JP7449192B2 - Timetable information management system, timetable information management method, and operation guidance system - Google Patents

Description

The present invention relates to a technology for providing operation information, that is, timetable information, according to actual operating conditions in transportation systems, particularly public transportation systems such as trains and buses.

In transportation, unfavorable weather or accidents can cause so-called ``schedule disruptions and service changes,'' such as early arrivals and departures, delays, and service cancellations, as well as changes in destinations and train types. In such cases, current general route guidance services generate and provide operation information in conjunction with a pre-planned timetable. Therefore, the user may obtain operation information that differs from the actual operation situation. Therefore, the user is required to receive guidance linked to the predicted timetable based on the actual operating conditions.

For example, Patent Document 1 states that if a user is unable to use the originally planned route or time due to a delay in public transportation, and if the delay threshold is exceeded, the stopover station of each line and the A technology has been disclosed that refers to a scheduled timetable database in which scheduled arrival/departure times are registered, and re-searches and redistributes routes to the relevant person.

Japanese Patent Application Publication No. 2007-199900

In the technology described in Patent Document 1, the delay threshold is defined by the delay time of each station, but if a delay or service suspension actually occurs, the cause of the delay, the location of the problem, the weather, and the delay recovery rules for each operator are defined. Various factors come into play. Additionally, delays can increase, decrease, or stall, making it difficult to reliably predict future arrival and departure times. For this reason, if only a state value at one time, ie, delay time, is used as a threshold value, there is a risk of giving incorrect guidance. Furthermore, there is no guarantee that the delay time can be obtained immediately and accurately, and since the timing and accuracy at which delay time can be obtained differs depending on the operator, it is necessary to define threshold values individually for each operator, which requires operational costs. As described above, in the conventional technology including Patent Document 1, there were many cases where the scheduled arrival and departure times could not be reliably predicted due to unstable operating conditions.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to make it possible to generate timetable information according to the operating conditions even when the operating conditions of transportation facilities are unstable.

One aspect of the present invention for achieving the above object is a timetable information management system that generates dynamic timetable information indicating the operation status when the operation schedule of the transportation facility is changed, the system comprising: an input means that accepts input of operation data; and based on the input operation data, performs prediction processing on the operation status for each predetermined unit constituting the transportation facility, and indicates the degree of influence of the change in the operation schedule. a predicted value generation means for generating a predicted value; a prediction error calculation means for calculating a prediction error regarding the accuracy of the prediction process ; a predicted value threshold storage means for storing a predetermined predicted value threshold; disclosure determining means for comparing the predicted value thresholds and identifying the accuracy of the prediction error as the accuracy of the prediction process; and generating a plurality of prediction timetables for each of the predetermined units including the specified predicted values. and a dynamic timetable generating means that generates a dynamic timetable that summarizes operating times of the transportation facility as the dynamic timetable information, the dynamic timetable generating means , a timetable information management system that generates the dynamic timetable using a predicted timetable specified according to accuracy specified by the disclosure determination means according to predetermined conditions.

The present invention also includes a timetable information management method using a timetable information management system and a computer program for executing the method. Furthermore, the present invention also includes a travel guidance system that has a timetable information management system and that creates a dynamic timetable and outputs guidance information in accordance with the creation of a dynamic timetable, a method using the same, and a computer program therefor.

Even when operating conditions are unstable, timetable information can be generated according to the situation, and guidance can be provided with appropriate quality. Other objects and novel features will become apparent from the description of this specification and the accompanying drawings.

1 is a system configuration diagram of an entire embodiment of the present invention; FIG. 1 is a hardware configuration diagram of a timetable management system 10 in an embodiment of the present invention. FIG. 2 is a data configuration diagram of a predicted timetable 107 used in an embodiment of the present invention. It is a flowchart of the process performed by the dynamic timetable generation part 103 in one Example of this invention. FIG. 3 is a diagram for explaining a method of calculating a prediction error. 3 is a diagram illustrating an example data structure of a disclosure threshold 108. FIG. This is an example of the data structure of the disclosure determination result 109. 10 is a flowchart of a process for creating a disclosure determination result 109. 3 is a flowchart of a process for generating a dynamic timetable 110. FIG. This is an example of a data structure of a dynamic timetable 110. This is an example of an administrator screen of the service systems 30a and 30b. This is an example of a guide screen of user terminals 40a to 40d. 3 is a flowchart of processing executed by the predicted timetable generation unit 101. FIG.

Hereinafter, one embodiment of the present invention will be described based on the drawings. In this embodiment, a railway will be used as an example of public transportation. Therefore, in this embodiment, when a timetable disruption such as a delay occurs in train operation, a recommended route is presented to the user (passenger) etc. according to the operating situation. Here, the term "disruption of the timetable" refers to an operating situation in which some kind of change has occurred in the scheduled operating schedule (for example, a static timetable, which will be described later). In addition to railways, public transportation includes route buses (including express buses and shared buses), ships, and airline routes.

<Figure 1: Overall system configuration diagram>
First, FIG. 1 shows an example of a system configuration diagram of the entire embodiment for executing the above-described processing. FIG. 1 shows an operation guidance system 1 that creates recommended routes, and user terminals 40a to 40d that output recommended routes via communication networks 53a and 53b. Here, the operation guidance system 1 includes a timetable management system 10, dynamic operation data distribution systems 20a, 20b, and service systems 30a, 30b.

The timetable management system 10 is connected to the dynamic operation data distribution system 20 via a communication network 51 and to the service systems 30a and 30b via a communication network 52. Note that the dynamic operation data distribution systems 20a and 20b are hereinafter referred to as the dynamic operation data distribution system 20. Further, the service systems 30a and 30b will be referred to as a service system 30. Further, the user terminals 40a to 40d are referred to as user terminals 40.

Further, the service system 30 is connected to a user terminal 40 used by a user using the system via communication networks 53a and 53b. Details of this timetable management system 10 will be described later.

Note that the communication networks 51, 52, 53a, and 53b may be a common communication network, or may be networks using different protocols. Furthermore, the communication networks 51, 52, and 53 may be either wired networks or wireless networks. Data is sent and received between each system and terminal via these networks.

The dynamic operation data distribution system 20 is a system that distributes data regarding the latest operation of public transportation such as trains and buses to the outside at any time. The dynamic operation data distribution system 20 is, for example, a system owned or used by a business operator that provides public transportation. The dynamic operation data distribution system 20 distributes operation data, which is data related to operation, every few seconds or minutes, or in response to an event, for a route operated by a corresponding business operator. Here, the above-mentioned event includes the dynamic operation data distribution system 20 detecting a timetable disruption such as a vehicle delay.

This operation data includes, for example, train travel position data. Traveling position data is data indicating where a vehicle is on a route at a certain time. In addition to the location of each vehicle, the traveling position data may also include information regarding timetable disruptions, such as error time (delay time or early arrival/departure time) with respect to the scheduled time. Note that it is assumed that one or more dynamic operation data distribution systems 20 are prepared for each operator. However, the dynamic operation data distribution system 20 may be jointly operated or utilized by a plurality of business operators.

Moreover, the dynamic operation data distribution system 20 can be realized by a so-called computer. Further, although not shown, a terminal device having an input section and a display section may be connected to the dynamic operation data distribution system 20. Further, the dynamic operation data distribution system 20 itself may have an output device or may be connected to an output device.

The service system 30 is, for example, a system owned or used by a service provider that provides a route search service. This service system 30 generates guidance information such as recommended routes and operating times based on route search conditions (for example, departure point, destination, date and time of use, etc.) received from any of the user terminals 40, and presents this information. . It is assumed that one or more service systems 30 are prepared for each business operator. However, the service system 30 may be jointly operated or used by a plurality of businesses.

Although not shown, a terminal device having an input section and a display section may be connected to the service system 30. Further, the service system 30 itself may have an output device or may be connected to an output device.

Generally, public transportation route search services provide guidance based on timetables (hereinafter referred to as "static timetables") that are established several times a year by public transportation operators. generate. In contrast, in this embodiment, as will be described later, the timetable management system 10 uses operation data acquired from the dynamic operation data distribution system 20 to periodically update the timetable (hereinafter referred to as , "dynamic timetables"). This allows the service system 30 to generate guide information based on the dynamic timetable.

Note that the timetable management system 10 is not limited to generating a dynamic timetable, but can also generate dynamic timetable information, which is some information related to operation. Therefore, the timetable management system 10 can also be realized as a timetable information management system.

Here, "dynamic" means changing according to some conditions. Therefore, the dynamic timetable may be generated and distributed aperiodically. Furthermore, generation of the dynamic timetable includes modifying the static timetable using operation data. Furthermore, generation of a dynamic timetable may be executed only when operation data indicates a timetable disruption. In this case, the dynamic timetable to be distributed will be one that was generated previously. Note that the static timetable and the dynamic timetable are not limited to so-called timetables, and it is possible to use timetable information indicating the operation status including flight times.

In this embodiment, the service system 30 is assumed to provide a guidance service in response to a request transmitted from any of the user terminals 40 at an arbitrary timing. However, the guidance information of each user terminal 40 may be updated periodically, periodically, or automatically from the service system 30 at an arbitrary timing.

Further, the user terminal 40 is not limited to a terminal owned or used by a general passenger, but may be a terminal used by a business providing public transportation in passenger guidance services. Here, the user terminal 40 may include, for example, a mobile phone (including a so-called smartphone), a personal digital assistant, a so-called wearable terminal such as an eyeglass type or a wristwatch type, a notebook type, a tablet type, or a desktop type personal computer. can. Alternatively, the user terminal 40 may be a guide display or a departure sign installed at a station. The service system 30 can be connected to a plurality of user terminals 40.

<Figure 2: Hardware configuration diagram of timetable management system 10>
FIG. 2 shows an example of a hardware configuration diagram of the timetable management system 10. The timetable management system 10 can be realized by a computer device (information processing device) having a storage device 61, a memory 62, a microprocessor 63, a user interface device (UI device) 64, and a communication device 65. Each of these components is communicatively connected via a bus 66. Note that the microprocessor 63 is realized using a CPU (Central Processing Unit). Note that, for example, it may be designed using a virtual machine and realized using a cloud system.

The storage device 61 can be realized by a nonvolatile storage device such as an SSD (Solid State Drive) or a hard disk drive. Here, the storage device 61 holds a computer program 67 for realizing the functional modules 101 to 105 (see FIG. 1) executed by the timetable management system 10. Note that each functional module 101 to 105 will be described later.

Furthermore, the storage device 61 holds, as data 68, data necessary for the execution of the functional modules or various data generated by the functional modules (predicted timetable 107 to dynamic timetable 110 (see FIG. 1)). Note that the data (predicted timetable 107 to dynamic timetable 110) will be described later.

The memory 62 is a volatile memory such as RAM (Random Access Memory). The CPU 63 reads the computer program 67 and data 68 held in the storage device 61 into the memory 62 and executes them. The UI device 64 is connected to an input device (not shown) such as a keyboard or a mouse, or an output device such as a display (not shown), and implements a GUI (Graphical User Interface). Here, the input device and the output device may be realized as a terminal device having an input section and a display section. Furthermore, in addition to the terminal device, the timetable management system 10 itself may have an output device or be connected to an output device.

The communication device 65 performs communication processing with external systems (dynamic operation data distribution system 20, service system 30) via the communication networks 51 and 52. Note that the dynamic operation data distribution system 20 and the service system 30 shown in FIG. 1 as related systems also have the same hardware configuration as the timetable management system 10. In other words, these can be realized using a so-called computer.

<Description of functional modules in Figure 1>
The functional modules 101 to 105 and data (predicted timetable 107 to dynamic timetable 110) of the timetable management system 10 shown in FIG. 1 will be described in detail below. Here, we will explain using railways as an example of public transportation. Note that a static timetable for railways will be described below on the assumption that information on arrival times and departure times at each route, each train, and each station, which are generally predetermined units in public transportation, is stored.

Here, in this embodiment, a prediction process is performed regarding the operating status according to the timetable disruption, and a dynamic timetable is generated based on the accuracy of this prediction process. For this purpose, a prediction error regarding accuracy is used, but this may be calculated by either the predicted timetable generation unit 101 or the dynamic timetable generation unit 103.

An overview of the processing in the predicted timetable generation unit 101 and the dynamic timetable generation unit 103 will be described below. Next, the predicted values 1077 to 1079 handled by the predicted timetable generation unit 101 will be explained. Next, details of the first example in which the dynamic timetable generation unit 103 calculates a prediction error will be explained using FIGS. 4 to 11. Thereafter, a second example in which the prediction error is calculated by the prediction timetable generation unit 101 will be explained using FIG. 12.

First, an overview of the process by which the prediction timetable generation unit 101 calculates a prediction error will be explained. The predicted timetable generation unit 101 receives operation data from the dynamic operation data distribution system 20. Then, the predicted timetable generation unit 101 performs a prediction process to predict predicted values 1077 to 1079, which are the values of future arrival times and departure times in predetermined units such as each route, each train, each station, etc., based on the received operation data. Execute. Then, the predicted timetable generation unit 101 generates a predicted timetable 107 based on the prediction error regarding the accuracy of the prediction process. In this example, the dynamic timetable generation unit 103 can generate a dynamic timetable according to the predicted value specified according to the accuracy of the prediction process of the predicted timetable generation unit 101.

Further, the predicted timetable generation unit 101 may be configured to generate the predicted timetable 107 when the operation data indicates a timetable disruption. In this case, if there is no timetable disruption, the predicted timetable 107 generated in the past may be used, or a static timetable or a dynamic timetable generated in the past may be used.

Note that the predicted value 1077 in this embodiment is the predicted arrival time and departure time, but it may be any information that indicates the degree or extent of the influence due to the schedule change due to the timetable disruption. In other words, it may be a change time such as a predicted delay time or predicted early arrival/departure time with respect to the scheduled arrival time or departure time as indicated by the predicted value 1078. Furthermore, the predicted required time as indicated by the predicted value 1079 may be used. Furthermore, as will be described later, when the dynamic timetable generation unit 103 calculates a prediction error, the prediction timetable generation unit 101 may omit calculation of the prediction error and generate the prediction timetable 107. desirable.

Here, the transmission and reception of operation data may be a PULL type starting from the timetable management system 10 or a PUSH type starting from the dynamic operation data distribution system 20. At this time, it is preferable that the timetable management system 10 checks the received operation data and cleans the operation data if there are any duplications or errors in the operation data. Additionally, when acquiring operation data from different operators or routes, it is desirable to convert the data into a predetermined format so that subsequent processing is possible even if there are differences in the format or items of the operation data. . Note that this cleansing may be performed by the predicted timetable generation unit 101, or may be performed by an input unit or a cleansing unit (not shown).

Further, the predicted timetable generation unit 101 may generate a plurality of patterns of the predicted timetable 107 in consideration of the instability of the operating situation. For example, in addition to the predicted timetable 107 that adopts the predicted value with the highest probability, the predicted timetable 107 that adopts the predicted value with a margin that reduces the possibility of missing or getting off the train may be generated. . The number of patterns may be further increased depending on the margin and size. Furthermore, there may be a predicted timetable 107 that includes only the error time (delay time or early arrival/departure time) with respect to the scheduled time or the required time between each station. Note that the details of the data configuration example of the predicted timetable 107 will be described later using FIG. 3.

Next, an outline of an example in which the dynamic timetable generation unit 103 generates the dynamic timetable 110 based on the accuracy of the prediction processing in the prediction timetable generation unit 101 will be explained. For example, the dynamic timetable generation unit 103 generates the dynamic timetable 110 based on the prediction error of the predicted timetable 107 as the accuracy of the predicted timetable. Details of the processing executed by the dynamic timetable generation unit 103, including this example, will be described later using FIG. 4. In addition to this example, a dynamic timetable may be generated from the specified predicted values themselves. Further, each function of the predicted timetable generation unit 101 and the dynamic timetable generation unit 103 may be executed by either one, or they may be configured as one functional module. Further, as described above, when the prediction error is calculated by the prediction timetable generation unit 101, the dynamic timetable generation unit 103 may omit calculation of the prediction error and generate the dynamic timetable 110. is desirable.

Next, the distribution unit 105 distributes the dynamic timetable 110 generated by the dynamic timetable generation unit 103 to the service system 30. The distribution method may vary depending on the service type of the service system 30, and may be a PUSH type or a PULL type, manual or automatic, email, file sharing service, or Web API. The delivery timing in the case of automatic delivery using the PUSH type can be assumed to be, for example, when the dynamic timetable is updated or when the delay time exceeds a threshold value. Although it has been described that the distribution unit 105 distributes the dynamic timetable 110, it may also distribute the predicted timetable 107 and other data. In this case, the service system 30 may generate the dynamic timetable 110 and the predicted timetable 107.

<Figure 3: Data structure of predicted timetable 107>
FIG. 3 is a data configuration diagram of the predicted timetable 107. Here, the entirety of FIGS. 3(a) to 3(c) may be treated as one predicted timetable 107, and each may be treated as a part of the predicted timetable 107, or at least one of them may be treated as a part of the predicted timetable 107. It may also be treated as a predicted timetable 107. In this embodiment, an example will be described in which each of FIGS. 3(a) to 3(c) is treated as a predicted timetable 107.

FIG. 3(a) is an example in which information about the arrival time and departure time of each route, each train, and each station, which is included in a general static timetable, is stored. It includes fields for storing each value of a category 1075 and a predicted value (predicted time) 1077. The route 1070 is a name or identification code that identifies the route. Train 1071 is a name or identification code that identifies the train. Station 1073 is a name or identification code that identifies the station. Note that the route 1070, the train 1071, and the station 1073, or the train 1071 and the station 1073, can be treated as a predetermined unit in the transportation system. This also applies to FIGS. 3(b) and 3(c).

Classification 1075 is a name or identification code that specifies whether the record has information regarding train arrival or departure. The predicted value (predicted time) 1077 is the value of the arrival time or departure time at the route, train, or station predicted by the predicted timetable generation unit 101. If the predicted value (predicted time) 1077 is distributed to the service system 30, the service system 30 can replace the predicted timetable 107 with the arrival time or departure time of the existing static timetable.

FIG. 3(b) is an example in which the error time (delay time or early arrival/departure time) of each route, each train, and each station with respect to the scheduled arrival time or departure time is stored. FIG. 3B includes fields for storing values of a route 1070, a train 1071, a station 1073, a classification 1075, and a predicted value (predicted delay time) 1078. Routes 1070 to 1075 are the same as those described in FIG. 3(a).

The predicted value (predicted delay time) 1078 is the value of the delay time for the route, train, and station predicted by the predicted timetable generation unit 101. Delays or early arrivals or departures can be distinguished by the positive or negative sign of the value. If the predicted value (predicted delay time) 1078 is distributed to the service system 30, the service system 30 can handle the value by adding or subtracting the arrival time or departure time of the static timetable.

FIG. 3(c) is an example in which the required time between each route, each train, and each station is stored. Contains fields that store values. The route 1070 to the train 1071 are the same as those described in FIG. 3(a). The departure station 1074 is a name or identification code that identifies the departure station. The arrival station 1076 is a name or identification code that specifies the arrival station.

The predicted value (predicted travel time) 1079 is the value of the route, the train, and the required time from the departure station to the arrival station predicted by the prediction timetable generation unit 101. If the predicted value (predicted required time) 1079 is distributed to the service system 30, the service system 30 can handle it by adding or subtracting the arrival time or departure time of the existing static timetable based on the value. Alternatively, when providing a service that provides guidance on the required time itself, the value of the predicted required time 1079 may be used as is.

Note that any method can be used to predict each value of the predicted timetable 107 based on the data received from the dynamic operation data distribution system 20. For example, the delay time may be predicted based on past performance statistics. Further, a change in the running position may be predicted from the relationship between the running positions of the train in question and the trains before and after. Further, expansion or reduction of the delay may be predicted based on the tendency of the delay time of the train. Further, expansion or reduction of the delay of the train in question may be predicted based on the tendency of the delay time of the preceding train. Furthermore, the influence on the train may be predicted from route operation information.

Furthermore, if the railway operator can obtain highly accurate predicted information from a control system that manages train operation, the prediction process may be performed using that information. External information such as weather information on typhoons, heavy snow, etc. may be used to predict whether the delay will increase or decrease. Furthermore, changes in transfer time and platform waiting time may be predicted using information on congestion inside the station, on trains, operating information on station equipment, etc., and added to the train delay information. It is also possible to collect and utilize the voices of local users from social data. You may combine two or more of these methods.

The predicted timetable 107 generates a pattern for each route, just like a general static timetable. Different patterns may be generated for each date or day of the week. In addition, there are many cases where trains are operated across multiple routes/sections or operators due to direct operation, and if a delay or suspension of service occurs, the operation may be changed depending on the situation, such as turning back at a connecting station. The boundaries of operation may change. For such routes, the predicted timetable 107 may be generated for each pattern by dividing one route into a plurality of routes, or the way the patterns are divided may be changed each time depending on the situation.

Note that, as described above, in the present invention, generation of the predicted timetable 107 shown in FIG. 3 is not essential, and predicted values and other information may be used.

The details of the first example in which the dynamic timetable generation unit 103 calculates a prediction error will be described below.

<Figure 4: Flowchart of processing executed by the dynamic timetable generation unit 103>
FIG. 4 is a processing flow executed by the dynamic timetable generation unit 103. That is, FIG. 4 shows a flow in which the dynamic timetable generation unit 103 calculates a prediction error and generates the dynamic timetable 110.

First, the prediction error calculation means (not shown) of the dynamic timetable generation unit 103 calculates the prediction error of the prediction timetable 107 generated by the prediction timetable generation unit 101 (step S1031). The prediction error is the predicted value (predicted time, predicted delay time, or predicted required time) of the predicted timetable 107, that is, an index regarding the accuracy of the prediction process, and is calculated for a predetermined unit (each route, each train, each station). be done. The smaller the prediction error is, the more likely the predicted value of the prediction timetable 107 is, and the larger the prediction error is, the more uncertain the predicted value is. A method for calculating the prediction error will be described later using FIG. 5.

Next, the disclosure determination unit (not shown) of the dynamic timetable generation unit 103 generates the disclosure determination result 109 based on the prediction error calculated in step S1031 (step S1033). That is, the disclosure determination means identifies the predicted timetable to be used for generating the dynamic timetable 110 from among the predicted timetables 107. For this purpose, the disclosure determination means compares the prediction error with the conditions and thresholds stored in the disclosure threshold 108. Then, the disclosure determination means generates a disclosure determination result 109 depending on whether the condition is satisfied or whether there is a predetermined relationship with the threshold value. Details of the data configuration example of the disclosure threshold 108 will be described later using FIG. 6. Details of the data structure example and generation method of the disclosure determination result 109 will be described later using FIG. 7. Note that the disclosure determination result 109 may be information indicating whether or not to be used for generating the dynamic timetable 110 for the predicted timetable 107.

Next, the dynamic timetable generation means (not shown) of the dynamic timetable generation unit 103 generates the dynamic timetable 110 based on the disclosure determination result 109 generated in step S1033 (step S1035). Details of step S1035 will be described later using FIG. 8B. Further, details of an example data structure of the dynamic timetable 110 will be described later using FIG. 9.

In the explanation of FIG. 4, each means of the dynamic timetable generation unit 103 is mainly described, but the dynamic timetable generation unit 103 itself may execute it, or it may be executed by a combination of means. good.

<Figure 5: Explanation of calculation method of prediction error>
Next, a method for calculating a prediction error will be explained using FIG. 5. Here, we will also explain the concept of calculating the prediction error.

In FIG. 5(a), the vertical axis 71 shows the predicted value (predicted time, predicted delay time, or predicted required time) of the predicted timetable 107, and the horizontal axis 72 shows the predicted value at an arbitrary time. This is a graph drawn as plots 73a to 73e. The passage of time on the horizontal axis 72 may be the time itself.

In addition, for example, it may be a list of stations arranged in the order of stops at an arbitrary train, or a list of trains arranged in the order of stops at an arbitrary station. Depending on the index used on the horizontal axis 72, the index on the vertical axis 71 also changes. For example, in the case of a graph in which the vertical axis 71 is the predicted delay time of train A, and the horizontal axis 72 is station a, station b, and station c, which are the stations where train A stops, the graph will be as follows. That is, the plot 73a is the predicted delay time of train A at station a, the plot 73b is the predicted delay time at station b, and the plot 73c is the predicted delay time at station c. In other words, the horizontal axis may be not only time but also location, transportation equipment, etc.

FIG. 5(b) is a graph in which a straight line 74 is drawn on the graph of FIG. 5(a), and the vertical distances between the plots 73a to 73e and the straight line 74 are illustrated as dotted lines 75a to 75e. The straight line 74 is an approximate straight line determined from each plot. The distance of the dotted line 75 is called a prediction error, and the smaller the prediction error is, the more likely the predicted value is, and the larger the prediction error is, the more uncertain the predicted value is. It is estimated that the closer the time on the horizontal axis 72 is to the present, the more likely the predicted value is, so for example, by weighted approximation, the closer the time is to the present, the shorter the distance, and the farther the time is from the present, the longer the distance on the dotted line 75. (The distance on the dotted line 75e is shorter than the distance on the dotted line 75c).

For example, the weight may be the reciprocal of the distance between stations or the time required between stations, or on routes where different train types (limited express trains and local trains, etc.) coexist, weights may be used depending on the train type. may be defined. Further, the straight line 74 does not necessarily have to be an approximate straight line, but may be obtained as an approximate curve. The prediction error is calculated for each route, each train, and each station.

That is, the above-mentioned prediction error calculation means compares the approximate line and the plot. Specifically, the approximation value indicated by the approximation line is compared with the values (plotted points) forming the prediction timetable 107, and the difference therebetween is calculated as the prediction error. Note that although an approximate line is used in this embodiment, other information may be used. For example, a predicted value generated by another algorithm may be used.

<Figure 6: Example data structure of disclosure threshold>
Next, FIG. 6 shows an example of the data structure of the disclosure threshold 108. The disclosure threshold 108 includes fields that store values of a priority 1081, a disclosure timetable 1083, a predicted timetable 1085, a predicted value item 1087, and a threshold 1089. A threshold value and a comparison condition are stored in each record of the disclosure threshold value 108.

The priority order 1081 is a value that specifies the order used for threshold comparison for each record of the disclosure threshold value 108. The disclosure timetable 1083 is a name or identification code that specifies the predicted timetable 107 that can be disclosed to the user when the threshold comparison conditions for each record are met.

The prediction timetable 1085 is a name or identification code that specifies the prediction timetable 107 that is the target for calculating the prediction error in step S1031 (see FIG. 4). The predicted value item 1087 is a name or identification code that specifies the predicted value item (predicted time, predicted delay time, or predicted required time) to be used when calculating the prediction error in step S1031.

The approximation value item 1088 is an item for identifying the approximation line explained in FIG. 5 or the approximation value that constitutes this. Note that in this embodiment, a predicted value item 1087 and an approximate value item 1088 are provided, but a prediction error item that is the difference between these items may also be provided. This also applies to FIGS. 6(b) and 6(c).

The threshold value 1089 is a threshold value that serves as a reference for comparison. That is, if the predicted value item 1087 in the predicted timetable 1085 satisfies the condition of the threshold value 1089, the disclosure timetable 1083 can be disclosed. It is checked whether the conditions are met in the order of records with the lowest priority 1081 value, and if the conditions are not met, it is checked whether the next record satisfies the conditions. If there is no record that satisfies the conditions, it is determined that there is no predicted timetable 107 that can be disclosed.

Specifically, the disclosure determination means compares the prediction error, which is the difference between the predicted delay time of the predicted value item 1087 and the approximate value of the approximate value item 1088, with a threshold value 1089. As a result, if the prediction error is less than or equal to the threshold value 1089, the disclosure determination means determines that accuracy is ensured, that is, the corresponding prediction timetable 107 can be used to generate the dynamic timetable 110. It is determined that As described above, when a prediction error item is provided instead of the predicted value item 1087 and the approximate value item 1088, the disclosure determination means uses the prediction error indicated by this prediction error item for comparison. Note that this also applies to FIGS. 6(b) and 6(c). The contents of this process will be specifically explained below.

FIG. 6A shows an example of one condition, that is, one record. In this case, when the prediction error of the predicted value item 1087 "Predicted delay time" in the predicted timetable 1085 "Predicted timetable pattern 1" is the threshold value 1089 "5 minutes or less", the disclosure timetable 1083 "Predicted timetable pattern 1" Indicates that it is possible to disclose.

FIG. 6(b) is an example of the disclosure threshold 108 in which there are multiple conditions (records). In this case, it is checked whether the conditions of each record are satisfied in the order of the values stored in the priority order 1081. In this example, when the prediction error of the predicted value item 1087 "Predicted delay time" in the predicted timetable 1085 "Predicted timetable pattern 1" is the threshold value 1089 "5 minutes or less", the disclosure timetable 1083 "Predicted timetable pattern 1" is ” can be disclosed. Then, if that condition is not met and the prediction error of the predicted value item 1087 "Predicted time required" in the predicted timetable 1085 "Predicted timetable pattern 2" is the threshold value 1089 "10 minutes or less", the disclosure timetable 1083 " "Predicted timetable pattern 2" can be disclosed.

FIG. 6C is an example in which there are multiple records in which the same value is stored in the priority order 1081. In this case, the corresponding disclosure timetable 1083 can be disclosed only when all the conditions of the plurality of records in which the same value is stored in the priority order 1081 are satisfied. This example shows that the "predicted timetable pattern 1" shown in the disclosure timetables 1083 of both records can be disclosed if the following two conditions are met.

Condition 1: The prediction error of the predicted value item 1087 “Predicted delay time” in the predicted timetable 1085 “Predicted timetable pattern 1” is the threshold 1089 “5 minutes or less”
Condition 2: The prediction error of the predicted value item 1087 “Predicted time required” in the predicted timetable 1085 “Predicted timetable pattern 2” is the threshold 1089 “5 minutes or less”
Instead of satisfying all of the conditions for multiple records, it may be sufficient to satisfy either one of the conditions. Although routes, trains, and stations are not specified in the example of FIG. 6, conditions and threshold values may be changed for each route, train, or station.

As described above, the above-described disclosure determining means uses the disclosure threshold 108 to determine the accuracy of the prediction process. In other words, the disclosure determination means compares the prediction error with the threshold value 1089, and determines whether the corresponding prediction timetable 107 is to be used for generating the dynamic timetable 110 according to the result. Here, in the example described so far, the accuracy of the prediction process was explained using the comparison result between the prediction error and the threshold value, but the comparison result between each prediction error may also be used. In this case, it is also possible to compare adjacent prediction errors (by station, time, etc.) and make a determination based on the difference.

<Figure 7: Example of data structure of disclosure determination result 109>
Next, the disclosure determination result 109 generated as a result of determination by the disclosure determination means will be explained. FIG. 7 shows an example of the data structure of the disclosure determination result 109.

In FIG. 7A, a predicted timetable 107 that can be disclosed to the user is set for each route, each train, and the arrival and departure of each station. Therefore, when the prediction error calculated in step S1031 (see FIG. 4) corresponds to the predicted time or predicted delay time, this data structure is generated. In this example, the disclosure determination result 109 includes fields that store values of a route 1090, a train 1091, a station 1093, a classification 1095, and a disclosure timetable 1097.

Route 1090 to segment 1095 are the same as route 1070 to segment 1075 of predicted timetable 107 shown in FIG. 3(a) or FIG. 3(b). The disclosure timetable 1097 is the name or identification code of the predicted timetable 107 determined to be disclosable by comparing the prediction error calculated in step S1033 (see FIG. 4) with the disclosure threshold 108. In this example, it is determined that the predicted value (predicted time or predicted delay time) when train 1091 "Train A" on route 1090 "Route 1" will arrive at station 1093 "Station a" in classification 1095 "Arrival" can be disclosed. to show that This means that as a result of comparing the conditions stored in the disclosure threshold value 108 and the threshold value, it has been determined that the disclosure timetable 1097 "predicted timetable pattern 1" can be disclosed.

In FIG. 7(b), a predicted timetable 107 that can be disclosed to the user is set for the combinations between each station of each train on each route, and the prediction error calculated in step S1031 corresponds to the predicted travel time. If so, generate this data structure. In this example, the disclosure determination result 109 includes fields that store values of a route 1090, a train 1091, a departure station 1094, a destination station 1096, and a disclosure timetable 1097. Route 1090 to destination station 1096 is the same as route 1070 to destination station 1076 in the predicted timetable 107 shown in FIG. 3(c), and disclosure timetable 1097 is the same as that explained in FIG. 3(a). It is. In this example, it has been determined that the predicted value (predicted travel time) of train 1091 "Train A" on route 1090 "Route 1" from departure station 1094 "Station a" to destination station 1096 "Station b" can be disclosed. shows. This means that as a result of comparing the conditions stored in the disclosure threshold value 108 and the threshold value, it has been determined that the disclosure timetable 1097 "predicted timetable pattern 1" can be disclosed.

Note that, as described above, the disclosure determination result 109 may be information indicating whether or not the predicted timetable 107 is used to generate the dynamic timetable 110.

<Figure 8A: Flowchart of processing for creating disclosure determination result 109>
Here, details of the accuracy determination process of the prediction process by the disclosure determination means, that is, step S1033 in FIG. 4 will be described using FIG. 8A. FIG. 8A is a flowchart of processing for creating the disclosure determination result 109. Note that although this flow uses the disclosure threshold 108, an example using a disclosure threshold different from the example of FIG. 6 will be described below. Another example is an example in which a plurality of threshold values 1089 (two in this example) are prepared. In this way, by preparing a plurality of threshold values 1089, it is possible to specify the predicted timetable 107 used for creating the dynamic timetable 110 more precisely in accordance with the actual situation.

First, the above-mentioned disclosure determination means calculates a prediction error (here, referred to as e) (step S10331). The contents are as described above.

Next, the disclosure determining means compares the prediction error e with two thresholds a and b (step S10332). Here, it is assumed that the threshold values a and b are stored in the same way as the disclosure threshold value 108 shown in FIG. Further, in this example, threshold a<threshold b.

As a result, if the prediction error e<threshold a, the process moves to step S10333. Further, if threshold a<prediction error e≦threshold b, the process moves to step S10334. Further, if the prediction error≧threshold b, the process moves to step S10335.

Next, the disclosure determining means determines that the accuracy of the prediction error is high (step S10333). Here, the disclosure determination means utilizes the predicted timetable 107 with "high accuracy" to create the dynamic timetable 110 using the correspondence relationship between the accuracy and the predicted timetable 107 that is stored in advance. It is specified as the predicted timetable 107.

Similarly, when the disclosure determining means determines that the accuracy of the prediction error is medium, it specifies the prediction timetable 107 with “medium accuracy” as the prediction timetable 107 to be used for creating the dynamic timetable 110. (Step S10334). Furthermore, when the disclosure determining means determines that the accuracy of the prediction error is low, it specifies the prediction timetable 107 with "low accuracy" as the prediction timetable 107 to be used for creating the dynamic timetable 110. (Step S10335). Note that in step S10335, the disclosure determining means may determine that there is no predicted timetable 107 to be used for creating the dynamic timetable 110.

This concludes the explanation of FIG. 8A, and next, the creation of the dynamic timetable 110 will be explained using FIG. 8B.

<Figure 8B: Flowchart of processing for generating dynamic timetable 110>
Next, details of the process in which the dynamic timetable generation means of the dynamic timetable generation unit 103 generates the dynamic timetable 110 in response to the processing result of FIG. 8A will be described. That is, details of step S1035 in FIG. 4 will be explained. FIG. 8B is a detailed flowchart of step S1035 (see FIG. 4) in the process executed by the dynamic timetable generation unit 103.

Here, in the disclosure determination result 109, a disclosure timetable is set for each line, each train, and each station (or between stations) as shown in FIG. In other words, each disclosable predicted value (predicted time, predicted delay time, or predicted required time) is obtained by referring to a different predicted timetable 107. However, as described above, since a general static timetable has a pattern generated for each route, it is desirable that the dynamic timetable 110 also generate a pattern for each route in order to be utilized by an external system. Therefore, the process for integrating the values set for each train and each station into one pattern data is step S1035, the details of which are shown in FIG. 8B.

The contents will be explained below with reference to FIG. 8B. The above-described dynamic timetable creation means determines which of the flows 10351a to 10351c is to be executed according to preset conditions (step S10351).

Here, the flow 10351a can generate a more reliable dynamic timetable 110, and the amount of processing increases accordingly. Conversely, flow 10351c reduces the reliability of the dynamic timetable 110, but requires less processing. Further, flow 10351b is intermediate processing between flows 10351a and 1035c.

Therefore, in step S10351, the dynamic timetable creation means determines which flow to use according to the load state of the timetable management system 10. In other words, the dynamic timetable creation means adopts flow 10351a when the load state is lower than the predetermined standard. The dynamic timetable creation means adopts flow 10351c when the load state is higher than a predetermined standard. Furthermore, if the load is at an intermediate level, such as at the same level as the predetermined standard, the dynamic timetable creation means adopts flow 10351b.

Furthermore, a time zone may be used as a preset condition. For example, by changing the flow for each time slot, it is possible to set the rush time slot with high demand as the flow 10351a, and the free time slot as the flow 10351c.

Furthermore, as a condition, facilities such as lines, stations, trains, etc. where the timetable disruption has occurred may be used. In this case, it is desirable to change the flow depending on the number of users (passengers) using the service.

Here, in flow 10351a, the dynamic timetable creation means specifies the predicted timetable 107 stored in the disclosure timetable 1097 of the disclosure determination result 109 for each train and for each station (or between stations) (step S10353).

Then, the dynamic timetable generation means generates the dynamic timetable 110 by integrating the specified predicted timetables for each train and station (or between stations). That is, the dynamic timetable creation means integrates each of the identified predicted timetables 107 as one pattern data (step S10357).

Further, in flow 10351b, the dynamic timetable creation means specifies the predicted timetable 107 for each train or for each station (or between stations) (S10355). For example, when referring to one predicted timetable 107 for each train, the pattern with the largest number of stored prediction timetables 107 stored in the disclosure timetable 1097 of the disclosure determination result 109 of each station for the relevant train The predicted timetable 107 is specified. Note that it is not necessary to select a pattern from the predicted timetable 107 that has a large number of patterns; it is sufficient to set a standard for selecting a plausible pattern and specify it according to this standard.

Furthermore, it is not necessary to select one pattern of the predicted timetable 107 for each train or station (or between stations); instead, the pattern of the predicted timetable 107 to be selected can be divided into several sections according to the arrangement of the stations. You can change it. Once the predicted timetables are identified for each train or station (or between stations) in this way, the timetables are integrated as one pattern data (step S10357).

In addition, in flow 10351c, the dynamic timetable generation means specifies a predicted route timetable that is one predicted timetable 107 for each train and each station (or between stations), that is, for the entire target route (step S10359). For example, the predicted timetables 107 stored in the disclosure timetables 1097 of the disclosure determination results 109 for each station (or between stations) of each train are acquired, and the predicted timetable 107 having the most stored pattern is identified. . In this case as well, it is not necessary to select a pattern from the predicted timetable 107 that has a large number of patterns; it is sufficient to select a plausible pattern. Further, the predicted route timetable may be information indicating a predetermined summary of the operation status such as "operation status unknown" or "significant timetable disruption."

<Figure 9: Example of data structure of dynamic timetable 110>
Next, the dynamic timetable 110 created in step SS1035 will be explained. FIG. 9 shows an example of the data structure of the dynamic timetable 110. Note that in this embodiment, two examples of the dynamic timetable 110 shown in FIGS. 9(a) and 9(b) are illustrated, but these may also be treated as one dynamic timetable 110.

In FIG. 9A, the predicted timetable 107 and its predicted values are set for each route, each train, and the arrival and departure of each station, which are referred to in order to generate the dynamic timetable 110. Then, when the predicted value is a predicted time or a predicted delay time, it is generated in the data structure shown in FIG. 9(a). In this example, the dynamic timetable 110 includes fields that store values of a route 1100, a train 1101, a station 1103, a segment 1105, a dynamic timetable pattern 1107, and a predicted value (predicted delay time) 1108.

Route 1100 to section 1105 are the same as route 1090 to section 1095 of disclosure determination result 109 shown in FIG. 7(a). The dynamic timetable pattern 1107 stores the name or identification code of the predicted timetable 107 referenced in step S10353, S10355, or S10359 of FIG. 8B. The predicted value (predicted delay time) 1108 stores the predicted value (predicted delay time in this example) stored for the corresponding route 1100, train 1101, station 1103, and section 1105 in the corresponding dynamic timetable pattern 1107. do. In this example, when train 1101 "Train A" on route 1100 "Route 1" arrives at station 1093 "Station a" in section 1105, dynamic timetable pattern 1107 "Predicted timetable pattern 1" can be disclosed. The predicted value (predicted delay time) 1108 is "1 minute".

In FIG. 9(b), the predicted timetable 107 referred to in order to generate the dynamic timetable 110 and its predicted value are set for the combination between each station of each train on each route. This data structure is generated when is the estimated required time. In this example, the dynamic timetable 110 includes fields that store values of a route 1100, a train 1101, a departure station 1104, a destination station 1106, a dynamic timetable pattern 1107, and a predicted value (predicted travel time) 1109. The route 1090 to the destination station 1106 is the same as the route 1090 to the destination station 1096 in the disclosure determination result 109 shown in FIG. 7(b).

The dynamic timetable pattern 1107 stores the name or identification code of the predicted timetable 107 referenced in step S10353, S10355, or S10359 of FIG. 8B. The predicted value (predicted time required) 1109 stores the predicted value (predicted required time) stored at the corresponding route 1100, train 1101, departure station 1104, and destination station 1106 in the corresponding dynamic timetable pattern 1107. Store. In this example, the dynamic timetable pattern 1107 "predicted timetable pattern" is used to determine the time required for train 1101 "train A" on route 1100 "route 1" to travel from departure station 1104 "station a" to destination station 1106 "station b". 1" can be disclosed. The predicted value (predicted required time) 1109 is "10 minutes".

In the examples shown in FIGS. 9A and 9B, each record has a different dynamic timetable pattern 1107. However, as shown in FIG. 8B, in step S1035, the predicted timetable 107 to be distributed is determined for each train or station, or one predicted timetable 107 is distributed at each train station (or between stations). Good too. Therefore, the dynamic timetable pattern 1107 of each record may be the same.

<Figure 10: Example of administrator screen of service systems 30a, 30b>
FIG. 10 shows an example of the administrator screen of the service system 30. This screen is displayed on a terminal device connected to the service system 30 described above. The administrator of the service system 30 operates this screen in order to set the display of services based on the dynamic timetable 110 distributed by the timetable management system 10. In other words, the terminal device receives input and instructions from the administrator. Here, the service system 30 is a system owned by a service provider that provides a route search service. Therefore, the service system 30 generates a recommended route and time based on the route search conditions received from the user terminal 40 (for example, starting point, destination, date and time of use, etc.), and displays the guidance on the screen of the user terminal 40. shall be displayed. Unless otherwise specified, the processing described below is realized as a management terminal, the service system 30, or a cooperative process thereof.

The administrator screen 300 includes a dynamic timetable pattern setting section 301 and a display setting section 303. The dynamic timetable pattern setting unit 301 stores a name or identification code representing a dynamic timetable pattern received from the timetable management system 10. The values stored in the dynamic timetable pattern setting section 301 correspond to each value stored in the dynamic timetable pattern 1107 of the dynamic timetable 110 distributed by the timetable management system 10.

The display setting section 303 sets the display contents of the guide screen of the user terminal 40 in the corresponding dynamic timetable pattern setting section 301. In the example shown in FIG. 10, parameters such as route, station, time, required time, and text are set in each section 3031 to 3037 of the display setting section 303.

Each value stored in the dynamic timetable 110 distributed by the timetable management system 10 is assigned to the parameters surrounded by solid line squares shown in 3031 to 3036. For example, the value of route 1100 of the dynamic timetable 110 is substituted for "route" shown in 3031. The value of the station 1103 of the dynamic timetable 110 is assigned to the "station" indicated by 3032.

The values of the departure station 1104 and the arrival station 1106 of the dynamic timetable 110 are substituted into the "departure station" and "destination station" shown in 3033. The values of the station 1103, the departure station 1104, and the arrival station 1106 of the dynamic timetable 110 are substituted into the "station (or) departure station, arrival station" shown in 3034. The "time" shown in 3035 stores the value obtained by adding or subtracting the predicted value (predicted delay time) 1108 of the dynamic timetable 110 from the arrival time or departure time of the static timetable, or the predicted time as a predicted value. If so, that value will be assigned.

The value of the predicted value (predicted required time) 1109 of the dynamic timetable 110 is substituted into the "required time" shown in 3036. Furthermore, text sentences are stored in the underlined parameters 3037 and displayed on the guide screen of the user terminal 40 together with the values of other parameters. A specific example of the guide screen of the user terminal 40 will be described later using FIG. 11. In the example of FIG. 10, the values of the train 1101 and the dynamic timetable pattern 1107 held by the dynamic timetable 110 are not displayed on the guide screen of the user terminal 40, but they may be displayed on the guide screen of the user terminal 40. Information may be displayed.

By designing the guidance screen in advance by the administrator of the service system 30 using the screen illustrated in FIG. 10, it becomes possible to provide the user terminal 40 with guidance that reflects the actual traffic situation. Note that this also allows the service system 30 itself to omit traffic situation monitoring and delay detection.

<Figure 11: Example of guidance screen of user terminals 40a to 40d>
FIG. 11 shows an example of a guide screen displayed when the user terminal 40 uses a service provided by the service system 30. Here, in accordance with the example of FIG. 10, it is assumed that the service system 30 is a system owned by a service provider that provides a route search service. Therefore, the recommended route will be displayed on this guidance screen. That is, FIG. 11 is an example of a screen displayed as a result of transmitting route search conditions (for example, departure point, destination, date and time of use, etc.) from the user terminal 40 to the service system 30. Note that each process described below is realized as a process of the user terminal 40.

In the example shown in FIG. 11, on the guidance screen 400, "get on" or "get off" is specified in 409, and then in each part of 401 to 407, various information such as route, station, time, required time, and text are specified. The value of the parameter is displayed. For example, each value set in 3031 of FIG. 10 is displayed in 401. Further, in 402, each value set in 3032 or 3034 in FIG. 10 is displayed.

Further, in 403, each value set in 3033 or 3034 in FIG. 10 is displayed. 404 displays each value set at 3034 in FIG. 10, 405 displays each value set at 3035 in FIG. 10, and 407 displays each value set at 3037 in FIG. Each value is displayed. In addition, the departure time and arrival time of each station determined by a static timetable may be displayed, or items displayed in general route search services (for example, fare, required time, boarding position, route summary, other route candidates, etc.) may also be displayed. Note that for areas without routes, it may be possible to display information on other modes of transportation such as private cars, taxis, and walking.

The screen illustrated in FIG. 11 allows the user to receive guidance that matches the actual traffic situation. Therefore, the user can judge the traffic situation by himself and can omit the selection of the service to be used.

<Figure 12: Flowchart of processing executed by the predicted timetable generation unit 101>
Next, a second example of calculation of a prediction error, that is, a process performed by the prediction timetable generation unit 101 will be described. FIG. 12 shows a flowchart of processing executed by the predicted timetable generation unit 101.

First, the predicted timetable generation unit 101 receives operation data from the dynamic operation data distribution system 20 (step S1011). Next, the predicted value generation means (not shown) of the predicted timetable generation unit 101 executes a process of predicting the operating status and calculates predicted values 1077 to 1079 (step S1012). At least one of these predicted values 1077 to 1079 may be generated, and the contents thereof are as described using FIG. 3.

Next, a prediction error calculation unit (not shown) of the prediction timetable generation unit 101 calculates a prediction error regarding the accuracy of the prediction process. This process can be executed by basically the same process as the prediction error calculation means of the dynamic timetable generation unit 103 (same as step S1031). However, in this example, since the predicted timetable 107 has not yet been generated, the predicted value calculated in step S1012 will be used.

Next, a disclosure determination unit (not shown) of the predicted timetable generation unit 101 generates a disclosure determination result 109 based on the prediction error calculated in step S1013 (step S1014). This is the same process as the process of the disclosure determining means of the dynamic timetable generation unit 103 (step S1033). That is, here, information indicating the accuracy of the prediction process is specified based on the prediction error. For this purpose, the disclosure determination means performs a comparison process with, for example, a pre-stored threshold value.

Next, a predicted timetable generation unit (not shown) of the predicted timetable generation unit 101 specifies predicted values to be used for creating the predicted timetable 107 and, ultimately, the dynamic timetable 110, based on the information indicating accuracy. Then, the predicted timetable generation means uses this to generate a predicted timetable 107 for each predetermined unit, that is, at least a portion of these (step S1015). As a result, the dynamic timetable generation unit 103 can generate the dynamic timetable 110 using the predicted timetable 107 created here.

<Summary>
The present invention has been described above along with examples. The present invention is not limited to the above embodiments, but includes various modifications. The above embodiments are described in detail to explain the present invention in an easy-to-understand manner, and the present invention is not limited to those having all the configurations described.

For example, the timetable management system 10 may also distribute information other than the timetable, such as information on service intervals. The frequency of distribution may be changed depending on the operation status. Moreover, the service system 30 may notify the user terminal 40 from the service system 30 when the operating situation changes, instead of displaying guidance according to the route search conditions from the user terminal 40.

Further, in the service system 30, the commuter pass section may be registered in advance as a search condition, or the route from the previous search may be saved and used for subsequent searches. Furthermore, the service system 30 does not necessarily have to be a system that provides route guidance, but may be a system that provides guidance on departure times, such as a departure sign installed at a station. Alternatively, the timetable management system 10 may distribute a plurality of patterns of dynamic timetables 110, and the service system 30 may determine which of the plurality of patterns to use.

If a plurality of patterns of the dynamic timetable 110 exist, the user may be confused if the guided time suddenly changes due to a change in the pattern of the predicted timetable 107 to be referred to. In order to avoid this, the reference pattern may be switched after satisfying the condition of the disclosure threshold 108 multiple times, or the time may be changed gradually (that is, in multiple steps).

Also, from an operator's perspective, if the forecasts are too accurate, trains could be flooded with passengers and delays could increase. In order to avoid this, it is possible to provide information on the prediction of later trains and guide passengers to skip one train, or to intentionally provide different guidance to each passenger. The disclosure threshold 108 may be determined from such a viewpoint. On the other hand, from a passenger's perspective, some people are more concerned about whether or not they will be able to make it on time when boarding a delayed train rather than the timetable itself. The disclosure threshold 108 may be determined taking such passengers into consideration. Alternatively, the threshold value may be set in the service system 30 or the user terminal 40.

As described above, according to the present embodiment, the disclosure of the predicted timetable is determined according to the prediction error calculated in a predetermined unit that constitutes a transportation system such as each route, each train, each station, etc., and the dynamic time that can be disclosed to the user is determined. Generate a table. As a result, even if the scheduled arrival and departure times cannot be reliably predicted due to unstable operating conditions, it is possible to provide services that maintain appropriate quality depending on the reliability of the predicted values. In other words, in order to prevent erroneous guidance to the user, it is possible to provide services that maintain appropriate quality even when the operating conditions are unstable and the scheduled arrival and departure times cannot be reliably predicted.

Furthermore, in this embodiment, the dynamic operation data distribution system 20, timetable management system 10, and service system 30 have been described as independent devices. However, a system may be constructed by combining at least some of these.

Furthermore, in this embodiment, the dynamic operation data distribution system 20, the timetable management system 10, and the service system 30 are each described as a framework operated and utilized by different businesses. However, these may be jointly operated and used by a plurality of businesses, or may be operated and managed collectively by a business.

Furthermore, the processing of this embodiment has been described with a configuration in which a dynamic timetable, which is a type of dynamic timetable information according to the operating conditions, is generated and a recommended route based on this is output to the user terminal. However, in this embodiment, a configuration may be created in which dynamic timetable information is generated from operation data of transportation facilities, and guidance information generated from the dynamic timetable information is distributed.

Here, the dynamic timetable information of this embodiment is information indicating the operating status when the operating schedule is changed such as a delay, and includes a dynamic timetable. The dynamic timetable information includes dynamic timetables that summarize the operating times of transportation facilities, as well as the operation status of predetermined units such as stations that make up transportation facilities, and details of changes to the operation schedule such as schedules and delay times. is included. Note that the dynamic timetable information may be expressed in any way, such as dynamic operation information.

Further, the guidance information indicates boarding guidance for users of transportation facilities, and includes route information such as so-called recommended routes such as transfer guidance.

Furthermore, the dynamic timetable generated in this embodiment may be used not only for guiding recommended routes but also for selling tickets such as train tickets and reserved seat tickets. In this case, a dynamic timetable may also be used to introduce dynamic pricing to fares and designated charges, and to vary prices to avoid delays due to congestion, for example. On the other hand, it is also conceivable to provide a dynamic timetable to a control system. Furthermore, the user may be provided with not only travel routes between stations, but also routes that include detours such as shopping, and suggestions for how to use time other than traveling.

1: Operation guidance system, 10: Timetable management system, 20a, 20b: Dynamic operation data distribution system, 30a, 30b: Service system, 40a to 40d: User terminal, 51, 52, 53: Communication network, 101: Prediction Timetable generation unit, 103: Dynamic timetable generation unit, 105: Distribution unit, 107: Prediction timetable, 108: Disclosure threshold, 109: Disclosure determination result, 110: Dynamic timetable

Claims (9)

A timetable information management system that generates dynamic timetable information indicating the operation status when the operation schedule of a transportation facility is changed,
an input means for accepting input of operation data of the transportation facility;
Predicted value generation means for generating a predicted value indicating the degree of influence of the change in the operation schedule by executing a prediction process on the operation status for each predetermined unit constituting the transportation facility based on the input operation data; ,
Prediction error calculation means for calculating a prediction error regarding the accuracy of the prediction process ;
Predicted value threshold storage means for storing a predetermined predicted value threshold;
disclosure determining means for comparing the prediction error with the prediction value threshold and identifying the accuracy of the prediction error as the accuracy of the prediction process;
Schedule timetable generation means for generating a plurality of predicted timetables for each of the predetermined units including the specified predicted values;
as the dynamic timetable information, a dynamic timetable generating means for generating a dynamic timetable that summarizes operating times in the transportation facility;
The dynamic timetable generation means is a timetable information management system that generates the dynamic timetable using a predicted timetable specified according to the accuracy specified by the disclosure determination means according to predetermined conditions. . The timetable information management system according to claim 1 ,
The means of transportation is a railway, the predetermined unit is a train and a station of the means of transportation,
The dynamic timetable generation means, according to the predetermined conditions,
(1) generating the dynamic timetable using the predicted timetable for each train and the station;
(2) generating the dynamic timetable using a predicted timetable for each train or station;
(3) A timetable information management system that uses a route prediction timetable for the entire target route including the predetermined unit as the dynamic timetable. In the timetable information management system according to claim 2 ,
The dynamic timetable generation means is a timetable information management system that generates the route prediction timetable using a predicted value corresponding to the accuracy of the most frequent prediction error among the accuracy of each prediction error in the predetermined unit. . The timetable information management system according to any one of claims 1 to 3 ,
The predicted value generation means generates a plurality of predicted values based on the predicted value in chronological order,
The prediction error calculation means specifies an approximate line based on the plurality of predicted values, and calculates a difference between the specified approximate line and the plurality of predicted values as the prediction error. A timetable information management method using a timetable information management system that generates dynamic timetable information indicating operation status when the operation schedule of a transportation facility is changed, the method comprising:
Accepting input of operation data of the transportation facility,
Based on the input operation data, perform a prediction process on the operation status for each predetermined unit constituting the transportation facility, and generate a predicted value indicating the degree of influence due to the change in the operation schedule,
By calculating the prediction error regarding the accuracy of the prediction process,
Generating the dynamic timetable information based on a predicted value identified from the generated predicted values according to the prediction error,
storing a predetermined predicted value threshold;
Comparing the prediction error with the prediction value threshold and identifying the accuracy of the prediction error as the accuracy of the prediction process,
generating a plurality of prediction timetables for each of the predetermined units including the specified predicted values;
The dynamic timetable information is a dynamic timetable that summarizes the operating times of the transportation facility, and the dynamic timetable is specified using a predicted timetable that is specified according to the accuracy according to predetermined conditions. A timetable information management method that generates a table . In the timetable information management method according to claim 5 ,
The means of transportation is a railway, the predetermined unit is a train and a station of the means of transportation,
According to the predetermined conditions,
(1) generating the dynamic timetable using the predicted timetable for each train and the station;
(2) generating the dynamic timetable using a predicted timetable for each train or station;
(3) A timetable information management method using a route prediction timetable for the entire target route including the predetermined unit as the dynamic timetable. In the timetable information management method according to claim 6 ,
A timetable information management method in which the route prediction timetable is generated using a predicted value according to the accuracy of the most frequent prediction error among the accuracy of each prediction error in the predetermined unit. The timetable information management method according to any one of claims 5 to 7 ,
Generate a plurality of predicted values according to the predicted value in time series,
A timetable information management method that identifies an approximation line based on the plurality of predicted values, and calculates a difference between the specified approximation line and the plurality of predicted values as the prediction error. An operation guidance system that generates dynamic timetable information indicating operation status when the operation schedule of a transportation facility is changed,
an input means for accepting input of operation data of the transportation facility;
Predicted value generation means for generating a predicted value indicating the degree of influence of the change in the operation schedule by executing a prediction process on the operation status for each predetermined unit constituting the transportation facility based on the input operation data; ,
Prediction error calculation means for calculating a prediction error regarding the accuracy of the prediction process;
Dynamic timetable information generation means for generating the dynamic timetable information based on a predicted value specified from the generated predicted values according to the prediction error ;
Predicted value threshold storage means for storing a predetermined predicted value threshold;
disclosure determining means for comparing the prediction error with the prediction value threshold and identifying the accuracy of the prediction error as the accuracy of the prediction process;
comprising a schedule timetable generating means for generating a plurality of predicted timetables for each of the predetermined units including the specified predicted values;
The dynamic timetable information generating means generates a dynamic timetable as the dynamic timetable information using a predicted timetable specified according to the accuracy specified by the disclosure determining means according to predetermined conditions. A traffic guidance system that is a means of generating dynamic timetables .

Images