JP2011118745A - Traffic information generation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a traffic information generation device which performs complementation with high accuracy by utilizing features of a plurality of complementing methods without depending on a current state. <P>SOLUTION: The traffic information generation device includes a current state information DB 102 storing current traffic information based on current probe car information, and a past information DB 101 storing current traffic information and storing past traffic information. One or both of the past traffic information recorded in the past information DB 101 and the current traffic information recorded in the current DB 102 are used to generate a plurality of kinds of complementing information for complementing traffic information of roads, which is missing due to missing of probe car information, in the current traffic information. Precision information of the plurality of kinds of generated complementing information is generated on the basis of the past traffic information recorded in the past information DB 101 and the plurality of kinds of generated complementing information, and complementing information for missing traffic information of roads is determined on the basis of the precision information and the plurality of kinds of generated complementing information. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a traffic information generating apparatus that generates traffic information for a road for which traffic information has not been collected by a probe car.

  In contrast to a system that collects traffic information using road sensors, a probe car system that collects a wide range of traffic information with a probe car at a lower cost is known. However, since the travel position and timing of the probe car are random, there is a spatial and temporal deficit in the collected traffic information.

  Therefore, for missing traffic information, there is a method of estimation complementation that calculates the correlation of traffic information between road links from past probe traffic information, and complements it from information of other road links based on such correlation (Patent Literature). 1). However, in the method of estimation complementation, when there is a lot of current state information of other links, it can be complemented with high accuracy. However, when the current state information is sparse, there are cases where the accuracy cannot be sufficiently guaranteed.

JP 2006-25194 A

  Therefore, there is a need for a complementing method that complements traffic information accurately and optimally in various situations for roads lacking traffic information.

  The invention of claim 1 is applied to a traffic information generating device for generating traffic information of each road based on probe car information, and records current traffic information database means for recording current traffic information based on current probe car information; At least one of past information database means for successively recording current traffic information and past traffic information, past traffic information recorded in past information database means, and current traffic information recorded in current information database means Is used to record supplementary information for creating multiple types of supplementary information for supplementing traffic information on roads that are missing due to missing probe car information in the current traffic information, and past information database means. Based on the past traffic information and the multiple types of supplementary information created by the supplementary information creation means, the supplementary information creation means Accuracy information generating means for generating accuracy information of a plurality of types of complementary information, accuracy information of a plurality of types of complementary information generated by the accuracy information generating means, and a plurality of types of complementary information generated by the complementary information generating means Based on the current traffic information, supplemental information determining means for determining supplemental information of traffic information on a road that is missing due to missing probe car information is provided.

  According to the present invention, it is possible to perform high-precision complementing that makes use of the features of a plurality of complementing methods without depending on current status information.

1 is a diagram showing a traffic information providing system in which a traffic information providing center 13 collects traffic information from a probe car 11 and provides traffic information to a traffic information user (automobile) 14. It is a figure which shows the structure of the traffic information generation apparatus. It is a figure which shows the data format of present condition information DB102 and past information DB101. It is a figure which shows the example of precision information DB110. It is a figure which shows the processing flow until it produces | generates complementary information.

  In the following, a traffic information generating device that generates highly accurate optimum supplemental information by dynamically selecting a plurality of supplementary processes for road traffic information for which current probe traffic information has not been collected will be described.

  FIG. 1 shows a traffic information providing center 13 having a traffic information generating device 12 according to an embodiment of the present invention, collecting traffic information from a probe car 11 and providing traffic information to a traffic information user (automobile) 14. It is a figure which shows the traffic information provision system. Data communication between the traffic information providing center 13 and the probe car 11 or the traffic information user 14 is performed wirelessly using, for example, a mobile phone.

  The probe car 11 is a vehicle contracted to provide the current location information of the vehicle to the traffic information providing center 13 periodically (for example, every 5 minutes). The probe car 11 is equipped with a car navigation device and uses a GPS sensor (GPS receiver) that detects a GPS signal from a GPS satellite, a direction sensor that detects a traveling direction of a vehicle, a vehicle speed sensor that detects a vehicle speed, and the like. Thus, it has a function of detecting the current location (vehicle position) of the vehicle, and periodically transmits the detected current location information (also referred to as probe information) of the vehicle to the traffic information providing center 13. At this time, the current position information of the vehicle is detected and transmitted together with vehicle identification information for identifying the vehicle. In addition, as the present location information of a vehicle, it is good also as only the positional information detected by the GPS signal.

  The traffic information providing center 13 has a server (computer system) 12, and the server 12 functions as a traffic information generating device described below. Hereinafter, the server 12 is referred to as a traffic information generation device 12. The traffic information generation device 12 generates traffic information based on the information transmitted from the probe car 11 and further supplements the missing traffic information, and then provides the traffic information to the traffic information user 14. The traffic information user 14 is a vehicle that uses traffic information, is equipped with a car navigation device, and uses the provided traffic information for navigation such as route guidance.

  FIG. 2 is a diagram illustrating a configuration of the traffic information generation device 12. The traffic information generation device 12 includes a processing unit that performs various types of processing and a database that stores various types of information (hereinafter simply referred to as DB). The current status information generation processing unit 100 generates current traffic information based on the probe information transmitted from the probe car 11. The traffic information referred to here is, for example, travel time, traffic jam information, and the like. Hereinafter, in this embodiment, the travel time will be described as an example.

  In the road map data, road intersections or specific points on the road are defined as nodes, and roads connecting the nodes are defined as links. The travel time as the traffic information in the present embodiment is the time required for the vehicle to pass through the road, that is, the link. The travel time can be obtained by calculation from vehicle position information transmitted from the probe car 11 every predetermined time (for example, 5 minutes).

  The current status information generation processing unit 100 generates current probe traffic information for each link based on the probe information (probe data) acquired from the probe car 11 and stores it in the current status information DB 102 and also in the past information DB 101. . In the current status information DB 102, the latest traffic information from about 30 minutes ago is accumulated. The past information DB 101 is a past information database in which past probe traffic information is recorded. In the past information DB 101, traffic information from the present to several years ago is accumulated.

  FIG. 3 is a diagram showing the data formats of the current status information DB 102 and the past information DB 101. The current information DB 102 and the past information DB 101 have the same data format. In the horizontal direction, links are arranged in ascending order of numbers, and in the vertical direction, information acquisition dates and times acquired in units of 5 minutes are arranged. The unit to be processed in this embodiment is a 10 km square quadratic mesh unit, and all the links included in the secondary mesh are described in the table of FIG.

  The current traffic information generated by the current status information generation processing unit 100 corresponds to one column in FIG. As described above, since the latest traffic information from the present to about 30 minutes before is accumulated in the current status information DB 102, data for seven columns is stored. As described above, the past information DB 101 stores traffic information from the present to several years ago. As can be seen from the table of FIG. 3, since the current traffic information (travel time in the present embodiment) is generated only from the information from the probe car 11 traveling at that time, the probe car 11 is traveling. You cannot get traffic information with no links. In FIG. 3, “None” is described in the link for which traffic information could not be obtained.

  Based on the vehicle position information transmitted from the probe car 11 every predetermined time (for example, 5 minutes) and the road map data stored in a database (not shown), the current status information generation processing unit 100 runs the probe car 11. A road (link) is specified, a travel distance is calculated, and a time taken for traveling on the road (link), that is, a travel time is obtained.

  The estimation / complementary processing unit 103 uses the past traffic information stored in the past information DB 101 to calculate and obtain correlation information between links by a principal component analysis method. Based on the past correlation information obtained in this way and the current traffic information stored in the current status information DB 102, that is, the latest traffic information, the estimated complement processing unit 103 determines the traffic information of the missing link. Is estimated to generate estimated complementary information, and the generated estimated complementary information is stored (recorded) in the estimated complementary information DB 103. That is, the traffic information of the missing link is estimated and complemented from the information of the other links not missing and the correlation information between the links.

  It is assumed that the estimated complementary information of the present embodiment is obtained using the correlation information obtained without distinction of day of the week or time zone. However, the correlation information may be obtained in consideration of the day of the week and the time zone, and the estimated supplement information may be obtained from the correlation information. In addition, the method for obtaining the estimated supplement information as described above is publicly known content, and details are referred to Patent Document 1. The description of the method for obtaining the estimated supplement information in Patent Document 1 constitutes a part of the present specification.

  The statistical processing unit 105 statistically processes past traffic information stored in the past information DB 101, and stores (records) statistical information generated by the statistical processing in the statistical information DB. For example, all the traffic information of link number 1 at about 10:00 am on the past Sunday is averaged, and the statistical information of link number 1 at about 10:00 am on Sunday is stored in the statistical information DB 106 as statistical traffic information. The statistical traffic information stored in the statistical information DB 106 is traffic information that is simply specified from the day of the week or time, regardless of the state of the current traffic information.

  When there is traffic information in the immediate vicinity of the same road, the front supplement processing unit 107 generates supplement information based on the information, and stores (records) the generated supplement information in the front supplement information DB. The pre-complementation processing unit 107 stores traffic information of a link existing within a predetermined time (for example, 10 minutes) from the current status information DB 102 in the pre-complementation information DB 108 as traffic information of the link. The value of the predetermined time is derived from an empirical rule, experiment, simulation, or the like as a range where the correlation can be obtained.

  Next, accuracy information which is a feature of the present invention will be described. The accuracy information generation processing unit 109 periodically calculates the past performance of the accuracy of the complementary information, and stores the calculated accuracy information of each complementary processing in the accuracy information DB 110 in advance. That is, the accuracy information generation processing unit 109 calculates an error rate between the past traffic information accumulation information and the supplemented information, and stores it in the accuracy information DB 110 in advance.

  FIG. 4 is a diagram illustrating an example of the accuracy information DB 110. The accuracy information DB 110 holds accuracy information for each of the estimated complement information, statistical complement information, and prefix complement information. Furthermore, accuracy information (error rate) is held for each time zone, day type, and link. The day type can be classified into weekday first day (Monday), weekday middle day (Tuesday, Wednesday, Thursday), weekday all day (Friday), Saturday, Sunday, GW, Bon Festival, etc. The time zone is a time zone unit for acquiring probe information. In addition to the above, classification with the weather taken into account is also conceivable. The accuracy information generation processing unit 109 calculates the error rate of the complementary information according to the day type and time from the past traffic information accumulation information, and stores (records) it in the accuracy information DB 110. The accuracy information generation processing unit 109 calculates the error rate as follows.

誤差率は、日種及び時間帯、地域により値が変化するため、時間帯、日種、リンクごとに算出する。統計補完情報との差分を求める実際の交通情報の所定期間は、例えば、ある年度の所定期間としたり、過去１年分としたり、統計補完情報を求めるために使用した実際の交通情報の期間としたり、適宜決定する。 For example, statistical supplement information will be described. The statistical supplement information is obtained by calculating the average travel time of each link for each day of the week and each time zone. The difference between the statistical supplement information obtained in this way and the actual traffic information for a predetermined period stored in the past information DB 101 is obtained to calculate the error rate. For example, the error rate err_h (t) when the day type is Sunday is the statistical supplement information at time t on Sunday as i (t), and the actual past Sunday traffic information stored in the past information DB 101 is f (t , N) (n: 0 to N), the following formula is used. It can be said that the smaller the error rate err_h (t), the smaller the error.

Since the error rate varies depending on the day type, time zone, and region, the error rate is calculated for each time zone, day type, and link. The predetermined period of the actual traffic information for which the difference from the statistical supplementary information is obtained is, for example, the predetermined period of a certain year, the past one year, or the period of the actual traffic information used for obtaining the statistical supplementary information. Or as appropriate.

  Next, estimation complement information is demonstrated. The estimated supplement information of this embodiment is obtained without distinguishing day of the week or time zone. The difference between the estimated supplement information thus obtained and the actual traffic information for a predetermined period stored in the past information DB 101 is obtained, and an error rate, which is accuracy information, is calculated. As described above, the estimated complementary information obtained without distinguishing the day of the week and the time zone is also obtained as the accuracy information by distinguishing the time zone and the day type.

  Specifically, for example, a case where traffic information for the past two months is stored in the past information DB 101 will be described. First, the accuracy information generation processing unit 109 calculates and obtains correlation information between links by the above-described principal component analysis method based on the traffic information for the first month. Based on the correlation information for the first month (first predetermined period) obtained in this way and the traffic information for the next month (second predetermined period), Estimate the traffic information of the missing link and generate estimated supplementary information.

  At this time, in the traffic information for the next month, the error rate is calculated for each case when the traffic information is sparse and when it is not sparse. When the traffic information is sparse, the percentage of traffic information based on the probe information (coverage rate) with respect to the total number of links in a predetermined range (for example, one secondary mesh) is 5% or less, for example. If it is lean, it is assumed that it is not lean if it is 5% better. You may change 5% of a threshold value suitably.

  In the estimated supplementary information obtained when the traffic information for the next month is sparse, the accuracy information is obtained by obtaining a difference from the actual traffic information for a predetermined period (for example, the next month) stored in the past information DB 101. The error rate is calculated. In the estimated supplementary information obtained when the traffic information for the next month is not sparse, the accuracy information is obtained by obtaining a difference from the actual traffic information for a predetermined period (for example, the next month) stored in the past information DB 101. The error rate is calculated. In this way, accuracy information is obtained separately for the lean case and the non-lean case. In this case, the accuracy information calculation result is naturally small in accuracy (large error rate) when it is sparse, and is naturally high in accuracy (small error rate) when it is sparse. In addition, since the specific arithmetic expression etc. of accuracy information are the same as that of the statistical complement information mentioned above, the description is abbreviate | omitted.

  Next, the prefix complement information will be described. The prefix supplement information of the present embodiment is the prefix supplement information of the link when there is traffic information of the corresponding link within a predetermined time (for example, 10 minutes) from the current time in the current status information DB 102. It is what. The prefix supplement information is not obtained by distinguishing day of the week or time zone. However, the prefix information obtained without distinguishing the day of the week and the time zone as described above is also obtained as the accuracy information by distinguishing the time zone, the day type, etc., as in the estimated complement information. Since the specific method of obtaining the accuracy information of the prefix complement information is the same as that of the statistical complement information described above, description thereof is omitted.

The weighting calculation processing unit 111 calculates weighting for generating complementary information based on the accuracy information stored in the accuracy information DB 110 and each complementary information. Assume that the values of estimation completion, statistical information, and prefix completion are i1 (t), i2 (t), and i3 (t), respectively, and the weighting coefficients are α (t), β (t), and γ (t). The estimated information P (t) calculated by the arithmetic device is calculated by the following equation.
P (t) = α (t) × i1 (t) + β (t) × i2 (t) + γ (t) × i3 (t)

  The weighting coefficient is determined by referring to the accuracy information DB 110 by using the link number to be supplemented with the traffic information and the current day of the week as a key, and giving more importance to weighting information with high accuracy. For weighting at this time, the most accurate weighting coefficient is set to “1”, and other weights are set to “0”. That is, the supplement information with the highest accuracy is selected and adopted from the estimated complement information, the statistical complement information, and the prefix complement information.

  In the case of estimated complement information, the reference database for estimated complement information is different depending on whether the current status information is sparse or not.

The weighting is not a binary value of “1” and “0”, but may be a multi-value or a continuous value in order to stabilize the estimated information calculated. For example, the weighting value may be determined according to the reciprocal of the error rate stored in the accuracy information DB. When the error rates of i1 (t), i2 (t), and i3 (t) are err1 (t), err2 (t), and err3 (t), respectively, α (t) is calculated by the following equation.

  β (t) and γ (t) may also be obtained by substituting each error rate into the error rate of the numerator in the above formula.

Note that there are cases where complementary information cannot be obtained during the weighting calculation. In particular, in the pre-complementation, when the number of traveling probe cars is small, the supplement information cannot be obtained in most cases. For example, when i1 (t) cannot be obtained, α (t) = 0, and β (t) is calculated by the following equation. Similarly, γ (t) is obtained.

  When neither i1 (t) nor i2 (t) is obtained, α (t) = 0, β (t) = 0, and γ (t) = 1. If none of i1 (t), i2 (t), and i3 (t) is obtained, complementation is impossible.

  The traffic information distribution processing unit 112 distributes the current state information and the complementary information obtained from the weighting calculation device. If both current status information and estimated information exist in the same link and the same time zone, the current status information is preferentially distributed.

  FIG. 5 is a diagram showing a processing flow until complementary information is generated. The process of FIG. 5 is performed when the server 12 which is a traffic information generation apparatus executes a predetermined program. Elements corresponding to those in FIG. 2 are given the same reference numerals. Each processing unit described in FIGS. 2 and 5 is realized by the server 12 executing a predetermined program corresponding to each processing unit. Each processing unit may be an independent computer system (processing device).

  First, the current status information generation processing unit 100 is activated at a fixed cycle (for example, 5 minutes) in synchronization with the timing of acquiring probe information from the probe car 11. The current status information generation processing unit 100 generates current traffic information based on the received probe information and stores it in the past information DB 101 and the current status information DB 102.

  When the current traffic information is stored in the past information DB 101 and the current status information DB 102, the estimated complement processing unit 103 is activated. The estimated complement processing unit 103 refers to the latest current status information in the past information DB 101 and the current status information DB 102, performs the above-described estimated complementary processing, and outputs the estimated complementary information that is the estimated complementary processing result to the estimated complementary information DB 104.

  When the estimated complement information is stored in the estimated complement information DB 104, the statistical information processing unit 106 is activated. The statistical information processing unit refers to the past information DB 101, generates statistical information as described above, and outputs the statistical information to the statistical information DB 106.

  When the statistical information is stored in the statistical information DB 106, the prefix complement processing unit 107 is activated. The pre-complementation processing unit 107 generates pre-complementation information by the pre-complementation process and stores it in the pre-storage information DB 108. Here, the estimation complement processing unit, the statistical information processing unit, and the prefix complement processing unit can be activated in parallel.

  When the above processing is completed, the weighting calculation processing unit 111 is activated. The weighting calculation processing unit 111 refers to the estimated complementary information DB 104, the statistical information DB 106, the prefix complementary information DB 108, and the accuracy information DB, generates optimal complementary information, and outputs the result to the traffic information distribution processing unit 112. At this time, when referring to the accuracy information DB, if the current traffic information is sparse, the above-described accuracy information data of the estimated supplementary information when the current traffic information is sparse, and if the current traffic information is not sparse, the above-mentioned current traffic information The accuracy information data of the estimated supplement information when is not sparse shall be referred to.

  The traffic information distribution processing unit 112 reads the supplementary information output from the weighting calculation processing unit 111 and the latest current status information from the current status information DB 102, integrates them, and outputs them as traffic information to the traffic information user 14.

  Note that the accuracy information generation processing unit 109 periodically updates the accuracy information DB 110. By updating, it is possible to prevent deterioration of the accuracy of complementation. For example, even if the accuracy of statistical information is high in the past, new roads will be created and traffic congestion will be less than before, or on holidays due to the ETC percentage of expressways, traffic congestion will be large and statistical information accuracy will be poor The case occurs. At that time, by updating the accuracy information DB 110, it is possible to preferentially adopt the estimated complement information and the prefix complement information over the statistical information whose accuracy has deteriorated, and to generate highly accurate complement information. .

According to the present embodiment described above, the following operational effects can be obtained.
(1) The traffic information generation device 12 that generates traffic information of each road based on the probe car information sequentially records the current traffic information and the current traffic information DB 102 that records the current traffic information based on the current probe car information. Using the past information DB 101 that records and records past traffic information, and the past traffic information recorded in the past information DB 101 and / or the current traffic information recorded in the current status information DB 102, current traffic Complementary information creation processing unit (estimated supplementary processing unit 103, statistical processing unit 105, previous type) that creates a plurality of types of supplementary information for supplementing traffic information of roads (links) that are missing due to missing probe car information in the information Complementary processing unit 107), and past traffic information recorded in past information DB 101 and a plurality of types of supplements created by the complementary information creation processing unit Accuracy information generation processing unit 109 that generates accuracy information of a plurality of types of complementary information created by the complementary information creation processing unit based on the information, and accuracy information of the plurality of types of complementary information generated by the accuracy information generation processing unit 109 Based on the multiple types of supplementary information created by the supplementary information creation processing unit, weighting is performed as a supplementary information determination unit that determines supplementary information for traffic information on missing roads due to missing probe car information in the current traffic information An arithmetic processing unit 111 is provided.

  As a result, it is possible to generate highly accurate complementary information by dynamically selecting a plurality of complementary processes for road traffic information for which current probe traffic information has not been collected. In the traffic information service, it can be used to provide estimated traffic information on roads (links) for which probe traffic information has not been collected. By using such a traffic information generation device 12, it is possible to provide highly accurate estimated traffic information without depending on the defect rate of probe traffic information.

(2) The plurality of types of supplementary information created by the supplementary information creation processing unit is the correlation information between the roads obtained by the principal component analysis method using the past traffic information recorded in the past information DB 101 and the current situation information DB 102. Is generated by statistical processing by the statistical processing unit 105 using the estimated complementary information obtained by the estimated complementary processing unit 103 based on the current traffic information recorded in the past and the past traffic information recorded in the past information DB 101. The pre-complementation information generated by the pre-complementation processing unit 107 using the statistical supplement information that has been performed and the traffic information within a predetermined period from the current time recorded in the past information DB 101 is used. This makes it possible to select appropriate complementary information according to the situation.

(3) The accuracy information generation processing unit 109 compares the plurality of types of complementary information created by the complementary information creation processing unit with the past traffic information recorded in the past information DB 101 on the corresponding road (link) ( Based on (difference), the accuracy information of each of the plural types of complementary information is generated. Thereby, appropriate accuracy information can be easily generated.

(4) As the complementary information determination unit, the weighting calculation processing unit 111 includes the traffic information of the road that lacks the complementary information indicating that the accuracy information has the highest accuracy among the plurality of types of complementary information in the current traffic information. It was decided as supplementary information. Thereby, highly accurate complementary information can be generated by an easy method.

(5) In addition, the weighting calculation processing unit 111 serving as the supplementary information determining unit performs the current traffic by calculating the weighting according to the accuracy information of the plurality of types of supplemental information on the plurality of types of supplemental information. It may be determined as supplementary information for traffic information of roads that are missing in the information. Thereby, the estimated information calculated can be stabilized.

(6) The accuracy information is obtained for each of a plurality of types of complementary information in a road and a predetermined time condition unit.

(7) Further, the accuracy information is obtained for each of a plurality of types of complementary information in units of roads, day types, and time zones. Thereby, in each road (link), precision information can be set finely for every day type and time zone, and appropriate complementary information can be generated for each day type and time zone.

(8) The accuracy information can be said to indicate the degree of variation between each type of complementary information created by the complementary information creation processing unit and past traffic information recorded in the past information DB 101 on the corresponding road. . The smaller the degree of variation, the higher the accuracy.

(9) The accuracy information generation processing unit 109 estimates the estimated supplementary information obtained on the basis of the correlation information and the current traffic information according to whether the current traffic information is sparse or not. Generates accuracy information of supplementary information, and weighting calculation processing unit 111 generates a case-by-case classification based on whether the current traffic information is sparse or not, depending on whether the current traffic information is sparse or not. The accuracy information of the estimated supplementary information was used. Thereby, even if the current traffic information is sparse, highly accurate supplementary information can be provided.

-Modification-
(1) In the above-described embodiment, the case where the probe information from the probe car 11 is lost has been described. However, in addition to the probe information from the probe car 11, information from the road sensor may be considered. That is, when no information is collected from either the probe car 11 or the road sensor, the complementary information described above may be obtained. When information from the road sensor is also taken into consideration, traffic information such as travel time may be calculated from information such as the number of vehicles passing in a certain time transmitted from the road sensor at a specific position.

(2) In the above-described embodiment, an example of travel time has been described as traffic information. However, traffic information need not be limited to travel time. Various information based on information transmitted from the probe car 11 may be used as traffic information. For example, traffic jam information may be traffic information. In the case of traffic jam information, a distinction such as traffic jam, congestion, and smoothness is made. And the traffic jam information of the road (link) that could not be acquired because the probe information of the probe car 11 is lost may be complemented by the same method as described above.

(3) In the above-described embodiment, the description has been made by taking the three types of the supplementary information, that is, the estimated supplemental information, the statistical information, and the prefix supplemental information as an example. However, it is not necessarily limited to three types. Any two of these three types may be used. Further, there may be three or more types by adding supplemental information by other methods.

(4) In the above-described embodiment, an example has been described in which correlation information between links is calculated by a principal component analysis method using past traffic information stored in the past information DB 101 when generating estimated supplementary information. . However, it is not necessarily limited to the principal component analysis method. In addition, statistical methods such as independent component analysis and factor analysis may be applied.

  Note that the present invention is not limited to the configurations in the above-described embodiments as long as the characteristic functions of the present invention are not impaired. Moreover, it is good also as a structure which combined the above-mentioned embodiment and a some modification.

100. . . Current status information generation processing unit 101. . . Past information database, 102. . . Current status information database, 103. . . Estimated complement processing unit, 104. . . Estimated supplementary information database, 105. . . Statistical processing unit, 106. . . Statistical information database, 107. . . Pre-complementation processing unit, 108. . . Prefix completion information database, 109. . . Accuracy information generation processing unit, 110. . . Accuracy information database, 111. . . Weighting calculation processing unit, 112. . . Traffic information distribution processing department

Claims (9)

A traffic information generating device that generates traffic information of each road based on probe car information,
A current information database means for recording current traffic information based on current probe car information;
Past information database means for successively recording the current traffic information and recording past traffic information;
Probe car information is missing in the current traffic information using at least one of the past traffic information recorded in the past information database means and the current traffic information recorded in the current status information database means. Supplementary information creating means for creating a plurality of types of supplementary information for supplementing traffic information of roads that are lost due to
Based on the past traffic information recorded in the past information database unit and the plurality of types of complementary information created by the complementary information creation unit, the accuracy of the plurality of types of complementary information created by the complementary information creation unit Accuracy information generating means for generating information;
Probe car information is missing in the current traffic information based on the accuracy information of the plurality of types of complementary information generated by the accuracy information generating unit and the plurality of types of complementary information generated by the complementary information generating unit. A traffic information generating apparatus comprising: complementary information determining means for determining complementary information of traffic information of a road that is missing due to the above. In the traffic information generation device according to claim 1,
The plurality of types of supplementary information created by the supplementary information creating means includes correlation information between roads obtained by principal component analysis using the past traffic information recorded in the past information database means and the current status information. Complementary information obtained based on the current traffic information recorded in the database means, complementary information generated by statistical processing using past traffic information recorded in the past information database means, and the past A traffic information generating apparatus comprising: complementary information generated by using traffic information within a predetermined period from the current time recorded in the information database means. In the traffic information generating device according to claim 1 or 2,
The accuracy information generating means is based on a difference between the plurality of types of complementary information created by the complementary information creating means and the past traffic information recorded in the past information database means on the corresponding road. A traffic information generating apparatus characterized by generating accuracy information of a plurality of types of complementary information. In the traffic information generating device according to any one of claims 1 to 3,
The supplementary information determining means determines supplemental information indicating that the accuracy information has the highest accuracy among the plurality of types of supplementary information as supplemental information of road traffic information that is missing in the current traffic information. A traffic information generation device characterized by the above. In the traffic information generating device according to any one of claims 1 to 3,
The supplementary information determining means is a road traffic that is missing in the current traffic information by performing an operation on the plurality of types of supplemental information by weighting the plurality of types of supplemental information according to the accuracy information. A traffic information generating apparatus characterized in that the information is determined as complementary information. In the traffic information generating device according to any one of claims 1 to 5,
The said accuracy information is calculated | required for a road and a predetermined time condition unit for every said several types of complementary information, The traffic information generation apparatus characterized by the above-mentioned. In the traffic information generating device according to any one of claims 1 to 5,
The accuracy information is obtained for each of the plurality of types of complementary information in units of roads, day types, and time zones. In the traffic information generation device according to any one of claims 1 to 7,
The accuracy information indicates a degree of variation between each of the plurality of types of supplementary information created by the supplementary information creating unit and the past traffic information recorded in the past information database unit on the corresponding road,
A traffic information generating apparatus characterized in that the smaller the degree of variation, the higher the accuracy. In the traffic information generation device according to claim 2,
The accuracy information generating means is configured to divide a case into a case where the current traffic information is sparse and a case where the current traffic information is not sparse in the estimated complementary information which is complementary information obtained based on the correlation information and the current traffic information. And generating accuracy information of the estimated complementary information,
The supplementary information determining means is configured to determine the estimated supplementary information generated by dividing the current traffic information into a case where the current traffic information is sparse and a case where the traffic information is not sparse according to whether the current traffic information is sparse or not. A traffic information generating apparatus using accuracy information.

Images