JP4487114B2

JP4487114B2 - Method for calculating feed pitch of strip rubber member, method for producing tire constituent member using the same, and program

Info

Publication number: JP4487114B2
Application number: JP2001057827A
Authority: JP
Inventors: 一則小野; 和夫宮坂
Original assignee: Yokohama Rubber Co Ltd
Current assignee: Yokohama Rubber Co Ltd
Priority date: 2001-03-02
Filing date: 2001-03-02
Publication date: 2010-06-23
Anticipated expiration: 2021-03-02
Also published as: JP2002254531A

Description

【０００１】
【発明の属する技術分野】
本発明は、長尺状のストリップゴム部材をドラム支持装置のドラム周上で部分的に重ねつつ巻き回して所定の断面形状を有するタイヤ構成部材を成形する際の、ストリップゴム部材の送りピッチを自動的に算出し、さらに、最適に算出するストリップゴム部材の送りピッチ算出方法、これを用いたタイヤ構成部材の作製方法およびプログラムに関する。
【０００２】
【従来の技術】
現在、空気入りタイヤは、ドラム支持装置のドラム表面上にカーカス部材やビード部材等を配置し、さらに、この上層にサイドゴム部材やトレッドゴム部材等の所定の断面形状を成したゴム部材を巻き付け貼り合わせることによって、加硫前の生タイヤを成形している。
【０００３】
一方、このような所定の断面形状を有するゴム部材からなるタイヤ構成部材を、長尺状のストリップゴム部材をドラム支持装置のドラム周上で部分的に重ねつつスパイラル状に巻き回すことによって、成形するタイヤストリップワインディング工法が提案されている。
従来のタイヤ構成部材の製造方法は、ゴム部材からなるタイヤ構成部材を、所定の開口形状を持った開口部を有する押出機の金型からゴム部材を押し出すことによって作製する方法であり、タイヤ構成部材の断面形状が変わる度に煩雑な金型を作製する必要があるのに対し、上記ストリップワインディング工法は、所望の断面形状を持つタイヤ構成部材を、ドラム支持装置の上で直接成形して生タイヤを作製することができるため、工数やコストの面で大きな低減が期待されている。
【０００４】
このようなストリップワインディング工法では、長尺状のストリップゴム部材を部分的に重ねつつスパイラル状に巻き回してタイヤ構成部材を成形するため、タイヤ構成部材の断面の目標形状からずれないように正確に成形することが必要である。従って、長尺状のストリップゴム部材を部分的に重ねつつスパイラル状に巻き回すストリップゴム部材の送りピッチを、目標形状とするタイヤ構成部材の断面の設計形状に一致するように設定することが重要な課題となっている。特に、ストリップゴム部材を巻き回して作られるタイヤ構成部材の断面形状の設計形状に対する誤差の程度が、タイヤ性能の良し悪し、例えば、車両に装着した際に発生するタイヤの振動成分（ユニフォーミティ成分）の大小に大きな影響を与える。従って、長尺状のストリップゴム部材の送りピッチをどのように設定するかが、タイヤ性能の良し悪しを決定する一要因にもなっている。
【０００５】
【発明が解決しようとする課題】
ところが、上記ストリップワインディング工法では、特開昭６１−１３００３０号公報に記載されているように、ドラム支持装置上でストリップゴムを巻き回しつつ、巻き厚センサー等を用いてタイヤ構成部材の厚みを測定してタイヤ構成部材を成形し、定められた厚みになるとドラム支持装置のドラム軸方向にストリップゴム部材を送っていた。
また、このストリップゴム部材をドラム上に巻き付ける周回毎の送りピッチを予め設定する場合、この送りピッチの設定は作業者の技能や経験に頼って手作業で行われており、成形されるタイヤ構成部材の断面形状は、粗荒な断面形状となる場合が多かった。そのため、目標形状に近い送りピッチを設定するには多くの試行錯誤の時間を必要とし、たとえ比較的正確に目標形状に近くなるように送りピッチを設定できたとしても、この送りピッチが最適な送りピッチとなっているかどうかについても判断することができないといった問題があった。
【０００６】
そこで、本発明は、上記問題点を解決すべく、長尺状のストリップゴム部材をドラム支持装置のドラム周上で部分的に重ねつつ巻き回してタイヤ構成部材を成形する際のストリップゴム部材の送りピッチを自動的に算出し、さらには、ストリップゴム部材を巻き回して得られる断面形状の、設計形状に対する誤差が最も小さくなる最適な送りピッチに調整するストリップゴム部材の送りピッチ算出方法、およびこれを用いたタイヤ構成部材の作製方法および上記送りピッチを算出するコンピュータが実行可能なプログラムを提供することを目的とする。
【０００７】
【課題を解決するための手段】
上記目的を達成するために、本発明は、長尺状のストリップゴム部材をドラム支持装置のドラム幅方向に沿って少なくとも部分的に重ねつつドラム周上で巻き回して所定の断面形状を有するタイヤ構成部材を作製する際の、前記ドラム支持装置のドラム幅方向に沿ったストリップゴム部材の送りピッチを算出する方法であって、
送りピッチ算出部が、前記送りピッチを、ストリップゴム部材の巻き回しによって得られる、ストリップゴム部材のドラム周方向に沿った平均周状厚みに基づいて算出し、ストリップゴム部材の巻き回し開始端から、ストリップゴム部材を巻き回す周回毎に算出する送りピッチ算出工程と、最適送りピッチ算出部が、最適な送りピッチを算出する最適送りピッチ算出工程と、を有し、
前記最適送りピッチ算出工程は、
前記送りピッチ算出工程を少なくとも経て得られた複数の送りピッチからなる送りピッチ群を用いて複数組の送りピッチ群を生成し、これらの送りピッチ群を解集合として初期設定する初期解集合設定工程と、
送りピッチ群の送りピッチから得られる平均周状厚みを合計加算することによって得られる合計加算厚みの分布形状の、前記タイヤ構成部材の断面形状に対する誤差をフィットネス関数として設定するフィットネス関数設定工程と、
送りピッチ群の解集合から、前記フィットネス関数に基づいて、送りピッチ群の対を複数組設定する対設定工程と、
この設定された送りピッチ群の対の各々が有する送りピッチ情報の一部を所定の確率でお互いに交叉させる交叉工程と、
この交叉工程で交叉して得られた送りピッチ情報の一部を所定の確率で変更する変更工程と、
前記交叉工程前の送りピッチ群の解集合を、前記変更工程を経て得られた送りピッチ群に置き換える置き換え工程とを備え、
この置き換え工程で得られた解集合のすべての送りピッチ群に対して前記フィットネス関数を適用して得られる前記フィットネス関数の値の最小値が所定の値以下になるまで、前記対設定工程、前記交叉工程、前記変更工程および前記置き換え工程を繰り返すことによって、最適な送りピッチを算出することを特徴とするストリップゴム部材の送りピッチ算出方法を提供するものである。
【０００８】
ここで、前記送りピッチ算出工程は、ストリップゴム部材を巻き回す周回の順番に送りピッチを順次算出するのが好ましい。
また、前記送りピッチ算出工程は、
算出すべき送りピッチの巻き回し周回後のストリップゴム部材の断面内の、前記ドラム幅方向における巻き回し開始側の端点位置において、前記算出すべき送りピッチの算出前に既に算出された送りピッチから求まる平均周状厚みの累積加算厚みにストリップゴム部材の厚みを加算することによって得られる新たな厚みが、前記端点位置に対応した前記タイヤ構成部材の対応位置における断面の厚みに略等しくなるように、前記算出すべき送りピッチを算出するのが好ましい。その際、前記送りピッチ算出工程で前記累積加算厚みに加算するストリップゴム部材の厚みが、ストリップゴム部材の断面積および横幅と同じ断面積および横幅を持つ等価矩形断面形状の厚みであるのがよい。
【０００９】
また、上記ストリップゴム部材の送りピッチ算出方法は、前記送りピッチ算出工程で算出された送りピッチの各々を調整する第１の送りピッチ調整工程を有し、
この第１の送りピッチ調整工程は、調整すべき送りピッチを除き送りピッチの値をそのまま保持するとともに、前記調整すべき送りピッチの値を所定の範囲内で変化させることによって得られる合計加算厚みが、前記調整すべき送りピッチの巻き回し周回後のストリップゴム部材の断面内の、前記ドラム幅方向における巻き回し開始側の端点位置において、この端点位置に対応した前記タイヤ構成部材の対応位置における断面の厚みに略等しくなるように、前記調整すべき送りピッチを調整するのが好ましい。
【００１０】
さらに、上記ストリップゴム部材の送りピッチ算出方法は、前記第１の送りピッチ調整工程で調整された送りピッチの各々を調整する第２の送りピッチ調整工程を有し、
この第２の送りピッチ調整工程は、前記第１の送りピッチ調整工程で調整された送りピッチを、調整すべき送りピッチを除き送りピッチの値をそのまま保持するとともに、前記調整すべき送りピッチの値を所定の範囲内で変化させることによって得られる平均周状厚みの合計加算厚みの分布形状を、前記タイヤ構成部材の断面形状に略等しくするように、前記調整すべき送りピッチを調整するのが好ましい。その際、前記第２の送りピッチ調整工程は、ストリップゴム部材を巻き回す周回の順番に、あるいは周回の順番と逆の順番に、送りピッチを順次調整するのが好ましい。
【００１１】
さらに、上記ストリップゴム部材の送りピッチ算出方法は、前記第２の送りピッチ調整工程で調整された送りピッチの各々を、前記第２の送りピッチ調整工程で調整される送りピッチの調整の順番と逆の順番に調整する第３の送りピッチ調整工程を有し、
この第３の送りピッチ調整工程は、前記第２の送りピッチ調整工程で調整された送りピッチを、調整すべき送りピッチを除き送りピッチの値をそのまま保持するとともに、調整すべき送りピッチの値を所定の範囲内で変化させることによって得られる合計加算厚みの分布形状を、前記タイヤ構成部材の断面形状に略等しくするように、送りピッチを調整するのが好ましい。
【００１２】
また、前記第２の送りピッチ調整工程および前記第３の送りピッチ調整工程の少なくとも一方は、前記合計加算厚みの分布形状の、前記タイヤ構成部材の断面形状に対する最大誤差、および、前記合計加算厚みの分布形状の、前記タイヤ構成部材の断面形状に対する平均誤差、および、前記最大誤差と前記平均誤差との加重平均、の内の一つを用いて前記調整すべき送りピッチの調整を行うのが好ましい。
【００１４】
その際、前記対設定工程は、前記解集合からランダムに選択された２つの送りピッチ群の内、前記フィットネス関数の値が小さい方の送りピッチ群を、送りピッチ群の対の一方とし、さらに、前記解集合から新たにランダムに選択された２つの送りピッチ群の内、前記フィットネス関数の値が小さい方の送りピッチ群を送りピッチ群の対の他方とするのが好ましく、前記フィットネス関数は、前記合計加算厚みの分布形状の、タイヤ構成部材の断面形状に対する最大誤差、あるいは、前記合計加算厚みの分布形状の、タイヤ構成部材の断面形状に対する平均誤差と前記最大誤差との加重平均であるのが好ましい。
【００１５】
上記ストリップゴム部材の送りピッチ算出方法は、少なくとも長尺状のストリップゴム部材を複数本用いてドラム周上で巻き回す送りピッチを、ストリップゴム部材毎に算出してもよい。
この場合、タイヤ構成部材の断面形状を２分割した各々の分割断面形状に対してストリップゴム部材の送りピッチを算出する際、
前記分割断面形状の各々に対して設定される送りピッチから作られる各々の前記合計加算厚みの分布形状を１つにまとめた分布形状の、タイヤ構成部材の断面形状に対する最大誤差が最小になるように、タイヤ構成部材の断面形状を２分割する分割位置が設定されてもよく、
あるいは、前記分割断面形状の各々に対して設定される、ストリップゴム部材の巻き回しの順番に対する送りピッチが、高次多項式で最適に近似されるように分割位置が設定されてもよい。
【００１６】
また、本発明は、上記ストリップゴム部材の送りピッチ算出方法を用いて算出された送りピッチを用いて、タイヤ構成部材を作製することを特徴とするタイヤ構成部材の作製方法を提供するものである。
【００１７】
さらに、本発明は、長尺状のストリップゴム部材をドラム支持装置上で少なくとも部分的に重ねつつドラム周上で巻き回して所定の断面形状を有するタイヤ構成部材を作製する際の、前記ドラム支持装置のドラム幅方向に沿ったストリップゴム部材の送りピッチを求める方法をコンピュータに実行させるプログラムであって、
前記送りピッチを、前記ストリップゴム部材のドラム周方向に沿った平均周状厚みに基づいて、コンピュータの演算手段によって算出させ、前記ストリップゴム部材を巻き回す周回毎に、前記演算手段を用いて前記送りピッチを求める手順と、
この求められた複数の送りピッチからなる送りピッチ群を用いて複数組の送りピッチ群を生成し、これらの送りピッチ群を解集合として初期設定し、この解集合をコンピュータの記憶手段に記録させる初期解集合設定手順と、
送りピッチ群の送りピッチから得られる平均周状厚みを合計加算することによって得られる合計加算厚みの分布形状の、前記タイヤ構成部材の断面形状に対する誤差をフィットネス関数として設定するフィットネス関数設定手順と、
前記送りピッチ群の解集合から、前記フィットネス関数に基づいて、送りピッチ群の対を複数組設定し、この複数組の対を前記記憶手段に記録させる対設定手順と、
この設定された送りピッチ群の対の各々が有する送りピッチ情報の一部を、コンピュータの乱数発生手段を用いて、所定の確率でお互いに交叉させる交叉手順と、
この交叉手順で交叉して得られた送りピッチ情報の一部を、前記乱数発生手段を用いて、所定の確率で変更する変更手順と、
前記交叉手順が実行される前の送りピッチ群の解集合を、前記変更手順を経て得られた送りピッチ群に置き換え、この置き換えた解集合を前記記憶手段に記録させる置き換え手順とを有し、
この置き換え手順で得られた送りピッチ群の解集合に対して前記フィットネス関数を適用して得られる前記フィットネス関数の値の最小値が所定の値以下になるまで、前記対設定手順、前記交叉手順、前記変更手順および前記置き換え手順を繰り返すことによって、最適な送りピッチ群を算出することを特徴とするプログラムを提供するものである。
【００１９】
【発明の実施の形態】
以下、本発明のストリップゴム部材の送りピッチ算出方法およびこれを用いたタイヤ構成部材の作製方法およびプログラムについて、添付の図面に示される好適実施例を基に詳細に説明する。
【００２０】
図１（ａ）は、本発明のストリップゴム部材の送りピッチ算出方法を用いたタイヤ構成部材の作製方法を説明する説明図である。
タイヤ構成部材、例えばサイドゴム部材やトレッドゴム部材は、長尺状のストリップゴム部材Ｇを少なくとも部分的に重ねつつドラム周上で巻き回して作製される。すなわち、カーカス部材やビード部材等をドラム支持装置１０のドラム表面１２に巻き付けて得られるタイヤ中間組立体１４の上から、ストリップゴム部材Ｇを少なくとも部分的に重ねつつ、ドラム支持装置１０のドラム周上に巻き回し、図１（ｂ）に示す様に、ストリップゴム部材Ｇの輪郭形状が目標形状１６に略一致するようにタイヤ構成部材は成形される。ここで、ストリップゴム部材Ｇの巻き回しは、少なくとも１部分が重なっており、一方向に巻かれる。勿論、タイヤ構成部材の断面形状に応じて、ストリップゴム部材Ｇは、１つ前の巻き回しのストリップゴム部材Ｇの全体とドラム幅方向で重なってもよい。
【００２１】
ここで、ストリップゴム部材Ｇの断面形状１８は、図１（ｃ）に示す様に、矩形形状の上両端部が切り欠かれた形状となっており（図１（ａ），（ｂ）や図２のストリップゴム部材Ｇは簡略化のために矩形断面形状のストリップゴム部材として表している）、このストリップゴム部材Ｇをドラム支持装置１０のドラム幅方向（ｘ方向）に沿って巻き回す送りピッチＰ_n（ｎ＝１，２，・・・Ｎ）が、巻き回し毎に設定されている。すなわち、図１（ａ）に示す様に、ストリップゴム部材Ｇをスパイラル状にｘ方向に巻き回していく場合、ストリップゴム部材Ｇの巻き回し開始端から３６０度ずつ巻き回す周回毎（ｎ＝１，２，・・・Ｎ）にストリップゴム部材Ｇのｘ方向への送り量であるｎ巻目の送りピッチＰ_n（ｎ＝１，２，・・・Ｎ）が設定されている。従って、ドラム表面１２に沿って巻き付けられたストリップゴム部材Ｇを展開した図２上で説明すると、ストリップゴム部材Ｇの巻き回しの傾斜角は、目標形状１６に応じて変化する。すなわち、目標形状１６の厚い部分は、重ね合わせ部分を大きくするために送りピッチＰ_nの値は小さくなり、薄い部分は、重ね合わせ部分を小さくするために送りピッチＰ_nの値は大きくなる。
また、図１（ｄ）に示す様に、ストリップゴム部材Ｇの断面形状１８は、送りピッチＰ_nの算出の際、必要に応じて断面形状１８の断面積と横幅Ｂと同じ断面積と横幅を持つ等価矩形形状１９として取り扱われる。
【００２２】
このような送りピッチＰ_ｎは、図３に示す様な、本発明のストリップゴム部材の送りピッチ算出方法を実施する送りピッチ算出装置２０によって、最適送りピッチ群（Ｐ_１，Ｐ_２，・・・，Ｐ_Ｎ）の要素として算出される。
送りピッチ算出装置２０は、入力設定部２２、平均周状厚み算出部２４、送りピッチ算出部２６、第１送りピッチ調整部２８、第２送りピッチ調整部３０、第３送りピッチ調整部３２、最適送りピッチ算出部３４、分割位置算出部３６、およびメモリ部３８を有し、その他、各種画面を表示するディスプレイやこれらの各部位を制御する制御部等を有して構成される。このような送りピッチ算出装置２０は、プログラムを実行することで上記各部位の機能を実行するコンピュータによって構成される。
【００２３】
入力設定部２２は、ストリップゴム部材Ｇの断面形状１８を、図１（ｃ）中のストリップゴム部材Ｇの断面形状１８のパラメータＡ’，Ｂ’，Ｃ’およびＤ’や、実際に巻かれる際のストリップゴム部材Ｇの断面形状を求める際に必要とする引張り率ｑが作業者から入力され、また、タイヤ構成部材の断面の目標形状が入力され、また、後述する様なタイヤ構成部材を１本のストリップゴム部材Ｇで成形するか、２本のストリップゴム部材Ｇでタイヤ構成部材の断面形状を成形するかといった１筆の選択や２筆の選択、さらには、後述する最適送りピッチ群の算出に用いられる各種パラメータが入力され、送りピッチ算出装置２０における各種パラメータを設定する部位である。
【００２４】
入力された断面形状１８のパラメータＡ’，Ｂ’，Ｃ’およびＤ’は、引張り率ｑから、下記式（１）〜（４）によって、実際に巻かれる際の断面形状のパラメータＡ，Ｂ，ＣおよびＤに修正される。
Ａ＝Ａ’／（ｑ）＾^(1/2) （１）
Ｂ＝Ｂ’／（ｑ）＾^(1/2) （２）
Ｃ＝Ｃ’／（ｑ）＾^(1/2) （３）
Ｄ＝Ｄ’／（ｑ）＾^(1/2) （４）
修正されたパラメータＡ、Ｂ、Ｃ、及びＤは、平均周状厚み算出部２４に送られる。また、作業者から入力されるタイヤ構成部材の断面の目標形状は、ドラム幅方向（ｘ方向）における所定間隔毎の厚みの数値データがメモリ部３８に送られて記憶される。なお、タイヤ構成部材の断面の目標形状は、ドラム幅方向における所定間隔の厚みの数値データが目標形状の識別符号とともに記憶されており、この識別符号を入力することで目標形状を設定してもよいし、目標形状を表すドラム幅方向における所定間隔の厚みの数値データを直接入力してもよい。
また、１筆の選択や２筆の選択の指示に応じて、例えば、２筆が選択された場合は、その指示が分割位置算出部３６に送られる。１筆が選択された場合、その指示が送りピッチ算出部２６に送られる。
【００２５】
平均周状厚み算出部２４は、図４（ａ）に示す様に、ストリップゴム部材Ｇの巻き回し開始端Ｇ_sの断面形状の開始側端点位置ｇ₁をｘ方向の原点とし、ストリップゴム部材Ｇのｎ巻目の巻き回しの送りピッチをＰ_n、ｎ巻目のストリップゴム部材Ｇの開始側端点位置ｇ_nの原点からの距離をＳ_nとした時、このｎ巻目のストリップゴム部材Ｇの巻き回しによって得られるストリップゴム部材Ｇのドラム周方向（図４（ａ）中のｙ方向）に沿った平均周状厚みｔ_nを算出する部位である。
【００２６】
ここで、ストリップゴム部材Ｇはｙ方向に傾斜して巻き回されるので、図４（ｂ）に示す様に、ストリップゴム部材Ｇの傾斜、すなわち送りピッチＰ_nと、ストリップゴム部材Ｇの巻き回し位置を表す距離Ｓ_nとを情報として与え、入力設定部２２から送られてきたストリップゴム部材Ｇの断面形状のパラメータＡ，Ｂ，Ｃ，Ｄを用いることで、ｘ方向の任意の位置ｍにおける平均周状厚みｔ_nを算出することができる。
例えば、図４（ｃ）〜（ｆ）に、位置ｍが種々変化した際の平均周状厚みｔ_nを示している。図４（ｃ）および（ｆ）に示す位置ｍの場合、ｎ巻目の巻き回しによる平均周状厚みｔ_nはゼロとなり、ｎ巻目のストリップゴム部材Ｇは、位置ｍにおいて後述する合計加算厚みに寄与しないことを意味する。
平均周状厚み算出部２４は、後述する送りピッチ算出部２６や第１送りピッチ調整部２８、第２送りピッチ調整部３０、第３送りピッチ調整部３２から、位置ｍ、距離Ｓ_nおよび送りピッチＰ_nが与えられると、平均周状厚みｔ_nを自動的に算出し、この平均周状厚みｔ_nの値を送り返す構成となっている。
【００２７】
送りピッチ算出部２６は、ストリップゴム部材Ｇを巻き回す順番に（巻き回す順番をｎ＝１，２，・・・，Ｎで表す）送りピッチＰ_nを順次算出する部位である。
具体的には、まず、最初の巻き回し（ｎ＝１）、すなわち１巻目における送りピッチＰ₁は、図５（ａ）に示す様に、ストリップゴム部材Ｇの断面形状を図１（ｄ）に示す等価矩形形状１９として巻き回して１周させた時の開始側端点位置ｇ₂において、目標形状１６の厚みが等価矩形形状１９の高さＣ^*に略一致する送りピッチを見つけ出すことによって得られる。
【００２８】
より具体的に説明すると、送りピッチの値を０から例えば０．１ｍｍ刻み幅でずらして、１巻目の巻き回し周回後の開始側端点位置ｇ₂における目標形状１６の高さが等価矩形形状１９の高さＣ^*以上であり、かつ、この高さＣ^*に最も近い送りピッチの値を送りピッチＰ₁とする。図５（ａ）に示す開始側端点位置ｇ₁をｘ方向の原点としたｘ軸に沿ったタイヤ構成部材の断面の目標形状１６の厚み分布をＴ_g（ｘ）とすると、下記式（５）で表される値ｖ₁を、値ｖ₁が正または０とする範囲内で最も小さくするｘを求めることになる。
ｖ₁ ＝Ｔ_g（ｘ） − Ｃ^* （５）
【００２９】
次に、２巻目は、１巻目の平均周状厚みｔ₁に少なくとも部分的に重ねてストリップゴム部材Ｇを巻くので、下記式（６）で表される値ｖ₂を、値ｖ₂が正または０とする範囲内で最も小さくするｘを求め、この値から、式（５）で求めた送りピッチＰ₁を差し引くことによって送りピッチＰ₂を求める。
ｖ₂ ＝Ｔ_g（ｘ） − （Ｃ^*＋ｔ₁（ｘ））（６）
従って、ｎ巻目は、１巻目、２巻目、・・・（ｎ−１）巻目の平均周状厚みｔ₁、ｔ₂、・・・ｔ_n-1を加算した累積加算厚みΣｔ_kに少なくとも部分的に重ねてストリップゴム部材Ｇを巻くので、図５（ｂ）に示すようにこのストリップゴム部材Ｇの厚みＣ^*を、既に算出された送りピッチによって定まる平均周状厚みの累積加算厚みΣｔ_kに加算することで得られる新たな厚みが、この巻き回し周回後の（ｎ＋１）巻目の開始側端点位置ｇ_n+1において、目標形状１６の厚みに略一致する送りピッチを見つけ出すことによって得られる。すなわち、下記式（７）で表される値ｖ_nを、値ｖ_nが正または０とする範囲内で最も小さくするｘを求め、この値から、ｎ巻目の開始側端点位置ｇ_nの原点からの距離Ｓ_n（＝Ｐ₁＋Ｐ₂＋・・・＋Ｐ_n-1）を差し引くことによって送りピッチＰ_nを求める。
ｖ_n ＝Ｔ_g（ｘ） − （Ｃ^*＋（ｔ₁（ｘ）＋ｔ₂（ｘ）＋・・・＋ｔ_n-1（ｘ））（７）
ここで、ｔ_i（ｘ）（ｉ＝１・・・ｎ−１）は、平均周状厚み算出部２４に位置ｘと距離Ｓ_iおよび送りピッチＰ_iを与えることによって平均周状厚み算出部２４から送り返される平均周状厚みｔ_iの値である。
また、ストリップゴム部材Ｇの巻き数Ｎは、目標形状の断面積を等価矩形形状１９の横幅Ｂおよび高さＣ^*を除算し、小数点１位が０．１以上の場合は切り上げて整数とすることによって求められる。
送りピッチ算出部２６は、Ｎ個の送りピッチＰ₁、Ｐ₂、・・・Ｐ_Nを算出し、第１送りピッチ調整部２８に送る。
【００３０】
第１送りピッチ調整部２８は、送りピッチ算出部２６で算出された送りピッチＰ₁、Ｐ₂、・・・Ｐ_Nをより正確に調整する部位である。
例えば、送りピッチＰ_nの調整は、送りピッチ算出部２６で算出されたｎ巻目の送りピッチＰ_nの２分の１の値を下限とし、送りピッチ算出部２６で算出されたｎ巻目の送りピッチＰ_nと送りピッチ算出部２６で算出されたｎ＋１巻目の送りピッチＰ_n+1の２分の１の値を加算した値を上限とする範囲内を、例えば０．１ｍｍ刻みで送りピッチの値を変化させ、この変化させたｎ巻目の送りピッチとｎ巻目以外の送りピッチとから各周回毎の平均周状厚みｔ_nを加算合計することによって作られる分布形状（合計加算厚みの分布形状）が、各巻き回し周回後の開始側端点位置ｇ_n+1（ｎ＝１，２・・・Ｎ−１）において目標形状１６に略等しくなる（最も良く近似する）送りピッチを見出し、この送りピッチを送りピッチＰ_nとするものである。
ここで、ｎ巻目の送りピッチＰ_nを調整の対象とする際に用いる平均周状厚みは、図４（ｃ）〜（ｆ）で示されるような、図１（ｃ）に示す断面形状１８を用いた平均周状厚みｔ_nが用いられる。
【００３１】
このような送りピッチの調整は、巻き回しの順番、すなわち、１巻目から順番に行われるが、本発明においては、送りピッチの調整の順番は限定されず、Ｎ巻目から１巻目に向かって調整してもよいし、ランダムの順番に調整してもよい。また、上記例では、ｎ巻目の送りピッチＰ_nの２分の１の値を下限とし、ｎ巻目の送りピッチＰ_nとｎ＋１巻目の送りピッチＰ_n+1の２分の１の値を加算した値を上限とする範囲内で送りピッチの値を変化させて調整するが、この送りピッチの値を変化させる範囲は、必ずしも上記範囲に限定されない。
調整された送りピッチＰ_nは、第２送りピッチ調整部３０に送られる。
【００３２】
第２送りピッチ調整部３０は、第１送りピッチ調整部２８で算出された送りピッチＰ₁、Ｐ₂、・・・Ｐ_Nをより正確に調整する部位である。
例えば、送りピッチＰ_nの調整は、第１送りピッチ調整部２８で調整されたｎ巻目の送りピッチＰ_nの２分の１の値を下限とし、第１送りピッチ調整部２８で調整されたｎ巻目の送りピッチＰ_nとｎ＋１巻目の送りピッチＰ_n+1の２分の１の値を加算した値を上限とする範囲内を、例えば０．１ｍｍ刻みで送りピッチの値を変化させ、この変化させたｎ巻目の送りピッチとｎ巻目以外の送りピッチとから各周回毎の平均周状厚みｔ_nを加算合計することによって作られる合計加算厚みの分布形状が、目標形状１６に略等しくなるような送りピッチの値を見出し、これを送りピッチＰ_nとするものである。ここで目標形状１６に略等しいかどうかの判断は、上記合計加算厚みの分布形状の目標形状１６に対する最大誤差あるいは平均誤差あるいは最大誤差と平均誤差の加重平均を求めることによって行う。例えば、ｘ方向において１ｍｍ刻みで目標形状１６に対する誤差を求め、この誤差から最大誤差あるいは平均誤差あるいは最大誤差と平均誤差の加重平均した加重平均誤差を求め、この最大誤差、平均誤差、あるいは、加重平均誤差を最小とする送りピッチを見出し、この送りピッチを送りピッチＰ_nとするものである。
【００３３】
ここで、ｎ巻目の送りピッチＰ_nを調整の対象とする場合に用いる平均周状厚みは、図４（ｃ）〜（ｆ）で示されるような断面形状１８を用いた平均周状厚みｔ_nが用いられる。このような送りピッチの調整は、巻き回しの周回の順番、すなわち、１巻目から順番に行われる。あるいは、Ｎ巻目から１巻目の順番に調整してもよい。また、上記例では、ｎ巻目の送りピッチＰ_nの２分の１の値を下限とし、ｎ巻目の送りピッチＰ_nとｎ＋１巻目の送りピッチＰ_n+1の２分の１の値を加算した値を上限とする範囲内で送りピッチの値を変化させて調整するが、この送りピッチの値を変化させる範囲は、必ずしも上記範囲に限定されない。
調整された送りピッチＰ_nは、第３送りピッチ調整部３２に送られる。
【００３４】
第３送りピッチ調整部３２は、第２送りピッチ調整部３０で算出された送りピッチＰ₁、Ｐ₂、・・・Ｐ_Nをより正確に調整する部位である。
具体的には、第２送りピッチ調整部３０で調整された送りピッチの順番と逆の順番に送りピッチが調整される。このように、第２送りピッチ調整部３０の調整と逆の順番に調整することで、一方向から調整される場合に比べて、平均周状厚みｔ_nから作られる合計加算厚みの分布形状が目標形状１６に近づくからである。
調整された送りピッチＰ₁、Ｐ₂、・・・Ｐ_Nは、最適送りピッチ算出部３４に送られる。
【００３５】
最適送りピッチ算出部３４は、遺伝的アルゴリズム（ＧＡ）を用いて最適送りピッチＰ₁、Ｐ₂、・・・Ｐ_Nの算出を行う部位である。
最適送りピッチＰ₁、Ｐ₂、・・・Ｐ_Nの算出方法は、後述する。
こうして得られた最適送りピッチＰ₁、Ｐ₂、・・・Ｐ_Nは、送りピッチ算出装置２０の出力として出力される。
この最適送りピッチＰ₁、Ｐ₂、・・・Ｐ_Nは、上述したタイヤ構成部材の作製の際の送りピッチとして用いられる。
【００３６】
分割位置算出部３６は、入力設定部２２で２筆が選択される場合、その設定に応じてタイヤ構成部材の断面の目標形状１６を２分割し、その最適分割位置を算出する部位である。
【００３７】
具体的には、図６（ａ）に示すようなタイヤ構成部材の断面の目標形状１６が設定される場合、図６（ｂ）に示す様に、例えば距離１００ｍｍの位置近傍で目標形状１６を部分目標形状（分割断面形状）１６Ｌと部分目標形状（分割断面形状）１６Ｒに分割すると、この時の部分目標形状１６Ｌ、１６Ｒのそれぞれがメモリ部３８に送られて記憶され、さらに、この部分目標形状１６Ｌ，１６Ｒについての送りピッチＰ_n（ｎ＝１，２・・・Ｎ）が送りピッチ算出部２６でそれぞれ算出され、この算出された送りピッチＰ_nが、分割位置算出部３６に戻される様に構成される。すなわち、分割位置算出部３６で分割位置を設定すると、この分割位置により分割されて得られた部分目標形状１６Ｌ、１６Ｒをメモリ部３８に送るとともに、送りピッチ算出部２６で得られたそれぞれの送りピッチＰ_n（ｎ＝１，２・・・Ｎ）が分割位置算出部３６に返され、部分目標形状１６Ｌ、１６Ｒのそれぞれの送りピッチＰ_nから作られる合計加算厚みの分布形状が１つにまとめられ、図６（ｃ）に示す様な目標形状１６に対応する１つの断面形状１７が作成される。しかし、図６（ｃ）に示す様に、断面形状１７は、必ずしも分割位置付近（図６（ｃ）中の距離９０〜１１０ｍｍ付近）で、目標形状１６に対して一致せず、大きな誤差（最大誤差）が生じてしまう。
そのため、分割位置算出部３６は、予め指定された範囲内で分割位置を変化させ、上記最大誤差が最も小さくなる最適分割位置を算出する。
【００３８】
なお、分割位置は、図６（ｂ）に示すように、厚み方向で変化する領域を形成している。そのため、分割位置を予め指定する範囲も、領域で設定される。例えば、分割開始位置Ｄ_sと厚み方向の中心に位置する中心分割位置Ｄ_mを指定し、中心分割位置Ｄ_mと分割開始位置Ｄ_sを結ぶ直線の傾斜角度を限度とし、中心分割位置Ｄ_mを通り距離軸（横軸）に対する垂線を対称線として傾斜角度を変化させた直線を分割領域線として分割する。
【００３９】
なお、上記例における最適分割位置の算出は、目標形状１６に対する最大誤差が最も小さくなる分割位置を抽出することによって行われるが、本発明において、最適分割位置の算出は、最大誤差を最小とする分割位置の算出に限定されず、ストリップゴム部材の巻き回しの順番に対する送りピッチを、高次多項式で最適に近似する、つまり、３次や４次や５次等の所定の高次多項式に最も相関係数が高く回帰する、つまり、滑らかに送りピッチが変動する、分割位置を抽出してもよい。
なお、入力設定部２２で１筆が選択される場合、分割位置算出部３６は機能せず、送りピッチ算出部２６でタイヤ構成部材の断面の目標形状１６がそのまま用いられる。
【００４０】
メモリ部３８は、タイヤ構成部材の断面形状の目標形状１６や部分断面形状１６Ｌ、１６Ｒや、各部位で設定される各種情報を記憶保持する部位である。
なお、送りピッチ算出装置２０は、目標形状１６や上記各部位で得られた送りピッチＰ_nから作られる断面形状１８等の合計加算厚みの断面形状、さらには、上記最大誤差や平均誤差の数値を表示するディスプレイ（図示されず）、および、各部位の動作を制御する制御部（図示されず）を有する。
【００４１】
次に、本発明のストリップゴム部材の送りピッチ算出方法について、図７に示す流れに基づいて説明する。
まず、ストリップゴム部材Ｇの断面形状１８のパラメータＡ’、Ｂ’、Ｃ’、Ｄ’、ストリップゴム部材Ｇの引っ張り率ｑや、タイヤ構成部材の断面の目標形状１６の設定や、１筆や２筆の選択、さらには、２筆が選択された場合の分割位置の範囲、後述する最適送りピッチの算出に用いられる各種パラメータが、入力設定部２２において入力されて設定される（ステップ１００）。
次に、入力設定部２２において、１筆が選択されたか、２筆が選択されたか判別され（ステップ１０２）、１筆が選択された場合、メモリ部３８に目標形状１６の厚みの数値データが記録され、この目標形状１６に対して送りピッチ算出部２６において上述した方法で送りピッチＰ_nの算出が行われる（ステップ１０４）。
【００４２】
一方、２筆が選択された場合、分割位置算出部３６で最適分割位置の算出が行われる（ステップ１０８）。具体的には、分割位置が予め設定された範囲内で、分割点位置を変化させ、その度に、分割位置で分割された部分目標形状各々について送りピッチ算出部２６で送りピッチＰ_nが求められ、得られた各々の送りピッチＰ_nから作られる合計加算厚みの分布形状が１つにまとめられる。このような１つにまとめられた分布形状の中から、この分布形状の、目標形状１６に対する最大誤差が最小となる、あるいは送りピッチＰ_nが所定の高次多項式で滑らかにカーブフィッティングされる分割位置が求められ、この分割位置が最適分割位置とされる。そして、この最適分割位置で分割して得られる２つの部分目標形状の各々に対して送りピッチ算出部２６で送りピッチＰ_nが算出される。なお、最適分割位置の算出の際、２つの部分目標形状の各々に対して送りピッチＰ_nがすでに求められている場合は、送りピッチの算出が省略され、第１送りピッチ調整（ステップ１０６）および第２送りピッチ調整（ステップ１１０）が行われる。
【００４３】
次に、算出された送りピッチＰ_nは、第１送りピッチ調整部２８に送られ、第１送りピッチ調整が行われる（ステップ１０６）。
すなわち、ステップ１０４で算出されたｎ巻目の送りピッチＰ_nの２分の１の値を下限とし、ステップ１０４で算出されたｎ巻目の送りピッチＰ_nと送りピッチ算出部２６で算出されたｎ＋１巻目の送りピッチＰ_n+1の２分の１の値を加算した値を上限とする範囲内で、例えば０．１ｍｍ刻みで送りピッチの値を変化させ、この変化させたｎ巻目の送りピッチとｎ巻目以外の送りピッチとによって作られる平均周状厚みｔ_nの合計加算厚みの分布形状が、各巻き回し周回後の開始側端点位置ｇ_n+1（ｎ＝１，２．・・・，Ｎ−１）において目標形状に略一致するような送りピッチの値が見出され、この送りピッチが送りピッチＰ_nとされる。
【００４４】
第１送りピッチ調整が行われた後、第２送りピッチ調整部３０において、第２送りピッチ調整が行われる（ステップ１１０）。
ステップ１０６で調整されたｎ巻目の送りピッチＰ_nの２分の１の値を下限とし、ステップ１０６で調整されたｎ巻目の送りピッチＰ_nとｎ＋１巻目の送りピッチＰ_n+1の２分の１の値を加算した値を上限とする範囲内を、例えば０．１ｍｍ刻みで送りピッチの値を変化させ、この変化させたｎ巻目の送りピッチとｎ巻目以外の送りピッチによる平均周状厚みｔ_nから作られる合計加算厚みの分布形状が、目標形状１６に略一致する送りピッチの値が見出され、送りピッチＰ_nとされる。目標形状１６に略一致するかどうかの判断は、平均周状厚みｔ_nによって作られる合計加算厚みの分布形状の目標形状１６に対する最大誤差あるいは平均誤差あるいは最大誤差と平均誤差を加重平均した加重平均誤差を最小とするｎ巻目の送りピッチを求めることによって行われる。
このような送りピッチの調整は１巻目の送りピッチＰ₁から順番に行われる。また、最後のＮ巻目の送りピッチＰ_Nから逆の順番に調整されてもよい。
【００４５】
次に、第２送りピッチ調整が行われた後、第３送りピッチ調整部３２において、第３送りピッチ調整が行われる（ステップ１１２）。
具体的には、第２送りピッチ調整部３０で調整される送りピッチの順番と逆の順番に送りピッチが調整される。例えば第２送りピッチ調整が１巻目の送りピッチＰ₁から順番に行われる場合、第３送りピッチ調整は、最後のＮ巻目の送りピッチＰ_Nから逆の順番に調整され、第２送りピッチ調整がＮ巻目の送りピッチＰ_Nから送りピッチＰ₁に向けて順番に行われる場合、１巻目の送りピッチＰ₁から順番に調整される。このように、第２送りピッチ調整と逆の順番に送りピッチＰ_nを調整することで、一方向から調整される場合に比べて、平均周状厚みｔ_nによって作られる合計加算厚みの分布形状が目標形状１６に近づく。
【００４６】
こうして解析的に算出あるいは調整された送りピッチＰ_nから、遺伝的アルゴリズム（ＧＡ）を用いた最適設計手法を用いて、最適送りピッチＰ_nが算出される（ステップ１１４）。
ここで、ステップ１１２で調整された送りピッチＰ_n（ｎ＝１，２・・・Ｎ）をＰ_n’（ｎ＝１，２・・・Ｎ）とし、さらに、この送りピッチＰ_n’（ｎ＝１，２・・・Ｎ）を１つにまとめて送りピッチ群（Ｐ_1,’Ｐ₂’，・・・Ｐ_N’）とする。最適送りピッチの算出は、送りピッチ群（Ｐ₁’，Ｐ₂’，・・・Ｐ_N’）を初期状態として、タイヤ構成部材の断面の目標形状１６に最も近く近似される最適送りピッチ群（Ｐ₁，Ｐ₂，・・・Ｐ_N）を算出する。このような最適送りピッチの算出方法の流れを図８に示す。
【００４７】
まず、ステップ１１２で求められた送りピッチ群（Ｐ₁’，Ｐ₂’，・・・Ｐ_N’）からＫ組の送りピッチ群が生成され、このＫ組の送りピッチ群が初期の解集合とされる（ステップ１１４ａ）。
具体的には、まず、送りピッチ群（Ｐ₁’，Ｐ₂’，・・・Ｐ_N’）を用いて作られる合計加算厚みの分布形状の、目標形状に対する最大誤差をΔ_maxとし、このΔ_maxの２倍の値を２¹⁶−１分割して得られる２¹⁶個の離散化された数値を用いて、送りピッチＰ_n’が表現される。すなわち、送りピッチ群（Ｐ₁’，Ｐ₂’，・・・Ｐ_N’）の各送りピッチＰ_n’を、下記式（８）によって表されるＶに置き換える。
【００４８】
【数１】

ここで、ｘｂ_kは、送りピッチＰ_n’を２進法で表した際のｋビット目の０又は１の値である。
【００４９】
このようにして、例えば、１６ビットの２進数で離散化された０と１のみで表されたビット形式の送りピッチＰ_n’から成る送りピッチ群を、０と１との乱数を発生させることによってＫ組の送りピッチ群を生成し、このＫ組の送りピッチ群が初期の解集合とされる。ここで、１つの送りピッチを表すビット数は、１６ビットに限定されず、１２〜２４ビットが好ましい。なお、初期の解集合の送りピッチ群の組数（Ｋ組）は、最適な送りピッチを求めるために、ストリップゴム部材Ｇの巻き数Ｎの１０倍程度とするのがよい。例えば、ストリップゴム部材Ｇの巻き数Ｎが２０であれば送りピッチ群の組数を２００程度とするのが好ましい。
こうして生成されたＫ組の送りピッチ群の送りピッチを用いて合計加算厚みの分布形状が、図４（ａ）に示す巻き回し開始端Ｇ_sの断面形状の開始側端点位置ｇ₁から所定の間隔で設けられた確認点、例えば１ｍｍ刻み毎に設けられた確認点における厚みの数値データとして求められる。
【００５０】
次に、フィットネス関数Ｆ_i（ｉ＝１，２，・・・Ｋ）がステップ１００の入力に応じて設定される（ステップ１１４ｂ）。
フィットネス関数Ｆ_iは、ステップ１１４ａで求められた上記確認点における断面形状の厚みと上記確認点の位置に対応する位置における目標形状１６の厚みとの誤差のうち、最大誤差を取り出す関数である。
すなわち、フィットネス関数Ｆ_iは、Ｋ組の送りピッチ群の送りピッチを用いて作られる平均周状厚みの合計加算厚みの分布形状と目標形状１６との最大誤差を求める。なお、フィットネス関数Ｆ_iとして、上記断面形状と目標形状１６の平均自乗誤差としてもよいが、断面形状と目標形状１６の誤差が部分的に大きくなる場合があるため、フィットネス関数Ｆ_iとして最大誤差を用いることが好ましい。
【００５１】
次に、親選定の工程が行われる（ステップ１１４ｃ）。
親選定では、図９（ａ）および（ｂ）に示す様に、解集合であるＫ組の送りピッチ群からＫ／２組の親Ａと親Ｂを、例えば、トーナメント方式を用いて選定する。
トーナメント方式では、Ｋ組の送りピッチ群について第１番目から第Ｋ番目まで番号を付すとともに、最適送りピッチ算出部３４で１〜Ｋまでの整数の乱数を２回発生させ、この乱数発生によって出た２つの数に対応した２つの送りピッチ群を取り出し、この２つの送りピッチ群について得られる上記フィットネス関数Ｆ_iの値の大小を比較し、フィットネス関数Ｆ_iの値の小さい方を親Ａとして選出する。同様に、１〜Ｋまでの乱数を２回発生させ、この乱数の発生によって出た２つの数に対応した２つの送りピッチ群を取り出し、この２つの送りピッチ群について得られるフィットネス関数Ｆ_iの値の大小を比較し、フィットネス関数Ｆ_iの値の小さい方を親Ｂとして選出する。こうして１組の親Ａと親Ｂが選出される。このような親Ａと親Ｂの組をＫ／２組選出する。親Ａと親Ｂの送りピッチ群は、１〜Ｋまでの乱数を２回発生させ、この乱数発生によって出た２つの数に対応した２つの送りピッチ群がＫ組の解集合から選出されるので、親Ａと親Ｂが、同一の送りピッチによって構成される同一の送りピッチ群が選出される場合がある他、別の組の親Ａ、Ｂに、同一の送りピッチから構成される同一の送りピッチ群が選出される場合もある。
【００５２】
図９（ａ）に示す様に、送りピッチ群（Ｐ₁’，Ｐ₂’，・・・Ｐ_N’）の送りピッチＰ_n’の値を識別するために、第ｉ番目を付した送りピッチ群の送りピッチＰ_n’の値をＰ_n ⁽ⁱ⁾（ｉ＝１〜Ｋ）と記載すると、例えば、図９（ｂ）に示す様に、第１組目の親Ａには第５番目が付された送りピッチ群が、親Ｂには第３番目が選出され、２組目の親Ａには３番目の送りピッチ群が、親Ｂには７番目の送りピッチ群が選出されている。すなわち、第３番目の送りピッチ群が親として２回選出されている。
【００５３】
次に、親Ａと親Ｂの交配（交叉）工程が各組の親Ａと親Ｂの間で行われる（ステップ１１４ｄ）。
親Ａと親Ｂの交配は、最適送りピッチ算出部３４で０〜１までの乱数を発生させ、この乱数によって出た値が、所定の値、例えば０．７以下の場合、以下の交配が行われる。所定の値より大きい場合、交配はおこなわれず、親Ａ、Ｂの２つの送りピッチ群が、そのまま保持される。
親Ａと親Ｂの交配は、図１０に示す様に、送りピッチＰ_n ⁽ⁱ⁾からなる親Ａと送りピッチＰ_n ^(j)からなる親Ｂの各送りピッチを、１６ビットの２進法で表し、即ち、式（８）で示されるｘｂ_kの値で表し、この数値を１〜Ｎ個の送りピッチの順番に一列に並べ、合計１６・Ｎビットの０または１の数値から成る数値列を作り送りピッチ情報とする。さらに、１〜（１６・Ｎ−１）までの整数の乱数を発生させ、発生した値に対応した数値列のビット以降の下位ビットの数値列Ａ^'およびＢ’が入れ替えられて交配が行われる。
【００５４】
さらに、突然変異の工程が行われる（ステップ１１４ｅ）。
突然変異の工程は、まず、０〜１までの乱数を発生させ、この乱数によって出た値が、所定の値、例えば０．０８以下の場合、ビットの数値が反転される。すなわち、ビットの値が０の場合１となり、ビットの値が１の場合０となる。このような突然変異の工程は、交配されて得られた１６・Ｎビットの数値列のビット毎に行われる。このように突然変異の工程を設けるのは、解集合の送りピッチ群を所定の範囲でばらつかせることによって、最適送りピッチ群を的確に見つけ出すようにするためである。
【００５５】
次に、得られた１６・Ｎビットの数値列から、数値列の順番に１６ビット毎に分けて、送りピッチの値を１０進法で表して、Ｎ個の送りピッチの値からなる送りピッチ群を式（８）に従って生成する。
このような送りピッチ群は、各組の親Ａ、Ｂを交配させることによって２つの送りピッチ群が生成され、親Ａ、Ｂの交配がない場合でも、親Ａ、Ｂの２つの送りピッチ群がそのまま生成される送りピッチ群として保持されるので、Ｋ／２組の親Ａ、ＢからＫ組の送りピッチ群が生成される（ステップ１１４ｆ）。
【００５６】
次に、ステップ１１４ｆで生成されたＫ組の送りピッチ群について、フィットネス関数Ｆ_iの計算が行われる（ステップ１１４ｇ）。すなわち、Ｋ組の送りピッチ群に、フィットネス関数Ｆ_iを適用してＫ個の値を算出する。
このようなＫ組の送りピッチ群が次世代の解集合として置き換えられる（ステップ１１４ｈ）。
【００５７】
次に、ステップ１１４ｇで計算されたフィットネス関数Ｆ_iの値の最小値が、所定の値、例えば、０．０５（ｍｍ）以下となるか判断し（ステップ１１４ｉ）、所定の値以下の場合、送りピッチ群が収束したものと判断し、上記フィットネス関数Ｆ_iの値が最小となる送りピッチ群が、最適送りピッチ群（Ｐ₁，Ｐ₂，・・・，Ｐ_N）とされる。
上記フィットネス関数Ｆ_iの値の最小値が上記所定の値より大きい場合、送りピッチ群が依然として収束しないものと判断し、ステップ１１４ｈで置き換えられたＫ組の解集合が次世代の解集合として、ステップ１１４ｃに戻される。この場合、次世代への世代交代の交代数が予め設定された世代交代数を超えない範囲でステップ１１４ｃに戻され、ステップ１１４ｈで生成された次世代の解集合を用いて、再度親選定、交配工程・・・が繰り返される。
この場合、フィットネス関数Ｆ_iの値を最大にする送りピッチ群が、フィットネス関数Ｆ_iの値を最小にする送りピッチ群に置き換えられる。
【００５８】
なお、世代交代数が所定の世代交代数、例えば１０世代数に達成すると、最適送りピッチは見出せなかったものとし、新たな試行による最適送りピッチの算出が行われる。すなわち、試行回数が試行回数の上限とする最大試行回数、例えば２０回以下であるか判断される（ステップ１１４ｋ）。なお、試行回数が１回目の場合、式（８）の中のΔ_maxは、上述した様にステップ１１２の第３送りピッチ調整で得られた送りピッチ群の最大誤差を用いるが、試行回数が２回目以上の場合、Ｋ組の送りピッチ群の解集合は、直前のステップ１１４ｈで置き換えられたＫ組の送りピッチ群の解集合が用いられ、さらに、式（８）におけるΔ_maxは、フィットネス関数Ｆ_iで算出された最大誤差が用いられる。
【００５９】
最適送りピッチ群が最大試行回数以内で算出されない場合、最適送りピッチ群は得られなかったものと判断されて終了するが、この場合、フィットネス関数Ｆ_iを最小にする送りピッチ群を最適送りピッチ群に準ずる送りピッチ群として求めてもよい。
以上が最適送りピッチの算出（ステップ１１４）である。
このようにして求められた最適送りピッチ群（Ｐ₁，Ｐ₂，・・・，Ｐ_N）の各送りピッチＰ_nは、上述したタイヤ構成部材の製造方法におけるストリップゴム部材Ｇの巻き回しにおける送りピッチとして用いられる。
【００６０】
一例として、図６（ｃ）に示される２筆の選択によって算出された送りピッチについて、第１送りピッチ調整後、最適送りピッチの算出を行った例を図１１（ａ）、（ｂ）に示している。具体的に説明すると、図１（ｃ）に示すストリップゴム部材Ｇの断面形状のパラメータＡ’，Ｂ’，Ｃ’，Ｄ’およびｑをそれぞれ、１０ｍｍ、２０ｍｍ、１ｍｍ、０．３ｍｍおよび１．０とすると、図６（ａ）に示す目標形状１６に対してストリップゴム部材Ｇの巻き数Ｎは２０となる。図１１（ａ）には、第１送りピッチ調整後の平均周状厚みの合計加算厚みの分布形状（鎖線）が、図１１（ｂ）には、最適送りピッチの算出を行った後の平均周状厚みの合計加算厚みの分布形状（鎖線）が、目標形状１６（実線）と共に示されている。図６（ｃ）の合計加算厚みの分布形状と目標形状１６との最大誤差は０．５２０５ｍｍであるが、図１１（ａ）に示す第１送りピッチ調整後の合計加算厚みの分布形状（鎖線）と目標形状１６（実線）との最大誤差は、０．４９７２ｍｍに縮小し、図１１（ｂ）に示す最適送りピッチ算出後の合計加算厚みの分布形状（鎖線）と目標形状１６（実線）との最大誤差は、０．４０００ｍｍに縮小している。
【００６１】
また、表１には、６種の目標形状（６ケース）に対して、１筆選択、あるいは２筆選択の区分けと、巻き数Ｎの情報と、作業者がマニュアルで設定した送りピッチから作られる合計加算厚みの分布形状と目標形状との最大誤差Ｅ₁と、上記送りピッチの算出および第１〜第３送りピッチ調整を行って得られた送りピッチから作られる合計加算厚みの分布形状と目標形状との最大誤差Ｅ₂と、さらに、上記ＧＡを用いた最適送りピッチの算出方法によって最適化された送りピッチから作られる合計加算厚みの分布形状と目標形状との最大誤差Ｅ₃とが示されている。
【００６２】
【表１】

【００６３】
表１より、最大誤差Ｅ₂は、最大誤差Ｅ₁より略小さく、最大誤差Ｅ₃は最大誤差Ｅ₂よりすべてのケースで小さくなっていることが判る。なお、ケース６の場合、最大誤差Ｅ₁に比べて最大誤差Ｅ₂が略同一となっているが、作業者の送りピッチの設定が比較的良好であったことに起因するものである。ケース１〜５ではすべて最大誤差Ｅ₂が最大誤差Ｅ₁より小さくなっている。
【００６４】
このように、本実施例は、ストリップゴム部材Ｇの送りピッチをステップ１０４〜ステップ１１２で解析的に算出し、その後、平均周状厚みを加算合計した加算合計厚みの分布形状の、目標形状に対する誤差が最も小さくなる最適送りピッチをＧＡを用いて算出するものであるが、このＧＡによる最適送りピッチの算出は、予め送りピッチを算出して調整しなければ、送りピッチを最適に収束させることが難しく、また得られた最適送りピッチも算出の度に異なり、最適送りピッチの解が安定性に欠ける場合があるからである。
【００６５】
なお、本発明は、必ずしも第１〜第３送りピッチ調整を経た後、最適送りピッチの算出を行う上記例に限定されるものではなく、第１送りピッチ調整後、あるいは、第２送りピッチ調整後に、ステップ１１４における最適送りピッチの算出を行ってもよい。
また、上記例では、ストリップゴム部材Ｇの送りピッチが巻き回し毎に変化し、巻き回し毎の送りピッチを求めるものであるが、本発明では上記例に限定されず、２巻き目毎や３巻き目毎に送りピッチを変化させるものであってもよく、この送りピッチを求めるものであってもよい。
【００６６】
また、本発明は、図７に示すストリップゴム部材の送りピッチ算出方法をコンピュータ内で実行するプログラムを提供する。このようなプログラムは、フロッピーディスクやＣＤ−ＲＯＭ等の公知の記録媒体に記録されて提供され、あるいは、インターネット等のネットワークを利用して搬送波によって搬送される。
すなわち、プログラムは、長尺状のストリップゴム部材Ｇをドラム支持装置１０上で少なくとも部分的に重ねつつ巻き回して断面が所定の形状を有するタイヤ構成部材を作製する際のストリップゴム部材Ｇの送りピッチを、平均周状厚みｔ_nに基づいて、コンピュータのＣＰＵ（中央演算ユニット）等の演算手段によって算出させるプログラムである。
【００６７】
このようなプログラムは、図７に示すステップ１０４からステップ１１２までのうち少なくともステップ１０４を、上記ＣＰＵを用いて行う手順と、この手順で算出、調整された複数の送りピッチからなる送りピッチ群から、複数組の送りピッチ群を解集合として初期設定し、この解集合をコンピュータのメモリに記録させる初期解集合設定手順と、平均周状厚みｔ_nを合計加算して得られる合計加算厚みの分布形状の、タイヤ構成部材の断面の目標形状に対する誤差をフィットネス関数Ｆ_iとして設定するフィットネス関数設定手順と、送りピッチ群の解集合から上記フィットネス関数Ｆ_iに基づいて、送りピッチ群の対を複数組設定し、この複数組の対をメモリに記録させ対設定手順と、この設定された送りピッチ群の対の各々が有する送りピッチ情報の一部を、コンピュータの乱数発生関数を用いて、所定の確率で交叉させる交叉手順と、この交叉手順で交叉して得られた送りピッチ情報の一部を、コンピュータの乱数発生関数を用いて、所定の確率で変更する変更手順と、交叉手順の実行前の送りピッチ群の解集合を、変更手順を実行して得られる送りピッチ群に置き換え、この置き換えた解集合をメモリに記録させる置き換え手順を有し、この置き換え手順で生成された解集合にフィットネス関数Ｆ_Iを適用して得られるフィットネス関数Ｆ_iの値の最小値が所定の値以下になるまで、対設定手順、交叉手順、変更手順および置き換え手順を繰り返し行う、コンピュータに実行させるプログラムである。このような手順は、上述したストリップゴム部材の送りピッチ算出方法と同一の工程で行われるのでその説明は省略する。
【００６８】
以上、本発明のストリップゴム部材の送りピッチ算出方法、これを用いたタイヤ構成部材の作製方法およびプログラムについて詳細に説明したが、本発明は上記実施例に限定はされず、本発明の要旨を逸脱しない範囲において、各種の改良および変更を行ってもよいのはもちろんである。
【００６９】
【発明の効果】
以上、詳細に説明したように、本発明によれば、ストリップゴム部材の平均周状厚みを求め、この平均周状厚みを加算合計して得られる加算合計厚みの分布形状が、タイヤ構成部材の断面の目標形状に略等しくなるように送りピッチを求めるので、ストリップゴム部材を用いてタイヤ構成部材を作製する際、ストリップゴム部材の送りピッチを作業者が経験に基づきマニュアルで設定する場合に比べて、目標形状に対する誤差を低減することができる。
しかも、作業者がマニュアルで設定する場合と異なり、特殊な技能や慣れを必要としない。特に、ＧＡによる最適設計手法を用いることで、目標形状に対する誤差を飛躍的に低減することができる。
【図面の簡単な説明】
【図１】（ａ）〜（ｄ）は、本発明のタイヤ構成部材の作製方法の一例を説明する説明図である。
【図２】本発明のタイヤ構成部材の作製方法の一例を説明する説明図である。
【図３】本発明のストリップゴム部材の送りピッチ算出方法の一例を実施する最適送りピッチ算出装置の構成を示す構成図である。
【図４】（ａ）〜（ｆ）は、本発明のストリップゴム部材の送りピッチ算出方法の一工程の一例を説明する説明図である。
【図５】（ａ）および（ｂ）は、本発明のストリップゴム部材の送りピッチ算出方法の他の工程の一例を説明する説明図である。
【図６】（ａ）〜（ｃ）は、本発明のストリップゴム部材の送りピッチ算出方法の一工程で得られる断面形状を説明する説明図である。
【図７】本発明のストリップゴム部材の送りピッチ算出方法の一例の流れを説明すフローチャートである。
【図８】本発明のストリップゴム部材の送りピッチ算出方法における最適送りピッチ算出の流れを説明すフローチャートである。
【図９】（ａ）および（ｂ）は、図８に示す最適送りピッチ算出の一工程を説明する説明図である。
【図１０】図８に示す最適送りピッチ算出の他の工程を説明する説明図である。
【図１１】（ａ）および（ｂ）は、本発明のストリップゴム部材の送りピッチ算出方法によって得られる断面形状の一例を示す図である。
【符号の説明】
１０ドラム支持装置
１２ドラム表面
１４タイヤ中間組立体
１６目標形状
１７，１８断面形状
１９等価矩形形状
２０送りピッチ算出装置
２２入力設定部
２４平均周状厚み算出部
２６送りピッチ算出部
２８第１送りピッチ調整部
３０第２送りピッチ調整部
３２第３送りピッチ調整部
３４最適送りピッチ算出部
３６分割位置算出部
３８メモリ部[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a feed pitch of a strip rubber member when a tire component having a predetermined cross-sectional shape is formed by winding a long strip rubber member while being partially overlapped on the drum circumference of a drum support device. The present invention relates to a method for calculating a feed pitch of a strip rubber member that is automatically calculated and further optimally calculated, a method for producing a tire component using the same, and a program.
[0002]
[Prior art]
At present, a pneumatic tire has a carcass member, a bead member or the like disposed on the drum surface of a drum support device, and a rubber member having a predetermined cross-sectional shape such as a side rubber member or a tread rubber member is wound and pasted on the upper layer. By combining them, a raw tire before vulcanization is formed.
[0003]
On the other hand, a tire constituent member made of a rubber member having a predetermined cross-sectional shape is formed by winding a long strip rubber member in a spiral shape while partially overlapping the drum circumference of the drum support device. A tire strip winding method has been proposed.
A conventional method for manufacturing a tire constituent member is a method for producing a tire constituent member made of a rubber member by extruding the rubber member from a mold of an extruder having an opening having a predetermined opening shape. While it is necessary to produce a complicated mold every time the cross-sectional shape of the member changes, the above-described strip winding method produces a tire component having a desired cross-sectional shape by directly forming it on a drum support device. Since a tire can be produced, a great reduction in man-hours and costs is expected.
[0004]
In such a strip winding method, a tire component is formed by winding a long strip rubber member in a spiral shape while partially overlapping it, so that it is accurate so as not to deviate from the target shape of the cross section of the tire component. It is necessary to mold. Therefore, it is important to set the feed pitch of the strip rubber member that is wound spirally while partially overlapping the long strip rubber member so as to match the design shape of the cross section of the target tire component member. It is a difficult issue. In particular, the degree of error with respect to the design shape of the cross-sectional shape of the tire component produced by winding the strip rubber member is poor or bad in tire performance. For example, the tire vibration component (uniformity component) generated when the tire is mounted on a vehicle ) Has a big effect on the size. Therefore, how to set the feed pitch of the long strip rubber member is one factor that determines whether the tire performance is good or bad.
[0005]
[Problems to be solved by the invention]
However, in the above-described strip winding method, as described in Japanese Patent Application Laid-Open No. 61-130030, the thickness of the tire component is measured using a winding thickness sensor or the like while winding the strip rubber on the drum support device. Then, the tire constituent member was molded, and when the thickness reached, the strip rubber member was sent in the drum axial direction of the drum support device.
In addition, when a feed pitch for each turn for winding the strip rubber member on the drum is set in advance, the setting of the feed pitch is performed manually depending on the skill and experience of the operator, and the tire configuration to be molded In many cases, the cross-sectional shape of the member was a rough cross-sectional shape. Therefore, setting a feed pitch close to the target shape requires a lot of trial and error, and even if the feed pitch can be set relatively close to the target shape, this feed pitch is optimal. There was a problem that it was impossible to determine whether or not the feed pitch was reached.
[0006]
Therefore, in order to solve the above-described problems, the present invention provides a strip rubber member for forming a tire constituent member by winding a long strip rubber member while being partially overlapped on the drum circumference of a drum support device. A method for calculating a feed pitch of a strip rubber member that automatically calculates a feed pitch, and further adjusts the cross-sectional shape obtained by winding the strip rubber member to an optimum feed pitch that minimizes an error with respect to the design shape, and An object of the present invention is to provide a method for producing a tire constituent member using this and a computer-executable program for calculating the feed pitch.
[0007]
[Means for Solving the Problems]
  In order to achieve the above object, the present invention provides a tire having a predetermined cross-sectional shape by winding a long strip rubber member on the drum circumference while at least partially overlapping the drum support device along the drum width direction. A method for calculating a feed pitch of a strip rubber member along a drum width direction of the drum support device when producing a component member,
  The feed pitch calculatorThe feed pitchTheCalculated based on the average circumferential thickness along the drum circumferential direction of the strip rubber member obtained by winding the strip rubber memberA feed pitch calculation step for calculating from each winding start end of the strip rubber member, and an optimum feed pitch calculation step for calculating an optimum feed pitch by an optimum feed pitch calculation unit; Have
  The optimum feed pitch calculation step includes
  An initial solution set setting step of generating a plurality of sets of feed pitch groups using a feed pitch group consisting of a plurality of feed pitches obtained at least through the feed pitch calculation step, and initializing these feed pitch groups as a solution set When,
  A fitness function setting step of setting, as a fitness function, an error with respect to the cross-sectional shape of the tire constituent member of the distribution shape of the total added thickness obtained by total addition of the average circumferential thickness obtained from the feed pitch of the feed pitch group;
  From a solution set of feed pitch groups, a pair setting step for setting a plurality of pairs of feed pitch groups based on the fitness function;
  A crossover step of crossing a part of the feed pitch information included in each of the set feed pitch group pairs with a predetermined probability;
  A changing step of changing a part of the feed pitch information obtained by crossing in this crossing step with a predetermined probability;
  A replacement step of replacing the solution set of the feed pitch group before the crossover step with the feed pitch group obtained through the change step,
  The pair setting step, until the minimum value of the fitness function value obtained by applying the fitness function to all the feed pitch groups of the solution set obtained in the replacement step is equal to or less than a predetermined value, The optimum feed pitch is calculated by repeating the crossover process, the change process and the replacement process.The present invention provides a method for calculating the feed pitch of a strip rubber member.
[0008]
  here,in frontIn the feed pitch calculation step, it is preferable to calculate the feed pitch sequentially in the order of the wrapping around the strip rubber member.
  The feed pitch calculation step includes
  From the feed pitch already calculated before the calculation of the feed pitch to be calculated at the end position on the winding start side in the drum width direction in the cross section of the strip rubber member after winding around the feed pitch to be calculated The new thickness obtained by adding the thickness of the strip rubber member to the cumulative added thickness of the obtained average circumferential thickness is substantially equal to the thickness of the cross section at the corresponding position of the tire component corresponding to the end point position. It is preferable to calculate the feed pitch to be calculated. At this time, the thickness of the strip rubber member to be added to the cumulative added thickness in the feed pitch calculating step may be an equivalent rectangular cross-sectional thickness having the same cross-sectional area and lateral width as the cross-sectional area and lateral width of the strip rubber member. .
[0009]
The strip rubber member feed pitch calculation method includes a first feed pitch adjustment step of adjusting each of the feed pitches calculated in the feed pitch calculation step,
This first feed pitch adjusting step maintains the feed pitch value as it is except for the feed pitch to be adjusted, and the total added thickness obtained by changing the feed pitch value to be adjusted within a predetermined range. However, at the end position on the winding start side in the drum width direction in the cross section of the strip rubber member after the winding around the feed pitch to be adjusted, at the corresponding position of the tire constituent member corresponding to this end point position It is preferable to adjust the feed pitch to be adjusted so as to be substantially equal to the thickness of the cross section.
[0010]
Furthermore, the strip rubber member feed pitch calculation method includes a second feed pitch adjustment step of adjusting each of the feed pitches adjusted in the first feed pitch adjustment step,
In the second feed pitch adjustment step, the feed pitch adjusted in the first feed pitch adjustment step holds the feed pitch value as it is except for the feed pitch to be adjusted, and the feed pitch to be adjusted. The feed pitch to be adjusted is adjusted so that the distribution shape of the total added thickness of the average circumferential thickness obtained by changing the value within a predetermined range is substantially equal to the cross-sectional shape of the tire constituent member. Is preferred. In that case, it is preferable that the second feed pitch adjusting step sequentially adjusts the feed pitch in the order of the winding of the strip rubber member or in the order opposite to the order of the winding.
[0011]
The strip rubber member feed pitch calculation method may further include adjusting the feed pitch adjusted in the second feed pitch adjustment step to each of the feed pitches adjusted in the second feed pitch adjustment step. Having a third feed pitch adjustment step to adjust in reverse order;
In the third feed pitch adjustment step, the feed pitch adjusted in the second feed pitch adjustment step holds the feed pitch value as it is except for the feed pitch to be adjusted, and the feed pitch value to be adjusted. It is preferable to adjust the feed pitch so that the distribution shape of the total added thickness obtained by changing the value within a predetermined range is approximately equal to the cross-sectional shape of the tire constituent member.
[0012]
In addition, at least one of the second feed pitch adjustment step and the third feed pitch adjustment step includes a maximum error of a distribution shape of the total added thickness with respect to a cross-sectional shape of the tire constituent member, and the total added thickness. The feed pitch to be adjusted is adjusted using one of an average error of the distribution shape of the tire configuration member with respect to a cross-sectional shape of the tire component and a weighted average of the maximum error and the average error. preferable.
[0014]
At this time, the pair setting step sets one of the feed pitch groups having the smaller fitness function value as one of the pairs of feed pitch groups among the two feed pitch groups randomly selected from the solution set, and Of the two feed pitch groups newly selected at random from the solution set, the feed pitch group with the smaller fitness function value is preferably the other of the pair of feed pitch groups, and the fitness function is A maximum error of the distribution shape of the total added thickness with respect to the cross-sectional shape of the tire constituent member, or a weighted average of the average error of the cross-sectional shape of the tire constituent member with respect to the cross-sectional shape of the tire constituent member and the maximum error. Is preferred.
[0015]
  The strip rubber member feed pitch calculation method may calculate, for each strip rubber member, a feed pitch that is wound around the drum circumference using at least a plurality of long strip rubber members.
  In this case, when calculating the feed pitch of the strip rubber member for each divided sectional shape obtained by dividing the sectional shape of the tire constituent member into two,
  The maximum error with respect to the cross-sectional shape of the tire constituent member of the distribution shape obtained by combining the distribution shapes of the total added thicknesses formed from the feed pitch set for each of the divided cross-sectional shapes into one is minimized. In addition,Dividing the cross-sectional shape of the tire component into twoThe division position may be set,
  Alternatively, the division position may be set so that the feed pitch with respect to the winding order of the strip rubber members set for each of the divided sectional shapes is optimally approximated by a high-order polynomial.
[0016]
The present invention also provides a method for producing a tire constituent member, characterized in that a tire constituent member is produced using the feed pitch calculated using the above-described feed pitch calculation method for strip rubber members. .
[0017]
  Further, the present invention provides the drum support when the long strip rubber member is wound around the drum circumference while at least partially overlapping the drum support device to produce a tire constituent member having a predetermined cross-sectional shape. A program for causing a computer to execute a method for obtaining a feed pitch of a strip rubber member along the drum width direction of the apparatus,
  The feeding pitch is calculated based on an average circumferential thickness along the drum circumferential direction of the strip rubber member.ByLet me calculate, For each round of winding the strip rubber member, a procedure for obtaining the feed pitch using the computing means;
  A plurality of feed pitch groups are generated using the obtained feed pitch group consisting of a plurality of feed pitches, these feed pitch groups are initialized as a solution set, and the solution set is recorded in the storage means of the computer. Initial solution set setting procedure,
  A fitness function setting procedure for setting, as a fitness function, an error with respect to the cross-sectional shape of the tire constituent member of the distribution shape of the total added thickness obtained by total addition of the average circumferential thickness obtained from the feed pitch of the feed pitch group;
  From the solution set of the feed pitch group, a pair setting procedure for setting a plurality of pairs of feed pitch groups based on the fitness function and recording the plurality of pairs in the storage means;
  A crossover procedure for crossing a part of the feed pitch information of each of the set feed pitch group pairs with each other with a predetermined probability using a computer random number generation means;
  A change procedure for changing a part of the feed pitch information obtained by crossing by this crossing procedure at a predetermined probability using the random number generating means,
  A replacement procedure in which the solution set of the feed pitch group before the crossover procedure is executed is replaced with the feed pitch group obtained through the change procedure, and the replacement solution set is recorded in the storage means;
  The pair setting procedure and the crossover procedure until the minimum value of the fitness function value obtained by applying the fitness function to the solution set of the feed pitch group obtained by the replacement procedure is equal to or less than a predetermined value. By repeating the change procedure and the replacement procedure, the optimum feed pitch group is calculated.A program characterized by the above is provided.
[0019]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, a method for calculating a feed pitch of a strip rubber member, a method for producing a tire component using the same, and a program according to the present invention will be described in detail based on preferred embodiments shown in the accompanying drawings.
[0020]
Fig.1 (a) is explanatory drawing explaining the manufacturing method of the tire structural member using the feed pitch calculation method of the strip rubber member of this invention.
A tire constituent member, for example, a side rubber member or a tread rubber member is produced by winding a long strip rubber member G on the drum circumference while at least partially overlapping. That is, while the strip rubber member G is at least partially overlapped on the tire intermediate assembly 14 obtained by winding a carcass member, a bead member or the like around the drum surface 12 of the drum support device 10, As shown in FIG. 1 (b), the tire constituent member is molded such that the contour shape of the strip rubber member G substantially matches the target shape 16. Here, the winding of the strip rubber member G is at least partly overlapped and wound in one direction. Of course, depending on the cross-sectional shape of the tire constituent member, the strip rubber member G may overlap the entire previous strip rubber member G in the drum width direction.
[0021]
Here, as shown in FIG. 1C, the cross-sectional shape 18 of the strip rubber member G is a shape in which the upper and lower ends of the rectangular shape are notched (FIGS. 1A and 1B). The strip rubber member G in FIG. 2 is shown as a strip rubber member having a rectangular cross-sectional shape for simplification), and the strip rubber member G is wound around the drum width direction (x direction) of the drum support device 10. Pitch P_n(N = 1, 2,... N) is set for each winding. That is, as shown in FIG. 1A, when the strip rubber member G is spirally wound in the x direction, every round (n = 1) that is wound 360 degrees from the winding start end of the strip rubber member G. , 2,... N), the feed pitch P of the nth roll, which is the feed amount of the strip rubber member G in the x direction._n(N = 1, 2,... N) is set. Therefore, when the strip rubber member G wound along the drum surface 12 is developed on FIG. 2, the inclination angle of the winding of the strip rubber member G changes according to the target shape 16. That is, the thick portion of the target shape 16 is fed to the feed pitch P in order to enlarge the overlapping portion._nThe value of becomes smaller, and the thin part is fed to the feed pitch P in order to reduce the overlapping part._nThe value of increases.
Further, as shown in FIG. 1 (d), the cross-sectional shape 18 of the strip rubber member G has a feed pitch P_nIs calculated, it is handled as an equivalent rectangular shape 19 having the same cross-sectional area and width as the cross-sectional area and width B of the cross-sectional shape 18 as necessary.
[0022]
  Such feed pitch P_nIs a feed pitch for carrying out the method for calculating the feed pitch of the strip rubber member of the present invention as shown in FIG.CalculationBy the device 20, the optimum feed pitch group (P₁, P₂, ..., P_N).
  The feed pitch calculation device 20 includes an input setting unit 22, an average circumferential thickness calculation unit 24, a feed pitch calculation unit 26, a first feed pitch adjustment unit 28, a second feed pitch adjustment unit 30, a third feed pitch adjustment unit 32, It has an optimum feed pitch calculation unit 34, a division position calculation unit 36, and a memory unit 38, and further includes a display that displays various screens, a control unit that controls each of these parts, and the like. Such a feed pitch calculation device 20 is configured by a computer that executes the function of each part by executing a program.
[0023]
  The input setting unit 22 actually winds the cross-sectional shape 18 of the strip rubber member G, parameters A ′, B ′, C ′ and D ′ of the cross-sectional shape 18 of the strip rubber member G in FIG. An operator inputs a tensile rate q required to obtain a cross-sectional shape of the strip rubber member G, and inputs a target shape of a cross-section of the tire constituent member. Selection of one brush, selection of two brushes such as molding with one strip rubber member G or molding of the cross-sectional shape of the tire component member with two strip rubber members G, and an optimum feed pitch group to be described later Various parameters used to calculate the feed pitch are inputCalculationThis is a part for setting various parameters in the apparatus 20.
[0024]
The parameters A ′, B ′, C ′ and D ′ of the input cross-sectional shape 18 are parameters A, B of the cross-sectional shape when actually wound by the following formulas (1) to (4) from the tensile rate q. , C and D.
A = A '/ (q) ^^(1/2)                    (1)
B = B '/ (q) ^^(1/2)                    (2)
C = C ′ / (q) ^^(1/2)                    (3)
D = D ′ / (q) ^^(1/2)                    (4)
The corrected parameters A, B, C, and D are sent to the average circumferential thickness calculator 24. Further, the target shape of the cross section of the tire constituent member input by the operator is sent to the memory unit 38 and stored therein numerical data on the thickness at predetermined intervals in the drum width direction (x direction). As for the target shape of the cross section of the tire component member, numerical data of thicknesses at predetermined intervals in the drum width direction is stored together with an identification code of the target shape, and even if the target shape is set by inputting this identification code Alternatively, numerical data of thicknesses at predetermined intervals in the drum width direction representing the target shape may be directly input.
Further, for example, when two brushes are selected in response to the selection of one brush or the selection of two brushes, the instruction is sent to the division position calculation unit 36. When one brush is selected, the instruction is sent to the feed pitch calculation unit 26.
[0025]
As shown in FIG. 4A, the average circumferential thickness calculation unit 24 starts the winding start end G of the strip rubber member G._sStart side end point position g of the cross-sectional shape of₁Is the origin in the x direction, and the feed pitch of the nth turn of the strip rubber member G is P_n, End position g of the start side of the strip rubber member G of the nth roll_nThe distance from the origin of S_nThe average circumferential thickness t along the drum circumferential direction (y direction in FIG. 4 (a)) of the strip rubber member G obtained by winding the n-th roll of the strip rubber member G._nIt is a part which calculates.
[0026]
Here, since the strip rubber member G is wound while being inclined in the y direction, as shown in FIG. 4B, the inclination of the strip rubber member G, that is, the feed pitch P_nAnd a distance S representing the winding position of the strip rubber member G_nAs information, and using the parameters A, B, C, and D of the cross-sectional shape of the strip rubber member G sent from the input setting unit 22, the average circumferential thickness t at an arbitrary position m in the x direction is used._nCan be calculated.
For example, in FIGS. 4C to 4F, the average circumferential thickness t when the position m is variously changed._nIs shown. In the case of the position m shown in FIGS. 4C and 4F, the average circumferential thickness t by winding the nth turn_nIs zero, which means that the n-th strip rubber member G does not contribute to the total added thickness described later at the position m.
The average circumferential thickness calculation unit 24 includes a position m, a distance S from a feed pitch calculation unit 26, a first feed pitch adjustment unit 28, a second feed pitch adjustment unit 30, and a third feed pitch adjustment unit 32, which will be described later._nAnd feed pitch P_nIs given, the average circumferential thickness t_nIs automatically calculated, and this average circumferential thickness t_nIt is configured to send back the value of.
[0027]
The feed pitch calculating unit 26 feeds the strip rubber member G in the turn order (the turn order is represented by n = 1, 2,..., N)._nIs a part for sequentially calculating.
Specifically, first, the first winding (n = 1), that is, the feed pitch P in the first roll₁As shown in FIG. 5 (a), the start side end point position g when the strip rubber member G is wound around as an equivalent rectangular shape 19 shown in FIG.₂, The thickness of the target shape 16 is equal to the height C of the equivalent rectangular shape 19.^*Can be obtained by finding a feed pitch that substantially matches.
[0028]
More specifically, the value of the feed pitch is shifted from 0 by, for example, 0.1 mm increments, and the starting side end point position g after the first winding is wound.₂The height of the target shape 16 in FIG.^*It is above and this height C^*The feed pitch value closest to₁And Start side end point position g shown in FIG.₁The thickness distribution of the target shape 16 of the cross-section of the tire constituent member along the x axis with the origin in the x direction as T_gAssuming (x), the value v represented by the following formula (5)₁To the value v₁X which is the smallest within the range where is positive or 0 is obtained.
v₁  = T_g(X) -C^*                      (5)
[0029]
Next, the second volume is the average circumferential thickness t of the first volume.₁Since the strip rubber member G is wound at least partially on the surface, the value v expressed by the following formula (6)₂To the value v₂X which is the smallest within the range where is positive or 0 is obtained, and from this value, the feed pitch P obtained by the equation (5) is obtained.₁Pitch P by subtracting₂Ask for.
v₂  = T_g(X)-(C^*+ T₁(X)) (6)
Therefore, the nth roll is the average circumferential thickness t of the first roll, the second roll, ... (n-1).₁, T₂, ... t_n-1Cumulative added thickness Σt_kSince the strip rubber member G is wound at least partially on the strip, the thickness C of the strip rubber member G is shown in FIG.^*Is the cumulative added thickness Σt of the average circumferential thickness determined by the already calculated feed pitch_kThe new thickness obtained by adding to is the start side end point position g of the (n + 1) th turn after this winding turn._{n + 1}In this case, the feed pitch is obtained by finding a feed pitch that substantially matches the thickness of the target shape 16. That is, the value v represented by the following formula (7)_nTo the value v_nX which is the smallest in the range where is positive or 0 is obtained, and from this value, the starting side end point position g of the nth turn_nThe distance S from the origin_n(= P₁+ P₂+ ... + P_n-1) Is subtracted from the feed pitch P_nAsk for.
v_n  = T_g(X)-(C^*+ (T₁(X) + t₂(X) + ... + t_n-1(X)) (7)
Where t_i(X) (i = 1... N−1) is the average circumferential thickness calculator 24 with the position x and the distance S._iAnd feed pitch P_iThe average circumferential thickness t sent back from the average circumferential thickness calculation unit 24_iIs the value of
Further, the number N of windings of the strip rubber member G is determined so that the cross-sectional area of the target shape is equal to the lateral width B and height C of the equivalent rectangular shape 19.^*When the first decimal place is 0.1 or more, it is calculated by rounding up to an integer.
The feed pitch calculation unit 26 is configured to generate N feed pitches P.₁, P₂・・・・・・ P_NIs sent to the first feed pitch adjustment unit 28.
[0030]
The first feed pitch adjusting unit 28 is a feed pitch P calculated by the feed pitch calculating unit 26.₁, P₂・・・・・・ P_NIt is the part which adjusts more accurately.
For example, feed pitch P_nIs adjusted with the feed pitch P calculated by the feed pitch calculator 26._nThe feed pitch P of the nth turn calculated by the feed pitch calculation unit 26 with the value of 1/2 of_nAnd the feed pitch P of the (n + 1) th roll calculated by the feed pitch calculation unit 26_{n + 1}For example, by changing the value of the feed pitch in increments of 0.1 mm within the range where the value obtained by adding one-half of the value is the upper limit, the changed feed pitch of the nth roll and the feed other than the nth roll Average circumferential thickness t for each lap from the pitch_nThe distribution shape (distribution shape of the total added thickness) created by adding and summing is the starting side end point position g after each winding._{n + 1}In (n = 1, 2,..., N−1), a feed pitch that is substantially equal (best approximate) to the target shape 16 is found, and this feed pitch is determined as the feed pitch P._nIt is what.
Here, the feed pitch P of the nth roll_nThe average circumferential thickness t used when adjusting is the average circumferential thickness t using the cross-sectional shape 18 shown in FIG. 1 (c) as shown in FIGS. 4 (c) to 4 (f)._nIs used.
[0031]
Such adjustment of the feed pitch is performed in the winding order, that is, in order from the first volume. However, in the present invention, the order of adjustment of the feed pitch is not limited, and the Nth to the first volume. You may adjust toward you and may adjust in random order. In the above example, the nth turn feed pitch P_nThe lower limit is half the value of the feed pitch P of the nth winding._nAnd n + 1 roll feed pitch P_{n + 1}The feed pitch value is changed and adjusted within a range where the upper limit is a value obtained by adding one-half of the value. However, the range in which the feed pitch value is changed is not necessarily limited to the above range.
Adjusted feed pitch P_nIs sent to the second feed pitch adjusting unit 30.
[0032]
The second feed pitch adjusting unit 30 is a feed pitch P calculated by the first feed pitch adjusting unit 28.₁, P₂・・・・・・ P_NIt is the part which adjusts more accurately.
For example, feed pitch P_nIs adjusted by the first feed pitch adjusting unit 28._nThe feed pitch P of the nth turn adjusted by the first feed pitch adjustment unit 28 with the value of 1/2 of the lower limit as the lower limit_nAnd n + 1 roll feed pitch P_{n + 1}For example, by changing the value of the feed pitch in increments of 0.1 mm within the range where the value obtained by adding one-half of the value is the upper limit, the changed feed pitch of the nth roll and the feed other than the nth roll Average circumferential thickness t for each lap from the pitch_nThe value of the feed pitch is found such that the distribution shape of the total added thickness created by adding and summing is substantially equal to the target shape 16, and this is calculated as the feed pitch P._nIt is what. Here, whether or not the target shape 16 is substantially equal is determined by obtaining a maximum error or an average error with respect to the target shape 16 of the distribution shape of the total added thickness or a weighted average of the maximum error and the average error. For example, an error with respect to the target shape 16 is obtained in increments of 1 mm in the x direction, and a maximum error, an average error, or a weighted average error obtained by weighted averaging of the maximum error and the average error is obtained from this error, and the maximum error, average error, or weight Find the feed pitch that minimizes the average error, and use this feed pitch as the feed pitch P._nIt is what.
[0033]
Here, the feed pitch P of the nth roll_nIs the average circumferential thickness t using a cross-sectional shape 18 as shown in FIGS. 4C to 4F._nIs used. Such adjustment of the feed pitch is performed in the winding order, that is, in order from the first volume. Or you may adjust in order of the 1st volume from the Nth volume. In the above example, the nth turn feed pitch P_nThe lower limit is half the value of the feed pitch P of the nth winding._nAnd n + 1 roll feed pitch P_{n + 1}The feed pitch value is changed and adjusted within a range where the upper limit is a value obtained by adding one-half of the value. However, the range in which the feed pitch value is changed is not necessarily limited to the above range.
Adjusted feed pitch P_nIs sent to the third feed pitch adjusting unit 32.
[0034]
The third feed pitch adjusting unit 32 is a feed pitch P calculated by the second feed pitch adjusting unit 30.₁, P₂・・・・・・ P_NIt is the part which adjusts more accurately.
Specifically, the feed pitch is adjusted in the order opposite to the order of the feed pitch adjusted by the second feed pitch adjustment unit 30. Thus, by adjusting in the reverse order to the adjustment of the second feed pitch adjusting unit 30, the average circumferential thickness t is compared with the case of adjusting from one direction._nThis is because the distribution shape of the total added thickness made from the above approaches the target shape 16.
Adjusted feed pitch P₁, P₂・・・・・・ P_NIs sent to the optimum feed pitch calculation unit 34.
[0035]
The optimum feed pitch calculation unit 34 uses the genetic algorithm (GA) to obtain the optimum feed pitch P.₁, P₂・・・・・・ P_NIt is a part which calculates.
Optimum feed pitch P₁, P₂・・・・・・ P_NThe calculation method of will be described later.
The optimum feed pitch P obtained in this way₁, P₂・・・・・・ P_NIs output as the output of the feed pitch calculation device 20.
This optimum feed pitch P₁, P₂・・・・・・ P_NIs used as a feed pitch in the production of the tire constituent member described above.
[0036]
When two strokes are selected by the input setting unit 22, the division position calculation unit 36 divides the target shape 16 of the cross section of the tire constituent member into two according to the setting, and calculates the optimal division position.
[0037]
Specifically, when the target shape 16 of the cross section of the tire constituent member as shown in FIG. 6A is set, as shown in FIG. 6B, the target shape 16 is set in the vicinity of a position of a distance of 100 mm, for example. When divided into the partial target shape (divided cross-sectional shape) 16L and the partial target shape (divided cross-sectional shape) 16R, each of the

partial target shapes

16L and 16R at this time is sent to and stored in the memory unit 38. Feed pitch P for

shapes

16L and 16R_n(N = 1, 2,... N) are respectively calculated by the feed pitch calculation unit 26, and the calculated feed pitch P is calculated._nIs returned to the division position calculation unit 36. That is, when the division position is set by the division position calculation unit 36, the

partial target shapes

16L and 16R obtained by division at this division position are sent to the memory unit 38, and the respective feeds obtained by the feed pitch calculation unit 26 are sent. Pitch P_n(N = 1, 2,... N) is returned to the division position calculation unit 36, and the respective feed pitches P of the partial target shapes 16L and 16R._nThe total added thickness distribution shapes created from the above are combined into one, and one cross-sectional shape 17 corresponding to the target shape 16 as shown in FIG. 6C is created. However, as shown in FIG. 6C, the cross-sectional shape 17 does not necessarily coincide with the target shape 16 in the vicinity of the division position (near the distance of 90 to 110 mm in FIG. 6C), and a large error ( Maximum error).
Therefore, the division position calculation unit 36 changes the division position within a range designated in advance, and calculates the optimum division position where the maximum error is minimized.
[0038]
As shown in FIG. 6B, the division position forms a region that changes in the thickness direction. Therefore, a range in which the division position is designated in advance is also set in the area. For example, the division start position D_sAnd the center division position D located at the center in the thickness direction_mAnd the center division position D_mAnd division start position D_sThe center split position D is limited to the inclination angle of the straight line connecting_mA straight line whose inclination angle is changed with a perpendicular line to the distance axis (horizontal axis) as a symmetrical line is divided as a divided region line.
[0039]
The calculation of the optimal division position in the above example is performed by extracting the division position where the maximum error with respect to the target shape 16 is the smallest. In the present invention, the calculation of the optimal division position minimizes the maximum error. It is not limited to the calculation of the division position, and the feed pitch for the winding order of the strip rubber members is optimally approximated by a high-order polynomial, that is, the highest to a predetermined high-order polynomial such as a third order, fourth order or fifth order. A division position where the correlation coefficient recurs high, that is, the feed pitch fluctuates smoothly may be extracted.
In addition, when one brush is selected by the input setting unit 22, the division is performed.

position

The calculation unit 36 does not function, and the target shape 16 of the cross section of the tire constituent member is used as it is in the feed pitch calculation unit 26.
[0040]
The memory unit 38 is a part that stores and holds the target shape 16 of the cross-sectional shape of the tire constituent member, the partial

cross-sectional shapes

16L and 16R, and various information set in each part.
Note that the feed pitch calculation device 20 is configured so that the feed pitch P obtained from the target shape 16 or each of the above-described parts._nThe cross-sectional shape of the total added thickness such as the cross-sectional shape 18 made from the above, a display (not shown) for displaying the numerical values of the maximum error and the average error, and a control unit (not shown) for controlling the operation of each part )).
[0041]
Next, a method for calculating the feed pitch of the strip rubber member of the present invention will be described based on the flow shown in FIG.
First, the parameters A ′, B ′, C ′, D ′ of the cross-sectional shape 18 of the strip rubber member G, the pull rate q of the strip rubber member G, the setting of the target shape 16 of the cross-section of the tire constituent member, Selection of two brushes, and further, a range of division positions when two brushes are selected, and various parameters used for calculating an optimum feed pitch, which will be described later, are input and set in the input setting unit 22 (step 100). .
Next, in the input setting unit 22, it is determined whether one brush is selected or two brushes are selected (step 102). When one brush is selected, numerical data of the thickness of the target shape 16 is stored in the memory unit 38. The feed pitch P is recorded by this method in the feed pitch calculation unit 26 for the target shape 16._nIs calculated (step 104).
[0042]
On the other hand, when two strokes are selected, the division position calculation unit 36 calculates the optimum division position (step 108). Specifically, the dividing point position is changed within a preset range of the dividing position, and each time the partial target shape divided at the dividing position is fed by the feed pitch calculating unit 26 at the feed pitch P._nFor each of the obtained feed pitches P_nThe distribution shape of the total added thickness made from is combined into one. Among the distribution shapes combined into one, the maximum error of the distribution shape with respect to the target shape 16 is minimized, or the feed pitch P_nIs obtained as a division position at which curve fitting is smoothly performed with a predetermined high-order polynomial, and this division position is set as the optimum division position. Then, a feed pitch P is calculated by the feed pitch calculation unit 26 for each of the two partial target shapes obtained by division at the optimum division position._nIs calculated. When calculating the optimum division position, the feed pitch P for each of the two partial target shapes_nIs already calculated, the calculation of the feed pitch is omitted, and the first feed pitch adjustment (step 106) and the second feed pitch adjustment (step 110) are performed.
[0043]
Next, the calculated feed pitch P_nIs sent to the first feed pitch adjusting unit 28, and the first feed pitch adjustment is performed (step 106).
That is, the feed pitch P of the nth roll calculated in step 104_nThe feed pitch P of the nth turn calculated in step 104, with the value of 1/2 of_nAnd the feed pitch P of the (n + 1) th roll calculated by the feed pitch calculation unit 26_{n + 1}For example, the feed pitch value is changed in increments of 0.1 mm within a range in which the value obtained by adding one-half of the value is the upper limit, and the changed n-th feed pitch and feed other than the n-th turn are changed. The average circumferential thickness t produced by the pitch_nThe distribution shape of the total added thickness is the start side end point position g after each winding._{n + 1}In (n = 1, 2,..., N−1), a feed pitch value that substantially matches the target shape is found, and this feed pitch is the feed pitch P._nIt is said.
[0044]
After the first feed pitch adjustment is performed, the second feed pitch adjustment unit 30 performs the second feed pitch adjustment (step 110).
N-th turn feed pitch P adjusted in step 106_nThe feed pitch P of the nth turn adjusted in step 106, with the value of 1/2 of_nAnd n + 1 roll feed pitch P_{n + 1}For example, by changing the value of the feed pitch in increments of 0.1 mm within the range where the value obtained by adding one-half of the value is the upper limit, the changed feed pitch of the nth roll and the feed other than the nth roll Average circumferential thickness t by pitch_nThe value of the feed pitch in which the distribution shape of the total added thickness produced from the above is substantially identical to the target shape 16 is found, and the feed pitch P_nIt is said. Whether or not the target shape 16 substantially matches is determined by determining the average circumferential thickness t_nIs obtained by determining the feed pitch of the nth turn that minimizes the weighted average error by weighted average of the maximum error or the average error or the maximum error and the average error with respect to the target shape 16 of the distribution shape of the total added thickness.
Such adjustment of the feed pitch is performed by adjusting the feed pitch P of the first roll₁It is done in order. Also, the feed pitch P of the last Nth roll_NMay be adjusted in the reverse order.
[0045]
Next, after the second feed pitch adjustment is performed, the third feed pitch adjustment unit 32 performs the third feed pitch adjustment (step 112).
Specifically, the feed pitch is adjusted in the reverse order of the order of the feed pitch adjusted by the second feed pitch adjusting unit 30. For example, the second feed pitch adjustment is the first feed pitch P₁The third feed pitch is adjusted in order from the feed pitch P of the last Nth roll._NThe second feed pitch adjustment is adjusted in the reverse order from the feed pitch P of the Nth roll._NFrom feed pitch P₁When it is performed in order toward₁Are adjusted in order. Thus, the feed pitch P in the reverse order of the second feed pitch adjustment._nBy adjusting the average circumferential thickness t as compared with the case of adjusting from one direction_nThe distribution shape of the total added thickness created by the above approaches the target shape 16.
[0046]
The feed pitch P thus analytically calculated or adjusted_nFrom the optimum feed pitch P using an optimum design method using a genetic algorithm (GA)_nIs calculated (step 114).
Here, the feed pitch P adjusted in step 112_n(N = 1, 2,... N) to P_n′ (N = 1, 2,... N), and this feed pitch P_n′ (N = 1, 2,... N) are combined into one feed pitch group (P_1,‘P₂', ... P_N′). The optimum feed pitch is calculated by the feed pitch group (P₁', P₂', ... P_N′) As an initial state, the optimum feed pitch group (P) that is closest to the target shape 16 of the cross section of the tire component₁, P₂, ... P_N) Is calculated. FIG. 8 shows the flow of such an optimum feed pitch calculation method.
[0047]
First, the feed pitch group (P₁', P₂', ... P_N'), K sets of feed pitch groups are generated, and these K sets of feed pitch groups are set as an initial solution set (step 114a).
Specifically, first, the feed pitch group (P₁', P₂', ... P_N′) Is used to calculate the maximum error with respect to the target shape of the total added thickness distribution shape_maxAnd this Δ_max2 times the value of 2¹⁶-1 divided by 2¹⁶Using the discrete values, the feed pitch P_n'Is expressed. That is, the feed pitch group (P₁', P₂', ... P_N') Each feed pitch P_n′ Is replaced with V represented by the following formula (8).
[0048]
[Expression 1]

Where xb_kIs the feed pitch P_nThis is a 0 or 1 value of the k-th bit when 'is represented in binary.
[0049]
In this way, for example, a feed pitch P in bit format represented by only 0 and 1 discretized by a 16-bit binary number._nBy generating

random numbers

0 and 1 from the feed pitch group consisting of ', K feed pitch groups are generated, and the K feed pitch groups are set as an initial solution set. Here, the number of bits representing one feed pitch is not limited to 16 bits, and is preferably 12 to 24 bits. Note that the number of sets of feed pitch groups (K sets) in the initial solution set is preferably about 10 times the number N of windings of the strip rubber member G in order to obtain an optimum feed pitch. For example, if the number of windings N of the strip rubber member G is 20, it is preferable to set the number of sets of feed pitch groups to about 200.
The distribution shape of the total added thickness using the feed pitches of the K sets of feed pitch groups generated in this way is the winding start end G shown in FIG._sStart side end point position g of the cross-sectional shape of₁To the confirmation points provided at predetermined intervals, for example, as thickness numerical data at the confirmation points provided every 1 mm.
[0050]
Next, fitness function F_i(I = 1, 2,... K) is set according to the input in step 100 (step 114b).
Fitness function F_iIs a function for extracting the maximum error from the errors between the thickness of the cross-sectional shape at the confirmation point obtained in step 114a and the thickness of the target shape 16 at the position corresponding to the position of the confirmation point.
That is, fitness function F_iFinds the maximum error between the distribution shape of the total added thickness of the average circumferential thicknesses created using the feed pitches of the K sets of feed pitch groups and the target shape 16. Fitness function F_iAs the mean square error between the cross-sectional shape and the target shape 16, the error between the cross-sectional shape and the target shape 16 may partially increase._iIt is preferable to use the maximum error as
[0051]
Next, a parent selection process is performed (step 114c).
In parent selection, as shown in FIGS. 9A and 9B, K / 2 sets of parent A and parent B are selected from the K sets of feed pitch groups as the solution set using, for example, a tournament method. .
In the tournament method, the first to Kth numbers are assigned to the K groups of feed pitch groups, and an integer random number from 1 to K is generated twice by the optimum feed pitch calculation unit 34. The two feed pitch groups corresponding to the two numbers are taken out, and the fitness function F obtained for the two feed pitch groups is obtained._iThe fitness function F is compared_iThe one with the smaller value is selected as parent A. Similarly, random numbers 1 to K are generated twice, two feed pitch groups corresponding to the two numbers generated by the generation of the random numbers are extracted, and the fitness function F obtained for the two feed pitch groups is obtained._iThe fitness function F is compared_iThe one with the smaller value is selected as parent B. In this way, a pair of parent A and parent B is selected. A pair of such parent A and parent B is selected as K / 2. The feed pitch groups of parent A and parent B generate random numbers 1 to K twice, and two feed pitch groups corresponding to the two numbers generated by this random number generation are selected from the K sets of solution sets. Therefore, there may be cases where the same feed pitch group composed of the same feed pitch is selected by the parent A and the parent B, and another group of parents A and B are configured by the same feed pitch. In some cases, a feed pitch group is selected.
[0052]
As shown in FIG. 9A, the feed pitch group (P₁', P₂', ... P_N′) Feed pitch P_nIn order to identify the value of ′, the feed pitch P of the feed pitch group given i-th_nThe value of ‘P’_n ⁽ⁱ⁾When describing (i = 1 to K), for example, as shown in FIG. 9B, the first set of parent A has a fifth feed pitch group, and parent B has a third. The third feed pitch group is selected for the second set of parent A, and the seventh feed pitch group is selected for parent B. That is, the third feed pitch group is selected twice as a parent.
[0053]
Next, a mating (crossing) process of parent A and parent B is performed between each group of parent A and parent B (step 114d).
In the mating of parent A and parent B, random numbers from 0 to 1 are generated by the optimum feed pitch calculation unit 34, and when the value obtained by this random number is a predetermined value, for example, 0.7 or less, the following mating is performed. Done. If it is larger than the predetermined value, the mating is not performed and the two feed pitch groups of the parents A and B are maintained as they are.
As shown in FIG. 10, the cross between parent A and parent B_n ⁽ⁱ⁾Parent A and feed pitch P_n ^(j)Is expressed in a 16-bit binary system, that is, xb represented by Expression (8)_kThese numerical values are arranged in a line in the order of 1 to N feed pitches, and a numerical string composed of 0 or 1 values of a total of 16 · N bits is made as feed pitch information. Further, an integer random number from 1 to (16 · N−1) is generated, and the numerical sequence A of the lower bits after the bit of the numerical sequence corresponding to the generated value^'And B 'are interchanged and mating is performed.
[0054]
Further, a mutation process is performed (step 114e).
In the mutation step, first, random numbers from 0 to 1 are generated, and when the value obtained by the random number is a predetermined value, for example, 0.08 or less, the numerical value of the bit is inverted. That is, it is 1 when the bit value is 0, and 0 when the bit value is 1. Such a mutation process is performed for each bit of the 16 · N-bit numeric string obtained by crossing. By providing the mutation process in this way, the optimal feed pitch group is determined by varying the feed pitch group of the solution set within a predetermined range.AccurateThis is to help you find out.
[0055]
Next, the obtained 16 · N-bit numeric value sequence is divided into 16-bit numbers in the order of the numeric value sequence, and the feed pitch value is expressed in decimal notation, and the feed pitch consisting of N feed pitch values Groups are generated according to equation (8).
In such a feed pitch group, two feed pitch groups are generated by crossing the parents A and B of each set, and even if there is no cross between the parents A and B, the two feed pitch groups of the parents A and B Are held as a feed pitch group that is generated as it is, K set feed pitch groups are generated from K / 2 sets of parents A and B (step 114f).
[0056]
Next, for the K sets of feed pitch groups generated in step 114f, the fitness function F_iIs calculated (step 114g). That is, the fitness function F is added to the K sets of feed pitch groups._iIs applied to calculate K values.
Such K sets of feed pitch groups are replaced with next-generation solution sets (step 114h).
[0057]
Next, the fitness function F calculated in step 114g_iIt is determined whether the minimum value of the value is a predetermined value, for example, 0.05 (mm) or less (step 114i). If it is equal to or less than the predetermined value, it is determined that the feed pitch group has converged, and the fitness Function F_iIs the optimum feed pitch group (P₁, P₂, ..., P_N).
The fitness function F_iIf the minimum value is larger than the predetermined value, it is determined that the feed pitch group still does not converge, and the K solution sets replaced in step 114h are returned to step 114c as the next-generation solution set. . In this case, the number of generation changes to the next generation is returned to step 114c within a range that does not exceed the preset generation change number, and the parent selection is performed again using the next generation solution set generated in step 114h. The mating process is repeated.
In this case, the fitness function F_iThe feed pitch group that maximizes the value of the fitness function F_iIs replaced with a feed pitch group that minimizes the value of.
[0058]
When the number of generation changes reaches a predetermined number of generation changes, for example, the number of generations, for example, it is assumed that the optimum feed pitch has not been found, and the optimum feed pitch is calculated by a new trial. That is, it is determined whether the number of trials is the maximum number of trials, which is the upper limit of the number of trials, for example, 20 or less (step 114k). When the number of trials is the first, Δ in equation (8)_maxUses the maximum error of the feed pitch group obtained by the third feed pitch adjustment in step 112 as described above, but when the number of trials is 2 or more times, the solution set of K feed pitch groups is A solution set of K feed pitch groups replaced in step 114h is used, and Δ in Equation (8) is used._maxIs the fitness function F_iThe maximum error calculated in is used.
[0059]
If the optimum feed pitch group is not calculated within the maximum number of trials, it is determined that the optimum feed pitch group has not been obtained, and the process ends. In this case, the fitness function F_iA feed pitch group that minimizes the value may be obtained as a feed pitch group that conforms to the optimum feed pitch group.
The above is the calculation of the optimum feed pitch (step 114).
The optimum feed pitch group (P₁, P₂, ..., P_N) Each feed pitch P_nIs used as a feed pitch in winding of the strip rubber member G in the method for manufacturing a tire constituent member described above.
[0060]
As an example, with respect to the feed pitch calculated by the selection of two strokes shown in FIG. 6C, an example in which the optimum feed pitch is calculated after the first feed pitch adjustment is shown in FIGS. 11A and 11B. Show. More specifically, parameters A ′, B ′, C ′, D ′ and q of the cross-sectional shape of the strip rubber member G shown in FIG. 1 (c) are set to 10 mm, 20 mm, 1 mm, 0.3 mm and 1. Assuming 0, the number of turns N of the strip rubber member G is 20 with respect to the target shape 16 shown in FIG. FIG. 11 (a) shows the distribution shape (dashed line) of the total added thickness of the average circumferential thickness after the first feed pitch adjustment, and FIG. 11 (b) shows the average after calculating the optimum feed pitch. The distribution shape (dashed line) of the total added thickness of the circumferential thickness is shown together with the target shape 16 (solid line). Although the maximum error between the distribution shape of the total added thickness in FIG. 6C and the target shape 16 is 0.5205 mm, the distribution shape of the total added thickness (dashed line) after the first feed pitch adjustment shown in FIG. ) And the target shape 16 (solid line) is reduced to 0.4972 mm, and the distribution shape (dashed line) of the total added thickness after calculation of the optimum feed pitch shown in FIG. 11B and the target shape 16 (solid line). The maximum error is reduced to 0.4000 mm.
[0061]
In addition, Table 1 shows 6 types of target shapes (6 cases) created from the selection of 1-stroke or 2-stroke selection, information on the number of windings N, and the feed pitch manually set by the operator. Maximum error E between the distribution shape of the total added thickness and the target shape₁And the maximum error E between the distribution shape of the total added thickness created from the feed pitch obtained by calculating the feed pitch and adjusting the first to third feed pitches and the target shape₂Further, the maximum error E between the distribution shape of the total added thickness created from the feed pitch optimized by the calculation method of the optimum feed pitch using the GA and the target shape_ThreeIs shown.
[0062]
[Table 1]

[0063]
From Table 1, the maximum error E₂Is the maximum error E₁Smaller, maximum error E_ThreeIs the maximum error E₂It can be seen that it is smaller in all cases. In case 6, the maximum error E₁Maximum error E compared to₂Are substantially the same, but this is because the setting of the feed pitch of the worker was relatively good. Cases 1-5 all have maximum error E₂Is the maximum error E₁It is getting smaller.
[0064]
As described above, in this embodiment, the feed pitch of the strip rubber member G is analytically calculated in steps 104 to 112, and thereafter, the distribution shape of the added total thickness obtained by adding and summing the average circumferential thickness to the target shape. The optimal feed pitch with the smallest error is calculated using GA, but the calculation of the optimal feed pitch by GA is to converge the feed pitch optimally unless the feed pitch is calculated and adjusted in advance. This is because the obtained optimum feed pitch is different every time it is calculated, and the solution of the optimum feed pitch may lack stability.
[0065]
The present invention is not necessarily limited to the above-described example in which the optimum feed pitch is calculated after the first to third feed pitch adjustments, but after the first feed pitch adjustment or the second feed pitch adjustment. Later, the optimum feed pitch in step 114 may be calculated.
Further, in the above example, the feed pitch of the strip rubber member G is changed every time it is wound, and the feed pitch is obtained every time it is wound. However, the present invention is not limited to the above example, and every second roll or 3 The feed pitch may be changed for each winding, or the feed pitch may be obtained.
[0066]
Further, the present invention provides a program for executing in the computer the method for calculating the feed pitch of the strip rubber member shown in FIG. Such a program is provided by being recorded on a known recording medium such as a floppy disk or a CD-ROM, or is carried by a carrier wave using a network such as the Internet.
That is, the program feeds the strip rubber member G when a long strip rubber member G is wound on the drum support device 10 while being at least partially overlapped to produce a tire constituent member having a predetermined cross section. The pitch is the average circumferential thickness t_nIs a program that is calculated by a calculation means such as a CPU (central processing unit) of the computer.
[0067]
Such a program includes a procedure for performing at least step 104 of steps 104 to 112 shown in FIG. 7 using the CPU, and a feed pitch group composed of a plurality of feed pitches calculated and adjusted by this procedure. , An initial solution set setting procedure for initializing a plurality of sets of feed pitch groups as a solution set and recording the solution set in a memory of a computer, and an average circumferential thickness t_nThe error of the distribution shape of the total added thickness obtained by adding the total to the target shape of the cross section of the tire constituent member is expressed as a fitness function F_iFitness function F from the fitness function setting procedure to be set as_iBased on the above, a plurality of pairs of feed pitch groups are set, the plurality of pairs are recorded in a memory, a pair setting procedure, and a part of the feed pitch information included in each of the set feed pitch group pairs are stored. A crossover procedure for crossing at a predetermined probability using a computer random number generation function, and a part of the feed pitch information obtained by crossover by this crossover procedure using a random number generation function for a computer And a replacement procedure that replaces the solution set of the feed pitch group before execution of the crossover procedure with a feed pitch group obtained by executing the change procedure and records the replaced solution set in the memory. The fitness function F is added to the solution set generated by this replacement procedure._IFitness function F obtained by applying_iThis is a program to be executed by a computer that repeats the pair setting procedure, the crossover procedure, the change procedure, and the replacement procedure until the minimum value of the value becomes equal to or less than a predetermined value. Since such a procedure is performed in the same process as the above-described method for calculating the feed pitch of the strip rubber member, the description thereof is omitted.
[0068]
As described above, the method for calculating the feed pitch of the strip rubber member of the present invention, the method for producing the tire constituent member using the same, and the program have been described in detail. However, the present invention is not limited to the above-described embodiment, and Of course, various improvements and modifications may be made without departing from the scope.
[0069]
【The invention's effect】
As described above in detail, according to the present invention, the average circumferential thickness of the strip rubber member is obtained, and the distribution shape of the added total thickness obtained by adding and summing the average circumferential thickness is the tire component member. Since the feed pitch is calculated so as to be substantially equal to the target shape of the cross section, when producing a tire component using a strip rubber member, the feed pitch of the strip rubber member is compared with the case where the operator manually sets the feed pitch based on experience. Thus, an error with respect to the target shape can be reduced.
Moreover, unlike the manual setting by the operator, no special skill or familiarity is required. In particular, by using an optimum design method based on GA, it is possible to drastically reduce errors with respect to the target shape.
[Brief description of the drawings]
FIGS. 1A to 1D are explanatory views illustrating an example of a method for producing a tire constituent member of the present invention.
FIG. 2 is an explanatory view illustrating an example of a method for producing a tire constituent member of the present invention.
FIG. 3 is a configuration diagram showing a configuration of an optimum feed pitch calculation apparatus that implements an example of a feed pitch calculation method for a strip rubber member according to the present invention.
FIGS. 4A to 4F are explanatory views for explaining an example of one step of a method for calculating a feed pitch of a strip rubber member according to the present invention.
FIGS. 5A and 5B are explanatory diagrams illustrating an example of another process of the strip rubber member feed pitch calculation method of the present invention. FIGS.
FIGS. 6A to 6C are explanatory views illustrating a cross-sectional shape obtained in one step of the strip rubber member feed pitch calculation method of the present invention.
FIG. 7 is a flowchart illustrating a flow of an example of a method for calculating a feed pitch of a strip rubber member according to the present invention.
FIG. 8 is a flowchart for explaining the flow of calculating the optimum feed pitch in the method for calculating the feed pitch of the strip rubber member of the present invention.
FIGS. 9A and 9B are explanatory diagrams for explaining a process of calculating an optimum feed pitch shown in FIG.
FIG. 10 is an explanatory diagram for explaining another process of calculating the optimum feed pitch shown in FIG.
FIGS. 11A and 11B are diagrams showing an example of a cross-sectional shape obtained by the method for calculating the feed pitch of the strip rubber member of the present invention.
[Explanation of symbols]
10 Drum support device
12 Drum surface
14 Tire intermediate assembly
16 Target shape
17, 18 Cross-sectional shape
19 Equivalent rectangular shape
20 Feed pitch calculation device
22 Input setting section
24 Average circumferential thickness calculator
26 Feed pitch calculator
28 First feed pitch adjustment section
30 Second feed pitch adjustment section
32 Third feed pitch adjustment section
34 Optimal feed pitch calculator
36 Division position calculation unit
38 Memory section

Claims

The drum support device when producing a tire constituent member having a predetermined cross-sectional shape by winding a long strip rubber member around the drum circumference while at least partially overlapping the drum support device in the drum width direction A method for calculating the feed pitch of the strip rubber member along the drum width direction of
Feed pitch calculation unit, the feed pitch is obtained by winding a strip rubber member, calculated based on the average circumferential thickness along the drum circumferential direction of the strip rubber member, the wrapping starting end of the strip rubber member A feed pitch calculating step for calculating each round of winding the strip rubber member, and an optimum feed pitch calculating step for calculating an optimum feed pitch by the optimum feed pitch calculating unit,
The optimum feed pitch calculation step includes
An initial solution set setting step of generating a plurality of sets of feed pitch groups using a feed pitch group consisting of a plurality of feed pitches obtained at least through the feed pitch calculation step, and initializing these feed pitch groups as a solution set When,
A fitness function setting step of setting, as a fitness function, an error with respect to the cross-sectional shape of the tire constituent member of the distribution shape of the total added thickness obtained by total addition of the average circumferential thickness obtained from the feed pitch of the feed pitch group;
From a solution set of feed pitch groups, a pair setting step for setting a plurality of pairs of feed pitch groups based on the fitness function;
A crossover step of crossing a part of the feed pitch information included in each of the set feed pitch group pairs with a predetermined probability;
A changing step of changing a part of the feed pitch information obtained by crossing in this crossing step with a predetermined probability;
A replacement step of replacing the solution set of the feed pitch group before the crossover step with the feed pitch group obtained through the change step,
The pair setting step, until the minimum value of the fitness function value obtained by applying the fitness function to all the feed pitch groups of the solution set obtained in the replacement step is equal to or less than a predetermined value, crossover process, the change process and the replacement by repeating the process, the feed pitch calculation method of the strip rubber member characterized that you calculate an optimum feed pitch.

The feed pitch calculation method for a strip rubber member according to claim 1 , wherein the feed pitch calculation step sequentially calculates the feed pitch in the order of the rounds of winding the strip rubber member.

It is a feed pitch calculation method of the strip rubber member according to claim 2 ,
The feed pitch calculation step includes
From the feed pitch already calculated before the calculation of the feed pitch to be calculated at the end position on the winding start side in the drum width direction in the cross section of the strip rubber member after winding around the feed pitch to be calculated The new thickness obtained by adding the thickness of the strip rubber member to the cumulative added thickness of the obtained average circumferential thickness is substantially equal to the thickness of the cross section at the corresponding position of the tire component corresponding to the end point position. A method for calculating a feed pitch of a strip rubber member, wherein the feed pitch to be calculated is calculated.

The thickness of the strip rubber member added to the cumulative added thickness in the feed pitch calculating step is a thickness of an equivalent rectangular cross-sectional shape having the same cross-sectional area and lateral width as the cross-sectional area and lateral width of the strip rubber member. Item 4. The method for calculating the feed pitch of the strip rubber member according to Item 3 .

A first feed pitch adjustment step of adjusting each of the feed pitches calculated in the feed pitch calculation step;
This first feed pitch adjusting step maintains the feed pitch value as it is except for the feed pitch to be adjusted, and the total added thickness obtained by changing the feed pitch value to be adjusted within a predetermined range. However, at the end position on the winding start side in the drum width direction in the cross section of the strip rubber member after the winding around the feed pitch to be adjusted, at the corresponding position of the tire constituent member corresponding to this end point position The feed pitch calculation method for a strip rubber member according to any one of claims 1 to 4 , wherein the feed pitch to be adjusted is adjusted so as to be substantially equal to a thickness of a cross section.

A second feed pitch adjustment step for adjusting each of the feed pitches adjusted in the first feed pitch adjustment step;
In the second feed pitch adjustment step, the feed pitch adjusted in the first feed pitch adjustment step holds the feed pitch value as it is except for the feed pitch to be adjusted, and the feed pitch to be adjusted. Adjusting the feed pitch to be adjusted so that the distribution shape of the total added thickness of the average circumferential thickness obtained by changing the value within a predetermined range is substantially equal to the cross-sectional shape of the tire constituent member The feed pitch calculation method of the strip rubber member according to claim 5 .

7. The strip rubber according to claim 6 , wherein the second feed pitch adjusting step sequentially adjusts the feed pitch in a turn order in which the strip rubber member is wound or in an order opposite to the turn order. Method for calculating the feed pitch of a member.

A third feed pitch adjustment step of adjusting each of the feed pitches adjusted in the second feed pitch adjustment step in an order opposite to the order of adjustment of the feed pitch adjusted in the second feed pitch adjustment step. Have
In the third feed pitch adjustment step, the feed pitch adjusted in the second feed pitch adjustment step holds the feed pitch value as it is except for the feed pitch to be adjusted, and the feed pitch value to be adjusted. the distribution shape of the total sum thickness obtained by changing the within a predetermined range, so as to be substantially equal to the cross-sectional shape of the tire component members, according to claim 7, characterized in that to adjust the feed pitch A method for calculating the feed pitch of the strip rubber member.

At least one of the second feed pitch adjustment step and the third feed pitch adjustment step includes a maximum error of a distribution shape of the total added thickness with respect to a cross-sectional shape of the tire constituent member, and a distribution of the total added thickness. shape, the average error for the cross-sectional shape of the tire component member, and a weighted average, claims 6 to adjusting the feed pitch to be the adjustment using one of the the maximum error and the average error 9. A method for calculating a feed pitch of a strip rubber member according to any one of claims 8 to 10.

In the pair setting step, a feed pitch group having a smaller fitness function value out of two feed pitch groups randomly selected from the solution set is set as one of a pair of feed pitch groups, and the solution the two feed pitch groups newly selected randomly from the set of claims 1-9, characterized in that the other pair of pitch groups feed the feed pitch group towards the value of the fitness function is less The feed pitch calculation method of the strip rubber member in any one .

The fitness function is a maximum error of a distribution shape of the total added thickness with respect to a cross-sectional shape of a tire component, or an average error of the distribution shape of the total added thickness with respect to a cross-sectional shape of a tire component and the maximum error. It is a weighted average, The feed pitch calculation method of the strip rubber member in any one of Claims 1-10 characterized by the above-mentioned.

The strip rubber member according to any one of claims 1 to 11 , wherein a feed pitch for winding the drum rubber around at least a plurality of long strip rubber members is calculated for each strip rubber member. Feed pitch calculation method.

When calculating the feed pitch of the strip rubber member for each divided sectional shape obtained by dividing the sectional shape of the tire constituent member into two,
The maximum error with respect to the cross-sectional shape of the tire constituent member of the distribution shape obtained by combining the distribution shapes of the total added thicknesses formed from the feed pitch set for each of the divided cross-sectional shapes into one is minimized. The strip rubber member feed pitch calculation method according to claim 12 , wherein a division position for dividing the cross-sectional shape of the tire constituent member into two is set.

When calculating the feed pitch of the strip rubber member for each divided sectional shape obtained by dividing the sectional shape of the tire constituent member into two,
The division position is set so that a feed pitch with respect to the winding order of the strip rubber members set for each of the divided cross-sectional shapes is optimally approximated by a high-order polynomial. 12. A method for calculating a feed pitch of the strip rubber member according to 12 .

A method for producing a tire constituent member, comprising producing a tire constituent member using a feed pitch calculated by using the feed pitch calculation method for a strip rubber member according to any one of claims 1 to 14 .

When producing a tire component having a predetermined cross-sectional shape by winding a long strip rubber member on the drum circumference while at least partially overlapping a long strip rubber member on the drum support device, in the drum width direction of the drum support device A program for causing a computer to execute a method for obtaining a feed pitch of the strip rubber member along the line,
The feed pitch is calculated by a calculation means of a computer based on an average circumferential thickness along the drum circumferential direction of the strip rubber member, and the calculation means is used for each round of winding the strip rubber member. The procedure to find the feed pitch,
A plurality of feed pitch groups are generated using the obtained feed pitch group consisting of a plurality of feed pitches, these feed pitch groups are initialized as a solution set, and the solution set is recorded in the storage means of the computer. Initial solution set setting procedure,
A fitness function setting procedure for setting, as a fitness function, an error with respect to the cross-sectional shape of the tire constituent member of the distribution shape of the total added thickness obtained by total addition of the average circumferential thickness obtained from the feed pitch of the feed pitch group;
From the solution set of the feed pitch group, a pair setting procedure for setting a plurality of pairs of feed pitch groups based on the fitness function and recording the plurality of pairs in the storage means;
A crossover procedure for crossing a part of the feed pitch information of each of the set feed pitch group pairs with each other with a predetermined probability using a computer random number generation means;
A change procedure for changing a part of the feed pitch information obtained by crossing by this crossing procedure at a predetermined probability using the random number generating means,
A replacement procedure in which the solution set of the feed pitch group before the crossover procedure is executed is replaced with the feed pitch group obtained through the change procedure, and the replacement solution set is recorded in the storage means;
The pair setting procedure and the crossover procedure until the minimum value of the fitness function value obtained by applying the fitness function to the solution set of the feed pitch group obtained by the replacement procedure is equal to or less than a predetermined value. by repeating the changing procedure and the replacement procedure, a program which is characterized that you calculate an optimum feed pitch group.