JP3642931B2 - Paper sheet thickness detection apparatus and correction method thereof - Google Patents

Description

になる。
【００５４】
このことから、基準ローラ１１及び厚さ検出ローラ１２が各回転位置にあるときの偏心補正値Ｈ（ｋ）（ｋ＝１、２、…、ｎ）は次の式を用いて算出される。
Ｈ（ｋ）＝ｑ_PA−ｑ_PA（ｋ）
次に、前記構成の紙葉類厚さ検出装置によって紙葉類の厚さを検出する方法について説明する。
【００５５】
前記搬送系を駆動し、基準ローラ１１と厚さ検出ローラ１２との間に紙葉類を挿入する前に、ＣＰＵ２０は、ＤＡコンバータ１８に前記ゲイン補正値ｇｙをセットし、ＤＡコンバータ１６にセットするオフセット補正値を変更して、ＡＤコンバータ１９の出力Ｑを出力目標値Ｍにする。次に、ＣＰＵ２０は、基準ローラ１１及び厚さ検出ローラ１２が１回転するごとに前記出力信号ＳＧ１にパルスＰＬを検出し、各タイミングで偏心補正を行うことができるように準備する。
【００５６】
続いて、紙葉類が基準ローラ１１と厚さ検出ローラ１２との間に到達すると、タイミングｔにおけるＡＤコンバータ１８の出力Ｑの値ｑ（ｔ）（ｔ＝１、２、…、ｎ）に対する偏心補正を開始する。この場合、偏心補正が行われた後の値をｑ^* （ｔ）とすると、値ｑ^* （ｔ）（ｔ＝１、２、…、ｎ）を次の式を用いて算出することができる。
【００５７】
ｑ^* （ｔ）＝ｑ（ｔ）＋Ｈ（ｔ）
したがって、該値ｑ^* （ｔ）に基づいて、紙葉類の搬送枚数、重走及び紙葉類に貼（ちょう）付されたテープ等を検出することができる。
このように、基準媒体Ａの厚さｄ_A と基準媒体Ｂの厚さｄ_B との差Δｄによって生じる出力Ｑの変化量Δｑと、所定の値ｑｐとの比に基づいてゲイン補正値ｇｙを算出するようにしているので、ゲイン補正値ｇｙを正確に検出することができる。したがって、紙葉類の厚さを検出するに当たり、ミクロンオーダーの検出精度を得ることができる。
【００５８】
次に、本発明の第２の実施の形態について説明する。
図８は本発明の第２の実施の形態におけるゲイン補正値及びオフセット補正値を算出するための説明図である。なお、図において、横軸に基準媒体の状態を、縦軸にＡＤコンバータ１９（図２）の出力Ｑを採ってある。
この場合、ゲイン補正値及びオフセット補正値を算出するために、厚さが異なる２枚の基準媒体Ａ、Ｂ、及び前記基準媒体Ａと同じ厚さを有する基準媒体Ｅが使用される。そして、基準媒体Ａは基準媒体Ｂより薄くされる。
【００５９】
まず、基準媒体Ａ、Ｂを基準ローラ１１と厚さ検出ローラ１２との間に挿入しない状態において、ＤＡコンバータ１６によってオフセット補正値を変更して、ＡＤコンバータ１９の出力Ｑが規定範囲内の所定の値ｑ１０になるようにし、そのときのオフセット補正値をオフセット補正初期値としてセットする。このとき、ＤＡコンバータ１８によってゲイン補正値に任意のゲイン補正初期値ｇ１１をセットしておく。そして、ＣＰＵ２０は、このときのＡＤコンバータ１９の出力Ｑの値ｑ１０を読み込み、ＥＥＰＲＯＭ２１に格納する。
【００６０】
次に、オフセット補正値をオフセット補正初期値にしたまま、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ｂを挿入し、ＤＡコンバータ１８にセットするゲイン補正値を変更して、ＡＤコンバータ１９の出力Ｑが所定の値ｑ２２になるゲイン補正値ｇ１２を求め、値ｑ２２及び前記ゲイン補正値ｇ１２をＥＥＰＲＯＭ２１に格納する。
【００６１】
そして、オフセット補正初期値及び前記ゲイン補正値ｇ１２を維持したまま、同様にして基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ａを挿入し、ＡＤコンバータ１９の出力Ｑの値ｑ２３を読み込み、該値ｑ２３をＥＥＰＲＯＭ２１に格納する。
ここで、基準媒体Ａの厚さｄ_A と基準媒体Ｂの厚さｄ_B との差Δｄ
Δｄ＝ｄ_A −ｄ_B
に対応させて決められたＡＤコンバータ１９の出力Ｑの変化量をΔｑとして、値ｑ２２、ｑ２３及びゲイン補正値ｇ１２により、次の式により所期のゲイン補正値ｇｙを算出する。
【００６２】
ｇｙ＝ｇ１２×Δｑ／（ｑ２２−ｑ２３）
すなわち、ＤＡコンバータ１８にセットされるゲイン補正値がｇｙである場合、基準媒体Ａの厚さｄ_A と基準媒体Ｂの厚さｄ_B との差がΔｄであると、ＡＤコンバータ１９の出力Ｑの変化量がΔｑになることを示す。
次に、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ａを挿入したまま、ＤＡコンバータ１８に所期のゲイン補正値ｇｙをセットし、オフセット補正値を変化させて、ＡＤコンバータ１９の出力Ｑが規定範囲内の所定の値ｑｐとなるようにする。そして、ＣＰＵ２０は、前記出力Ｑが前記値ｑｐになったときのオフセット補正値ｆ５を算出し、該オフセット補正値ｆ５をＤＡコンバータ１６にセットする。このとき、ＣＰＵ２０は、前記オフセット補正値ｆ５をＡＤコンバータ１９の出力Ｑの値ｑｐと共にＥＥＰＲＯＭ２１に格納する。
【００６３】
続いて、オフセット補正値ｆ５及びゲイン補正値ｇｙを維持したまま、基準ローラ１１と厚さ検出ローラ１２との間から基準媒体Ａを取り除く。ＣＰＵ２０は、このときのＡＤコンバータ１９の出力Ｑの値Ｍを出力目標値としてＥＥＰＲＯＭ２１に格納する。
次に、偏心補正値を算出する場合、基準媒体Ａと同じ厚さの基準媒体Ｅを基準ローラ１１と厚さ検出ローラ１２との間に挿入し、図示しないモータ等の搬送系を駆動して基準媒体Ｅを定速で搬送する。そして、ＣＰＵ２０は、前記フォトインタラプタ２２の出力信号ＳＧ１（図７）にパルスＰＬを検出してから、１回転をｎ個に等分割した各タイミングｔ１〜ｔｎをＴ回転分（Ｔは正の整数）採り、各ｎ×Ｔ個のタイミングでＡＤコンバータ１９の出力Ｑの値ｑ（１）、…、ｑ（ｎ）、ｑ（ｎ＋１）、…、ｑ（ｎ×Ｔ）を読み込む。
【００６４】
ここで、基準ローラ１１及び厚さ検出ローラ１２を１回転させたときに、ｎ個のタイミングについて前記出力Ｑの値ｑ（１）、…、ｑ（ｎ）を得ることができるので、基準ローラ１１及び厚さ検出ローラ１２が同じ回転位置にあるときに、値ｑ（ｋ）（ｋ＝１、２、…、ｎ）と値ｑ（ｋ＋ｎ×（Ｕ−１））（Ｕ＝１、２、…、）とは等しくなる。
【００６５】
したがって、基準ローラ１１及び厚さ検出ローラ１２が同じ回転位置にあるときの、前記出力Ｑの各値ｑ（ｋ）、ｑ（ｋ＋ｎ）、…、ｑ（ｋ＋ｎ×（Ｕ−１））の平均値ｑ_PA（ｋ）は、

になる。
【００６６】
このことから、基準ローラ１１及び厚さ検出ローラ１２が各回転位置にあるときの偏心補正値Ｈ（ｋ）（ｋ＝１、２、…、ｎ）は次の式を用いて算出される。
Ｈ（ｋ）＝ｑ_PA−ｑ_PA（ｋ）
次に、前記構成の紙葉類厚さ検出装置によって紙葉類の厚さを検出する方法について説明する。
【００６７】
前記搬送系を駆動し、基準ローラ１１と厚さ検出ローラ１２との間に紙葉類を挿入する前に、ＣＰＵ２０は、ＤＡコンバータ１８に前記ゲイン補正値ｇｙをセットし、ＤＡコンバータ１６にセットするオフセット補正値を変更して、ＡＤコンバータ１９の出力Ｑを出力目標値Ｍにする。次に、ＣＰＵ２０は、基準ローラ１１及び厚さ検出ローラ１２が１回転するごとに前記出力信号ＳＧ１にパルスＰＬを検出し、各タイミングで偏心補正を行うことができるように準備する。
【００６８】
続いて、紙葉類が基準ローラ１１と厚さ検出ローラ１２との間に到達すると、タイミングｔにおけるＡＤコンバータ１８の出力Ｑの値ｑ（ｔ）（ｔ＝１、２、…、ｎ）に対する偏心補正を開始する。この場合、偏心補正が行われた後の値をｑ^* （ｔ）とすると、値ｑ^* （ｔ）（ｔ＝１、２、…、ｎ）を次の式を用いて算出することができる。
【００６９】
ｑ^* （ｔ）＝ｑ（ｔ）＋Ｈ（ｔ）
したがって、該値ｑ^* （ｔ）に基づいて、紙葉類の搬送枚数、重走及び紙葉類に貼付されたテープ等を検出することができる。
このように、基準ローラ１１と厚さ検出ローラ１２との間に厚い順に基準媒体Ｂ及び基準媒体Ａを挿入するようにしているので、基準媒体Ａ、Ｂを挿入する回数を少なくすることができる。
【００７０】
したがって、ゲイン補正値ｇｙ及びオフセット補正値ｆ５を算出するための作業を簡素化することができるだけでなく、算出するための時間を短くすることができる。また、基準媒体Ａ、Ｂを挿入する順序を誤ることがなくなるので、ゲイン補正値ｇｙ及びオフセット補正値ｆ５を正確に算出することができる。
次に、本発明の第３の実施の形態について説明する。
【００７１】
図９は本発明の第３の実施の形態におけるゲイン補正値及びオフセット補正値を算出するための説明図、図１０は本発明の第３の実施の形態における正規方程式の各項を説明するためのマトリクスである。なお、図９において、横軸に基準媒体の状態を、縦軸にＡＤコンバータ１９（図２）の出力Ｑを採ってある。
この場合、ゲイン補正値及びオフセット補正値を算出するために、厚さが異なるＪ枚の基準媒体Ｆ_r （ｒ＝１、２、…、Ｊ）を使用し、厚さの薄い順に基準媒体Ｆ₁ 、Ｆ₂ 、…、Ｆ_J とする。また、基準媒体Ｆ_r の厚さを薄い順にｄ₁ 、ｄ₂ 、…、ｄ_J とする。
【００７２】
まず、基準媒体Ｆ_r を基準ローラ１１と厚さ検出ローラ１２との間に挿入しない状態において、ＤＡコンバータ１６によってオフセット補正値を変更して、ＡＤコンバータ１９の出力Ｑが規定範囲内の所定の値ｑ_F0とし、そのときのオフセット補正値をオフセット補正初期値としてセットする。このとき、ＤＡコンバータ１８によってゲイン補正値に任意のゲイン補正初期値ｇ１１をセットしておく。そして、ＣＰＵ２０は、このときのＡＤコンバータ１９の出力Ｑの値ｑ_F0を読み込み、ＥＥＰＲＯＭ２１に格納する。
【００７３】
次に、オフセット補正値をオフセット補正初期値にしたまま、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ｆ₁ を挿入し、ＤＡコンバータ１８にセットするゲイン補正値を変更し、ＡＤコンバータ１９の出力Ｑが所定の値ｑ_F1になるゲイン補正値ｇ１２を求め、値ｑ_F1及び前記ゲイン補正値ｇ１２をＥＥＰＲＯＭ２１に格納する。
【００７４】
そして、オフセット補正初期値及び前記ゲイン補正値ｇ１２を維持したまま、同様にして基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ｆ₂ を挿入し、ＡＤコンバータ１９の出力Ｑの値ｑ_F2を読み込み、該値ｑ_F2をＥＥＰＲＯＭ２１に格納する。以降、オフセット補正初期値及び前記ゲイン補正値ｇ１２を維持したまま、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ｆ₃ 、Ｆ₄ 、…、Ｆ_J を挿入し、ＡＤコンバータ１９の出力Ｑの値ｑ_F3、ｑ_F4、…、ｑ_FJをＥＥＰＲＯＭ２１に格納する。
【００７５】
次に、最小２乗法により厚さＸと出力Ｙとの関係を表す近似式を求める。
ここで、厚さＸと出力Ｙとの関係が
Ｙ＝ａ＋ｂＸ
であるとすると、定数ａ、ｂを求めるためのＪ個の条件方程式
Ｙ_i ＝ａ＋ｂＸ_i （ｉ＝１、２、…、Ｊ）
が得られ、正規方程式は、図１０のマトリクスに従って、
ａ×Ｊ＋ｂ×Σｄ_r ＝Σｑ_Fr ……（１）
ａ×Σｄ_r ＋ｂ×Σ（ｄ_r ×ｄ_r ）＝Σ（ｄ_r ×ｑ_Fr） ……（２）
で表すことができる。したがって、前記式（１）、（２）の連立方程式を解いて定数ａ、ｂを求めることができる。
【００７６】
このようにして、最小２乗法による厚さＸと出力値Ｙとの関係を表す近似式を求めることができる。該近似式による基準媒体Ｆ₁ 及び基準媒体Ｆ_J の出力値ｑ_F1 ^* 、ｑ_FJ ^* は、
ｑ_F1 ^* ＝ａ＋ｂ×ｄ₁
ｑ_FJ ^* ＝ａ＋ｂ×ｄ_J
になる。そして、基準媒体Ｆ₁ 、Ｆ_J のそれぞれ厚さｄ₁ 、ｄ_J の差Δｄ_F に対応させて決められたＡＤコンバータ１９の出力Ｑの値の変化量をΔｑ_F とすると、ゲイン補正値ｇ１１に基づいて、次の式を用いて所期のゲイン補正値ｇｚを算出することができる。
【００７７】
ｇｚ＝ｇ１１×Δｑ_F ／（ｑ_FJ ^* −ｑ_F1 ^* ）
すなわち、ＤＡコンバータ１８にセットされるゲイン補正値がｇｚである場合、基準媒体Ａの厚さｄ_A と基準媒体Ｂの厚さｄ_B との差がΔｄ_F であると、ＡＤコンバータ１９の出力Ｑの変化量がΔｑ_F になることを示す。
次に、基準媒体Ｆ₁ を挿入したまま、ＤＡコンバータ１８に所期のゲイン補正値ｇｚをセットし、オフセット補正値を変更して、ＡＤコンバータ１９の出力Ｑが規定範囲内の所定の値ｑｐとなるようにする。そして、ＣＰＵ２０は、前記出力Ｑが前記値ｑｐになったときのオフセット補正値ｆ６を算出し、該オフセット補正値ｆ６をＤＡコンバータ１６にセットする。このとき、ＣＰＵ２０は、前記オフセット補正値ｆ６をＡＤコンバータ１９の出力Ｑの値ｑｐと共にＥＥＰＲＯＭ２１に格納する。
【００７８】
続いて、オフセット補正値ｆ６及びゲイン補正値ｇｚを維持したまま、基準ローラ１１と厚さ検出ローラ１２との間から基準媒体Ｆ₁ を取り除く。ＣＰＵ２０は、このときのＡＤコンバータ１９の出力Ｑの値Ｍを出力目標値としてＥＥＰＲＯＭ２１に格納する。
次に、偏心補正値を算出する場合、基準媒体Ａと同じ厚さの基準媒体Ｅを基準ローラ１１と厚さ検出ローラ１２との間に挿入し、図示しないモータ等の搬送系を駆動して基準媒体Ｅを定速で搬送する。そして、ＣＰＵ２０は、フォトインタラプタ２２の出力信号ＳＧ１（図７）にパルスＰＬを検出してから、１回転をｎ個に等分割した各タイミングｔ１〜ｔｎをＴ回転分（Ｔは正の整数）採り、各ｎ×Ｔ個のタイミングでＡＤコンバータ１９の出力Ｑの値ｑ（１）、…、ｑ（ｎ）、ｑ（ｎ＋１）、…、ｑ（ｎ×Ｔ）を読み込む。
【００７９】
ここで、基準ローラ１１及び厚さ検出ローラ１２を１回転させたときに、ｎ個のタイミングについて前記出力Ｑの値ｑ（１）、…、ｑ（ｎ）を得ることができるので、基準ローラ１１及び厚さ検出ローラ１２が同じ回転位置にあるときに、値ｑ（ｋ）（ｋ＝１、２、…、ｎ）と値ｑ（ｋ＋ｎ×（Ｕ−１））（Ｕ＝１、２、…、）とは等しくなる。
【００８０】
したがって、基準ローラ１１及び厚さ検出ローラ１２が同じ回転位置にあるときの、前記出力Ｑの各値ｑ（ｋ）、ｑ（ｋ＋ｎ）、…、ｑ（ｋ＋ｎ×（Ｕ−１））の平均値ｑ_PA（ｋ）は、

になる。
【００８１】
このことから、基準ローラ１１及び厚さ検出ローラ１２が各回転位置にあるときの偏心補正値Ｈ（ｋ）（ｋ＝１、２、…、ｎ）は次の式を用いて算出される。
Ｈ（ｋ）＝ｑ_PA−ｑ_PA（ｋ）
次に、前記構成の紙葉類厚さ検出装置によって紙葉類の厚さを検出する方法について説明する。
【００８２】
前記搬送系を駆動し、基準ローラ１１と厚さ検出ローラ１２との間に紙葉類を挿入する前に、ＣＰＵ２０は、ＤＡコンバータ１８に前記ゲイン補正値ｇｚをセットし、ＤＡコンバータ１６にセットするオフセット補正値を変化させて、ＡＤコンバータ１９の出力Ｑを出力目標値Ｍにする。次に、ＣＰＵ２０は、基準ローラ１１及び厚さ検出ローラ１２が１回転するごとに前記出力信号ＳＧ１にパルスＰＬを検出し、各タイミングで偏心補正を行うことができるように準備する。
【００８３】
続いて、紙葉類が基準ローラ１１と厚さ検出ローラ１２との間に到達すると、タイミングｔにおけるＡＤコンバータ１９の出力Ｑの値ｑ（ｔ）（ｔ＝１、２、…、ｎ）に対する偏心補正を開始する。この場合、偏心補正が行われた後の値をｑ^* （ｔ）とすると、値ｑ^* （ｔ）（ｔ＝１、２、…、ｎ）を次の式を用いて算出することができる。
【００８４】
ｑ^* （ｔ）＝ｑ（ｔ）＋Ｈ（ｔ）
したがって、該値ｑ^* （ｔ）に基づいて、紙葉類の搬送枚数、重走及び紙葉類に貼付されたテープ等を検出することができる。
このように、基準ローラ１１と厚さ検出ローラ１２との間に厚さが異なるＪ枚の基準媒体Ｆ_r （ｒ＝１、２、…、Ｊ）を薄い順に挿入し、各ＡＤコンバータ１９の出力Ｑの値ｑ_Fr（ｒ＝１、２、…、Ｊ）を読み取り、値ｑ_Frに基づいて最小２乗法により厚さＸと出力Ｙとの関係を表す近似式を求め、該近似式に基づいてゲイン補正値ｇｚを算出するようになっているので、基準媒体Ｆ_r の数が多い分だけゲイン補正値ｇｚを正確に算出することができる。
【００８５】
次に、本発明の第４の実施の形態について説明する。
図１１は本発明の第４の実施の形態におけるゲイン補正値及びオフセット補正値を算出するための説明図である。なお、図において、横軸に基準媒体の状態を、縦軸にＡＤコンバータ１９（図２）の出力Ｑを採ってある。
この場合、ゲイン補正値及びオフセット補正値を算出するために、厚さが異なる２枚の基準媒体Ａ、Ｂ、及び前記基準媒体Ａと同じ厚さを有する基準媒体Ｅが使用される。そして、基準媒体Ａは基準媒体Ｂより薄くされる。
【００８６】
まず、基準媒体Ａ、Ｂを基準ローラ１１と厚さ検出ローラ１２との間に挿入しない状態において、ＤＡコンバータ１６によってオフセット補正値を変更して、ＡＤコンバータ１９の出力Ｑが規定範囲内の所定の値ｑ１０になるようにし、そのときのオフセット補正値をオフセット補正初期値としてセットする。このとき、ＤＡコンバータ１８によってゲイン補正値に任意のゲイン補正初期値ｇ１１をセットしておく。そして、ＣＰＵ２０は、このときのＡＤコンバータ１９の出力Ｑの値ｑ１０及び基準ローラ１１を読み込み、ＥＥＰＲＯＭ２１に格納する。
【００８７】
次に、オフセット補正値をオフセット補正初期値にしたまま、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ｂを挿入し、図示しないモータ等の搬送系を駆動することによって前記基準媒体Ｂを搬送する。そして、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ｂがあるときに、ＣＰＵ２０は、所定間隔を置いたタイミングでＡＤコンバータ１９の出力Ｑを複数個読み込み、該出力Ｑの平均値ｑ_A1を算出する。そして、この作業を複数回繰り返し、同時にＤＡコンバータ１８にセットするゲイン補正値を変更し、前記平均値ｑ_A1が所定の値ｑ２２になるゲイン補正値ｇ１２を求め、前記値ｑ２２及び前記ゲイン補正値ｇ１２をＥＥＰＲＯＭ２１に格納する。
【００８８】
次に、オフセット補正初期値及び前記ゲイン補正値ｇ１２を維持したまま、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ａを挿入し、前記搬送系を駆動することによって前記基準媒体Ａを搬送する。そして、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ａがあるときに、ＣＰＵ２０は、所定間隔を置いたタイミングでＡＤコンバータ１９の出力Ｑを複数個読み込み、該出力Ｑの平均値ｑ_A2を算出する。このとき、該平均値ｑ_A2をＥＥＰＲＯＭ２１に格納する。
【００８９】
ここで、基準媒体Ａの厚さｄ_A と基準媒体Ｂの厚さｄ_B との差Δｄ
Δｄ＝ｄ_A −ｄ_B
に対応させて決められたＡＤコンバータ１９の出力Ｑの変化量をΔｑとして、値ｑ２２、平均値ｑ_A2及びゲイン補正値ｇ１２により、次の式を用いて所期のゲイン補正値ｇａを算出する。
【００９０】
ｇａ＝ｇ１２×Δｑ／（ｑ２２−ｑ_A2）
すなわち、ＤＡコンバータ１８にセットされるゲイン補正値がｇｙ_A である場合、基準媒体Ａの厚さｄ_A と基準媒体Ｂの厚さｄ_B との差がΔｄであると、ＡＤコンバータ１９の出力Ｑの変化量がΔｑになることを示す。
次に、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ａを挿入し、かつ、オフセット補正値を維持したまま、ＤＡコンバータ１８に所期のゲイン補正値ｇｙ_A をセットし、所定間隔を置いたタイミングでＡＤコンバータ１９の出力Ｑを複数個読み込み、該出力Ｑの平均値ｑ_A3を算出する。そして、この作業を複数回繰り返し、同時にＤＡコンバータ１６にセットするオフセット補正値を変更し、前記平均値ｑ_A3が所定の値ｑｐ_A になるオフセット補正値ｆ_A を求め、平均値ｑ_A3及びオフセット補正値ｆ_A をＥＥＰＲＯＭ２１に格納する。なお、以上の処理は、基準ローラ１１と厚さ検出ローラ１２との間に基準媒体Ａがあるうちに終了させる。
【００９１】
続いて、オフセット補正値ｆ_A 及びゲイン補正値ｇｙ_A を維持したまま、所定間隔を置いたタイミングでＡＤコンバータ１９の出力Ｑを複数個読み込み、該出力Ｑの平均値ｑ_A4を算出する。ＣＰＵ２０は、平均値ｑ_A4を出力目標値としてＥＥＰＲＯＭ２１に格納する。
以下、第１の実施の形態と同様に、基準媒体Ａと同じ厚さの基準媒体Ｅを基準ローラ１１と厚さ検出ローラ１２との間に挿入し、搬送系を駆動して基準媒体Ｅを定速で搬送してＡＤコンバータ１９の出力Ｑの値ｑ（１）、…、ｑ（ｎ）、ｑ（ｎ＋１）、…、ｑ（ｎ×Ｔ）を読み込む。
【００９２】
そして、基準ローラ１１及び厚さ検出ローラ１２が同じ回転位置にあるときの、前記出力Ｑの各値ｑ（ｋ）、ｑ（ｋ＋ｎ）、…、ｑ（ｋ＋ｎ×（Ｕ−１））の平均値ｑ_PA（ｋ）を算出することができ、基準ローラ１１及び厚さ検出ローラ１２が各回転位置にあるときの偏心補正値Ｈ（ｋ）（ｋ＝１、２、…、ｎ）は次の式を用いて算出することができる。
【００９３】
Ｈ（ｋ）＝ｑ_PA−ｑ_PA（ｋ）
したがって、前記構成の紙葉類厚さ検出装置によって紙葉類の厚さを検出する場合は、偏心補正が行われた後の出力Ｑの値ｑ^* （ｔ）を次の式を用いて算出することができる。
ｑ^* （ｔ）＝ｑ（ｔ）＋Ｈ（ｔ）
このように、２枚の厚さが異なる基準媒体Ａ、Ｂを、１枚ずつ基準ローラ１１と厚さ検出ローラ１２との間を搬送するだけでゲイン補正値ｇｙ_A 及びオフセット補正値ｆ_A を算出することができるので、作業を簡素化することができる。
【００９４】
次に、本発明の第５の実施の形態について説明する。
図１２は本発明の第５の実施の形態における基準媒体を示す図である。
図において、Ｋは基準媒体であり、厚さが異なる２枚の基準媒体Ｋ₁ 、Ｋ₂ を端部において貼付することによって形成される。この場合、基準媒体Ｋ₂ は基準媒体Ｋ₁ より薄くされる。
【００９５】
そして、前記基準媒体Ｋを、基準媒体Ｋ₁ 側を前方にして搬送することによって、第７の実施の形態と同じ方法でゲイン補正値ｇｙ_A （図１１）、オフセット補正値ｆ_A 及び偏心補正値Ｈ（ｋ）を算出することができる。
この場合、１枚の基準媒体Ｋを挿入して搬送するだけでよいので、作業を一層簡素化することができる。
【００９６】
次に、本発明の第６の実施の形態について説明する。
この場合、第１の実施の形態と同じ方法でオフセット補正値ｆ５及びゲイン補正値ｇｙ（図６）を算出した後、偏心補正値を算出するに当たり、基準媒体Ａと同じ厚さの基準媒体Ｅを基準ローラ１１と厚さ検出ローラ１２との間に挿入し、図示しないモータ等の搬送系を駆動して基準媒体Ｅを定速で第１、第２の方向に搬送する。そのために、前記搬送系は正方向及び逆方向に駆動することができるようになっている。
【００９７】
そして、基準媒体Ｅを第１の方向に搬送したときの、ＡＤコンバータ１９の出力Ｑの値ｑ（１）、…、ｑ（ｎ）、ｑ（ｎ＋１）、…、ｑ（ｎ×Ｔ）を読み込み、基準ローラ１１及び厚さ検出ローラ１２が同じ回転位置にあるときの、前記出力Ｑの各値ｑ（ｋ）、ｑ（ｋ＋ｎ）、…、ｑ（ｋ＋ｎ×（Ｕ−１））の平均値ｑ_PA（ｋ）を算出する。また、基準ローラ１１及び厚さ検出ローラ１２が各回転位置にあるときの偏心補正値Ｈ（ｋ）
Ｈ（ｋ）＝ｑ_PA−ｑ_PA（ｋ）
を算出し、偏心補正が行われた後の出力Ｑの値ｑ^* （ｔ）
ｑ^* （ｔ）＝ｑ（ｔ）＋Ｈ（ｔ）
を算出する。
【００９８】
次に、基準媒体Ｅを第２の方向に搬送したときの、ＡＤコンバータ１９の出力Ｑの値ρ（１）、…、ρ（ｎ）、ρ（ｎ＋１）、…、ρ（ｎ×Ｔ）を読み込み、基準ローラ１１及び厚さ検出ローラ１２が同じ回転位置にあるときの、前記出力Ｑの各値ρ（ｋ）、ρ（ｋ＋ｎ）、…、ρ（ｋ＋ｎ×（Ｕ−１））の平均値ρ_PA（ｋ）を算出する。また、基準ローラ１１及び厚さ検出ローラ１２が各回転位置にあるときの偏心補正値Ｈ（ｋ）
Ｈ（ｋ）＝ρ_PA−ρ_PA（ｋ）
を算出し、偏心補正が行われた後の出力Ｑの値ρ^* （ｔ）
ρ^* （ｔ）＝ρ（ｔ）＋Ｈ（ｔ）
を算出する。
【００９９】
したがって、紙葉類の厚さを検出する場合、紙葉類を第１の方向に搬送したときには、
ｑ^* （ｔ）＝ｑ（ｔ）＋Ｈ（ｔ）
の式を用いて、紙葉類を第２の方向に搬送したときには、
ρ^* （ｔ）＝ρ（ｔ）＋Ｈ（ｔ）
の式を用いて、偏心補正を行うことができる。
【０１００】
このように、紙葉類を搬送する方法に対応させて偏心補正値Ｈ（ｋ）を異ならせることができるので、紙葉類の厚さを正確に検出することができる。
なお、本発明は前記実施の形態に限定されるものではなく、本発明の趣旨に基づいて種々変形させることが可能であり、それらを本発明の範囲から排除するものではない。
【０１０１】
【発明の効果】
以上詳細に説明したように、本発明によれば、紙葉類厚さ検出装置の補正方法においては、基準ローラと厚さ検出ローラとの間に厚さの異なる複数の基準媒体を挿入して、前記厚さ検出ローラの変位を検出し、出力発生手段によって前記変位に対応する出力を発生させ、前記基準媒体の厚さの差、及び該差に基づいて生じる前記出力の変化量に基づいてゲイン補正値を算出し、該ゲイン補正値に従って前記出力を調整する。
【０１０７】
この場合、２種類の基準媒体の厚さの差、及び該厚さの差によって生じる出力の変化量に基づいてゲイン補正値を算出するようにしているので、ゲイン補正値を正確に検出することができる。したがって、紙葉類の厚さを検出するに当たり、ミクロンオーダーの検出精度を得ることができる。
【図面の簡単な説明】
【図１】本発明の説明の前提となる技術の第１の形態におけるゲイン補正値及びオフセット補正値を算出するための説明図である。
【図２】従来の紙葉類厚さ検出装置の概略図である。
【図３】従来のゲイン補正値及びオフセット補正値を算出するための説明図である。
【図４】従来の偏心補正値を算出するための説明図である。
【図５】本発明の説明の前提となる技術の第１の形態における偏心補正値を算出するための説明図である。
【図６】本発明の第１の実施の形態におけるゲイン補正値及びオフセット補正値を算出するための説明図である。
【図７】本発明の第１の実施の形態におけるタイミングを示す図である。
【図８】本発明の第２の実施の形態におけるゲイン補正値及びオフセット補正値を算出するための説明図である。
【図９】本発明の第３の実施の形態におけるゲイン補正値及びオフセット補正値を算出するための説明図である。
【図１０】本発明の第３の実施の形態における正規方程式の各項を説明するためのマトリクスである。
【図１１】本発明の第４の実施の形態におけるゲイン補正値及びオフセット補正値を算出するための説明図である。
【図１２】本発明の第５の実施の形態における基準媒体を示す図である。
【符号の説明】
１１ 基準ローラ
１２ 厚さ検出ローラ
１６、１８ ＤＡコンバータ
１９ ＡＤコンバータ[0001]
BACKGROUND OF THE INVENTION
  The present invention relates to a paper sheet thickness detection apparatus and a correction method thereof.
[0002]
[Prior art]
Conventionally, in automatic transaction devices, printing devices, OCR, etc., the thickness of paper sheets such as cards and banknotes is detected, and based on the detection results, the presence or absence of tape affixed to the paper sheets, paper The authenticity of the leaves is determined. For this purpose, a paper sheet thickness detection device is provided, the paper sheet thickness detection device includes a reference roller and a thickness detection roller arranged to face each other, and a paper sheet is inserted between the two. The thickness of the paper sheet is detected based on the displacement of the thickness detection roller.
[0003]
Then, the gain correction value, the offset correction value, and the eccentricity correction value of the paper sheet thickness detection device are calculated so that a minute variation in the thickness of the paper sheet can be captured by the sensor, and based on the calculation result Gain correction, offset correction, and eccentricity correction are performed.
FIG. 2 is a schematic view of a conventional paper sheet thickness detection apparatus.
[0004]
In the figure, 11 is a reference roller, 12 is a thickness detection roller disposed opposite to the reference roller 11, 31 and 32 are bearings disposed at both ends of the reference roller 11, and 33 and 34 are the thickness. Bearings 35 and 36 disposed at both ends of the detection roller 12 are fulcrum brackets that rotatably support the bearings 33 and 34. The bearings 31 and 32 are rotatably supported with respect to a frame (not shown) without changing the position, whereas the bearings 33 and 34 are rotatable so that the position can be changed with respect to the frame. Supported.
[0005]
The fulcrum brackets 35 and 36 are biased by springs 37 and 38, respectively, and press the bearings 33 and 34 against the bearings 31 and 32. At this time, there is a gap t between the reference roller 11 and the thickness detection roller 12._GIs set.
The gap t_GT is the minimum value of the thickness of the paper sheet handled in the paper thickness detecting device._SThe minimum value t_SVariation and sheet thickness detection device manufacturing error (± t_E)
t_G= T_S-A × t_E
a: Safety factor
Given in. Therefore, the thickness of the paper sheet inserted between the reference roller 11 and the thickness detection roller 12 can be detected without fail.
[0006]
The potentiometer 14 is connected to the fulcrum bracket 36, and the potentiometer 14 detects the displacement of the thickness detection roller 12 when a sheet is inserted between the reference roller 11 and the thickness detection roller 12. The displacement is converted into an electric signal. An amplifier 15 is connected to the potentiometer 14 and the electric signal is amplified by the amplifier 15.
[0007]
The output of the amplifier 15 and the output of the DA converter 16 are input to the amplifier 17, and the output of the amplifier 17 is input to the DA converter 18. Further, the output of the DA converter 18 is further input to an AD converter 19, and the output Q of the AD converter 19 is captured and processed by the CPU 20. The CPU 20 is connected to the DA converter 16, the DA converter 18, and the EEPROM 21, respectively. The DA converters 16 and 18 and the AD converter 19 constitute output generating means.
[0008]
A photo interrupter 22 is provided to detect the rotation of the thickness detection roller 12, and an output signal of the photo interrupter 22 is input to the CPU 20. The rotation of the reference roller 11 can also be detected by the photo interrupter 22.
Next, a method for calculating the gain correction value, the offset correction value, and the eccentricity correction value by the paper sheet thickness detection device having the above-described configuration will be described.
[0009]
FIG. 3 is an explanatory diagram for calculating a conventional gain correction value and offset correction value, and FIG. 4 is an explanatory diagram for calculating a conventional eccentricity correction value. In FIG. 3, the horizontal axis indicates the state of the reference medium, the vertical axis indicates the output Q of the AD converter 19 (FIG. 2), the horizontal axis indicates the rotational position of the thickness detection roller 12 in FIG. The output Q is taken.
[0010]
  In this case, in order to calculate the gain correction value, the offset correction value, and the eccentricity correction value, the two reference media A and B having different thicknesses, and the reference media having the same thickness as normal paper sheets without foreign matter D is used. The reference medium A is made thinner than the reference medium B.
  First, the reference medium A is inserted between the reference roller 11 and the thickness detection roller 12, and the offset correction value is changed by the DA converter 16 so that the output Q of the AD converter 19 becomes a predetermined value q1 within a specified range. Then, the offset correction value at that time is set as the offset correction initial value. At this time, the DA converter 18 sets an arbitrary gain correction initial value g1 as the gain correction value. Then, the CPU 20 reads the value q1 of the output Q of the AD converter 19 at this time and stores it in the EEPROM 21.
[0011]
  The gain correction value is a correction value for adjusting the slope of the output Q in the graph of FIG. 3, and the gain correction value is changed by adjusting the DA converter 18 to change the slope of the output Q. be able to. The offset correction value is a correction value for adjusting the intercept of the graph of FIG. 3, and the offset correction value can be changed by adjusting the DA converter 16 to change the intercept.
[0012]
  Next, similarly, the reference medium B is inserted between the reference roller 11 and the thickness detection roller 12, and the gain correction value is changed so that the output Q of the AD converter 19 becomes a predetermined value q2. A gain correction value g2 is obtained. Further, the CPU 20 stores the gain correction value g2 at this time in the EEPROM 21.
  Then, the reference medium A is inserted again between the reference roller 11 and the thickness detection roller 12 while maintaining the gain correction value g2. At this time, the CPU 20 reads the value q 3 of the output Q of the AD converter 19 and stores it in the EEPROM 21.
[0013]
  Subsequently, the CPU 20 calculates an intended gain correction value gx of the paper sheet thickness detection device according to a predetermined formula based on the values q1 and q3, the gain correction initial value g1 and the gain correction value g2.
  The CPU 20 sets the calculated desired gain correction value gx in the DA converter 18.
[0014]
Next, in order to calculate the offset correction value, the CPU 20 sets the value qx of the output Q of the AD converter 19 with the reference medium A being inserted between the reference roller 11 and the thickness detection roller 12. The offset correction value is changed so that the value qx becomes the value q1. Then, the CPU 20 calculates an offset correction value f2 when the value qx becomes the value q1, and sets the offset correction value f2 in the DA converter 16. At this time, the CPU 20 stores the offset correction value f2 in the EEPROM 21 together with the value qx of the output Q of the AD converter 19.
[0015]
Subsequently, the reference medium B is again inserted between the reference roller 11 and the thickness detection roller 12. Then, the CPU 20 reads the value q 4 of the output Q of the AD converter 19 and stores it in the EEPROM 21.
Next, as shown in the following equation, the CPU 20 determines that no sheet is inserted between the reference roller 11 and the thickness detection roller 12 based on the values q0, q1, and q4, that is, no medium is present. Time gap t_GOIs calculated. As shown in FIG. 3, the value q0 is the value of the output Q of the AD converter 19 when there is no medium.
[0016]
t_GO= (Q0-α) / β
β = (q4-q0) / (d_B-D_A)
α = q4−β × d_B
Or
α = q0−β × d_A
However,
d_A= Thickness of reference medium A
d_B= Thickness of reference medium B
And
[0017]
Next, in order to perform eccentricity correction, the reference media A and B are removed from between the reference roller 11 and the thickness detection roller 12, and a conveyance system such as a motor (not shown) is driven to drive the reference roller 11 and the thickness detection roller. 12 is rotated. During this time, the CPU 20 reads the output signal of the photo interrupter 22.
The CPU 20 determines the rotation position of the thickness detection roller 12 when a pulse is generated in the output signal of the photo interrupter 22, and the timing H at which the thickness detection roller 12 starts one rotation.₀And the timing H₀The value q0 of the output Q of the AD converter 19 at₀Is stored in the EEPROM 21.
[0018]
Next, a reference medium D having the same thickness as a normal paper sheet without foreign matter is inserted between the reference roller 11 and the thickness detection roller 12, and the reference system D is driven at a constant speed by driving the transport system. Transport by. At this time, after detecting a pulse in the output signal, the CPU 20 divides each rotation equally into n (n is a positive integer) each timing H₀~ H_nThe value qd of the output Q of the AD converter 19_i(I = 0, 1,..., N) is read and stored in the EEPROM 21. And each said value qd_iThe timing H₀Initial value qd at₀Output difference (qd_i-Qd₀) And the output difference (qd_i-Qd₀) Is stored in the EEPROM 21 as the reference medium eccentricity correction value HDi when the reference medium D is conveyed.
[0019]
Further, the reference medium D is removed from between the reference roller 11 and the thickness detection roller 12, and the conveyance system is driven to rotate the reference roller 11 and the thickness detection roller 12. At this time, after detecting a pulse in the output signal, the CPU 20 detects each timing H₀~ H_nThe value q0 of the output Q of the AD converter 19_i(I = 0, 1,..., N) is read and stored in the EEPROM 21. And the value q0_iThe timing H₀Initial value q0 at₀Output difference (q0_i-Q0₀) And the output difference (q0_i-Q0₀) Is stored in the EEPROM 21 as the idling correction value KDi.
[0020]
[Problems to be solved by the invention]
However, in the correction method of the conventional paper sheet thickness detection device, when the offset correction value f2 is calculated, the reference roller 11 and the thickness detection roller 12 and the thickness detection roller 12 cannot be rotated. The value of the output Q of the AD converter 19 varies depending on the eccentricity existing in the roller 12, and the offset correction value f2 cannot be calculated accurately.
[0021]
Therefore, since offset correction is not properly performed, when the thickness of the paper sheet is detected, the detected thickness varies depending on the amount of eccentricity existing in the reference roller 11 and the thickness detection roller 12. End up.
Further, when the reference medium eccentricity correction value HDi and the idling rotation eccentricity correction value KDi are calculated, the reference medium D having the same thickness as the paper sheet whose thickness is to be detected is conveyed. In addition, the thickness variation of the reference medium D is added to the amount of eccentricity existing in the thickness detection roller 12, and the reference medium eccentricity correction value HDi and the idling eccentricity correction value KDi cannot be accurately calculated.
[0022]
  Accordingly, since the eccentricity correction is not properly performed, when the thickness of the paper sheet is detected, the detected thickness varies depending on the thickness variation of the reference medium D.
  The present invention solves the problems of the correction method of the conventional paper sheet thickness detection device, can appropriately perform offset correction and eccentricity correction, and detects when detecting the thickness of paper sheets. It is an object of the present invention to provide a paper sheet thickness detection device and a correction method thereof that do not cause variations in the thickness to be generated.
[0023]
[Means for Solving the Problems]
  Therefore, in the correction method of the paper sheet thickness detection device of the present invention, a plurality of reference media having different thicknesses are inserted between the reference roller and the thickness detection roller, and the displacement of the thickness detection roller is And generating an output corresponding to the displacement by the output generation means, calculating a gain correction value based on the difference in thickness of the reference medium and the amount of change in the output generated based on the difference, The output is adjusted according to the gain correction value.
[0027]
DETAILED DESCRIPTION OF THE INVENTION
  DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the technology as a premise of the description of the present invention and embodiments of the present invention will be described in detail with reference to the drawings. Since the paper sheet thickness detection device has the same structure as the conventional one, FIG. 2 is used.
  FIG. 1 is an explanatory diagram for calculating a gain correction value and an offset correction value in the first embodiment of the technology that is the premise of the description of the present invention, and FIG. 5 is a first embodiment of the technology that is the premise of the description of the present invention. It is explanatory drawing for calculating the eccentricity correction value in. In FIG. 1, the horizontal axis indicates the state of the reference medium, the vertical axis indicates the output Q of the AD converter 19 (FIG. 2), the horizontal axis indicates the rotational position of the thickness detection roller 12 in FIG. The output Q of the AD converter 19 is taken.
[0028]
In this case, in order to calculate the gain correction value and the offset correction value, two reference media A and B having different thicknesses and a reference medium E having the same thickness as the reference medium A are used. The reference medium A is made thinner than the reference medium B.
First, the reference medium A is inserted between the reference roller 11 and the thickness detection roller 12, the offset correction value is changed by the DA converter 16, and the output Q of the AD converter 19 becomes a predetermined value q1 within a specified range. Thus, the offset correction value at that time is set as the offset correction initial value. At this time, the DA converter 18 sets an arbitrary gain correction initial value g1 as the gain correction value. Then, the CPU 20 reads the value q1 of the output Q of the AD converter 19 at this time and stores it in the EEPROM 21.
[0029]
Next, similarly, the reference medium B is inserted between the reference roller 11 and the thickness detection roller 12, the gain correction value is changed, and the gain correction in which the output Q of the AD converter 19 becomes a predetermined value q2. The value g2 is obtained. Further, the CPU 20 stores the gain correction value g2 at this time in the EEPROM 21.
Then, the reference medium A is inserted again between the reference roller 11 and the thickness detection roller 12 while maintaining the gain correction value g2. At this time, the CPU 20 reads the value q 3 of the output Q of the AD converter 19 and stores it in the EEPROM 21.
[0030]
  Subsequently, the CPU 20 calculates an intended gain correction value gx of the paper sheet thickness detection device according to a predetermined formula based on the values q1 and q3, the gain correction initial value g1 and the gain correction value g2.
  The CPU 20 sets the calculated desired gain correction value gx in the DA converter 18.
[0031]
Next, in order to calculate an offset correction value, the reference medium E is inserted between the reference roller 11 and the thickness detection roller 12, and a transport system such as a motor (not shown) is driven to move the reference medium E at a constant speed. Transport by. The CPU 20 reads the value qx of the output Q of the AD converter 19 when the reference roller 11 and the thickness detection roller 12 are rotated once, and the average value qx of the values qx._AAnd the average value qx_AThe offset correction value is changed so that becomes the value q1. Further, the CPU 20 determines that the average value qx_AThe offset correction value f3 is calculated when the value becomes the value q1, and the offset correction value f3 is set in the DA converter 16. Then, the CPU 20 determines that the offset correction value f3 and the average value qx_AIs stored in the EEPROM 21.
[0032]
In this way, the reference roller 11 and the thickness detection roller 12 are rotated once, and the average value qx of the value qx of the output Q of the AD converter 19 is determined._AAnd the average value qx_ASince the offset correction value f3 is calculated so that becomes the value q1, even if the reference roller 11 or the thickness detection roller 12 is decentered, it is not affected, and the offset correction value f3 is accurately calculated. be able to.
[0033]
Therefore, since offset correction is performed appropriately, when detecting the thickness of the paper sheet, the detected thickness does not vary depending on the amount of eccentricity existing in the reference roller 11 and the thickness detection roller 12. . As a result, the thickness of the paper sheet can be accurately detected.
Next, in order to calculate an eccentricity correction value, the reference medium E is inserted between the reference roller 11 and the thickness detection roller 12, and the transport system is driven to transport the reference medium E at a constant speed. Then, in order to detect the first displacement of the thickness detection roller 12, the CPU 20 detects a pulse in the output signal of the photo interrupter 22, and then each timing H in which one rotation is equally divided into n.₀~ H_nThe value qe of the output Q of the AD converter 19 as the first output_i(I = 0, 1,..., N) is read and stored in the EEPROM 21. Further, each value qe_iThe timing H₀Initial value qe at₀Output difference (qe_i-Qe₀) And the output difference (qe)_i-Qe₀) Is stored in the EEPROM 21 as the reference medium eccentricity correction value HEi when the reference medium E is conveyed.
[0034]
Next, the reference medium E is removed from between the reference roller 11 and the thickness detection roller 12, and the reference roller 11 and the thickness detection roller 12 are rotated. Then, in order to detect the second displacement of the thickness detection roller 12, the CPU 20 detects a pulse in the output signal in the same manner as when the reference medium E is conveyed, and then detects each timing H₀~ H_nThe value q0 of the output Q of the AD converter 19 as the second output_i(I = 0, 1,..., N) is read and stored in the EEPROM 21. The value q0_iThe timing H₀Initial value q0 at₀Output difference (q0_i-Q0₀) And the output difference (q0_i-Q0₀) And is stored in the EEPROM 21 as the eccentricity correction value KEi.
[0035]
Next, a method for detecting the thickness of a paper sheet by the paper thickness detecting device having the above-described configuration will be described.
When the transport system is driven and a paper sheet is transported between the reference roller 11 and the thickness detection roller 12, the thickness detection roller 12 is displaced, and the displacement is mechanically transmitted to the potentiometer 14. And converted into an electrical signal by the potentiometer 14. The electric signal is amplified by the amplifier 15 and input to the amplifier 17 via one terminal.
[0036]
At this time, the CPU 20 reads the offset correction value f3 from the EEPROM 21 and inputs the offset correction value f3 to the amplifier 17 via the other terminal to perform offset correction. Then, the output of the amplifier 17 is input to the DA converter 18 with the offset correction being performed. Further, the CPU 20 reads the desired gain correction value gx from the EEPROM 21, sets it in the DA converter 18, and performs gain correction.
[0037]
Then, the CPU 20 rotates each of the timings H while rotating the reference roller 11 and the thickness detection roller 12.₀~ H_nThe value q of the output Q of each AD converter 19 for each_iAnd the reference medium eccentricity correction value HEi and the idle rotation eccentricity correction value KEi stored in the EEPROM 21 are read, and the value q_iΔq between the value and the idle rotation eccentricity correction value KEi_iAnd calculate each difference Δq_iThe eccentricity correction is performed by adding the reference medium eccentricity correction value HEi.
[0038]
  Next, a second embodiment of the technology that is the premise for describing the present invention will be described. In addition, about the thing which has the same structure as a 1st form, the description is abbreviate | omitted by providing the same code | symbol.
  In this case, in order to calculate the offset correction value, a reference medium D having the same thickness as the paper sheet for detecting the thickness is inserted between the reference roller 11 and the thickness detection roller 12, and a transport system such as a motor is used. To convey the reference medium D at a constant speed.
[0039]
Then, the CPU 20 outputs the value Qx of the output Q of the AD converter 19 when the reference roller 11 and the thickness detection roller 12 are rotated once._DAnd the value qx_DMean value qx_MDAnd the average value qx_MDThe offset correction value is changed so that becomes the value q1. Further, the CPU 20 determines that the average value qx_MDThe offset correction value f4 when the value becomes the value q1 is calculated, and the offset correction value f4 is set in the DA converter 16.
[0040]
Thus, since the thickness of the reference medium D for calculating the offset correction value f4 is made equal to the paper sheet whose thickness is to be detected, other than the eccentricity of the reference roller 11 or the thickness detection roller 12 The offset correction value f4 can be accurately calculated without being affected.
Therefore, since the offset correction is performed appropriately, when the thickness of the paper sheet is detected, the detected thickness does not vary. As a result, the thickness of the paper sheet can be accurately detected.
[0041]
  Next, a third embodiment of the technology that is the premise for describing the present invention will be described. In addition, about the thing which has the same structure as a 1st form, the description is abbreviate | omitted by providing the same code | symbol.
  In this case, a transport system such as a motor is driven to rotate the reference roller 11 and the thickness detection roller 12 to transport the reference medium A at a constant speed. Then, the CPU 20 determines the position of the home position set in advance in the reference roller 11 and the thickness detection roller 12 with respect to the reference medium A based on the output signal from the photo interrupter 22. A value q of the output Q of the AD converter 19 detected based on the waveform and at the home position_HPAnd the value q_HPIs changed to a predetermined value q1 within the specified range. Further, the CPU 20 determines that the value q_HPOffset correction value f when becomes the value q1_HPIs stored in the EEPROM 21.
[0042]
Then, when detecting the thickness of the paper sheet, the transport system is driven to rotate the reference roller 11 and the thickness detection roller 12, thereby transporting the paper sheet at a constant speed. At this time, the CPU 20 calculates the amount of eccentricity at each rotational position of the reference roller 11 and the thickness detection roller 12 with respect to the home position, and uses the amount of eccentricity as the value q of the output Q of the AD converter 19._iAdd to.
[0043]
  Therefore, even if the reference roller 11 and the thickness detection roller 12 are decentered, they are not affected, and the thickness of the paper sheet can be accurately detected.
  In the first to third embodiments, since the gain correction value gx is calculated based on the values q1, q3, the gain correction initial value g1, and the gain correction value g2, the thickness of the paper sheet In detecting the thickness, it is not possible to obtain detection accuracy on the order of microns.
[0044]
In addition, in order to calculate the gain correction value gx and the offset correction values f3 and f4, it is necessary to repeatedly insert the two types of reference media A and B between the reference roller 11 and the thickness detection roller 12, which makes the operation Not only is it bothersome, but it takes time to calculate. If the order of inserting the reference media A and B is incorrect, the gain correction value gx and the offset corrections f3 and f4 cannot be accurately calculated.
[0045]
Further, when calculating the gain correction value gx, two output Q values when two types of reference media A and B are inserted between the reference roller 11 and the thickness detection roller 12 are used. In some cases, the value of one output Q may be large. In this case, the gain correction value gx cannot be accurately calculated.
Further, when calculating the reference medium eccentricity correction value HEi and the idle rotation eccentricity correction value KEi, the reference roller 11 and the thickness detection roller 12 are rotated only in one direction, so that the paper sheet is conveyed in both directions. However, if it is attempted to detect the thickness of the paper sheet, the reference medium eccentricity correction value HEi and the idling eccentricity correction value KEi cannot be accurately detected depending on the rotation direction.
[0046]
  Next, a first embodiment of the present invention will be described.
  FIG. 6 is an explanatory diagram for calculating the gain correction value and the offset correction value in the first embodiment of the present invention. In the figure, the horizontal axis represents the state of the reference medium, and the vertical axis represents the output Q of the AD converter 19 (FIG. 2). Moreover, about the thing which has the same structure as a 1st form, the description is abbreviate | omitted by providing the same code | symbol.
[0047]
In this case, in order to calculate the gain correction value and the offset correction value, two reference media A and B having different thicknesses and a reference medium E having the same thickness as the reference medium A are used. The reference medium A is made thinner than the reference medium B.
First, in a state in which the reference media A and B are not inserted between the reference roller 11 and the thickness detection roller 12, the offset correction value is changed by the DA converter 16, and the output Q of the AD converter 19 is within a predetermined range. The offset correction value at that time is set as an offset correction initial value. At this time, the DA converter 18 sets an arbitrary gain correction initial value g11 as the gain correction value. The CPU 20 reads the value q10 of the output Q of the AD converter 19 at this time and stores it in the EEPROM 21.
[0048]
Next, with the offset correction value set to the offset correction initial value, the reference medium A is inserted between the reference roller 11 and the thickness detection roller 12, and the gain correction value set in the DA converter 18 is changed to change the AD correction value. A gain correction value g12 at which the output Q of the converter 19 becomes a predetermined value q12 is obtained, and the value q12 and the gain correction value g12 are stored in the EEPROM 21.
[0049]
Then, while maintaining the offset correction initial value and the gain correction value g12, the reference medium B is similarly inserted between the reference roller 11 and the thickness detection roller 12, and the value q13 of the output Q of the AD converter 19 is set. The value q13 is read and stored in the EEPROM 21.
Here, the thickness d of the reference medium A_AAnd the thickness d of the reference medium B_BDifference Δd from
Δd = d_A-D_B
Assuming that the change amount of the output Q of the AD converter 19 determined in accordance with is Δq, the desired gain correction value gy is calculated from the values q12, q13 and the gain correction value g12 by the following equation.
[0050]
gy = g12 × Δq / (q13−q12)
That is, when the gain correction value set in the DA converter 18 is gy, the thickness d of the reference medium A_AAnd the thickness d of the reference medium B_BIf the difference between the two is Δd, the change amount of the output Q of the AD converter 19 is Δq.
Next, the reference medium A is inserted between the reference roller 11 and the thickness detection roller 12, the desired gain correction value gy is set in the DA converter 18, the offset correction value is changed, and the AD converter 19 The output Q is set to a predetermined value qp within a specified range. Then, the CPU 20 calculates an offset correction value f5 when the output Q of the AD converter 19 becomes the value qp, and sets the offset correction value f5 in the DA converter 16. At this time, the CPU 20 stores the offset correction value f5 in the EEPROM 21 together with the value qp of the output Q of the AD converter 19.
[0051]
Subsequently, the reference medium A is removed from between the reference roller 11 and the thickness detection roller 12 while maintaining the offset correction value f5 and the gain correction value gy. The CPU 20 stores the value M of the output Q of the AD converter 19 at this time in the EEPROM 21 as an output target value.
Next, when calculating an eccentricity correction value, the reference medium E is inserted between the reference roller 11 and the thickness detection roller 12, and a transport system such as a motor (not shown) is driven to move the reference medium E at a constant speed. Transport. Then, after detecting a pulse in the output signal of the photo interrupter 22, the CPU 20 takes T rotations (T is a positive integer) for each timing t1 to tn obtained by equally dividing one rotation into n, and each n × T The values q (1),..., Q (n), q (n + 1),..., Q (n × T) of the output Q of the AD converter 19 are read at each timing.
[0052]
  FIG. 7 is a diagram showing the timing in the first embodiment of the present invention.
  In the figure, SG1 is an output signal of the photo interrupter 22 (FIG. 2), PL is a pulse generated in the output signal SG1, t1, t2,..., 2n + 1 are timings.
  Here, when the reference roller 11 and the thickness detection roller 12 are rotated once, values q (1),..., Q (n) of the output Q of the AD converter 19 can be obtained for n timings. When the reference roller 11 and the thickness detection roller 12 are at the same rotational position, the value q (k) (k = 1, 2,..., N) and the value q (k + n × (U−1)) (U = 1, 2, ...)).
[0053]
Therefore, when the reference roller 11 and the thickness detection roller 12 are at the same rotational position, the average of the values q (k), q (k + n),..., Q (k + n × (U−1)) of the output Q. Value q_PA(K)

become.
[0054]
From this, the eccentricity correction value H (k) (k = 1, 2,..., N) when the reference roller 11 and the thickness detection roller 12 are at the respective rotational positions is calculated using the following equation.
H (k) = q_PA-Q_PA(K)
Next, a method for detecting the thickness of a paper sheet by the paper thickness detecting device having the above-described configuration will be described.
[0055]
The CPU 20 sets the gain correction value gy in the DA converter 18 and sets it in the DA converter 16 before driving the transport system and inserting paper sheets between the reference roller 11 and the thickness detection roller 12. The offset correction value to be changed is changed so that the output Q of the AD converter 19 becomes the output target value M. Next, the CPU 20 detects the pulse PL in the output signal SG1 every time the reference roller 11 and the thickness detection roller 12 make one rotation, and prepares to perform eccentricity correction at each timing.
[0056]
Subsequently, when the sheet reaches between the reference roller 11 and the thickness detection roller 12, the value Q (t) (t = 1, 2,..., N) of the output Q of the AD converter 18 at the timing t. Start eccentricity correction. In this case, the value after the eccentricity correction is q^*If (t), the value q^*(T) (t = 1, 2,..., N) can be calculated using the following equation.
[0057]
q^*(T) = q (t) + H (t)
Therefore, the value q^*Based on (t), it is possible to detect the number of transported paper sheets, heavy running, tapes attached to the paper sheets, and the like.
Thus, the thickness d of the reference medium A_AAnd the thickness d of the reference medium B_BSince the gain correction value gy is calculated based on the ratio of the change amount Δq of the output Q caused by the difference Δd to the predetermined value qp, the gain correction value gy can be accurately detected. Therefore, in detecting the thickness of the paper sheet, detection accuracy on the order of microns can be obtained.
[0058]
  Next, a second embodiment of the present invention will be described.
  FIG. 8 is an explanatory diagram for calculating the gain correction value and the offset correction value in the second embodiment of the present invention. In the figure, the horizontal axis represents the state of the reference medium, and the vertical axis represents the output Q of the AD converter 19 (FIG. 2).
  In this case, in order to calculate the gain correction value and the offset correction value, two reference media A and B having different thicknesses and a reference medium E having the same thickness as the reference medium A are used. The reference medium A is made thinner than the reference medium B.
[0059]
First, in a state in which the reference media A and B are not inserted between the reference roller 11 and the thickness detection roller 12, the offset correction value is changed by the DA converter 16, and the output Q of the AD converter 19 is within a predetermined range. The offset correction value at that time is set as an offset correction initial value. At this time, the DA converter 18 sets an arbitrary gain correction initial value g11 as the gain correction value. The CPU 20 reads the value q10 of the output Q of the AD converter 19 at this time and stores it in the EEPROM 21.
[0060]
Next, with the offset correction value set to the offset correction initial value, the reference medium B is inserted between the reference roller 11 and the thickness detection roller 12, and the gain correction value set in the DA converter 18 is changed, and AD A gain correction value g12 at which the output Q of the converter 19 becomes a predetermined value q22 is obtained, and the value q22 and the gain correction value g12 are stored in the EEPROM 21.
[0061]
Then, while maintaining the offset correction initial value and the gain correction value g12, the reference medium A is similarly inserted between the reference roller 11 and the thickness detection roller 12, and the value q23 of the output Q of the AD converter 19 is set. The value q23 is read and stored in the EEPROM 21.
Here, the thickness d of the reference medium A_AAnd the thickness d of the reference medium B_BDifference Δd from
Δd = d_A-D_B
Assuming that the amount of change in the output Q of the AD converter 19 determined in accordance with is Δq, the desired gain correction value gy is calculated by the following equation using the values q22 and q23 and the gain correction value g12.
[0062]
gy = g12 × Δq / (q22−q23)
That is, when the gain correction value set in the DA converter 18 is gy, the thickness d of the reference medium A_AAnd the thickness d of the reference medium B_BIf the difference between the two is Δd, the change amount of the output Q of the AD converter 19 is Δq.
Next, with the reference medium A being inserted between the reference roller 11 and the thickness detection roller 12, the intended gain correction value gy is set in the DA converter 18 and the offset correction value is changed to change the AD converter 19 Is set to a predetermined value qp within a specified range. Then, the CPU 20 calculates an offset correction value f5 when the output Q becomes the value qp, and sets the offset correction value f5 in the DA converter 16. At this time, the CPU 20 stores the offset correction value f5 in the EEPROM 21 together with the value qp of the output Q of the AD converter 19.
[0063]
Subsequently, the reference medium A is removed from between the reference roller 11 and the thickness detection roller 12 while maintaining the offset correction value f5 and the gain correction value gy. The CPU 20 stores the value M of the output Q of the AD converter 19 at this time in the EEPROM 21 as an output target value.
Next, when calculating the eccentricity correction value, a reference medium E having the same thickness as the reference medium A is inserted between the reference roller 11 and the thickness detection roller 12, and a conveyance system such as a motor (not shown) is driven. The reference medium E is conveyed at a constant speed. Then, the CPU 20 detects the pulse PL in the output signal SG1 (FIG. 7) of the photo interrupter 22, and then converts the timings t1 to tn obtained by equally dividing one rotation into n parts by T rotations (T is a positive integer). ), And the values q (1),..., Q (n), q (n + 1),..., Q (n × T) of the output Q of the AD converter 19 are read at n × T timings.
[0064]
Here, when the reference roller 11 and the thickness detection roller 12 are rotated once, the values q (1),..., Q (n) of the output Q can be obtained for n timings. 11 and the thickness detection roller 12 are at the same rotational position, the value q (k) (k = 1, 2,..., N) and the value q (k + n × (U−1)) (U = 1, 2) , ...,) is the same.
[0065]
Therefore, when the reference roller 11 and the thickness detection roller 12 are at the same rotational position, the average of the values q (k), q (k + n),..., Q (k + n × (U−1)) of the output Q. Value q_PA(K)

become.
[0066]
From this, the eccentricity correction value H (k) (k = 1, 2,..., N) when the reference roller 11 and the thickness detection roller 12 are at the respective rotational positions is calculated using the following equation.
H (k) = q_PA-Q_PA(K)
Next, a method for detecting the thickness of a paper sheet by the paper thickness detecting device having the above-described configuration will be described.
[0067]
The CPU 20 sets the gain correction value gy in the DA converter 18 and sets it in the DA converter 16 before driving the transport system and inserting paper sheets between the reference roller 11 and the thickness detection roller 12. The offset correction value to be changed is changed so that the output Q of the AD converter 19 becomes the output target value M. Next, the CPU 20 detects the pulse PL in the output signal SG1 every time the reference roller 11 and the thickness detection roller 12 make one rotation, and prepares to perform eccentricity correction at each timing.
[0068]
Subsequently, when the sheet reaches between the reference roller 11 and the thickness detection roller 12, the value Q (t) (t = 1, 2,..., N) of the output Q of the AD converter 18 at the timing t. Start eccentricity correction. In this case, the value after the eccentricity correction is q^*If (t), the value q^*(T) (t = 1, 2,..., N) can be calculated using the following equation.
[0069]
q^*(T) = q (t) + H (t)
Therefore, the value q^*Based on (t), it is possible to detect the number of transported paper sheets, heavy running, tapes attached to the paper sheets, and the like.
Thus, since the reference medium B and the reference medium A are inserted in the order of thickness between the reference roller 11 and the thickness detection roller 12, the number of insertions of the reference media A and B can be reduced. .
[0070]
  Therefore, not only can the work for calculating the gain correction value gy and the offset correction value f5 be simplified, but also the time for calculation can be shortened. In addition, since the order of inserting the reference media A and B is not mistaken, the gain correction value gy and the offset correction value f5 can be accurately calculated.
  Next, a third embodiment of the present invention will be described.
[0071]
  FIG. 9 is an explanatory diagram for calculating the gain correction value and the offset correction value in the third embodiment of the present invention, and FIG. 10 is for describing each term of the normal equation in the third embodiment of the present invention. It is a matrix. In FIG. 9, the horizontal axis represents the state of the reference medium, and the vertical axis represents the output Q of the AD converter 19 (FIG. 2).
  In this case, in order to calculate the gain correction value and the offset correction value, J reference media F having different thicknesses are used._r(R = 1, 2,..., J) and the reference medium F in ascending order of thickness.₁, F₂... F_JAnd Reference medium F_rD in ascending order of thickness₁, D₂, ..., d_JAnd
[0072]
First, the reference medium F_rIs not inserted between the reference roller 11 and the thickness detection roller 12, the offset correction value is changed by the DA converter 16, and the output Q of the AD converter 19 is a predetermined value q within a specified range._F0And the offset correction value at that time is set as an offset correction initial value. At this time, the DA converter 18 sets an arbitrary gain correction initial value g11 as the gain correction value. Then, the CPU 20 determines the value q of the output Q of the AD converter 19 at this time._F0Is stored in the EEPROM 21.
[0073]
Next, the reference medium F is interposed between the reference roller 11 and the thickness detection roller 12 while keeping the offset correction value at the offset correction initial value.₁Is inserted, the gain correction value set in the DA converter 18 is changed, and the output Q of the AD converter 19 is set to a predetermined value q._F1A gain correction value g12 to be_F1The gain correction value g12 is stored in the EEPROM 21.
[0074]
Then, while maintaining the offset correction initial value and the gain correction value g12, the reference medium F is placed between the reference roller 11 and the thickness detection roller 12 in the same manner.₂And the value q of the output Q of the AD converter 19_F2And the value q_F2Is stored in the EEPROM 21. Thereafter, the reference medium F is placed between the reference roller 11 and the thickness detection roller 12 while maintaining the offset correction initial value and the gain correction value g12._Three, F_Four... F_J, And the value q of the output Q of the AD converter 19_F3, Q_F4, ..., q_FJIs stored in the EEPROM 21.
[0075]
Next, an approximate expression representing the relationship between the thickness X and the output Y is obtained by the method of least squares.
Here, the relationship between the thickness X and the output Y is
Y = a + bX
, J conditional equations for obtaining constants a and b
Y_i= A + bX_i  (I = 1, 2,..., J)
The normal equation is obtained according to the matrix of FIG.
a × J + b × Σd_r= Σq_Fr                            ...... (1)
a × Σd_r+ B × Σ (d_rXd_r) = Σ (d_r× q_Fr) (2)
Can be expressed as Therefore, the constants a and b can be obtained by solving the simultaneous equations of the expressions (1) and (2).
[0076]
In this way, an approximate expression representing the relationship between the thickness X and the output value Y by the least square method can be obtained. Reference medium F by the approximate expression₁And reference medium F_JOutput value q_F1 ^*, Q_FJ ^*Is
q_F1 ^*= A + b × d₁
q_FJ ^*= A + b × d_J
become. And the reference medium F₁, F_JEach thickness d₁, D_JDifference Δd_FThe amount of change in the value of the output Q of the AD converter 19 determined corresponding to_FThen, based on the gain correction value g11, the intended gain correction value gz can be calculated using the following equation.
[0077]
gz = g11 × Δq_F/ (Q_FJ ^*-Q_F1 ^*)
That is, when the gain correction value set in the DA converter 18 is gz, the thickness d of the reference medium A_AAnd the thickness d of the reference medium B_BΔd is_FThe change amount of the output Q of the AD converter 19 is Δq_FShow that.
Next, the reference medium F₁The desired gain correction value gz is set in the DA converter 18 while the signal is inserted, and the offset correction value is changed so that the output Q of the AD converter 19 becomes a predetermined value qp within a specified range. Then, the CPU 20 calculates an offset correction value f6 when the output Q becomes the value qp, and sets the offset correction value f6 in the DA converter 16. At this time, the CPU 20 stores the offset correction value f6 in the EEPROM 21 together with the value qp of the output Q of the AD converter 19.
[0078]
Subsequently, while maintaining the offset correction value f6 and the gain correction value gz, the reference medium F is interposed between the reference roller 11 and the thickness detection roller 12.₁Remove. The CPU 20 stores the value M of the output Q of the AD converter 19 at this time in the EEPROM 21 as an output target value.
Next, when calculating the eccentricity correction value, a reference medium E having the same thickness as the reference medium A is inserted between the reference roller 11 and the thickness detection roller 12, and a conveyance system such as a motor (not shown) is driven. The reference medium E is conveyed at a constant speed. Then, the CPU 20 detects the pulse PL in the output signal SG1 (FIG. 7) of the photo interrupter 22, and then converts each timing t1 to tn obtained by equally dividing one rotation into n parts by T rotation (T is a positive integer). The values q (1),..., Q (n), q (n + 1),..., Q (n × T) of the AD converter 19 are read at n × T timings.
[0079]
Here, when the reference roller 11 and the thickness detection roller 12 are rotated once, the values q (1),..., Q (n) of the output Q can be obtained for n timings. 11 and the thickness detection roller 12 are at the same rotational position, the value q (k) (k = 1, 2,..., N) and the value q (k + n × (U−1)) (U = 1, 2) , ...,) is the same.
[0080]
Therefore, when the reference roller 11 and the thickness detection roller 12 are at the same rotational position, the average of the values q (k), q (k + n),..., Q (k + n × (U−1)) of the output Q. Value q_PA(K)

become.
[0081]
From this, the eccentricity correction value H (k) (k = 1, 2,..., N) when the reference roller 11 and the thickness detection roller 12 are at the respective rotational positions is calculated using the following equation.
H (k) = q_PA-Q_PA(K)
Next, a method for detecting the thickness of a paper sheet by the paper thickness detecting device having the above-described configuration will be described.
[0082]
Prior to driving the transport system and inserting paper sheets between the reference roller 11 and the thickness detection roller 12, the CPU 20 sets the gain correction value gz in the DA converter 18 and sets it in the DA converter 16. The offset correction value to be changed is changed to set the output Q of the AD converter 19 to the output target value M. Next, the CPU 20 detects the pulse PL in the output signal SG1 every time the reference roller 11 and the thickness detection roller 12 make one rotation, and prepares to perform eccentricity correction at each timing.
[0083]
Subsequently, when the sheet reaches between the reference roller 11 and the thickness detection roller 12, the value Q (t) (t = 1, 2,..., N) of the output Q of the AD converter 19 at the timing t. Start eccentricity correction. In this case, the value after the eccentricity correction is q^*If (t), the value q^*(T) (t = 1, 2,..., N) can be calculated using the following equation.
[0084]
q^*(T) = q (t) + H (t)
Therefore, the value q^*Based on (t), it is possible to detect the number of transported paper sheets, heavy running, tapes attached to the paper sheets, and the like.
Thus, J reference media F having different thicknesses between the reference roller 11 and the thickness detection roller 12 are used._r(R = 1, 2,..., J) are inserted in ascending order, and the value q of the output Q of each AD converter 19_Fr(R = 1, 2,..., J) is read and the value q_Fr, An approximate expression representing the relationship between the thickness X and the output Y is obtained by the least square method, and the gain correction value gz is calculated based on the approximate expression._rThe gain correction value gz can be calculated accurately as much as the number of.
[0085]
  Next, a fourth embodiment of the present invention will be described.
  FIG. 11 is an explanatory diagram for calculating a gain correction value and an offset correction value in the fourth embodiment of the present invention. In the figure, the horizontal axis represents the state of the reference medium, and the vertical axis represents the output Q of the AD converter 19 (FIG. 2).
  In this case, in order to calculate the gain correction value and the offset correction value, two reference media A and B having different thicknesses and a reference medium E having the same thickness as the reference medium A are used. The reference medium A is made thinner than the reference medium B.
[0086]
First, in a state in which the reference media A and B are not inserted between the reference roller 11 and the thickness detection roller 12, the offset correction value is changed by the DA converter 16, and the output Q of the AD converter 19 is within a predetermined range. The offset correction value at that time is set as an offset correction initial value. At this time, the DA converter 18 sets an arbitrary gain correction initial value g11 as the gain correction value. Then, the CPU 20 reads the value Q10 of the output Q of the AD converter 19 and the reference roller 11 at this time, and stores them in the EEPROM 21.
[0087]
Next, the reference medium B is inserted between the reference roller 11 and the thickness detection roller 12 while the offset correction value is kept at the offset correction initial value, and the reference medium is driven by driving a transport system such as a motor (not shown). Transport B. When there is the reference medium B between the reference roller 11 and the thickness detection roller 12, the CPU 20 reads a plurality of outputs Q of the AD converter 19 at a predetermined interval, and averages the outputs Q. q_A1Is calculated. Then, this operation is repeated a plurality of times, and the gain correction value set in the DA converter 18 is changed at the same time, and the average value q_A1The gain correction value g12 that becomes a predetermined value q22 is obtained, and the value q22 and the gain correction value g12 are stored in the EEPROM 21.
[0088]
Next, while maintaining the offset correction initial value and the gain correction value g12, the reference medium A is inserted between the reference roller 11 and the thickness detection roller 12 and the transport system is driven to thereby drive the reference medium A. Transport. When there is the reference medium A between the reference roller 11 and the thickness detection roller 12, the CPU 20 reads a plurality of outputs Q of the AD converter 19 at a predetermined interval, and the average value of the outputs Q q_A2Is calculated. At this time, the average value q_A2Is stored in the EEPROM 21.
[0089]
Here, the thickness d of the reference medium A_AAnd the thickness d of the reference medium B_BDifference Δd from
Δd = d_A-D_B
Assuming that the change amount of the output Q of the AD converter 19 determined in accordance with is Δq, the value q22, the average value q_A2Then, the desired gain correction value ga is calculated from the gain correction value g12 using the following equation.
[0090]
ga = g12 × Δq / (q22−q_A2)
That is, the gain correction value set in the DA converter 18 is gy._AThe thickness d of the reference medium A_AAnd the thickness d of the reference medium B_BIf the difference between the two is Δd, the change amount of the output Q of the AD converter 19 is Δq.
Next, the reference medium A is inserted between the reference roller 11 and the thickness detection roller 12, and the desired gain correction value gy is supplied to the DA converter 18 while maintaining the offset correction value._AAre read, and a plurality of outputs Q of the AD converter 19 are read at a predetermined interval, and an average value q of the outputs Q is read._A3Is calculated. Then, this operation is repeated a plurality of times, and the offset correction value set in the DA converter 16 is changed at the same time, and the average value q_A3Is the predetermined value qp_AOffset correction value f_AFor the average value q_A3And offset correction value f_AIs stored in the EEPROM 21. The above processing is completed while the reference medium A is between the reference roller 11 and the thickness detection roller 12.
[0091]
  Subsequently, the offset correction value f_AAnd gain correction value gy_AIn this manner, a plurality of outputs Q of the AD converter 19 are read at a predetermined interval and the average value q of the outputs Q is maintained._A4Is calculated. The CPU 20 calculates the average value q_A4Are stored in the EEPROM 21 as output target values.
  Hereinafter, as in the first embodiment, a reference medium E having the same thickness as that of the reference medium A is inserted between the reference roller 11 and the thickness detection roller 12, and the conveyance system is driven so that the reference medium E is inserted. The values Q (1),..., Q (n), q (n + 1),..., Q (n × T) of the output Q of the AD converter 19 are read at a constant speed.
[0092]
When the reference roller 11 and the thickness detection roller 12 are at the same rotational position, the average of the values q (k), q (k + n),..., Q (k + n × (U−1)) of the output Q. Value q_PA(K) can be calculated, and the eccentricity correction value H (k) (k = 1, 2,..., N) when the reference roller 11 and the thickness detection roller 12 are at the respective rotational positions is expressed by the following equation. Can be used to calculate.
[0093]
H (k) = q_PA-Q_PA(K)
Therefore, when the thickness of the paper sheet is detected by the paper thickness detecting device having the above-described configuration, the value q of the output Q after the eccentricity correction is performed.^*(T) can be calculated using the following equation.
q^*(T) = q (t) + H (t)
In this way, the gain correction value gy can be obtained simply by transporting two reference media A and B having different thicknesses between the reference roller 11 and the thickness detection roller 12 one by one._AAnd offset correction value f_ATherefore, the operation can be simplified.
[0094]
  Next, a fifth embodiment of the present invention will be described.
  FIG. 12 is a diagram showing a reference medium according to the fifth embodiment of the present invention.
  In the figure, K is a reference medium, and two reference media K having different thicknesses.₁, K₂Is formed by sticking at the end. In this case, the reference medium K₂Is the reference medium K₁Made thinner.
[0095]
Then, the reference medium K is changed to the reference medium K.₁By carrying the front side forward, the gain correction value gy is obtained in the same manner as in the seventh embodiment._A(FIG. 11), offset correction value f_AAnd the eccentricity correction value H (k) can be calculated.
In this case, since only one reference medium K needs to be inserted and conveyed, the operation can be further simplified.
[0096]
  Next, a sixth embodiment of the present invention will be described.
  In this case, after calculating the offset correction value f5 and the gain correction value gy (FIG. 6) by the same method as in the first embodiment, the reference medium E having the same thickness as the reference medium A is calculated in calculating the eccentricity correction value. Is inserted between the reference roller 11 and the thickness detection roller 12, and a transport system such as a motor (not shown) is driven to transport the reference medium E in the first and second directions at a constant speed. For this purpose, the transport system can be driven in the forward and reverse directions.
[0097]
Then, when the reference medium E is conveyed in the first direction, values Q (1), q (n), q (n + 1), q, (n × T) of the output Q of the AD converter 19 are obtained. Reading, the average of the values q (k), q (k + n),..., Q (k + n × (U−1)) of the output Q when the reference roller 11 and the thickness detection roller 12 are at the same rotational position. Value q_PA(K) is calculated. Further, the eccentricity correction value H (k) when the reference roller 11 and the thickness detection roller 12 are at the respective rotational positions.
H (k) = q_PA-Q_PA(K)
And the value q of the output Q after the eccentricity correction is performed^*(T)
q^*(T) = q (t) + H (t)
Is calculated.
[0098]
Next, values ρ (1),..., Ρ (n), ρ (n + 1),..., Ρ (n × T) of the output Q of the AD converter 19 when the reference medium E is conveyed in the second direction. , And when the reference roller 11 and the thickness detection roller 12 are at the same rotational position, the values ρ (k), ρ (k + n),..., Ρ (k + n × (U−1)) Average value ρ_PA(K) is calculated. Further, the eccentricity correction value H (k) when the reference roller 11 and the thickness detection roller 12 are at the respective rotational positions.
H (k) = ρ_PA−ρ_PA(K)
And the value Q of the output Q after the eccentricity correction is performed^*(T)
ρ^*(T) = ρ (t) + H (t)
Is calculated.
[0099]
Therefore, when detecting the thickness of the paper sheet, when the paper sheet is conveyed in the first direction,
q^*(T) = q (t) + H (t)
When the paper sheet is conveyed in the second direction using
ρ^*(T) = ρ (t) + H (t)
The eccentricity correction can be performed using the following equation.
[0100]
As described above, since the eccentricity correction value H (k) can be varied in accordance with the method of conveying the paper sheet, the thickness of the paper sheet can be accurately detected.
In addition, this invention is not limited to the said embodiment, It can change variously based on the meaning of this invention, and does not exclude them from the scope of the present invention.
[0101]
【The invention's effect】
  As described above in detail, according to the present invention, in the correction method of the paper sheet thickness detection device, a plurality of reference media having different thicknesses are inserted between the reference roller and the thickness detection roller. Detecting a displacement of the thickness detection roller, generating an output corresponding to the displacement by an output generating means, and based on a difference in thickness of the reference medium and a change amount of the output generated based on the difference. A gain correction value is calculated, and the output is adjusted according to the gain correction value.
[0107]
In this case, the gain correction value is calculated based on the difference between the thicknesses of the two types of reference media and the amount of change in output caused by the difference in thickness, so that the gain correction value can be accurately detected. Can do. Therefore, in detecting the thickness of the paper sheet, detection accuracy on the order of microns can be obtained.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram for calculating a gain correction value and an offset correction value according to a first embodiment of a technique that is a premise for explaining the present invention;
FIG. 2 is a schematic view of a conventional paper sheet thickness detection device.
FIG. 3 is an explanatory diagram for calculating a conventional gain correction value and offset correction value;
FIG. 4 is an explanatory diagram for calculating a conventional eccentricity correction value.
FIG. 5 is an explanatory diagram for calculating an eccentricity correction value in the first embodiment of the technology that is the premise for explaining the present invention;
FIG. 6 is an explanatory diagram for calculating a gain correction value and an offset correction value according to the first embodiment of the present invention.
FIG. 7 is a diagram showing timing in the first embodiment of the present invention.
FIG. 8 is an explanatory diagram for calculating a gain correction value and an offset correction value according to the second embodiment of the present invention.
FIG. 9 is an explanatory diagram for calculating a gain correction value and an offset correction value according to the third embodiment of the present invention.
FIG. 10 is a matrix for explaining each term of a normal equation in the third embodiment of the present invention.
FIG. 11 is an explanatory diagram for calculating a gain correction value and an offset correction value according to the fourth embodiment of the present invention.
FIG. 12 is a diagram illustrating a reference medium according to a fifth embodiment of the present invention.
[Explanation of symbols]
11 Standard roller
12 Thickness detection roller
16, 18 DA converter
19 AD converter

Claims (8)

(A) by inserting a plurality of reference media of different thicknesses between the criteria roller and thickness detection roller, detects the displacement of the thickness detection roller,
(B) generating an output corresponding to the displacement by the output generating means;
(C) the difference between the thickness of the reference medium, and calculates a gain correction value based on the change amount of the output produced based on the difference, paper and adjusts the output in accordance with the gain correction value Correction method of leaf thickness detection apparatus. Correction method of the sheet thickness detecting device according to claim 1, 2 types of reference media of different thicknesses are inserted between the reference roller and the thickness detection roller. The correction method of the paper sheet thickness detection apparatus according to claim 2 , wherein the reference medium is inserted first with a thick reference medium and then with a thin reference medium. (A) Three or more types of reference media having different thicknesses are inserted between the reference roller and the thickness detection roller,
(B) The correction method of the paper sheet thickness detection device according to claim 1 , wherein the gain correction value is calculated by a least square method based on the output generated by the output generation means. Wherein the plurality of reference media, the correction method of a sheet thickness detecting device according to claim 1 which is conveyed between the reference roller and the thickness detection roller. The correction method of the paper sheet thickness detection device according to claim 5 , wherein ends of the plurality of reference media are attached to each other. Wherein the plurality of reference media, first between the reference roller and the thickness detection roller, the correction method of a sheet thickness detecting device according to claim 5 which is conveyed in the second direction. (A) means for detecting displacement of the thickness detection roller when a plurality of reference media having different thicknesses are inserted between the reference roller and the thickness detection roller;
(B) output generating means for generating an output corresponding to the displacement;
(C) having a means for calculating a gain correction value based on a difference in thickness of the reference medium and a change amount of the output generated based on the difference, and adjusting the output in accordance with the gain correction value. Characteristic sheet thickness detection device.

Images