Information processing device, information processing program, and information processing method

Abstract

This information processing device recursively repeats processing in which each congested section indicated by training data including the congestion start position and the congestion end position is classified in each congestion direction and adjacent congested sections are connected such that the congested sections are formed as a single integrated congested section, and thereby determines which of the congested sections is the last of the the congested sections forming the integrated congested section, classifies, for each integrated congested section indicating the same congestion, the integrated congested section according to the determined demarcation, and sets, for each integrated congested section indicating the same congestion and for each determined demarcation, the extension scale of said integrated congested section.

Description

Information processing device, information processing program, and information processing method

[0001] The present disclosure relates to an information processing device, an information processing program, and an information processing method for predicting the length of congestion that occurs suddenly due to a change in demand.

[0002] 2. Description of the Related Art There is known a traffic congestion prediction service that predicts the location, time, and length of road congestion and provides users with the predicted information (for example, Non-Patent Document 1).

[0003] When a user inputs a date and time into the traffic congestion prediction service, a road map such as that shown in FIG. 19 is displayed on the screen, and sections where traffic congestion is predicted to occur at the input date and time are displayed on the road map by, for example, arrows 50.

[0004] Furthermore, when the user selects a detail icon 51 displayed corresponding to the arrow 50, a traffic jam detail screen 52 such as that shown in Fig. 20 is displayed. The traffic jam detail screen 52 displays the traffic jam section, the time period during which the traffic jam occurs, the location of the bottleneck of the traffic jam, the length of the traffic jam at peak times, the time required to pass through the traffic jam, etc.

[0005] <Non-Patent Document 1> East Nippon Expressway Company Limited, Zenrin Datacom Co., Ltd., "Drive Traffic" <URL: https: / / www.drivetraffic.jp / .>

[0006] These congestion prediction services predict traffic congestion based on past actual values ​​of steady increases in traffic volume (for example, over a year or more). Therefore, when it comes to congestion that has newly occurred due to changes in behavior caused by the COVID-19 pandemic, such as an increase in the number of drive-throughs or increased demand for visits to home improvement stores, the number of actual values ​​of traffic volume increases that can be used for prediction is limited.

[0007] Therefore, with conventional congestion prediction services, when new congestion occurs in a location where congestion has not previously occurred due to a change in demand (hereinafter referred to as "sudden congestion"), it is difficult to predict how far the congestion will extend once it has begun to occur.

[0008] On the other hand, it is possible to obtain the situation of unexpected congestion from traffic congestion images captured by cameras installed on roads. However, since it is impossible to predict in advance where unexpected congestion will occur, it is necessary to install cameras in various locations, which increases the cost of congestion prediction. Furthermore, obtaining the situation of unexpected congestion from traffic congestion images necessarily obtains the situation after the congestion has occurred. Therefore, obtaining the situation of unexpected congestion from traffic congestion images delays the provision of information to users compared to traffic congestion forecast information that notifies users in advance of the possibility of congestion. Therefore, it is preferable to predict the occurrence of unexpected congestion from observation data on past unexpected congestion (hereinafter referred to as "learning data") rather than relying on traffic congestion images. In the verification described herein, information obtained from the Japan Road Traffic Information Center (JARTIC), a public interest incorporated foundation, was used as the observation data on unexpected congestion.

[0009] In view of the above, an object of the present invention is to provide an information processing device, an information processing program, and an information processing method that can predict the length of a sudden traffic jam even if the sudden traffic jam occurs for which there is not enough learning data available to be used for traffic jam prediction using conventional traffic jam prediction services.

[0010] A first aspect of the present disclosure is an information processing device, comprising: a classification unit that classifies each congestion section represented by learning data including a congestion start position and a congestion end position, into a congestion direction represented by the start position and the end position; and a process of connecting adjacent congestion sections, whose start position and end position are within a predetermined range, as a continuous integrated congestion section for each congestion section classified by the congestion direction by the classification unit, when the start position or the end position of the adjacent congestion section does not fall within the predetermined range. and a setting unit that acquires the integrated congestion sections determined by the determination unit for each integrated congestion section that represents the same congestion, classifies the integrated congestion section for each integrated congestion section that represents the same congestion according to each determination category that is divided by day of the week, by work attribute that represents whether the congestion occurs on a weekday or a holiday, and by time of day, and sets the extension scale of the integrated congestion section for each integrated congestion section that represents the same congestion and for each determination category.

[0011] A second aspect of the present disclosure is an information processing device that uses the extension lengths of consecutive identical congestion sections, which are classified into judgment categories by day of the week, by work attributes indicating whether the occurrence time is on a weekday or a holiday, and by time period, and the extension scale for each judgment category, to determine whether there is a bias in the extension scale of a specified congestion section for each judgment category, and if there is a bias in the extension scale of the specified congestion section within the judgment category, selects the judgment category to be used for predicting the extension length of the congestion section depending on the combination of judgment categories in which the bias in extension scale exists, and a prediction unit that predicts the extension length of the specified congestion section at a specified date and time using the extension length of the specified congestion section included in the judgment category selected by the selection unit.

[0012] A third aspect of the present disclosure is an information processing program that causes a computer to function as each unit of an information processing device.

[0013] A fourth aspect of the present disclosure is an information processing method for an information processing device including a classification unit, a determination unit, a setting unit, a selection unit, and a prediction unit, the method including: a classification step in which the classification unit classifies each congestion section represented by learning data including a start position and an end position of a congestion, by congestion direction represented by the start position and the end position; a determination step in which the determination unit, for each congestion section classified by congestion direction, connects adjacent congestion sections whose start position and end position are within a predetermined range as a continuous integrated congestion section, recursively repeating this process until the start position or the end position of the adjacent congestion section is no longer within the predetermined range; and a determination step in which the setting unit, for each integrated congestion section representing the same congestion, The method includes a setting step of classifying integrated congestion sections according to each judgment category, which is divided into categories for when congestion occurs, by day of the week, by employment attribute indicating whether it is a weekday or a holiday, and by time of day, and setting the extension scale of the integrated congestion section for each integrated congestion section that represents the same congestion and for each judgment category; a selection step in which the selection unit uses the extension scale to determine whether there is a bias in the extension scale of the specified congestion section within the judgment category, and if there is a bias in the extension scale of the specified congestion section within the judgment category, selects the judgment category to be used for predicting the extension length of the congestion section according to the combination of judgment categories in which there is a bias in extension scale; and a prediction step in which the prediction unit predicts the extension length of the specified congestion section at a specified date and time using the extension length of the specified congestion section included in the selected judgment category.

[0014] The information processing device, information processing program, and information processing method disclosed herein have the effect of being able to predict the length of sudden congestion even if a sudden congestion occurs for which conventional congestion prediction services do not have enough learning data to be used for congestion prediction.

[0015] 1 is a diagram showing an example of the functional configuration of an information processing device. FIG. 1 is a diagram showing an example of a congested section. FIG. 2 is a diagram showing an example of a connection process of congested sections. FIG. 3 is a diagram showing an example of a branched congested section. FIG. 4 is a diagram showing an example of a day of the week time slot classification table. FIG. 5 is a diagram showing an example of a working time slot classification table. FIG. 6 is a diagram showing an example of an extension scale determination table. FIG. 7 is a diagram showing an example of the configuration of the main parts of the electrical system of a computer. FIG. 8 is a flowchart showing an example of the flow of a prediction process. FIG. 9 is a flowchart showing an example of the flow of a congested section determination process. FIG. 10 is a flowchart showing an example of the flow of an extension length prediction process. FIG. 11 is a diagram showing an example of a bias in the extension tendency of congested sections classified by congestion direction. FIG. 12 is a diagram showing an example of the RMSE for each combination of determination sections when the usage rate of learning data is changed. FIG. 13 is a diagram showing an example of an analysis of a congested section with the lowest prediction accuracy of extension length. FIG. 14 is a diagram showing an example of an analysis of a congested section with a non-zero extension length but with a relatively high prediction accuracy of extension length. FIG. 15 is a diagram showing an example of a congested section in an area with complicated roads. FIG. 16 is a diagram showing an example of the RMSE for each combination of determination sections when the usage rate of learning data is changed for an area with complicated roads. 1 is a diagram showing an example of a traffic congestion display in an existing traffic congestion prediction service. FIG. 2 is a diagram showing an example of a traffic congestion detail screen in an existing traffic congestion prediction service.

[0016] The present embodiment will be described below with reference to the drawings. Note that the same components and processes are given the same reference numerals throughout the drawings, and redundant description will be omitted.

[0017] 1 is a diagram illustrating an example of a functional configuration of an information processing device 10 according to the present disclosure. The information processing device 10 is a device that predicts the future extension length of a congested section 3 based on past congestion information recorded in chronological order.

[0018] The congestion information is information that records the occurrence status of each congestion occurring at a predetermined interval, for example, every five minutes, in chronological order over a predetermined period of time. The congestion information corresponding to each congestion includes at least time series information that indicates the time series of the congestion information, such as the date and time when the congestion occurrence status was recorded, the start and end positions of the congestion. In other words, although the information that records the occurrence status of the congestion also includes other information, such as the length of the congestion, the congestion information in this embodiment only needs to include the time series information, the start and end positions of the congestion.

[0019] Traffic congestion information recorded on the same date and time contains the same time-series information. The start and end points of a traffic congestion are represented by two-dimensional coordinate values ​​using, for example, latitude and longitude.

[0020] Since the information processing device 10 uses such congestion information to predict the future extension length of the congested section 3, each piece of congestion information in chronological order will hereinafter be referred to as "learned data."

[0021] In this embodiment, the extension length of the congested section 3 refers to the absolute value of the increase or decrease between the congestion length of the congested section 3 represented by specific learning data and the congestion length of the congested section 3 represented by the immediately preceding learning data that is chronologically adjacent to the learning data and recorded at the same congestion occurrence point. In other words, if the learning data is recorded every five minutes, the extension length of the congested section 3 is the absolute value of the increase or decrease in the congestion length relative to the congested section 3 five minutes earlier.

[0022] As shown in FIG. 1, the information processing device 10 includes functional units of a congestion classification unit 11, a congestion identity determination unit 12, an extension scale setting unit 13, a section selection unit 14, and an extension length prediction unit 15, as well as a storage device 16.

[0023] The congestion classification unit 11 classifies each congestion section 3 represented by the learning data by congestion direction, with reference to the start and end positions of the congestion included in the learning data. In other words, the congestion classification unit 11 is an example of a classification unit that classifies each learning data by congestion direction.

[0024] Although there are no restrictions on the congestion direction to be classified, in this embodiment, as an example, the congestion section 3 represented by the learning data is classified into four directions: west to east, east to west, south to north, and north to south. Therefore, just as there are uphill and downhill sections on the same road, the congestion section 3 occurring between opposite lanes is classified into different directions, for example, west to east and east to west.

[0025] The congestion classification unit 11 stores the learning data classified by congestion direction in the storage device 16 as direction-specific learning data information 16A.

[0026] On the other hand, when congestion occurs intermittently, a congestion section 3 that forms a single congestion as a whole may be recognized as a plurality of small, divided congestion sections 3. In such a case, it is preferable to treat each congestion section 3 as a fragment of the congestion section 3 that constitutes a continuous congestion section 3, rather than treating each congestion section 3 as an independent congestion section 3.

[0027] Therefore, the congestion identity determination unit 12 acquires learning data classified by congestion direction from the direction-specific learning data information 16A and determines up to which congestion sections 3 constitute a continuous congestion section 3. Such a continuous congestion section 3 represented by a plurality of congestion sections 3 is referred to as an "integrated congestion section 5." Hereinafter, the integrated congestion section 5 will be referred to as a "congestion section 5." The congestion identity determination unit 12, which determines the range of the congestion section 5, is an example of a determination unit.

[0028] The traffic congestion identity determination unit 12 acquires learning data classified by traffic congestion direction from the direction-specific learning data information 16A, and converts the start and end points of the traffic congestion contained in each learning data into geohashes 1.

[0029] A geohash 1 is an example of a predetermined range, and refers to each divided area on Earth based on latitude and longitude. Each geohash 1 is represented by a different multi-digit symbol, and as the number of digits representing the geohash 1 increases, the range of the area represented by the geohash 1 becomes narrower and the accuracy of the geohash 1 increases.

[0030] 2 is a diagram showing an example of a display on a map of the start and end positions of a traffic jam represented by a traffic jam section 3. The example in FIG. 2 shows how the endpoints of the traffic jam included in the learning data, i.e., the start and end positions of the traffic jam, have been converted into geohash 1A represented by the symbol "xn76q41" and geohash 1B represented by the symbol "xn76q2c," respectively. When describing each geohash 1 separately, such as geohash 1A and geohash 1B, they are distinguished by adding an alphabet to the end of geohash 1.

[0031] The congestion identity determination unit 12 selects learning data for which it has not yet been determined whether the congestion section 3 is a congestion section 3 that constitutes a congestion section 5. The congestion identity determination unit 12 performs a process of linking another congestion section 3 that has a geohash 1 (referred to as a "start geohash 1") corresponding to the start position of the congestion section 3 (referred to as a "selected congestion section 3") represented by the selected learning data as a geohash 1 (referred to as a "end geohash 1") corresponding to the end position of the congestion, to the selected congestion section 3. In other words, the congestion identity determination unit 12 performs a process of linking another congestion section 3 that has the start geohash 1 of the selected congestion section 3 as its end geohash 1 to the selected congestion section 3. Then, the congestion identity determination unit 12 regards the linked congestion section 3 as the newly selected congestion section 3 and recursively repeats the process of linking congestion sections 3 until there is no other congestion section 3 that has the start geohash 1 of the selected congestion section 3 as its end geohash 1.

[0032] The congestion identity determination unit 12 also performs a process of linking another congested section 3 having the end geohash 1 of the selected congested section 3 as its start geohash to the selected congested section 3, using the end geohash 1 of the selected congested section 3 as the starting point geohash. Then, the congestion identity determination unit 12 sets the linked congested section 3 as the newly selected congested section 3 and recursively repeats the process of linking congested sections 3 until there are no more congested sections 3 having the end geohash 1 of the selected congested section 3 as their start geohash 1.

[0033] By performing the above process on each piece of learning data classified into the same congestion direction, a congestion section 5 is generated.

[0034] 3 is a diagram showing an example of the process of connecting congested sections 3 displayed on a map. The example of FIG. 3 shows the process of generating congested section 5 having geohash 1G represented by the symbol "xn76v8b" as its start point geohash 1 and geohash 1E represented by the symbol "xn76tpv" as its end point geohash 1 by connecting congested section 3B having geohash 1C represented by the symbol "xn76v23" as its start point geohash 1, which is the end point geohash 1 of congested section 3A, congested section 3C having geohash 1D represented by the symbol "xn76v0p" as its start point geohash 1, which is the end point geohash 1 of congested section 3B, and congested section 3D having geohash 1F represented by the symbol "xn76v27" as its end point geohash 1, which is the start point geohash 1 of congested section 3A, to congested section 3A.

[0035] 3 shows an example in which the start position of the congestion represented by one of the congested sections 3 to be connected overlaps with the end position of the congestion represented by the other congested section 3. On the other hand, even if the start position of the congestion represented by one congested section 3 and the end position of the other congested section 3 are far apart, it goes without saying that adjacent congested sections 3 may be connected as long as the respective start and end positions are included in the same geohash 1. Furthermore, even if the start position of the congestion represented by the newly connected congested section 3 is not included in the same geohash 1 as the end position of the congestion represented by the other congested section 3, the congested section 3 may be connected as long as the start position is included in the geohash 1 of any of the congested sections 3 that were connected as a continuous congested section 3 in the previous connection operation.

[0036] In other words, for each congestion section 3 classified by congestion direction, the congestion identity determination unit 12 recursively repeats the process of connecting adjacent congestion sections 3 whose starting point and ending point are within a predetermined range as a continuous congestion section 3 until there are no starting point or ending point positions of adjacent congestion sections 3 within the predetermined range, thereby determining which congestion sections 3 represent a continuous congestion.

[0037] On the other hand, Figure 4 is a diagram showing an example of a congested section 5 displayed on a map. For example, in urban areas where roads are densely packed, the congested section 5 may not be represented by a single line, but may be represented by multiple lines as the congested section 5 branches off at point P3, which corresponds to a road branch point, as shown in Figure 4. The existence of a branch point on a road means that one of the roads branching off from the branch point is managed as a separate road from the other roads.

[0038] Therefore, it is preferable to treat the branching congested section 5 as a different congested section 5 with the branching point as the boundary.

[0039] Specifically, in the case of the congested section 5 shown in FIG. 4, it is preferable to handle it as being divided into a congested section 5A connecting points P1 and P4 and a congested section 5B connecting points P3 and P5.

[0040] 1 includes a dividing unit 12A. The dividing unit 12A divides the congestion section 5 using the azimuth angles of adjacent congestion sections 3 that make up the congestion section 5. The azimuth angle of the congestion section 3 is a value that represents the direction of travel of the congestion section 3 as an angle, where the north direction is 0 degrees, the east direction is 90 degrees, the south direction is 180 degrees, and the west direction is 270 degrees.

[0041] The dividing unit 12A calculates the difference in azimuth angle between adjacent congestion sections 3 that make up the congestion section 5, i.e., the azimuth angle difference. If the azimuth angle difference between the congestion sections 3 is equal to or greater than a predetermined angle, the dividing unit 12A determines that the adjacent congestion sections 3 are congestion sections 3 that make up different congestion sections 5, and divides the congestion section 5 at the junction point between the congestion sections 3.

[0042] In the congested section 5 shown in Figure 4, if the azimuth angle of one adjacent congested section 3 at point P2 is 28.54 degrees and the azimuth angle of the other congested section 3 is 58.49 degrees, the azimuth angle difference of the congested sections 3 is 29.95 degrees. Also, if the azimuth angle of one adjacent congested section 3 at point P3 is 18.07 degrees and the azimuth angle of the other congested section 3 is 81.79 degrees, the azimuth angle difference of the congested sections 3 is 63.72 degrees.

[0043] Therefore, if the threshold value of the azimuth angle difference of the congested section 3 is set to, for example, 40 degrees, it will be clear that adjacent congested sections 3 at point P2 constitute a continuous congested section 5, and adjacent congested sections 3 at point P3 constitute different congested sections 5. In this way, the dividing unit 12A divides the congested section 5 into congested sections 5A and 5B. Note that the threshold value of the azimuth angle difference of the congested section 3 is an example and is set according to road conditions.

[0044] In other words, when the difference in azimuth angles of adjacent congestion sections 3 that make up the congestion section 5 is greater than or equal to a predetermined angle, the dividing unit 12A divides the congestion section 5 into two different congestion sections 5 at the junction of adjacent congestion sections 3 where the difference in azimuth angles is greater than or equal to the predetermined angle.

[0045] In this way, the congestion identity determination unit 12, in cooperation with the division unit 12A, determines the range of the congestion section 5 for the learning data classified by congestion direction, and calculates the extension length of each congestion section 5 from the change in the congestion section 5 over time. The extension length of the congestion section 5 is expressed in predetermined units, for example, 10 m units.

[0046] The congestion identity determination unit 12 performs the above-determined congestion section 5 on the learning data in each time series classified by congestion direction. Then, the congestion identity determination unit 12 classifies the congestion sections 5 into groups of congestion sections 5 that can be considered to be the same congestion, and stores information associating time series information and extension length for each congestion section 5 as congestion section information 16B in the storage device 16. Note that whether or not the congestion is the same can be determined, for example, by whether or not the congestion sections 5 are connected starting from the same geohash 1. Hereinafter, a set of congestion sections 5 connected starting from the same geohash 1 will be referred to as the "same congestion section 5."

[0047] The extension scale setting unit 13 acquires the congested section 5 for each of the same congested sections 5 from the congested section information 16B.

[0048] In order to acquire the extension tendency of each congested section 5, the extension scale setting unit 13 classifies each congested section 5 into a predetermined judgment category according to the time of occurrence of congestion for each congested section 5. The judgment category is a category set for determining whether or not a significant difference, i.e., a bias, is observed in the extension tendency of the congested sections 5 belonging to that category, and is set in advance by, for example, a user.

[0049] In this embodiment, as an example, the congested section 5 is classified into the judgment categories of day of the week, work, and time period, but other judgment categories, such as classification into days that are multiples of 5, such as the 5th and 10th, and other days, may also be used. Classification may also be by month, by season (spring, summer, fall, winter), or by year. Classifying the congested section 5 by season makes it easier to predict the length of a congested section 5 that is seasonal, such as sudden congestion caused by cherry blossom viewing visitors that began occurring only in the spring two years ago.

[0050] The day-of-week classification is a classification indicating on which of the seven days from Sunday to Saturday each congested section 5 occurred, and the time-of-day classification is a classification indicating in which of the 24 time periods into which a day is divided into one hour each congested section 5 occurred. Also, the work classification is a classification indicating whether each congested section 5 occurred on a weekday or on a holiday.

[0051] Therefore, the extension scale setting unit 13 uses a day of the week and time zone classification table 19A, which is classified by a combination of day of the week classification and time zone classification, and classifies each congestion section 5 into one of the 168 classifications in the day of the week and time zone classification table 19A for each congestion section 5.

[0052] 5 is a diagram showing an example of the day-of-week and time-slot classification table 19A. For example, if congested section 5 occurs after 7:00 and before 8:00 on Friday, it is classified into the category represented by "F7."

[0053] By referring to the day-of-week and time-slot division table 19A, it is possible to obtain congested sections 5 classified by day-of-week division and time-slot division. Although the day is divided into one-hour units in the day-of-week and time-slot division table 19A shown in Fig. 5, it goes without saying that the day may be divided into other units, such as two-hour units.

[0054] In addition, the extension scale setting unit 13 uses a working time zone classification table 19B, which is classified by a combination of working zones and time zone classifications, to classify each of the same congested sections 5 into one of the 48 classifications in the working time zone classification table 19B.

[0055] 6 is a diagram showing an example of the work time zone classification table 19B. For example, if a congested section 5 occurs after 14:00 and before 15:00 on a holiday, it is classified into the category represented by "H14." While what constitutes a holiday varies from person to person, in this embodiment, Saturdays, Sundays, and public holidays are defined as holidays, and days other than holidays are defined as weekdays.

[0056] By referring to the work hour division table 19B, it is possible to obtain congested sections 5 classified by time period division and work category. Note that, since the congested sections 5 classified by time period division can also be obtained from the day of week time period division table 19A, it is not necessary to divide the work categories by time period as in the work hour division table 19B shown in Figure 6, and they may be simply divided into two categories: weekdays and holidays.

[0057] In this way, the extension scale setting unit 13 stores information classifying each congestion section 5 into a judgment category for each congestion section 5 as category-specific congestion section information 16C in the storage device 16. In other words, category-specific congestion section information 16C exists for each congestion section 5. For ease of explanation, the day of the week and time zone category table 19A and the working time zone category table 19B may be collectively referred to as the "category table 19."

[0058] Meanwhile, the extension scale setting unit 13 sorts the extension lengths of the congestion sections 5 in ascending or descending order for each congestion section 5 that is chronologically continuous among the same congestion sections 5 in each judgment category, and identifies the extension length located in the middle of the sorting, i.e., the median of the extension lengths. A congestion section 5 that is chronologically continuous refers to a congestion section 5 that is included in the range from when congestion occurs to when it disappears, and hereafter, a congestion section 5 that is chronologically continuous is referred to as a "continuous congestion section 5."

[0059] Then, the extension scale setting unit 13 calculates the average of the median extension length across all sections and the standard deviation of the median extension length for each judgment category, and generates an extension scale determination table 4 for each judgment category. The extension scale determination table 4 is a table for determining the extension scale of the congestion section 5 classified into the corresponding judgment category.

[0060] FIG. 7 is a diagram showing an example of the extension scale determination table 4. In the extension scale determination table 4, determination criteria are defined for each extension scale. In the example of the extension scale determination table 4 shown in FIG. 7, if the extension length of the congested section 5 is equal to or greater than (all-section average of median extension length + standard deviation), the extension scale is set to "large." If the extension length of the congested section 5 is equal to or greater than (all-section average of median extension length - standard deviation) and less than (all-section average of median extension length + standard deviation), the extension scale is set to "medium." Furthermore, if the extension length of the congested section 5 is equal to or greater than 0 and less than (all-section average of median extension length - standard deviation), the extension scale is set to "small." Note that the "average median" in FIG. 7 refers to the all-section average of median extension length. In this embodiment, the extension scale of the congested section 5 is set to three levels: "large," "medium," and "small." However, it goes without saying that other granularities, such as five levels or two levels, may also be set.

[0061] Since the extension length of the congested section 5 is expressed in units of 10 m, the extension scale setting unit 13 rounds up the units of each of the values ​​representing (all section average of median extension length - standard deviation) and (all section average of median extension length + standard deviation) so that the extension length serving as the judgment criterion is expressed in units of 10 m. Furthermore, if the value of (all section average of median extension length - standard deviation) is 0 or less, the value of (all section average of median extension length - standard deviation) is set to "10".

[0062] The extension scale setting unit 13 compares the extension length of each congestion section 5 classified into a judgment category with the extension scale judgment table 4 corresponding to the judgment category, identifies the extension scale for each congestion section 5, and sets the extension scale with the largest number as the extension scale for the congestion section 5 in that judgment category.

[0063] Specifically, the extension scale setting unit 13 classifies the extension scale of each congested section 5 that occurs for each day of the week and time period into either "large," "medium," or "small" by referring to the extension scale determination table 4, tallies the number of cases classified as "large," "medium," or "small," and then determines the extension scale for that determination category as either "large," "medium," or "small" by majority vote of the number of cases. There may be two extension scales with the most classifications. For example, if the number of cases classified as "small" and the number of cases classified as "medium" on Monday are tied for first place, with both being 10, the extension scale of the congested section 5 on Monday is determined to be "medium." If the number of cases classified as "small" and the number of cases classified as "large" are tied for first place, the extension scale of the congested section 5 on Monday is determined to be "medium." If the number of cases classified as "small" and the number of cases classified as "large" are tied for first place, the extension scale of the congested section 5 on Monday is determined to be "medium." If the number of cases classified as "medium" and the number of cases classified as "large" are tied for first place, the extension scale of the congested section 5 on Monday is determined to be "large."

[0064] That is, the extension scale setting unit 13 sets the extension scale of the congested section 5 for each congested section 5 and for each judgment category. The extension scale setting unit 13 stores the extension scale of the congested section 5 set for each congested section 5 and for each judgment category in the storage device 16 as extension scale information 16D.

[0065] The extension scale setting unit 13 that sets the extension scale of the congested section 5 for each congested section 5 and for each judgment division in this manner is an example of a setting unit.

[0066] When the category selection unit 14 receives prediction request information including the congested section 5 whose extension length is to be predicted and the date and time, the category selection unit 14 uses the extension scale information 16D set by the extension scale setting unit 13 to determine whether or not there is a bias within the judgment category in the extension scale of the congested section 5 specified by the prediction request information, i.e., the congested section 5 whose extension length is to be predicted. Specifically, the category selection unit 14 determines whether or not there is a bias in at least one of the judgment categories by day of the week, by time of day, and by employment attribute regarding the extension scale of the specified congested section 5.

[0067] If there is a bias in at least one judgment section, the section selection unit 14 selects a judgment section to be used for predicting the extension length of the congested section 5 according to the combination of judgment sections in which there is a bias in the extension scale.

[0068] For example, if there is a bias in the day-of-week classification, the classification selection unit 14 selects the day-of-week classification; if there is a bias in the work classification, the classification selection unit 14 selects the work classification; and if there is a bias in the time-slot classification, the classification selection unit 14 selects the time-slot classification. Furthermore, if there is a bias in the day-of-week classification and the time-slot classification, the classification selection unit 14 selects the day-of-week classification and the time-slot classification; if there is a bias in the work classification and the time-slot classification, the classification selection unit 14 selects the work classification and the time-slot classification. Note that since both the day-of-week classification and the work classification are classifications based on the day of the week, if there is a bias in the day-of-week classification and the work classification, it is sufficient to consider that there is a bias only in the classification with the stronger degree of bias. If the bias is the same between the day-of-week classification and the work classification, the work classification is given priority. This increases the amount of data used for prediction, making it possible to improve the prediction accuracy of the extension length of the congested section 5.

[0069] As will be described later, the determination category selected by the category selection unit 14 is used to predict the extension length of the congested section 5 specified by the prediction request information.

[0070] In other words, the category selection unit 14 uses the extension scale for each congestion section and for each judgment category obtained from the average of all sections and the standard deviation of the median extension length for each consecutive congestion section 5, which are classified into judgment categories based on the occurrence time by day of the week, employment attribute, and time of day, to determine whether there is a bias in the extension scale of the specified congestion section 5 within the judgment category, and if there is a bias in the extension scale of the specified congestion section 5 within the judgment category, selects the judgment category to be used for predicting the extension length of the congestion section 5 depending on the combination of judgment categories in which there is a bias in the extension scale.

[0071] The category selection unit 14 that selects the determination category used to predict the extension length of the congested section 5 is an example of a selection unit.

[0072] The extension length prediction unit 15 obtains the extension length of each specified congested section 5 classified according to the judgment category selected by the category selection unit 14 from the category-specific congested section information 16C, and predicts the extension length of the specified congested section 5 at the specified date and time.

[0073] For example, assume that the determination category selected by the category selection unit 14 is a day-of-week category, and the date and time specified in the prediction request information is Monday. In this case, the extension length prediction unit 15 refers to the day-of-week / time-zone category table 19A of the specified congested section 5 constituting the category-specific congested section information 16C, and outputs the average of the extension lengths of all congested sections 5 classified as Monday as the predicted value of the extension length of the specified congested section 5. In the example of the day-of-week / time-zone category table 19A shown in FIG. 5 , the average of the extension lengths of the congested sections 5 classified into the categories "M0" to "M23" becomes the predicted value of the extension length of the congested section 5 specified in the prediction request information.

[0074] Also, for example, assume that the judgment category selected by the category selection unit 14 is a time slot category, and the date and time specified by the prediction request information corresponds to a time slot from 18:00 onwards to less than 19:00. In this case, the extension length prediction unit 15 references the day of week and time slot category table 19A of the specified congested section 5 constituting the category-specific congested section information 16C, and outputs the average of the extension lengths of all congested sections 5 classified into the time slot from 18:00 onwards to less than 19:00 as the predicted value of the extension length of the specified congested section 5. In the example of the day of week and time slot category table 19A shown in FIG. 5 , the average of the extension lengths of the congested sections 5 classified into the categories “M18”, “Tu18”, “W18”, “Th18”, “F18”, “Sa18”, and “Su18” becomes the predicted value of the extension length of the congested section 5 specified by the prediction request information.

[0075] Also, for example, suppose the determination categories selected by the category selection unit 14 are day-of-week categories and time-slot categories, and the date and time specified by the prediction request information corresponds to a time slot from 18:00 onwards to 19:00 on Monday. In this case, the extension length prediction unit 15 references the day-of-week and time-slot category table 19A of the specified congested section 5 constituting the category-specific congested section information 16C, and outputs the average of the extension lengths of all congested sections 5 classified in the time slot from 18:00 onwards to 19:00 on Monday as the predicted value of the extension length for the specified congested section 5. In the example of the day-of-week and time-slot category table 19A shown in FIG. 5 , the average of the extension lengths of the congested sections 5 classified in the "M18" category becomes the predicted value of the extension length for the congested section 5 specified by the prediction request information.

[0076] The extension length prediction unit 15 that predicts the extension length of the congested section 5 specified by the prediction request information is an example of a prediction unit.

[0077] In the present embodiment, the information processing device 10 includes the storage device 16. However, the information processing device 10 does not necessarily need to include the storage device 16. In this case, the information processing device 10 may store and acquire various information between the storage device 16 connected to a communication line, for example, via a communication unit 27 (described later). Alternatively, the information processing device 10 may be separated into a first device including a congestion classification unit 11, a congestion identity determination unit 12, and an extension scale setting unit 13, and a second device including a section selection unit 14 and an extension length prediction unit 15. The first device is an example of an information processing device 10 that receives learning data as input and outputs section-specific congestion section information 16C and extension scale information 16D. The second device is an example of an information processing device 10 that receives section-specific congestion section information 16C, extension scale information 16D, and prediction request information as input and outputs a predicted value of the extension length in the congestion section 5 specified by the prediction request information.

[0078] In the example shown above, the extension length of the congested section 5 is predicted, but the present invention can also be applied to predicting the congestion length itself of the congested section 5. If the congestion length is predicted, it becomes possible to notify the driver of the vehicle of the occurrence of the congested section 5 without referring to past congestion information.

[0079] The information processing device 10 having such functions is configured using, for example, a computer 20. Fig. 8 is a diagram showing an example of the configuration of the main parts of the electrical system of the computer 20 applied to the information processing device 10.

[0080] 1 . The computer 20 also includes a ROM (Read Only Memory) 22 that stores an information processing program that causes the computer 20 to function as the information processing device 10, and a RAM (Random Access Memory) 23 that is used as a temporary work area for the CPU 21. The computer 20 also includes a non-volatile memory 24 and an input / output interface (I / O) 25. The CPU 21, ROM 22, RAM 23, non-volatile memory 24, and I / O 25 are each connected by a bus 26.

[0081] The nonvolatile memory 24 is an example of a storage device 16 that maintains stored information even if the power supplied to the nonvolatile memory 24 is cut off, and is, for example, a semiconductor memory or a hard disk. The nonvolatile memory 24 does not necessarily have to be included in the computer 20; for example, a portable nonvolatile memory 24 that is detachable from the computer 20 may be used.

[0082] The non-volatile memory 24 stores, for example, direction-specific learning data information 16A, congested section information 16B, category-specific congested section information 16C, and extension scale information 16D.

[0083] To the I / O 25, for example, a communication unit 27, an input unit 28, and a display unit 29 are connected.

[0084] The communication unit 27 is connected to a communication line such as the Internet or a local area network (LAN), and includes a communication protocol for performing data communication with an external device (not shown) also connected to the communication line. The CPU 21 receives, for example, learning data from the external device through the communication line connected to the communication unit 27. The communication line connected to the communication unit 27 may be either wired or wireless.

[0085] The input unit 28 is a device that receives instructions from the user and notifies the CPU 21 of the contents of the received instructions, and may be, for example, a button, a touch panel, a keyboard, or a mouse.

[0086] The display unit 29 is an example of a device that visually displays information processed by the CPU 21, and may be, for example, a liquid crystal display or an organic EL (Electro Luminescence) display.

[0087] When the information processing device 10 receives user instructions from an external device via a communication line and transmits information processed in accordance with the instructions to the external device via the communication line, the input unit 28 and the display unit 29 do not necessarily need to be connected to the I / O 25.

[0088] Next, the operation of the information processing device 10 of the present disclosure will be described.

[0089] FIG. 9 is a flowchart showing an example of the flow of the prediction process executed by the CPU 21 of the information processing device 10.

[0090] An information processing program that defines the prediction process is stored in advance in, for example, the ROM 22 of the information processing device 10. The CPU 21 of the information processing device 10 executes the prediction process by reading the information processing program stored in the ROM 22. It is assumed that learning data is stored in advance in the non-volatile memory 24.

[0091] In step S10 of FIG. 9, the CPU 21 executes a congested section determination process to determine the congested section 3 from the learning data and generate the congested section information 16B.

[0092] Next, in step S20, the CPU 21 executes an extension length prediction process using the congested section information 16B generated in step S10 to predict the extension length of the congested section 3 specified by the prediction request information at the specified date and time, and then terminates the prediction process shown in Figure 9.

[0093] FIG. 10 is a flowchart showing an example of the flow of the congested section determination process executed in step S10 of the prediction process shown in FIG.

[0094] 10, the CPU 21 acquires learning data from the nonvolatile memory 24 and classifies each piece of acquired learning data by traffic congestion direction, thereby generating direction-specific learning data information 16A.

[0095] In step S110, the CPU 21 selects one of the traffic congestion directions into which the learning data has been classified. The traffic congestion direction selected in step S110 is referred to as a selected traffic congestion direction.

[0096] In step S120, the CPU 21 selects from the direction-specific learning data information 16A the learning data containing the oldest time series information among the unselected learning data classified as the selected congestion direction. If there are multiple learning data containing time series information representing the same date and time, multiple learning data will be selected. The learning data selected in step S120 is called selected learning data.

[0097] In step S130, the CPU 21 converts the start and end points of the traffic jam included in the selected learning data into geohashes 1.

[0098] In step S140, the CPU 21 uses the geohash 1 converted in step S130 to determine which of the congested sections 3 represented by each selected learning data is a congested section 5 that constitutes a continuous congested section 3, and generates the congested section 5.

[0099] In step S150, the CPU 21 calculates the length of each congested section 5, classifies the congested sections 5 into groups of congested sections 5 that can be considered to be the same congestion, and compares the length of each congested section 5 with the length of the congested section 5 observed in the immediately preceding time series to calculate the extension length of each congested section 5 included in the same congested section 5. For each congested section 5, the CPU 21 associates the calculated extension length with the time series information included in the selected learning data.

[0100] The length of the congested section 5 can be obtained by, for example, referring to map data stored in advance in the non-volatile memory 24 and calculating the distance along the road between the start point and the end point of the congested section 5. If there is no congested section 5 observed in the immediately preceding time series for the generated congested section 5, the CPU 21 may set the extension length to "0".

[0101] In step S160, the CPU 21 determines whether or not there is unselected learning data among the learning data classified in the selected congestion direction, and if there is unselected learning data, the CPU 21 proceeds to step S120 and repeatedly executes steps S120 to S160. As a result, the learning data classified in the selected congestion direction are selected in chronological order, and the time-series information and the extension length are associated with each congestion section 5 in the selected congestion direction.

[0102] On the other hand, if it is determined in the determination process of step S160 that there is no unselected learning data, the process proceeds to step S170.

[0103] In step S170, the CPU 21 determines whether or not there is an unselected congestion direction among the congestion directions for which the learning data has been classified, and if there is an unselected congestion direction, the CPU 21 proceeds to step S110 and repeatedly executes steps S110 to S170. As a result, the time-series information and the extension length are associated with each congestion section 5 in all congestion directions, and congestion section information 16B is generated.

[0104] This completes the congested section determination process shown in FIG.

[0105] On the other hand, FIG. 11 is a flowchart showing an example of the flow of the extension length prediction process executed in step S20 of the prediction process shown in FIG.

[0106] 11, the CPU 21 selects the same congested section 5 from the congested section information 16B. The congested section 5 selected in step S200 is referred to as a selected congested section 5.

[0107] In step S210, the CPU 21 classifies the selected congested section 5 in accordance with the classification table 19, and generates classification-specific congested section information 16C.

[0108] In step S220, the CPU 21 generates an extension scale determination table 4 for each determination category using the average of all sections of the median extension length calculated from the extension lengths of the congested sections 5 classified into the determination category and the standard deviation of the median extension length.

[0109] In step S230, the CPU 21 compares the extension length of each consecutive congestion section 5 among the congestion sections 5 classified into each judgment category with the extension scale determination table 4 for the judgment category into which the consecutive congestion sections 5 are classified, and determines which extension scale the extension length of the consecutive congestion sections 5 is classified. Then, the CPU 21 sets the extension scale for the congestion section 5 in that judgment category to the extension scale for the congestion section 5. Note that if there are multiple extension scales with the most extension lengths, the CPU 21 sets the extension scale for the congestion section 5 in the judgment category based on a predetermined rule. For example, if the number of extension scale classifications is tied for first place between "small" and "medium," or "medium" and "large," the larger extension scale is adopted. If the number of extension scale classifications is tied for first place between "large" and "small," the CPU 21 adopts "medium," which is an extension scale between "large" and "small." That is, the CPU 21 sets the extension scale of the selected congested section 5 for each determination section.

[0110] In step S240, the CPU 21 determines whether or not there is an unselected congested section 5 in the congested section information 16B, and if there is an unselected congested section 5, the CPU 21 proceeds to step S200 and repeatedly executes steps S200 to S240. As a result, the extension scale of the congested section 5 for each congested section 5 and for each judgment category based on the learning data is obtained.

[0111] In step S250, the CPU 21 determines whether or not there is a bias in the extension scale of the congested section 5, the extension length of which is to be predicted, that is, the prediction target congested section 6, for each determination section.

[0112] Specifically, the CPU 21 determines that no bias exists if the extension scales of the prediction target congestion sections 6 in the determination section are all the same, and determines that bias exists if at least one extension scale is different from the other extension scales. Note that the determination of the presence or absence of bias in the determination section is not limited to this, and needless to say, the determination may be made by other methods.

[0113] In step S260, the CPU 21 selects a combination of the judgment sections for which it was determined in step S250 that a bias exists. If there are no judgment sections for which a bias exists, the CPU 21 does not select any judgment sections. In this case, the length of the congested section 5 is predicted by referring to the congested section 5 for all days of the week and all time periods.

[0114] The CPU 21 predicts, in the selected judgment category, the average of the extension lengths of all the congested sections 5 classified into the category corresponding to the specified date and time of the prediction target congested section 6 as the extension length at the specified date and time of the prediction target congested section 6. Then, the CPU 21 outputs the predicted extension length.

[0115] The CPU 21 may output the predicted extension length in any form as long as the form allows the user to recognize the predicted extension length on the information processing device 10. For example, the CPU 21 may display the predicted extension length on the display unit 29, or may transmit the predicted extension length to an external device connected to a communication line via the communication unit 27. The CPU 21 may also print the predicted extension length on paper or store the predicted extension length in the non-volatile memory 24. If the predicted extension length is equal to or greater than a certain value, it may be possible to use the predicted extension length by changing the route guided in the road guidance navigation to a detour route that bypasses the congested section 5.

[0116] As a result, the extension length prediction process shown in FIG. 11 is completed, and the prediction process shown in FIG. 9 is also completed.

[0117] <Verification Results> Next, the verification results of the extension length of the congested section 5 predicted using the information processing device 10 according to the present embodiment will be shown.

[0118] To conduct this verification, learning data from October 2019 to February 2021 in a predetermined verification area A was collected from congestion information from the Japan Road Traffic Information Center (JARTIC), a public interest incorporated foundation, and the extension length of the congested section 5 predicted using each extension scale information 16D generated by changing the number of learning data was compared with the actual extension length of the congested section 5. The congested section 5 illustrated on the maps in Figures 14, 15, 16, and 17 is the learning result obtained by learning using congestion information obtained from the Japan Road Traffic Information Center (JARTIC), a public interest incorporated foundation.

[0119] Specifically, the RMSE between the extension length of the congested section 5 predicted using the extension scale information 16D, i.e., the predicted extension length of the congested section 5, and the actual extension length of the congested section 5, i.e., the actual extension length of the congested section 5, was calculated, and the prediction accuracy of the congested section 5 in the information processing device 10 was evaluated.

[0120] RMSE is an abbreviation for "Root Mean Squared Error." RMSE is an example of an accuracy evaluation index expressed as the root mean square of the difference between the predicted extension length of the congested section 5 and the actual extension length of the congested section 5, and is calculated by equation (1).

[0121]

[0122] In equation (1), "n" is the predicted extension length of the congestion section 5, "k" is the index of the predicted extension length, and "f ｋ " is the kth predicted extension length, and "y ｋ " represents the kth actual extension length.

[0123] As can be seen from equation (1), RMSE is an accuracy evaluation index that indicates that the closer the value is to 0, the higher the prediction accuracy.

[0124] The information processing device 10 classified the learning data representing the congestion information of the verification area A by congestion direction and determined the congestion section 5 using the 8-digit geohash 1, and the congestion section 5 was classified into 63 sections.

[0125] FIG. 12 is a diagram showing an example of the number of congested sections 5 with bias in extension tendency for each judgment section, in which 63 congested sections 5 are grouped by congestion direction.

[0126] The denominator of the numerical value in each column in Figure 12 represents the number of congestion sections 5 classified into congestion directions corresponding to the row direction, and the numerator of the numerical value represents the number of congestion sections 5 in each congestion direction where there is a bias in the judgment classification corresponding to the column direction.

[0127] 13 is a diagram showing an example of RMSE for each combination of judgment sections taking into consideration bias in the extension tendency when predicting the extension length of congested section 5 while changing the usage rate of learning data from 10% to 100% in 10% increments. Note that the RMSE was 163.7245 when all extension lengths of congested section 5 were set to "0." Note that the unit of RMSE is meters.

[0128] 13 , when the proportion of learning data used to predict the extension length of the congested section 5 is in the range from 10% to 40% of the total learning data counted from the oldest, prediction accuracy of the extension length tends to be higher when predicting the extension length of the congested section 5 taking into account the bias in the extension trend than when predicting the extension length of the congested section 5 without taking into account the bias in the extension trend. In other words, it is shown that as the number of learning data used to predict the extension length of the congested section 5 decreases, prediction accuracy of the extension length increases when predicting the extension length of the congested section 5 taking into account the bias in the extension trend.

[0129] Next, among the 63 classified congested sections 5, the characteristics of congested sections 5 with relatively low RMSE, i.e., congested sections 5 with low prediction accuracy, and the characteristics of congested sections 5 with relatively high RMSE, i.e., congested sections 5 with high prediction accuracy, are analyzed.

[0130] FIG. 14 is a diagram showing an example of analysis of the congested section 5 that had the lowest prediction accuracy of the extension length out of the 63 classified congested sections 5.

[0131] The congested section 5 analyzed in Figure 14 is a congested section 5 heading from east to west, and is the congested section 5 included in geohash 1 of "xn76s76g". In addition, the number of congestion occurrences by extension scale in the congested section 5 analyzed in Figure 14 was 50 "small", 24 "medium", and 105 "large", and the RMSE was "399.4".

[0132] 14, in the congested section 5 being analyzed, an extension length exceeding 1000 m occurred in the time period from after 18:00 to before 21:00, which is thought to have resulted in a relatively large error in the predicted extension length of the congested section 5 when biases in extension trends by day of the week and by time period are taken into consideration, and a high RMSE. In other words, in the congested section 5, which is more likely to experience congestion with an extension length exceeding 1000 m than the other congested sections 5, there is a tendency for the prediction accuracy of the extension length to decrease.

[0133] FIG. 15 is a diagram showing an example of analysis of the congested section 5 that had the highest prediction accuracy of the extension length among the 63 classified congested sections 5.

[0134] The congested section 5 to be analyzed shown in Figure 15 is a congested section 5 heading from north to south, and is the congested section 5 included in geohash 1 of "xn76svef". Furthermore, the number of congestion occurrences by extension scale in the congested section 5 to be analyzed shown in Figure 15 was 89 "small", 0 "medium", and 0 "large", and the RMSE was "0.0".

[0135] In the congested section 5 analyzed in Figure 15, the number of congestion occurrences was only 89, which is less than half the number of congestion occurrences in the congested section 5 analyzed in Figure 14. Moreover, the extension lengths of the congested sections 5 analyzed in Figure 15 were all "0", and the extension scale of the congested sections 5 was all classified as "small". In other words, there is a tendency that the fewer the number of congestion occurrences and the closer the extension length of the congested section 5 is to "0", the higher the prediction accuracy of the extension length.

[0136] On the other hand, among the congested sections 5, there are congested sections 5 whose extension length is not "0" but whose extension length prediction accuracy is higher than that of the other congested sections 5.

[0137] FIG. 16 is a diagram showing an example of analysis of such a congested section 5. The congested section 5 to be analyzed shown in FIG. 16 is a congested section 5 heading from west to east, and is the congested section 5 included in geohash 1 of "xn76ubgy". Furthermore, the number of congestion occurrences by extension scale in the congested section 5 to be analyzed shown in FIG. 16 was 496 for "small", 1608 for "medium", and 1143 for "large", with an RMSE of "79.4".

[0138] In the congested section 5 analyzed in Figure 16, congestion occurs between 8:00 and 18:00 on weekdays, and the number of congestion occurrences is greater than in the congested section 5 analyzed in each of Figures 14 and 15, and there was a tendency for extension to occur, with the most congestion in the congested section 5 having a "medium" extension scale. In other words, it is thought that there was a tendency for congestion classified as a "medium" extension scale to occur, and there was a bias in the time periods and days of the week when extension occurred, so that in the congested section 5, there was a tendency for the prediction accuracy of the extension length to be high.

[0139] It is preferable that the congested section 5 shown in Figure 16 be displayed along the dotted line 7 representing the road, but due to errors in the start and end positions of the congested section 3 contained in the learning data, the congested section 5 is displayed offset from the dotted line 7.

[0140] So far, we have described the verification results of the extension length of the congested section 5 predicted in the verification area A. Next, we will describe the verification results of the extension length of the congested section 5 predicted in the verification area B, which is different from the verification area A.

[0141] The period for collecting learning data in verification area B was the same as the period for collecting learning data in verification area A, but the number of learning data collected in verification area B during the same period was approximately 6.7 times the number of learning data in verification area A. Furthermore, verification area B has roads that are more complicated than verification area A.

[0142] When the information processing device 10 classified the learning data representing congestion information for the verification area B by congestion direction and determined the congested sections 5 using the 8-digit geohash 1, the data was classified into 204 congested sections 5. However, of the 204 classified congested sections 5, 25 had branch points.

[0143] Fig. 17 is a diagram showing an example of a congested section 5 in a verification area B. The congested section 5 shown in Fig. 17 includes points P6, P7, and P8 that correspond to branch points.

[0144] If the azimuth angle of one adjacent congested section 3 at point P6 is 15.21 degrees and the azimuth angle of the other congested section 3 is 74.76 degrees, the azimuth angle difference of the congested sections 3 is 59.55 degrees. If the azimuth angle of one adjacent congested section 3 at point P7 is 62.28 degrees and the azimuth angle of the other congested section 3 is 6.64 degrees, the azimuth angle difference of the congested sections 3 is 55.64 degrees. If the azimuth angle of one adjacent congested section 3 at point P8 is 6.64 degrees and the azimuth angle of the other congested section 3 is 80.06 degrees, the azimuth angle difference of the congested sections 3 is 73.42 degrees.

[0145] Therefore, if the threshold value of the azimuth angle difference of the congested section 3 is set to, for example, 40 degrees, the congested section 5 shown in FIG.

[0146] The threshold value of the azimuth angle difference of the congested sections 3 may be set to an angle smaller than the smallest azimuth angle difference among the azimuth angle differences between the congested sections 3 that constitute different congested sections 5, for example.

[0147] By dividing the congested sections 5, the congested sections 5 in the verification area B were classified into 235 congested sections 5.

[0148] 18 is a diagram showing an example of the RMSE for each combination of judgment sections taking into consideration the bias in the extension tendency when predicting the extension length of congested section 5 while changing the usage rate of the learning data from 10% to 100% in 10% increments starting from the oldest. Note that the RMSE was 179.8048 when the extension length of congested section 5 was all set to "0."

[0149] According to the example of RMSE shown in Figure 18, when the proportion of learning data used to predict the extension length of congested section 5 is 10% of the total learning data, the prediction accuracy of the extension length is higher when the extension length of congested section 5 is predicted taking into account the bias in the extension trend than when the extension length of congested section 5 is predicted without taking into account the bias in the extension trend.

[0150] In the example of RMSE in verification area A shown in Figure 13, when the proportion of learning data used to predict the extension length of congested section 5 ranged from 10% to 40% of the total learning data counted from the oldest, the prediction accuracy of the extension length tended to be higher when predicting the extension length of congested section 5 taking into account the bias in the extension trend than when predicting the extension length of congested section 5 without taking the bias in the extension trend into account. This is a phenomenon resulting from the fact that the number of learning data collected in verification area B was greater than the number of learning data collected in verification area A.

[0151] In other words, the smaller the number of learning data used to predict the extension length of congested section 5, the higher the accuracy of predicting the extension length when taking into account the bias in the extension trend. This is the same trend as the verification results for verification area A, and the verification results for verification area B also show this trend.

[0152] While one embodiment of the information processing device 10 has been described above, the disclosed embodiment of the information processing device 10 is merely an example, and the embodiment of the information processing device 10 is not limited to the scope described in the present embodiment. Various modifications or improvements can be made to the present embodiment without departing from the gist of the present disclosure, and embodiments incorporating such modifications or improvements are also included in the technical scope of the disclosure. For example, the order of the prediction processes, including the congested section determination process shown in FIG. 10 and the extension length prediction process shown in FIG. 11, may be changed without departing from the gist of the present disclosure.

[0153] In addition, in the present disclosure, an embodiment in which the prediction process is realized by software has been described as an example. However, the process equivalent to the flowcharts shown in Figures 10 and 11 may be implemented in, for example, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a PLD (Programmable Logic Device) and processed by hardware. In this case, the processing speed can be increased compared to when the prediction process is realized by software.

[0154] In this way, the CPU 21 of the information processing device 10 may be replaced with a dedicated processor specialized for a specific process, such as an ASIC, FPGA, PLD, GPU (Graphics Processing Unit), or FPU (Floating Point Unit).

[0155] The prediction process may be implemented by one CPU 21, or may be implemented by a combination of two or more processors of the same type or different types, such as a plurality of CPUs 21 or a combination of a CPU 21 and an FPGA.

[0156] Furthermore, the prediction process may be realized by cooperation of processors located in physically separate locations connected via the Internet, for example.

[0157] Furthermore, in the present embodiment, an example has been described in which the information processing program is stored in the ROM 22 of the information processing device 10, but the storage destination of the information processing program is not limited to the ROM 22. The information processing program of the present disclosure can also be provided in a form recorded on a storage medium readable by the computer 20. For example, the information processing program may be provided in a form recorded on an optical disc such as a CD-ROM (Compact Disk Read Only Memory) or a DVD-ROM (Digital Versatile Disk Read Only Memory). Furthermore, the information processing program may be provided in a form recorded on a portable semiconductor memory such as a USB (Universal Serial Bus) memory or a memory card.

[0158] ROM 22, non-volatile memory 24, CD-ROM, DVD-ROM, USB, and memory cards are examples of non-transitory storage media.

[0159] Furthermore, the information processing device 10 may download an information processing program from an external device via the communication unit 27 and store the downloaded information processing program, for example, in the non-volatile memory 24. In this case, the information processing device 10 reads the information processing program downloaded from the external device and executes the prediction process.

[0160] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[0161] The following additional notes are provided regarding the above-described embodiment.

[0162] (Supplementary Item 1) A system including a memory and at least one processor connected to the memory, wherein the processor classifies each congestion section represented by learning data including a congestion start position and a congestion end position into a congestion direction represented by the start position and the end position, and for each congestion section classified into a congestion direction, recursively repeats a process of connecting adjacent congestion sections whose start position and end position are within a predetermined range as a continuous integrated congestion section until the start position or the end position of the adjacent congestion sections no longer falls within the predetermined range, thereby determining which congestion sections constitute the integrated congestion section, An information processing device configured to obtain the integrated congestion section for each integrated congestion section that represents the same congestion, classify the integrated congestion section for each integrated congestion section that represents the same congestion according to each judgment category that is divided by day of the week, by employment attribute that represents whether the congestion occurs on a weekday or a holiday, and by time of day, and to set the extension scale of the integrated congestion section for each integrated congestion section that represents the same congestion and for each judgment category.

[0163] (Supplementary Item 2) An information processing device comprising: a memory; and at least one processor connected to the memory, wherein the processor is configured to: determine, for each judgment category, whether there is a bias in the extension scale of a designated congested section, using the extension scale for each congestion section obtained from the extension lengths of consecutive identical congestion sections classified into judgment categories by day of the week, by employment attribute indicating whether the occurrence occurs on a weekday or a holiday, and by time of day; and, if there is a bias in the extension scale of the designated congested section within the judgment category, select the judgment category to be used for predicting the extension length of the congested section according to the combination of judgment categories in which the bias in extension scale exists; and predict the extension length of the designated congested section at a designated date and time using the extension length of the designated congested section included in the selected judgment category.

[0164] (Supplementary Item 3) A system including a memory and at least one processor connected to the memory, wherein the processor classifies each congestion section represented by learning data including a congestion start position and a congestion end position into a congestion direction represented by the start position and the end position, and for each congestion section classified into a congestion direction, recursively repeats a process of connecting adjacent congestion sections whose start position and end position are within a predetermined range as a continuous integrated congestion section until the start position or the end position of the adjacent congestion sections no longer falls within the predetermined range, thereby determining which congestion sections constitute the integrated congestion section, An information processing device configured to obtain the integrated congestion section for each integrated congestion section that represents the same congestion, classify the integrated congestion section for each integrated congestion section that represents the same congestion according to each judgment category that is divided by day of the week, by employment attribute that represents whether the congestion occurs on a weekday or a holiday, and by time of day, set an extension scale of the integrated congestion section for each integrated congestion section that represents the same congestion and for each judgment category, use the extension scale to determine whether there is a bias in the extension scale of the specified congestion section within the judgment category, and if there is a bias in the extension scale of the specified congestion section within the judgment category, select the judgment category to be used for predicting the extension length of the congestion section according to the combination of judgment categories in which there is a bias in extension scale, and predict the extension length of the specified congestion section at a specified date and time using the extension length of the specified congestion section included in the selected judgment category.

[0165] (Addendum 4) An information processing program for causing a computer to execute the following process: classifying each congestion section represented by learning data including the start position and the end position of the congestion by the congestion direction represented by the start position and the end position; for each congestion section classified by congestion direction, connecting adjacent congestion sections whose start position and end position are within a predetermined range as a continuous integrated congestion section, recursively repeating this process until there are no start positions or end positions of the adjacent congestion sections within the predetermined range, thereby determining which congestion sections constitute the integrated congestion section; obtaining the integrated congestion section for each integrated congestion section that represents the same congestion; classifying the integrated congestion section for each integrated congestion section that represents the same congestion according to each judgment category that is divided by day of the week, by employment attribute that represents whether the congestion occurs on a weekday or a holiday, and by time of day; and setting the extension scale of the integrated congestion section for each integrated congestion section that represents the same congestion and for each judgment category.

[0166] (Appendix 5) An information processing program for causing a computer to execute a process of: determining, for each judgment category, whether there is a bias in the extension scale of a designated congested section, using the extension lengths of consecutive identical congested sections, which are classified into judgment categories according to the day of the week, the employment attribute indicating whether the occurrence occurs on a weekday or a holiday, and the judgment category according to the time of day, and the extension scale of each judgment category; if there is a bias in the extension scale of the designated congested section within the judgment category, selecting the judgment category to be used for predicting the extension length of the congested section according to the combination of judgment categories in which the bias in extension scale exists; and predicting the extension length of the designated congested section at a designated date and time, using the extension length of the designated congested section included in the selected judgment category.

[0167] (Supplementary Item 6) A non-transitory storage medium storing a program executable by a computer to execute a prediction process, wherein the prediction process comprises: a classification step of classifying each congestion section represented by learning data including a congestion start position and a congestion end position into a congestion direction represented by the start position and the end position; and a determination step of determining which congestion sections constitute the integrated congestion section by recursively repeating a process of connecting adjacent congestion sections whose start position and end position are within a predetermined range for each congestion section classified into each congestion direction as a continuous integrated congestion section until the start position or the end position of the adjacent congestion sections no longer falls within the predetermined range. a setting step of classifying each integrated congestion section representing the same congestion according to each judgment category which is divided by day of the week, by employment attribute which indicates whether the congestion occurs on a weekday or a holiday, and by time of day, and setting an extension scale of the integrated congestion section for each integrated congestion section representing the same congestion and for each judgment category; a selection step of using the extension scale to determine whether there is a bias within the judgment category in the extension scale of a specified congestion section, and if there is a bias within the judgment category in the extension scale of the specified congestion section, selecting the judgment category to be used for predicting the extension length of the congestion section according to the combination of judgment categories in which there is a bias in extension scale; and a prediction step of predicting the extension length of the specified congestion section at a specified date and time using the extension length of the specified congestion section included in the selected judgment category.

Claims

1. A determination unit that determines which congested sections constitute the integrated congestion section by repeating a process of linking together intermittently occurring congestion sections whose starting and ending points are within a predetermined range for each congestion section classified by congestion direction until the starting or ending point of the congestion section is no longer within the predetermined range. An information processing device comprising:

2. a division unit that, when a difference in azimuth angle between congestion sections branching off from a junction point of a continuous integrated congestion section composed of a plurality of congestion sections is equal to or greater than a predetermined angle, divides the integrated congestion section into two different integrated congestion sections at the junction point between the congestion sections where the difference in azimuth angle is equal to or greater than the predetermined angle. An information processing device comprising:

3. a selection unit which uses the extension scale of each congestion section obtained from the extension lengths of consecutive identical congestion sections classified into each judgment division according to the day of the week, the working attribute indicating whether the occurrence time is a weekday or a holiday, and the time of day, and the extension scale of each of the judgment divisions to determine whether there is a bias in the extension scale of a designated congestion section for each of the judgment divisions, and if there is a bias in the extension scale of the designated congestion section within the judgment division, selects the judgment division to be used for predicting the extension length of the congestion section according to the combination of the judgment divisions in which the bias in the extension scale exists; a prediction unit that predicts an extension length of the designated congestion section at a designated date and time using an extension length of the designated congestion section included in the judgment section selected by the selection unit; An information processing device comprising:

4. The prediction unit predicts the extension length of the specified congestion section at the specified date and time by using the extension length of the specified congestion section included in the combination of the judgment section corresponding to the specified date and time among all the combinations of the judgment section in which a bias exists in the extension scale of the specified congestion section. The information processing device according to claim 3 .

5. An information processing program for causing a computer to function as each part of the information processing device according to any one of claims 1 to 4, 7, and 8.

6. An information processing method in an information processing device including a classification unit, a determination unit, a setting unit, a selection unit, and a prediction unit, a classification step in which the classification unit classifies each congestion section represented by learning data including a start point position and an end point position of a congestion into a congestion direction represented by the start point position and the end point position; a determination step in which the determination unit recursively repeats a process of connecting, for each congestion section classified by congestion direction, congestion sections whose start point position and end point position are within a predetermined range as a continuous integrated congestion section until the start point position or end point position of the congestion section is no longer within the predetermined range, thereby determining which congestion sections constitute the integrated congestion section; A setting step in which the setting unit classifies each of the integrated congestion sections representing the same congestion according to each judgment category that is divided into each day of the week, each work attribute that represents whether the congestion occurs on a weekday or a holiday, and each time zone, and sets an extension scale of the integrated congestion section for each integrated congestion section representing the same congestion and for each judgment category; a selection step in which the selection unit uses the extension scale to determine whether or not there is a bias in the extension scale of the designated congestion section within the judgment section, and when there is a bias in the extension scale of the designated congestion section within the judgment section, selects the judgment section to be used for predicting the extension length of the congestion section according to a combination of the judgment sections in which the bias in extension scale exists; a prediction step in which the prediction unit predicts an extension length of the designated congestion section at a designated date and time using an extension length of the designated congestion section included in the selected judgment section; An information processing method comprising:

7. The determination unit determines the congestion sections that constitute the integrated congestion section without limiting the position of the start point of the integrated congestion section to a specific position on a road. The information processing device according to claim 1 .

8. The dividing unit divides the integrated congestion section, in which a difference in azimuth angle between congestion sections branching off from a junction point of the integrated congestion section is equal to or greater than a predetermined angle, into a first integrated congestion section passing through the junction point of the integrated congestion section and a second integrated congestion section having the junction point of the integrated congestion section as a starting point. The information processing device according to claim 2 .