JP3568768B2 - Vehicle position identification device - Google Patents

Description

【００６１】
例えば、図１５（ａ）に示すように、車両の現在位置が地点Ｐｊであり、その地点Ｐｊに対応するリンクＬ６の始点ノードがＮ５であり、終点ノードがＮ６であり、幅員が１０メートルであり、車線数が４であり、上下線が分離されておらず、規制方向が逆方向である場合、すなわち、リンクＬ６に対応する道路が、１０メートル幅員の片側２車線の対面通行道路である場合、位置同定手段１２は、図１５（ｂ）に示す２．５メートル幅員の車線Ｌａｎｅ１から車線Ｌａｎｅ４に対応する４つの仮想車線リンクＶＬ１からＶＬ４を図１６および図１７に示すように生成する。
【００６２】
このとき、４つの車線Ｌａｎｅ１からＬａｎｅ４のうち、車線Ｌａｎｅ３，Ｌａｎｅ４は、その通行方向（規制されていない方向）が車両の進行方向の逆方向であるので、車線Ｌａｎｅ３を車両が走行している確率ｋ３と、車線Ｌａｎｅ４を車両が走行している確率ｋ４は、ともに０％に設定される。
【００６３】
次に、例えば、走行軌跡が車線Ｌａｎｅ１と車線Ｌａｎｅ２との境界から路側へ０．７５メートルのところを通過している場合、ΔＷ１は、０．５（＝１．２５−０．７５）メートルになり、ΔＷ２は２．０（＝１．２５＋０．７５）メートルになり、図１７に示すように、位置同定手段１２は、式（１）に従って、車線Ｌａｎｅ１を車両が走行している確率ｋ１を、８０％（＝（１−０．５／２．５）×１００％）と計算し、車線Ｌａｎｅ２を車両が走行している確率ｋ２を、２０％（＝（１−２．０／２．５）×１００％）と計算する。なお、リンクＬ６に対応する道路と走行軌跡が平行である場合、図１８に示すように、ΔＷ１およびΔＷ２は一定であるので、ｋ１およびｋ２も一定のままである。
【００６４】
したがって、位置同定手段１２は、４つの車線に対応する確率ｋ１からｋ４のうち、最も高い確率ｋ１に対応する車線である車線Ｌａｎｅ１を、車両が走行している車線であると同定する。
【００６５】
以上のように、この実施の形態１によれば、道路データから、各車線に対応する仮想車線リンクを生成し、その仮想車線リンクのデータと走行軌跡から、各車線について、車両が走行している確率（存在確率）を計算し、その確率に基づいて走行車線を同定しているので、道路の幅員方向における車両位置を同定することができるという効果が得られる。
【００６６】
なお、このようにして同定した走行車線をユーザに報知するようにしてもよい。その際、各車線についての車両の存在確率を表示するようにしてもよい。
【００６７】
また、上述の走行車線の同定を利用して、走行車線を考慮した経路探索、経路案内などを行うようにしてもよい。
【００６８】
また、上述の存在確率は式（１）に従って算出されるが、存在確率を導出する式は特にこの式に限定されるものではなく、各車線における車両の存在確率を適切に導出することができる式であれば、他の式を使用してもよい。
【００６９】
実施の形態２．
この発明の実施の形態２による車両位置同定装置は、実施の形態１の道路網記憶手段１と車線算出手段１１を変更したものであるので、それらについて説明し、その他の構成要素の説明を省略する。
【００７０】
実施の形態２による車両位置同定装置においては、道路網記憶手段１は、実施の形態１の道路網記憶手段１が記憶する道路網（第１の道路網）の道路データの他、詳細な道路網（第２の道路網）に対応するリンクの道路データを記憶し、車線算出手段１１は、車両の現在位置付近に関する詳細な道路網の道路データが記憶されているか否かを判断し、適宜、詳細な道路網に対応するリンクの道路データを読み出して、その道路データに基づいて仮想車線リンクを計算するものである。
【００７１】
次に動作について説明する。
図１９および図２０は、詳細な道路網への切り替えの例を説明する図である。なお、ここでは、図１９および図２０の例に沿って動作を説明する。
【００７２】
道路網記憶手段１は、図１９（ａ）に示す道路に対応するリンクＬｂ１からＬｂ４（図１９（ｂ））の道路データと、図１９（ｂ）の範囲６１の詳細な道路網に対応するリンクＬｃ１からＬｃ１０（図１９（ｃ））の道路データを記憶している。なお、範囲６１は、その各頂点の緯度と経度が道路網記憶手段１に予め記憶されることにより設定される。
【００７３】
車線算出手段１１は、現在位置計測手段５からの現在位置の情報に基づき、車両の現在位置が、詳細な道路網の道路データが記憶されている範囲（例えば範囲６１）内にあるか否かを判定し、現在位置がそのような範囲内にあると判定した場合、その範囲６１のリンクを、詳細な道路網に対応するリンクに変更する。
【００７４】
例えば、図１９（ｂ）に示す道路で車両が地点Ｐｉ−Ｎに位置するときには、現在位置（地点Ｐｉ−Ｎ）は範囲６１の外側にあるので、リンクすなわち道路網の変更は行われない。この場合、リンクＬｂ１の道路データに基づいて仮想車線リンクが算出される。
【００７５】
一方、車両が地点Ｐｉに位置するときには、現在位置（地点Ｐｉ）は範囲６１内にあるので、リンクＬｂ１からＬｂ４は、図１９（ｃ）に示す詳細な道路地図に対応するリンクＬｃ１からＬｃ１０に変更される。
【００７６】
そして、車線算出手段１１は、それらのリンクＬｃ１からＬｃ１０のうちの現在位置に対応するリンク（今の場合、リンクＬｃ９）の道路データに基づいて仮想車線リンクを計算する。例えば、図１９（ｂ）および図１９（ｃ）に示す地点Ｐｉに車両が位置する場合、詳細な道路地図に対応するリンクＬｃ９の道路データ（幅員＝Ｗ２、車線数＝１、上下分離＝有）に基づいて仮想車線リンクが算出される。
【００７７】
車線算出手段１１により算出された仮想車線リンクは位置同定手段１２に供給され、その仮想車線リンクに基づいて車両の位置が同定される。
【００７８】
このように道路網を切り換えることにより、範囲６１内においては、リンクＬｂ１からＬｂ４の代わりに、詳細な道路地図に対応するリンクＬｃ１からＬｃ１０に基づいて仮想車線リンクが算出されるので、図１９（ｂ）に示すように車両が道路から離脱したと判定されることなく、図１９（ｃ）に示すように道路上を継続的に走行していると判定される。
【００７９】
また、例えば図２０（ａ）に示す道路を走行する場合、道路網記憶手段１に、図２０（ａ）に示す道路に対応するリンクＬｂ１からＬｂ３（図２０（ｂ））の道路データと、図２０（ｂ）の範囲６２の詳細な道路地図に対応するリンクＬｃ１からＬｃ５（図２０（ｃ））の道路データが記憶されているとすると、車両が範囲６２の外側を走行しているときには、車線算出手段１１は、そのリンクＬｂ１またはリンクＬｂ３に基づいて仮想車線リンクを算出するが、車両が範囲６２内を走行しているときには、リンクＬｂ１からＬｂ３を、詳細な道路地図に対応するリンクＬｃ１からＬｃ５に切り換えて、それらのリンクＬｃ１からＬｃ５のいずれかのリンク（今の場合、リンクＬｃ３）に基づいて仮想車線リンクを算出する。
【００８０】
このように道路地図を切り換えることにより、範囲６２内においては、詳細な道路地図に対応するリンクＬｃ１からＬｃ１５に基づいて仮想車線リンクが算出されるので、図２０（ｂ）に示すように車両が車線から外れていると判定されることなく、図２０（ｃ）に示すように１つの車線上を継続的に走行していると判定される。
【００８１】
以上のように、この実施の形態２によれば、同一範囲について、縮尺の異なる複数の道路網を記憶しておき、詳細な道路網の道路データを使用して、走行車線を同定するようにしたので、より適切な車線を走行車線と同定することができるという効果が得られる。
【００８２】
なお、上記実施の形態２においては、範囲６１，６２は、その各頂点の緯度と経度が道路網記憶手段１に予め記憶されることにより設定されているが、地図データベースの管理距離単位で正規化された座標情報を利用して範囲６１，６２を設定してもよい。その場合、車両の現在位置を正規化された座標値に変換した後、現在位置が範囲６１，６２内にあるか否かが判定される。
【００８３】
また、図２１に示すように、範囲６１，６２をリンク６４の形状に対応して予め設定しておくようにしてもよい。このように、リンク６４に関連づけて範囲６１，６２を設定することにより、車両が走行中の道路に対応するリンクが特定されると、車両が範囲６１，６２内に位置すると判定される。
【００８４】
さらに、図２２に示すように、外縁がノード６５を通過するように、範囲６１，６２を設定し、その範囲６１，６２内のリンクＬｂ５からＬｂ８に関連づけて、その範囲６１，６２の詳細な道路網に対応するリンクを記憶しておき、車両が走行中のリンクがその範囲６１，６２内のリンクＬｂ５からＬｂ８である場合に、リンクＬｂ５からＬｂ８を詳細な道路網に対応するリンクに変更するようにしてもよい。
【００８５】
実施の形態３．
この発明の実施の形態３による車両位置同定装置は、実施の形態１の位置同定手段１２を変更したものであるので、位置同定手段１２について説明し、その他の構成要素の説明を省略する。
【００８６】
実施の形態３による車両位置同定装置においては、位置同定手段１２は、車両が車線変更する場合、変更元の車線（仮想車線リンクＶＬｉ）に対する上述の確率ｋｉと変更先の車線（仮想車線リンクＶＬｊ）に対する確率ｋｊを比較し、変更先の車線に対する確率ｋｊが、変更元の車線に対する上述の確率ｋｉと所定のヒステリシス幅Δｋとの和より大きくなった（ｋｊ＞（ｋｉ＋Δｋ））とき、変更先の車線を車両が走行していると判断する。すなわち、変更先の車線に対する確率ｋｊが、元の車線に対する確率ｋｉと所定のヒステリシス幅Δｋとの和以下である時点では、位置同定手段１２は、変更元の車線を車両が走行していると判断する。
【００８７】
次に動作について説明する。
図２３は、車両が車線変更したときの位置同定手段１２の動作の一例を説明する図である。ここでは、図２３に沿って動作を説明する。
【００８８】
図２３（ａ）に示す道路は、幅員がＷである、２車線の一方通行道路である。そして、車両は、走行軌跡７３のとおり、最初に車線Ｌａｎｅ１を走行し、地点７１で車線変更を開始し、地点７２で車線変更を終了し、その後車線Ｌａｎｅ２を走行したものとする。
【００８９】
このとき、車線Ｌａｎｅ１の仮想車線リンク７４と走行軌跡７３との距離、および車線Ｌａｎｅ２の仮想車線リンク７５と走行軌跡７３との距離は図２３（ｂ）に示すように変化していき、車線Ｌａｎｅ１を車両が走行している確率および車線Ｌａｎｅ２を車両が走行している確率は、式（１）に従って図２３（ｃ）に示すように計算される。
【００９０】
このとき、車線Ｌａｎｅ１，Ｌａｎｅ２に対する車両の存在確率は図２３（ｃ）に示すようにそれぞれ変化していき、時刻ｔｉ−ｃにおいて一致する。それ以後においては、それらの存在確率の大小関係が逆転し、車線Ｌａｎｅ２に対する車両の存在確率が増加していき、時刻ｔｉ−ｂにおいて、車線Ｌａｎｅ２に対する車両の存在確率と車線Ｌａｎｅ１に対する車両の存在確率との差が所定のヒステリシス幅より大きくなる。
【００９１】
位置同定手段１２は、時刻ｔｉ−ｃまでは、実施の形態１のものと同様にして走行車線を車線Ｌａｎｅ１と同定し、それ以降において、車線Ｌａｎｅ２に対する車両の存在確率と車線Ｌａｎｅ１に対する車両の存在確率との差が所定のヒステリシス幅より大きくなる時刻ｔｉ−ｂまでは、図２３（ｄ）に示すように、走行車線を車線Ｌａｎｅ１と同定し、時刻ｔｉ−ｂ以降においては、走行車線を車線Ｌａｎｅ２と同定する。
【００９２】
以上のように、この実施の形態３によれば、例えば駐車車両を避ける場合など、一時的に隣りの車線にはみ出して走行する場合においても、所定のヒステリシス幅以上、他の車線に対する車両の存在確率が大きくなるまで、同定する走行車線が変更されないので、走行車線を誤って同定する可能性を低減することができるという効果が得られる。
【００９３】
実施の形態４．
この発明の実施の形態４による車両位置同定装置は、実施の形態１の位置同定手段１２を変更したものであるので、位置同定手段１２について説明し、その他の構成要素の説明を省略する。
【００９４】
実施の形態４による車両位置同定装置においては、位置同定手段１２は、走行軌跡の幅員方向の変化量が、車線の幅員に所定の係数を乗じた幅以上になった場合に車線変更が行われたと判断する。
【００９５】
また、位置同定手段１２は、交差点での右折または左折時において、走行軌跡との距離が短い仮想道路リンクから順番に、式（２）に従って、その仮想道路リンクに対応する車線を車両が走行している確率を算出し、最も大きい確率を有する仮想道路リンクに対応する車線を走行車線と同定する。
【数２】

【００９６】
次に動作について説明する。
図２４は、車両が車線変更したときの位置同定手段１２の動作の一例を説明する図であり、図２５は、車両が交差点で左折したときの位置同定手段１２の動作の一例を説明する図である。ここでは、図２４と図２５に沿って動作を説明する。
【００９７】
図２４に示すように、車両が地点Ｐｉ−ｃから右側の地点Ｐｉ−ａへ移動して車線変更する場合、位置同定手段１２は、走行軌跡の幅員方向の変化量が車線の幅員に所定の係数を乗じた値以上になる地点Ｐｉ−ｂを車両が通過した時点ｔｉ−ｂに走行車線が車線Ｌａｎｅ２に変更されたと判断する。
【００９８】
また、図２５（ａ）に示すように、例えば、車両が交差点で左折して、上下線が分離された３車線の道路に進入した後、地点Ｐｉ−１と地点Ｐｉで車線変更した場合、位置同定手段１２は、まず、地点Ｐｉ−２付近において、式（２）に従って計算した確率（図２５（ｂ））に従って、走行車線を車線Ｌａｎｅ１と同定する。このとき、車線Ｌａｎｅ１を車両が走行している確率は５０％であり、車線Ｌａｎｅ２を車両が走行している確率は３３％であり、車線Ｌａｎｅ３を車両が走行している確率は１７％である。
【００９９】
次に、地点Ｐｉ−１において、走行軌跡の幅員方向の変化量が、車線の幅員に所定の係数を乗じた幅以上になると、位置同定手段１２は、車両が右側へ車線変更したと判断する。そして、それ以降においては、位置同定手段１２は、最も左側にある車線Ｌａｎｅ１を車両が走行している確率を０％に設定した後、車線Ｌａｎｅ２を車両が走行している確率と、車線Ｌａｎｅ３を車両が走行している確率とを計算する。このとき、車線Ｌａｎｅ２を車両が走行している確率は６７％であり、車線Ｌａｎｅ３を車両が走行している確率は３３％である。なお、この場合、車線Ｌａｎｅ１を車両が走行している確率を０％に設定したので、位置同定手段１２は、車線数Ｎを２として確率を計算する。
【０１００】
さらに、地点Ｐｉにおいて、同様に、走行軌跡の幅員方向の変化量が、車線の幅員に所定の係数を乗じた幅以上になると、位置同定手段１２は、車両がさらに右側へ車線変更したと判断する。そして、それ以降においては、位置同定手段１２は、車線Ｌａｎｅ１の次に左側にある車線Ｌａｎｅ２を車両が走行している確率を０％に設定する。したがって、車線Ｌａｎｅ３を車両が走行している確率が１００％になる。
【０１０１】
一方、図２５（ｃ）に示すように、左折後の走行車線が車線Ｌａｎｅ２であると誤って判断された場合、位置同定手段１２は、地点Ｐｊ−３において、式（２）に従って計算した確率（図２５（ｄ））に従って、走行車線を車線Ｌａｎｅ２と同定する。このとき、走行軌跡から近い車線順に確率を計算すると、車線Ｌａｎｅ２を車両が走行している確率は５０％になり、車線Ｌａｎｅ１を車両が走行している確率は３３％になり、車線Ｌａｎｅ３を車両が走行している確率は１７％になる。なお、この確率を計算する順番は、慣例的に車両が走行することが多い車線順としてもよい。
【０１０２】
次に、地点Ｐｊ−２において、走行軌跡の幅員方向の変化量が、車線の幅員に所定の係数を乗じた幅以上になると、位置同定手段１２は、車両が右側へ車線変更したと判断する。そして、それ以降においては、位置同定手段１２は、最も左側にある車線Ｌａｎｅ１を車両が走行している確率を０％に設定した後、車線Ｌａｎｅ２を車両が走行している確率と、車線Ｌａｎｅ３を車両が走行している確率とを計算する。なお、位置同定手段１２は、走行軌跡に基づいて車両が車線Ｌａｎｅ２から右側へ車線変更したと判断して、変更先の車線Ｌａｎｅ３、元の車線Ｌａｎｅ２の順番で確率を計算する。このとき、車線Ｌａｎｅ３を車両が走行している確率は６７％であり、車線Ｌａｎｅ２を車両が走行している確率は３３％である。なお、この場合、車線Ｌａｎｅ１を車両が走行している確率を０％に設定したので、位置同定手段１２は、車線数Ｎを２として確率を計算する。
【０１０３】
さらに、地点Ｐｊ−１において、同様に、走行軌跡の幅員方向の変化量が、車線の幅員に所定の係数を乗じた幅以上になると、位置同定手段１２は、車両がさらに右側へ車線変更したと判断するが、車線Ｌａｎｅ３の右側には車線が存在しないので、車線Ｌａｎｅ３を車両が走行している確率を１００％に設定し、その他の車線を車両が走行している確率を０％に設定する。すなわち、位置同定手段１２は、この場合、走行車線を車線Ｌａｎｅ３と同定し、地点Ｐｊ−３と地点Ｐｊ−２における走行車線が１つ右の車線にずれていたと判断する。
【０１０４】
なお、その後、地点Ｐｊにおいて、車両が左側に車線変更した場合、位置同定手段１２は、走行軌跡の幅員方向の変化量が、車線の幅員に所定の係数を乗じた幅以上になると、車両が車線変更したと判断し、車線Ｌａｎｅ２を車両が走行している確率を１００％に設定し、その他の車線を車両が走行している確率を０％に設定する。
【０１０５】
以上のように、この実施の形態４によれば、走行軌跡に基づいて車両の車線変更を検出し、そのときに、各車線について車両の存在確率を設定し直し、その存在確率に応じて走行車線を同定するようにしたので、誤った車線が走行車線と同定された場合において、車両の車線変更を利用して、適切な走行車線を同定することができるという効果が得られる。
【０１０６】
実施の形態５．
この発明の実施の形態５による車両位置同定装置は、実施の形態１の位置同定手段１２を変更したものであるので、位置同定手段１２について説明し、その他の構成要素の説明を省略する。
【０１０７】
実施の形態５による車両位置同定装置においては、位置同定手段１２は、車両が交差点で右折または左折した場合、方位計測手段３の例えばジャイロにより計測された車両の角速度ω（ｔ）［ｒａｄ／ｓｅｃ］および距離計測手段４により計測されたパルス信号に基づく車両の速度Ｖ（ｔ）［ｍ／ｓｅｃ］に基づいて右折または左折中の旋回半径Ｒ（ｔ）［ｍ］（＝Ｖ（ｔ）／ω（ｔ））を計算し、さらに、旋回半径Ｒ（ｔ）の平均値Ｒを計算し、その平均値Ｒに所定の係数を乗じた値に、路側から仮想車線リンクまでの距離が最も近い車線を、走行車線と同定する。なお、以下、旋回半径Ｒ（ｔ）の平均値Ｒを単に旋回半径という。
【０１０８】
次に動作について説明する。
図２６は、実施の形態５による車両位置同定装置により、車両が交差点で左折した場合における走行車線の同定の一例を示す図である。図２６（ａ）に示すように、旋回半径Ｒに所定の係数を乗じた値が最も左側の車線Ｌａｎｅ１の幅員より小さいので、位置同定手段１２は、走行車線を車線Ｌａｎｅ１と同定する。また、図２６（ｂ）に示すように、旋回半径Ｒに所定の係数を乗じた値が最も左側の車線Ｌａｎｅ１の幅員以上であるので、位置同定手段１２は、走行車線を車線Ｌａｎｅ２と同定する。すなわち、位置同定手段１２は、路側から仮想車線リンクまでの距離が、旋回半径Ｒに所定の係数を乗じた値に近い車線を走行車線と同定する。
【０１０９】
以上のように、この実施の形態５によれば、旋回時の車両の角速度と速度から計算される旋回半径に基づいて、右折または左折後の走行車線を同定するようにしたので、特に幅員の大きな道路から右折または左折したときの、走行車線の同定の精度を上げることができるという効果が得られる。
【０１１０】
実施の形態６．
この発明の実施の形態６による車両位置同定装置は、実施の形態１の位置同定手段１２を変更したものであるので、位置同定手段１２について説明し、その他の構成要素の説明を省略する。
【０１１１】
実施の形態６による車両位置同定装置においては、位置同定手段１２は、車両が交差点で右折または左折した場合、右折または左折前の道路に関する道路データを参照して走行可能な車線のうちの最も旋回方向寄りの車線（例えば、右折の場合、走行可能な車線のうちの最も右側の車線）から右折または左折したときの位置に、走行軌跡を平行移動させて車両の現在位置を修正する。
【０１１２】
次に動作について説明する。
図２７は、実施の形態６による車両位置同定装置により、車両が交差点で右折または左折した場合における走行車線の同定の一例を示す図である。図２７（ａ）に示すように、上下線が分離された３車線の道路６７の最も左側の車線Ｌａｎｅ１から片側２車線ずつの４車線の道路６８に車両が右折した場合、位置同定手段１２は、車両の位置を、地点Ｐｉから、道路６７の最も右側の車線Ｌａｎｅ３から右折した場合に車両が位置する地点Ｐｉａに修正する。なお、位置同定手段１２は、走行軌跡を平行移動させることにより地点Ｐｉａを算出する。
【０１１３】
また、図２７（ｂ）に示すように、車両が道路６７の最も右側の車線Ｌａｎｅ３から道路６８に左折した場合、位置同定手段１２は、車両の位置を、地点Ｐｊから、道路６７の最も左側の車線Ｌａｎｅ１から左折した場合に車両が位置する地点Ｐｊａに修正する。なお、位置同定手段１２は、走行軌跡を平行移動させることにより地点Ｐｊａを算出する。
【０１１４】
以上のように、この実施の形態６によれば、車両が交差点で右折または左折した場合、右折または左折前の道路に関する道路データを参照して走行可能な車線のうちの最も旋回方向寄りの車線から右折または左折したときの位置に、走行軌跡を平行移動させて車両の現在位置を修正するので、運転者が慣例的に選択する車線が、走行車線に同定されるようになり、即ち、右折または左折前の道路における慣例的に右左折に適した車線が走行車線と判断され、右折または左折後における走行車線の同定の処理を簡単にすることができるという効果が得られる。
【０１１５】
実施の形態７．
この発明の実施の形態７による車両位置同定装置は、実施の形態１の道路網記憶手段１を変更したものであるので、道路網記憶手段１について説明し、その他の構成要素の説明を省略する。
【０１１６】
実施の形態７による車両位置同定装置においては、道路網記憶手段１は、実施の形態１のものとは異なる道路データを記憶する。この道路網記憶手段１に記憶される道路データは、始点ノード、終点ノード、道路の幅員、道路の車線数および通行条件で構成される。
【０１１７】
なお、この実施の形態７においては、始点ノードから終点ノードへの方向は常に通行可能であるものとし、通行条件は「一方通行」または「対面通行」に設定される。「一方通行」が設定されている場合、その道路は、始点ノードから終点ノードへの方向の通行だけが可能であることが表され、「対面通行」が設定されている場合、その道路は、始点ノードと終点ノードとの間の両方向の通行が少なくとも１つの車線を有していることが表されている。
【０１１８】
次に動作について説明する。
図２８は、道路データから算出される仮想車線リンクを示す図である。
【０１１９】
車線算出手段１１は、道路網記憶手段１に記憶されている上述の道路データを読み出し、道路の幅員を車線数で等分割して仮想車線リンクを生成する。
【０１２０】
図２８（ａ）に示すリンクが、図２８（ｂ）に示すように、始点ノードがＮＳｉであり、終点ノードがＮＤｉであり、幅員がＷｉであり、車線数が１であり、通行条件が「対面通行」である道路データを有する場合、車線算出手段１１は、仮想始点ノードと仮想終点ノードを、道路データの始点ノードと終点ノードと同一にそれぞれ設定し、幅員をＷｉに設定し、通行条件を「対面通行」に設定した仮想車線リンクを生成する。
【０１２１】
図２８（ａ）に示すリンクが、図２８（ｃ）に示すように、始点ノードがＮＳｉであり、終点ノードがＮＤｉであり、幅員がＷｉであり、車線数が１であり、通行条件が「一方通行」である道路データを有する場合、車線算出手段１１は、仮想始点ノードと仮想終点ノードを、道路データの始点ノードと終点ノードと同一にそれぞれ設定し、幅員をＷｉに設定し、通行条件を「一方通行」に設定した仮想車線リンクを生成する。
【０１２２】
図２８（ａ）に示すリンクが、図２８（ｄ）に示すように、始点ノードがＮＳｉであり、終点ノードがＮＤｉであり、幅員がＷｉであり、車線数が２であり、通行条件が「対面通行」である道路データを有する場合、車線算出手段１１は、仮想始点ノードから仮想終点ノードへの方向を、道路データの始点ノードから終点ノードへの方向として、始点ノードと一方の路側との中間に仮想始点ノードを配置し、終点ノードと一方の路側との中間に仮想終点ノードを配置し、幅員をＷｉ／２に設定し、通行条件を「一方通行」に設定した第１の仮想車線リンクと、仮想始点ノードから仮想終点ノードへの方向を、道路データの終点ノードから始点ノードへの方向として、終点ノードと他方の路側との中間に仮想始点ノードを配置し、始点ノードと他方の路側との中間に仮想終点ノードを配置し、幅員をＷｉ／２に設定し、通行条件を「一方通行」に設定した第２の仮想車線リンクを生成する。
【０１２３】
図２８（ａ）に示すリンクが、図２８（ｅ）に示すように、始点ノードがＮＳｉであり、終点ノードがＮＤｉであり、幅員がＷｉであり、車線数が２であり、通行条件が「一方通行」である道路データを有する場合、車線算出手段１１は、仮想始点ノードから仮想終点ノードへの方向を、道路データの始点ノードから終点ノードへの方向として、始点ノードと一方の路側との中間に仮想始点ノードを配置し、終点ノードと一方の路側との中間に仮想終点ノードを配置し、幅員をＷｉ／２に設定し、通行条件を「一方通行」に設定した第１の仮想車線リンクと、仮想始点ノードから仮想終点ノードへの方向を、道路データの始点ノードから終点ノードへの方向として、始点ノードと他方の路側との中間に仮想始点ノードを配置し、終点ノードと他方の路側との中間に仮想終点ノードを配置し、幅員をＷｉ／２に設定し、通行条件を「一方通行」に設定した第２の仮想車線リンクを生成する。
【０１２４】
なお、上述のように、リンクの道路データの車線数が２以上である場合には、仮想車線リンクのデータの通行条件は、すべて「一方通行」に設定される。また、図中の「ＯＫ」と矢印は、その矢印の方向への通行が可能であることを表している。
【０１２５】
このように、車両位置同定装置で使用する道路データは、実施の形態１による車両位置同定装置で使用する道路データに限定されるものではなく、この実施の形態７による車両位置同定装置において使用されたものや、その他の形式のものを使用してもよい。ただし、その場合、車線算出手段１１は、その道路データの形式に対応して、その道路データから仮想車線リンクを生成するものに変更する。
【０１２６】
実施の形態８．
実施の形態１または実施の形態７による車両位置同定装置においては、位置同定手段１２は道路データの幅員を車線数で等分割して車線の幅を計算しているが、道路網記憶手段１に、道路データの一部として車線毎の車線幅を予め記憶するようにしてもよい。また、上り線と下り線との車線数が異なる道路では、上り線および下り線毎に道路データを設けるようにしてもよい。
【０１２７】
実施の形態９．
実施の形態１による車両位置同定装置においては、交通の規制方向が順方向または逆方向のいずれか１つであるとしているが、道路網記憶手段１に、道路データの一部として車線毎の規制方向を記憶したり、上り線および下り線の通行方向毎の規制方向を記憶するようにして、それらのデータに基づいて車線の規制方向を判断するようにしてもよい。
【０１２８】
実施の形態１０．
実施の形態３による車両位置同定装置においては、位置同定手段１２は、走行車線の同定を行う場合にヒステリシス幅を利用しているが、所定の車線における車両の存在確率が、所定時間が経過する間にあるいは所定距離を移動する間に継続して最大である場合、その車線を走行車線と同定するようにしてもよい。
【０１２９】
実施の形態１１．
実施の形態４による車両位置同定装置においては、位置同定手段１２は、各車線での車両の存在確率を式（２）により計算しているが、車線変更した方向にある車線での車両の存在確率が大きくなるものであれば、式（２）の代わりに他の計算式を使用してもよい。また、位置同定手段１２は、車線変更した方向にある車線での車両の存在確率が大きくなるように車線変更の度に定数を設定するようにしてもよい。
【０１３０】
また、位置同定手段１２は、車両の走行軌跡の変化から車線変更を検知して、そのときに、各車線での車両の存在確率を計算しているが、車両の進行方位が変化したときに、その変化した方向にある車線での車両の存在確率を増加させるようにしてもよい。さらに、運転者により車両が操作された方向を表す信号（左折あるいは右折信号）を車両に設けられている方向指示器から検出し、その方向にある車線での車両の存在確率を増加させるようにしてもよい。
【０１３１】
実施の形態１２．
実施の形態５による車両位置同定装置においては、位置同定手段１２は、交差点での右折または左折時における旋回半径を速度Ｖ（ｔ）を角速度ω（ｔ）で除した値Ｒ（ｔ）の平均値として計算しているが、右折または左折時における３箇所での車両の位置に基づいて、その３箇所を通る円の半径を旋回半径として算出するようにしてもよい。また、他の計算式で旋回半径を計算するようにしてもよい。
【０１３２】
さらに、位置同定手段１２は、旋回半径に所定係数を乗じた値と車線の幅員とを比較して走行車線を同定しているが、旋回半径に所定係数を乗じた値に基づいて各車線での車両の存在確率を計算し、各車線での車両の存在確率の大小関係から走行車線を同定するようにしてもよい。
【０１３３】
実施の形態１３．
実施の形態６による車両位置同定装置においては、位置同定手段１２は、車両が交差点で右折または左折する場合、右折または左折後の走行車線として、慣例的に選択されやすい車線を選択しているが、交差点での右左折前後や直進前後の走行車線の関係を表す情報を道路網記憶手段１に記憶しておき、位置同定手段１２が、その情報に基づいて走行車線および車両の現在位置を同定したり、あるいは、交差点での車両の右左折時において慣例的に選択されやすい車線ほど存在確率が大きくなるように車線毎に存在確率を設定して各車線での車両の存在確率の大小関係から走行車線を同定するようにしてもよい。
【０１３４】
【発明の効果】
以上のように、この発明によれば、車両の現在位置を計測する現在位置計測手段と、時間の経過とともに計測された現在位置の履歴を車両の走行軌跡として記憶する走行軌跡記憶手段と、道路の交差点あるいは折曲点等の位置を示すノードと、ノード間を直線で接続するリンクとにより構成された所定の範囲の道路網のリンク毎に、道路の形状、道路の位置、道路の幅員、車線数および車両の通行方向を、そのリンクに対応する道路の属性として記憶する道路網記憶手段と、その道路網のリンクおよび走行軌跡に基づいて、車両が走行している道路を特定する走行道路特定手段と、特定された道路の属性を道路網記憶手段から読み出し、その道路の属性から、その道路を構成する車線毎の属性として、車線の位置、車線の幅員、および車両の通行方向を算出する車線算出手段と、その車線の属性および走行軌跡に基づいて車線毎に車両が走行している確率を計算し、その確率に基づいて車両が走行している走行車線を同定する位置同定手段とを備えたので、道路の幅員方向における車両位置を同定することができる効果がある。
【０１３５】
この発明によれば、道路網記憶手段に、ノードとリンクとにより構成された所定の範囲の第１の道路網のリンク毎に、そのリンクに対応する道路の属性を記憶するとともに、所定の範囲と同一の範囲における、ノードとリンクより多いノードとリンクにより構成された第２の道路網のリンク毎に、そのリンクと対応する道路の属性として記憶し、走行道路特定手段で、第１の道路網のリンクと対応する道路の属性を使用し、車線算出手段および位置同定手段で、第２の道路網のリンクに対応する道路の属性を使用するように構成したので、より適切な車線を走行車線と同定することができる効果がある。
【０１３６】
この発明によれば、位置同定手段により、前回同定された走行車線とは別の車線を車両が走行している確率が、前回同定された走行車線を車両が走行している確率より所定の値以上高くなったときに、別の車線を車両の走行車線と同定するように構成したので、走行車線を誤って同定する可能性を低減することができる効果がある。
【０１３７】
この発明によれば、位置同定手段により、道路の幅員方向の走行軌跡の変化量に基づいて車両が車線変更したことを検知し、車線変更を検知したときに、車線毎に車両が走行している確率を設定し直すように構成したので、誤った車線が走行車線と同定された場合において、車両の車線変更を利用して、適切な走行車線を同定することができる効果がある。
【０１３８】
この発明によれば、位置同定手段により、車両の進行方位の変化あるいは走行軌跡から車両の右折または左折を検知し、右折または左折を検知した場合、右折または左折時の車両の進行方位の変化量あるいは走行軌跡から車両の旋回半径を計算し、その旋回半径に基づいて、右折または左折後の道路の車線毎に車両が走行している確率を計算し、その確率に基づいて車両が走行している走行車線を同定するように構成したので、特に幅員の大きな道路から右折または左折したときの、走行車線の同定の精度を上げることができる効果がある。
【０１３９】
この発明によれば、位置同定手段により、車両の進行方位の変化あるいは走行軌跡から車両の右折または左折を検知し、右折または左折を検知した場合、右折または左折前の道路における慣例的に右左折に適した車線を判断して、右折または左折後の道路の車線毎に車両が走行している確率を計算し、その確率に基づいて車両が走行している走行車線を同定するように構成したので、運転者が慣例的に選択する車線が、走行車線に同定されるようになり、右折または左折後における走行車線の同定の処理を簡単にすることができる効果がある。
【図面の簡単な説明】
【図１】この発明の実施の形態１による車両位置同定装置の構成を示すブロック図である。
【図２】実際の通過コースおよび計測された走行軌跡の一例を示す図である。
【図３】走行軌跡記憶手段の動作を説明するフローチャートである。
【図４】車両が位置Ｐｉに位置する場合における走行軌跡記憶手段に記憶される走行軌跡を説明する図である。
【図５】位置候補設定手段により設定された２つの位置候補ａ，ｂを示す図である。
【図６】（ａ）は図５の候補ａの通過コースを示し、（ｂ）は順番に記憶されたその通過コースに対応する道路データを示す図である。
【図７】（ａ）は図５の候補ｂの通過コースを示し、（ｂ）は順番に記憶されたその通過コースに対応する道路データを示す図である。
【図８】通過コースを特定する動作を説明する図である。
【図９】（ａ）は走行軌跡を修正するときの動作を説明する図であり、（ｂ）は修正された走行軌跡を表示したときの表示手段の表示画面を示す図である。
【図１０】車両が位置Ｐｊに位置する場合における走行軌跡記憶手段に記憶された走行軌跡を示す図である。
【図１１】通過コースを特定する動作を説明する図である。
【図１２】通過コースに基づく車両の走行軌跡と修正後の走行軌跡の一例を示す図である。
【図１３】誤った通過コース候補を選択した場合における修正後の走行軌跡の一例を示す図である。
【図１４】道路データから計算される仮想車線リンクのデータを説明する図である。
【図１５】仮想車線リンクと車両の走行軌跡に基づいて車両が走行している車線を同定するときの動作を説明する図である。
【図１６】仮想車線リンクと車両の走行軌跡に基づいて車両が走行している車線を同定するときの動作を説明する図である。
【図１７】図１５に示す各車線についての車両の存在確率を示す図である。
【図１８】図１５に示す各車線についての車両の存在確率を示す図である。
【図１９】この発明の実施の形態２による詳細な道路網への切り替えの例を説明する図である。
【図２０】詳細な道路網への切り替えの例を説明する図である。
【図２１】詳細な道路網の道路データを有する範囲の他の設定について説明する図である。
【図２２】詳細な道路網の道路データを有する範囲のさらに他の設定について説明する図である。
【図２３】この発明の実施の形態３による車両が車線変更したときの位置同定手段の動作の一例を説明する図である。
【図２４】この発明の実施の形態４による車両が車線変更したときの位置同定手段の動作の一例を説明する図である。
【図２５】車両が交差点で左折したときの位置同定手段の動作の一例を説明する図である。
【図２６】この発明の実施の形態５による車両位置同定装置により、車両が交差点で左折した場合における走行車線の同定の一例を示す図である。
【図２７】この発明の実施の形態６による車両位置同定装置により、車両が交差点で右折または左折した場合における走行車線の同定の一例を示す図である。
【図２８】この発明の実施の形態７による道路データから算出される仮想車線リンクを示す図である。
【図２９】従来の車両位置同定装置の構成を示すブロック図である。
【図３０】従来の車両位置同定装置におけるパターンマッチング手段の動作を説明する図である。
【図３１】従来の車両位置同定装置における投影マッチング手段の動作を説明する図である。
【符号の説明】
１ 道路網記憶手段、３ 方位計測手段、５ 現在位置計測手段、６ 走行軌跡記憶手段、１０ 走行道路特定手段、１１ 車線算出手段、１２ 位置同定手段、２３ 走行軌跡。[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a vehicle position identification device that identifies a current position of a running vehicle on a road, and more particularly to a vehicle position identification device that identifies a traveling lane as a position in a width direction of a road.
[0002]
[Prior art]
FIG. 29 is a block diagram showing a configuration of a conventional vehicle position identification device described in, for example, JP-A-8-159788. In FIG. 29, reference numeral 1 denotes a node indicating a position such as a road intersection or a turning point; Road network storage means for storing data such as links connecting nodes in a straight line (collectively referred to as road data), and 2 receives a radio wave from a GPS (Global Positioning System) satellite and A GPS receiver 3 for measuring the position and outputting the position data of the vehicle, and 3 for measuring the traveling direction of the vehicle based on, for example, the angular velocity of the vehicle detected by a gyro (not shown) and outputting the direction as direction data 4 is a pulse signal output from a distance sensor (not shown) according to the traveling distance of the vehicle, for example, and The release is measured, the distance measuring means for outputting the distance data.
[0003]
5 calculates the current position of the vehicle in longitude and latitude based on the azimuth data from the azimuth measuring means 3 and the distance data from the distance measuring means 4, and uses the position data from the GPS receiver 2 to calculate the current position of the vehicle. Current position measuring means 81 for correcting the position as appropriate and outputting the current position of the vehicle as current position data. Numeral 81 indicates a predetermined signal when the distance data from the distance measuring means 4 is integrated and the traveling distance reaches 150 meters. Is a 150-meter determining means that outputs the data and resets the integrated value to zero, and then restarts the integration, and 82 integrates the distance data from the distance measuring means 4 and, when the running distance reaches 30 meters, a predetermined value. This is a 30-meter determination unit that outputs a signal and restarts integration after resetting the integrated value to zero.
[0004]
Reference numeral 6 denotes a traveling locus storage unit that stores the traveling locus of the vehicle based on the direction data from the direction measuring unit 3 and the current position data from the current position measuring unit 5. The traveling locus storage means 6 generates a traveling vector having a size of 15 and an orientation of the orientation measuring means 3 every time the vehicle travels 15 meters, and sequentially stores the generated traveling vectors.
[0005]
7 reads the road data from the road network storage unit 1 as appropriate, and when the vehicle passes through the intersection, the road data read from the road network storage unit 1 and the azimuth data from the azimuth measurement unit 3, the distance data from the distance measurement unit 4 Position candidate setting means for setting the position candidates of the vehicles in the extending directions of all the roads extending from the intersection every time the vehicle travels for 15 meters, and 8 is supplied from the position candidate setting means 7. This is a traveling road shape generation unit that generates road shape vectors corresponding to the position candidates and sequentially stores the generated road shape vectors.
[0006]
Each time a predetermined signal is supplied from the 150-meter determining means 81, the reference 91 compares the traveling locus from the traveling locus storage means 6 with the road shape vector from the traveling road shape generating means 8, and correlates most with the traveling locus. Is a pattern matching unit that selects a road shape vector having a high road shape and corrects the current position of the vehicle to a position candidate point on the road corresponding to the selected road shape vector. Each time it is supplied, the current position data from the current position measuring means 5 and each position candidate from the position candidate setting means 7 are sequentially compared, and a position candidate having a distance close to the current position of the vehicle and a short traveling direction is selected. Then, the current position of the vehicle is corrected to the position of the selected position candidate, and the traveling locus stored in the traveling locus storage unit 6 is modified. That is a projection matching means.
[0007]
101 reads out map data around the current position of the vehicle from the road network storage unit 1 in accordance with the current position data from the current position measurement unit 5, and displays a map around the current position together with a vehicle mark indicating the current position of the vehicle. This is a display control means for displaying on the display means 102.
[0008]
Next, the operation will be described.
FIG. 30 is a diagram for explaining the operation of the pattern matching unit 91, and FIG. 31 is a diagram for explaining the operation of the projection matching unit 92.
[0009]
Each time a predetermined signal is supplied from the 150-meter determining means 81, the pattern matching means 91 converts the travel locus stored in the travel locus storage means 6 and each road shape vector from the travel road shape generating means 8 into Read only 2 kilometers.
[0010]
Next, the pattern matching unit 91 calculates the correlation between each road shape vector and the traveling locus, and selects the road shape vector having the highest correlation. It is determined that the route of the road corresponding to the selected road shape vector is the actually traveling route, and the current position is corrected to the position candidate position on the road corresponding to the road shape vector, and the traveling locus is determined. The traveling locus from the storage means 6 is corrected.
[0011]
At this time, the pattern matching means 91 removes, from the road shape vectors, those having no correlation between the road shape vector and the travel locus, and deletes the position candidates on the road corresponding to the road shape vector.
[0012]
For example, in FIG. 30, the route indicated by the broken line is a travel locus from the time when the vehicle departed from the departure point Ps to the time when the vehicle traveled 150 meters, and the triangular marks on each road are position candidates. In this case, the current position P0 is corrected to the position of the position candidate P0-1 on the road corresponding to the road shape vector having the highest correlation with the travel locus among these position candidates.
[0013]
On the other hand, when the vehicle has traveled 30 meters after departure from the departure point Ps, the projection matching means 92 selects one of the position candidates P2-1, P2-2, and P2-3 set on the road near the current position P2. Then, P2-3, which is the position candidate closest to the current position P2 and in the direction of travel, is selected, the current position is corrected to that position, and the running locus stored in the running locus storage means 6 is corrected. Thereafter, similarly, every time the vehicle travels 30 meters, the current position is corrected to position candidates P3, P4, and P5, and the traveling locus is similarly corrected.
[0014]
Then, the display control means 101 reads map data around the current position of the vehicle from the road network storage means 1 in accordance with the current position data from the current position measurement means 5, displays a map around the current position, The vehicle mark indicating the current position of the vehicle is displayed on the display at the position on the map corresponding to the current position appropriately corrected by the pattern matching unit 91 and the projection matching unit 92 as described above.
[0015]
[Problems to be solved by the invention]
Since the conventional vehicle position identification device is configured as described above, the position of the vehicle in the width direction of the road is not identified, and the position candidate is set at a preset position (for example, the center of the road). Was. Therefore, when the current position of the vehicle is corrected to the selected position candidate, the position of the vehicle is corrected in the center of the road in the width direction, and it is difficult to determine the lane in which the vehicle is traveling, and for example, When a vehicle makes a right turn or a left turn from a road having a large width such as an arterial road in a large city, there is a problem that an error in the vehicle position occurs at most almost only the width of the road.
[0016]
The present invention has been made to solve the above-described problem, and has as its object to provide a vehicle position identification device that can more accurately identify a vehicle position and a travel locus in a width direction of a road.
[0017]
It is another object of the present invention to provide a vehicle position identification device capable of more accurately identifying a vehicle position and a traveling locus in the width direction of a road when changing lanes or turning right or left at an intersection.
[0018]
[Means for Solving the Problems]
A vehicle position identification device according to the present invention includes a current position measurement unit that measures a current position of a vehicle, and a travel path storage that stores a history of the current position measured over time by the current position measurement unit as a travel path of the vehicle. Means, a node representing a position such as an intersection or a bending point of the road, and a link of a road network within a predetermined range formed by a link connecting the nodes with a straight line, the shape of the road, the position of the road, A road network storage unit that stores the width of the road, the number of lanes, and a traffic direction of the vehicle as an attribute of a road corresponding to the link; a road network link stored in the road network storage unit; Based on the stored traveling trajectory, the traveling road identification means for identifying the road on which the vehicle is traveling and the attributes of the road identified by the traveling road identification means are read from the road network storage means. Lane calculating means for calculating a lane position, a lane width, and a traveling direction of a vehicle as attributes for each lane constituting the road, and an attribute of the lane calculated by the lane calculating means. And a position identification unit that calculates a probability that the vehicle is traveling for each lane based on the traveling locus stored in the traveling locus storage unit, and identifies a traveling lane in which the vehicle is traveling based on the probability. It is provided with.
[0019]
The vehicle position identification device according to the present invention stores, for each link of a first road network in a predetermined range formed by nodes and links, an attribute of a road corresponding to the link in a road network storage unit. For each link of the second road network constituted by more nodes and links than the nodes and links in the same range as the predetermined range, the link is stored as the attribute of the road corresponding to the link, The attribute of the road corresponding to the link of the first road network is used, and the attribute of the road corresponding to the link of the second road network is used by the lane calculating means and the position identifying means.
[0020]
In the vehicle position identification device according to the present invention, the probability that the vehicle is traveling in a lane different from the previously identified traveling lane by the position identification means is the probability that the vehicle is traveling in the previously identified traveling lane. When it becomes higher than a predetermined value, another lane is identified as the traveling lane of the vehicle.
[0021]
The vehicle position identification device according to the present invention, by the position identification means, detects that the vehicle has changed lanes based on the amount of change in the traveling locus in the width direction of the road, and when the lane change is detected, the vehicle is identified for each lane. Is to reset the probability that the vehicle is traveling.
[0022]
The vehicle position identification device according to the present invention includes a direction measurement unit that measures a change in the traveling direction of the vehicle. The position identification unit detects a right or left turn of the vehicle from a change in the traveling direction of the vehicle or a traveling locus, and turns right. Or, when a left turn is detected, the turning radius of the vehicle is calculated from the change amount of the heading or the traveling locus of the vehicle when turning right or left, and based on the turning radius, the vehicle is turned for each lane of the road after the right or left turn. The running probability is calculated, and the traveling lane in which the vehicle is running is identified based on the calculated probability.
[0023]
The vehicle position identification device according to the present invention includes a direction measurement unit that measures a change in the traveling direction of the vehicle. The position identification unit detects a right or left turn of the vehicle from a change in the traveling direction of the vehicle or a traveling locus, and turns right. Or, when a left turn is detected, it is customary to determine a suitable lane on the road before the right or left turn, and calculate the probability that the vehicle is running for each lane on the road after the right or left turn. The lane in which the vehicle is traveling is identified based on the probability.
[0024]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of the present invention will be described.
Embodiment 1 FIG.
FIG. 1 is a block diagram showing a configuration of a vehicle position identification device according to Embodiment 1 of the present invention. In the figure, reference numeral 1 denotes a road network storage means for storing nodes indicating positions such as intersections or bends of roads, links connecting the nodes with straight lines, and road data such as road width and road attributes such as one-way traffic. Reference numeral 2 denotes a GPS receiver which receives a radio wave from a GPS satellite via the antenna 2-1, measures a current position based on the received radio wave, and outputs position data of the vehicle, and 3 is, for example, not shown. An azimuth measuring means for measuring the traveling azimuth of the vehicle based on the angular velocity of the vehicle detected by the gyro, and outputting the azimuth data. Reference numeral 4 denotes a pulse signal output from a distance sensor (not shown) according to the traveling distance of the vehicle. This is a distance measuring unit that counts and measures the traveling distance of the vehicle, and outputs it as distance data.
[0025]
Reference numeral 5 denotes a current position of the vehicle which is calculated based on the direction data from the direction measuring means 3, the distance data from the distance measuring means 4 and the position data from the GPS receiver 2 so as to represent the current position of the vehicle in longitude and latitude. Is output as current position data.
[0026]
Reference numeral 6 denotes a traveling locus storage means for storing the current position data from the current position measuring means 5 as a traveling locus of the vehicle every time the vehicle travels a predetermined distance or more or changes the course by a predetermined angle or more.
[0027]
Reference numeral 7 reads the travel locus from the travel locus storage means 6, reads the road data from the road network storage means 1 as appropriate, compares the travel locus with the road data, and searches for a plurality of courses in a road network similar to the travel locus. Position candidate setting means for setting the tip as a position candidate of the vehicle.
[0028]
Reference numeral 9 records the history of the road data (start point node, end point node, width, number of lanes, vertical separation, regulation direction) of the course through which each position candidate passes, and stores the candidates of the passing course corresponding to each position candidate. A passing course candidate storage unit 10 selects a candidate determined as an actual vehicle passing course from a plurality of passing course candidates stored in the passing course candidate storage unit 9, and selects a vehicle on the passing course. Is a traveling road identification unit that identifies a road link portion where is located as a traveling road.
[0029]
Reference numeral 11 denotes a lane calculating unit that reads road data of the traveling road identified by the traveling road identifying unit 10 from the road network storage unit 1 and calculates a virtual lane link corresponding to each lane of the road based on the attribute of the road. And 12 calculates the existence probabilities of the vehicles for each virtual lane link based on the virtual lane link calculated by the lane calculation means 11 and the traveling locus stored in the traveling locus storage means 6, respectively. The position identification means identifies the lane corresponding to the virtual lane link having the highest existence probability as the traveling lane, and identifies the current position of the vehicle in the traveling lane.
[0030]
Reference numeral 13 denotes display means for superimposing and displaying the current position of the vehicle identified by the position identification means 12 on the road data of the road network storage means 1.
[0031]
Next, the operation will be described.
FIG. 2 is a diagram illustrating an example of an actual passage course and a measured traveling locus.
[0032]
FIG. 3 is a flowchart for explaining the operation of the traveling locus storage means 6, FIG. 4 is a diagram for explaining the traveling locus stored in the traveling locus storage means 6 when the vehicle is located at the position Pi, and FIG. FIG. 4 is a diagram showing two position candidates a and b set by a position candidate setting unit 7.
[0033]
FIG. 6 is a diagram showing a passing course (FIG. 6A) of the candidate a in FIG. 5 and road data (FIG. 6B) stored in order and corresponding to the passing course. FIG. 7B is a diagram showing a passing course (FIG. 7A) of the candidate b in FIG. 5 and road data (FIG. 7B) stored in order and corresponding to the passing course.
[0034]
FIG. 8 is a diagram for explaining the operation for specifying the passing course, and FIG. 9 is a diagram for explaining the operation when the traveling locus is corrected (FIG. 9A) and the diagram for when the corrected traveling locus is displayed. FIG. 10 is a diagram showing a display screen of a display unit 13 (FIG. 9B).
[0035]
FIG. 10 is a diagram showing a running locus stored in the running locus storage means 6 when the vehicle is located at the position Pj, FIG. 11 is a diagram for explaining an operation for specifying a passing course, and FIG. FIG. 13 is a diagram illustrating an example of a traveling locus of a vehicle based on the vehicle and a modified traveling locus, and FIG. 13 is a diagram illustrating an example of a modified traveling locus when an incorrect passing course candidate is selected.
[0036]
FIG. 14 is a diagram illustrating virtual lane link data calculated from road data. FIGS. 15 and 16 identify a lane in which the vehicle is traveling based on the calculated virtual lane link and the traveling locus of the vehicle. FIG. 17 and FIG. 18 are diagrams illustrating the existence probability of the vehicle in each lane illustrated in FIG. 15.
[0037]
In FIG. 2, reference numeral 21 denotes a departure point (double circle in the figure), reference numeral 22 denotes an actual passage course of the vehicle, and reference numeral 23 denotes a measured traveling locus. All circles in the figure are nodes, and links are those connecting the nodes with straight lines. In the following, the operation at that time will be described by taking as an example the case of the actual passing course 22 and the running locus 23 as shown in FIG.
First, the operation of the traveling locus storage means 6 will be described with reference to the flowchart of FIG.
[0038]
In step ST1, the traveling locus storage means 6 stores the traveling distance from the time when the vehicle position (current position of the vehicle) supplied from the current position measuring means 5 is stored as the past position of the traveling locus (past vehicle position). It is determined whether or not the travel distance is equal to or longer than the predetermined distance Lv1, and if it is determined that the travel distance is equal to or longer than the predetermined distance Lv1, the process proceeds to step ST2, and the current position of the vehicle at that time is set to one of the past positions of the travel locus. Remember as one.
[0039]
On the other hand, when it is determined that the traveling distance is not longer than the predetermined distance Lv1, the process proceeds to step ST3. In step ST3, the traveling locus storage means 6 determines that the traveling distance from the time when the current position of the vehicle is stored as one of the past positions is the predetermined distance Lv2 or more and the traveling direction has changed by the predetermined angle αv1 or more. If it is determined that the travel distance is equal to or greater than the predetermined distance Lv2 and the traveling direction has changed by the predetermined angle αv1 or more, the process proceeds to step ST2, and the current position of the vehicle at that time is determined by the travel locus. Is stored as one of the past positions.
[0040]
On the other hand, if it is determined in step ST3 that the traveling distance is not longer than the predetermined distance Lv2 or that the traveling direction has not changed by the predetermined angle αv1 or more, the process proceeds to step ST4. In step ST4, the traveling locus storage means 6 stores the current position of the vehicle as the current position in the traveling locus. After the process of step ST2 ends, the process of step ST4 is executed. In other words, the traveling locus is composed of a past position sequentially stored and a current position indicating the vehicle position at that time.
[0041]
For example, when the traveling trajectory from the departure point PR0 in FIG. 4A to the current position Pi is measured, as shown in FIG. 4B, PR0 to PR6 are sequentially stored as past positions, and Pi is stored. Is stored as the current position. The past positions PR0 to PR6 and the current position Pi are stored in the traveling locus storage means 6 in the order shown in FIG.
[0042]
As described above, the traveling locus storage means 6 stores the vehicle position at each time point as the past position or the current position of the traveling locus.
[0043]
Next, the position candidate setting unit 7 reads out the past position and the current position of the traveling locus from the traveling locus storing unit 6 and reads out the road data from the road network storing unit 1 to convert the road data having high similarity to the traveling locus. Set position candidates.
[0044]
For example, for the traveling locus shown in FIG. 4B, as shown in FIG. 5, there is road data that substantially overlaps with the traveling locus. Therefore, a position candidate a is set on the road data. Then, on a nearby road, a position candidate b is set on road data having the second similarity to the travel locus. Therefore, in this case, the correlation between the candidate a and the travel locus is higher than the correlation between the candidate b and the travel locus.
[0045]
Then, the passing course candidate storage unit 9 receives, from the position candidate setting unit 7, road data of a predetermined road (that is, a link) constituting a course having a high similarity to the traveling locus, and stores the road data as the passing course candidate. Are sequentially stored.
[0046]
For example, as shown in FIG. 5, when a candidate a and a candidate b are set, as shown in FIG. 6A, road data of links L1, L2, and L3, which are road data corresponding to the candidate a, As the data of the passing course candidate a, it is additionally stored in order as shown in FIG. 6B as the vehicle travels. As shown in FIG. 7A, the road data of the links L10, L11, L12, and L13, which are the road data corresponding to the candidate b, are used as the data of the passage course candidate b, as the vehicle travels. Additional storage is performed in order as shown in FIG. As shown in FIG. 6B and FIG. 7B, the road data of each link includes nodes (node numbers) at both ends of the link, the width of the actual road corresponding to the link, the number of lanes, The information includes information indicating whether the road is vertically separated (information indicating whether the road is only an up line or a down line) and a regulation direction (a traffic direction regulated by a road rule or the like).
[0047]
Next, the traveling road identifying means 10 reads the road data of the above-mentioned passing course candidates from the passing course candidate storage means 9 and, based on the road data, actual road course candidates as shown in FIG. A template that is substantially similar to the shape of the road is generated, and the past position and the current position of the traveling locus are read from the traveling locus storage unit 6 to determine whether the past position and the current position fall within the template. If it is determined that the past position and the current position fall within the template, it is determined that the similarity between the template and the running locus is high. Further, the traveling road identification unit 10 compares the shape of the traveling locus with the template shape for the course change point, the length of the straight traveling section, the traveling direction, and the like, and determines the most similar one of all the templates. The selected passing course candidate corresponding to the selected template is specified as an actual passing course.
[0048]
For example, when passing course candidates a and b shown in FIG. 6 and FIG. 7 are stored in the passing course candidate storage means 9 as the passing course candidates, the traveling road specifying means 10 Templates 41a and 41b as shown in FIG. 8 are generated from the data, respectively, and compared with the data of the traveling locus read out from the traveling locus storage means 6 to determine the similarity. (The area indicated by the broken line in FIG. 6A). In the case of the candidates a and b in FIG. 8, since the candidate a is closer to the template shape than the candidate b with respect to the course change point and the length of the straight section, it is determined that the similarity of the candidate a is high. You.
[0049]
After specifying the passing course in this way, the traveling road identification means 10 translates or rotates the past position or the current position of the traveling locus so that the vehicle mark indicating the current position is displayed on the road. Let it. For example, in the case of the candidate a shown in FIG. 8, as shown in FIG. 9A, the traveling road identification unit 10 corrects PR5, PR6, and Pi so that PR5, PR6, and Pi fall within the template 41a. The amount (moving amount) is calculated, and PR5, PR6, and Pi are moved by the corrected amount. By doing so, the vehicle mark is displayed on the road in the road map displayed by the display unit 13 as shown in FIG. 9B. After the point in time when this Pi is measured, the passing course is specified appropriately using the correction amount of the traveling locus and the traveling direction.
[0050]
Further, as shown in FIG. 10 (a), when the vehicle further travels a predetermined distance and travels to, for example, a point Pj, the vehicle travels to a point Pj in FIG. 10 (c). The past positions PR0 to PR10 shown and data Pi of the current position are stored. At this time, as shown in FIG. 11, the traveling road identifying means 10 determines that the previously executed modification is successful and the identified traveling path 43 is within the template 44, as shown in FIG. It is determined that the course is correct, and thereafter, as shown in FIG. 12, the past position and the current position of the traveling locus corrected in the template 44 are stored in the traveling locus storage means 6.
[0051]
When the passing course candidate b is erroneously selected as the actual passing course, the similarity between the traveling path 42 and the template 44b corresponding to the passing course candidate b is low as shown in FIG. The current position of the vehicle deviates from the candidate b for the passing course (point I1 in the figure). As described above, when the vehicle deviates from the passing course, and when the similarity between the template of another candidate and the traveling locus 42 is high, the candidate is selected.
[0052]
In FIG. 13, the vehicle has deviated from the candidate b of the passing course after making a right turn, but at this time, a link 45 extending in the same direction as the direction of the vehicle after turning near the right turn point on the traveling locus. If there is, the link 45 is selected as the passing course candidate b. However, thereafter, as in the case of the point I2, the candidate b of the passing course cannot be generated, and the candidate b and the traveling trajectory 42 do not completely coincide with each other. Similarity is reduced, and an appropriate candidate with high similarity is selected.
[0053]
Next, the lane calculating means 11 obtains the road data (start point node, end point node, width, number of lanes, presence / absence of separation of upper and lower lines, regulation direction) of each link of the specified passing course from the traveling road specifying means 10. Upon reception, the width of the road data is equally divided by the number of lanes, and a virtual lane link set at the center of the equally divided road (lane) is generated. The road data of the virtual lane link includes a virtual start node, a virtual end node, a width of the lane, and a regulation direction of the lane located at the center of the lane.
[0054]
For example, as shown in FIG. 14A, road data of a link whose start point node is NSi and whose end point node is NDi, as shown in FIG. 14B, whose width is Wi and the number of lanes is 2, If there is no separation of the up and down lines and the regulation direction is the opposite direction, it is determined that the road of this link is a two-way road with one up lane and one down lane. Therefore, an intermediate point between the original link start node NSi and one road side of the road is a virtual start node VNSi-1, and an intermediate point between the original link end node NDi and one road side of the road is a virtual end node VNDi. The first virtual lane link 51b whose width is Wi / 2 and the regulation direction is the reverse direction, and the intermediate point between the start node NSi of the original link and the other road side of the road are the virtual start node VNSi-2, and the intermediate point between the end node NDi of the original link and the other road side of the road is the virtual end node VNDi-2. A virtual lane link 52b is generated.
[0055]
In addition, as shown in FIG. 14C, in the case of a link having a width of Wi, two lanes, separation of upper and lower lines, and a restriction direction in the opposite direction, the road of this link is one side. It is determined that the road is a traffic road and has two lanes, and that the direction from the end node to the start node is restricted. Therefore, an intermediate point between the original link start node NSi and one road side of the road is a virtual start node VNSi-1, and an intermediate point between the original link end node NDi and one road side of the road is a virtual end node VNDi. A first virtual lane link 51c whose width is Wi / 2 and whose regulation direction is the reverse direction, and an intermediate point between the start point node NSi of the original link and the other road side of the road are assumed to be virtual start point nodes. VNSi-2, the intermediate point between the end node NDi of the original link and the other road side of the road is the virtual end node VNDi-2, and the width is Wi / 2 and the regulation direction is the second direction. The virtual lane link 52c is generated.
[0056]
Further, as shown in FIG. 14D, in the case of a link having a width of Wi, two lanes, a separation of upper and lower lines, and a regulation direction of a forward direction, the road of this link has one side. It is determined that the road is a traffic road and has two lanes, and that the direction from the start node to the end node is restricted. Therefore, an intermediate point between the original link start node NSi and one road side of the road is a virtual start node VNSi-1, and an intermediate point between the original link end node NDi and one road side of the road is a virtual end node VNDi. The first virtual lane link 51d whose width is Wi / 2 and whose regulation direction is the forward direction, and the intermediate point between the start point node NSi of the original link and the other road side of the road are assumed to be virtual start point nodes. VNSi-2, the middle point between the end point node NDi of the original link and the other road side of the road is the virtual end point node VNDi-2, the width is Wi / 2, and the second restriction direction is the forward direction. A virtual lane link 52d is generated.
[0057]
Further, as shown in FIG. 14E, in the case of a link having a width of Wi, a number of lanes of one, no separation of upper and lower lines, and no regulation direction, a start node NSi of the original link is set as a virtual start point. A first virtual lane link 53 having a width of Wi and no regulation direction is generated as a node, and an end point node NDi of the original link as a virtual end point node. That is, a one-lane road indicated by the road link is used both up and down.
[0058]
The road data of the virtual lane link generated in this way is supplied to the position identification unit 12.
[0059]
Next, the position identification means 12 reads the data of the traveling locus from the traveling locus storage means 6 and, based on the road data of the virtual lane link from the lane calculating means 11, the distance between the traveling locus and each virtual lane link in the width direction. Is calculated, and the probability that the vehicle exists in the lane corresponding to the virtual lane link is calculated from the distances, and the lane with the highest probability is identified as the lane in which the vehicle is traveling.
[0060]
When the distance between the traveling locus and the j-th virtual lane link VLj in the width direction is ΔWj and the number of lanes of the road on which the vehicle is traveling (that is, the number of virtual lane links) is N, The probability kj that the vehicle is traveling in the lane corresponding to the virtual lane link VLj is calculated by equation (1). According to this equation (1), kN is calculated from the probability k1 that the vehicle is traveling in the lanes corresponding to the first to Nth virtual lane links VL1 to VLN. However, the probability that the vehicle is traveling in a lane having a traffic direction opposite to the direction of the traveling locus is set to 0%.
(Equation 1)

[0061]
For example, as shown in FIG. 15A, the current position of the vehicle is a point Pj, the start node of the link L6 corresponding to the point Pj is N5, the end node is N6, and the width is 10 meters. Yes, the number of lanes is 4, the upper and lower lines are not separated, and the regulation direction is the opposite direction, that is, the road corresponding to the link L6 is a 10-meter wide two-lane two-sided facing road. In this case, the position identification unit 12 generates four virtual lane links VL1 to VL4 corresponding to the lanes Lane1 to Lane4 having a width of 2.5 meters shown in FIG. 15B, as shown in FIGS. 16 and 17.
[0062]
At this time, among the four lanes Lane1 to Lane4, the lanes Lane3 and Lane4 have a traffic direction (unregulated direction) opposite to the traveling direction of the vehicle, and therefore the probability that the vehicle is traveling in the lane Lane3. Both k3 and the probability k4 that the vehicle is traveling in the lane Lane4 are set to 0%.
[0063]
Next, for example, when the traveling locus passes 0.75 meters from the boundary between the lane Lane1 and the lane Lane2 to the road side, ΔW1 becomes 0.5 (= 1.25-0.75) meters. ΔW2 is 2.0 (= 1.25 + 0.75) meters, and as shown in FIG. 17, the position identification means 12 calculates the probability k1 that the vehicle is traveling in the lane Lane1 according to the equation (1). , 80% (= (1−0.5 / 2.5) × 100%), and the probability k2 that the vehicle is traveling in the lane Lane2 is calculated as 20% (= (1−2.0 / 2. 5) x 100%). When the road corresponding to the link L6 and the traveling locus are parallel, as shown in FIG. 18, since ΔW1 and ΔW2 are constant, k1 and k2 also remain constant.
[0064]
Therefore, the position identification means 12 identifies the lane Lane1 which is the lane corresponding to the highest probability k1 among the probabilities k1 to k4 corresponding to the four lanes as the lane in which the vehicle is traveling.
[0065]
As described above, according to the first embodiment, the virtual lane link corresponding to each lane is generated from the road data, and the vehicle travels in each lane from the virtual lane link data and the travel locus. Since the running lane is identified based on the calculated probability (existence probability), the vehicle position in the width direction of the road can be identified.
[0066]
The traveling lane identified in this way may be notified to the user. At this time, the existence probability of the vehicle in each lane may be displayed.
[0067]
Further, route identification, route guidance, and the like may be performed in consideration of the traveling lane by using the above-described identification of the traveling lane.
[0068]
Further, the above-mentioned existence probability is calculated according to the equation (1), but the equation for deriving the existence probability is not particularly limited to this equation, and the existence probability of the vehicle in each lane can be appropriately derived. If it is an expression, another expression may be used.
[0069]
Embodiment 2 FIG.
The vehicle position identification device according to the second embodiment of the present invention is a modification of the road network storage unit 1 and the lane calculation unit 11 of the first embodiment, and therefore will be described, and description of other components will be omitted. I do.
[0070]
In the vehicle position identification device according to the second embodiment, the road network storage unit 1 stores detailed road data in addition to the road data of the road network (first road network) stored by the road network storage unit 1 of the first embodiment. The road data of the link corresponding to the network (the second road network) is stored, and the lane calculating means 11 determines whether or not the road data of the detailed road network relating to the vicinity of the current position of the vehicle is stored, and appropriately. The road data of a link corresponding to a detailed road network is read, and a virtual lane link is calculated based on the road data.
[0071]
Next, the operation will be described.
FIGS. 19 and 20 are diagrams illustrating an example of switching to a detailed road network. Here, the operation will be described with reference to the examples of FIGS. 19 and 20.
[0072]
The road network storage unit 1 corresponds to the road data of the links Lb1 to Lb4 (FIG. 19B) corresponding to the road shown in FIG. 19A and the detailed road network in the range 61 in FIG. 19B. Road data of links Lc1 to Lc10 (FIG. 19C) is stored. The range 61 is set by storing the latitude and longitude of each vertex in the road network storage unit 1 in advance.
[0073]
The lane calculating means 11 determines whether or not the current position of the vehicle is within a range (for example, the range 61) in which the road data of the detailed road network is stored, based on the information on the current position from the current position measuring means 5. Is determined, and if it is determined that the current position is within such a range, the link in the range 61 is changed to a link corresponding to a detailed road network.
[0074]
For example, when the vehicle is located at the point Pi-N on the road shown in FIG. 19B, the link, that is, the road network is not changed because the current position (point Pi-N) is outside the range 61. In this case, a virtual lane link is calculated based on the road data of the link Lb1.
[0075]
On the other hand, when the vehicle is located at the point Pi, since the current position (point Pi) is within the range 61, the links Lb1 to Lb4 are changed to the links Lc1 to Lc10 corresponding to the detailed road map shown in FIG. Be changed.
[0076]
Then, the lane calculating means 11 calculates a virtual lane link based on the road data of the link (in this case, the link Lc9) corresponding to the current position among the links Lc1 to Lc10. For example, when the vehicle is located at the point Pi shown in FIGS. 19B and 19C, the road data of the link Lc9 corresponding to the detailed road map (width = W2, number of lanes = 1, vertical separation = Yes) ) Is calculated based on the virtual lane link.
[0077]
The virtual lane link calculated by the lane calculating unit 11 is supplied to the position identifying unit 12, and the position of the vehicle is identified based on the virtual lane link.
[0078]
By switching the road network in this manner, in the range 61, instead of the links Lb1 to Lb4, the virtual lane links are calculated based on the links Lc1 to Lc10 corresponding to the detailed road map. It is determined that the vehicle is continuously traveling on the road as shown in FIG. 19C without being determined that the vehicle has left the road as shown in FIG.
[0079]
For example, when traveling on the road shown in FIG. 20A, the road network storage unit 1 stores road data of links Lb1 to Lb3 (FIG. 20B) corresponding to the road shown in FIG. Assuming that road data of links Lc1 to Lc5 (FIG. 20C) corresponding to the detailed road map of the range 62 in FIG. 20B is stored, when the vehicle is traveling outside the range 62, The lane calculating means 11 calculates a virtual lane link based on the link Lb1 or the link Lb3, but when the vehicle is traveling in the range 62, the link Lb1 to Lb3 is changed to a link corresponding to a detailed road map. Switching from Lc1 to Lc5, a virtual lane link is calculated based on any of the links Lc1 to Lc5 (in this case, link Lc3).
[0080]
By switching the road map in this manner, within the range 62, the virtual lane link is calculated based on the links Lc1 to Lc15 corresponding to the detailed road map, so that the vehicle is switched as shown in FIG. 20C, it is determined that the vehicle is continuously traveling on one lane without being determined to be out of the lane.
[0081]
As described above, according to the second embodiment, a plurality of road networks of different scales are stored for the same range, and the driving lane is identified using the road data of the detailed road network. Therefore, an effect is obtained that a more appropriate lane can be identified as a traveling lane.
[0082]
In the second embodiment, the ranges 61 and 62 are set by storing the latitude and longitude of each vertex in the road network storage unit 1 in advance. The ranges 61 and 62 may be set using the converted coordinate information. In that case, after converting the current position of the vehicle into normalized coordinate values, it is determined whether or not the current position is within the ranges 61 and 62.
[0083]
Further, as shown in FIG. 21, the ranges 61 and 62 may be set in advance corresponding to the shape of the link 64. By setting the ranges 61 and 62 in association with the link 64 as described above, when the link corresponding to the road on which the vehicle is traveling is specified, it is determined that the vehicle is located within the ranges 61 and 62.
[0084]
Further, as shown in FIG. 22, the ranges 61 and 62 are set so that the outer edge passes through the node 65, and the links 61 and 62 in the ranges 61 and 62 are linked to the links Lb5 to Lb8, and the details of the ranges 61 and 62 are described. The links corresponding to the road network are stored, and when the links on which the vehicle is traveling are links Lb5 to Lb8 in the ranges 61 and 62, the links Lb5 to Lb8 are changed to links corresponding to the detailed road network. You may make it.
[0085]
Embodiment 3 FIG.
Since the vehicle position identification device according to the third embodiment of the present invention is a modification of the position identification device 12 of the first embodiment, the position identification device 12 will be described, and the description of other components will be omitted.
[0086]
In the vehicle position identification device according to the third embodiment, when the vehicle changes lanes, the position identification means 12 determines the above-described probability ki with respect to the change-source lane (virtual lane link VLi) and the change-destination lane (virtual lane link VLj). ) Are compared, and when the probability kj for the lane of the change destination becomes larger than the sum of the above-mentioned probability ki and the predetermined hysteresis width Δk for the lane of the change source (kj> (ki + Δk)), It is determined that the vehicle is traveling in the lane. That is, at the point in time when the probability kj for the destination lane is equal to or less than the sum of the probability ki for the original lane and the predetermined hysteresis width Δk, the position identification unit 12 determines that the vehicle is traveling in the source lane. to decide.
[0087]
Next, the operation will be described.
FIG. 23 is a diagram illustrating an example of the operation of the position identification unit 12 when the vehicle changes lanes. Here, the operation will be described with reference to FIG.
[0088]
The road shown in FIG. 23A is a two-lane one-way road having a width of W. Then, it is assumed that the vehicle first travels on the lane Lane1 as indicated by the travel locus 73, starts changing lanes at the point 71, ends the lane change at the point 72, and then travels on the lane Lane2.
[0089]
At this time, the distance between the virtual lane link 74 of the lane Lane1 and the travel locus 73 and the distance between the virtual lane link 75 of the lane Lane2 and the travel locus 73 change as shown in FIG. And the probability that the vehicle is traveling in the lane Lane2 is calculated as shown in FIG. 23C according to the equation (1).
[0090]
At this time, the existence probabilities of the vehicles with respect to the lanes Lane1 and Lane2 change as shown in FIG. 23C, and coincide at time ti-c. Thereafter, the magnitude relation of the existence probabilities is reversed, and the existence probability of the vehicle with respect to the lane Lane2 increases, and at time ti-b, the existence probability of the vehicle with respect to the lane Lane2 and the existence probability of the vehicle with respect to the lane Lane1. Is larger than a predetermined hysteresis width.
[0091]
The position identification unit 12 identifies the traveling lane as the lane Lane1 in the same manner as in the first embodiment until the time ti-c, and thereafter, the existence probability of the vehicle with respect to the lane Lane2 and the existence of the vehicle with respect to the lane Lane1. Until time ti-b at which the difference from the probability is larger than the predetermined hysteresis width, as shown in FIG. 23D, the traveling lane is identified as lane Lane1, and after time ti-b, the traveling lane is identified as lane Lane. Lane2 is identified.
[0092]
As described above, according to the third embodiment, even when the vehicle temporarily runs out of the adjacent lane, such as when avoiding a parked vehicle, the existence of the vehicle in another lane is larger than a predetermined hysteresis width. Until the probability increases, the traveling lane to be identified is not changed, so that the effect of reducing the possibility of erroneously identifying the traveling lane can be obtained.
[0093]
Embodiment 4 FIG.
Since the vehicle position identification device according to the fourth embodiment of the present invention is a modification of the position identification device 12 of the first embodiment, the position identification device 12 will be described, and the description of other components will be omitted.
[0094]
In the vehicle position identification device according to the fourth embodiment, position identification means 12 performs lane change when the amount of change in the width direction of the traveling locus is equal to or greater than the width obtained by multiplying the width of the lane by a predetermined coefficient. Judge that
[0095]
Further, when the vehicle turns right or left at the intersection, the position identifying means 12 sequentially drives the vehicle along the lane corresponding to the virtual road link according to the equation (2) in order from the virtual road link having the shortest distance to the travel locus. Is calculated, and the lane corresponding to the virtual road link having the largest probability is identified as the traveling lane.
(Equation 2)

[0096]
Next, the operation will be described.
FIG. 24 is a diagram illustrating an example of the operation of the position identification unit 12 when the vehicle changes lanes, and FIG. 25 is a diagram illustrating an example of the operation of the position identification unit 12 when the vehicle turns left at an intersection. It is. Here, the operation will be described with reference to FIGS. 24 and 25.
[0097]
As shown in FIG. 24, when the vehicle moves from the point Pi-c to the right point Pi-a and changes lanes, the position identifying unit 12 determines that the amount of change in the width direction of the traveling locus is a predetermined value for the width of the lane. It is determined that the traveling lane has been changed to the lane Lane2 at the time point ti-b when the vehicle has passed the point Pi-b at which the value is multiplied by the coefficient or more.
[0098]
Further, as shown in FIG. 25A, for example, when the vehicle turns left at an intersection and enters a three-lane road where the upper and lower lines are separated, and then changes lanes at points Pi-1 and Pi, First, the position identification means 12 identifies the traveling lane as the lane Lane1 in the vicinity of the point Pi-2 according to the probability (FIG. 25B) calculated according to the equation (2). At this time, the probability that the vehicle is traveling in lane Lane1 is 50%, the probability that the vehicle is traveling in lane Lane2 is 33%, and the probability that the vehicle is traveling in lane Lane3 is 17%. .
[0099]
Next, when the change amount in the width direction of the traveling locus at the point Pi-1 is equal to or greater than the width obtained by multiplying the width of the lane by a predetermined coefficient, the position identification unit 12 determines that the vehicle has changed lanes to the right. . After that, the position identification unit 12 sets the probability that the vehicle is traveling on the leftmost lane Lane1 to 0%, and then determines the probability that the vehicle is traveling on the lane Lane2 and the lane Lane3. Calculate the probability that the vehicle is running. At this time, the probability that the vehicle is traveling in lane Lane2 is 67%, and the probability that the vehicle is traveling in lane Lane3 is 33%. In this case, since the probability that the vehicle is traveling in the lane Lane1 is set to 0%, the position identification unit 12 calculates the probability with the number of lanes N being 2.
[0100]
Further, when the amount of change in the width direction of the traveling locus at the point Pi is equal to or larger than the width obtained by multiplying the width of the lane by a predetermined coefficient, the position identification unit 12 determines that the vehicle has changed lane further to the right side. I do. After that, the position identification unit 12 sets the probability that the vehicle is traveling in the lane Lane2 on the left side of the lane Lane1 to 0%. Therefore, the probability that the vehicle is traveling in the lane Lane3 is 100%.
[0101]
On the other hand, as shown in FIG. 25C, when it is erroneously determined that the traveling lane after the left turn is the lane Lane2, the position identification unit 12 calculates the probability calculated at the point Pj-3 according to the equation (2). According to (FIG. 25D), the traveling lane is identified as the lane Lane2. At this time, when the probabilities are calculated in the order of lanes closest to the travel locus, the probability that the vehicle is traveling in lane Lane2 is 50%, the probability that the vehicle is traveling in lane Lane1 is 33%, and the probability that the vehicle is traveling in lane Lane1 is lane Lane3. Is 17%. Note that the order in which the probability is calculated may be a lane order in which the vehicle often travels by convention.
[0102]
Next, at the point Pj-2, when the amount of change in the width direction of the traveling trajectory becomes equal to or greater than the width obtained by multiplying the width of the lane by a predetermined coefficient, the position identification unit 12 determines that the vehicle has changed lane to the right. . After that, the position identification unit 12 sets the probability that the vehicle is traveling on the leftmost lane Lane1 to 0%, and then determines the probability that the vehicle is traveling on the lane Lane2 and the lane Lane3. Calculate the probability that the vehicle is running. The position identification unit 12 determines that the vehicle has changed lanes from the lane Lane2 to the right based on the traveling locus, and calculates the probability in the order of the lane Lane3 of the change destination and the original lane Lane2. At this time, the probability that the vehicle is traveling in lane Lane3 is 67%, and the probability that the vehicle is traveling in lane Lane2 is 33%. In this case, since the probability that the vehicle is traveling in the lane Lane1 is set to 0%, the position identification unit 12 calculates the probability with the number of lanes N being 2.
[0103]
Further, at the point Pj-1, similarly, when the amount of change in the width direction of the traveling locus becomes equal to or larger than the width obtained by multiplying the width of the lane by a predetermined coefficient, the position identification unit 12 further changes the lane to the right side. However, since there is no lane to the right of lane Lane3, the probability that the vehicle is traveling in lane Lane3 is set to 100%, and the probability that the vehicle is traveling in other lanes is set to 0%. I do. That is, in this case, the position identification unit 12 identifies the traveling lane as the lane Lane3, and determines that the traveling lanes at the points Pj-3 and Pj-2 have shifted to the right lane.
[0104]
After that, when the vehicle changes lane to the left at the point Pj, the position identification unit 12 determines that the vehicle has moved when the amount of change in the width direction of the traveling locus is equal to or greater than the width obtained by multiplying the width of the lane by a predetermined coefficient. It is determined that the lane has been changed, and the probability that the vehicle is traveling in lane Lane2 is set to 100%, and the probability that the vehicle is traveling in other lanes is set to 0%.
[0105]
As described above, according to the fourth embodiment, the lane change of the vehicle is detected based on the traveling locus, and at that time, the existence probability of the vehicle is reset for each lane, and the traveling is performed according to the existence probability. Since the lane is identified, it is possible to obtain an effect that, when the wrong lane is identified as the traveling lane, an appropriate traveling lane can be identified by using the lane change of the vehicle.
[0106]
Embodiment 5 FIG.
Since the vehicle position identification device according to the fifth embodiment of the present invention is a modification of the position identification device 12 of the first embodiment, the position identification device 12 will be described, and description of other components will be omitted.
[0107]
In the vehicle position identification device according to the fifth embodiment, when the vehicle turns right or left at an intersection, the position identification means 12 determines the angular velocity ω (t) [rad / sec of the vehicle measured by the azimuth measurement means 3 using a gyro, for example. ] And a turning radius R (t) [m] (= V (t) / during a right or left turn based on the vehicle speed V (t) [m / sec] based on the pulse signal measured by the distance measuring means 4. ω (t)), an average value R of the turning radius R (t) is calculated, and the distance from the road side to the virtual lane link is closest to a value obtained by multiplying the average value R by a predetermined coefficient. The lane is identified as the driving lane. Hereinafter, the average value R of the turning radius R (t) is simply referred to as the turning radius.
[0108]
Next, the operation will be described.
FIG. 26 is a diagram illustrating an example of identification of a traveling lane when the vehicle makes a left turn at an intersection by the vehicle position identification device according to the fifth embodiment. As shown in FIG. 26A, since the value obtained by multiplying the turning radius R by a predetermined coefficient is smaller than the width of the leftmost lane Lane1, the position identification unit 12 identifies the traveling lane as the lane Lane1. Further, as shown in FIG. 26 (b), the value obtained by multiplying the turning radius R by a predetermined coefficient is equal to or larger than the width of the leftmost lane Lane1, so that the position identifying means 12 identifies the traveling lane as the lane Lane2. . That is, the position identification unit 12 identifies a lane whose traveling distance from the road side to the virtual lane link is close to a value obtained by multiplying the turning radius R by a predetermined coefficient.
[0109]
As described above, according to the fifth embodiment, the traveling lane after turning right or left is identified based on the turning radius calculated from the angular velocity and the speed of the vehicle at the time of turning. The effect is obtained that the accuracy of identifying the traveling lane when turning right or left from a large road can be improved.
[0110]
Embodiment 6 FIG.
Since the vehicle position identification device according to the sixth embodiment of the present invention is a modification of the position identification device 12 of the first embodiment, only the position identification device 12 will be described, and description of other components will be omitted.
[0111]
In the vehicle position identification device according to the sixth embodiment, when the vehicle makes a right or left turn at an intersection, the position identification means 12 refers to the road data on the road before the right or left turn to turn the most of the lanes that can be traveled. The current position of the vehicle is corrected by translating the traveling locus to a position where a right turn or a left turn is made from a lane closer to the direction (for example, in the case of a right turn, the rightmost lane of the lanes that can travel).
[0112]
Next, the operation will be described.
FIG. 27 is a diagram illustrating an example of identification of a traveling lane when the vehicle turns right or left at an intersection by the vehicle position identification device according to the sixth embodiment. As shown in FIG. 27 (a), when the vehicle turns right from the leftmost lane Lane1 of a three-lane road 67 where the upper and lower lanes are separated to a four-lane road 68 with two lanes on each side, the position identifying means 12 The position of the vehicle is corrected from the point Pi to the point Pia where the vehicle is located when the vehicle turns right from the rightmost lane Lane3 of the road 67. Note that the position identification unit 12 calculates the point Pia by translating the traveling locus in parallel.
[0113]
In addition, as shown in FIG. 27B, when the vehicle turns left from the rightmost lane Lane3 of the road 67 to the road 68, the position identification unit 12 determines the position of the vehicle from the point Pj to the leftmost of the road 67. Is corrected to the point Pja where the vehicle is located when the vehicle turns left from the lane Lane1. In addition, the position identification means 12 calculates the point Pja by translating the traveling locus in parallel.
[0114]
As described above, according to the sixth embodiment, when the vehicle makes a right or left turn at an intersection, the lane closest to the turning direction among the lanes that can be run with reference to the road data on the road before the right or left turn Since the current position of the vehicle is corrected by translating the traveling trajectory to the position at which the vehicle turns right or left from, the lane that the driver routinely selects becomes identified as the traveling lane, that is, the right turn Alternatively, the lane on the road before the left turn that is conventionally suited to the right or left turn is determined as the travel lane, and the effect of identifying the travel lane after the right or left turn can be simplified is obtained.
[0115]
Embodiment 7 FIG.
Since the vehicle position identification device according to the seventh embodiment of the present invention is a modification of the road network storage unit 1 of the first embodiment, only the road network storage unit 1 will be described, and the description of other components will be omitted. .
[0116]
In the vehicle position identification device according to the seventh embodiment, the road network storage unit 1 stores road data different from that of the first embodiment. The road data stored in the road network storage unit 1 includes a start point node, an end point node, a width of a road, the number of lanes of the road, and traffic conditions.
[0117]
In the seventh embodiment, it is assumed that the direction from the start node to the end node is always passable, and the pass condition is set to “one-way” or “face-to-face”. If "one-way" is set, it indicates that the road can only travel in the direction from the start node to the end node. If "two-way" is set, the road is It is shown that traffic in both directions between the start node and the end node has at least one lane.
[0118]
Next, the operation will be described.
FIG. 28 is a diagram illustrating virtual lane links calculated from road data.
[0119]
The lane calculating means 11 reads the above-mentioned road data stored in the road network storing means 1 and equally divides the width of the road by the number of lanes to generate a virtual lane link.
[0120]
As shown in FIG. 28B, the link shown in FIG. 28A is such that the start node is NSi, the end node is NDi, the width is Wi, the number of lanes is 1, and the traffic condition is If the road data is “face-to-face traffic”, the lane calculating means 11 sets the virtual start point node and the virtual end point node to be the same as the start point node and the end point node of the road data, sets the width to Wi, and sets the traffic width to Wi. A virtual lane link in which the condition is set to “face-to-face traffic” is generated.
[0121]
As shown in FIG. 28 (c), the link shown in FIG. 28 (a) is such that the start node is NSi, the end node is NDi, the width is Wi, the number of lanes is 1, and the traffic condition is If the road data is “one-way”, the lane calculating means 11 sets the virtual start node and the virtual end node to be the same as the start node and the end node of the road data, sets the width to Wi, and sets the width to Wi. A virtual lane link with the condition set to “one-way” is generated.
[0122]
As shown in FIG. 28 (d), the link shown in FIG. 28 (a) is such that the start node is NSi, the end node is NDi, the width is Wi, the number of lanes is 2, and the traffic condition is When there is road data that is “face-to-face traffic”, the lane calculation unit 11 sets the direction from the virtual start node to the virtual end node as the direction from the start node to the end node of the road data, and A first virtual node in which a virtual start node is arranged in the middle of the route, a virtual end node is arranged in the middle between the end node and one roadside, the width is set to Wi / 2, and the traffic condition is set to "one-way" The lane link and the direction from the virtual start node to the virtual end node are set as the direction from the end node to the start node of the road data, and the virtual start node is arranged between the end node and the other road side. Square of placing the virtual end-point node in the middle of the roadside, to set the width to Wi / 2, to generate a second virtual lane link with the traffic condition in the "one way".
[0123]
As shown in FIG. 28 (e), the link shown in FIG. 28 (a) is such that the start node is NSi, the end node is NDi, the width is Wi, the number of lanes is 2, and the traffic condition is When there is road data that is “one-way”, the lane calculating means 11 sets the direction from the virtual start node to the virtual end node as the direction from the start node to the end node of the road data, and A first virtual node in which a virtual start node is arranged in the middle of the route, a virtual end node is arranged in the middle between the end node and one roadside, the width is set to Wi / 2, and the traffic condition is set to "one-way" The lane link and the direction from the virtual start node to the virtual end node are set as the direction from the start node to the end node of the road data, and the virtual start node is arranged between the start node and the other road side. Square of placing the virtual end-point node in the middle of the roadside, to set the width to Wi / 2, to generate a second virtual lane link with the traffic condition in the "one way".
[0124]
As described above, when the number of lanes in the road data of the link is two or more, all the traffic conditions of the virtual lane link data are set to “one-way”. Further, “OK” and the arrow in the figure indicate that the vehicle can pass in the direction of the arrow.
[0125]
As described above, the road data used in the vehicle position identification device is not limited to the road data used in the vehicle position identification device according to the first embodiment, but is used in the vehicle position identification device according to the seventh embodiment. Or other types may be used. However, in that case, the lane calculating means 11 changes to a method for generating a virtual lane link from the road data in accordance with the format of the road data.
[0126]
Embodiment 8 FIG.
In the vehicle position identification device according to the first or seventh embodiment, the position identification means 12 calculates the lane width by equally dividing the width of the road data by the number of lanes. Alternatively, the lane width for each lane may be stored in advance as a part of the road data. In addition, on a road having different numbers of lanes between an up line and a down line, road data may be provided for each of the up line and the down line.
[0127]
Embodiment 9 FIG.
In the vehicle position identification device according to the first embodiment, it is assumed that the traffic regulation direction is one of the forward direction and the backward direction. However, the road network storage unit 1 stores the regulation for each lane as a part of the road data. The direction may be stored, or the restriction direction for each of the traveling directions of the up line and the down line may be stored, and the restriction direction of the lane may be determined based on the data.
[0128]
Embodiment 10 FIG.
In the vehicle position identification device according to the third embodiment, the position identification unit 12 uses the hysteresis width when identifying the traveling lane, but the existence probability of the vehicle in the predetermined lane elapses by a predetermined time. In the case where the maximum value is maintained continuously during the travel or a predetermined distance, the lane may be identified as the traveling lane.
[0129]
Embodiment 11 FIG.
In the vehicle position identification device according to the fourth embodiment, the position identification means 12 calculates the existence probability of the vehicle in each lane by the equation (2). As long as the probability increases, another calculation formula may be used instead of the formula (2). Further, the position identification unit 12 may set a constant every time the lane is changed so that the existence probability of the vehicle in the lane in the direction of the lane change is increased.
[0130]
Further, the position identification means 12 detects a lane change from a change in the traveling locus of the vehicle, and calculates the existence probability of the vehicle in each lane at that time, but when the traveling direction of the vehicle changes, Alternatively, the existence probability of the vehicle in the lane in the changed direction may be increased. Further, a signal indicating a direction in which the vehicle is operated by the driver (a left turn signal or a right turn signal) is detected from a direction indicator provided in the vehicle, and the existence probability of the vehicle in the lane in that direction is increased. You may.
[0131]
Embodiment 12 FIG.
In the vehicle position identification device according to the fifth embodiment, position identification means 12 calculates an average of values R (t) obtained by dividing speed V (t) by angular velocity ω (t) by a turning radius at the time of right or left turn at an intersection. Although calculated as a value, a radius of a circle passing through the three locations may be calculated as a turning radius based on the positions of the vehicle at the three locations when turning right or left. Further, the turning radius may be calculated by another calculation formula.
[0132]
Further, the position identification means 12 identifies the traveling lane by comparing the value obtained by multiplying the turning radius by a predetermined coefficient with the width of the lane, but in each lane based on the value obtained by multiplying the turning radius by the predetermined coefficient. May be calculated, and the traveling lane may be identified from the magnitude relationship between the vehicle existence probabilities in each lane.
[0133]
Embodiment 13 FIG.
In the vehicle position identification device according to the sixth embodiment, when the vehicle turns right or left at an intersection, position identification means 12 selects a lane that is conventionally easily selected as a traveling lane after right or left turn. The information indicating the relationship between the driving lanes before and after turning right and left at the intersection and before and after going straight ahead is stored in the road network storage means 1, and the position identification means 12 identifies the driving lane and the current position of the vehicle based on the information. Or, set the existence probability for each lane so that the lane that is customarily easier to select when turning right or left at the intersection is larger than the lane, and from the magnitude relationship between the existence probabilities of vehicles in each lane The driving lane may be identified.
[0134]
【The invention's effect】
As described above, according to the present invention, a current position measuring unit that measures a current position of a vehicle, a traveling locus storage unit that stores a history of the current position measured over time as a traveling locus of the vehicle, For each link of the road network within a predetermined range constituted by nodes indicating positions such as intersections or bending points and links connecting the nodes with straight lines, the shape of the road, the position of the road, the width of the road, Road network storage means for storing the number of lanes and the direction of travel of the vehicle as attributes of the road corresponding to the link; The identification means and the attribute of the identified road are read from the road network storage means, and from the attribute of the road, the attributes of each lane constituting the road include the position of the lane, the width of the lane, and the A lane calculating means for calculating a line direction, and calculating a probability that the vehicle is running for each lane based on the attribute and the running locus of the lane, and identifying a running lane on which the vehicle is running based on the probability. Since the position identification means is provided, the vehicle position in the width direction of the road can be identified.
[0135]
According to the present invention, for each link of the first road network in the predetermined range constituted by the nodes and the links, the attribute of the road corresponding to the link is stored in the road network storage means, For each link of the second road network composed of more nodes and links than the nodes and links in the same range as above, it is stored as the attribute of the road corresponding to the link, and the traveling road specifying means stores the first road Since the attribute of the road corresponding to the link of the second road network is used by the lane calculating means and the position identifying means using the attribute of the road corresponding to the link of the network, the vehicle travels in a more appropriate lane. There is an effect that can be identified as a lane.
[0136]
According to the present invention, the probability that the vehicle is traveling in a lane different from the previously identified traveling lane by the position identifying means is a predetermined value from the probability that the vehicle is traveling in the previously identified traveling lane. Since the configuration is such that another lane is identified as the traveling lane of the vehicle when the traffic lane becomes higher, there is an effect that the possibility of erroneously identifying the traveling lane can be reduced.
[0137]
According to the present invention, the position identifying means detects that the vehicle has changed lanes based on the amount of change in the travel locus in the width direction of the road, and when the lane change is detected, the vehicle travels for each lane. Since the configuration is made so as to reset the probability that the vehicle is running, when the wrong lane is identified as the traveling lane, there is an effect that an appropriate traveling lane can be identified using the lane change of the vehicle.
[0138]
According to the present invention, the position identification means detects a right turn or a left turn of the vehicle from a change in the heading of the vehicle or a traveling locus, and when a right turn or a left turn is detected, the amount of change in the heading of the vehicle when turning right or left. Alternatively, the turning radius of the vehicle is calculated from the traveling locus, and based on the turning radius, the probability that the vehicle is running for each lane of the road after turning right or left is calculated, and the vehicle is running based on the probability. Since the present embodiment is configured so as to identify the traveling lane, there is an effect that the accuracy of the identification of the traveling lane can be improved particularly when the vehicle turns right or left from a road having a large width.
[0139]
According to the present invention, the position identification means detects a right or left turn of the vehicle from a change in the heading direction or a traveling locus of the vehicle. It is configured to determine the lane suitable for the vehicle, calculate the probability that the vehicle is traveling for each lane of the road after turning right or left, and identify the traveling lane in which the vehicle is traveling based on the probability. Therefore, the lane conventionally selected by the driver is identified as the traveling lane, and there is an effect that the process of identifying the traveling lane after turning right or left can be simplified.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a vehicle position identification device according to a first embodiment of the present invention.
FIG. 2 is a diagram illustrating an example of an actual passage course and a measured traveling locus.
FIG. 3 is a flowchart illustrating an operation of a traveling locus storage unit.
FIG. 4 is a diagram illustrating a traveling locus stored in a traveling locus storage unit when the vehicle is located at a position Pi.
FIG. 5 is a diagram showing two position candidates a and b set by a position candidate setting unit.
6A is a diagram showing a passage course of a candidate a in FIG. 5, and FIG. 6B is a diagram showing road data corresponding to the passage course stored in order.
7A is a diagram showing a passage course of a candidate b in FIG. 5, and FIG. 7B is a diagram showing road data corresponding to the passage course stored in order.
FIG. 8 is a diagram illustrating an operation of specifying a passing course.
FIG. 9A is a diagram illustrating an operation when a traveling locus is corrected, and FIG. 9B is a diagram illustrating a display screen of a display unit when the corrected traveling locus is displayed.
FIG. 10 is a diagram illustrating a traveling locus stored in a traveling locus storage unit when the vehicle is located at a position Pj.
FIG. 11 is a diagram illustrating an operation of specifying a passing course.
FIG. 12 is a diagram illustrating an example of a traveling locus of a vehicle based on a passing course and a corrected traveling locus.
FIG. 13 is a diagram illustrating an example of a corrected traveling locus when an incorrect passing course candidate is selected.
FIG. 14 is a diagram illustrating virtual lane link data calculated from road data.
FIG. 15 is a diagram illustrating an operation when a lane in which a vehicle is traveling is identified based on a virtual lane link and a traveling locus of the vehicle.
FIG. 16 is a diagram illustrating an operation when a lane in which a vehicle is traveling is identified based on a virtual lane link and a traveling locus of the vehicle.
FIG. 17 is a diagram showing a vehicle existence probability for each lane shown in FIG. 15;
18 is a diagram showing a vehicle existence probability for each lane shown in FIG.
FIG. 19 is a diagram illustrating a detailed example of switching to a road network according to the second embodiment of the present invention;
FIG. 20 is a diagram illustrating a detailed example of switching to a road network.
FIG. 21 is a diagram illustrating another setting of a range having road data of a detailed road network.
FIG. 22 is a diagram illustrating still another setting of a range having road data of a detailed road network.
FIG. 23 is a diagram illustrating an example of an operation of the position identification unit when the vehicle changes lanes according to the third embodiment of the present invention.
FIG. 24 is a diagram for explaining an example of the operation of the position identification means when the vehicle changes lanes according to the fourth embodiment of the present invention.
FIG. 25 is a diagram illustrating an example of the operation of the position identification unit when the vehicle turns left at an intersection.
FIG. 26 is a diagram showing an example of identification of a traveling lane when a vehicle makes a left turn at an intersection by the vehicle position identification device according to the fifth embodiment of the present invention.
FIG. 27 is a diagram showing an example of identification of a traveling lane when a vehicle turns right or left at an intersection by the vehicle position identification device according to the sixth embodiment of the present invention.
FIG. 28 is a diagram showing virtual lane links calculated from road data according to the seventh embodiment of the present invention.
FIG. 29 is a block diagram showing a configuration of a conventional vehicle position identification device.
FIG. 30 is a diagram illustrating the operation of a pattern matching unit in a conventional vehicle position identification device.
FIG. 31 is a diagram illustrating the operation of a projection matching unit in a conventional vehicle position identification device.
[Explanation of symbols]
1 road network storage means, 3 azimuth measurement means, 5 current position measurement means, 6 travel trajectory storage means, 10 travel road identification means, 11 lane calculation means, 12 position identification means, 23 travel trajectory.

Claims (6)

Current position measuring means for measuring the current position of the vehicle,
Travel path storage means for storing the history of the current position measured over time by the current position measurement means as the travel path of the vehicle,
For each of the links of the road network within a predetermined range formed by nodes representing the positions of intersections or bending points of the road and links connecting the nodes with straight lines, the shape of the road, the position of the road, and the width of the road Road network storage means for storing the number of lanes and the traffic direction of the vehicle as attributes of a road network corresponding to the link;
Traveling road identification means for identifying a road on which the vehicle is traveling, based on the link of the road stored in the road network storage means, and a traveling locus stored in the traveling locus storage means,
The attribute of the road specified by the traveling road specifying unit is read out from the road network storage unit, and from the attribute of the road, the position of the center of the lane, the width of the lane, and the vehicle Lane calculating means for calculating a traffic direction;
The attribute of the lane calculated by the lane calculating means, and the probability that the vehicle is running for each lane is calculated based on the running locus stored in the running locus storage means, based on the probability, A vehicle position identification device comprising: a position identification unit that identifies a traveling lane on which the vehicle is traveling. The road network storage unit stores, for each link of the first road network in a predetermined range formed by nodes and links, the attribute of the road corresponding to the link, and stores the attribute in the same range as the predetermined range. Storing, for each link of the second road network constituted by more nodes and links than the nodes and links, as an attribute of a road corresponding to the link;
The driving road specifying means uses a road attribute corresponding to the link of the first road network,
The vehicle position identification device according to claim 1, wherein the lane calculation unit and the position identification unit use a road attribute corresponding to the link of the second road network. The position identification means, when the probability that the vehicle is traveling in a different lane from the previously identified traveling lane is higher than the probability that the vehicle is traveling in the previously identified traveling lane by a predetermined value or more. 2. The vehicle position identification device according to claim 1, wherein the another lane is identified as a traveling lane of the vehicle. The position identification means detects that the vehicle has changed lanes based on the amount of change in the running locus in the width direction of the road, and resets the probability that the vehicle is running for each lane when the lane change is detected. The vehicle position identification device according to claim 1, wherein: A direction measuring means for measuring a change in the traveling direction of the vehicle is provided,
The position identification means detects a right turn or a left turn of the vehicle from a change in the heading direction or a traveling locus of the vehicle measured by the heading measuring means, and detects a right turn or a left turn when detecting a right turn or a left turn. The turning radius of the vehicle is calculated from the change amount or the traveling locus, and based on the turning radius, the probability that the vehicle is running for each lane of the road after right or left turn is calculated, and the vehicle is calculated based on the probability. The vehicle position identification device according to claim 1, wherein the traveling lane on which the vehicle is traveling is identified. A direction measuring means for measuring a change in the traveling direction of the vehicle is provided,
The position identification means detects a right turn or a left turn of the vehicle from a change in the traveling direction of the vehicle measured by the azimuth measurement means or a traveling locus. A lane suitable for turning left or right is determined, a probability that the vehicle is traveling is calculated for each lane of the road after right or left turn, and a traveling lane in which the vehicle is traveling is identified based on the probability. The vehicle position identification device according to claim 1, wherein:

Images