JP2021098425A - Operation forecasting system and method for the same - Google Patents

Abstract

To provide an operation forecasting system which can generate a prediction diagram that has some latitude in both a departure time and an arrival time.SOLUTION: An operation forecasting system in this invention comprises a storage part to store operation performance data indicating an operation performance of a movable body, at least one of a time when the movable body arrives at a prescribed location and a time when the movable body departs the prescribed location being included in the operation performance data, an acquisition part to acquire the operation performance data of an operation state similar to an operation state of the movable body of a prediction object from the storage part, a calculation part to calculate at least one of both a latitude in an arrival time indicating an arrival of the movable body of the prediction object with a constant reliability and a latitude in a departure time indicating a departure of the movable body of the prediction object with a constant reliability, at a prescribed location based on the operation performance data acquired by the acquisition part and an output part to output a result calculated by the calculation part.SELECTED DRAWING: Figure 1

Description

The present invention relates to an operation prediction system and an operation prediction method, and is suitable for being applied to, for example, an operation prediction system and an operation prediction method for predicting the operation of a moving object to be operated.

In the operation management system represented by trains, forecast schedules are created and used to make operational prospects. The forecast timetable predicts the arrival and departure times of each train at each station at a time earlier than the current time, based on the actual arrival and departure times before the current time, and the arrival time of each train. Is calculated as a specific time.

In addition to making a forecast of operation, the forecast schedule is for the person in charge (hereinafter sometimes referred to as "station staff") who monitors the train and guides passengers at each station according to the arrival or departure of the train. It is also used as a reference for business execution. In the work of station staff, if the arrival and departure times of trains to each station are different, it may interfere with the arrival preparation of trains in the station premises, passenger guidance, etc., so it is necessary to have a forecast schedule that is as accurate as possible. ..

Traditional forecast schedules have been calculated assuming that the train arrives or departs at the station of interest at exactly a particular time. However, train arrival and departure times depend on variable factors such as the speed of travel between stations, the number of passengers getting on and off at each station, and the occupancy rate. The forecast (hereinafter referred to as "operation forecast") is likely to deviate from the actual operation (hereinafter referred to as "operation record"). In particular, when there is a delay in train operation, the discrepancy between the operation forecast and the operation record tends to be remarkable. In order to accurately grasp the forecast schedule of trains, in addition to predicting arrival and departure times, it is necessary to have a mechanism that can grasp how early or late the train may arrive or arrive late.

Regarding the method of simultaneously calculating and utilizing the deviation between the operation prediction and the operation record, the on-line probability distribution is calculated for the preceding train and the following train, and the results are superimposed to calculate the collision probability on the ground for position correction. A design simulator for designing the installation position of the child has been proposed (see Patent Document 1).

In addition, train operation management that calculates multiple forecast schedules up to the maximum delay width that can be expected when a delay occurs, and lets the commander select a forecast schedule that takes into account the events that should be prioritized and considered at present. A display system has been proposed (see Patent Document 2).

Japanese Unexamined Patent Publication No. 2016-002777 Japanese Unexamined Patent Publication No. 2001-001905

When a train is delayed, the prediction of the arrival and departure time of the train at each station cannot be uniquely determined, and there is a certain probability that the train will shift back and forth. If the predicted arrival / departure time of a certain train deviates, it may affect the following train. Therefore, it is desirable to correct the predicted timetable of the following train depending on the predicted timetable and the value of the expected deviation width. That is, in order to grasp the forecast schedule of the train more accurately, it is effective to create a forecast schedule including the expected time lag width and consider the influence on the following train.

In the design simulator described in Patent Document 1, an appropriate correction ground element installation position is calculated from the line probability between two trains, but the arrival time can be adjusted according to the line probability distribution of the preceding train, or the following train can be used. Do not correct the prediction timetable.

Further, in the train operation management display system described in Patent Document 2, the possibility that the prediction is deviated, the event that the following train is delayed due to the delay of the preceding train, and the like are not taken into consideration. The method of selecting the prediction pattern also depends on the experience and feeling of the commander.

The present invention has been made in consideration of the above points, and an object of the present invention is to propose an operation prediction system or the like capable of creating a prediction timetable having a range of arrival and departure times.

In order to solve such a problem, in the present invention, it is an operation prediction system that predicts the operation of a moving body to be operated, and is a storage unit that stores operation record data indicating the operation record of the moving body, and the operation record data. Includes at least one of the time when the moving body arrives at a predetermined point and the time when the moving body departs from the predetermined point, and the operation record of the operation status similar to the operation status of the mobile body to be predicted. Based on the acquisition unit that acquires data from the storage unit and the operation record data acquired by the acquisition unit, it indicates that the mobile object to be predicted arrives at the predetermined point with a certain degree of reliability. A calculation unit that calculates at least one of the arrival time width and the departure time width that indicates that the mobile object to be predicted departs with a certain degree of reliability, and an output that outputs the result calculated by the calculation unit. I tried to provide a part and.

In the above configuration, the arrival time or departure time of the moving object at a predetermined point is calculated as a prediction (arrival time width or departure time width) having a time width.

In the above configuration, by calculating and outputting the arrival time width or the departure time width in consideration of the deviation that may occur in the arrival time or the departure time of the moving body, for example, a prediction with a width in the arrival / departure time. You can create a diamond. Further, for example, it is possible to assist the work of the person in charge of considering the subsequent operation reorganization plan, the work of the staff member who confirms the arrival of the moving body at the place where the moving body stops, opens / closes the door, guides passengers, and the like.

In particular, even if it is difficult to predict the subsequent operation of the moving object due to delays caused by bad weather, accidents, equipment inspections, etc., the arrival time width and / or departure time width can be used based on past operation results. In addition, it is possible to make a certain degree of operational outlook, which can be useful for facilitating the work of the person in charge, staff, etc.

According to the present invention, it is possible to realize a highly accurate prediction diamond that reflects the interaction between trains.

It is a figure which shows an example of the structure which concerns on the operation prediction system by 1st Embodiment. It is a figure which shows an example of the operation performance data by 1st Embodiment. It is a figure which shows an example of the situation data between stations by 1st Embodiment. It is a figure which shows an example of the function which the operation prediction system by 1st Embodiment has. It is a figure which shows an example of the distribution of the similar operation performance data by 1st Embodiment. It is a figure which shows an example of the flowchart which concerns on the correction processing by 1st Embodiment. It is a figure which shows an example of the aspect which outputs the operation prediction distribution by 1st Embodiment. It is a figure which shows an example of the timetable display by 1st Embodiment. It is a figure which shows an example of the timetable display by 1st Embodiment. It is a figure which shows an example of the timetable display by 1st Embodiment. It is a figure which shows an example of the timetable display by 1st Embodiment. It is a figure which shows an example of the timetable display by 1st Embodiment. It is a figure which shows an example of the timetable display by 1st Embodiment. It is a figure which shows an example of the timetable display by 1st Embodiment. It is a figure which shows an example of the timetable display by 1st Embodiment. It is a figure which shows the concept of the operation prediction distribution correction formula by 1st Embodiment. It is a figure which shows a part of the structure of the operation prediction system by 2nd Embodiment. It is a figure which shows an example of the function provided in the operation prediction system by 2nd Embodiment. It is a figure which shows an example of the table by 2nd Embodiment. It is a figure which shows an example of the relationship between the confidence interval and the operation prediction distribution by the 2nd Embodiment.

Hereinafter, one embodiment of the present invention will be described in detail with reference to the drawings. The operation prediction system shown in the present embodiment has, for example, a distribution showing how much the arrival / departure time predicted for a certain train and the arrival / departure time deviate from each other with a certain probability or more based on the past operation results. Calculate and output a forecast with a time range.

According to the embodiment described below, the arrival time and / or departure time of the train to be predicted to each station is predicted from the past operation results with a time range (hereinafter referred to as "operation prediction distribution"). It can be calculated as).

The calculated operation forecast distribution will be used for train arrival preparation, passenger guidance, etc. at each station by displaying it on top of the conventional forecast schedule that assumes that the train will arrive and / or depart at a specific time. be able to.

In addition, by considering the overlap between the operation prediction distributions of the train arrival time at each station and / or the operation prediction distributions of the train departure time at each station, the operation interval, stop time, and running speed of the front and rear trains are taken into consideration. Etc. can be corrected to optimize the operation of the train when a train delay occurs.

In the following, the calculation method and display method of the statistical operation prediction distribution in railways will be described, but the present invention is not limited to this. Each method is widely applicable to systems in which it is necessary to create a subsequent operation forecast using the operation results for the train operation plan. For example, instead of trains, it can be applied to transportation means such as aircraft and buses.

(1) First Embodiment In FIG. 1, 100 indicates an operation prediction system according to the first embodiment as a whole.

FIG. 1 is a diagram showing an example of a configuration related to the operation prediction system 100. The operation prediction system 100 includes a calculation device 110, an operation plan import device 120, a display device 130, and a storage device 140. The calculation device 110, the operation plan acquisition device 120, the display device 130, and the storage device 140 are communicably connected via the network 150.

The computing device 110 is composed of, for example, a general-purpose computer. The calculation device 110 includes an operation prediction distribution calculation program 111, operation record data 112, and station-to-station status data 113. The operation prediction distribution calculation program 111 is stored in, for example, a storage device (not shown) included in the calculation device 110.

The operation prediction distribution calculation program 111 uses the operation record data 112 to predict the future (future) operation status of the train, and how much the prediction deviates when the past operation results are statistically viewed. This is a program for creating an operation prediction distribution with a time range for which the data is calculated. Further, the operation prediction distribution calculation program 111 is a program that uses the station-to-station situation data 113 to correct the prediction schedule and the operation prediction distribution according to the congestion status of each station, the wind speed between stations, and the like.

The operation record data 112 is data in which the operation record of the past train and the conditions at that time (hereinafter, referred to as “operation conditions”) are accumulated up to the present time. The details of the operating conditions will be described later with reference to FIG. 2, but refer to various conditions related to the operation of the train.

Operating conditions can be freely added, changed, deleted, accumulated, and utilized by the user. Regarding changes in operating conditions, the system may select conditions that have a large impact when creating forecasts, and may automatically add, change, or delete them.

The inter-station status data 113 includes congestion status of platforms at each station, presence / absence of specific passengers (for example, passengers in wheelchairs) that require time to get on / off at each station, wind speed between stations, presence / absence of speed restrictions between stations, and the like. It is data showing the current situation of each station related to the operation and the current situation between stations. The inter-station status data 113 will be described later with reference to FIG.

The status items acquired as data can be freely added, changed, or deleted by the user. Alternatively, the system may select conditions that have a large impact when creating a forecast, and automatically add, change, or delete them.

The operation plan import device 120 is a device that acquires operation plan data indicating a train operation plan. In the present embodiment, the operation plan import device 120 is a device that acquires operation plan data from the outside of the operation prediction system 100. For example, the operation plan import device 120 may be a device that reads the operation plan data from the recording medium on which the operation plan data is recorded.

In addition, instead of the operation plan import device 120, or in addition to the operation plan import device 120, a device that receives the operation plan data directly from the system having the operation plan data or via the network 150 may be used. Good. The system having the operation plan data is, for example, an operation plan creation system that creates the operation plan data, an operation management system that manages the train operation based on the operation plan data, and the like.

The display device 130 displays the prediction timetable and the operation prediction distribution which are the calculation results of the calculation device 110. The display device 130 includes a display 131, an input device 132, a screen display program 133, and an input content processing program 134. The screen display program 133 and the input content processing program 134 are stored in, for example, a storage device (not shown) included in the display device 130.

The display 131 is a display device capable of displaying various screen information. The input device 132 is a device that receives various information from the user who uses the operation prediction system 100. The input device 132 is a general-purpose input device for a computer such as a mouse, a keyboard, and a touch panel.

The screen display program 133 is a program for generating screen information to be displayed on the display 131. The input content processing program 134 is a program for processing input information which is information received by the input device 132.

The storage device 140 is composed of, for example, a general-purpose computer or a storage device. The storage device 140 is a device that stores the prediction diamond and the operation prediction distribution, which are the calculation results of the calculation device 110. The storage device 140 may store information other than the prediction timetable and the operation prediction distribution.

The configuration shown in FIG. 1 is an example and is not limited to this configuration. For example, the operation prediction system 100 may be realized by a single computer having the functions of the calculation device 110, the operation plan acquisition device 120, the display device 130, and the storage device 140. Further, for example, the inter-station status data 113 may be captured when the data from the past to the present is stored in an external storage device and used for the calculation of the operation prediction distribution.

Further, for example, when it is not necessary for a human to confirm the calculation result of the calculation device 110, the display device 130 may not be provided. A case where it is not necessary for a human to confirm the calculation result of the calculation device 110 includes a case where the calculation device 110 directly transmits the calculation result to a system that automatically creates an operation plan.

FIG. 2 is a diagram showing an example of operation record data 112 (operation record table 200).

The operation record table 200 stores the information of the operation condition 202 and the information of the operation record 203 in association with each train number 201.

The train number 201 is identification information that can identify the train.

The operation condition 202 is a condition that can affect the operation of the train according to the schedule, and refers to various conditions related to the operation of the train. The operating conditions 202 include, for example, day of the week 211, weather 212, operating time zone 213, delay degree 214, vehicle formation 215, speed limit in a certain section (speed type 216), deterrence 217, and the like.

Day of the week 211 indicates which day of the week the train with train number 201 is operating from Sunday to Saturday. Weather 212 refers to weather information including wind speed, temperature, rainfall, snowfall, etc. on the day of operation of the train. The operating time zone 213 indicates which time zone the prediction target corresponds to among the time zones classified into early morning, morning commuting rush hour, daytime, evening rush hour, night to midnight, and the like. Not only a time zone having a range of several tens of minutes to several hours, but also a specific time may be classified.

The delay degree 214 is, for example, a delay of less than 30 seconds (including "early arrival" arriving at the station earlier than planned. Early arrival can be regarded as a negative delay), a delay of less than 10 minutes, a delay of less than 30 minutes. Refers to an index (for example, "A", "B", "C") that classifies the operation status based on the delay time, such as the delay of 30 minutes or more. The method of classifying the delay degree 214 does not have to be based on time, but may be based on the number of passengers affected by the delay, the number of trains affected by the schedule, and the like. The classification method of the delay degree 214 may be configured to be freely set by the user of the operation prediction system 100, or may be configured so that the system automatically sets and adjusts a suitable classification method. ..

The vehicle formation 215 can be classified by, for example, the number of trains, the type of vehicle, the performance of the vehicle, and the like. The speed type 216 refers to the type of speed limit, its reason, and the section in which the speed limit is issued. Deterrence 217 is an instruction to stop the train for a certain period of time for various reasons and prohibit departure, and as the operation condition used for calculating the operation prediction distribution, the type of event that caused the deterrence (personal injury, personal injury, Invasion of animals into railroad tracks, strong winds, ice and snow, etc.), the section to be deterred, the time taken from the issuance of deterrence to the end of deterrence, etc.

The operation condition 202 shown in the operation record table 200 is an example, and is not limited thereto. The operating condition 202 can be freely added or changed.

The operation record 203 indicates the operation record of the train. For example, the operation record 203 includes information on the time of arrival at the stop station and the time of departure from the stop station.

Regarding the operation record data 112, for example, the operation record (arrival time) can be accumulated for each train according to the conditions of each station (weather, next station, up / down, track, etc.). Regarding the operation record data 112, for example, for each train, the operation record (departure time of the previous station, arrival of the next station, etc.) is met with respect to the conditions between stations (weather, up or down, speed limit, wind speed, rainfall, etc.). Time) can be accumulated.

FIG. 3 is a diagram showing an example (situation table group 300) of the situation data 113 between stations. The status table group 300 includes each station status table 310 and an inter-station status table 320.

Each station status table 310 stores information indicating the current status of each station. More specifically, in each station status table 310, the station number 311 that can identify the station, the weather 312 that indicates the weather at the station, the next station 313 that can identify the next station of the station, and the platform of the station go up. Specific to the upper and lower 314 indicating whether the platform is in the direction or the platform in the downward direction, the number 315 indicating the platform number of the station, the congestion status 316 indicating the congestion status of the platform of the station, and the platform of the station. The information of the specific passenger presence / absence 317 indicating whether or not there is a passenger of the above is stored in association with each other.

The inter-station status table 320 stores information indicating the current status between each station. More specifically, in the inter-station status table 320, the inter-station number 321 that can identify the section between the previous station and the next station, the weather 322 that indicates the weather in the section, and the traveling direction of the train in the section are in the upward direction. Information of up and down 323 indicating whether there is or in the downward direction, wind speed 324 indicating the wind speed in the section, rainfall 325 indicating the amount of rainfall in the section, and speed limit 326 indicating the speed limit in the section are stored in association with each other.

FIG. 4 is a diagram showing an example of the functions included in the operation prediction system 100.

As shown in FIG. 4, the operation prediction system 100 includes an operation plan data input unit 401, an operation record data input unit 402, a delay time calculation unit 403, an operation condition classification unit 404, and a similar operation record data extraction unit 405. The operation prediction distribution calculation unit 406, the operation prediction distribution correction unit 407, the operation prediction distribution calculation result processing unit 408, and the operation prediction distribution output unit 409 are provided.

The operation plan data input unit 401 is realized by, for example, the operation plan import device 120 of FIG. The function of the operation plan import device 120 (operation plan data input unit 401, etc.) is executed by, for example, the CPU (Central Processing Unit) reading the program stored in the ROM (Read Only Memory) into the RAM (Random Access Memory). It may be realized by doing (software), it may be realized by hardware such as a dedicated circuit, or it may be realized by combining software and hardware. Further, a part of the functions of the operation plan import device 120 may be realized by another computer capable of communicating with the operation plan import device 120.

The operation plan data input unit 401 acquires the operation plan data 410 of the train (the train to be predicted) for which the future operation status is predicted.

The operation plan data 410 includes data indicating the planned arrival / departure time of each station through which the train passes. Stations through which trains pass include starting stations, intermediate stations, and ending stations. The intermediate station may include a stop station where the train stops and a transit station where the train passes without stopping. The arrival / departure time includes an arrival time, which is the time when the train arrives at the station, and a departure time, which is the time when the train departs from the station. Further, the arrival / departure time may include a stop time, which is the time when the train is stopped at the station.

In the present embodiment, the unit of time is the hour and minute, but a unit finer than the hour and minute may be used, or another unit system may be used.

Operation record data input unit 402, delay time calculation unit 403, operation condition classification unit 404, similar operation record data extraction unit 405, operation prediction distribution calculation unit 406, operation prediction distribution correction unit 407, and operation prediction distribution calculation result processing unit 408. Is realized by, for example, the operation prediction distribution calculation program 111 of the calculation device 110 of FIG.

Functions of the calculation device 110 (operation record data input unit 402, delay time calculation unit 403, operation condition classification unit 404, similar operation record data extraction unit 405, operation prediction distribution calculation unit 406, operation prediction distribution correction unit 407, operation prediction distribution The calculation result processing unit 408, etc.) may be realized, for example, by the CPU reading the operation prediction distribution calculation program 111 stored in the ROM into the RAM and executing it (software), or hardware such as a dedicated circuit. It may be realized by a combination of software and hardware. Further, a part of the function of the arithmetic unit 110 may be realized by another computer capable of communicating with the arithmetic unit 110.

The operation record data input unit 402 acquires the operation record data 411, which is a record diagram showing the operating status of the trains that have operated in the past and each time of arrival, stop, departure, passage, etc. at each station. For example, the operation record data input unit 402 acquires the operation record data 112 shown in FIG. 1 as the operation record data 411.

The operation record data 411 includes data indicating the actual arrival / departure time at the time of operation of each station via which the train has passed. In addition, information related to train operation is added to the operation record data 411.

The delay time calculation unit 403 compares the operation plan data 410 of the train to be predicted and the operation record data 411 acquired from the operation plan data input unit 401 and the operation record data input unit 402, and operates up to now. Calculate the discrepancy between the plan and the actual operation, that is, the current delay time of the train to be predicted.

The operation condition classification unit 404 acquires the current operation condition of the train to be predicted. The information on the current operating conditions of the train to be predicted may be obtained by accessing an external system that stores the information, or stored for creating the operation record data 112. It may be configured to be acquired from the existing information.

The similar operation record data extraction unit 405 searches the operation record data 411 according to the operation conditions of the train to be predicted, which is acquired by the operation condition classification unit 404, and operates operation record data with similar operation conditions (hereinafter, "similar"). "Operation record data") is extracted.

As a similar determination method, for example, the current operation conditions and the operation conditions of the operation record data are compared, it is individually determined whether each condition is similar, and the prediction target is based on the number of similar operation conditions. There is a method of determining the similarity of the operation conditions of the train and the operation record data. Further, for example, the similar operation record data extraction unit 405 weights the operation conditions that have a large influence on the subsequent operation of the train to be predicted and scores the similarity, and performs the operation record data of a certain number of points or more as similar data. It may be determined as. Further, for example, the similar operation record data extraction unit 405 may simply determine the similarity based on the matching ratio of the operating conditions, such as when all the operating conditions match or when the operating conditions match by 80% or more. ..

Further, for example, the similar operation record data extraction unit 405 may perform similar determination for each operation condition by dividing it into station units or inter-station units instead of train units.

The operation prediction distribution calculation unit 406 performs statistical processing using the similar operation record data extracted by the similar operation record data extraction unit 405, and creates an operation prediction distribution. Here, one method of calculating and expressing the operation prediction distribution will be described. Regarding the arrival time or departure time of the train to be predicted at each station, similar operation record data is decomposed by station, arranged in ascending order of delay time (including minus), and trains with operation conditions similar to the operation conditions of the train to be predicted. Find the time (mode) with the highest probability of arrival in the past operation record.

FIG. 5 is a diagram showing an example (distribution example 500) of distribution in which similar operation record data are arranged in ascending order of arrival time or departure time delay time of a train to be predicted at a certain station.

In FIG. 5, the mode has a delay time of 20 seconds shown in the graph display 501, which has the largest number of data. Next, data of a predetermined ratio (hereinafter, described as 80%) of the extracted similar operation record data (for example, if there are 100 extracted similar operation record data, 80 of them). Find the time to settle. In the time when 80% of the data is stored, it is assumed that both the time on the side where the delay time is shorter than the mode and the time on the side where the delay time is longer than the mode appear. Depending on the situation, it is assumed that only one of them may appear. In the example of FIG. 5, the delay time on the side where the delay time is shorter than the mode value is −10 seconds, and the delay time on the side where the delay time is longer than the mode value is +120 seconds.

By the above processing, in the operation prediction distribution calculation unit 406, regarding the estimated arrival time and the estimated departure time of each station of the train to be predicted, the mode of the arrival time and the departure time of the trains that have traveled under similar operating conditions in the past The mode and the minimum and maximum values of the time when 80% of the trains arrive or depart (hereinafter referred to as "predicted margin") are calculated. In the example of FIG. 5, the predicted margin width is indicated by the width 502. In addition, the minimum value and the maximum value of the extracted similar operation record data as a whole can be acquired.

Here, in the present embodiment, a point n% from the minimum value (minimum value of the extracted similar operation record data as a whole) is set as a start point of the width 502, and a point (100-n)% is set as an end point of the width 502. To do. In the above example, when it is desired to find the width of the delay time in which 80% of the data of the past trains can be accommodated, it is assumed that there are 100 data of the past trains. They are arranged in ascending order, and the delay time corresponding to 10% (10th item) from the smallest is the start point of the width 502, and the delay time corresponding to 90% (90th item) is the end point of the width 502.

However, the method of setting the width 502 is not limited to the above contents. For example, the width 502 may be set using a so-called box plot. In this case, the width 502 is set from the first quartile to the third quartile (interquartile range) in the boxplot. The first quartile is a value located at 25% of the total, and the third quartile is a value located at 75% of the total.

Further, for example, the 100 (1-α)% credible interval in Bayesian statistics may be set as the width 502. However, 0 <α <1. As a method of selecting a credit interval, the concept of an equilateral posterior credit interval may be used as a basis, and the concept of the highest posterior credible interval (HPD interval: Highest Posterior Density Interval) may be used if necessary. This point will be described later using the second embodiment.

The operation prediction distribution correction unit 407 corrects the mode value acquired by the operation prediction distribution calculation unit 406 and the prediction margin width. In the operation prediction distribution correction unit 407, correction conditions (operation prediction distribution) are used to optimize the relationship between the front and rear trains and the operation interval of each station based on the operation prediction distribution calculated for each train or for each station and between stations. The operation prediction distribution is shifted back and forth based on the correction formula 412) to improve the accuracy of the operation prediction.

For example, the operation prediction distribution correction unit 407 calculates the operation prediction distribution for all trains after the current time, and as a result, when the prediction margin widths of the front and rear trains that stop at station A overlap, the train on the rear side The train interval at station A can be optimized by moving down the operation prediction distribution of the above to a time when the prediction margins do not overlap.

Further, for example, the operation prediction distribution correction unit 407 overlaps 10% or more of the operation prediction distribution (hereinafter referred to as "maximum width prediction distribution") including the minimum value and the maximum value of the entire similar operation record data of the front and rear trains. Then, the maximum width prediction distribution of the rear train may be optimized by moving it down to a time that does not overlap.

In addition, the operation prediction distribution correction unit 407 also improves the accuracy of the operation prediction by shifting the operation prediction distribution back and forth using the station-to-station situation data 413 or by directly adjusting the prediction timetable. Such processing will be described later with reference to FIG.

By correcting the statistical operation forecast distribution obtained from past operation results using information showing the current status of each station and the current status between stations, operation is more accurate and reliable than before. You can make predictions.

In this embodiment, the operation prediction distribution correction formula 412 and the station-to-station situation data 413 are used to shift the operation prediction distribution obtained by the operation prediction distribution calculation unit 406 back and forth, or to extend the stop time or extend the stop time. An example of making corrections such as shortening is shown. As a result of adding the situation data 413 between stations, if the operation prediction distribution fluctuates significantly, return to the similar operation record data extraction unit 405 and return to the operation record data similar to the current conditions of each station, or the current operation record data between stations. Correction may be made by re-extracting the operation record data similar to the conditions and recalculating the operation prediction distribution.

The operation prediction distribution calculation result processing unit 408 compares the operation prediction with the operation result, and corrects the operation prediction distribution correction formula 412 based on the degree of agreement between the operation prediction and the operation result.

Further, the operation prediction distribution calculation result processing unit 408 outputs the result corrected by the operation prediction distribution correction unit 407 to the operation prediction distribution output unit 409.

The operation prediction distribution output unit 409 is realized by, for example, the display device 130 shown in the operation prediction system 100.

The function of the display device 130 (operation prediction distribution output unit 409, etc.) may be realized by, for example, the CPU reading the screen display program 133 stored in the ROM into the RAM and executing it (software), or may be dedicated. It may be realized by hardware such as a circuit, or it may be realized by combining software and hardware. Further, a part of the functions of the display device 130 may be realized by another computer capable of communicating with the display device 130.

Regarding the operation prediction distribution corrected by the operation prediction distribution correction unit 407 displayed on the display device 130, the person in charge of creating the operation reorganization plan that monitors the train operation and reviews the train operation schedule when the timetable is disturbed. 414 confirms and reviews the operation reorganization plan 415 based on the operation forecast distribution. The person in charge of creating the operation reorganization plan 414 may be a system other than a human being, for example, a system that automatically creates the operation reorganization plan. In that case, the operation reorganization plan 415 is the operation prediction distribution corrected by the operation prediction distribution correction unit 407. It is automatically created by the system based on the above.

FIG. 6 is a diagram showing an example of a flowchart relating to a correction process in which the operation prediction distribution correction unit 407 corrects the stop time. The correction process is executed by changing the processing target for each station and for each train scheduled to stop at the station.

In step S601, the operation prediction distribution correction unit 407 determines whether or not the number of passengers waiting for boarding at the platform of the station to be processed is larger than the number of passengers on the same day of the week and the same time zone.

For example, the operation prediction distribution correction unit 407 is a ratio of the number of passengers waiting for boarding at the platform of the station to be processed to the number of people waiting for boarding on the same day of the week and the same time zone (for example, the average number of passengers). If it is determined that the (passenger waiting ratio) is 120% or more, the process is transferred to step S602, and if it is determined that the passenger waiting ratio is less than 120%, the process is transferred to step S604.

In step S602, the operation prediction distribution correction unit 407 determines whether or not the ratio of the number of passengers waiting for boarding is equal to or greater than a predetermined ratio (for example, 150%). For example, when the operation prediction distribution correction unit 407 determines that the ratio of the number of passengers waiting for boarding is 150% or more, the process shifts to step S603, and when it is determined that the ratio of the number of passengers waiting for boarding is less than 150%, the process proceeds to step S607. Move the process.

In step S603, if the operation prediction distribution correction unit 407 predicts that the stop time in the operation prediction distribution will be shortened as planned or due to delay reduction, the operation prediction distribution correction unit 407 will set the stop time to a predetermined time (for example, 1 minute) from the planned time. ) Correct the prediction to be extended. In addition, the operation prediction distribution correction unit 407 corrects the operation prediction distribution related to the extension to be shifted later.

In step S604, the operation prediction distribution correction unit 407 determines whether or not the ratio of the number of passengers waiting for boarding is less than a predetermined ratio (for example, 60%). For example, if the operation prediction distribution correction unit 407 determines that the ratio of waiting passengers is less than 60%, the process shifts to step S605, and if it determines that the ratio of waiting passengers is 60% or more, the operation prediction distribution correction unit 407 proceeds to step S607. Move the process.

In step S605, the operation prediction distribution correction unit 407 determines whether or not the train scheduled to stop at the station to be processed is delayed by a predetermined time (for example, 5 minutes) or more. For example, when the operation prediction distribution correction unit 407 determines that the train scheduled to stop at the station to be processed is delayed by 5 minutes or more, the process is transferred to step S606, and the train scheduled to stop at the station to be processed is delayed by 5 minutes or more. If it is determined that there is no such processing, the process proceeds to step S607.

In step S606, when the stop time in the operation prediction distribution is predicted as planned, the operation prediction distribution correction unit 407 shortens the stop time by a predetermined time (for example, 30 seconds) in order to recover the delay. Correct to the prediction. In addition, the operation prediction distribution correction unit 407 corrects the operation prediction distribution related to the shortening to be shifted forward.

In step S607, the operation prediction distribution correction unit 407 is a train (limited express, Shinkansen, etc., hereinafter referred to as "reserved seat type train") in which the train scheduled to stop at the station to be processed has reserved seats. Judge whether or not. When the operation prediction distribution correction unit 407 determines that the train scheduled to stop at the station to be processed is a reserved seat type train, the process is transferred to step S608, and the train scheduled to stop at the processing target station is a reserved seat type train. If it is determined that the train is not a train of, the correction process is terminated.

In step S608, the operation prediction distribution correction unit 407 determines whether or not the train scheduled to stop at the station to be processed is delayed by a predetermined time (for example, 15 minutes) or more. For example, when the operation prediction distribution correction unit 407 determines that the train scheduled to stop at the station to be processed is delayed by 15 minutes or more, the process is transferred to step S609, and the train scheduled to stop at the station to be processed is delayed by 15 minutes or more. If it is determined that there is no correction process, the correction process is terminated.

In step S609, the operation prediction distribution correction unit 407 is based on the reservation information, and the number of passengers scheduled to disembark from the train scheduled to stop at the station to be processed and the number of passengers disembarking from the train on the same day of the week and at the same time. It is determined whether or not the ratio to (for example, the average number of people) (the ratio of the number of people scheduled to get off) is equal to or greater than a predetermined ratio (for example, 150%). For example, when the operation prediction distribution correction unit 407 determines that the ratio of the number of people scheduled to disembark is 150% or more, the process is transferred to step S610, and when it is determined that the ratio of the number of people scheduled to disembark is less than 150%, the process proceeds to step S611. Move the process.

In step S610, the operation prediction distribution correction unit 407 corrects the stop time to a prediction that extends the stop time by a predetermined time (for example, 1 minute) from the planned time. In addition, the operation prediction distribution correction unit 407 corrects the operation prediction distribution related to the extension to be shifted later.

In step S611, the operation prediction distribution correction unit 407 determines whether or not the ratio of the number of people scheduled to get off is less than a predetermined ratio (for example, 60%). For example, when the operation management system prediction distribution correction unit 407 determines that the ratio of the number of people scheduled to disembark is less than 60%, it shifts the process to step S612, and when it determines that the ratio of the number of people scheduled to disembark is 60% or more, it corrects it. End the process.

In step S612, the operation prediction distribution correction unit 407 corrects the stop time to a prediction that shortens the stop time by a predetermined time (for example, 30 seconds) for delay recovery. In addition, the operation prediction distribution correction unit 407 corrects the operation prediction distribution related to the shortening to be shifted forward.

Further, although not shown, the operation prediction distribution correction unit 407 determines the distance between stations based on the conditions related to the operation acquired from the information along the railway line, such as the information of the anemometer and the rain gauge. It is corrected to a prediction that extends the time required to pass (hereinafter referred to as "travel time between stations"), or to a prediction that shortens the travel time between stations. In addition, the operation prediction distribution correction unit 407, for example, makes a correction for shifting the operation prediction distribution related to the extension later, or makes a correction for shifting the operation prediction distribution related to the shortening forward.

FIG. 7 is a diagram showing an example (operation distribution display 700) of an aspect of outputting an operation prediction distribution.

The operation distribution display 700 indicates the maximum value, the minimum value, and the mode value of the past operation record and the minimum time and the maximum time in which 80% of the similar operation record data is stored in the operation prediction distribution. The display 701 shows the minimum value of the entire similar operation record data. The display 702 shows the maximum value of the entire similar operation record data. Display 703 shows the mode value of similar operation record data. Display 704 indicates the minimum time for which 80% of similar operation record data can be stored. The display 705 indicates the maximum time during which 80% of similar operation record data can be stored.

When the similar operation record data shown in the operation distribution display 700 is represented by a graph showing the probability distribution, it becomes as shown in display 706. In order to make it easier to understand how the operation prediction distributions overlap, the operation prediction distribution may be shown by using a probability distribution graph such as the display 706 instead of the operation distribution display 700.

The person in charge of creating the operation reorganization plan 414 who sees the operation distribution display 700 can easily determine, for example, that the train to be confirmed may arrive later than the mode. In the actual output result, since the operation distribution display 700 is shown on the graph with the station on the vertical axis and the time on the horizontal axis, it is easy to determine how much time may be delayed.

In FIG. 7, the arrival of the train has been described as an example, but the same applies to the departure of the train.

FIG. 8 shows an example of a train operating from A station to F station, in which a planned timetable, an actual timetable, and a predicted timetable are displayed on a graph with stations on the vertical axis and time zones on the horizontal axis (diamond display 800). ). In the following, planned timetables, actual timetables, and forecast timetables may be collectively referred to as "diamonds".

In the diamond display 800, the planned diamond is indicated by a thin solid line, the actual diamond is indicated by a thick solid line, and the predicted diamond is indicated by a dotted line. Note that these notations are the same in other drawings.

Currently, the train is leaving station B, and there is no particular delay. In addition, it is not predicted that delays will occur at future stops.

In this way, when the delay is not expected, the operation prediction distribution output unit 409 may not display the operation prediction distribution but display the predicted time in order to simplify the display.

FIG. 9 is a diagram showing an example (diamond display 900) displaying a planned timetable, an actual timetable, and a predicted timetable for a train scheduled to depart from the same station B as in FIG.

The timetable display 900 is the same as the timetable display 800 in terms of the situation at station B (the planned timetable up to station B and the actual timetable match), and no delay has occurred.

However, at the next station C, information has been obtained that the ratio of passengers waiting for boarding is 200%. Information about the number of passengers at station C is included in, for example, the situation data 413 between stations. At this time, even if the operation prediction distribution calculated by the operation prediction distribution calculation unit 406 matches the plan schedule, the operation prediction distribution correction unit 407 extends the plan and stops the train at station C. It is corrected to the prediction, and the prediction after C station is also corrected. The operation prediction distribution calculated by the correction is displayed on the prediction timetable as shown in the display 901. At this time, for example, the operation prediction distribution is displayed so that the mode value of the operation prediction distribution overlaps on the prediction timetable. Further, the maximum delay that can occur with a probability of 80% on the operation prediction distribution may be shown lightly on the prediction timetable as shown in the display 902.

Here, the representation that visualizes the operation prediction distribution as shown in FIG. 7 does not necessarily have to be always displayed on the screen. In the first place, when the planning schedule is overcrowded, the operation distribution display 700, the figure expressing the operation prediction distribution as shown in the display 706, etc., and the graph suggesting the maximum possible delay such as the display 902 are superimposed. If it is displayed, it will be difficult to see the original operation plan, the current operation record, and the most probable forecast schedule. Therefore, in the timetable display that displays all trains, the operation prediction distribution output unit 409 does not display the operation prediction distribution, and when confirming the operation prediction distribution of only one train, the operation prediction distribution is predicted by paying attention to one station. If you want to check the distribution, or if you want to check the operation prediction distribution focusing on one train and one station, you may display the operation prediction distribution.

FIG. 10 is a diagram showing an example (diamond display 1000) in which a planned timetable, an actual timetable, and a predicted timetable are displayed on a graph with stations on the vertical axis and time zones on the horizontal axis from station A to station F. Is.

The timetable display 1000 is an example of the result of outputting the forecast by the operation prediction system 100, assuming that the equipment inspection occurs between the DE stations shortly before the current time and the train cannot pass temporarily. Is shown.

In the timetable display 1000, the section and the time zone in which the train cannot run are shown as, for example, display 1001. The end of the section and time zone in which the train cannot pass indicated by the display 1001 indicates the estimated time for canceling the impassability instruction at the present time.

When the equipment inspection between the DE stations is started, the train 1002 is on the E station, so it is predicted that the train will run according to the planned schedule without being affected by the equipment inspection.

The train 1003, which was running between the CD stations at the time of the equipment inspection, was stopped at the D station at the current time, and was scheduled to depart soon on the planned schedule, but the DE It is predicted that departures will be awaited until after the expected completion time of facility inspections between stations.

The display 1004 is an example of an icon suggesting that the operation prediction distribution is calculated and a deviation from the prediction timetable is sufficiently assumed.

In the timetable display 1000, the planned timetable is overcrowded, and if a delay occurs and a large difference occurs between the planned timetable and the predicted timetable, it tends to be difficult to visually recognize the planned timetable and the predicted timetable.

That is, in the timetable display 1000, since the plan timetable is overcrowded, if the operation prediction distribution is displayed on the plan timetable, the actual timetable, and the prediction timetable, it becomes difficult to visually recognize the information that should be originally confirmed. Take advantage of icons like.

The display 1004 is just an example, and as a method of suggesting the existence of the operation prediction distribution, the prediction diamond may be blinked or the color for displaying the prediction diamond may be changed.

FIG. 11 is a diagram showing an example (diamond display 1100) in which a part of the operation prediction distribution suggested by the display 1004 shown in FIG. 10 is displayed by paying attention to the train 1003 and the next train.

In the timetable display 1100, the operation prediction distribution display 1101 is shown for the prediction of the time when the train 1003, which has the greatest influence on the subsequent operation prediction, departs from D station. Focusing on the operation forecast distribution display 1101, the time that train 1003 can depart from D station may be significantly delayed from the time displayed in the current forecast schedule, for example, 80% of the past operation record. If the maximum delay that the data can appear occurs, the time when train 1003 departs from D station and the time when the next train arrives at D station will be close, so the prediction of the next train is , It is considered that it needs to be corrected. That is, the timetable display 1100 indicates that the prediction of the following train after the D station may be delayed according to the maximum value of the prediction margin width of the train 1003.

In the timetable display 1100, based on the operation prediction distribution calculated and corrected for each station and each train, the prediction display 1102 (dark black dotted line showing the prediction timetable) connecting the modes and 80% of the data are displayed. Predictive display 1103 (thin black dotted line) connecting the maximum delays that can occur is drawn respectively. In this example, for the purpose of preventing the prediction diamonds from overlapping each other and reducing the visibility, the time (prediction width) surrounded by the prediction displays for the same train is expressed by coloring it as shown in display 1104. ing.

In this way, by displaying the forecast width by paying attention to some trains, it is possible to visualize the operation prediction distribution of the train to be noticed and the influence on the following trains when the maximum delay occurs. As shown in this example, visualizing how much delay can be made from the current forecast when the restart time of train operation cannot be accurately predicted is for the person in charge of planning the recovery of the schedule thereafter. It can be useful for work, monitoring the arrival or departure of trains at stations, and the work of the person in charge of providing passenger guidance. In this example, the operation prediction distribution display 1101 is displayed at only one place, but if necessary, the operation prediction distribution of the arrival / departure time at another point or time point may be shown.

FIG. 12 shows the operation prediction distributions of the arrival time and departure time of each train after the current time, focusing on the D station and the E station, with respect to the operation prediction distribution suggested by the display 1004 shown in FIG. It is a figure which shows an example (diamond display 1200) which displayed about.

In the timetable display 1200, in order to prevent the visibility from being lowered due to the overlap between the operation prediction distributions, the operation prediction distribution output unit 409 shows the mode and the prediction margin as shown in the display 1201, and is similar. The scale indicating the maximum value and the minimum value in the operation record data is not displayed.

Further, in the timetable display 1200, when the operation prediction distributions overlap each other as shown in the display 1202, the operation prediction distribution output unit 409 prevents the deterioration of visibility by displaying the operation prediction distribution by shifting it up and down. There is. The operation prediction distribution focusing on a specific station can be useful information especially for the station staff of each station.

In this way, the output of the operation prediction distribution can be changed according to the needs of the user.

FIG. 13 is an example of a planned route (diamond display 1300) in which a large station is partially present and a rapid train 1301 and a local train 1302 are present. FIG. 14 shows the result of outputting the operation prediction distribution using the operation prediction system 100 for the timetable display 1300.

FIG. 14 is a diagram showing an example (diamond display 1400) displaying a planned timetable, an actual timetable, and a predicted timetable of a route in which a rapid train and a train stopped at each station exist.

In the timetable display 1400, a train between PQ stations cannot run because a personal injury occurred between PQ stations shortly before the current time. At the current time, the local train 1401 has already passed between the QP stations, which is the accident section, and is waiting for the subsequent rapid train 1402 to pass at the R station.

The rapid train 1402 was scheduled to have already departed on the planned schedule at the current time, but it is stopped at P station due to the accident. The operation prediction distribution of the rapid train 1402 is calculated according to the display 1403 indicating the section and the time zone in which the train cannot travel, and the operation prediction distribution of the departure time of the P station is shown by the display 1404.

According to the current forecast schedule, the rapid train 1402 actually passes through the R station where each station stop train 1401 is waiting for the rapid train 1402 to pass from the R station departure time of each station stop train 1401 on the planned schedule. Also significantly later. Further, from the operation prediction width display 1405 regarding the transit time of the rapid train 1402 at the R station, it can be seen that the transit time of the rapid train 1402 at the R station is likely to be significantly delayed from the current forecast schedule. Based on the above, the operation prediction distribution correction unit 407 can create a prediction that each station-stopped train 1401 departs from R station according to the planned schedule without waiting for the delayed rapid train 1402 to pass at R station. it can.

In the situation as shown in FIG. 14, it is conventionally determined whether to give priority to the passage of the following rapid train even if the waiting train is delayed, or to cancel the waiting for passage and change the passing order of the trains in the operation plan. It has been carried out by relying on the experience of the person in charge of making a recovery plan for the diamond. By utilizing the operation prediction system 100 of the present embodiment, when the estimated passage time of the following train is after a predetermined time (for example, 5 minutes or more) of the operation plan, the waiting for passage is canceled and the order of the trains is changed. It can be judged by a quantitative standard.

Regarding the operation prediction distribution, the operation prediction system 100 not only predicts the operation after the current time, but also compares the timetable (planned timetable) which is the basis of train operation with the arrival time and stop time of each station in the planned timetable. It is also possible to correct the transit time, departure time, etc.

FIG. 15 is a diagram showing an example (diamond display 1500) displaying a planned timetable from W station to Z station of a certain train and an operation prediction distribution of arrival time at each station.

Checking the difference between the planned timetable of each station and the operation forecast distribution, at W station, trains tend to arrive late, and although there are trains that arrived by the arrival time on the planned timetable, only a few trains were found. Understand. Therefore, the arrival time of the train at W station may be delayed from the current planned timetable.

At station X, trains tend to arrive late. The probability of arriving according to the planned schedule is higher than in the case of W station, but on the other hand, the probability of a significant delay is also high. Therefore, the arrival time of the train at X station may be delayed from the current planned timetable. Furthermore, since it can be seen from the operation prediction distribution that the arrival time is likely to be delayed, the stop time of the X station may be set longer than the current planned timetable in order to easily cover the delay.

At Y station, trains tend to pass earlier than the transit time on the planned schedule, and the probability of delay is not high. Therefore, the transit time of the train at Y station may be earlier than the present, taking into account the departure time of X station.

At Z station, the arrival time of the train is almost the same as the planned timetable. On the other hand, there is a possibility of a slight delay. Therefore, the arrival time of the train at Z station is generally not a problem according to the current planned schedule, but the stop time of the train at Z station may be set slightly longer than the present.

In FIG. 15, the arrival of the train has been described as an example, but the same applies to the departure of the train.

In this way, in the operation prediction system 100, the planned timetable can be corrected from the difference between the planned timetable and the operation prediction distribution. For example, when the operation prediction distribution is later than the planned time by a predetermined time or more, the operation prediction distribution calculation result processing unit 408 extends the arrival time, the stop time, the passing time, the departure time, etc. of the planned timetable by a predetermined time. Further, for example, when the operation prediction distribution is earlier than the planned time by a predetermined time, the operation prediction distribution calculation result processing unit 408 shortens the arrival time, stop time, transit time, departure time, etc. of the planned timetable by a predetermined time. To do.

FIG. 16 is a diagram showing the concept 1600 of the operation prediction distribution correction formula 412.

FIG. 16 shows an overlapping pattern of the operation prediction distribution of the front and rear trains at a certain station. In this example, the operation prediction distribution correction unit 407 corrects the operation prediction distribution by comparing the difference between the modes in the operation prediction distribution with the time intervals of the front and rear trains on the planned schedule.

In the overlapping example 1610, the time interval 1613 between the most frequent value in the operation prediction distribution 1611 of the preceding train A and the most frequent value in the operation prediction distribution 1612 of the following train B (time interval between the preceding train A and the following train B). ) Is shorter than the time interval on the planned timetable. Further, in the overlapping example 1610, if the predicted margin widths overlap each other and the arrival time of the preceding train A is close to the maximum value of the predicted margin width, the following train B needs to advance the arrival time. There are some disadvantages such as having to wait for the departure of the preceding train A on the track. Therefore, when a prediction like the overlap example 1610 ((time interval between modes) <(planned time interval) and overlap between prediction margin widths) is obtained, the operation prediction distribution correction unit 407 may perform the operation prediction distribution correction unit 407. At least one of the correction of deferring the arrival time of the following train B and the correction of deferring the predicted margin of the following train B is performed until the predicted margin widths do not overlap each other.

In the overlapping example 1620, the mode 1623 between the mode in the operation prediction distribution 1621 of the preceding train A and the mode value in the operation prediction distribution 1622 of the following train B is longer than the time interval on the planning schedule. In other words, the predicted margins overlap, but when paying attention to the mode, (planned time interval) <(time interval between the modes), it is possible to secure sufficient train spacing. The operation prediction distribution correction unit 407 does not perform the correction.

In the overlapping example 1630, the mode 1633 between the mode in the operation prediction distribution 1631 of the preceding train A and the mode value in the operation prediction distribution 1632 of the following train B is shorter than the time interval on the planning schedule. That is, since (time interval between modes) <(planned time interval), the train interval is shorter than the standard, but the prediction margin width does not overlap, so the operation prediction distribution correction unit 407 Do not make corrections.

The operation prediction distribution correction formula 412 with the above contents is used to correct the operation prediction distribution and the like.

Further, when the operation prediction distribution correction unit 407 compares the operation result and the operation prediction distribution and determines that there is a discrepancy between the two, the correction condition (parameter) is changed to make it suitable.

For example, by adding the criteria for carrying down the following train to the time interval between the modes, it is not whether or not the predicted margins of the preceding train and the following train overlap, but whether or not they overlap by a predetermined ratio or more. It may be used as a reference. Further, for example, the standard for carrying down the following train is not the time interval between the modes, but the overlapping time of the predicted margin width of the preceding train and the succeeding train, and a predetermined ratio of the predicted margin width time (for example, 10%). If the above overlaps, the arrival time of the following train may be postponed. In addition, machine learning may be performed so that the discrepancy between the predicted margin and the operation record becomes small, and the predetermined ratio may be changed.

In FIG. 16, the arrival of the train has been described as an example, but the same applies to the departure of the train.

According to this embodiment, the train operation prediction can be regarded as a distribution with a time range based on the past operation results. In the present embodiment, in addition to the prediction that the train will arrive or depart at a specific time, it is possible to express the probability of early arrival or late arrival from that time. By grasping the time width that may arrive early and the time width that may arrive late, it can be used for subsequent operation planning, passenger guidance in the station yard, and the like. It is also possible to predict the influence of the train running interval, the delay of a certain train, etc. on the following trains from the overlap of the operation prediction distributions, and to optimize the entire train operation.

(2) Second Embodiment FIG. 17 is a diagram showing a part of the configuration of the operation prediction system 1700 according to the second embodiment. In the present embodiment, the same reference numerals as those of the first embodiment will be used and the description thereof will be omitted.

The operation prediction system 1700 compares the operation prediction distribution calculated using the concept of the credit interval of Bayesian statistics with the operation record data (results for the predicted operation), and uses the credit interval when calculating the prediction margin. Adjust how to take.

For example, in the operation prediction system 1700, when comparing the operation prediction distribution calculated using the equal-tailed post-credible interval of Bayesian statistics with the operation record data, the number of cases deviating from the credit interval exceeds 100 α% ( This threshold may be set arbitrarily by the user), and the operation prediction distribution (in the past case that deviates from the credible interval) is calculated again using the concept of the HPD interval. That is, the operation prediction system 1700 recalculates the prediction of the target train whose operation has already been completed by using the concept of the HPD section.

The operation prediction system 1700 compares the post-credible interval with the actual operation record data and the HPD section with the actual operation record data, respectively, and the error from the operation record is smaller or the mode is the mode. The credible interval with the shorter distance from the operation record is used to calculate the forecast margin for the next and subsequent times.

The value of α in the 100 (1-α)% confidence interval can be freely set by the user.

More specifically, the operation prediction system 1700 includes an operation prediction accuracy evaluation unit 1701, an operation prediction distribution calculation method comparison unit 1702, and an operation prediction distribution calculation method change unit 1703 as functions of the calculation device 110.

The operation prediction accuracy evaluation unit 1701 compares the operation prediction distribution calculation result data 1704 with the operation record data 411 of the train (station or between stations) to be predicted, and verifies whether or not the prediction is correct. The operation prediction distribution calculation result data 1704 is data output from the operation prediction distribution calculation result processing unit 408, and is the similar operation result data extracted by the similar operation result data extraction unit 405 and used for calculating the operation prediction distribution. Each station inter-station situation data 413 and the like are included.

The operation prediction distribution calculation method comparison unit 1702 calculates the credible interval (equal tail post-credit interval or HPD section) that is not used for calculating the operation prediction distribution, and compares it with the original method. After the confidence interval is calculated, corrections are also made.

The operation prediction distribution calculation method change unit 1703 sets the operation prediction distribution calculation unit 406 to use the credit interval having the smaller difference between the operation result and the operation prediction.

FIG. 18 is a diagram showing an example of the functions included in the operation prediction system 1700.

As shown in FIG. 18, the operation prediction system 1700 incorporates an operation prediction accuracy evaluation unit 1701, an operation prediction distribution calculation method comparison unit 1702, and an operation prediction distribution calculation method change unit 1703.

FIG. 19 is a diagram showing an example (comparison table 1900) of a table created by the operation prediction accuracy evaluation unit 1701 for comparing a train operation prediction and an operation record at a predetermined station. The comparison table 1900 shows an example of arrival time, but a similar table is created for departure time.

The comparison table 1900 shows train number 1901, operating day 1902, day of the week 1903, plan (arrival) 1904, forecast (minimum) 1905, forecast (most frequent) 1906, forecast (maximum) 1907, actual 1908, (actual-most frequent). Information of (value) 1909 and (actual-predicted end) 1910 is stored in association with each other. The prediction end is selected from the prediction (minimum) 1905 and the prediction (maximum) 1907, whichever is closer to the actual 1908.

The train number 1901 is identification information that can identify a train at a predetermined station. The operating day 1902 is information indicating the date on which the train was operated. The day of the week 1903 is information indicating the day of the week when the train was operated. The plan (arrival) 1904 is information indicating the time when the train is scheduled to arrive at a predetermined station (arrival time on the plan schedule). Prediction (minimum) 1905 is information indicating the minimum time for 80% of similar operation record data to be settled. Prediction (mode) 1906 indicates the mode of similar operation record data. Prediction (maximum) 1907 is information indicating the maximum time during which 80% of similar operation record data can be stored.

The actual result 1908 is information indicating the actual time (arrival time on the actual timetable) when the train arrived at a predetermined station. (Actual-mode) 1909 is information indicating a value obtained by subtracting the predicted (mode) 1906 from the actual 1908. (Actual-predicted end) 1910 is information indicating a value obtained by subtracting predicted (minimum) 1905 or predicted (maximum) 1907 from actual 1908.

In the comparison table 1900, as shown in (actual-predicted end) 1910, the actual 1908 is not within the predicted margin (= confidence interval). Here, it is assumed that, for example, an 80% credible interval (100 (1-α)% credible interval: α = 0.2) and an equilateral post-credit interval are used as the credit interval. At this time, if the case where the actual result 1908 is outside the forecast margin exceeds a predetermined standard (for example, 20%), it is considered to change the method of taking the credit interval. The reference numerical value may be determined by the user.

FIG. 20 is a diagram showing an example of the relationship between the confidence interval and the operation prediction distribution (relationship diagram 2000). The relationship diagram 2000 shows Example 2010, where the equilateral post-credible interval is more appropriate, and Example 2020, where the HPD interval is more appropriate.

Operation prediction distribution calculation method The comparison unit 1702 recalculates the operation prediction of the train at a predetermined station using another confidence interval (here, the HPD interval) under the same conditions as at that time. That is, the operation prediction distribution calculation method comparison unit 1702 recreates a comparison table such as the comparison table 1900 using another credit interval.

The operation forecast distribution calculation method change unit 1703 compares the operation forecast distribution recalculated in the HPD section with the original operation forecast distribution calculated in the equilateral post-confidence interval, and the operation record does not fit within the confidence interval. If is less than the equi-tailed post-confidence interval, the HPD interval is adopted, and if it is more than the equi-tailed post-confidence interval, the equi-tailed post-confidence interval is retained.

As shown in Example 2010, when the operation record falls within the equi-tailed post-credible interval and the operation record does not fall within the HPD section, the equi-tailed post-section is adopted for the calculation of the operation prediction distribution. On the other hand, as shown in Example 2020, when the operation record falls within the HPD section and the operation record does not fall within the equilateral post-credible interval, the HPD section is adopted for the calculation of the operation prediction distribution.

The cause of the discrepancy between the operation forecast and the operation record is not only the method of taking the confidence interval, but also the correction based on the situation data 413 between each station and the correction by the operation prediction distribution correction formula 412. If the correction of the credible interval is not enough, the operation prediction distribution calculation method changing unit 1703 changes the parameters of the correction by the operation prediction distribution correction formula 412 and the correction by the operation prediction distribution correction formula 412.

According to the present embodiment, the arrival time width and / or the departure time width is calculated by a more appropriate method of taking a credit interval, so that it is possible to reduce the discrepancy between the operation prediction and the operation record.

In addition, regardless of this method, the method of determining which credit interval to adopt may be in a form that can be arbitrarily determined by the user.

(3) Addendum The above-described embodiment includes, for example, the following contents.

In the above-described embodiment, the case where the present invention is applied to the operation prediction system has been described, but the present invention is not limited to this, and is widely applied to various other systems, devices, methods, and programs. Can be done.

Further, in the above-described embodiment, the configuration of each table is an example, and one table may be divided into two or more tables, or all or a part of the two or more tables is one table. You may.

Further, in the above-described embodiment, various data have been described using the XX table for convenience of explanation, but the data structure is not limited and may be expressed as XX information or the like.

Further, in the above description, information such as programs, tables, and files that realize each function is recorded in a memory, a hard disk, a storage device such as an SSD (Solid State Drive), or an IC card, an SD card, a DVD, or the like. Can be placed on the medium.

The above-described embodiment has, for example, the following characteristic configurations.

The operation prediction system (for example, operation prediction system 100) that predicts the operation of the moving object to be operated is a storage unit (for example, operation record data 112) that stores operation record data (for example, operation record data 112) indicating the operation record of the moving object. The calculation device 110, the storage device 140) and the operation record data include the time when the moving body arrives at a predetermined point (for example, a station) (for example, the arrival time) and the moving body departs from the predetermined point. Operation record data (eg, operation status data) that includes at least one of the time (for example, departure time) and is similar to the operation status of the moving object to be predicted (for example, operation condition 202, condition of each station, condition between stations). , Similar operation record data) is acquired from the storage unit (for example, calculation device 110, similar operation record data extraction unit 405), and based on the operation record data acquired by the acquisition unit, the above-mentioned predetermined Probability such as arrival time width (for example, prediction margin width at the time of arrival, operation distribution display 700 at the time of arrival, display 706 at the time of arrival) indicating that the moving object to be predicted arrives at a certain point with a certain reliability. (Distribution graph) and the departure time width (for example, the predicted margin width at the time of departure, the operation distribution display 700 at the time of departure, the display 706 at the time of departure) indicating that the moving object to be predicted departs with a certain degree of reliability. A calculation unit (for example, calculation device 110, operation prediction distribution calculation unit 406) that calculates at least one of the above calculation units, and an output unit (for example, a display device) that outputs the result calculated by the above calculation unit. 130, operation prediction distribution output unit 409).

The mobile body is, for example, a transportation means such as a train, an aircraft, or a bus. The stops are stations, airports, bus stops, etc.

Further, the output unit may display the result calculated by the calculation unit on a display device, transmit it to another system, or print it on a medium such as paper. It may be output as audio by a speaker.

In the above configuration, the arrival time or departure time of the moving object at a predetermined point is calculated as a prediction (arrival time width or departure time width) having a time width.

In the above configuration, by calculating and outputting the arrival time width or the departure time width in consideration of the deviation that may occur in the arrival time or the departure time of the moving body, for example, a prediction with a width in the arrival / departure time. You can create a diamond. Further, for example, it is possible to assist the work of the person in charge of considering the subsequent operation reorganization plan, the work of the staff member who confirms the arrival of the moving body at the place where the moving body stops, opens / closes the door, guides passengers, and the like.

In particular, even if it is difficult to predict the subsequent operation of the moving object due to delays caused by bad weather, accidents, equipment inspections, etc., the arrival time width and / or departure time width can be used based on past operation results. In addition, it is possible to make a certain degree of operational outlook, which can be useful for facilitating the work of the person in charge, staff, etc.

The calculation unit calculates the first arrival time width of the moving body to be predicted and the second arrival time width of the moving body following the moving body to be predicted at the predetermined point. Based on the first arrival time width and the second arrival time width, the operation prediction system makes a correction for shifting the arrival time of the subsequent moving object later and a correction for shifting the second arrival time width later. A correction unit (for example, a calculation device 110, an operation prediction distribution correction unit 407) that performs at least one of the above is provided.

According to the above configuration, for example, when the arrival time widths of the front and rear moving bodies overlap, the distance between the front and rear moving bodies and the arrival time of the succeeding moving bodies are corrected to optimize the operation of the future moving bodies. be able to.

The first arrival time width includes the first mode value of the operation record data used for the calculation in the calculation unit (for example, the mode value of the operation prediction distribution 1611 of the preceding train A). The second arrival time width includes the second mode value of the operation record data used for the calculation in the calculation unit (for example, the mode value in the operation prediction distribution 1612 of the following train B). The correction unit determines whether or not the time interval between the first mode and the second mode is smaller than a predetermined time interval (for example, the time interval on the planning schedule), and if it is small, When it is determined, it is determined whether or not the first arrival time width and the second arrival time width overlap by a predetermined value or more (for example, whether or not they overlap), and when it is determined that they overlap, the above Correction to shift the arrival time of the following moving object later (correction to lower the arrival time of the following train B until the time when the predicted margin widths do not overlap) and correction to shift the arrival time width of the second arrival time later (predicted margin width) At least one of the corrections (correction to lower the predicted margin width of the following train B) is performed until the time when they do not overlap each other.

The correction unit has the first arrival time width (for example, the prediction margin width in the operation prediction distribution 1631 of the preceding train A) and the second arrival time width (for example, the prediction margin in the operation prediction distribution 1632 of the following train B). It is determined whether or not the width) and the above predetermined value overlap, and if it is determined that they do not overlap, no correction is performed.

The correction unit has the first mode (for example, the mode in the operation prediction distribution 1621 of the preceding train A) and the second mode (for example, the mode in the operation prediction distribution 1622 of the following train B). It is determined whether or not the time interval with the value) is larger than the above-mentioned predetermined time interval, and if it is determined that the time interval is larger, no correction is performed.

If the time interval between the first mode and the second mode is equal to the predetermined time interval, the correction may be performed to shift the arrival time of the subsequent moving object later. Alternatively, the correction for shifting the arrival time of the subsequent moving object later may not be performed.

In the above configuration, for example, the operation prediction system calculates the arrival time width using the past operation record data similar to the operation status of the mobile object to be predicted, and moves back and forth depending on the degree of overlap between the arrival time widths. Correct the arrival time of the moving body in consideration of the relationship between the bodies (operation interval, etc.).

Conventionally, when a delay occurs in a moving body, the operation interval has been adjusted. In this regard, by calculating the arrival time width, it becomes possible to quantitatively grasp how much the operation interval of the front and rear moving objects may be narrowed. For example, the correction unit can use a quantitative value of the degree of overlap of arrival time widths as a reference for correcting the operation interval. Therefore, more suitable operation is possible in consideration of the relationship between the front and rear moving bodies.

The calculation unit calculates the first departure time width of the moving body to be predicted and the second departure time width of the moving body following the moving body to be predicted at the predetermined point. Based on the first departure time width and the second departure time width, the operation prediction system makes a correction for shifting the departure time of the subsequent moving object later and a correction for shifting the second departure time width later. A correction unit (for example, a calculation device 110, an operation prediction distribution correction unit 407) that performs at least one of the above is provided.

According to the above configuration, for example, when the departure time widths of the front and rear moving bodies overlap, the distance between the front and rear moving bodies and the departure time of the succeeding moving bodies are corrected to optimize the operation of the future moving bodies. be able to.

The operation prediction system is based on information that affects the operation of the moving body at the predetermined point (for example, each station status table 310), and is the time when the moving body to be predicted stops at the predetermined point. A correction unit (for example, calculation device 110, operation prediction distribution correction unit 407) for correcting a certain stop time (for example, stop time) is provided.

According to the above configuration, for example, by adding information that affects the operation of the moving body at a predetermined point where the moving body is stopped to the stopping time of the moving body, a more accurate operation prediction result can be obtained. ..

In the operation prediction system, the moving body to be predicted is moved to the predetermined point based on the information affecting the operation of the moving body on the route on which the moving body moves (for example, the inter-station situation table 320). A correction unit (for example, a calculation device 110, an operation prediction distribution correction unit 407) for correcting a stop time (for example, a stop time), which is a stop time, is provided.

According to the above configuration, for example, by adding information that affects the operation of the moving body on the route on which the moving body moves to the dwell time of the moving body, a more accurate operation prediction result can be obtained. In addition, even if there is no delay and there is no need to refer to past operation results, the stop time can be corrected according to the status of the travel route, and a more accurate operation outlook can be established. it can.

The correction unit (for example, the calculation device 110, the operation prediction distribution calculation result processing unit 408) corrects based on the result of comparing the first arrival time width with the operation record data of the mobile body to be predicted. Change the conditions.

According to the above configuration, for example, when the arrival time width of the front and rear moving bodies is very large, a value having a small degree of overlap as a reference for correction of the arrival time width of the succeeding moving body (for example, 10% overlap degree). Even so, the arrival time width of the succeeding moving body will be significantly reduced by the arrival time width of the preceding moving body. On the contrary, when the arrival time width of the moving body in the front-rear direction is small, the effect of the correction may not be sufficiently obtained, and a suitable correction result may not be obtained. However, such disadvantages can be eliminated by updating the correction conditions based on the operation record data.

The correction unit (for example, the calculation device 110, the operation prediction distribution calculation result processing unit 408) corrects based on the result of comparing the first departure time width with the operation record data of the mobile body to be predicted. Change the conditions.

According to the above configuration, for example, when the departure time width of the front and rear moving bodies is very large, the overlap degree that is the reference for the correction of the departure time width of the succeeding moving body is a small value (for example, 10% overlap). Even so, the departure time width of the subsequent moving body will be significantly reduced with the departure time width of the preceding moving body. On the contrary, when the departure time width of the moving body in the front-rear direction is small, the effect of the correction may not be sufficiently obtained, and a suitable correction result may not be obtained. However, such disadvantages can be eliminated by updating the correction conditions based on the operation record data.

The calculation unit calculates the arrival time width of the moving body to be predicted, and the operation prediction system is a planning diagram showing the arrival time width calculated by the calculation unit and the operation plan of the moving body to be predicted. A correction unit (for example, a calculation device 110, an operation prediction distribution calculation result processing unit 408) for correcting the planned schedule is provided.

In the above configuration, for example, the planned timetable is corrected by utilizing the arrival time width obtained from the past operation record data. The correction of the planning schedule has conventionally been carried out based on the experience and feeling of the person in charge of creating the planning schedule. In this respect, according to the above configuration, it is possible to quantitatively correct the planning schedule based on the past operation record data without relying on the skilled human sense.

The calculation unit calculates the departure time width of the moving body to be predicted, and the operation prediction system is a planning diagram showing the departure time width calculated by the calculation unit and the operation plan of the moving body to be predicted. A correction unit (for example, a calculation device 110, an operation prediction distribution calculation result processing unit 408) for correcting the planned schedule is provided.

In the above configuration, for example, the planned schedule is corrected by utilizing the departure time width obtained from the past operation record data. The correction of the planning schedule has conventionally been carried out based on the experience and feeling of the person in charge of creating the planning schedule. In this respect, according to the above configuration, it is possible to quantitatively correct the planning schedule based on the past operation record data without relying on the skilled human sense.

The output unit displays at least one of the arrival time width and the departure time width calculated by the calculation unit together with the prediction timetable of the moving object to be predicted (see, for example, FIGS. 9 to 12).

For example, an operation prediction distribution having a width of the calculated arrival time or departure time is displayed on a prediction diamond having no width (a diamond that predicts that a moving object will arrive at a specific time) to arrive. It is possible to indicate how much deviation may occur in time or departure time.

The calculation unit (for example, calculation device 110, operation prediction accuracy evaluation unit 1701, operation prediction distribution calculation method comparison unit 1702, operation prediction distribution calculation method change unit 1703) is based on the operation record data acquired by the acquisition unit. In addition, for the time when the moving object arrived, the equal-tailed post-credible interval and the highest post-density credit interval of the Bayesian statistics were calculated, and the difference between the above-mentioned equal-footed post-credible interval and the operation record of the above-mentioned predicted moving object. , Compare the difference between the highest posterior density confidence interval and the operation record of the vehicle to be predicted, and take the credit interval with the higher probability that the record is included in the credit interval. Calculate the width.

According to the above configuration, since the arrival time width is calculated by a more appropriate method of taking a credit interval, it is possible to reduce the discrepancy between the operation prediction and the operation record.

The calculation unit (for example, calculation device 110, operation prediction accuracy evaluation unit 1701, operation prediction distribution calculation method comparison unit 1702, operation prediction distribution calculation method change unit 1703) is based on the operation record data acquired by the acquisition unit. In addition, for the time when the above-mentioned moving object departed, the equal-tailed post-credible interval and the highest post-density credit interval of the Bayesian statistics were calculated, and the difference between the above-mentioned equal-footed post-credible interval and the operation record of the above-mentioned predicted moving object. , Compare the difference between the highest posterior density confidence interval and the operation record of the vehicle to be predicted, and take the credit interval with the higher probability that the record is included in the credit interval. Calculate the width.

According to the above configuration, since the departure time width is calculated by a more appropriate method of taking a credit interval, it is possible to reduce the discrepancy between the operation prediction and the operation record.

Further, the above-described configuration may be appropriately changed, rearranged, combined, or omitted as long as it does not exceed the gist of the present invention.

The items included in the list in the form of "at least one of A, B, and C" are (A), (B), (C), (A and B), (A and C), (B). And C) or (A, B, and C) can be understood to mean. Similarly, the items listed in the form of "at least one of A, B, or C" are (A), (B), (C), (A and B), (A and C), Can mean (B and C) or (A, B, and C).

100 …… Operation prediction system, 110 …… Calculation device.

Claims (14)

It is an operation prediction system that predicts the operation of moving objects that are operated.
A storage unit that stores operation record data that shows the operation record of the mobile body,
The operation record data includes at least one of the time when the moving body arrives at a predetermined point and the time when the moving body departs from the predetermined point.
An acquisition unit that acquires operation record data of operation status similar to the operation status of the mobile object to be predicted from the storage unit, and an acquisition unit.
Based on the operation record data acquired by the acquisition unit, the arrival time width indicating that the mobile body to be predicted arrives with a certain reliability and the mobile body to be predicted are determined at the predetermined point. A calculation unit that calculates at least one of the departure time widths indicating that the vehicle departs with a certain degree of reliability.
An output unit that outputs the result calculated by the calculation unit, and
Operation prediction system equipped with. The calculation unit calculates, at the predetermined point, the first arrival time width of the moving body to be predicted and the second arrival time width of the moving body following the moving body to be predicted.
At least one of a correction for shifting the arrival time of the subsequent moving object later and a correction for shifting the arrival time width of the second arrival time later based on the first arrival time width and the second arrival time width. It is equipped with a correction unit that performs
The operation prediction system according to claim 1. The first arrival time width includes the first mode value of the operation record data used for the calculation in the calculation unit.
The second arrival time width includes the second mode value of the operation record data used for the calculation by the calculation unit.
The correction unit determines whether or not the time interval between the first mode and the second mode is smaller than a predetermined time interval, and if it is determined that the time interval is smaller, the first arrival time. It is determined whether or not the width and the second arrival time width overlap by a predetermined value or more, and when it is determined that they overlap, a correction for shifting the arrival time of the subsequent moving body later and the second arrival time are performed. Make at least one with a correction that shifts the width later,
The operation prediction system according to claim 2. The calculation unit calculates the first departure time width of the mobile body to be predicted and the second departure time width of the subsequent mobile body of the mobile body to be predicted at the predetermined point.
At least one of a correction for shifting the departure time of the subsequent moving body later and a correction for shifting the departure time width of the second departure time later based on the first departure time width and the second departure time width. It is equipped with a correction unit that performs
The operation prediction system according to claim 1. Based on the information affecting the operation of the moving body at the predetermined point, a correction unit for correcting the staying time, which is the time when the moving body to be predicted stays at the predetermined point, is provided.
The operation prediction system according to claim 1. A correction unit for correcting the dwell time, which is the time for the moving body to be predicted to stay at the predetermined point, based on the information affecting the operation of the moving body on the route on which the moving body moves.
The operation prediction system according to claim 1. The correction unit changes the correction conditions based on the result of comparing the first arrival time width with the operation record data of the moving body to be predicted.
The operation prediction system according to claim 2. The correction unit changes the correction conditions based on the result of comparing the first departure time width with the operation record data of the moving body to be predicted.
The operation prediction system according to claim 4. The calculation unit calculates the arrival time width of the moving object to be predicted.
A correction unit is provided that compares the arrival time width calculated by the calculation unit with the planning timetable indicating the operation plan of the moving object to be predicted, and corrects the planning timetable.
The operation prediction system according to claim 1. The calculation unit calculates the departure time width of the moving object to be predicted.
It is provided with a correction unit that compares the departure time width calculated by the calculation unit with the planning timetable indicating the operation plan of the moving body to be predicted and corrects the planning timetable.
The operation prediction system according to claim 1. The output unit displays at least one of the arrival time width and the departure time width calculated by the calculation unit together with the prediction timetable of the moving object to be predicted.
The operation prediction system according to claim 1. Based on the operation record data acquired by the acquisition unit, the calculation unit calculates the equal-tailed posterior credible interval and the highest posterior density credible interval of Bayesian statistics for the time when the moving object arrives, and the above-mentioned etc. Comparing the difference between the post-mortem credit interval and the operation record of the forecasted moving object with the difference between the highest post-density credit interval and the operating record of the predicted moving object, the credit interval with the smaller difference is compared. Calculate the arrival time width of the prediction target by the method of taking
The operation prediction system according to claim 1. Based on the operation record data acquired by the acquisition unit, the calculation unit calculates the equal-tailed posterior credible interval and the highest posterior density credible interval of Bayesian statistics for the time when the moving object departs, and the above-mentioned etc. Comparing the difference between the post-mortem credit interval and the operation record of the forecasted moving object with the difference between the highest post-density credit interval and the operating record of the predicted moving object, the credit interval with the smaller difference is compared. Calculate the departure time width of the forecast target by the method of taking
The operation prediction system according to claim 1. It is an operation prediction method in an operation prediction system that predicts the operation of a moving object to be operated.
The operation prediction system includes a storage unit that stores operation record data indicating the operation record of the moving body.
The operation record data includes at least one of the time when the moving body arrives at a predetermined point and the time when the moving body departs from the predetermined point.
The acquisition unit acquires the operation record data of the operation status similar to the operation status of the mobile object to be predicted from the storage unit.
Based on the operation record data acquired by the acquisition unit, the calculation unit has an arrival time width indicating that the moving object to be predicted arrives with a certain degree of reliability at the predetermined point, and the prediction target. To calculate at least one of the departure time widths that indicate that the moving object departs with a certain degree of reliability.
The output unit outputs the result calculated by the calculation unit.
Operation prediction method equipped with.

Images