JP3545598B2

JP3545598B2 - Traffic information prediction method and apparatus, and recording medium recording traffic information prediction program

Info

Publication number: JP3545598B2
Application number: JP12237098A
Authority: JP
Inventors: 文夫安達; 智章小川; 仁士毛利; 力堀越; 英朋境野; 智鈴木
Original assignee: Nippon Telegraph and Telephone Corp
Current assignee: Nippon Telegraph and Telephone Corp
Priority date: 1998-05-01
Filing date: 1998-05-01
Publication date: 2004-07-21
Anticipated expiration: 2018-05-01
Also published as: JPH11316890A

Description

【０００１】
【発明の属する技術分野】
本発明は交通情報予測方法および装置に関する。
【０００２】
【従来の技術】
従来の交通情報予測方法として、予測するリンクの過去の交通情報より、将来の交通情報を予測する方法があった。具体的な予測方法には、自己回帰法やニューラルネットワークが適用される。しかし、この方法では、予測対象のリンクの情報だけを使用しているため、予測精度が低いという問題があった。ここで、交通情報として、速度、旅行時間、渋滞度などが考えられる。
【０００３】
これを解決する従来の交通情報予測方法を図６に示す。交通情報を予測するリンクに接続された隣接リンクの情報を利用する。すなわち、交通情報を予測するリンクの過去の交通情報と、交通情報を予測するリンクに接続された隣接リンクの過去の交通情報から、交通情報を予測するリンクの将来の交通情報を予測する。具体的な予測方法には、自己回帰法やニューラルネットワークが適用される。しかし、この方法では、複雑に変化する交通情報を予測するには、予測精度が低いという問題があった。
【０００４】
さらにこれを解決するために、図７に示す方法がある。この方法では予測の対象とする予測対象リンクを含む領域内のリンクの交通情報を用いて、予測対象リンクの交通情報の予測を行う。この方法によれば、広い範囲のリンクの交通情報を利用することができるので、複雑に変化する交通情報を精度よく予測できる。
【０００５】
【発明が解決しようとする課題】
しかしながら、この方法では、領域に含まれるリンクの数が多くなると、交通情報の蓄積のための記憶装置の容量が増加するとともに、予測に使用する交通情報の数が膨大となり、予測部を装置で実現した場合は、その構成が複雑になるという問題があり、また予測部をプログラム処理で実現する場合は、予測処理に時間がかかるという問題があった。
【０００６】
本発明の目的は、複雑に変化する交通情報を精度よく、かつ効率的に予測する交通情報予測方法および装置を提供することにある。
【０００７】
【課題を解決するための手段】
本発明の第１の交通情報予測方法は、
選択手段が、複数のリンクの交通情報を入力し、選択することにより、予測対象リンクの交通情報と、該予測対象リンクを含むあらかじめ定めた領域内のリンクの交通情報とを出力する選択手順と、
領域交通情報算出手段が、前記選択手段から出力された前記領域内のリンクの交通情報から前記領域における領域の交通情報を算出する領域交通情報算出手順と、
第１の記憶手段に前記領域交通情報算出手段で算出された前記領域における領域の交通情報を現在から過去Ｎ個（Ｎは１以上の整数）記憶し、前記領域交通情報算出手段から前記領域における領域の交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、前記領域における領域の交通情報を更新する第１の記憶手順と、
第２の記憶手段に前記選択手段から出力された前記予測対象リンクの交通情報を現在から過去Ｎ個記憶し、前記選択手段から前記予測対象リンクの交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、前記予測対象リンクの交通情報を更新する第２の記憶手順と、
予測手段が、前記第１および第２の記憶手段に、それぞれ新たな、前記領域における領域の交通情報、および前記予測対象リンクの交通情報が入力される毎に、前記第１および第２の記憶手段から、それぞれ前記領域における領域の交通情報、および前記予測対象リンクの交通情報を各Ｎ個ずつ読み出し、これらの交通情報から前記予測対象リンクの交通情報の予測を行う予測手順を有する。
【０００８】
本発明の第２の交通情報予測方法は、
選択手段が、複数のリンクの交通情報を入力し、選択することにより、予測対象リンクの交通情報と、該予測対象リンクに隣接する１つ以上のリンクからなる周辺リンクの交通情報と、該予測対象リンクを含むあらかじめ定めた領域内のリンクの交通情報とを出力する選択手順と、
領域交通情報算出手段が、前記選択手段から出力された前記領域内のリンクの交通情報から前記領域における領域の交通情報を算出する領域交通情報算出手順と、
第１の記憶手段に前記領域交通情報算出手段で算出された前記領域における領域の交通情報を現在から過去Ｎ個（Ｎは１以上の整数）記憶し、前記領域交通情報算出手段から前記領域における領域の交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、前記領域における領域の交通情報を更新する第１の記憶手順と、
第２の記憶手段に前記選択手段から出力された前記予測対象リンクの交通情報を現在から過去Ｎ個記憶し、前記選択手段から前記予測対象リンクの交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、前記予測対象リンクの交通情報を更新する第２の記憶手順と、
第３の記憶手段に前記選択手段から出力された前記周辺リンクの交通情報を該周辺リンクに含まれる各リンク毎に現在から過去Ｎ個記憶し、前記選択手段から前記周辺リンクの交通情報が入力される毎に、最も古いＮ番目の各交通情報を破棄して、前記周辺リンクの交通情報を更新する第３の記憶手順と、
予測手段が、前記第１、第２、および第３の記憶手段に、それぞれ新たな、前記領域における領域の交通情報、前記予測対象リンクの交通情報、および前記周辺リンクの交通情報が入力される毎に、前記第１、第２、および第３の記憶手段から、それぞれ前記領域における領域の交通情報、前記予測対象リンクの交通情報、および前記周辺リンクに含まれる各リンクの交通情報を各Ｎ個ずつ読み出し、これらの交通情報から前記予測対象リンクの交通情報の予測を行う予測手順とを有する。
【０００９】
また、本発明の第１の交通情報予測装置は、
複数のリンクの交通情報を入力し、選択することにより、予測対象リンクの交通情報と、該予測対象リンクを含むあらかじめ定めた領域内のリンクの交通情報とを出力する選択手段と、
前記選択手段から出力された前記領域内のリンクの交通情報から前記領域における領域の交通情報を算出する領域交通情報算出手段と、
前記領域交通情報算出手段で算出された前記領域における領域の交通情報を現在から過去Ｎ個（Ｎは１以上の整数）記憶し、前記領域交通情報算出手段から前記領域における領域の交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、前記領域における領域の交通情報を更新する第１の記憶手段と、
前記選択手段から出力された前記予測対象リンクの交通情報を現在から過去Ｎ個記憶し、前記選択手段から前記予測対象リンクの交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、前記予測対象リンクの交通情報を更新する第２の記憶手段と、
前記第１および第２の記憶手段に、それぞれ新たな、前記領域における領域の交通情報、および前記予測対象リンクの交通情報が入力される毎に、前記第１および第２の記憶手段から、それぞれ前記領域における領域の交通情報、および前記予測対象リンクの交通情報を各Ｎ個ずつ読み出し、これらの交通情報から前記予測対象リンクの交通情報の予測を行う予測手段とを有する。
【００１０】
本発明の第２の交通情報予測装置は、
複数のリンクの交通情報を入力し、選択することにより、予測対象リンクの交通情報と、該予測対象リンクに隣接する１つ以上のリンクからなる周辺リンクの交通情報と、該予測対象リンクを含むあらかじめ定めた領域内のリンクの交通情報とを出力する選択手段と、
前記選択手段から出力された前記領域内のリンクの交通情報から前記領域における領域の交通情報を算出する領域交通情報算出手段と、
前記領域交通情報算出手段で算出された前記領域における領域の交通情報を現在から過去Ｎ個（Ｎは１以上の整数）記憶し、前記領域交通情報算出手段から前記領域における領域の交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、前記領域における領域の交通情報を更新する第１の記憶手段と、
前記選択手段から出力された前記予測対象リンクの交通情報を現在から過去Ｎ個記憶し、前記選択手段から前記予測対象リンクの交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、前記予測対象リンクの交通情報を更新する第２の記憶手段と、
前記選択手段から出力された前記周辺リンクの交通情報を該周辺リンクに含まれる各リンク毎に現在から過去Ｎ個記憶し、前記選択手段から前記周辺リンクの交通情報が入力される毎に、最も古いＮ番目の各交通情報を破棄して、前記周辺リンクの交通情報を更新する第３の記憶手段と、
前記第１、第２、および第３の記憶手段に、それぞれ新たな、前記領域における領域の交通情報、前記予測対象リンクの交通情報、および前記周辺リンクの交通情報が入力される毎に、前記第１、第２、および第３の記憶手段から、それぞれ前記領域における領域の交通情報、前記予測対象リンクの交通情報、および前記周辺リンクに含まれる各リンクの交通情報を各Ｎ個ずつ読み出し、これらの交通情報から前記予測対象リンクの交通情報の予測を行う予測手段とを有する。
【００１１】
本発明は、予測対象リンクの交通情報（と予測対象リンクの周辺リンクの交通情報）の他に、予測対象リンクを含む領域の交通情報を求め、これを用いて予測対象リンクの交通情報を予測するものである。
【００１２】
図７に示す方法に比べ予測手段に入力する交通情報が少なくなるため、効率的に予測でき、また図６の方法に比べ、予測に使用する交通情報が多いため、精度良く予測できる。
【００１３】
【発明の実施の形態】
次に、本発明の実施の形態について図面を参照して説明する。
【００１４】
図１を参照すると、本発明の第１の実施形態の交通情報予測装置は選択部１と領域交通情報算出部２と記憶部３_１、３_２と予測部４で構成されている。本実施形態は、図７で説明した領域内のリンクの交通情報が領域外の他のリンクの情報とともに、時間的に一定の間隔で入力される場合である。
【００１５】
選択部１は、入力された複数のリンクの交通情報から、あらかじめ定めた領域内のリンクの交通情報を選択し、選択したリンクの交通情報を領域交通情報算出部２へ出力するとともに、予測対象リンクの交通情報を記憶部３_２へ出力する。
【００１６】
領域交通情報算出部２では、選択部１からリンクの交通情報が入力されると領域の交通情報を算出し、記憶部３_１へ送出する。領域の交通情報は、具体的には、例えば、リンクの交通情報がリンクの平均速度で表されている場合は、リンクの平均速度の領域の全リンクに対する平均をとることができ、リンクの交通情報がリンクの旅行時間で表されている場合は、領域内の全リンクに対する旅行時間の総和をとることができる。このように、リンクの交通情報の表す量によって、領域の交通情報として適切な量を定めればよい。
【００１７】
記憶部３_１では、領域の交通情報を現在から過去Ｎ個（Ｎは１以上の整数）記憶し、領域交通情報算出部２から領域の交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、領域の交通情報を更新する。
【００１８】
記憶部３_２では、予測対象リンクの交通情報を現在から過去Ｎ個記憶し、選択部１から予測対象リンクの交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、予測対象リンクの交通情報を更新する。
【００１９】
予測部４は、記憶部３_１と３_２に、それぞれ領域の新たな交通情報および予測対象リンクの新たな交通情報が入力される毎に、領域のＮ個の交通情報と予測対象リンクのＮ個の交通情報を読み出し、これら交通情報に基づいて予測を行う。予測の具体的な方法としては、自己回帰法やニューラルネットワークを適用できる。
【００２０】
このようにして、複数のリンクの交通情報が入力されると、領域内のリンクの交通情報より領域の交通情報が算出され記憶されるとともに、予測対象リンクの交通情報も記憶され、この２種類の交通情報をもとに予測を行う。
【００２１】
図２を参照すると、本発明の第２の実施形態の交通情報予測装置は選択部１１と領域交通情報算出部１２と記憶部１３₁〜１３₈と予測部１４で構成されている。本実施形態は、周辺リンクが図６の例の予測対象リンクに隣接する６つのリンクからなる場合である。
【００２２】
選択部１１は、入力された複数のリンクの交通情報から、あらかじめ定めた領域内のリンクの交通情報を選択し、選択したリンクの交通情報を領域交通情報算出部１２へ出力するとともに、予測対象リンクの交通情報を記憶部１３_２へ、予測対象リンクの周辺リンクの交通情報を記憶部１３_３〜１３_８へ出力する。
【００２３】
記憶部１３_１，１３_２は図１中の記憶部３_１，３_２と同じである。
【００２４】
記憶部１３_３〜１３_８では、周辺リンクの交通情報を現在から過去Ｎ個（Ｎは１以上の整数）記憶し、選択部１１から周辺リンクの交通情報が入力される毎に、最も古いＮ番目の交通情報を破棄して、周辺リンクの交通情報を更新する。
【００２５】
予測部１４は、記憶部１３_１と記憶部１３_２、および記憶部１３_３〜１３_８に、それぞれ領域の交通情報、予測対象リンクの交通情報、および周辺リンクの交通情報が入力される毎に、領域のＮ個の交通情報と予測対象リンクのＮ個の交通情報と各々の周辺リンクのＮ個の交通情報を読み出し、予測を行う。予測の具体的な方法としては、自己回帰法やニューラルネットワークを適用できる。予測部１４の構成は予測部４の構成と同様である。
【００２６】
このようにして、複数のリンクの交通情報が入力されると、領域内のリンクの交通情報より領域の交通情報が算出され記憶されるとともに、予測対象リンクと周辺リンクの交通情報も記録され、この３種類の交通情報をもとに予測を行う。
【００２７】
図３を参照すると、本発明の第３の実施形態の交通情報予測装置は、リンクの交通情報を入力する、モデム等の入力装置２１と、領域の交通情報を現在から過去Ｎ個（Ｎは１以上の整数）記憶する記憶装置２２_１と、予測対象リンクの交通情報を現在から過去Ｎ個記憶する記憶装置２２_２と、予測情報を出力する、プリンタ、ディスプレイ等の出力装置２３と、第１の実施の形態中の選択部１、領域交通情報算出部２、予測部４および記憶部３_１，３_２への書き込み、読み出しの各処理からなる交通情報予測プログラムを記録した、ＦＤ、ＣＤ−ＲＯＭ、半導体メモリ等の記録媒体２４から構成される。
【００２８】
図４を参照すると、本発明の第４の実施形態の交通情報予測装置は、リンクの交通情報を入力する、モデム等の入力装置３１と、領域の交通情報を現在から過去Ｎ個（Ｎは１以上の整数）記憶する記憶装置３２、予測対象リンクの交通情報を現在から過去Ｎ個記憶する記録装置３２_２と、各周辺リンクの交通情報を現在から過去Ｎ個記憶する記録装置３２_３〜３２_８と、予測情報を出力する、プリンタ、ディスプレイ等の出力装置３３と、第２の実施の形態中の選択部１１、領域交通情報算出部１２、予測部１４および記憶部１３_１〜１３_８への書き込み、読み出しの処理からなる交通情報予測プログラムを記録した、ＦＤ、ＣＤ−ＲＯＭ、半導体メモリ等の記録媒体３４から構成される。
【００２９】
図５は予測部３の一例を示す構成図である。
【００３０】
保持回路４１_１１〜４１_１Ｎは記憶部３_１からの情報を係数器４２_１１１，４２_１１２，４２_１２１，４２_１２２，・・・，４２_１Ｎ１，４２_１Ｎ２へ出力するとともに、加算器４３_１または（隣接する）保持回路４１_１１〜４１_１Ｎー１から情報を受けとると、それまで保持していた情報を他の隣接する保持回路４１_１２〜４１_１Ｎへ出力するとともに、受信した情報を係数器４２_１１１，４２_１１２，４２_１２１，４２_１２２，４２_１Ｎ１，４２_１Ｎ２へ出力する。係数器４２_１１１，４２_１１２，４２_１２１，４２_１２２，４２_１Ｎ１，４２_１Ｎ２は保持回路４１_１１〜４１_１Ｎの情報の値を係数（Ｋ_１１１，Ｋ_１１２，Ｋ_１２１，Ｋ_１２２，・・・，Ｋ_１Ｎ１，Ｋ_１Ｎ２）倍して出力する。加算器４３_１は、各係数器４２_１１１，４２_１２１，・・・，４２_１Ｎ１，４２_２１１，４２_２２１，・・・，４２_２Ｎ１の出力を加算して予測情報４４_１として出力する。保持回路４１_２１〜４１_２Ｎ，係数器４２_２１１〜４２_２Ｎ２，加算器４３_２，予測情報４４_２は記憶部３に対応する。
【００３１】
係数器４２_１１１〜４２_１Ｎ２，４２_２１１〜４２_２Ｎ２の係数Ｋ_１１１〜Ｋ_１Ｎ２，Ｋ_２１１〜Ｋ_２Ｎ２は、それぞれの予測情報４４_１，４４_２の値と実際の情報の値との２乗平均誤差が最小となるように予め定めておく。このようにすることにより、交通情報の入力間隔をＴとすると、記憶部３_１，３_２より情報を受けとると、時間Ｔ後の予測情報４４_１，４４_２が算出される。
【００３２】
この予測された情報を保持回路４１_１，４１_２１へ入力することにより、時間２Ｔ後の予測情報４４_１，４４_２が算出される。以下これを繰り返して交通情報が予測できる。
【００３３】
【発明の効果】
以上説明したように、本発明は、予測対象のリンクとそれに接続した隣接リンクだけでなく、予測対象のリンクを含む領域のリンクの交通情報を使用することにより、予測手段に入力する交通情報を少なくして予測ができるので、複雑に変化する交通情報を精度よく、かつ効率的に予測することができる。
【図面の簡単な説明】
【図１】本発明の第１の実施形態の交通情報予測装置の構成図である。
【図２】本発明の第２の実施形態の交通情報予測装置の構成図である。
【図３】本発明の第３の実施形態の交通情報予測装置の構成図である。
【図４】本発明の第４の実施形態の交通情報予測装置の構成図である。
【図５】予測部４の構成例を示す図である。
【図６】従来の交通情報予測方法の説明図である。
【図７】従来の交通情報予測方法の説明図である。
【符号の説明】
１，１１選択部
２，１２領域交通情報算出部
３_１，３_２，１３_１〜１３_８記憶部
４，１４予測部
２１，３１入力装置
２２_１，２２_２，３１_１〜３１_８記憶装置
２３，３３出力装置
２４，３４記録媒体
４１_１１〜４１_１Ｎ，４１_２１〜４１_２Ｎ保持回路
４２_１１１〜４２_１Ｎ２，４２_２１１〜４２_２Ｎ２係数器
４３_１，４３_２加算器
４４_１，４４_２予測情報[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a traffic information prediction method and device.
[0002]
[Prior art]
As a conventional traffic information prediction method, there is a method of predicting future traffic information from past traffic information of a link to be predicted. As a specific prediction method, an autoregressive method or a neural network is applied. However, this method has a problem that prediction accuracy is low because only information of a link to be predicted is used. Here, speed, travel time, congestion degree, and the like can be considered as traffic information.
[0003]
FIG. 6 shows a conventional traffic information prediction method for solving this problem. The information of the adjacent link connected to the link for which traffic information is predicted is used. That is, future traffic information of a link for which traffic information is predicted is predicted from past traffic information of a link for which traffic information is predicted and past traffic information of an adjacent link connected to the link for which traffic information is predicted. As a specific prediction method, an autoregressive method or a neural network is applied. However, this method has a problem that prediction accuracy is low in predicting traffic information that changes in a complicated manner.
[0004]
To further solve this, there is a method shown in FIG. In this method, traffic information of a link to be predicted is predicted using traffic information of a link in a region including the link to be predicted to be predicted. According to this method, traffic information of a wide range of links can be used, so that traffic information that changes in a complicated manner can be accurately predicted.
[0005]
[Problems to be solved by the invention]
However, in this method, when the number of links included in the area increases, the capacity of a storage device for storing traffic information increases, and the number of traffic information used for prediction becomes enormous. When it is realized, there is a problem that its configuration becomes complicated, and when it is realized by a program processing, there is a problem that it takes time for the prediction process.
[0006]
An object of the present invention is to provide a traffic information prediction method and apparatus that accurately and efficiently predict traffic information that changes in a complicated manner.
[0007]
[Means for Solving the Problems]
A first traffic information prediction method according to the present invention includes:
A selection step of inputting and selecting traffic information of a plurality of links to output traffic information of a link to be predicted and traffic information of links in a predetermined area including the link to be predicted, and ,
Area traffic information calculation means for calculating traffic information of an area in the area from traffic information of links in the area output from the selection means,
From the present time, N pieces of past traffic information (N is an integer of 1 or more) of the area in the area calculated by the area traffic information calculating means are stored in the first storage means. A first storage procedure for discarding the oldest N-th traffic information every time traffic information of the area is input and updating the traffic information of the area in the area;
N pieces of past traffic information of the prediction target link output from the selection means are stored in the second storage means from the present time to the past, and each time traffic information of the prediction target link is input from the selection means, the oldest N A second storage procedure for discarding the first traffic information and updating the traffic information of the link to be predicted;
The prediction unit is configured to input the first and second storage units each time new traffic information of a region in the region and traffic information of the link to be predicted are input to the first and second storage units. A prediction procedure is provided for reading traffic information of an area in the area and N pieces of traffic information of the link to be predicted from the means, and predicting traffic information of the link to be predicted from the traffic information .
[0008]
The second traffic information prediction method of the present invention includes:
The selection unit inputs and selects traffic information of a plurality of links, and thereby selects the traffic information of the link to be predicted, the traffic information of one or more links adjacent to the link to be predicted, and the traffic information of the link. A selection step of outputting traffic information of links in a predetermined area including the target link, and
Area traffic information calculation means for calculating traffic information of an area in the area from traffic information of links in the area output from the selection means,
From the present time, N pieces of past traffic information (N is an integer of 1 or more) of the area in the area calculated by the area traffic information calculating means are stored in the first storage means. A first storage procedure for discarding the oldest N-th traffic information every time traffic information of the area is input and updating the traffic information of the area in the area;
N pieces of past traffic information of the prediction target link output from the selection means are stored in the second storage means from the present time to the past, and each time traffic information of the prediction target link is input from the selection means, the oldest N A second storage procedure for discarding the first traffic information and updating the traffic information of the link to be predicted;
The traffic information of the peripheral link output from the selecting means is stored in the third storage means for each link included in the peripheral link in the past N times from the present, and the traffic information of the peripheral link is inputted from the selecting means. A third storage procedure of discarding the Nth oldest traffic information each time the traffic information of the neighboring link is updated,
The prediction means inputs new traffic information of the area in the area, the traffic information of the link to be predicted, and the traffic information of the surrounding link to the first, second, and third storage means, respectively. For each time, the traffic information of the area in the area, the traffic information of the link to be predicted, and the traffic information of each link included in the surrounding link are stored in the first, second, and third storage units, respectively, by N A prediction procedure for reading out the individual traffic information and predicting the traffic information of the prediction target link from the traffic information .
[0009]
Further, the first traffic information prediction device of the present invention includes:
Enter the traffic information of a plurality of links, by selecting the traffic information prediction target link, selection means for outputting the traffic information of the links predetermined area including the prediction target link,
Area traffic information calculation means for calculating traffic information of an area in the area from traffic information of links in the area output from the selection means ,
The area traffic information calculated by the calculating means the region in the area traffic information from the current last N (N is an integer of 1 or more) is stored, the traffic information of the area in the region from the area traffic information calculating means input First storage means for discarding the Nth oldest traffic information every time the traffic information is updated and updating traffic information of the area in the area;
Traffic information of the prediction target link output from the selection means to the last N number storage current, every time the traffic information of the prediction target link from said selecting means is input, discards the oldest N-th traffic information and, a second storage means for updating the traffic information of the prediction target link,
Said first and second memory means, respectively new, traffic information of the area in the region, and each time the traffic information of the prediction target link is inputted, from said first and second memory means, respectively traffic information of the area in the region, and the read out transportation information of the prediction target link by the N pieces, and a prediction means for these traffic information to predict the traffic information of the prediction target link.
[0010]
The second traffic information prediction device of the present invention comprises:
By inputting and selecting traffic information of a plurality of links, traffic information of a link to be predicted, traffic information of a peripheral link including one or more links adjacent to the link to be predicted, and the link to be predicted are included. Selecting means for outputting traffic information of links in a predetermined area;
Area traffic information calculation means for calculating traffic information of an area in the area from traffic information of links in the area output from the selection means ,
The area traffic information calculated by the calculating means the region in the area traffic information from the current last N (N is an integer of 1 or more) is stored, the traffic information of the area in the region from the area traffic information calculating means input each time it is, discards the oldest N-th traffic information, a first storage means for updating the traffic information of the area in the region,
The past N traffic information of the link to be predicted output from the selecting means is stored from the present to the past N times, and every time the traffic information of the link to be predicted is inputted from the selecting means, the oldest Nth traffic information is discarded. and, a second storage means for updating the traffic information of the prediction target link,
The traffic information of the peripheral link output from the selecting means is stored in the past N times for each link included in the peripheral link from the present, and each time the traffic information of the peripheral link is input from the selecting means, discard the old N-th the traffic information, and the third storage means to update the traffic information of the peripheral link,
Said first, second, and third storage means, each new traffic information of the area in the region, the prediction target link of the traffic information, and each time the traffic information of the peripheral link is input, the first, read from the second, and third storage means, the traffic information for each area in the region, the transportation information of the prediction target link, and the traffic information of each link included in said peripheral link portions each N or Prediction means for predicting the traffic information of the link to be predicted from the traffic information.
[0011]
The present invention obtains traffic information of an area including a link to be predicted, in addition to traffic information of a link to be predicted (and traffic information of a link around the link to be predicted), and predicts traffic information of the link to be predicted using the obtained information. Is what you do.
[0012]
The traffic information input to the prediction means is smaller than in the method shown in FIG. 7, so that the prediction can be made efficiently. In addition, compared with the method of FIG.
[0013]
BEST MODE FOR CARRYING OUT THE INVENTION
Next, embodiments of the present invention will be described with reference to the drawings.
[0014]
Referring to FIG. 1, and a first the embodiment of the traffic information prediction apparatus selecting section 1 and the area traffic information calculating section 2 and the storage unit 3 _1, 3 ₂ and the prediction unit 4 of the present invention. The present embodiment is a case where traffic information of a link in the area described with reference to FIG. 7 is input together with information of another link outside the area at a constant time interval.
[0015]
The selecting unit 1 selects traffic information of a link in a predetermined area from the input traffic information of a plurality of links, outputs the traffic information of the selected link to the area traffic information calculating unit 2, and sets a prediction target. and outputs the link of traffic information to the storage unit 3 _2.
[0016]
In the region the traffic information calculation section 2, the traffic information of the link from the selecting unit 1 is input to calculate the traffic information area, and sends it to the storage unit 3 _1. Specifically, for example, when the traffic information of the link is represented by the average speed of the link, the traffic information of the area can be averaged for all the links in the area having the average speed of the link, and the traffic of the link can be calculated. When the information is represented by the travel time of the link, the sum of the travel time for all the links in the area can be obtained. As described above, an appropriate amount may be determined as the traffic information of the area based on the amount represented by the traffic information of the link.
[0017]
The storage unit 3 _1, (the N 1 or more integer) past N pieces of traffic information area from the current store, the area traffic information calculating section 2 each time the traffic information of the area is input, the oldest N th Discard the traffic information and update the traffic information for the area.
[0018]
The storage unit 3 ₂ and past the N store traffic information of the prediction target link from the current, every time the traffic information prediction target link selection unit 1 is input, discards the oldest N-th traffic information, Update the traffic information of the predicted link.
[0019]
Prediction unit 4, the storage unit 3 ₁ and 3 _2, each time a new traffic information of a new traffic information and the prediction target link each region is input, the prediction target link and the N traffic information area N Individual traffic information is read, and prediction is performed based on the traffic information. As a specific method of prediction, an autoregressive method or a neural network can be applied.
[0020]
When the traffic information of a plurality of links is input in this way, the traffic information of the area is calculated and stored from the traffic information of the links in the area, and the traffic information of the link to be predicted is also stored. Predict based on traffic information.
[0021]
Referring to FIG. 2, and a prediction unit 14 the traffic information prediction apparatus of the second embodiment of the present invention the selector 11 and the area traffic information calculating unit 12 and the storage unit _131-134 _8. This embodiment is a case where the peripheral link is composed of six links adjacent to the prediction target link in the example of FIG.
[0022]
The selection unit 11 selects traffic information of a link in a predetermined area from the input traffic information of the plurality of links, outputs the traffic information of the selected link to the area traffic information calculation unit 12, and sets a prediction target. the link of the traffic information to the storage unit 13 _2, and outputs the traffic information around the link of the prediction target link to the storage unit ₁₃ 3-13 _8.
[0023]
Storage unit ₁₃ 1, 13 ₂ is the same as the storage unit ₃ 1, _{3 2} in Fig.
[0024]
The storage unit 13 _3-13 _8, the traffic information around the link past N-number from the current (N is an integer of 1 or more) stored, every time the traffic information around the link is input from the selection unit 11, the oldest N Discard the second traffic information and update the traffic information of surrounding links.
[0025]
Prediction unit 14, the storage unit 13 ₁ and the memory unit 13 _2, and the storage unit ₁₃ 3-13 _8, traffic information respectively area, the traffic information of the prediction target link, and every of the traffic information around the link is input , N traffic information of the area, N traffic information of the link to be predicted, and N traffic information of each peripheral link are read and predicted. As a specific method of prediction, an autoregressive method or a neural network can be applied. The configuration of the prediction unit 14 is the same as the configuration of the prediction unit 4.
[0026]
In this way, when the traffic information of a plurality of links is input, the traffic information of the area is calculated and stored from the traffic information of the links in the area, and the traffic information of the prediction target link and the surrounding links are also recorded. Prediction is performed based on these three types of traffic information.
[0027]
Referring to FIG. 3, the traffic information prediction device according to the third embodiment of the present invention includes an input device 21 such as a modem for inputting traffic information of a link, and N past traffic information of a region (N is an integer of 1 or more) and the storage device 22 ₁ for storing, a storage device 22 ₂ for the last N number storing traffic information of the prediction target link from the current, and outputs the prediction information, the printer, an output device 23 such as a display, the selecting unit 1 in one embodiment, the area traffic information calculator 2, writing into the prediction unit 4 and the storage unit 3 _1, 3 _2, was recorded traffic information prediction program of the processes of reading, FD, CD -A recording medium 24 such as a ROM and a semiconductor memory.
[0028]
Referring to FIG. 4, a traffic information prediction device according to a fourth embodiment of the present invention includes an input device 31 such as a modem for inputting traffic information of a link, and N past traffic information of a region (N is an integer of 1 or more) memory for storing device 32, a recording device 32 ₂ for the last N number storing traffic information of the prediction target link from the current, the recording apparatus 32 ₃ to the last N number storing traffic information of each peripheral link from the current 32 _8, and outputs the prediction information, the printer, an output device 33 such as a display, the second selector 11 in the embodiment, area traffic information calculating unit 12, prediction unit 14 and the storage unit _131-134 ₈ It is composed of a recording medium 34 such as an FD, a CD-ROM, a semiconductor memory, etc., on which a traffic information prediction program composed of writing and reading processes to and from a computer is recorded.
[0029]
FIG. 5 is a configuration diagram illustrating an example of the prediction unit 3.
[0030]
Holding circuit ₄₁ 11 _{to 41 1N} is a coefficient unit information from the storage unit _{_{_{_{3 1 42 111, 42 112,}}}} 42 121, 42 122, ···, 42 1N1, 42 and outputs to _1N2, adder 43 ₁ or ( When the information is received from the (adjacent) holding circuits 41 _{11 to} 41 _1N-1 , the information held so far is output to the other adjacent holding circuits 41 _{12 to} 41 _1N , and the received information is converted into the coefficient units 42 ₁₁₁ _{_{_{_{, 42 112, 42 121, 42}}}} 122, 42 1N1, 42 outputs to _1N2. Coefficient unit _{_{_{_{42 111, 42 112, 42 121}}}} , 42 122, 42 1N1, 42 1N2 is the value of the holding circuit ₄₁ 11 _{to 41 1N} information coefficient _{_{_{_{(K 111, K 112, K}}}} 121, K 122, ···, _K _1N1, K _1N2) times and outputs. The adder 43 _1, the coefficient unit _{_{_{_{42 111, 42 121, ···,}}}} 42 1N1, 42 211, 42 221, and _outputs., As prediction information 44 ₁ adds the output of the _{42 2N1.} Holding circuit ₄₁ 21 _{to 41 2N,} coefficient unit ₄₂ 211 _{through 42 2N2,} adder 43 _2, the prediction information 44 ₂ corresponding to the storage unit 3.
[0031]
Coefficient unit _{_₄₂} 111 _~42 _1N2, ₄₂ 211 _~42 coefficient of _{_{_{_{2N2 K 111 ~K 1N2, K 211}}}} ~K 2N2 is the mean square of the value of the actual information and respective prediction information ₄₄ 1, 44 ₂ values It is determined in advance so that the error is minimized. In this way, when the input interval of the traffic information is T, when receiving the information from the storage unit 3 _1, 3 _2, prediction information 44 1 after time _T, 44 ₂ are calculated.
[0032]
By inputting the predicted information to the holding circuit ₄₁ 1, _{41 21,} prediction information ₄₄ 1 after time 2T, 44 ₂ are calculated. Hereinafter, traffic information can be predicted by repeating this.
[0033]
【The invention's effect】
As described above, the present invention uses traffic information not only for a link to be predicted and an adjacent link connected thereto but also for a link in an area including the link to be predicted, so that traffic information to be input to the prediction unit is obtained. Since prediction can be performed with a small amount, traffic information that changes in a complicated manner can be predicted accurately and efficiently.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of a traffic information prediction device according to a first embodiment of the present invention.
FIG. 2 is a configuration diagram of a traffic information prediction device according to a second embodiment of the present invention.
FIG. 3 is a configuration diagram of a traffic information prediction device according to a third embodiment of the present invention.
FIG. 4 is a configuration diagram of a traffic information prediction device according to a fourth embodiment of the present invention.
FIG. 5 is a diagram illustrating a configuration example of a prediction unit 4.
FIG. 6 is an explanatory diagram of a conventional traffic information prediction method.
FIG. 7 is an explanatory diagram of a conventional traffic information prediction method.
[Explanation of symbols]
1,11 selector 2, 12 area traffic information calculating unit ₃ _1, 3 _2, 131-134 ₈ storage unit 4, 14 prediction unit 21, 31 input device ₂₂ _1, 22 _2, 31 1 to 31 ₈ storage device 23 , 33 output device 24, 34 recording medium _{_{_{_{41 11 ~41 1N, 41 21 ~41}}}} 2N holding circuits _{_{_{_{42 111 ~42 1N2, 42 211 ~42}}}} 2N2 coefficient unit ₄₃ 1, 43 ₂ adders ₄₄ 1, 44 ₂ prediction information

Claims

In a method in which a traffic information prediction device predicts traffic information of a link to be predicted,
A selection step of inputting and selecting traffic information of a plurality of links to output traffic information of a link to be predicted and traffic information of links in a predetermined area including the link to be predicted, and ,
Area traffic information calculation means for calculating traffic information of an area in the area from traffic information of links in the area output from the selection means,
From the present time, N pieces of past traffic information (N is an integer of 1 or more) of the area in the area calculated by the area traffic information calculating means are stored in the first storage means. A first storage procedure for discarding the oldest N-th traffic information every time traffic information of the area is input and updating the traffic information of the area in the area;
N pieces of past traffic information of the prediction target link output from the selection means are stored in the second storage means from the present time to the past, and each time traffic information of the prediction target link is input from the selection means, the oldest N A second storage procedure for discarding the first traffic information and updating the traffic information of the link to be predicted;
The prediction unit is configured to input the first and second storage units each time new traffic information of a region in the region and traffic information of the link to be predicted are input to the first and second storage units. A prediction procedure for reading traffic information of an area in the area and traffic information of the link to be predicted from the means, and predicting traffic information of the link to be predicted from the traffic information;
Traffic information prediction method having

In a method in which a traffic information prediction device predicts traffic information of a link to be predicted,
The selection unit inputs and selects traffic information of a plurality of links, and thereby selects the traffic information of the link to be predicted, the traffic information of one or more links adjacent to the link to be predicted, and the traffic information of the link. A selection step of outputting traffic information of links in a predetermined area including the target link, and
Area traffic information calculation means for calculating traffic information of an area in the area from traffic information of links in the area output from the selection means,
From the present time, N pieces of past traffic information (N is an integer of 1 or more) of the area in the area calculated by the area traffic information calculating means are stored in the first storage means. A first storage procedure for discarding the oldest N-th traffic information every time traffic information of the area is input and updating the traffic information of the area in the area;
N pieces of past traffic information of the prediction target link output from the selection means are stored in the second storage means from the present time to the past, and each time traffic information of the prediction target link is input from the selection means, the oldest N A second storage procedure for discarding the first traffic information and updating the traffic information of the link to be predicted;
The traffic information of the peripheral link output from the selecting means is stored in the third storage means for each link included in the peripheral link in the past N times from the present, and the traffic information of the peripheral link is inputted from the selecting means. A third storage procedure of discarding the Nth oldest traffic information each time the traffic information of the neighboring link is updated,
The prediction means inputs new traffic information of the area in the area, the traffic information of the link to be predicted, and the traffic information of the surrounding link to the first, second, and third storage means, respectively. For each time, the traffic information of the area in the area, the traffic information of the link to be predicted, and the traffic information of each link included in the surrounding link are stored in the first, second, and third storage units, respectively, by N A prediction procedure of reading out the traffic information one by one and predicting the traffic information of the prediction target link from the traffic information;
Traffic information prediction method having

Enter the traffic information of a plurality of links, by selecting the traffic information of the prediction target link, selection means for outputting the traffic information of the links predetermined area including the prediction target link,
Area traffic information calculation means for calculating traffic information of an area in the area from traffic information of links in the area output from the selection means ,
The area traffic information calculated by the calculating means the region in the area traffic information from the current last N (N is an integer of 1 or more) is stored, the traffic information of the area in the region from the area traffic information calculating means input First storage means for discarding the Nth oldest traffic information every time the traffic information is updated and updating traffic information of the area in the area;
Traffic information of the prediction target link output from the selection means to the last N number storage current, every time the traffic information of the prediction target link from said selecting means is input, discards the oldest N-th traffic information and, a second storage means for updating the traffic information of the prediction target link,
Said first and second memory means, respectively new, traffic information of the area in the region, and each time the traffic information of the prediction target link is inputted, from said first and second memory means, respectively traffic information of the area in the region, and the read out transportation information of the prediction target link by the N pieces, traffic with <br/> a prediction means for these traffic information to predict the traffic information of the prediction target link Information prediction device.

By inputting and selecting traffic information of a plurality of links, the traffic information of the link to be predicted, the traffic information of a peripheral link including one or more links adjacent to the link to be predicted, and the link to be predicted are included. Selecting means for outputting traffic information of links in a predetermined area;
Area traffic information calculation means for calculating traffic information of an area in the area from traffic information of links in the area output from the selection means ,
The area traffic information calculated by the calculating means the region in the area traffic information from the current last N (N is an integer of 1 or more) is stored, the traffic information of the area in the region from the area traffic information calculating means input each time it is, discards the oldest N-th traffic information, a first storage means for updating the traffic information of the area in the region,
The past N traffic information of the link to be predicted output from the selecting means is stored from the present to the past N times, and every time the traffic information of the link to be predicted is inputted from the selecting means, the oldest Nth traffic information is discarded. and, a second storage means for updating the traffic information of the prediction target link,
The traffic information of the peripheral link output from the selecting means is stored in the past N times for each link included in the peripheral link from the present, and each time the traffic information of the peripheral link is input from the selecting means, discard the old N-th the traffic information, and the third storage means to update the traffic information of the peripheral link,
Said first, second, and third storage means, each new traffic information of the area in the region, the prediction target link of the traffic information, and each time the traffic information of the peripheral link is input, the first, read from the second, and third storage means, the traffic information for each area in the region, the transportation information of the prediction target link, and the traffic information of each link included in said peripheral link portions each N or , the traffic information prediction apparatus having <br/> a prediction means for these traffic information to predict the traffic information of the prediction target link.

By inputting and selecting traffic information of a plurality of links, a selection process of outputting traffic information of a link to be predicted and traffic information of links in a predetermined area including the link to be predicted ,
Area traffic information calculation processing for calculating traffic information of an area in the area from traffic information of links in the area output in the selection processing ;
From the present, N pieces of past traffic information (N is an integer of 1 or more) of the area in the area calculated in the area traffic information calculation processing are stored in the first storage unit, and are calculated in the area traffic information calculation processing . each time the traffic information of the area in the region is input, discards the oldest N-th traffic information, a first storage processing for updating the traffic information of the area in the region,
Traffic information of the prediction target link output in the selection process the second current from and past the N stored in the memory unit, each time the traffic information of the prediction target link output in the selection process is input, most discard the old N-th traffic information, a second storage process of updating the traffic information of the prediction target link,
Said first and second storage portions, respectively a new, traffic information of the area in the region, and each time the traffic information of the prediction target link is inputted, from said first and second storage portions, respectively traffic information of the area in the region, and transportation information of the prediction target link read by the N pieces, these traffic information from the prediction target link traffic information computer <br/> a prediction process for predicting the A recording medium on which a traffic information prediction program to be executed is recorded.

By inputting and selecting information on a plurality of links, traffic information on a link to be predicted, traffic information on a peripheral link including one or more links adjacent to the link to be predicted, and A selection process of outputting traffic information of links within the determined area,
Area traffic information calculation processing for calculating traffic information of an area in the area from traffic information of links in the area output in the selection processing ;
From the present, N pieces of past traffic information (N is an integer of 1 or more) of the area in the area calculated in the area traffic information calculation processing are stored in the first storage unit, and calculated in the area traffic information calculation processing . A first storage process of discarding the oldest N-th traffic information and updating the traffic information of the area in the area each time traffic information of the area in the area is input;
The traffic information of the prediction target link output in the selection processing is stored in the second storage unit in the past N pieces from the present, and every time the traffic information of the prediction target link output in the selection processing is input , most discard the old N-th traffic information, a second storage process of updating the traffic information of the prediction target link,
The traffic information of the peripheral link output in the selection process is stored in the third storage unit for each link included in the peripheral link in the past N times from the present and the traffic information of the peripheral link output in the selection process is output. each time the information is input, and the most to discard old N-th the traffic information, the third storage processing to update the traffic information of the peripheral link,
Said first, second, and third storage unit, respectively new, traffic information of the area in the region, the prediction target link of the traffic information, and each time the traffic information of the peripheral link is input, the first, read from the second, and the third storage unit, the traffic information of each region in the region, the transportation information of the prediction target link, and the traffic information of each link included in said peripheral link portions each N or , a recording medium which records the traffic information prediction program for executing the predictive processing and <br/> the computer for predicting the traffic information of the prediction target link from these traffic information.