidates are the same, the lower second evaluation index value is selected. Using the index value, the best solution is selected by determining the superiority or inferiority of the solution candidates.

  Furthermore, it is possible to calculate a comprehensive evaluation index value by combining two or more second evaluation index values, and to select the best solution based on the comprehensive evaluation index value. For example, a weighted average calculation is performed on two types of second evaluation index values of boarding efficiency and transportation revenue to calculate a comprehensive evaluation index value, and a solution candidate having the maximum comprehensive evaluation index value is selected as the best solution. May be.

  Of course, the selection of the best solution using the plurality of evaluation index values is the same for the solution candidates for the number of formation vehicles using the first evaluation index value.

(D) Setting the number of seats to be supplied In the above-described embodiment, the number of seats to be supplied as a solution candidate has been described as being set at random. The number of seats to be supplied may be set based on the number of seats. Specifically, the number of seats in demand for each section is calculated from the seat candidate demand data obtained by performing seat type demand calculation. That is, the number of seat-type demand seats in the section unit is calculated from the number of seat-type demand seats in the OD unit. And about each section, the demand seat number of the said section is set as a supply seat number.

  Further, instead of setting the demand seat number as the supply seat number, the sales seat number may be set as the supply seat number. That is, the seat type sales seat number candidate data for each section is generated from the seat type sales seat number candidate data for each OD unit. That is, the number of seats sold by OD unit is calculated. For each section, the number of seats sold in that section is set as the number of seats supplied.

(E) Best Solution Search Method In the above-described embodiment, the update process is repeatedly performed on a plurality of solution candidates (L number of trained vehicle solution candidates and N seat type solution candidates), and the best evaluation is performed. It was described that the best solution was searched by selecting the resulting solution candidates. This best solution searching method is merely an example, and the setting can be changed as appropriate.

  For example, regarding the seat type solution candidate, N = 1 may be set, and only one seat type solution candidate may be set to search for the best solution. At this time, among the values included in the seat type solution candidates, the method for selecting the value to be changed and the method for determining the amount to be changed may be as follows, for example. That is, when the boarding efficiency is set as the first evaluation index value, the number of seats supplied by the seat type, the seat type fee, and the seat type sales coefficient are changed in the direction of increasing the boarding efficiency (for example, increased by a predetermined value). ) Method.

(F) Setting of Solution Candidates In the above-described embodiment, the solution candidates for the number of trained vehicles are four types of candidates, that is, the number of trained vehicles, the number of seat type supply seats, the seat type charge, and the seat type sales coefficient. It has been described that the solution candidate for the seat type is set to include three types of candidate data of the seat type supply seat number, the seat type charge, and the seat type sales coefficient. May be included as a solution candidate or may be set by replacement, and the setting can be changed as appropriate.

  In the above-described embodiment, the seat type supply seat number candidate data is set as candidate data for each section, and the seat type charge candidate data and the seat type sales coefficient candidate data are set as candidate data for each OD between stations. However, the format of the candidate data can also be arbitrarily determined. For example, all candidate data may be set as candidate data for each section, or all candidate data may be set as candidate data for each OD.

(G) Setting of seat type sales coefficient In the above-described embodiment, for each seat type sales system number candidate data of solution candidates (solution candidates for the number of trained vehicles and solution candidates for seat type), “0.0 to The description has been made assuming that the value is set at random within the range of 1.0 ”. This is because there are various passengers depending on the boarding section, so if the sales coefficient is set to “1.0” for all ODs, seats may not be allocated to all passengers depending on the boarding section. is there.

  However, for example, for seat type sales coefficient candidate data of 20% of a plurality of solution candidates, the sales coefficient is set to “1.0” for all ODs, and the remaining 80% of solution candidates For the seat type sales coefficient candidate data, the sales coefficient is set to a value in the range of “0.0 to 1.0” at random, and calculation is performed, for example, so that the sales type is “1.0”. Adjustment may be performed so that the probability that the sales coefficient candidate data is selected as the best solution is high.

(H) Calculation of the best solution In the above-described embodiment, all Shinkansen trains operating on a certain day are targeted, and a plurality of solution candidates are set for each of these Shinkansen trains to calculate the best solution. However, the best solution may be calculated by setting solution candidates collectively for a plurality of Shinkansen trains instead of one train.

Specifically, all the Shinkansen trains operating on a certain day are divided into a plurality of train groups composed of one or a plurality of Shinkansen trains. For example, 15 trains are divided into three train groups: a first train group consisting of four trains, a second train group consisting of five trains, and a third train group consisting of six trains. As a way of dividing the train group, it may be divided according to time zones, or may be classified in consideration of vehicle operation such as whether or not the same vehicle operation.
And the preparation process (step A1-A11 of FIG. 3, FIG. 14) of the formation vehicle number plan is performed with respect to each of these train groups. In other words, the process for creating a train schedule is performed a number of times equal to the number of train groups.

  In this case, one train candidate for the number of trains (one solution candidate data for trains 530 in FIG. 4) is generated for each train group. The number of trains for trains included in this solution for train number of trains (number of trains for trains 532) may be set differently for each train included in the corresponding train group, or for all trains in the train group. You may set as common (same). In the former case, in one set candidate number solution data 530, a combination of the number of trains for each train included in the corresponding train group (the train number candidate data 532) is set. Even included trains can have different numbers of trains. In the latter case, the number of trains included in the same train group is the same.

  Moreover, when evaluating the solution candidate for vehicle organization by using a plurality of Shinkansen trains as one train group, the following may be performed. That is, for each train included in the train group corresponding to the vehicle composition solution candidate, the "supply-residual residual square value", which is the square of the difference between the number of demand seats and the number of supply seats in each section, is calculated. Calculate the sum of the squares of supply and demand residuals, which is the sum of all sections. Next, a “total sum of supply and demand residual squares” that is the sum of the “sum of supply and demand residual squares” of each train included in the corresponding train group is calculated, and this is calculated as the “first evaluation index for the vehicle composition solution candidate”. Value ". By doing in this way, the solution candidate for vehicle organization with respect to a train group can be evaluated.

(I) Creation of Seat Allocation Plan In the second embodiment described above, a seat allocation plan creation device based on the number of organized vehicles / seat type setting plan created by the number of organized vehicles / seat type setting plan creation device 1 2 has been described as creating a seat allocation plan, but the number of trained vehicles / seat type setting plan creation device 1 may create a seat allocation plan, and the seat allocation plan creation device 2 may be omitted.

(J) Guest Room Electronic Display Board In the second embodiment described above, a description has been given assuming that a mobile guest room electronic display board that can move in the front-rear direction is provided in the guest room of the Shinkansen train. However, some high-quality trains currently operating in Japan have a non-mobile electronic bulletin board at the forefront of each guest room. Therefore, information such as seat assignment and expected congestion rate may be displayed on the non-movable cabin electronic display board without providing the mobile cabin electronic display board.

(K) Unreserved Seat Guidance In the second embodiment described above, it has been described that information regarding seat assignment in each OD included in the passenger riding section is printed on the ticket as the seat guidance. Only information related to seat allocation in the section may be printed. For example, the non-reserved seat limited express ticket in FIG. 17 does not print the free seat guide for all sections of “A station → B station”, “B station → C station”, and “A station → C station”. Only the free seat guidance for the section “A station → C station”, which is the boarding section, may be printed.

DESCRIPTION OF SYMBOLS 1 Formation vehicle number / seat type setting plan creation apparatus 10 Control part, 20 Input part, 30 Display part, 40 Communication part 50 Storage part 510 Number of formation vehicle / seat type setting plan creation program 511 Formation vehicle number plan creation program 512 Seat type Setting plan creation program 521 User demand OD data, 522 Seat type selection model data 523 Other data 530 Solution candidate data for number of trained vehicles (plan candidate)
550 seat type solution candidate data (seat type plan candidate)
541,561 Seat type selection probability candidate data 542,562 Seat type demand seat number candidate data 543,563 Seat type sales seat number candidate data 544,564 Best solution data 545 Trained vehicle number plan data, 565 Seat type setting plan data 611 Designation Seat selection probability candidate data, 612 Designated seat demand candidate data 613 Designated seat sales seat number candidate data, 614 Unoccupied seat demand candidate data 100 Shinkansen operation system 3 Number of trains and seat type setting plan creation system 2 Seat allocation plan creation device 5 Boarding Ticket sales system 51 Server, 53 Ticket vending machine 7 Seat allocation notification system 71 Ground communication control device, 73 Vehicle operation DB
9 Shinkansen vehicles 91 Seat type display system, 93 seat type individual guidance device, 95 control device

Claims (4)

Computer system,
A plan for generating a plan candidate including at least the number of trains of the target train to be planned for the number of trains, the number of seats supplied for each seat between the stations where the target train stops, and the seat fee for each seat between the stations A candidate generation step;
Using the seat type selection model that models the probability of the seat type selected by the passenger using the boarding time and the seat type charge, the boarding time between each station of the target train, and the plan candidate, the plan candidate A first seat type selection probability calculating step for calculating a seat type selection probability;
Estimating the number of seat type users between the stations of the target train and estimating the number of seat type users between the stations of the target train using the predetermined user demand between the stations of the target train and the seat type selection probability. Steps,
A first evaluation step of evaluating the plan candidate using the seat type supply seat number and the seat type user number;
A program for executing the arbitrary number of the planned candidates for target train generated by the candidate generation step, said first seat type selection probability calculation step, the first seat type user the number estimating step, the first evaluation is performed and the step, the program for determining the organization number of vehicles of the target train based on the planned candidate was the best evaluation result. The first evaluation step is a step of evaluating the plan candidate by using a difference between the seat type supply seat number and the seat type user number between adjacent stop stations of the target train.
The program according to claim 1. Computer system,
The number of seats supplied by each seat between the stations where the target train stops set based on the number of trains determined by executing the program according to claim 1 or 2, and the seat type charge between the stations A seat type plan candidate generating step for generating a seat type plan candidate including at least
A second seat type selection probability calculating step for calculating a seat type selection probability for the seat type plan candidate using the seat type selection model, a boarding time between stations of the target train, and the seat type plan candidate. When,
Using the predetermined user demand between the stations of the target train and the seat type selection probability calculated in the second seat type selection probability calculation step, the seat type between the stations for the seat type plan candidate A second seat type user number estimating step for estimating the number of users;
1) Seat type utilization rate obtained from the seat type supply seat number included in the seat type plan candidate and the seat type user number estimated in the second seat type user number estimation step, 2) the second The seat type plan candidate using at least one of the number of seat type users estimated in the seat type user number estimation step and the charge revenue obtained from the seat type charge candidate included in the seat type plan candidate A second evaluation step for evaluating
A program for executing the arbitrary number of the seat species plan candidates for target train generated in the seat species plan candidate generation step, and the second seat type selection probability calculation step, the second A program for causing a seat type setting plan for the target train to be executed on the basis of a seat type plan candidate having the best evaluation result by executing the seat type user number estimating step and the second evaluation step. A plan for generating a plan candidate including at least the number of trains of the target train to be planned for the number of trains, the number of seats supplied for each seat between the stations where the target train stops, and the seat fee for each seat between the stations Candidate generation means;
Using the seat type selection model that models the probability of the seat type selected by the passenger using the boarding time and the seat type charge, the boarding time between each station of the target train, and the plan candidate, the plan candidate First seat type selection probability calculating means for calculating a seat type selection probability;
Using the predetermined user demand between the stations of the target train and the seat type selection probability,
First seat type user number estimating means for estimating the number of seat type users between the stations of the target train;
First evaluation means for evaluating the plan candidate using the seat type supply seat number and the seat type user number;
The candidate generation means generates an arbitrary number of the plan candidates related to the target train, and for each plan candidate, the first seat type selection probability calculation means, the first seat type user number estimation means, A train number determination system that determines the number of trains of the target train based on a plan candidate that is executed by the first evaluation unit and has the best evaluation result.