JP6375588B2 - Intelligent train scheduling method - Google Patents

Description

  The present invention relates to an intelligent train scheduling method.

  With the development of train networks such as railways and subways, and the development of bus networks linked therewith, research on methods that can efficiently manage the schedule of each means of transportation has been actively conducted. For example, a product related to software or hardware for scheduling such means of transportation is being developed.

  However, when looking at conventional scheduling methods for transportation, the train schedule and the vehicle operation plan are configured as separate programs. Some schemes operate to create train schedules around the user experience, while other programs create train schedules through automated algorithms.

  In addition, the conventional train scheduling program cannot predict the passenger capacity and the congestion rate. That is, regarding the train schedule, it has not been possible to provide information on how many passengers board each section, time zone, or train.

  On the other hand, in this regard, Korean Published Patent No. 10-2012-0129344 (Title of Invention: Train Operation Plan Establishing Method and Device) inputs data relating to train schedule restriction conditions relating to the train operation plan establishing method and device. Thus, a component that can increase or decrease the allowable range of forwarding is disclosed, such as the schedule planning unit of the subject invention that plans an execution schedule in addition to the basic schedule of a train.

  The present invention provides an intelligent train scheduling method that provides prediction information related to the amount of traffic for each section of a train and for each train in the train schedule, and adjusts the train schedule based on the prediction information.

  As a technical means for achieving the technical problem described above, the train scheduling method of the intelligent train scheduling system according to one aspect of the present invention includes a train schedule including train route information and information on the operation time of each train. A step in which information is provided, a station where the train included in the train schedule information stops, as a node, a route between each station included in the train schedule information as an arc, and included in the train schedule information Generating a time extension network including operation time information for each train, and calculating prediction information related to traffic for each section and for each train on the train schedule based on the time extension network.

  According to the above-described embodiment of the present invention, it is possible to calculate prediction information related to the traffic volume for each section and for each time zone with respect to the already generated train schedule. Therefore, the user who is in charge of creating the train schedule can adjust the train schedule in consideration of the prediction result on how many passengers board the train schedule that he / she has established in which section and time zone. .

1 shows an intelligent train scheduling system according to an embodiment of the present invention. It is a flowchart which shows the calculation process of traffic prediction information which concerns on one Example of this invention. It is a figure which shows the example of a normal traffic network. It is a figure which shows the example of the time expansion network applied to this invention. The concept of the division | segmentation of the time expansion node type based on one Example of this invention is shown. The concept of the division | segmentation of the time expansion arc type type | mold which concerns on one Example of this invention is shown. It is a figure for demonstrating the concept of the discretization of the demand which concerns on one Example of this invention. It is a flowchart which shows the traffic prediction information calculation method which concerns on one Example of this invention. It is a figure for demonstrating the traffic amount prediction information calculation method which concerns on one Example of this invention. It is a figure for demonstrating the concept of the traffic calculation which considered the priority between the paths based on one Example of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that a person having ordinary knowledge in the technical field to which the present invention belongs can be easily implemented. However, the present invention can be embodied in various different forms and is not limited to the embodiments described herein. In the drawings, in order to clearly describe the present invention, portions not related to the description are omitted, and similar portions are denoted by similar reference numerals throughout the specification.

  In the entire specification, when a part is “connected” to another part, this is not only “directly connected” but also “electrically” via another element in the middle. This includes cases where they are connected. In addition, when a part “comprises” a component, this means that the component may further include other components rather than excluding other components unless specifically stated to the contrary.

  FIG. 1 shows an intelligent train scheduling system according to an embodiment of the present invention.

  The intelligent train scheduling system 10 includes a demand analysis unit 110, a train schedule generation unit 120, a traffic allocation unit 130, and a database 140.

  The demand analysis unit 110 collects statistically collected user traffic performance data, and predicts demand for each route including origin-destination pairs. For example, information such as a user's ticket purchase record or traffic card usage is received from the database 140, and based on this information, a demand for each route is predicted. The demand analysis unit 110 calculates the transfer traffic between routes in consideration of the transfer time between routes and the train time interval (headway), and considers the calculated transfer traffic in total. Calculate traffic volume. At this time, the amount of traffic under direct transfer (transfer from the route to another route) and directly under transfer (transfer from the other route to the route) is converted into the traffic amount of the route through the shortest route calculation method. The demand for the route from the departure point to the destination of the route can be predicted in consideration of the operation interval between trains on the route and the time required for transfer at the transfer station.

  The train schedule generation unit 120 generates a basic schedule and an execution schedule for a train. First, the train schedule is a plan table for setting the arrival time and departure time for each station of an individual train that operates on a railway line. In addition, the basic schedule is a train schedule that reflects the experience and subjectivity of the user, and is a train schedule in the form of a draft that can contain train competition. The execution schedule is a train schedule in which all the conflicts of trains included in the train schedule are resolved (conflict free), and is a train schedule applicable in reality.

  In order to generate a basic schedule, a basic schedule heuristic algorithm can be applied to generate a draft, which can be modified according to the user's intention by the train timetable editing function. On the other hand, since a specific method for generating a train schedule is beyond the scope of the present invention, a detailed description thereof will be omitted.

  The traffic allocation unit 130 calculates prediction information related to the traffic volume for each section and for each time zone with respect to the train schedule. At this time, the train schedule may be generated by the train schedule generation unit 120 or may be received from the outside. The calculation process of traffic volume prediction information by the traffic allocator 130 will be described later.

  The database 140 includes basic data for demand analysis in the demand analysis unit 110, basic data for train schedule generation in the train schedule generation unit 120, and basic data for calculation of traffic volume prediction information in the traffic allocation unit 130. And manage each. The calculation data of the demand analysis unit 110, the train schedule generation unit 120, and the traffic allocation unit 130 are received and managed.

  For example, it collects and manages ticket purchase information or traffic guard use record information. It also manages information about the entire route such as track sections, stations, railway gradients, railway curves, speed limit sections, and depots. In addition, it manages information related to train operation, such as conventional schedule information, standard operation diagrams, operation intervals, operation hours, and stop times. It also manages information about the vehicle such as the vehicle type, knitting state, creation, and the like. Furthermore, the route of the departure point-destination pair between stations, the traffic demand information for each route, and the like are managed.

  Meanwhile, the components shown in FIG. 1 according to the embodiment of the present invention mean software components or hardware components such as FPGA (Field Programmable Gate Array) or ASIC (Application Specific Integrated Circuit). Fulfill.

  However, “components” are not meant to be limited to software or hardware, and each component may be configured to reside on a storage medium that can be addressed to play one or more processors. It may be configured as follows.

  Thus, by way of example, components include software components, object-oriented software components, class components and task components such as processes, functions, attributes, procedures, subroutines, program code segments, drivers, Includes firmware, microcode, circuits, data, databases, data structures, tables, arrays and variables.

  A component and the functionality provided within that component can be combined into fewer components or further separated into additional components.

  FIG. 2 is a flowchart illustrating a process of calculating traffic amount prediction information according to an embodiment of the present invention.

  First, in the present invention, train schedule information already generated for traffic volume prediction information is used (S210). Here, the train schedule information includes train route information and information related to operation time for each train. The train schedule information can be a basic schedule or an execution schedule.

  Next, a time extension network is generated based on the train schedule information (S220).

  The time expansion network includes the station where the train included in the train schedule information stops as a node, includes the route between each station included in the train schedule information as an arc, and the operation time for each train included in the train schedule information Contains information. At this time, the repetition number n is set to 0, and an initial value is set for the cost of each arc.

  FIG. 3 is a diagram showing an example of a normal traffic network, and FIG. 4 is a diagram showing an example of a time extension network applied to the present invention.

  In the case of FIG. 3, a node indicating a station and a link indicating a route connecting the stations are included, but the concept of time is not included.

  In the case of FIG. 4, the traffic network is expanded by adding a time dimension. In the case of such a time expansion network, it is possible to accurately grasp the waiting time of the passenger, the train that arrives first, and the like, unlike the conventional traffic allocation system based on the number of times of entry. However, since all stations must be expressed differently according to the time of the schedule, the size of the network becomes very large and complicated, thereby deriving a solution within a realistic time. Can be difficult.

  Such a time extension network (VXE) includes a time extension node set V and a time extension arc set E.

  FIG. 5 shows the concept of time extension node type division according to an embodiment of the present invention, and FIG. 6 shows the concept of time extension arc type division according to an embodiment of the present invention.

  The time expansion node set V includes attributes of a node number, a train line number, a station number, and node time information, and is classified into types such as station arrival and departure, demand generation, and demand end. The time extension arc set E includes attributes of arc number, arc start node number, arc end node number, time information between the last and first of the arc, movement between stations, stop, transfer, demand generation and It is divided into types such as end. Here, the demand generation node and the demand end node correspond to virtual nodes for applying a demand value to each node, not a node representing an actual physical station.

  FIG. 7 is a diagram for explaining the concept of discretization of demand according to an embodiment of the present invention.

  The graph shown in the lower part of the drawing is a graph showing demand generation statistics according to time for each departure point-destination pair. Such demand generation statistical data is based on the above-described information such as the user's ticket purchase results or traffic card usage. It can be confirmed that the demand changes with the passage of time for each departure point-destination pair. For example, demand is relatively concentrated in work hours (07: 0 to 09:00) and work hours (18: 0 to 20:00). Thus, in order to discretize the state in which the generation of demand continuously changes, a virtual demand generation node is considered. That is, referring to the demand generation statistics by each departure point-destination pair, virtual demand generation nodes 300 and 310 are generated for each route, and the demand amount at the time is set for the node. For example, the demand amount of the demand generation node 300 is determined by adding the demand amount of each departure point-destination pair at a specific time (07: 0 to 08:00), and this value is the boarding amount of the route. To be considered as. That is, a time interval between demand generation nodes and a demand generation amount are set based on statistically collected user traffic data. With such a configuration, the time interval between the demand generation nodes is reduced in a time zone where the demand is concentrated, and the time interval between the demand generation nodes is increased in a time zone where the demand is relatively low.

  The boarding amount and the degree of congestion for each section of the individual train represented by the time expansion network are optimally solved through the schedule-based traffic assignment algorithm presented below for the schedule-based transit assignment model. And an approximate optimal solution is calculated.

  Next, prediction information regarding the traffic volume for each section and time zone for the train schedule is calculated based on the time extension network (S230).

  FIG. 8 is a flowchart showing a traffic amount prediction information calculation method according to an embodiment of the present invention.

  First, an initial traffic allocation step is performed in which a demand amount for a route defined by a departure point-destination pair is set as a traffic amount of each route (S231). Calculate the initial cost of each arc e in the time extension network, and use Equation 1 for that.

  At the same time, the shortest route is searched for each departure point-destination pair, and initial traffic allocation is performed for each shortest route. At this time, as shown in the following mathematical formula 2, the initial traffic allocation is performed in a form in which the demand amount is set as the initial traffic amount for each shortest route.

Here, d _i indicates the demand amount, p _i * indicates the shortest route of the departure point-destination pair, and f _p indicates the traffic amount of the route p. In addition, I is a set of all starting point-destination pairs i, and the demand generation nodes are arranged in order of speed. At this time, the demand amount set in the above-described demand generation node can also be considered.

  Next, the arc cost is updated based on the type and traffic of each arc constituting the route (S232).

Here, the arc cost function is defined as c _e = c _e (x _e ), and the arc cost calculation may vary depending on the arc type as shown in Table 1 below.

  For example, as shown in the table, since the arc cost varies depending on the type of each arc, the arc cost is set to be different depending on the type of each arc defined in the time extension network. For example, when the amount of traffic passing through the arc “e” is 500 or less or 1500 or less, the cost of multiplying the time passing through the arc by the weighted value “1.00” is calculated and exceeds 1500 In this case, the cost for multiplying the weighted value “1.09” is calculated. When the type of the arc “e” is transfer, the arc cost is calculated by adding a constant value corresponding to the transfer type. Further, the cost for the demand generation node is calculated by multiplying the required time by the weighted value “0.1”.

  In this way, the user's traffic volume for each route is different, which may change the arc cost.

  Next, based on the updated arc cost, a shortest route that replaces the route defined by each starting point-destination pair is searched (S233), and traffic amount prediction information for each route is calculated by changing the route to the shortest route. (S234 to S236).

  FIG. 9 is a diagram for explaining a traffic amount prediction information calculation method according to an embodiment of the present invention.

  First, assume that there are a total of four routes that satisfy each origin-destination pair. For example, a first path (Path 1) that satisfies (B3 → A7), a second path (Path 2) that satisfies (A1 → C6), and a third path (Path 3) that satisfies (C1 → A4) ) And a fourth path (Path 4) that satisfies (B1 → D2). At this time, the shortest route for replacing the conventional route is searched, and for example, a new shortest route (New Path) for replacing the fourth route can be searched.

  In order to calculate the traffic amount prediction information due to the route change to the shortest route, the expandable traffic amount of each arc included in the shortest route is calculated (S234). For this purpose, the traffic amount of the route passing through the arc is subtracted from the maximum traffic amount of each arc, and the traffic amount of the route passing through the arc and having a lower priority than the shortest route is added up to be calculated. This is defined by Equation 3 below.

u _e represents the maximum possible traffic volume of each arc, and x _e represents the traffic volume of the arc (the sum of the traffic volumes of all routes including the arc e). Here, q (e) <p means that the priority of the route q passing through the arc e is lower than that of the route p. Thus, in the present invention, the traffic amount is predicted in consideration of the priority order between routes.

  FIG. 10 is a diagram for explaining the concept of traffic volume calculation in consideration of priority between routes according to an embodiment of the present invention.

  In the present invention, when a route passing through different routes passes through one route, priority is given to the route that enters the route first. For example, the route p has priority since it entered the route (Line 2) earlier than the route q.

  The traffic volume prediction will be described with reference to FIG. 9 again.

  The expandable traffic amount of each arc constituting the new shortest path (New Path) is as follows. The maximum possible traffic for each arc is 10 and the same. Further, the traffic amount of each arc is calculated based on the traffic amount shown in the drawing.

u _{B1 → B2} = 10−3 + 0 = 7, u _{B2 → B3} = 10−3 + 0 = 7

u _{B3 → B4} = 10−10 + 7 = 7, u _{C3 → C4} = 10−8 + 0 = 2

u _{C4 → C5} = 10−8 + 0 = 2, u _{C5 → C6} = 10−9 + 9 = 10

u _{D1 → D2} = 10-0 + 0 = 10

  For example, the expandable traffic volume between B3 and B4 is obtained by subtracting the traffic volume (3 + 7) of the arc from the maximum possible traffic volume 10 and the first path (Path 1 having a lower priority than the new shortest path) ) Traffic volume (7). Further, the expandable traffic volume between C5 and C6 is the second path (Path 2) having a lower priority than the new shortest path, obtained by subtracting the traffic volume (9) of the arc from the maximum possible traffic volume 10. ) Traffic volume (9).

  The minimum value is 2 among the expandable traffic volume of each arc thus calculated. That is, considering the minimum value of the expandable traffic amount of each arc constituting the new shortest route, the increase amount of the traffic amount due to the addition of the shortest route is calculated (S235). For example, it can be calculated by mathematical formula 4.

At this time, n represents the number of repetitions. In the (n-1) -th iteration, the amount obtained by dividing the minimum value of the expandable traffic amount by the number of iterations is added to the traffic amount of p _i * and added up.

  Next, the traffic volume of the route defined by the departure point-destination pair is adjusted by subtracting the amount of increase in traffic volume from the route with the higher route cost among the routes defined by the departure point-destination pair. (S236). For example, the traffic amount of the shortest route (New Path) is subtracted from the existing route (Path 4) having the same starting point and destination.

  On the other hand, if there is an excess traffic volume that exceeds the maximum traffic volume of each arc for all the arcs included in the shortest path, the path cost of the path that passes through the arc until the excess traffic volume of the arc becomes 0 Further, the step of reducing the traffic amount from the route with the largest is performed. For example, since the traffic amount of the second route (Path 2) is 9, and the traffic amount of the new shortest route is 2, the traffic amount of the arc (C5 → C6) is 11, and the maximum possible traffic amount 10 is It will exceed. Therefore, it is necessary to reduce the traffic amount by 1 from the arc (C5 → C6). Therefore, in order to correct this, the traffic amount of the second route (Path 2) is corrected to 8.

  Next, the step of updating the arc cost until the traffic amount prediction information converges to a preset level, the step of searching for the shortest route, and the step of calculating the traffic amount prediction information are repeated. At this time, n increases by 1 for each repetition.

  For example, convergence can be tested using Equation 5.

  At this time, if the calculated value is equal to or less than the threshold value, the repeating step is interrupted.

  In addition, the boarding amount and the degree of congestion for each section of the train calculated from the results of the timetable-based traffic allocation model are applied to the train schedule that has already been established in the train schedule planning department, and the traffic is displayed by emphasis processing such as color display. The amount can be displayed. For example, a section with a large traffic volume and a high congestion rate is displayed in red, and a section with a relatively low congestion rate is displayed in green. Thereby, train schedule adjustment etc. will be performed with respect to a section with a high congestion rate.

  The embodiment of the present invention may be embodied in the form of a recording medium including an instruction word executable by a computer such as a program module executed by the computer. Computer-readable media can be any available media that can be accessed by the computer and includes all volatile and non-volatile media, separated and non-separated media. In addition, computer readable media may include all computer storage media and communication media. The computer storage medium is volatile and non-volatile, separated and non-separated embodied in any method or technique for storing information such as computer readable instructions, data structures, program modules or other data. Including all media. Communication media typically include computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism and include any information delivery media.

  Although the method and system of the present invention have been described in connection with specific embodiments, some or all of those components or operations may be implemented using a computer system having a general purpose hardware architecture.

  The above description of the present invention is given for the purpose of illustration, and those who have ordinary knowledge in the technical field to which the present invention pertains do not change the technical idea or essential features of the present invention. It should be understood that it is easily deformable in a typical form. Therefore, it should be understood that the above-described embodiments are illustrative in all aspects and not limiting. For example, each component described in a single type can be implemented in a distributed manner, and similarly, components described as being distributed can also be implemented in a combined form.

  The scope of the present invention is defined by the following claims rather than the above detailed description, and all modifications or variations derived from the meaning and scope of the claims and equivalent concepts thereof are described in the present invention. It should be construed as being included in the scope of the invention.

Claims (9)

In the train scheduling method of the intelligent train scheduling system,
Train schedule information including train route information and information on train-specific operation times is provided;
The station included in the train schedule information includes a station where the train stops, the route between the stations included in the train schedule information is included as a link , and the operation time information for each train included in the train schedule information. Generating a time extension network including:
Calculating prediction information on traffic for each link and train for the train schedule information based on the time extension network ,
The step of calculating the prediction information related to the traffic volume includes
An initial traffic allocating step for setting a demand amount for a route defined by a departure point-destination pair as a traffic amount of each route;
Updating the link cost based on the type and traffic volume of each link constituting the route;
Searching for a shortest route to replace a route defined in the origin-destination pair based on the updated link cost;
Calculating the traffic volume prediction information of each route by the route change to the shortest route,
The step of calculating the traffic amount prediction information of each route by the route change to the shortest route,
Calculating an expandable traffic volume of each link included in the shortest path;
Calculating an increase in traffic volume due to a route change of the shortest path from a minimum value among the expandable traffic volumes;
By subtracting the amount of increase in traffic from the route defined by the departure point-destination pair and having a higher route cost, the traffic amount of the route defined by the departure point-destination pair is adjusted. And an intelligent train scheduling method. Further comprising displaying prediction information regarding the traffic volume on a user interface for displaying the train schedule information,
The train scheduling method according to claim 1, wherein a link with a relatively large amount of traffic is highlighted and displayed. The time expansion network includes a time expansion node set divided into a station arrival node, a station departure node, a demand generation node, or a demand end node, and a time divided into movement, stop, transfer, demand generation or demand end between stations. The train scheduling method according to claim 1, comprising an extended link set. The initial traffic allocation step includes:
4. The demand generation node is set based on statistically collected user traffic performance data, and the time interval and the demand generation amount between the demand generation nodes are set . The train scheduling method according to one item . Calculating the expandable traffic volume,
The traffic volume of route via the link from the maximum traffic amount of each link is subtracted, via the link, and is also calculated by summing the traffic amount of lower priority paths from the shortest path, claim 1 The train scheduling method according to claim 1 . The train scheduling method according to claim 5 , wherein the low priority route is a route that has entered a route that passes through the link later than the shortest route. If there is excess traffic amount exceeds the maximum traffic amount of each link for all links included in the shortest path, the path cost of the route the excess traffic amount of the link via the link until the zero The train scheduling method according to any one of claims 1 to 6 , further comprising a step of reducing a traffic amount from a maximum route. Updating the link cost until the traffic volume prediction information converges to a preset level, searching for the shortest route, and calculating the traffic volume prediction information are further performed. The train scheduling method according to any one of claims 1 to 7, further comprising: The computer program which performs each step of the method as described in any one of Claims 1-8 .

Images