Warning prediction device, warning prediction method, and warning prediction program

Abstract

To appropriately determine the probability of traffic restrictions occurring using quantitative rainfall forecast information. [Solution] The issuance prediction device 100 according to this embodiment has a reception unit 121, an acquisition unit 122, a calculation unit 123, and a determination unit 124. The reception unit 121 receives information on rainfall observation points to be predicted. The acquisition unit 122 acquires ensemble rainfall prediction amounts for predetermined time intervals, generated using ensemble information on multiple rainfall predictions at rainfall observation points received by the reception unit 121 and rainfall observation history recorded at the rainfall observation points. The calculation unit 123 calculates the probability that the ensemble rainfall prediction amount acquired by the acquisition unit 122 exceeds the standard value of rainfall set for each rainfall observation point. The determination unit 124 determines the possibility that information regarding railway operation restrictions at rainfall observation points will be issued based on the probability calculated by the calculation unit 123.

Description

Technical Field

[0001] The present invention relates to a departure prediction device, a departure prediction method, and a departure prediction program.

Background Art

[0002] Conventionally, when it is predicted that a disaster such as a typhoon that hinders railway operation will occur, a planned train suspension is carried out in advance to plan a train suspension or the like. In particular, in railways with a large number of passengers, since the impact on passengers depends on whether the railway operation is possible or not, it is required to accurately plan a planned train suspension at an early stage.

[0003] In such a situation, as a conventional technique, when disaster prediction information is received, a technique that enables the formulation of an appropriate operation plan by calculating the probability of occurrence of an operation restriction and creating a proposed operation plan change has been disclosed (for example, see Patent Document 1).

[0004] Also, regarding rainfall prediction, a technique called ensemble rainfall prediction has been disclosed (for example, see Non-Patent Document 1). Ensemble rainfall prediction is a method of making a plurality of predictions using a plurality of initial values with slight variations and predicting the most likely phenomenon and the worst scenario, and it also enables estimation of the possibility and degree of change in the prediction.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Non-Patent Documents

[0006]

Non-Patent Document 1

[0007] However, conventional technology cannot appropriately determine the probability of train operation restrictions using quantitative rainfall forecast information. For example, conventional technology uses the probability of disaster occurrence notified from an external source to decide on changes to the operation plan, and cannot accurately calculate the probability of train operation restrictions from rainfall forecasts at observation points. Furthermore, conventional technology cannot appropriately determine the probability of train operation restrictions at a target location using the calculated ensemble rainfall forecast amount. [Means for solving the problem]

[0008] To solve the above-mentioned problems and achieve the objective, the present invention's warning prediction device is characterized by comprising: a reception unit that receives information on rainfall observation points to be predicted; an acquisition unit that acquires an ensemble rainfall prediction amount for predetermined time intervals, generated using ensemble information relating to multiple rainfall predictions at rainfall observation points received by the reception unit and rainfall observation history recorded at the rainfall observation points; a calculation unit that calculates the probability that the ensemble rainfall prediction amount acquired by the acquisition unit exceeds a standard value for rainfall set for each of the rainfall observation points; and a determination unit that determines the possibility of issuing information regarding railway operation restrictions at rainfall observation points based on the probability calculated by the calculation unit. [Effects of the Invention]

[0009] According to the present invention, the probability of traffic restrictions occurring can be appropriately determined using quantitative rainfall forecast information. [Brief explanation of the drawing]

[0010] [Figure 1] Figure 1 shows an example of an order prediction system according to an embodiment. [Figure 2] Figure 2 is a block diagram showing an example configuration of a warning prediction device according to an embodiment. [Figure 3] Figure 3 shows an example of ensemble rainfall forecast according to the embodiment. [Figure 4] Figure 4 shows an example of a prediction confidence rank according to the embodiment. [Figure 5] Figure 5 shows an example of the correlation between the prediction confidence rank and the likelihood of issuing operational restrictions during rainfall, according to the embodiment. [Figure 6] Figure 6 shows an example of the process for determining the threshold of the prediction confidence rank according to the accuracy rate according to the embodiment. [Figure 7] Figure 7 is an example of an image showing routes with a high prediction confidence rank according to the embodiment. [Figure 8] Figure 8 is an example of an image showing the prediction confidence rank for each section according to the embodiment, displayed in chronological order. [Figure 9] Figure 9 shows a specific example of the process for obtaining ensemble rainfall forecasts according to the embodiment. [Figure 10] Figure 10 shows a specific example of the calculation process for the probability of exceeding the operational shutdown threshold according to the embodiment. [Figure 11] Figure 11 is a diagram showing a specific example of the process for determining the possibility of issuing operational restrictions during rainfall according to the embodiment. [Figure 12] Figure 12 is a flowchart showing the processing flow performed by the issuance prediction device according to the embodiment. [Figure 13] Figure 13 shows an example of a computer running a program. [Modes for carrying out the invention]

[0011] Hereinafter, embodiments of the issuance prediction apparatus, issuance prediction method, and issuance prediction program according to the present application will be described in detail based on the drawings. Note that the issuance prediction apparatus, issuance prediction method, and issuance prediction program according to the present application are not limited by this embodiment. Also, in the description of the drawings, the same parts are denoted by the same reference numerals, and duplicate descriptions are omitted as appropriate.

[0012] 〔First〕 (Outline of Issuance Prediction System 10) First, an outline of the issuance prediction system 10 according to this embodiment will be described. FIG. 1 is a diagram showing an example of the issuance prediction system according to the embodiment. The issuance prediction system 10 shown in FIG. 1 includes an issuance prediction apparatus 100 that determines the possibility of issuing a driving restriction during rainfall from an ensemble prediction of rainfall amounts at an observation point, and weather information distribution servers 200-1 and 200-2 that transmit an ensemble prediction of rainfall amounts at the observation point to the issuance prediction apparatus. Note that the number of weather information distribution servers 200 shown in FIG. 1 is not particularly limited and can be increased or decreased.

[0013] The issuance prediction apparatus 100 is an information processing apparatus that determines the possibility of issuing a driving restriction during rainfall at an observation point based on rainfall prediction information acquired from the weather information distribution server 200, and is realized by a computer or the like. For example, the issuance prediction apparatus 100 acquires an ensemble prediction of rainfall amounts at an observation point to be predicted from the weather information distribution server 200, and determines that the possibility of issuing a driving restriction is high when the probability that the ensemble rainfall prediction amount exceeds a reference value is high.

[0014] The weather information distribution server 200 is an information processing apparatus that outputs an ensemble prediction of rainfall amounts at a specific rainfall observation point a predetermined time ahead, and is realized by a server apparatus or a cloud system. For example, the weather information distribution server 200 transmits a plurality of ensemble rainfall prediction amounts output by the ensemble prediction model for the rainfall observation point notified from the issuance prediction apparatus 100 to the issuance prediction apparatus 100.

[0015] Here, the ensemble rainfall forecast amount according to this embodiment will be described. The ensemble rainfall forecast amount according to this embodiment is the forecast amount of rainfall at a specific location at a predetermined time in advance, which is predicted by generating multiple initial values ​​by slightly varying the weather data at a specific location and then using each of these initial values ​​to predict the amount of rainfall at a predetermined time in advance (see Non-Patent Document 1).

[0016] Furthermore, the number of predicted data points generated by ensemble prediction depends on the prediction model performing the ensemble prediction; for example, prediction data points with 21 members or 51 members may be generated. Additionally, the prediction model performing the ensemble prediction has a maximum prediction time set, for example, 39 hours or 15 days ahead.

[0017] (Overview of the processing by the order prediction device 100) Next, the processing details of the weather forecasting device 100 according to this embodiment will be described. In view of the aforementioned problems, the weather forecasting device 100 according to this embodiment receives information on rainfall observation points to be predicted, and obtains an ensemble rainfall prediction amount for each predetermined time period, which is generated using ensemble information relating to multiple rainfall predictions at the received rainfall observation points and the rainfall observation history recorded at the rainfall observation points.

[0018] The issuance prediction device 100 then calculates the probability that the acquired ensemble rainfall forecast exceeds the rainfall threshold set for each rainfall observation point, and based on the calculated probability, determines the likelihood that information regarding railway operation restrictions at the rainfall observation point will be issued.

[0019] For example, the weather forecasting device 100 notifies the weather information distribution server 200 of the received rainfall observation points, and uses the ensemble rainfall forecast amount from the weather information distribution server 200's prediction model and the rainfall observation history of the rainfall observation points to obtain hourly ensemble rainfall forecast amounts for at least 48 hours from the present.

[0020] The issuance prediction device 100 then calculates the probability that, for example, out of all members of the acquired ensemble rainfall prediction amount, the number of members whose predicted rainfall exceeds the predetermined threshold for triggering operational restrictions at the rainfall observation point is. The issuance prediction device 100 then determines the likelihood of issuing information regarding railway operational restrictions at the rainfall observation point based on the result of determining which of the predetermined prediction confidence ranks the calculated probability corresponds to.

[0021] As a result, the warning prediction device 100 can appropriately determine the reliability of the probability of train service suspension based on the ensemble rainfall forecast from rainfall observation points, and can appropriately determine the probability of train service restrictions occurring using quantitative rainfall forecast information.

[0022] [Configuration of the warning prediction device 100] The following describes an example of the functional configuration of the above-described order prediction device 100. Figure 2 is a block diagram showing an example of the configuration of the order prediction device according to the embodiment. As shown in Figure 2, the order prediction device 100 has a communication unit 110, a control unit 120, and a storage unit 130.

[0023] The communication unit 110 is implemented, for example, by a NIC (Network Interface Card). For example, the communication unit 110 controls communication regarding various types of information exchanged with connected devices, and mediates the processing of the receiving unit 121 and the acquisition unit 122, which will be described later.

[0024] The memory unit 130 is implemented by a storage device such as RAM (Random Access Memory) or a hard disk. The memory unit 130 stores data and programs necessary for various processes performed by the control unit 120. The storage device may be implemented by a storage system installed outside the order prediction device 100. The memory unit 130 includes, for example, an operation termination reference value storage unit 131 and a prediction reliability rank storage unit 132.

[0025] The train operation suspension criterion value storage unit 131 stores the rainfall threshold value at which a train operation restriction is issued for each rainfall observation point, based on the relationship between the past rainfall measurements of each rainfall observation point and the information on the issuance of train operation restrictions for railways near the rainfall observation point. For example, if a train operation restriction was issued for nearby railways when a rainfall of 80 mm was observed at rainfall observation point A in the past, the train operation suspension criterion value for A is stored as 80 mm.

[0026] The prediction confidence rank storage unit 132 stores information about prediction confidence ranks set according to the likelihood of railway operation restrictions being issued. For example, the prediction confidence rank storage unit 132 stores thresholds for the probability of exceeding the probability of operation suspension (described later) for each rank classified according to the likelihood of operation suspension. Specifically, the prediction confidence rank storage unit 132 stores that prediction confidence rank 3, which indicates a situation with a high probability of operation suspension, is classified when the probability of exceeding the probability of operation suspension is between 35% and 100%.

[0027] The control unit 120 is implemented by a processor such as an integrated circuit (CPU), MPU, ASIC (Application Specific Integrated Circuit), or FPGA (Field Programmable Gate Array), which executes various programs stored in the internal memory of the command prediction device 100 using RAM or the like as the working area. In the example shown in Figure 2, the control unit 120 includes a reception unit 121, an acquisition unit 122, a calculation unit 123, a determination unit 124, and a display unit 125.

[0028] The reception unit 121 receives information on the rainfall observation points to be predicted. For example, the reception unit 121 receives the name and identification number of the rainfall observation point for which the possibility of issuing driving restrictions during rainfall is predicted.

[0029] The acquisition unit 122 acquires ensemble rainfall forecast amounts for predetermined time intervals, which are generated using ensemble information regarding multiple rainfall forecasts at rainfall observation points received by the reception unit 121 and the rainfall observation history recorded at the rainfall observation points.

[0030] For example, the acquisition unit 122 notifies the weather information distribution server 200 of the rainfall observation points received by the reception unit 121, and uses the ensemble information, which is the initial value obtained by varying the weather data of the rainfall observation points, and the rainfall observation records recorded as weather data of the rainfall observation points, to acquire the ensemble rainfall forecast amount for each hour from the present onward, which is output by the ensemble prediction model.

[0031] Here, using Figure 3, a specific example of the ensemble rainfall forecast amount acquired by the acquisition unit 122 will be explained. Figure 3 is a diagram showing an example of the ensemble rainfall forecast amount according to the embodiment. As shown in Figure 3, the acquisition unit 122 acquires the ensemble rainfall forecast amount shown in the line graph of the "forecast 21 member".

[0032] Specifically, the acquisition unit 122 acquires the ensemble rainfall forecast amounts of 21 members output by the prediction model for 39 hours from the present (11th - 9am) to 13th - 0am.

[0033] Furthermore, the acquisition unit 122 acquires the ensemble rainfall forecast amount calculated by the forecast model that has been switched according to the forecast time. For example, if the acquisition unit 122 is connected to an ensemble forecast model with a maximum forecast time of 39 hours ahead and an ensemble forecast model with a maximum forecast time of 15 days ahead, it can acquire the ensemble rainfall forecast amount from the present to at least 48 hours ahead by switching the forecast model used for the relatively short forecast time of up to 39 hours ahead and from 39 hours ahead to 48 hours ahead.

[0034] The calculation unit 123 calculates the probability that the ensemble rainfall forecast acquired by the acquisition unit 122 exceeds the rainfall threshold set for each rainfall observation point. For example, the calculation unit 123 calculates the probability of exceeding the threshold for train service suspension, which is the probability that the ensemble rainfall forecast exceeds the rainfall threshold set for each rainfall observation point, by dividing the number of members of the ensemble rainfall forecast that exceed the threshold set for the amount of rainfall at which train service restrictions are issued by the total number of members of the ensemble rainfall forecast.

[0035] Here, referring again to Figure 3, we will explain a specific example of how to calculate the probability of exceeding the limit for operation suspension. In Figure 3, the operation suspension threshold used when calculating the probability of exceeding the limit for operation suspension is stored as 80 mm, and an example is shown in which the probability of a member exceeding 80 mm in ensemble rainfall forecast is calculated.

[0036] As shown in Figure 3, the calculation unit 123 calculates that for the ensemble rainfall forecast 36 hours ahead (12th - 21:00), all ensemble rainfall forecasts for all 21 members exceed the operational shutdown threshold of 80 mm. Therefore, the probability of exceeding the operational shutdown threshold 36 hours ahead is calculated as 21 members / 21 members = 100%.

[0037] The determination unit 124 determines the likelihood of issuing information regarding railway operation restrictions at rainfall observation points based on the probability of exceeding the limit of operation calculated by the calculation unit 123. For example, the determination unit 124 classifies the probability of exceeding the limit of operation into multiple classes, which are set based on the accuracy rate of operation restrictions in past probability of exceeding the limit of operation and for which thresholds are set, and determines the likelihood of issuing information regarding railway operation restrictions.

[0038] The information regarding train operation restrictions determined by the determination unit 124 includes, for example, train service suspensions, speed restrictions, and simple warning judgments. By associating each piece of information with the respective ranks of the prediction reliability rank described later, it is possible to determine which type of train operation restriction information is most likely to be issued.

[0039] Here, with reference to Figure 4, we will explain the prediction confidence rank used to classify the probability of exceeding the probability of operational shutdown. Figure 4 is a diagram showing an example of the prediction confidence rank according to this embodiment. The prediction confidence rank consists of multiple ranks set according to the accuracy rate of past predictions of the probability of exceeding the probability of operational shutdown. In Figure 4, an example is shown in which the region where the probability of exceeding the probability of operational shutdown is 60% or higher is set as rank 3, the region where the probability of exceeding the probability of operational shutdown is 18-59% is set as rank 2, the region where the probability of exceeding the probability of operational shutdown is 3-17% is set as rank 1, and the region where the probability of exceeding the probability of operational shutdown is 2% or less is excluded.

[0040] The vertical axis of the graph in Figure 4, "(Actual) Operation Suspension Criteria Ratio," shows the ratio of the actual rainfall observed in the past to the operation suspension criteria at the rainfall observation point where that rainfall was observed. For example, if the operation suspension criteria ratio is 1.5, it means that the actual rainfall observed was 1.5 times the operation suspension criteria. As mentioned above, the operation suspension criteria is the rainfall threshold at which operation restrictions are issued. Therefore, if the operation suspension criteria ratio is 1.0 or higher, an operation suspension order was issued, and if it is less than 1.0, an operation suspension order was not issued. The horizontal axis of the graph in Figure 4, "(Predicted) Operation Suspension Excess Probability (%)," shows the operation suspension excess probability calculated by the calculation unit 123.

[0041] Specifically, the points plotted around 3.3 for the shutdown threshold ratio and around 42% for the shutdown excess probability indicate that, when the shutdown excess probability based on ensemble rainfall forecasts was 42%, the actual rainfall observed at the predicted time was 3.3 times the shutdown threshold ratio.

[0042] Next, we will explain the accuracy rate. As shown in Figure 4, the accuracy rate is calculated by dividing the number of predictions in the prediction confidence rank where the ratio of the (actual) operational shutdown threshold value to 1 or higher is 1 by the total number of predictions in the prediction confidence rank. For example, in rank 3, A represents the number of predictions where the ratio of the operational shutdown threshold value to 1 or higher and the train was actually shut down, and B represents the number of predictions where the ratio of the operational shutdown threshold value to 1 and the prediction was incorrect, and the accuracy rate for rank 3 is calculated as "Rank 3 accuracy rate = A / (A+B)". As shown in the table in Figure 4, the accuracy rate for rank 3 is 56%.

[0043] The prediction confidence rank can be used to classify the calculated probability of exceeding the probability of operational shutdown by confidence rank, for example, by setting a threshold for the probability of exceeding the probability of operational shutdown based on the accuracy rate calculated from past observation data.

[0044] Next, we will explain the relationship between the prediction confidence rank and the likelihood of issuing driving restrictions. Figure 5 is a diagram showing an example of the correlation between the prediction confidence rank and the likelihood of issuing driving restrictions during rainfall according to the embodiment.

[0045] As shown in Figure 5, the prediction confidence rank shown in Figure 4 is set so that the accuracy rate increases with each rank increase, from "not applicable" to rank 3. Therefore, a prediction confidence rank of "not applicable" indicates a low probability of issuance, a prediction confidence rank of rank 1 indicates a slightly low probability of issuance, a prediction confidence rank of rank 2 indicates a slightly high probability of issuance, and a prediction confidence rank of rank 3 indicates a high probability of issuance.

[0046] In the example shown in Figure 5, the determination unit 124 can determine that there is a high probability of issuing a rain-related operational restriction on the railway at the rainfall observation point if it classifies the calculated probability of exceeding the limit for suspension of operations to rank 2 or rank 3, which are ranks where the probability of issuance is somewhat high or highly likely.

[0047] Note that the example in Figure 5 shows an example where all prediction confidence ranks are associated with the likelihood of issuing operational restrictions, but it is not limited to this. For example, when rank 2 is set to indicate a high probability of issuing speed restrictions, if rank 2 is also assigned to the probability of exceeding the limit for stopping operations, the determination unit 124 determines that there is a high probability of issuing railway speed restrictions at the rainfall observation point.

[0048] Here, the warning prediction device 100 can easily generate a prediction confidence rank by automatically extracting a threshold for the probability of exceeding the probability of train cancellation that results in a certain accuracy rate, based on statistical data that has accumulated the correlation between the probability of exceeding the probability of train cancellation calculated by the calculation unit 123 and the actual accuracy rate of train operation restrictions during rainfall.

[0049] Figure 6 shows an example of the process for determining the threshold of the prediction confidence rank according to the accuracy rate according to the embodiment. The issuance prediction device 100 automatically extracts a threshold that satisfies the set accuracy rate using the statistical data shown in the graph in Figure 6. Below, an example of the issuance prediction device 100 automatically extracting a threshold for rank 3 will be described.

[0050] The warning prediction device 100 first obtains a pre-set accuracy rate of 60% for rank 3. Then, the warning prediction device 100 identifies the range of excess operation probability from 35% to 100% where the accuracy rate is 60%, from a table showing the relationship between the range of excess operation probability set in 5% increments and the accuracy rate of past predicted cases within that range. The warning prediction device 100 then extracts the identified range of excess operation probability from 35% to 100% as the threshold for rank 3.

[0051] The warning prediction device 100 then identifies a range of 10% to 100% of the probability of exceeding the threshold for operation shutdown, for example, where the accuracy rate for rank 2 is a predetermined 30%. The warning prediction device 100 then extracts 10% to 34% of the identified 10% to 100% range, excluding the 35% to 100% threshold for rank 3, as the threshold for rank 2.

[0052] The warning prediction device 100 can automatically extract thresholds for each rank through the processing described above and easily create confidence ranks that satisfy a pre-set accuracy rate. Note that the process of automatically extracting thresholds for each rank shown in Figure 6 is merely an example, and the warning prediction device 100 can also perform the series of processes described above by dividing the range of the probability of exceeding the operational shutdown limit, which is set in steps of 5%, into arbitrary percentages such as 1% increments.

[0053] The display unit 125 displays, in chronological order, the routes related to rainfall observation points where the probability of exceeding the limit for train service suspension is classified into a rank with a higher accuracy rate than a predetermined value. For example, the display unit 125 refers to information on the correspondence between the rainfall observation points being predicted and the routes running near them, and displays the routes running near rainfall observation points where the probability of exceeding the limit for train service suspension is classified into rank 2 or rank 3, which has a relatively high probability of issuing train service restrictions during rainfall.

[0054] Here, with reference to Figure 7, an example of the route map display by the display unit 125 will be described. Figure 7 is a diagram showing an example of an image displaying routes with a high prediction reliability rank according to the embodiment. As shown in Figure 7, the display unit 125 can change the route display for each route within the prediction target area according to the reliability rank of the nearby rainfall observation point.

[0055] Specifically, the display unit 125 can indicate routes with a high probability of being ordered to operate during rainfall by showing routes near rank 2 and routes near rank 1 in corresponding display modes. In addition, the display unit 125 indicates that the rainfall observation points near the "Sobu Main Line," "Uchibo Line," "Sotobo Line," "Narita Line," "Kashima Line," and "Kururi Line" within the prediction range are rank 2 or rank 3.

[0056] This allows the display unit 125 to easily identify routes where there is a high probability of operational restrictions being imposed during rainfall at any given time.

[0057] Next, with reference to Figure 8, an example of a display showing the reliability rank for each section connecting specific points by the display unit 125 in chronological order will be described. Figure 8 is an example of an image showing the predicted reliability rank for each section according to the embodiment in chronological order.

[0058] As shown in Figure 8, the display unit 125 refers to the type of prediction confidence rank that classifies the probability of exceeding the operational suspension limit at rainfall observation points in the section connecting any two points, and displays the prediction confidence rank for any given section in chronological order. Specifically, for the "Yugawara" section, the display unit 125 displays rank 2 from 15:00 on May 28 to 4:00 on May 29, rank 1 from 5:00 to 6:00 on May 29, and excludes the period from 7:00 on May 29 onwards.

[0059] This allows the display unit 125 to understand the time-series changes in the prediction confidence rank for the interval within the prediction range.

[0060] [Specific examples] Next, with reference to Figures 9 to 11, a specific example of a series of processes performed by the rainfall prediction device 100 according to the embodiment will be described. In the specific example described below, the rainfall observation point A is received by the reception unit 121, and the process up to the determination of the possibility of issuing a rainfall-related operational restriction for 5 / 28-12:00 will be described.

[0061] Figure 9 shows a specific example of the process for acquiring ensemble rainfall forecasts according to the embodiment. The acquisition unit 122 notifies the weather information distribution server 200 of the received rainfall observation point A and acquires the ensemble rainfall forecasts for each member for each time period up to 48 hours ahead, generated by the weather model. Specifically, the acquisition unit 122 acquires that for member 1, the forecast is 80 mm for 5 / 28-10:00, 100 mm for 5 / 28-11:00, 130 mm for 5 / 28-12:00, and so on.

[0062] Figure 10 shows a specific example of the calculation process for the probability of exceeding the operational shutdown threshold according to the embodiment. The calculation unit 123 calculates the probability of exceeding the operational shutdown threshold of 33% for the ensemble rainfall forecast amount for 5 / 28-12:00, which is the time period to be judged, and which is the probability that the forecast amount exceeds the operational shutdown threshold of 100 mm. Specifically, the calculation unit 123 calculates the probability of exceeding the operational shutdown threshold of 33% for rainfall observation point A for 5 / 28-12:00 by using the formula "7 members / 21 members = 33%".

[0063] Figure 11 shows a specific example of the process for determining the possibility of issuing operational restrictions during rainfall according to the embodiment. The determination unit 124 classifies the calculated operational suspension probability of 33% into prediction confidence rank 2, where the threshold for operational suspension probability is 10-35%, and determines that the possibility of issuing operational restrictions during rainfall at rainfall observation point A from 12:00 on 5 / 28 is somewhat high.

[0064] As a result, the rainfall forecasting device 100 can determine the possibility of issuing a rainfall-related operational restriction based on the ensemble rainfall forecast amount at any given time at any given rainfall observation point, thereby enabling it to appropriately determine the probability of operational restrictions occurring using quantitative rainfall forecast information.

[0065] 〔flowchart〕 Next, an example of processing by the warning prediction device according to this embodiment will be described with reference to Figure 12. Figure 12 is a flowchart showing the flow of processing performed by the warning prediction device according to this embodiment. Note that each step in the flowchart shown in Figure 12 can be executed in a different order, and additional processing may be added or omitted.

[0066] First, the weather forecasting device 100 receives information from the rainfall observation points to be predicted (S101). Then, the weather forecasting device 100 uses the ensemble information of the rainfall observation points and the rainfall observation history to obtain the ensemble rainfall prediction amount for predetermined time intervals (S102). Then, the weather forecasting device 100 switches the prediction model according to the time of the prediction (S103).

[0067] The warning prediction device 100 then calculates the probability of exceeding the operational suspension threshold, which is the probability that the ensemble rainfall forecast exceeds the operational suspension threshold (S104). The warning prediction device 100 then determines the possibility of issuing operational restrictions during rainfall according to the prediction confidence rank into which the operational suspension excess probability was classified (S105), and then terminates the process.

[0068] 〔effect〕 The rainfall prediction device 100 according to this embodiment includes a reception unit 121, an acquisition unit 122, a calculation unit 123, and a determination unit 124. The reception unit 121 receives information on rainfall observation points to be predicted. The acquisition unit 122 acquires ensemble rainfall prediction amounts for predetermined time intervals, which are generated using ensemble information on multiple rainfall predictions at rainfall observation points received by the reception unit 121 and rainfall observation history recorded at the rainfall observation points.

[0069] The calculation unit 123 calculates the probability that the ensemble rainfall forecast acquired by the acquisition unit 122 exceeds the rainfall threshold set for each rainfall observation point. Based on the probability calculated by the calculation unit 123, the determination unit 124 determines the possibility of railway operation restrictions being issued at the rainfall observation points.

[0070] As a result, the rainfall forecasting device 100 can determine the possibility of issuing a rainfall-related operational restriction based on the ensemble rainfall forecast amount at any given time at any given rainfall observation point, thereby enabling it to appropriately determine the probability of operational restrictions occurring using quantitative rainfall forecast information.

[0071] Furthermore, the acquisition unit 122 acquires the ensemble rainfall forecast amount calculated by the prediction model that is switched according to the predicted time. As a result, the issuance prediction device 100 can acquire the ensemble rainfall forecast amount output by the prediction model with the highest accuracy for the predicted time.

[0072] Furthermore, the calculation unit 123 divides the number of members of the ensemble rainfall forecast that exceed the set threshold value for the amount of rainfall that triggers the issuance of operational restrictions by the total number of members of the ensemble rainfall forecast. This allows the issuance prediction device 100 to appropriately calculate the probability that operational suspension due to rainfall will occur at the predicted time, using the quantitative rainfall forecast output by the prediction model.

[0073] Furthermore, the determination unit 124 classifies the probability into multiple classes, which are set based on the accuracy rate of operation restrictions in past probabilities, and which have thresholds set for the probability. As a result, the issuance prediction device 100 can easily determine the likelihood of issuing operation restrictions during rainfall that corresponds to the prediction confidence rank into which the probability of exceeding the operation suspension is classified.

[0074] Furthermore, the warning prediction device 100 further includes a display unit 125. The display unit 125 displays, in chronological order, the routes related to rainfall observation points whose probability is classified into a category with a hit rate higher than a predetermined value. This allows the warning prediction device 100 to easily identify routes near rainfall observation points that have been determined to have a relatively high probability of issuing traffic restrictions during rainfall as routes where traffic restrictions are likely to be issued.

[0075] 〔others〕 Furthermore, among the processes described in the above embodiments, all or part of those described as being performed automatically may be performed manually. Conversely, all or part of those described as being performed manually may be performed automatically by known methods. In addition, the processing procedures, specific names, and various data and parameters shown in the above documents and drawings may be changed at will unless otherwise specified.

[0076] Furthermore, each component of the illustrated device is a functional concept and does not necessarily have to be physically or functionally configured as shown. In other words, the specific forms of distribution and integration of each device and functional configuration are not limited to those shown, and all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions. In addition, the embodiments described above and the processes carried out in each embodiment may be combined as appropriate, as long as the processing content is not contradictory.

[0077] 〔program〕 Furthermore, the order prediction device 100 according to the above-described embodiment is implemented by a computer 1000 having a configuration such as that shown in Figure 13. Figure 13 is a diagram showing an example of a computer that executes a program. The computer 1000 is connected to an output device 1010 and an input device 1020, and has a configuration in which a central processing unit 1030, memory 1040, storage 1050, output IF (Interface) 1060, input IF 1070, and communication interface 1080 are connected by a bus 1090.

[0078] The central processing unit 1030 operates based on programs stored in the memory 1040 and storage 1050, as well as programs read from the input device 1020, and executes various processes. The memory 1040 is a memory device, such as RAM, that temporarily stores data used by the central processing unit 1030 for various calculations. The storage 1050 is a storage device where data used by the central processing unit 1030 for various calculations and various databases are registered, and is implemented using ROM (Read Only Memory), HDD (Hard Disk Drive), flash memory, etc.

[0079] Output IF1060 is an interface for transmitting information to be output to output devices 1010 that output various types of information, such as monitors and printers. It can be implemented using connectors of standards such as USB (Universal Serial Bus), DVI (Digital Visual Interface), or HDMI (High Definition Multimedia Interface). Input IF1070 is an interface for receiving information from various input devices 1020, such as mice, keyboards, and cameras. It can be implemented using USB, for example. Input devices 1020 may also be devices that read information from optical recording media, magneto-optical recording media, tape media, magnetic recording media, or semiconductor memory, or external storage media such as USB memory.

[0080] The communication interface 1080 receives data from other devices via the network N and sends it to the central processing unit 1030, and also transmits data generated by the central processing unit 1030 to other devices via the network N. The central processing unit 1030 controls the output device 1010 and input device 1020 via the output IF 1060 and input IF 1070. For example, the central processing unit 1030 loads a program from the input device 1020 or storage 1050 into memory 1040 and executes the loaded program.

[0081] Although some embodiments of the present invention have been described above with reference to the drawings, these are illustrative examples, and the present invention may be implemented in other forms by various modifications and improvements based on the knowledge of those skilled in the art, starting with the embodiments described in the disclosure section of the invention. Furthermore, the terms "section, module, unit" used above can be read as "means" or "circuit," etc. For example, the "applying part" can be read as the "applying means" or "applying circuit." [Explanation of Symbols]

[0082] 100 Warning Prediction Device 110 Communications Department 120 Control Unit 121 Reception Department 122 Acquisition Department 123 Calculation Section 124 Judgment section 130 Storage section 200, 200-1, 200-2 Weather Information Distribution Servers

Claims

1. A reception desk that receives information from rainfall observation points targeted for prediction, An acquisition unit that acquires an ensemble rainfall forecast amount for a predetermined time interval, generated using ensemble information relating to multiple rainfall forecasts at the rainfall observation points received by the reception unit and the rainfall observation history recorded at the rainfall observation points. A calculation unit calculates the probability that the ensemble rainfall forecast amount acquired by the acquisition unit exceeds the rainfall threshold value set for each of the rainfall observation points, A determination unit determines the possibility that information regarding railway operation restrictions at the rainfall observation point will be issued based on the probability calculated by the calculation unit, An order prediction device characterized by having the following features.

2. The issuance prediction device according to claim 1, characterized in that the acquisition unit acquires the ensemble rainfall forecast amount calculated by a prediction model that is switched according to the predicted time.

3. The issuance prediction device according to claim 1, characterized in that the calculation unit divides the number of members of the ensemble rainfall prediction amount that exceeds the standard value set as the amount of rainfall for which the driving restriction is issued by the total number of members of the ensemble rainfall prediction amount.

4. The determination unit classifies the probability into a plurality of classes, which are classes set based on the accuracy rate of the driving restrictions in past probabilities, and which are set by setting a threshold for the probability, as described in claim 1.

5. The issuance prediction device according to claim 4, further comprising a display unit that displays in time series the routes relating to the rainfall observation points in which the probability is classified into the class in which the accuracy rate is higher than a predetermined value.

6. A method for predicting the issuance of an order, which is performed by an order issuance prediction device, The process involves receiving information from rainfall observation points that are the target of the prediction, An acquisition step to acquire an ensemble rainfall forecast amount for a predetermined time interval, which is generated using ensemble information relating to multiple rainfall forecasts at the rainfall observation points received in the aforementioned acceptance step and the rainfall observation history recorded at the rainfall observation points. A calculation step for calculating the probability that the ensemble rainfall forecast obtained by the acquisition step exceeds the rainfall threshold set for each of the rainfall observation points, A determination step, based on the probability calculated by the calculation step, determines the possibility that information regarding railway operation restrictions at the rainfall observation point will be issued, A method for predicting the issuance of an order, characterized by including the following:

7. The procedure for receiving information from rainfall observation points to be included in the prediction, An acquisition procedure for acquiring an ensemble rainfall forecast amount for a predetermined time interval, which is generated using ensemble information relating to multiple rainfall forecasts at the rainfall observation point received through the above-mentioned acceptance procedure and the rainfall observation history recorded at the rainfall observation point. A calculation procedure for calculating the probability that the ensemble rainfall forecast obtained by the acquisition procedure exceeds the rainfall threshold set for each of the rainfall observation points, A determination procedure for determining the possibility of issuing information regarding railway operation restrictions at the rainfall observation point based on the probability calculated by the calculation procedure, A program for predicting the issuance of an emergency order, characterized by having a computer execute it.

Images