JP2004056856A - Motor controller - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform the abnormality diagonosis of an encoder in a system which F/B-controls a motor as far as the target position, based on an encoder count value. <P>SOLUTION: When it becomes unable to perform the F/B control of the motor normally, it switches the control into open loop control. Also each time it outputs a motor drive signal in the open loop control, it counts up (or counts down) the positional count values at open loop control. At the point of time when this positional count value reaches the objective count value, it judges that a rotor has reached the target position, and finishes the open loop control. In this open loop control, it judges the existence of the wire breaking of the phase A/B of the encoder by comparing the count value of the encoder (the count value of the A-phase signal/B-phase signal of the encoder) with the positional count value (the count value of the motor drive signal), and also judges the existence of the abnormality of a Z-phase signal, based on the output interval of the Z-phase signal. <P>COPYRIGHT: (C)2004,JPO

Description

【０１３０】
上記位置カウント値Ｎｃｎｔｏｐの算出式において、Ｋｏｐ×６１／３．７５はディテントレバー１５の最大回転角Ｋｏｐをモータ駆動ステップ数（励磁回数）に換算した値である。「６１」は減速機構２６の減速比であり、「３．７５」は１ステップ（励磁１回当たり）のロータ３２の回転角である。この位置カウント値Ｎｃｎｔｏｐは、オープンループ制御開始時の位置カウント値Ｎｃｎｔｏｐの初期値として用いられる。
【０１３１】
位置カウント値Ｎｃｎｔｏｐの算出後、ステップ８０５に進み、リカバリ処理の実行回数をカウントするリカバリ処理実行回数カウンタＣｒｃｖをカウントアップし、次のステップ８０６で、オープンループ制御実行フラグＸｏｐｅｎをＯＮにセットして、後述する図３７のオープンループ制御ルーチンによりオープンループ制御を実行する。
【０１３２】
リカバリ処理（オープンループ制御）の実行中は、本ルーチンが起動される毎に、ステップ８０１で「Ｎｏ」と判定されて、ステップ８０７に進み、オープンループ制御完了フラグＸｏｐｃｍｐがオープンループ制御完了を意味するＯＮにセットされているか否かを判定し、Ｘｏｐｃｍｐ＝ＯＦＦであれば、そのまま本ルーチンを終了し、オープンループ制御を継続する。
【０１３３】
その後、オープンループ制御が完了した時点で、ステップ８０８に進み、初期駆動を再実行するために、初期駆動終了フラグＸｅｎｄ＝ＯＦＦ（初期駆動未終了）、レンジ切換操作フラグＸｃｈｇ＝ＯＦＦ（レンジ切換操作なし）、通電時間カウンタＣＴ＝０、励磁回数カウンタＣＡＳＥ＝０、エンコーダカウント値Ｎｃｎｔ＝０、初期位置ずれ学習値Ｇｃｎｔ＝０にセットする。そして、次のステップ８１１で、オープンループ制御実行フラグＸｏｐｅｎ＝ＯＦＦにセットしてオープンループ制御（リカバリ処理）を終了した後、ステップ８１２に進み、リカバリ処理実行フラグＸｒｃｖ＝ＯＦＦ、Ａ相信号、Ｂ相信号の異常検出フラグＸｆａｂ＝ＯＦＦ、Ｘｆｔａ＝ＯＦＦ、Ｘｆｔｂ＝ＯＦＦ、後述する第３の異常検出フラグＸｆｒｃｖ＝ＯＦＦにセットする。
【０１３４】
［Ｆ／Ｂ制御中ＡＢ相信号異常検出］
図３０に示すＦ／Ｂ制御中ＡＢ相信号異常検出ルーチンは、所定周期（例えば１ｍｓ周期）で起動され、Ｆ／Ｂ制御中にエンコーダ４６のＡ相信号とＢ相信号の異常の有無を次のようにして判定する。本ルーチンが起動されると、まずステップ８２１で、Ｆ／Ｂ許可フラグＸｆｂ＝ＯＮ（Ｆ／Ｂ制御中）であるか否かを判定し、Ｆ／Ｂ許可フラグＸｆｂ＝ＯＦＦであれば、ステップ８２２に進み、Ａ相信号カウンタＮａ＝０、Ｂ相信号カウンタＮｂ＝０、Ａ相信号エッジ間隔カウンタΔＴａ＝０、Ｂ相信号エッジ間隔カウンタΔＴｂ＝０にリセットして本ルーチンを終了する。
【０１３５】
これに対して、ステップ８２１でＦ／Ｂ許可フラグＸｆｂ＝ＯＮ（Ｆ／Ｂ制御中）と判定された場合は、ＡＢ相信号異常判定実行条件が成立して、ステップ８２３に進み、Ａ相信号カウント値ＮａとＢ相信号カウント値Ｎｂとの偏差ΔＮａｂを算出する。
ΔＮａｂ＝｜Ｎａ−Ｎｂ｜
ここで、Ａ相信号カウント値ＮａとＢ相信号カウント値Ｎｂは、Ａ相信号のエッジとＢ相信号のエッジを別々にカウントした値である。
【０１３６】
そして、次のステップ８２４で、ＡＢ相信号カウント値偏差ΔＮａｂを異常判定値Ｋａｂ（例えば３）と比較し、もし、ＡＢ相信号カウント値偏差ΔＮａｂが異常判定値Ｋａｂ以上であれば、Ａ相信号カウント値ＮａとＢ相信号カウント値Ｎｂのいずれかが異常であると判断して、ステップ８２５に進み、ＡＢ相信号異常検出フラグＸｆａｂをＯＮ（異常あり）にセットして、次のステップ８２６に進む。
【０１３７】
一方、ＡＢ相信号カウント値偏差ΔＮａｂが異常判定値Ｋａｂ未満の場合は、ＡＢ相信号異常検出フラグＸｆａｂをＯＦＦ（異常なし）に維持してステップ８２６に進む。
【０１３８】
Ｆ／Ｂ制御中にＡ相信号とＢ相信号が正常に出力されていれば、Ａ相信号カウント値ＮａとＢ相信号カウント値Ｎｂとの偏差ΔＮａｂが０又は１となるため、このＡＢ相信号カウント値偏差ΔＮａｂが２以上になれば、いずれか一方の信号の抜けやノイズパルスの混入が考えられる（図２５参照）。従って、ＡＢ相信号カウント値偏差ΔＮａｂが異常判定値Ｋａｂ以上であるか否かで、Ａ相信号とＢ相信号の異常の有無を判定することができる。本実施形態では、異常判定値Ｋａｂを余裕を見て３に設定したが、２又は４以上の値に設定しても良い。
【０１３９】
以上のようにして、ＡＢ相信号カウント値偏差ΔＮａｂによる異常判定処理を行った後、ステップ８２６に進み、Ａ相信号のエッジ間隔ΔＴａとＢ相信号のエッジ間隔ΔＴｂをカウントするカウンタをそれぞれカウントアップした後、ステップ８２７に進み、Ａ相信号エッジ間隔ΔＴａを異常判定値Ｋｔａ（例えばΔＴａの平均値の１０倍の値）と比較し、もし、Ａ相信号エッジ間隔ΔＴａが異常判定値Ｋｔａ以上であれば、Ａ相信号が異常であると判断して、ステップ８２８に進み、Ａ相信号異常検出フラグＸｆｔａをＯＮ（異常あり）にセットして、次のステップ８２９に進む。
【０１４０】
一方、Ａ相信号エッジ間隔ΔＴａが異常判定値Ｋｔａ未満の場合は、Ａ相信号異常検出フラグＸｆｔａをＯＦＦ（異常なし）に維持してステップ８２９に進む。そして、このステップ８２９で、Ｂ相信号エッジ間隔ΔＴｂを異常判定値Ｋｔｂ（例えばΔＴｂの平均値の１０倍の値）と比較し、もし、Ｂ相信号エッジ間隔ΔＴｂが異常判定値Ｋｔｂ以上であれば、Ｂ相信号が異常であると判断して、ステップ８３０に進み、Ｂ相信号異常検出フラグＸｆｔｂをＯＮ（異常あり）にセットして本ルーチンを終了する。
【０１４１】
一方、Ｂ相信号エッジ間隔ΔＴｂが異常判定値Ｋｔｂ未満の場合は、Ｂ相信号異常検出フラグＸｆｔｂをＯＦＦ（異常なし）に維持して、本ルーチンを終了する。
【０１４２】
各信号のエッジ間隔ΔＴａ，ΔＴｂは、ロータ回転速度に応じて変化するため、Ｆ／Ｂ制御中に、各信号のエッジ間隔ΔＴａ，ΔＴｂが長くなり過ぎる場合は、ロータ３２の異常停止、異常速度低下、パルス抜けが考えられるため（図２５参照）、各信号のエッジ間隔ΔＴａ，ΔＴｂが異常判定値Ｋｔａ，Ｋｔａ（例えばΔＴａ，ΔＴｂの平均値の１０倍）以上であるか否かで、Ａ相信号とＢ相信号の異常の有無を判定することができる。
【０１４３】
尚、各信号のエッジ間隔ΔＴａ，ΔＴｂが短くなり過ぎる場合は、ノイズパルスが正規のパルス信号と誤認されている可能性があるため、各信号のエッジ間隔ΔＴａ，ΔＴｂが所定の異常判定値以下であるか否かで、Ａ相信号とＢ相信号の異常の有無を判定するようにしても良い。
【０１４４】
また、ＡＢ相信号カウント値偏差ΔＮａｂによる異常判定処理（ステップ８２３〜８２５）と各信号のエッジ間隔ΔＴａ，ΔＴｂによる異常判定処理（ステップ８２６〜８３０）のいずれか一方の処理のみを行い、他方の処理を省略するようにしても良い。
【０１４５】
［逆転カウンタ］
図３１に示す逆転カウンタルーチンは、ＡＢ相割り込み処理により実行され、ロータ３２の回転方向が回転方向指示に反して連続して逆転した回数を次のようにしてカウントする。本ルーチンが起動されると、まずステップ８３１で、回転方向指示値Ｄが１（Ｐレンジ→ＮｏｔＰレンジの回転方向）、０（停止）、−１（ＮｏｔＰレンジ→Ｐレンジの回転方向）のいずれに該当するかを判定し、回転方向指示値Ｄが０（停止）であれば、以降の処理を行うことなく、本ルーチンを終了する。
【０１４６】
一方、回転方向指示値Ｄが１（Ｐレンジ→ＮｏｔＰレンジの回転方向）であれば、ステップ８３２に進み、前回のエンコーダカウント値Ｎｃｎｔ（ｉ−１）　と今回のエンコーダカウント値Ｎｃｎｔ（ｉ）　との差が１以上であるか否かで、ロータ３２の回転方向が回転方向指示に反して逆転したか否か（エンコーダカウント値Ｎｃｎｔが減少したか否か）を判定し、逆転していれば、ステップ８３４に進み、逆転回数をカウントする逆転カウンタＣｒをカウントアップし、逆転していなければ、ステップ８３５に進み、逆転カウンタＣｒを０にリセットする。
【０１４７】
また、回転方向指示値Ｄが−１（ＮｏｔＰレンジ→Ｐレンジの回転方向）であれば、ステップ８３３に進み、前回のエンコーダカウント値Ｎｃｎｔ（ｉ−１）　と今回のエンコーダカウント値Ｎｃｎｔ（ｉ）　との差が−１以下であるか否かで、ロータ３２の回転方向が回転方向指示に反して逆転したか否か（エンコーダカウント値Ｎｃｎｔが増加したか否か）を判定し、逆転していれば、ステップ８３６に進み、逆転カウンタＣｒをカウントアップし、逆転していなければ、ステップ８３５に進み、逆転カウンタＣｒを０にリセットする。
【０１４８】
［逆転判定］
図３２に示す逆転判定ルーチンは、所定周期（例えば１ｍｓ周期）で起動される。本ルーチンが起動されると、まずステップ８４１で、Ｆ／Ｂ許可フラグＸｆｂ＝ＯＮ（Ｆ／Ｂ制御中）であるか否かを判定し、Ｆ／Ｂ許可フラグＸｆｂ＝ＯＦＦであれば、ステップ８４４に進み、逆転検出フラグＸｒをＯＦＦにセットして、本ルーチンを終了する。
【０１４９】
これに対して、ステップ８４１でＦ／Ｂ許可フラグＸｆｂ＝ＯＮ（Ｆ／Ｂ制御中）と判定された場合は、逆転判定実行条件が成立して、ステップ８４２に進み、図３１の逆転カウンタルーチンで求めた逆転カウント値Ｃｒが異常判定値Ｋｒ（例えば４）以上であるか否かで、ロータ３２が回転方向指示に反して逆転し続けているか否かを判定し、逆転カウント値Ｃｒが異常判定値Ｋｒ以上であれば、ステップ８４３に進み、逆転検出フラグＸｒをＯＮにセットして逆転の判定を確定し、逆転カウント値Ｃｒが異常判定値Ｋｒ未満であれば、ステップ８４４に進み、逆転検出フラグＸｒをＯＦＦに維持する。
【０１５０】
［システム異常診断］
図３３に示すシステム異常診断ルーチンは、所定周期（例えば８ｍｓ周期）で起動される。本ルーチンが起動されると、まずステップ８５１で、次の▲１▼、▲２▼の異常診断実行条件が成立しているか否かを判定する。
▲１▼イグニッションスイッチ（図示せず）の操作によりＥＣＵ４１に電源が投入された後、所定時間（例えば１ｓ）が経過していること
▲２▼リカバリ処理実行中でないこと（リカバリ処理実行フラグＸｒｃｖ＝ＯＦＦであること）
【０１５１】
これら２つの条件▲１▼、▲２▼のうち、いずれか一方でも満たさない条件があれば、異常診断実行条件が不成立となり、ステップ８５２に進み、後述する仮異常状態継続時間カウンタＣｆｎｓｗ、Ｃｆｃｎｔを０にリセットして、本ルーチンを終了する。
【０１５２】
これに対し、上記２つの条件▲１▼、▲２▼を両方とも満たせば、異常診断実行条件が成立し、ステップ８５１からステップ８５４に進み、出力軸センサ１４が正常で、且つ、指令シフトレンジと、出力軸センサ１４の出力信号Ｖｎｓｗから検出したレンジとが一致しているか否かを判定し、その判定結果が「Ｙｅｓ」であれば、ステップ８５６に進み、第１の仮異常状態継続時間カウンタＣｆｎｓｗを０にリセットする。
【０１５３】
これに対し、上記ステップ８５４の判定結果が「Ｎｏ」であれば、ステップ８５５に進み、第１の仮異常状態継続時間カウンタＣｆｎｓｗをカウントアップして、指令シフトレンジと出力軸センサ１４の出力信号Ｖｎｓｗとが一致していない状態の継続時間を計測し、次のステップ８５７で、第１の仮異常状態継続時間カウンタＣｆｎｓｗの値が異常判定値Ｋｆｎｓｗよりも大きいか否かを判定する。ここで、異常判定値Ｋｆｎｓｗは、正常なレンジ切換動作に要する最長時間よりも十分に長い時間（例えば１０ｓ）に設定されている。従って、第１の仮異常状態継続時間カウンタＣｆｎｓｗの値が異常判定値Ｋｆｎｓｗよりも大きければ、レンジ切換動作が正常に行われない状態になっているため、システムの異常ありと判断して、ステップ８５８に進み、第１の異常検出フラグＸｆｎｓｗをＯＮにセットする。上記ステップ８５７の判定結果が「Ｎｏ」であれば、第１の異常検出フラグＸｆｎｓｗをＯＦＦ（異常なし）に維持する。
【０１５４】
以上のようにして、出力軸センサ１４の出力信号Ｖｎｓｗを用いて１回目の異常診断を行った後、図３４のステップ８５９に進み、エンコーダカウント値Ｎｃｎｔと目標カウント値Ａｃｎｔとの偏差｜Ｎｃｎｔ−Ａｃｎｔ｜が所定値Ｋｃｎｔ（例えば５°相当値）よりも大きいか否かを判定し、このカウント値偏差｜Ｎｃｎｔ−Ａｃｎｔ｜が所定値Ｋｃｎｔ以下であれば、ステップ８６１に進み、第２の仮異常状態継続時間カウンタＣｆｃｎｔを０にリセットする。
【０１５５】
これに対し、上記ステップ８５９の判定結果が「Ｙｅｓ」であれば、ステップ８６０に進み、第２の仮異常状態継続時間カウンタＣｆｃｎｔをカウントアップして、カウント値偏差｜Ｎｃｎｔ−Ａｃｎｔ｜が所定値Ｋｃｎｔよりも大きい状態の継続時間を計測し、次のステップ８６２で、第２の仮異常状態継続時間カウンタＣｆｃｎｔの値が異常判定値Ｋｆｃｎｔよりも大きいか否かを判定する。ここで、異常判定値Ｋｆｃｎｔは、正常なレンジ切換動作に要する最長時間より十分に長い時間（例えば１０ｓ）に設定されている。従って、第２の仮異常状態継続時間カウンタＣｆｃｎｔの値が異常判定値Ｋｆｃｎｔよりも大きければ、レンジ切換動作が正常に行われない状態になっているため、システムの異常ありと判断して、ステップ８６３に進み、第２の異常検出フラグＸｆｃｎｔをＯＮにセットする。上記ステップ８６２の判定結果が「Ｎｏ」であれば、第２の異常検出フラグＸｆｃｎｔをＯＦＦ（異常なし）に維持する。
【０１５６】
以上のようにして、カウント値偏差｜Ｎｃｎｔ−Ａｃｎｔ｜を用いて２回目の異常診断を行った後、ステップ８６４に進み、図２８のリカバリ処理ルーチンのステップ８０５でカウントされるリカバリ処理実行回数Ｃｒｃｖが異常判定値Ｋｃｒｖ（例えば３回）以上であるか否かを判定し、リカバリ処理実行回数Ｃｒｃｖが異常判定値Ｋｃｒｖ以上であれば、ステップ８６５に進み、システムの異常ありと判断して、ステップ８６５に進み、第３の異常検出フラグＸｆｒｃｖをＯＮにセットして本ルーチンを終了する。つまり、リカバリ処理が連続して異常判定値Ｋｃｒｖ（例えば３回）以上も繰り返される場合は、Ｆ／Ｂ制御系が正常に機能しない状態になっているため、システムの異常ありと判断するものである。
【０１５７】
一方、リカバリ処理実行回数Ｃｒｃｖが異常判定値Ｋｃｒｖ未満であれば、第３の異常検出フラグＸｆｒｃｖをＯＦＦ（異常なし）に維持して本ルーチンを終了する。
【０１５８】
［フェールセーフ処理］
図３５に示すフェールセーフ処理ルーチンは、所定周期（例えば８ｍｓ周期）で起動される。本ルーチンが起動されると、まずステップ９０１で、次の▲１▼、▲２▼のフェールセーフ処理実行条件が成立しているか否かを判定する。
▲１▼オープンループ制御によるレンジ切り換えが可能な状態であること（システム故障フラグＸｆａｉｌｏｆｆ＝ＯＦＦであること）
▲２▼図３３のシステム異常診断ルーチンでセットされる３つの異常検出フラグＸｆｎｓｗ、Ｘｆｃｎｔ、Ｘｆｒｃｖのうちの少なくとも１つがＯＮ（異常あり）であること、又は、後述する図４４のＡＢ相断線検出ルーチンでセットされるＡＢ相断線検出フラグＸｆａｉｌａｂがＯＮ（ＡＢ相断線あり）であること
【０１５９】
これら２つの条件▲１▼、▲２▼のうち、いずれか一方でも満たさない条件があれば、フェールセーフ処理実行条件が不成立となり、以降の処理を行うことなく、本ルーチンを終了する。
【０１６０】
これに対し、上記２つの条件▲１▼、▲２▼を両方とも満たせば、フェールセーフ処理実行条件が成立し、ステップ９０１からステップ９０２に進み、フェールセーフ処理実行フラグＸｆｓｏｐをＯＮにセットし、次のステップ９０３で、警告ランプを点灯したり、或は、メータパネルの情報表示部等に警告表示して、運転者に警告する。
【０１６１】
この後、ステップ９０４に進み、前回のフェールセーフ処理終了後の経過時間を計測するフェールセーフ処理間隔カウンタＣｆｓｏｐの値が所定値Ｋｆｓｏｐ未満であるか否かを判定する。ここで、所定値Ｋｆｓｏｐは、前回のフェールセーフ処理終了後にディテントバネ２３の弾性力によってディテントバネ２３の係合部２３ａがディテントレバー１５の保持凹部２４，２５の傾斜側壁に沿ってその底部に滑り落ちるのに要する時間よりもある程度長い時間（例えば５００ｍｓ）に設定されている。
【０１６２】
フェールセーフ処理間隔カウンタＣｆｓｏｐの値（前回のフェールセーフ処理終了後の経過時間）が所定値Ｋｆｓｏｐ未満であれば、ステップ９０５に進み、フェールセーフ処理間隔カウンタＣｆｓｏｐをカウントアップして前回のフェールセーフ処理終了後の経過時間をカウントする。この後、ステップ９０６に進み、全相の通電をオフした状態を継続し、次のステップ９０７で、通電フラグＸｏｎを通電オフを意味するＯＦＦにセットして本ルーチンを終了する。これにより、前回のフェールセーフ処理終了後の経過時間が所定値Ｋｆｓｏｐを越えるまでは、全相の通電をオフした状態に維持される。
【０１６３】
その後、フェールセーフ処理間隔カウンタＣｆｓｏｐの値（前回のフェールセーフ処理終了後の経過時間）が所定値Ｋｆｓｏｐ以上になった時点で、ステップ９０４からステップ９０８に進み、オープンループ制御実行フラグＸｏｐｅｎがＯＦＦであるか否かを判定し、ＯＦＦであれば、ステップ９０９に進み、指令シフトレンジｓｆｔが切り換えられたか否かを判定する。その結果、指令シフトレンジｓｆｔが切り換えられていないと判定されれば、ステップ９１０に進み、全相の通電をオフした状態を継続する。
【０１６４】
これに対して、上記ステップ９０９で、指令シフトレンジｓｆｔが切り換えられたと判定されれば、ステップ９１１に進み、前記図２８のリカバリ処理ルーチンのステップ８０４と同じ方法で、現在の出力軸センサ１４の出力信号Ｖｎｓｗを用いて現在の位置カウント値Ｎｃｎｔｏｐを算出する。この位置カウント値Ｎｃｎｔｏｐは、オープンループ制御開始時の位置カウント値Ｎｃｎｔｏｐの初期値として用いられる。
【０１６５】
この後、ステップ９１２に進み、オープンループ制御実行フラグＸｏｐｅｎをＯＮにセットして、後述する図３７のオープンループ制御ルーチンによりオープンループ制御を実行し、次のステップ９１３で、フェールセーフ処理の実行回数をカウントするフェールセーフ処理実行回数カウンタＣｆｓをカウントアップして、本ルーチンを終了する。
【０１６６】
また、前記ステップ９０８で、オープンループ制御実行フラグＸｏｐｅｎがＯＮであると判定されれば、図３６のステップ９１４に進み、オープンループ制御完了フラグＸｏｐｃｍｐがオープンループ制御完了を意味するＯＮにセットされているか否かを判定し、Ｘｏｐｃｍｐ＝ＯＦＦであれば、以降の処理を行うことなく、本ルーチンを終了する。
【０１６７】
これに対し、ステップ９１４で、オープンループ制御完了フラグＸｏｐｃｍｐがＯＮにセットされていると判定されれば、ステップ９１５に進み、指令シフトレンジと、出力軸センサ１４の出力信号Ｖｎｓｗから検出したレンジとが一致しているか否かを判定し、両者が一致していれば、ステップ９１６に進み、フェールセーフ処理実行回数カウンタＣｆｓを０にリセットする。そして、次のステップ９１８で、オープンループ制御完了フラグＸｏｐｃｍｐをＯＦＦにセットすると共に、次のステップ９１９で、オープンループ制御実行フラグＸｏｐｅｎをＯＦＦにセットして、本ルーチンを終了する。
【０１６８】
また、上記ステップ９１５で、指令シフトレンジと出力軸センサ１４の出力信号Ｖｎｓｗとが一致していないと判定された場合は、ステップ９１７に進み、フェールセーフ処理実行回数カウンタＣｆｓの値が故障判定値（例えば３回）以上であるか否かを判定し、フェールセーフ処理実行回数カウンタＣｆｓの値が故障判定値未満であれば、まだ故障とは判定せずに、オープンループ制御完了フラグＸｏｐｃｍｐ＝ＯＦＦ、オープンループ制御実行フラグＸｏｐｅｎ＝ＯＦＦにセットして（ステップ９１８、９１９）、本ルーチンを終了する。
【０１６９】
その後、フェールセーフ処理実行回数カウンタＣｆｓの値が故障判定値（例えば３回）以上になれば、その時点で、システムの故障（オープンループ制御でもレンジ切り換えが不可能な状態）と判断して、ステップ９１７からステップ９２０に進み、システム故障フラグＸｆａｉｌｏｆｆをＯＮにセットし、次のステップ９２１で、全相の通電をオフして、本ルーチンを終了する。
【０１７０】
［オープンループ制御］
図３７乃至図３９に示すオープンループ制御ルーチンは、所定周期（例えば１ｍｓ周期）で起動され、特許請求の範囲でいうオープンループ制御手段としての役割を果たす。本ルーチンが起動されると、まずステップ９３１で、オープンループ制御実行フラグＸｏｐｅｎがＯＮにセットされているか否かを判定し、Ｘｏｐｅｎ＝ＯＦＦであれば、以降の処理を行うことなく、本ルーチンを終了する。
【０１７１】
これに対し、オープンループ制御実行フラグＸｏｐｅｎがＯＮにセットされていれば、ステップ９３１からステップ９３２に進み、時間を計測する時間カウンタＴｏｐをカウントアップし、次のステップ９３３で、時間カウンタＴｏｐの値が現在の通電相の励磁時間ｔｍ以上になったか否かを判定し、励磁時間ｔｍ未満であれば、以降の処理を行うことなく、本ルーチンを終了する。これにより、ステップ９３４以降の処理は、励磁時間ｔｍ毎に実行される。
【０１７２】
その後、時間カウンタＴｏｐの値が現在の通電相の励磁時間ｔｍ以上になった時点で、ステップ９３３からステップ９３４に進み、目標カウント値Ａｃｎｔと現在の位置カウント値Ｎｃｎｔｏｐとを比較して、ロータ３２の駆動方向を判定する。その結果、現在の位置カウント値Ｎｃｎｔｏｐが目標カウント値Ａｃｎｔよりも小さい場合は、ロータ３２の駆動方向がＰレンジ側からＮｏｔＰレンジ側への回転方向（正回転方向）と判断して、ステップ９３５に進み、通電相判定値Ｍｐｔｎｏｐを１ずつカウントアップし、次のステップ９３６で、位置カウント値Ｎｃｎｔｏｐを２ずつカウントアップする。これは、１回の励磁でロータ３２が２カウント分（３．７５°）回転するためである。
【０１７３】
これに対し、現在の位置カウント値Ｎｃｎｔｏｐが目標カウント値Ａｃｎｔよりも大きい場合は、ロータ３２の駆動方向がＮｏｔＰレンジ側からＰレンジ側への回転方向（逆回転方向）と判断して、ステップ９３７に進み、通電相判定値Ｍｐｔｎｏｐを１ずつカウントダウンし、次のステップ９３８で、位置カウント値Ｎｃｎｔｏｐを２ずつカウントダウンする。
【０１７４】
この後、ステップ９３９に進み、通電相判定値Ｍｐｔｎｏｐを“６”で割り算した時の余りＭｐｔｎｏｐ％６を算出する。ここで、“６”は、オープンループ制御で通電相を一巡させる間の通電相判定値Ｍｐｔｎｏｐの増減量に相当する。
【０１７５】
Ｍｐｔｎｏｐ％６の算出後、ステップ９４０に進み、図４０のテーブルによってＭｐｔｎｏｐ％６に応じて通電相を選択し、その通電相を励磁する。本実施形態では、オープンループ制御中は、１−２相励磁方式で駆動するように通電相を選択し、更に、オープンループ制御を開始するときの最初の励磁が２相通電（ＷＵ相通電又はＵＶ相通電又はＶＷ相通電）、つまりＭｐｔｎｏｐ％６＝１又は３又は５となるように通電相判定値Ｍｐｔｎｏｐの初期値が設定されている。
【０１７６】
この後、図３８のステップ９４１に進み、位置偏差ｄｅｖを次式により算出する。
ｄｅｖ＝｜Ａｃｎｔ−Ｎｃｎｔｏｐ｜／２
【０１７７】
上式から明らかなように、位置偏差ｄｅｖは、目標カウント値Ａｃｎｔと現在の位置カウント値Ｎｃｎｔｏｐとの偏差（絶対値）の１／２である。
【０１７８】
そして、次のステップ９４２で、位置偏差ｄｅｖと回転速度パラメータｍｓｐｄとを比較する。この回転速度パラメータｍｓｐｄは、位置偏差ｄｅｖに応じて増減されるパラメータであり、この回転速度パラメータｍｓｐｄに応じて図４２のテーブルにより励磁時間ｔｍを設定することで、ロータ３２の回転速度、加速、減速を制御する。この場合、目標とする回転速度が高くなるほど、回転速度パラメータｍｓｐｄが大きくなり、励磁時間ｔｍが短い時間に設定されるようになっている。この回転速度パラメータｍｓｐｄの初期値は０である。
【０１７９】
上記ステップ９４２で、位置偏差ｄｅｖが回転速度パラメータｍｓｐｄより大きいと判定された場合は、ステップ９４３に進み、回転速度パラメータｍｓｐｄを１だけカウントアップし、次のステップ９４４で、最小回転速度パラメータｍｓｐｄｍｉｎを０にセットする。この最小回転速度パラメータｍｓｐｄｍｉｎは、回転速度パラメータｍｓｐｄの最小値を制限するパラメータ（下限ガード値）である。このように、位置偏差ｄｅｖが回転速度パラメータｍｓｐｄより大きいときに、最小回転速度パラメータｍｓｐｄｍｉｎを０にセットすることで、オープンループ制御開始からロータ３２が回転し始めるまでの期間に、励磁時間ｔｍを最長時間（５０ｍｓ）に設定できるようにしている。
【０１８０】
一方、上記ステップ９４２で、位置偏差ｄｅｖが回転速度パラメータｍｓｐｄより小さいと判定された場合は、ステップ９４５に進み、回転速度パラメータｍｓｐｄを１だけカウントダウンし、次のステップ９４６で、最小回転速度パラメータｍｓｐｄｍｉｎを４にセットする。これにより、ロータ３２の回転が立ち上がった後は、励磁時間ｔｍを例えば５ｍｓ以下に制限できるようにしている。
【０１８１】
また、上記ステップ９４２で、位置偏差ｄｅｖが回転速度パラメータｍｓｐｄと一致すると判定された場合は、回転速度パラメータｍｓｐｄを変化させずに、ステップ９４７に進み、最小回転速度パラメータｍｓｐｄｍｉｎを０にセットする。
【０１８２】
以上のようにして、回転速度パラメータｍｓｐｄと最小回転速度パラメータｍｓｐｄｍｉｎを設定した後、ステップ９４８に進み、図４１のテーブルを用いてバッテリ電圧（電源電圧）に応じて最大回転速度パラメータｍｓｐｄｍａｘを設定する。この最大回転速度パラメータｍｓｐｄｍａｘは、回転速度パラメータｍｓｐｄの最大値を制限するパラメータ（上限ガード値）であり、バッテリ電圧が高くなるほど、最大回転速度パラメータｍｓｐｄｍａｘを大きくして、ロータ３２の回転速度を速くできるようにしている。
【０１８３】
そして、次のステップ９４９で、回転速度パラメータｍｓｐｄを最小回転速度パラメータｍｓｐｄｍｉｎと最大回転速度パラメータｍｓｐｄｍａｘとの範囲内に収めるようにガード処理する。例えば、ステップ９４３又は９４５で更新した回転速度パラメータｍｓｐｄが最小回転速度パラメータｍｓｐｄｍｉｎよりも小さい場合は、ｍｓｐｄ＝ｍｓｐｄｍｉｎとし、また、回転速度パラメータｍｓｐｄが最大回転速度パラメータｍｓｐｄｍａｘよりも大きい場合は、ｍｓｐｄ＝ｍｓｐｄｍａｘとし、ｍｓｐｄｍｉｎ≦ｍｓｐｄ≦ｍｓｐｄｍａｘの場合は、その回転速度パラメータｍｓｐｄをそのまま用いる。
【０１８４】
この後、ステップ９５０に進み、図４２のテーブルを用いて、回転速度パラメータｍｓｐｄに応じて励磁時間ｔｍを設定する。図４２の例では、回転速度パラメータｍｓｐｄが０〜３までの範囲では、励磁時間ｔｍを最長時間（例えば５０ｍｓ）に設定し、起動時にロータ３２の回転位相と通電相との同期を確実にとることができるようにしている。そして、回転速度パラメータｍｓｐｄが４〜９までの範囲では、回転速度パラメータｍｓｐｄが大きくなるに従って、励磁時間ｔｍを例えば５ｍｓから３ｍｓまで３段階に小さくして、ロータ３２の回転速度を３段階に調整できるようになっている。尚、励磁時間ｔｍの切り換えは、２段階以下又は４段階以上であっても良いことは言うまでもない。
【０１８５】
励磁時間ｔｍの設定後、ステップ９５１に進み、通電フラグＸｏｎをＯＮにセットし、次のステップ９５２で、時間カウンタＴｏｐをクリアして本ルーチンを終了する。
【０１８６】
また、図３７のステップ９３４で、現在の位置カウント値Ｎｃｎｔｏｐが目標カウント値Ａｃｎｔと一致すると判定された場合は、図３９のステップ９５５に進み、時間カウンタＴｏｐの計測時間が所定時間Ｋｈｌｄ（例えば５００ｍｓ）以上になったか否かを判定する。その結果、時間カウンタＴｏｐの計測時間が所定時間Ｋｈｌｄ未満と判定された場合は、ステップ９５６に進み、全相の通電をオフし、次のステップ９５７で、励磁時間ｔｍを所定時間Ｋｈｌｄ（例えば５００ｍｓ）にセットする。
【０１８７】
この後、図３８のステップ９５２に進み、時間カウンタＴｏｐをクリアして本ルーチンを終了する。このような処理により、オープンループ制御終了後、所定時間Ｋｈｌｄ（例えば５００ｍｓ）が経過するまで、前記図３７のステップ９３４以降の処理が行われないため、次のオープンループ制御が開始されない。これは、オープンループ制御終了後に、ディテントバネ２３の弾性力によりディテントバネ２３の係合部２３ａがディテントレバー１５の保持凹部２４，２５の傾斜側壁に沿ってその底部に滑り落ちるまで待つためである。
【０１８８】
また、図３９のステップ９５５で、時間カウンタＴｏｐの値が所定値Ｋｈｌｄ以上になったと判定されれば、ステップ９５８に進み、通電相判定値Ｍｐｔｎｏｐ＝０、回転速度パラメータｍｓｐｄ＝０にセットし、次のステップ９５９で、オープンループ制御完了フラグＸｏｐｃｍｐをＯＮにセットする。この後、ステップ９６０に進み、励磁時間Ｔｍを０にセットし、次のステップ９６１で、通電フラグＸｏｎをＯＦＦにセットした後、図３８のステップ９５２に進み、時間カウンタＴｏｐをクリアして本ルーチンを終了する。
【０１８９】
以上説明した本ルーチンでは、オープンループ制御中の励磁方式を１−２相励磁方式に設定し、オープンループ制御開始時の最初の励磁を２相通電で行うと共に、図２７に示すように、オープンループ制御開始から例えば３ステップ分は、各通電相の励磁時間ｔｍを最長時間（例えば５０ｍｓ）に設定して、ロータ３２を確実に掴んでロータ３２の回転位相と通電相とを確実に同期させる。この後は、励磁時間ｔｍを大幅に短くするが、ロータ３２の回転速度の上昇に伴って励磁時間ｔｍを徐々に短くして、ロータ３２の回転速度を緩やかに上昇させる。
【０１９０】
このオープンループ制御中のロータ３２の回転位置を推定するために、オープンループ制御開始時のロータ３２の初期位置（位置カウント値Ｎｃｎｔｏｐの初期値）を出力軸センサ１４の出力信号に基づいて算出し、その後、通電相を切り換える毎に位置カウント値Ｎｃｎｔｏｐを２ずつカウントアップ（又はカウントダウン）して、該位置カウント値Ｎｃｎｔｏｐに基づいてロータ３２の回転位置を推定し、該位置カウント値Ｎｃｎｔｏｐと目標カウント値Ａｃｎｔとの偏差がある程度小さくなった段階で、減速領域に入ったと判断して、励磁時間ｔｍを徐々に長くして、ロータ３２の回転速度を緩やかに減速し、該位置カウント値Ｎｃｎｔｏｐが目標カウント値Ａｃｎｔに到達した時点で、ロータ３２が目標位置に到達したと判断して、オープンループ制御を終了する。
【０１９１】
尚、本ルーチンでは、励磁時間ｔｍを回転速度パラメータｍｓｐｄに応じて設定するようにしたが、駆動開始後の回転量、駆動開始後の経過時間、目標位置までの回転量、負荷トルクの少なくとも１つに基づいて励磁時間ｔｍを設定するようにしても良い。
【０１９２】
また、本ルーチンでは、オープンループ制御開始時の最初の励磁を２相通電としたが、最初の励磁を１相通電としても良く、更には、オープンループ制御中の励磁方式も１−２相励磁方式に限定されず、１相励磁方式又は２相励磁方式を採用しても良い。
【０１９３】
［オープンループ制御時エンジントルク抑制］
図４３に示すオープンループ制御時エンジントルク抑制ルーチンは、所定周期（例えば１ｍｓ周期）で起動される。本ルーチンが起動されると、まずステップ９７１で、オープンループ制御実行フラグＸｏｐｅｎ＝ＯＮ（オープンループ制御実行中）であるか否かを判定し、Ｘｏｐｅｎ＝ＯＮであれば、ステップ９７２に進み、エンジントルク抑制フラグＸｔｒｑをＯＮにセットし、Ｘｏｐｅｎ＝ＯＦＦであれば、ステップ９７３に進み、エンジントルク抑制フラグＸｔｒｑをＯＦＦにセットする。このエンジントルク抑制フラグＸｔｒｑの情報は、レンジ切換制御用のＥＣＵ４１からエンジン制御用のＥＣＵ（図示せず）に送信される。
【０１９４】
エンジン制御用のＥＣＵは、オープンループ制御実行フラグＸｏｐｅｎ＝ＯＮの情報を受信したときに、エンジントルクを抑制する方向にエンジン制御パラメータを制御して、エンジントルクを抑制した状態で、オープンループ制御によるレンジ切換動作を実行する。
【０１９５】
この際、エンジントルクを抑制する方向に制御するエンジン制御パラメータは、空気系、燃料系、点火系のうちの少なくとも１つの制御パラメータを用いれば良く、例えば、空気系の場合は、スロットルバルブの開駆動を禁止してスロットル開度を現状維持するようにしたり、或は、運転に支障を来さない範囲内でスロットル開度を閉じるようにしても良い。また、燃料系の場合は、運転に支障を来さない範囲内で燃料噴射量を制限するようにすれば良く、点火系の場合は、運転に支障を来さない範囲内で点火時期を遅角するようにすれば良い。
【０１９６】
［ＡＢ相断線検出］
図４４に示すＡＢ相断線検出ルーチンは、ＡＢ相割り込み処理により起動される。本ルーチンが起動されると、まずステップ５０１で、オープンループ制御実行フラグＸｏｐｅｎ＝ＯＮ（オープンループ制御実行中）であるか否かを判定し、オープンループ制御実行フラグＸｏｐｅｎ＝ＯＮ（オープンループ制御実行中）であれば、ステップ５０２に進み、オープンループ制御開始時（オープンループ制御実行フラグＸｏｐｅｎがＯＦＦからＯＮに切り換わった直後）であるか否かを判定する。その結果、オープンループ制御開始時と判定されれば、ステップ５０３に進み、オープンループ制御開始時のエンコーダカウント値Ｎｃｎｔと位置カウント値Ｎｃｎｔｏｐとの差分をオープンループ制御開始時のカウント値ずれ量ΔＮｉｎｉｔとして求める。
ΔＮｉｎｉｔ＝Ｎｃｎｔ−Ｎｃｎｔｏｐ
【０１９７】
これに対して、オープンループ制御開始後（オープンループ制御中）であれば、上記ステップ５０２で「Ｎｏ」と判定されて、ステップ５０４に進み、オープンループ制御中のエンコーダカウント値Ｎｃｎｔと位置カウント値Ｎｃｎｔｏｐとのずれ量ΔＮｅｒｒを算出する。
ΔＮｅｒｒ＝Ｎｃｎｔ−Ｎｃｎｔｏｐ
【０１９８】
この後、ステップ５０５に進み、オープンループ制御終了時（オープンループ制御実行フラグＸｏｐｅｎがＯＮからＯＦＦに切り換わった直後）であるか否かを判定し、オープンループ制御終了時でなければ、以降の処理を行うことなく、本ルーチンを終了するが、オープンループ制御終了時であれば、ステップ５０６に進み、オープンループ制御終了時のカウント値ずれ量ΔＮｅｒｒとオープンループ制御開始時のカウント値ずれ量ΔＮｉｎｉｔとの偏差の絶対値｜ΔＮｅｒｒ−ΔＮｉｎｉｔ｜を判定値と比較して、エンコーダ４６のＡ相・Ｂ相が断線しているか否かを判定する。
【０１９９】
例えば、オープンループ制御によってロータ３２が正常に駆動された場合、エンコーダ４６のＡ相・Ｂ相が断線していなければ、オープンループ制御終了時のカウント値ずれ量ΔＮｅｒｒとオープンループ制御開始時のカウント値ずれ量ΔＮｉｎｉｔとの偏差｜ΔＮｅｒｒ−ΔＮｉｎｉｔ｜が０になるが、Ａ相・Ｂ相のいずれか一方又は両方が断線していれば、｜ΔＮｅｒｒ−ΔＮｉｎｉｔ｜が０にならない。従って、｜ΔＮｅｒｒ−ΔＮｉｎｉｔ｜が判定値以上であれば、エンコーダ４６のＡ相・Ｂ相の断線ありと判断して、ステップ５０７に進み、ＡＢ相断線検出フラグＸｆａｉｌａｂをＡ相・Ｂ相の断線ありを意味するＯＮにセットして本ルーチンを終了する。
【０２００】
一方、｜ΔＮｅｒｒ−ΔＮｉｎｉｔ｜が判定値未満であれば、エンコーダ４６のＡ相・Ｂ相の断線無しと判断して、ステップ５０７に進み、ＡＢ相断線検出フラグＸｆａｉｌａｂをＡ相・Ｂ相の断線なしを意味するＯＦＦにセットし、次のステップ５０９で、ＡＢ相正常判定フラグＸｎｏｒｍａｂを正常意味するＯＮにセットして本ルーチンを終了する。
【０２０１】
尚、本ルーチンでは、オープンループ制御終了時のカウント値ずれ量ΔＮｅｒｒとオープンループ制御開始時のカウント値ずれ量ΔＮｉｎｉｔとの偏差の絶対値｜ΔＮｅｒｒ−ΔＮｉｎｉｔ｜を算出するようにしたが、オープンループ制御開始時から終了時までのエンコーダカウント値Ｎｃｎｔの変化量と位置カウント値Ｎｃｎｔｏｐの変化量を算出し、両者の偏差の絶対値が判定値以上であるか否かで、エンコーダ４６のＡ相・Ｂ相が断線しているか否かを判定するようにしても良い。これは、オープンループ制御開始時から終了時までのエンコーダカウント値Ｎｃｎｔの変化量と位置カウント値Ｎｃｎｔｏｐの変化量との偏差の絶対値｜Ｎｃｎｔ変化量−Ｎｃｎｔｏｐ変化量｜は、｜ΔＮｅｒｒ−ΔＮｉｎｉｔ｜と等しくなるためである。
【０２０２】
尚、エンコーダ４６のＡ相・Ｂ相の断線検出は、オープンループ制御時にしかできない。つまり、Ｆ／Ｂ制御中は、断線によりＡ相・Ｂ相信号が入力されなくなると、通電をホールドするため（Ａ相・Ｂ相信号のエッジで駆動パルスを出力するため）、モータ１２の停止を判定することはできるが、これがエンコーダ４６の故障に起因するものかどうかを特定することはできない。
【０２０３】
［Ｚ相断線検出］
図４５に示すＺ相断線検出ルーチンは、ＡＢ相割り込み処理により起動され、次のようにしてエンコーダ４６のＺ相の断線の有無を判定する。本ルーチンが起動されると、まずステップ５１１で、オープンループ制御実行フラグＸｏｐｅｎ＝ＯＮ（オープンループ制御実行中）であるか否かを判定し、オープンループ制御実行フラグＸｏｐｅｎ＝ＯＦＦであれば、ステップ５１２に進み、Ｚ相断線仮検出カウンタＣｚｎｇをクリアしてステップ５１８に進み、Ｚ相信号出力間隔カウンタＣｚｏｎをクリアする。
【０２０４】
また、オープンループ制御実行フラグＸｏｐｅｎ＝ＯＮ（オープンループ制御実行中）であれば、ステップ５１３に進み、Ｚ相信号の値が１（ハイレベル）であるか否かを判定し、Ｚ相信号の値が１であれば、ステップ５１４に進み、Ｚ相正常判定フラグＸｎｏｒｍｚを正常意味するＯＮにセットしてステップ５１８に進み、Ｚ相信号出力間隔カウンタＣｚｏｎをクリアしてステップ５１９に進む。
【０２０５】
これに対し、Ｚ相信号の値が０（ローレベル）であれば、ステップ５１５に進み、Ｚ相信号の出力間隔を計測するためのＺ相信号出力間隔カウンタＣｚｏｎをカウントアップする。本ルーチンは、ＡＢ相割り込み処理により起動されるため、Ｚ相信号出力間隔カウンタＣｚｏｎは、エンコーダカウント値Ｎｃｎｔと同じく、Ａ相信号とＢ相信号の立ち上がり／立ち下がりの両方のエッジに同期してカウントアップされる。
【０２０６】
本実施形態では、Ｚ相信号の出力間隔（ロータ３２の基準回転位置の間隔）がロータ回転角で４５°であり、Ａ相信号とＢ相信号のエッジ間隔が３．７５°であるため、Ｚ相信号の出力間隔をＺ相信号出力間隔カウンタＣｚｏｎのカウント値（エンコーダカウント値Ｎｃｎｔ）に換算すると、４５°÷３．７５°＝１２となる。従って、Ｚ相信号出力間隔カウンタＣｚｏｎのカウント値が１２を越えてもＺ相信号が出力されない場合は、Ｚ相信号が抜けたことを意味する。
【０２０７】
そこで、本ルーチンでは、ステップ５１６で、Ｚ相信号出力間隔カウンタＣｚｏｎのカウント値が２４を越えたか否かを判定することで、Ｚ相信号が２回以上連続して出力されていないか否かを判定し、その結果、Ｚ相信号が２回以上連続して出力されていないと判定されれば、ステップ５１７に進み、Ｚ相異常仮検出カウンタＣｚｎｇをカウントアップして、ステップ５１８に進み、Ｚ相信号出力間隔カウンタＣｚｏｎをクリアしてステップ５１９に進む。
上述したステップ５１６で、Ｚ相信号出力間隔カウンタＣｚｏｎのカウント値が２４を越えていないと判定されれば、ステップ５１９に進む。
【０２０８】
このステップ５１９では、Ｚ相異常仮検出カウンタＣｚｎｇのカウント値が判定値（例えば３）以上であるか否かで、エンコーダ４６のＺ相の断線の可能性があるか否かを判定し、Ｚ相異常仮検出カウンタＣｚｎｇのカウント値が判定値未満であれば、まだ、Ｚ相断線診断の途中であると判断して、本ルーチンを終了する。
【０２０９】
これに対し、Ｚ相異常仮検出カウンタＣｚｎｇのカウント値が判定値以上であれば、エンコーダ４６のＺ相の断線又はＡ相・Ｂ相信号の異常のどちらかであると判断して、ステップ５２０に進み、ＡＢ相正常判定フラグＸｎｏｒｍａｂがＯＮであるか否かで、エンコーダ４６のＡ相・Ｂ相が正常であるか否かを判定する。その結果、ＡＢ相正常判定フラグＸｎｏｒｍａｂ＝ＯＮ（Ａ相・Ｂ相が正常）であると判定されれば、Ｚ相の断線と判断して、ステップ５２１に進み、Ｚ相異常検出フラグＸｆａｉｌｚをＺ相断線（異常）を意味するＯＮにセットして、ステップ５２２に進み、Ｚ相補正禁止フラグＸｚｎｇをＯＮにセットして、本ルーチンを終了する。
【０２１０】
Ｚ相補正禁止フラグＸｚｎｇがＯＮにセットされた場合は、図２２のＺ相補正ルーチンによるＺ相補正が禁止される。
【０２１１】
［Ｚ相信号異常検出］
図４６に示すＺ相信号異常検出ルーチンは、Ｚ相割り込み処理により起動され、次のようにしてエンコーダ４６のＺ相信号の異常の有無を判定する。本ルーチンが起動されると、まずステップ５３１で、前回のＺ相信号から今回のＺ相信号までのＺ相信号出力間隔ΔＮｚを次式により算出する。
ΔＮｚ＝｜Ｎｃｎｔ−Ｎｃｎｔｏｌｄ｜
上式において、Ｎｃｎｔは今回のＺ相信号出力時のエンコーダカウント値であり、Ｎｃｎｔｏｌｄは前回のＺ相信号出力時のエンコーダカウント値である。
【０２１２】
そして、次のステップ５３２で、Ｚ相信号出力間隔ΔＮｚが０又は１２であるか否かを判定する。Ｚ相信号出力間隔ΔＮｚが０であれば、Ａ相・Ｂ相が断線していることを意味し、Ｚ相信号出力間隔ΔＮｚが１２であれば、Ｚ相信号出力間隔ΔＮｚが正常であることを意味する。
【０２１３】
従って、上記ステップ５３２で、Ｚ相信号出力間隔ΔＮｚが０又は１２であると判定されれば、ステップ５３３に進み、Ｚ相正常判定フラグＸｎｏｒｍｚを正常を意味するＯＮにセットして、ステップ５３８に進み、次回のＺ相信号出力間隔ΔＮｚを演算するために、今回のＺ相信号出力時のエンコーダカウント値Ｎｃｎｔを前回値Ｎｃｎｔｏｌｄとして記憶して本ルーチンを終了する。
【０２１４】
一方、Ｚ相信号出力間隔ΔＮｚが０又は１２でない場合は、Ｚ相信号が異常である可能性があるため、ステップ５３２からステップ５３４に進み、Ｚ相異常仮検出カウンタＣｚｏｎｎｇをカウントアップする。そして、次のステップ５３５で、Ｚ相異常仮検出カウンタＣｚｏｎｎｇのカウント値が判定値（例えば３）以上であるか否かで、エンコーダ４６のＺ相信号の異常の可能性があるか否かを判定し、Ｚ相異常仮検出カウンタＣｚｏｎｎｇのカウント値が判定値未満であれば、まだ、Ｚ相信号異常診断の途中であると判断して、ステップ５３８に進み、今回のＺ相信号出力時のエンコーダカウント値Ｎｃｎｔを前回値Ｎｃｎｔｏｌｄとして記憶して本ルーチンを終了する。
【０２１５】
これに対し、Ｚ相異常仮検出カウンタＣｚｏｎｎｇのカウント値が判定値以上であれば、エンコーダ４６のＺ相信号の異常であると判断して、ステップ５３６に進み、Ｚ相異常検出フラグＸｆａｉｌｚを異常を意味するＯＮにセットして、ステップ５３７に進み、Ｚ相補正禁止フラグＸｚｎｇをＯＮにセットして、図２２のＺ相補正ルーチンによるＺ相補正を禁止する。そして、ステップ５３８に進み、今回のＺ相信号出力時のエンコーダカウント値Ｎｃｎｔを前回値Ｎｃｎｔｏｌｄとして記憶して本ルーチンを終了する。
尚、図４４〜図４６の各ルーチンは、特許請求の範囲でいう異常診断手段としての役割を果たす。
【０２１６】
以上説明した本実施形態によれば、オープンループ制御中（フェールセーフ処理及びリカバリ処理中）に、エンコーダカウント値Ｎｃｎｔ（エンコーダ４６のＡ相信号・Ｂ相信号のカウント値）と位置カウント値Ｎｃｎｔｏｐ（モータ駆動信号のカウント値）とを比較して、エンコーダ４６のＡ相・Ｂ相の断線の有無を判定するようにしたので、オープンループ制御中にエンコーダ４６のＡ相・Ｂ相の断線が検出されたときは、オープンループ制御からＦ／Ｂ制御への復帰を禁止することができ、無駄なＦ／Ｂ制御を行わずに済むと共に、オープンループ制御によってロータ３２を目標位置まで回転させることができる。
【０２１７】
更に、本実施形態では、オープンループ制御中にＺ相信号出力間隔に基づいてエンコーダ４６のＺ相信号の異常の有無を判定するようにしたので、Ｚ相信号の異常が検出されときは、Ｚ相補正を禁止することができ、誤ったＺ相補正を行うことを回避することができる。
【０２１８】
尚、本発明に用いるエンコーダは、磁気式のエンコーダ４６に限定されず、例えば、光学式のエンコーダやブラシ式のエンコーダを用いても良い。
また、本発明に用いるモータは、ＳＲモータ１２に限定されず、エンコーダの出力信号のカウント値に基づいてロータの回転位置を検出してモータの通電相を順次切り換えるブラシレス型のモータであれば、ＳＲモータ以外のブラシレス型のモータを用いても良い。
【０２１９】
また、本実施形態のレンジ切換装置は、ＰレンジとＮｏｔＰレンジの２つのレンジを切り換える構成であるが、例えば、ディテントレバー１５の回動動作に連動して自動変速機のレンジ切換弁とマニュアルバルブを切り換えて、自動変速機のＰ、Ｒ、Ｎ、Ｄ、…の各レンジを切り換えるレンジ切換装置にも本発明を適用して実施できる。
【０２２０】
その他、本発明は、レンジ切換装置に限定されず、ＳＲモータ等のブラシレス型のモータを駆動源とする各種の装置に適用して実施できることは言うまでもない。
【図面の簡単な説明】
【図１】本発明の一実施形態を示すレンジ切換装置の斜視図
【図２】ＳＲモータの構成を説明する図
【図３】ＳＲモータを駆動する回路構成を示す回路図
【図４】レンジ切換装置の制御システム全体の構成を概略的に示す図
【図５】エンコーダのロータリマグネットの構成を説明する平面図
【図６】エンコーダの側面図
【図７】（ａ）はエンコーダの出力波形を示すタイムチャート、（ｂ）は通電相切り換えパターンを示すタイムチャート
【図８】初期駆動ルーチンの処理の流れを示すフローチャート（その１）
【図９】初期駆動ルーチンの処理の流れを示すフローチャート（その２）
【図１０】Ｐレンジ初期駆動ルーチンの処理の流れを示すフローチャート
【図１１】ＮｏｔＰレンジ初期駆動ルーチンの処理の流れを示すフローチャート
【図１２】Ｐレンジで初期駆動を行ったときの制御例を示すタイムチャート
【図１３】初期駆動中にレンジ切換操作されたときの制御例を示すタイムチャート
【図１４】初期駆動中にレンジ切換操作されたときの初期位置ずれ学習方法を説明するタイムチャート
【図１５】エンコーダカウンタルーチンの処理の流れを示すフローチャート
【図１６】カウントアップ値ΔＮ算出マップの一例を示す図
【図１７】指令レンジシフト、Ａ相信号、Ｂ相信号、エンコーダカウント値の関係を示すタイムチャート
【図１８】モータＦ／Ｂ制御ルーチンの処理の流れを示すフローチャート
【図１９】通電相設定ルーチンの処理の流れを示すフローチャート
【図２０】Ｍｐｔｎ％１２から通電相への変換テーブルの一例を示す図
【図２１】通電処理を説明するタイムチャート
【図２２】Ｚ相補正ルーチンの処理の流れを示すフローチャート
【図２３】Ｚ相補正を説明するタイムチャート
【図２４】リカバリ処理の一例を説明するタイムチャート
【図２５】エンコーダのＡ相信号・Ｂ相信号の異常を検出する方法を説明する図
【図２６】Ｆ／Ｂ制御中のロータの逆転を検出する方法を説明する図
【図２７】オープンループ制御の一例を説明するタイムチャート
【図２８】リカバリ処理ルーチンの処理の流れを示すフローチャート
【図２９】オープンループ制御開始時の位置カウント値Ｎｃｎｔの初期値を算出する方法を説明する図
【図３０】Ｆ／Ｂ制御中ＡＢ相信号異常検出ルーチンの処理の流れを示すフローチャート
【図３１】逆転カウンタルーチンの処理の流れを示すフローチャート
【図３２】逆転判定ルーチンの処理の流れを示すフローチャート
【図３３】システム異常診断ルーチンの処理の流れを示すフローチャート（その１）
【図３４】システム異常診断ルーチンの処理の流れを示すフローチャート（その２）
【図３５】フェールセーフ処理ルーチンの処理の流れを示すフローチャート（その１）
【図３６】フェールセーフ処理ルーチンの処理の流れを示すフローチャート（その２）
【図３７】オープンループ制御ルーチンの処理の流れを示すフローチャート（その１）
【図３８】オープンループ制御ルーチンの処理の流れを示すフローチャート（その２）
【図３９】オープンループ制御ルーチンの処理の流れを示すフローチャート（その３）
【図４０】Ｍｐｔｎｏｐ％６から通電相への変換テーブルの一例を示す図
【図４１】バッテリ電圧から最大回転速度パラメータｍｓｐｄｍａｘを求めるテーブルの一例を示す図
【図４２】回転速度パラメータｍｓｐｄから励磁時間ｔｍを求めるテーブルの一例を示す図
【図４３】オープンループ制御時エンジントルク抑制ルーチンの処理の流れを示すフローチャート
【図４４】ＡＢ相断線検出ルーチンの処理の流れを示すフローチャート
【図４５】Ｚ相断線検出ルーチンの処理の流れを示すフローチャート
【図４６】Ｚ相信号異常検出ルーチンの処理の流れを示すフローチャート
【符号の説明】
１１…レンジ切換機構、１２…ＳＲモータ、１４…出力軸センサ、１５…ディテントレバー、１８…パーキングロッド、２０…パーキングギヤ、２１…ロックレバー、２３…ディテントバネ、２３ａ…係合部、２４…Ｐレンジ保持凹部、２５…ＮｏｔＰレンジ保持凹部、２６…減速機構、２７…自動変速機、３１…ステータ、３２…ロータ、３３，３４…巻線、３５，３６…モータ励磁部、３７，３８…モータドライバ（駆動回路）、４１…ＥＣＵ（Ｆ／Ｂ制御手段，オープンループ制御手段，異常診断手段）、４３…Ｐレンジスイッチ、４４…ＮｏｔＰレンジスイッチ、４６…エンコーダ、４７…ロータリマグネット、４８…Ａ相信号用の磁気検出素子、４９…Ｂ相信号用の磁気検出素子、５０…Ｚ相信号用の磁気検出素子。[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a motor control device that detects a rotational position of a rotor based on a count value of a pulse signal of an encoder and sequentially switches a current-carrying phase of the motor to rotate the rotor to a target position.
[0002]
[Prior art]
In recent years, brushless motors such as switch reluctance motors, which have been increasing in demand as simple and inexpensive motors, are equipped with an encoder that outputs a pulse signal in synchronization with the rotation of the rotor. In some cases, the rotational position of the rotor is detected based on the encoder count value, and the rotor is rotationally driven by sequentially switching the energized phase. Since such a motor with an encoder can detect the rotation position of the rotor based on the encoder count value after startup, the position switching control (rotation of the rotor to the target position by a feedback control system (closed loop control system)) is performed. It is used as a drive source for various position switching devices that perform positioning control).
[0003]
[Problems to be solved by the invention]
As described above, in a system that performs feedback control (hereinafter, referred to as “F / B control”) for rotating a rotor to a target position based on an encoder count value, when an encoder output signal becomes abnormal, the energized phase (encoder count) Value) and the rotation phase of the rotor cannot be synchronized, so that the motor cannot be driven normally and control becomes impossible.
[0004]
The present invention has been made in view of such circumstances, and accordingly, the object is to enable the rotor to be rotated to the target position even when the output signal of the encoder becomes abnormal, and to diagnose the abnormality of the encoder. An object of the present invention is to provide a motor control device that can perform the control with high accuracy and can improve the reliability of the motor control.
[0005]
[Means for Solving the Problems]
In order to achieve the above object, a motor control device according to a first aspect of the present invention is configured such that when an F / B control execution condition is satisfied, a rotation position of a rotor is determined by an F / B control unit based on an encoder count value. Is detected and F / B control for rotating the rotor to the target position is performed by sequentially switching the energized phase of the motor. When the F / B control execution condition is not satisfied due to a failure or the like, the open loop control means Switching from F / B control to open loop control and outputting a drive signal to the drive circuit of the motor without feeding back information on the encoder count value, sequentially switching the energized phase, counting the drive signal and reducing the count value to the count value The rotor is driven to rotate to the target position based on the rotation. Then, the abnormality diagnosing unit compares the encoder count value counted during the open loop control with the count value of the drive signal to determine whether or not the encoder is abnormal.
[0006]
During the F / B control, if the encoder count value becomes abnormal due to, for example, missing pulses or noise from the encoder, or if the signal line of the encoder is broken, the synchronization between the energized phase (encoder count value) and the rotational phase of the rotor is synchronized. And the F / B control state (rotor state of the rotor) becomes abnormal. In such a state, the F / B control execution condition is not satisfied, and the F / B control is switched to the open loop control. With this configuration, even if the encoder count value becomes abnormal due to, for example, missing pulses or noise from the encoder, or the signal line of the encoder is broken, the rotor can be rotated to the target position by open-loop control. . When the encoder count value counted during the open-loop control is compared with the count value of the drive signal, the encoder count value (the amount of rotation detected by the encoder) becomes equal to the count value of the drive signal (the amount of drive rotation by the open-loop control). Can be accurately determined whether or not the encoder is abnormal.
[0007]
It should be noted that the failure determination of the encoder can be performed only during the open loop control. In other words, during the F / B control, when the pulse signal of the encoder is not input due to the disconnection, the stop of the motor is determined to hold the energization (to output a drive pulse in synchronization with the pulse signal of the encoder). Can be done, but it cannot be determined whether this is due to an encoder failure.
[0008]
In this case, it is preferable to prohibit the return from the open loop control to the F / B control when the abnormality of the encoder is detected. That is, when the encoder becomes abnormal, the encoder count value becomes abnormal. Therefore, even if the F / B control for feeding back the encoder count value is performed, the motor cannot be driven normally. Therefore, in this case, by prohibiting the return to the F / B control, unnecessary F / B control is not performed, and the rotor can be rotated to the target position by the open loop control.
[0009]
Further, the encoder outputs the A-phase signal and the B-phase signal having a predetermined phase difference in synchronization with the rotation of the rotor, and outputs the Z-phase signal at the reference rotation position of the rotor. During the F / B control, the edges of the A-phase signal and the B-phase signal are counted, and the energized phase is sequentially switched based on the encoder count value, and the rotational position of the rotor and the energized phase are determined based on the Z-phase signal. Is determined, and if there is a shift, the shift is corrected (hereinafter, this correction is referred to as “Z-phase correction”), and further, the encoder count value and the drive signal count value counted during the open loop control. To determine whether there is an abnormality in the A-phase signal and the B-phase signal of the encoder. In this case, when updating the encoder count value by counting the A-phase signal and the B-phase signal, the rotation direction of the rotor is determined based on the generation order of the A-phase signal and the B-phase signal, and the encoder is determined in accordance with the rotation direction. The count value may be counted up or down, and the energized phase may be selected and energized according to the encoder count value and the rotation direction of the rotor. With this configuration, even if the rotor rotates in either the forward rotation or the reverse rotation, the correspondence between the encoder count value and the rotation position of the rotor can be maintained. , The correct energized phase can be selected based on the encoder count value. Further, since the reference rotation position of the rotor can be accurately detected by the Z-phase signal output from the encoder, for example, the energized phase (encoder count value) when the Z-phase signal is output corresponds to the reference rotation position of the rotor. By determining whether or not the current is in the energized phase (encoder count value), it is possible to confirm whether or not the correspondence between the rotational position of the rotor and the energized phase (encoder count value) has shifted. If there is a deviation, it is possible to perform highly reliable motor control by correcting the deviation.
[0010]
Further, when it is determined that the A-phase signal and the B-phase signal of the encoder are normal, the output interval of the Z-phase signal is determined based on the encoder count value to determine whether the Z-phase signal is abnormal. The determination may be made. If the A-phase signal and the B-phase signal of the encoder are determined to be normal, the encoder count value obtained by counting them accurately corresponds to the rotation angle (rotation amount) of the rotor. By determining the signal output interval (rotation angle), it can be accurately determined whether the Z-phase signal is output at the reference rotation position of the rotor.
[0011]
In this case, it is preferable to prohibit the Z-phase correction when the abnormality of the Z-phase signal is detected. By doing so, it is possible to avoid erroneous Z-phase correction when an abnormality occurs in the Z-phase signal.
[0012]
Further, a switch reluctance motor may be used as the motor. Switched reluctance motors have the advantage that they are inexpensive and have high durability and reliability against temperature environments and the like because they do not require permanent magnets and have a simple structure.
[0013]
The inventions according to claims 1 to 6 described above can be applied to various position switching devices using a brushless type motor such as a switch reluctance motor as a drive source. The present invention may be applied to a motor control device that drives a range switching mechanism that switches the range of the machine. Thus, a highly reliable motor-driven range switching device can be configured.
[0014]
BEST MODE FOR CARRYING OUT THE INVENTION
An embodiment in which the present invention is applied to a range switching device for a vehicle will be described below with reference to the drawings.
[0015]
First, the configuration of the range switching mechanism 11 will be described with reference to FIG. The motor 12 serving as a drive source of the range switching mechanism 11 is constituted by, for example, a switch reluctance motor, has a built-in speed reduction mechanism 26 (see FIG. 4), and is provided with an output shaft sensor 14 for detecting the rotational position of the output shaft 13. Have been. A detent lever 15 is fixed to the output shaft 13.
[0016]
An L-shaped parking rod 18 is fixed to the detent lever 15, and a conical body 19 provided at the tip of the parking rod 18 is in contact with the lock lever 21. The lock lever 21 moves up and down about the shaft 22 in accordance with the position of the cone 19 to lock / unlock the parking gear 20. The parking gear 20 is provided on the output shaft of the automatic transmission 27. When the parking gear 20 is locked by the lock lever 21, the driving wheels of the vehicle are held in a state where they are prevented from rotating (parking state).
[0017]
On the other hand, a detent spring 23 for holding the detent lever 15 in a parking range (hereinafter referred to as “P range”) and another range (hereinafter referred to as “NotP range”) is fixed to the support base 17. When the engaging portion 23a provided at the tip of the spring 23 is fitted into the P range holding concave portion 24 of the detent lever 15, the detent lever 15 is held at the position of the P range. The detent lever 15 is held at the position of the NotP range when 23a is fitted into the NotP range holding recess 25 of the detent lever 15.
[0018]
In the P range, the parking rod 18 moves in a direction approaching the lock lever 21, and the thick portion of the cone 19 pushes up the lock lever 21, and the projection 21 a of the lock lever 21 fits into the parking gear 20 to park. The gear 20 is locked, whereby the output shaft (drive wheel) of the automatic transmission 27 is maintained in a locked state (parking state).
[0019]
On the other hand, in the NotP range, the parking rod 18 moves in a direction away from the lock lever 21, the thick portion of the cone 19 comes out of the lock lever 21, and the lock lever 21 descends. 21a is disengaged from the parking gear 20, the lock of the parking gear 20 is released, and the output shaft of the automatic transmission 27 is maintained in a rotatable state (runnable state).
[0020]
The output shaft sensor 14 described above is constituted by a rotation sensor (for example, a potentiometer) that outputs a voltage corresponding to the rotation angle of the output shaft 13 of the speed reduction mechanism 26 of the motor 12, and the current range is set to the P range by the output voltage. And the NotP range.
[0021]
Next, the configuration of the motor 12 will be described with reference to FIG. In the present embodiment, a switched reluctance motor (hereinafter referred to as “SR motor”) is used as the motor 12. The SR motor 12 is a motor in which both the stator 31 and the rotor 32 have a salient pole structure, and has an advantage that a permanent magnet is not required and the structure is simple. For example, twelve salient poles 31a are formed at equal intervals on the inner peripheral portion of the cylindrical stator 31, whereas, for example, eight salient poles 32a are formed at equal intervals on the outer peripheral portion of the rotor 32. The respective salient poles 32a of the rotor 32 are sequentially opposed to the respective salient poles 31a of the stator 31 via a minute gap with the rotation of the rotor 32. The twelve salient poles 31a of the stator 31 have a total of six windings 33 of U-phase, V-phase, and W-phase, and a total of six windings 34 of U'-, V'-, and W'-phases. It is wound in order. It goes without saying that the number of salient poles 31a and 32a of the stator 31 and the rotor 32 may be changed as appropriate.
[0022]
The winding order of the windings 33 and 34 in the present embodiment is, for example, V phase → W phase → U phase → V phase → W phase → U phase → V ′ for the twelve salient poles 31a of the stator 31. The phases are wound in the order of phase → W ′ phase → U ′ phase → V ′ phase → W ′ phase → U ′ phase. As shown in FIG. 3, a total of six windings 33 of U-phase, V-phase and W-phase and a total of six windings 34 of U′-phase, V′-phase and W′-phase are two-system motors. They are connected so as to form the excitation units 35 and 36. One motor excitation unit 35 is configured by connecting a total of six windings 33 of U-phase, V-phase, and W-phase in a Y-connection (two windings 33 of the same phase are connected in series). The other motor excitation unit 36 is configured by connecting a total of six windings 34 of U ′ phase, V ′ phase, and W ′ phase in a Y connection (two windings 34 of the same phase are connected in series, respectively). It is connected to the). In the two motor excitation units 35 and 36, the U phase and the U 'phase are simultaneously energized, the V phase and the V' phase are simultaneously energized, and the W phase and the W 'phase are simultaneously energized.
[0023]
These two motor excitation units 35 and 36 are driven by separate motor drivers 37 and 38 (motor drive circuits) using a battery 40 mounted on the vehicle as a power supply. In this way, by providing the motor excitation units 35 and 36 and the motor drivers 37 and 38 each in two systems, even if one system fails, the SR motor 12 can be rotated in the other system. ing. The circuit configuration example of the motor drivers 37 and 38 shown in FIG. 3 has a unipolar drive circuit configuration in which one switching element 39 such as a transistor is provided for each phase, but two switching elements are provided for each phase. Alternatively, a circuit configuration of a bipolar drive system provided for each may be adopted. It is needless to say that the present invention may be configured such that one system is provided for each of the motor excitation unit and the motor driver.
[0024]
ON / OFF of each switching element 39 of each motor driver 37, 38 is controlled by the ECU 41 (control means). As shown in FIG. 4, the ECU 41 and the motor drivers 37 and 38 are mounted on a range switching control device 42. The range switching control device 42 includes a P range switch 43 for performing a switching operation to a P range, An operation signal of a NotP range switch 44 for performing a switching operation to the NotP range is input. The range selected by operating the P range switch 43 or the NotP range switch 44 is displayed on a range display section 45 provided on an instrument panel (not shown). The P range switch 43 and the NotP range switch 44 are constituted by, for example, momentary switches.
[0025]
The SR motor 12 is provided with an encoder 46 for detecting the rotational position of the rotor 32. The encoder 46 is constituted by, for example, a magnetic rotary encoder. As shown in FIGS. 5 and 6, the specific configuration of the encoder 46 is such that north poles and south poles are alternately arranged at equal pitches in the circumferential direction. A magnetized annular rotary magnet 47 is coaxially fixed to the side surface of the rotor 32, and three magnetic detecting elements 48, 49, 50 such as Hall ICs are arranged at positions facing the rotary magnet 47. It has a configuration. In the present embodiment, the magnetization pitch between the N pole and the S pole of the rotary magnet 47 is set to 7.5 °. The magnetization pitch (7.5 °) of the rotary magnet 47 is set to be the same as the rotation angle of the rotor 32 per excitation of the SR motor 12. As will be described later, when the energized phase of the SR motor 12 is switched six times by the 1-2-phase excitation method, all energized phases are switched once, and the rotor 32 and the rotary magnet 47 are integrally 7.5 °. × 6 = 45 ° rotation. The number of N poles and S poles present in the rotation angle range of 45 ° of the rotary magnet 47 is a total of six poles.
[0026]
Further, the N pole (N ') at the position corresponding to the reference rotation position of the rotor 32 and the S poles (S') on both sides thereof are formed so as to be wider in the radial direction than the other magnetic poles. In the present embodiment, considering that the rotor 32 and the rotary magnet 47 rotate integrally by 45 ° during one cycle of the switching of the energized phase of the SR motor 12, a wide width corresponding to the reference rotation position of the rotor 32 is taken into consideration. The large magnetized portions (N ′) are formed at a 45 ° pitch, so that a total of eight wide magnetized portions (N ′) corresponding to the reference rotation position are formed as the rotary magnet 47 as a whole. The wide magnetized portion (N ′) corresponding to the reference rotation position may be configured such that only one magnet is formed as the whole rotary magnet 47.
[0027]
Three magnetic detecting elements 48, 49, 50 are arranged with respect to the rotary magnet 47 in the following positional relationship. The magnetic detecting element 48 that outputs the A-phase signal and the magnetic detecting element 49 that outputs the B-phase signal are composed of a narrow magnetized portion (N, S) and a wide magnetized portion (N ′, S ′) of the rotary magnet 47. ) Are arranged on the same circumference at positions that can face both. On the other hand, the magnetic detection element 50 that outputs the Z-phase signal is located at a position radially outside or inside the narrow magnetized portion (N, S) of the rotary magnet 47 and has a wide magnetized portion (N ′). , S ′) only. As shown in FIG. 7, the interval between the two magnetic detecting elements 48 and 49 for outputting the A-phase signal and the B-phase signal is such that the phase difference between the A-phase signal and the B-phase signal is 90 ° in electrical angle (mechanical angle). 3.75 °). Here, “electric angle” is an angle when the generation period of the A / B phase signal is one cycle (360 °), and “mechanical angle” is a mechanical angle (one rotation of the rotor 32 is 360 °). The angle at which the rotor 32 rotates from the fall (rise) of the A-phase signal to the fall (rise) of the B-phase signal corresponds to the mechanical angle of the phase difference between the A-phase signal and the B-phase signal. I do. The magnetic detection element 50 that outputs the Z-phase signal is arranged so that the phase difference between the Z-phase signal and the B-phase signal (or the A-phase signal) becomes zero.
[0028]
The output of each of the magnetic detection elements 48, 49, and 50 becomes a high level "1" when facing the N pole (N 'pole), and becomes a low level "0" when facing the S pole (S' pole). Become. The output of the magnetic detection element 50 for the Z-phase signal becomes high level "1" every time it faces the wide N 'pole corresponding to the reference rotation position of the rotor 32. At other positions, the output becomes low level "1". 0 ".
[0029]
In the present embodiment, the ECU 41 counts both rising and falling edges of the A-phase signal and the B-phase signal by an encoder counter routine described later, and switches the energized phase of the SR motor 12 according to the encoder count value. Drives the rotor 32 to rotate. At this time, the rotation direction of the rotor 32 is determined based on the generation order of the A-phase signal and the B-phase signal, and the encoder count value is counted up in the forward rotation (P direction → NotP range rotation direction), and the reverse rotation (NotP range → In the (P-range rotation direction), the encoder count value is counted down. Accordingly, even if the rotor 32 rotates in either the forward direction or the reverse direction, the correspondence between the encoder count value and the rotational position of the rotor 32 is maintained. However, the rotation position (rotation angle) of the rotor 32 is detected based on the encoder count value, and the windings 33 and 34 of the phase corresponding to the rotation position are energized to drive the rotor 32 to rotate.
[0030]
FIG. 7 shows an output waveform of the encoder 46 and a switching pattern of the energized phase when the rotor 32 is rotated in the reverse rotation direction (the rotation direction from the NotP range to the P range). In both the reverse rotation direction (the rotation direction from the NotP range to the P range) and the normal rotation direction (the rotation direction from the P range to the NotP range), one phase energization and two phases are performed every time the rotor 32 rotates 7.5 °. The energization and the energization are alternately switched, and during the rotation of the rotor 32 by 45 °, for example, in the order of U-phase energization → UW-phase energization → W-phase energization → VW-phase energization → V-phase energization → UV-phase energization The switching of the energized phase is performed once. Each time the energized phase is switched, the rotor 32 rotates by 7.5 ° so that the magnetic pole of the rotary magnet 47 facing the magnetic detection elements 48 and 49 for the A-phase and B-phase signals changes from N pole to S pole. (N ′ pole → S ′ pole) or S pole → N pole (S ′ pole → N ′ pole), and the levels of the A-phase signal and the B-phase signal are alternately inverted. The encoder count value is incremented (or decremented) by 2 for every .5 ° rotation. Each time the switching of the energized phase completes one cycle and the rotor 32 rotates 45 °, the magnetic detection element 50 for the Z phase faces the wide N ′ pole corresponding to the reference rotation position of the rotor 32, and The signal becomes high level "1". In this specification, the fact that the A-phase, B-phase, and Z-phase signals are at the high level “1” may be referred to as the output of the A-phase, B-phase, and Z-phase signals.
[0031]
By the way, since the encoder count value is stored in the RAM of the ECU 41, when the power of the ECU 41 is turned off, the storage of the encoder count value disappears. Therefore, the encoder count value (0) immediately after the power supply of the ECU 41 is turned on does not correspond to the actual rotational position (energized phase) of the rotor 32. Therefore, in order to switch the energized phase according to the encoder count value, it is necessary to make the encoder count value correspond to the actual rotational position of the rotor 32 after power-on, and to make the encoder count value correspond to the energized phase.
[0032]
Therefore, in the present embodiment, the initial drive routine shown in FIGS. 8 and 9 to be described later is executed by the ECU 41 of the range switching control device 42, so that the initial phase of the SR motor 12 is turned on when the ECU 41 is powered on. Is cycled in a predetermined time schedule to count the edges of the A-phase signal and the B-phase signal of the encoder 46, and learn the correspondence between the encoder count value at the end of the initial drive, the rotational position of the rotor 32, and the energized phase. Then, during the subsequent normal driving, the energized phase is determined based on the encoder count value and the learning result at the end of the initial driving.
[0033]
The learning at the time of the initial drive is specifically performed as follows. As shown in FIG. 12, when the initial drive is performed when the power is supplied to the ECU 41 in the P range, for example, W-phase energization → UW-phase energization → U-phase energization → UV-phase energization → V-phase energization → VW-phase The energization phase is switched in a predetermined time schedule in the energization order, and the rotor 32 is driven in the normal rotation direction (P range → Not P range rotation direction).
[0034]
On the other hand, when the initial drive is performed when the power is supplied to the ECU 41 in the NotP range, for example, energization is performed in the order of V-phase energization → UV-phase energization → U-phase energization → UW-phase energization → W-phase energization → VW-phase energization. The phase switching is performed once in a predetermined time schedule, and the rotor 32 is driven in the reverse rotation direction (the rotation direction from the NotP range to the P range).
[0035]
During this initial drive, the one-phase energization time T1 is set shorter than the two-phase energization time T2, for example, T1 = 10 ms and T2 = 100 ms. Even after the rotational position of the rotor 32 and the energized phase are synchronized during the initial drive, the rotor 32 vibrates in the one-phase energization with a small torque, so the time T1 of the one-phase energization is shortened, and as soon as possible. By switching to the next two-phase energization, the vibration of the rotor 32 is stopped immediately and the output signal of the encoder 46 is stabilized.
[0036]
As described above, if the switching of the energized phase is performed once during the initial drive, the rotational position of the rotor 32 and the energized phase always match in any energized phase before the end of the initial drive. The rotor 32 rotates in synchronization with the phase switching, and the A-phase signal and the B-phase signal are output from the encoder 46 in synchronization with the rotation of the rotor 32.
[0037]
During this initial drive, both rising and falling edges of the A-phase signal and the B-phase signal of the encoder 46 are counted. Therefore, by looking at the encoder count value at the end of the initial drive, the angle (rotation amount) at which the rotor 32 actually rotates in synchronization with the switching of the energized phase before the end of the initial drive can be determined. The relationship between the encoder count value at the time, the rotational position of the rotor 32, and the energized phase can be understood.
[0038]
In the example of FIG. 12, the rotor 32 rotates from the first energized phase (W phase) during the initial drive, and the rotor 32 rotates by 7.5 ° every time the energized phase is switched, and the encoder count value is incremented by two. At the end of the initial drive, the encoder count value becomes 12.
[0039]
On the other hand, for example, if the rotor 32 does not rotate in the first three excitations (W-phase energization → UW-phase energization → U-phase energization), that is, the fourth and subsequent excitations (UV-phase energization → V-phase energization → VW-phase) In the case where the rotation position of the rotor 32 and the energized phase are synchronized with each other in the energized state and the rotor 32 rotates by three times of excitation, the rotor 32 is 7.5 ° × 3 = 22.5 ° by the end of the initial drive. It rotates and the encoder count value becomes 2 × 3 = 6. Therefore, by looking at the encoder count value at the end of the initial drive, the angle (rotation amount) at which the rotor 32 actually rotates in synchronization with the switching of the energized phase before the end of the initial drive is determined.
[0040]
Although the last energized phase of the initial drive is always the VW phase, the encoder count value is not always 12 and may be, for example, 8 or 4. At the time of normal drive after the end of the initial drive, the energized phase is determined based on the encoder count value. Therefore, unless the deviation of the encoder count value due to the initial drive is corrected, the correct energized phase cannot be selected at the time of normal drive.
[0041]
Therefore, in the present embodiment, the encoder count value at the end of the initial drive is learned as the initial position deviation learning value, and the encoder count value at the time of the subsequent normal drive is corrected with the initial position deviation learning value. The deviation between the encoder count value and the energized phase (the rotational position of the rotor 32) is corrected so that the correct energized phase can be selected during normal driving.
[0042]
After the end of the initial drive, as shown in FIG. 12, the F / B control start position stop holding process described later energizes the same phase as the energized phase (VW phase) at the end of the initial drive, for example, for 10 ms, and the position of the rotor 32 becomes Is held at the position at the end of the initial driving, and thereafter, the energized phase is determined by feedback control (hereinafter referred to as “F / B control”) described later based on the encoder count value at that time and the initial position deviation learning value. By switching, the rotor 32 is rotated in the direction of the target position Acnt. Thus, when the rotational position (encoder count value) of the rotor 32 reaches, for example, within 0.5 ° from the target position Acnt, the switching of the energized phase is completed and the rotor 32 is stopped. In the position stop holding process, the same phase is continuously energized to hold the stopped state of the rotor 32, and this held state is continued for, for example, 50 ms. Thereafter, if the target position Acnt does not change, the energization is stopped.
[0043]
When the P range switch 43 or the NotP range switch 44 is operated to generate a range switching request during the initial drive, the next two-phase energization (when the range switching request occurs during the execution of the two-phase energization) is performed. After the completion of the two-phase energization, the operation shifts to the normal drive, and the correspondence between the encoder count value at the end of the two-phase energization, the rotational position of the rotor 32, and the energized phase is learned. The energized phase is determined based on the encoder count value and the learning result at the end of the two-phase energization.
[0044]
In the two-phase energization, since the torque is large, even if the position of the rotor 32 is slightly shifted from the position corresponding to the two-phase energization, the rotor 32 can be rotated to the position corresponding to the two-phase energization. Therefore, it is considered that there is a high probability that the rotation position of the rotor 32 and the energized phase can be synchronized only by performing the two-phase energization once or twice during the initial driving. Therefore, when a range switching request is generated during the initial drive, the normal two-phase energization (or the current two-phase energization) is terminated before the normal drive is started. After learning the correspondence between the count value, the rotational position of the rotor 32, and the energized phase, it is possible to quickly shift to the normal drive.
[0045]
For example, as shown in the time chart of FIG. 13, when a range switching request from the P range to the NotP range occurs during the second excitation (UW phase energization) of the initial drive, the initial drive is performed by the UW phase energization. At the same time, the process is shifted to the normal drive, and the correspondence between the encoder count value at the end of the UW phase energization, the rotational position of the rotor 32, and the energized phase is learned. In the present embodiment, even when shifting to the normal drive in the middle of the initial drive, the encoder count value at the end of the original initial drive is estimated assuming that the initial drive has been executed to the end, and the estimated value is initialized. The position deviation learning value is set. For example, as shown in FIG. 13, when the initial drive is ended by two excitations of the W-phase energization and the UW-phase energization, the initial drive is ended without performing the four scheduled excitations thereafter. Assuming that four unexcited excitations (U-phase energization → UV-phase energization → V-phase energization → VW-phase energization) have been performed, the count-up value (2) corresponding to the rotation angle for the four excitations × 4 = 8) is added to the encoder count value at the end of the UW-phase energization to obtain an initial position deviation learning value.
[0046]
By the way, during the F / B control, for example, the A-phase signal or the B-phase signal of the encoder 46 is temporarily missing for some reason, or a noise pulse superimposed on the signal line of the encoder 46 is erroneously recognized as a normal pulse signal. When the step-out or step-out occurs, the energized phase (encoder count value) and the rotation phase of the rotor 32 become out of synchronization, and the rotor 32 cannot be driven normally, whereby the rotor 32 stops. The F / B control may not be able to continue normally and the control may not be possible.
[0047]
Therefore, in the present embodiment, during the F / B control, whether the F / B control state (rotation state of the rotor 32) is normal or abnormal is monitored, and if an abnormality is detected, the F / B control is performed at that time. Is switched to open loop control and the recovery process is executed. In the execution example of the recovery processing shown in FIG. 24, during the F / B control for switching the shift range from the P range to the NotP range, the F / B control state becomes abnormal, and the rotor 32 abnormally stops or the encoder count value When an abnormal stop of the count operation of Ncnt occurs, the F / B control is switched to the open loop control to execute the recovery processing.
[0048]
During the open loop control (recovery processing), the rotor 32 is driven to rotate by sequentially switching the energized phases without feeding back the information of the encoder count value Ncnt, and the open loop control is performed each time the energized phase is switched during the open loop control. The position count value Ncntop at the time is counted up (or counted down), and when the position count value Ncntop reaches the target count value Actnt, it is determined that the rotor 32 has reached the target position, and the open loop control (recovery processing) is performed. ), And returns to the F / B control. As a result, in the present embodiment, even if an abnormality of the encoder count value Ncnt or step-out occurs during the F / B control, the F / B control state (rotation state of the rotor 32) temporarily becomes abnormal. The rotor 32 can be rotated to the target position by the open loop control.
[0049]
At this time, the abnormality detection method of the F / B control state may be any one or two or all of the following three F / B control abnormality detection methods (1) to (3). good.
[0050]
[F / B control abnormality detection method (1)]
As shown in FIG. 25, during the F / B control, the edge of the A-phase signal and the edge of the B-phase signal are separately counted, and the deviation between the count value Na of the A-phase signal and the count value Nb of the B-phase signal is calculated. The presence / absence of an abnormality in the F / B control state is monitored based on whether | Na−Nb | is equal to or greater than an abnormality determination value (for example, 3). If the A-phase signal and the B-phase signal are normally output during the F / B control, the difference between the count value Na of the A-phase signal and the count value Nb of the B-phase signal becomes 0 or 1, so that both signals are output. If the deviation | Na−Nb | of the count value of the second signal is 2 or more, it is possible that one of the signals is missing or a noise pulse is mixed. In such a state, the encoder count value Ncnt and the rotational position of the rotor 32 cannot be synchronized, and the rotational state (F / B control state) of the rotor 32 becomes abnormal. | Na-Nb | can be used to determine whether there is an abnormality in the rotation state (F / B control state) of the rotor 32.
[0051]
[F / B control abnormality detection method (2)]
As shown in FIG. 25, during the F / B control, the edge interval ΔTa of the A-phase signal and the edge interval ΔTb of the B-phase signal are measured, and the edge intervals ΔTa and ΔTb of each signal are determined to be abnormal determination values (for example, ΔTa, The presence or absence of an abnormality in the F / B control state is monitored based on whether the average is equal to or more than 10 times the average value of ΔTb).
[0052]
As the abnormality detection method (2), the abnormality may be detected by monitoring the rotor rotation speed. For example, the abnormality may be detected based on the number of output pulses of the encoder 46 per unit time. good.
[0053]
[F / B control abnormality detection method (3)]
As shown in FIG. 26, during the F / B system, the rotation direction of the rotor 32 is monitored from the change direction of the encoder count value Ncnt, and whether the rotation direction of the rotor 32 is reversed from the rotation direction toward the target position is determined. Monitor whether there is an abnormality in the F / B control state (rotation state of the rotor 32). In the present embodiment, the number of times that the encoder count value Ncnt is continuously updated in the direction opposite to the target position is counted, and F / B is determined based on whether or not the count value Cr is equal to or greater than an abnormality determination value (for example, 4). The presence or absence of an abnormality in the control state (rotation state of the rotor 32) is monitored. That is, when the encoder count value Ncnt is continuously updated in the direction opposite to the target position, for example, four times or more, it is determined that the rotor 32 is rotating in the reverse direction, and the reverse rotation detection flag Xr is turned ON.
[0054]
Further, in the present embodiment, the excitation method during the open loop control is set to the 1-2 phase excitation method, the first excitation at the start of the open loop control is performed by two-phase conduction, and as shown in FIG. For example, for three steps from the start of the loop control, the excitation time of each energized phase is set to the longest time (for example, 50 ms), and the rotor 32 is reliably gripped and the rotational phase of the rotor 32 and the energized phase are reliably synchronized. After that, the excitation time is greatly reduced, but the excitation time is gradually shortened as the rotation speed of the rotor 32 increases, so that the rotation speed of the rotor 32 is gradually increased.
[0055]
In order to estimate the rotational position of the rotor 32 during the open loop control, the initial position of the rotor 32 at the start of the open loop control (the initial value of the position count value Ncntop) is calculated based on the output signal of the output shaft sensor 14. Thereafter, every time a drive signal is output from the ECU 41 to the motor drivers 37 and 38 (motor drive circuits) to switch the energized phase, the position count value Ncntop is counted up (or down) and based on the position count value Ncntop. When the difference between the position count value Ncntop and the target count value Acnt (target position) becomes equal to or less than a predetermined value, it is determined that the vehicle has entered the deceleration region, and the excitation time is gradually increased. And the rotational speed of the rotor 32 is gradually reduced, and the position count value Ncntop is Upon reaching the count value Acnt, the rotor 32 is determined to have reached the target position, and ends the open-loop control, returns to the F / B control.
[0056]
By the way, even if the F / B control system fails and the SR motor 12 cannot be normally driven by the F / B control, there is a possibility that the SR motor 12 can be normally driven by switching to the open loop control. For example, when the encoder 46 fails, one of the two motor excitation units 35 and 36 (windings 33 and 34) of the SR motor 12 fails, or the two motor driver Even if one of the systems 37 and 38 fails and the driving force of the SR motor 12 is reduced, there is a possibility that the SR motor 12 can be driven normally by switching to the open loop control.
[0057]
Therefore, in the present embodiment, when a failure (system abnormality) of the F / B control system is detected by a system abnormality diagnosis routine of FIGS. 33 and 34 described later, a fail-safe processing routine of FIG. The SR motor 12 is driven by switching to open loop control during processing. The open loop control during the fail-safe processing is the same as the above-described open loop control during the recovery processing.
[0058]
In the present embodiment, the encoder count value Ncnt (the count value of the A-phase signal and the B-phase signal of the encoder 46) and the position count value Ncntop (the motor drive signal) during the open-loop control (during the fail-safe process and the recovery process). And the presence / absence of A / B phase disconnection of the encoder 46 is determined. Specifically, whether the difference between the change amount of the encoder count value Ncnt and the change amount of the position count value Ncntop from the start to the end of the open loop control is equal to or larger than a determination value is determined by determining whether the A phase of the encoder 46 It is determined whether there is a disconnection in phase B. For example, when both the A phase and the B phase of the encoder 46 are disconnected, the encoder count value Ncnt does not change at all even if the rotor 32 rotates during the open loop control. Further, when one of the A phase and the B phase of the encoder 46 is disconnected, the amount of change in the encoder count value Ncnt is の of the normal amount. If the SR motor 12 is normally driven during the open loop control, the change amount of the position count value Ncntop during the open loop control is a value corresponding to the rotation amount (rotation angle) of the rotor 32 during the open loop control. It becomes. Therefore, by comparing the encoder count value Ncnt and the position count value Ncntop during the open loop control, it is possible to determine whether or not the A-phase / B-phase disconnection of the encoder 46 has occurred.
[0059]
Further, in the present embodiment, in order to determine whether or not the Z-phase signal of the encoder 46 is abnormal during the open loop control, the output interval of the Z-phase signal is changed to the next Z-phase signal after the output of the previous Z-phase signal. The measurement is performed based on a change amount (rotation angle of the rotor 32) of the encoder count value Ncnt until the signal is output, and the presence or absence of abnormality of the Z-phase signal is determined based on the output interval of the Z-phase signal. That is, the amount of change (the rotation angle of the rotor 32) of the encoder count value Ncnt from the output of the previous Z-phase signal to the output of the next Z-phase signal is determined by the interval between the reference rotational positions of the rotor 32 (this embodiment). If the value is equivalent to 45 °), it is determined that the Z-phase signal is normal, and if the state in which the Z-phase signal is not output even several times beyond the reference rotational position of the rotor 32 continues several times, the encoder When it is determined that the Z-phase is disconnected (abnormal) at 46, and when the Z-phase signal is output at an interval shorter than the interval of the reference rotation position of the rotor 32, erroneous detection of the Z-phase signal due to noise or the like (abnormal) Is determined.
[0060]
With the above-described method, it is determined whether there is an abnormality in the A-phase, B-phase, and Z-phase of the encoder 46 during the open-loop control, and if the disconnection (abnormality) of the A-phase and the B-phase is detected, the open-loop control is performed. From the control to the F / B control is prohibited. That is, if the A-phase and B-phase signals of the encoder 46 become abnormal, the encoder count value Ncnt becomes abnormal. Therefore, even if the F / B control for feeding back the encoder count value Ncnt is performed, the SR motor 12 is controlled. Cannot drive normally. Therefore, in this case, by prohibiting the return to the F / B control, unnecessary F / B control is not performed, and the rotor 32 can be rotated to the target position by the open loop control.
[0061]
Further, when an abnormality of the Z-phase signal of the encoder 46 is detected during the open loop control, the Z-phase correction is prohibited. By doing so, it is possible to avoid erroneous Z-phase correction when an abnormality occurs in the Z-phase signal.
[0062]
The range switching control of the present embodiment described above is executed by the ECU 41 of the range switching control device 42 in accordance with each routine described later. Hereinafter, the processing contents of each of these routines will be described.
[0063]
[Initial drive]
This is executed according to the initial drive routine shown in FIGS. This routine is executed at a predetermined cycle (for example, 1 ms cycle) immediately after the power supply to the ECU 41 is turned on (immediately after the ignition switch is operated from the OFF position to the ACC position) until the end of the initial drive.
[0064]
When the routine is started, first, in step 101, it is determined whether the fail-safe processing execution flag Xfsop = ON or the recovery processing execution flag Xrcv = ON. Here, the fail-safe process execution flag Xfsop is a flag for determining whether or not a fail-safe process (range switching by open loop control) executed when the encoder 46 or the SR motor 12 has failed is executed. ON means being executed. The recovery process execution flag Xrcv is a flag for determining whether or not a recovery process (open loop control) that is temporarily executed when a temporary F / B control state abnormality occurs has been executed. , ON means being executed.
[0065]
If “Yes” is determined in step 101, the present routine is terminated without performing the subsequent processing, and if “No” is determined, the process proceeds to step 102 and the output from the initial processing of the ECU 41 is performed. It is determined whether a predetermined time (for example, 100 ms) for waiting for the output voltage of the axis sensor 14 to stabilize has elapsed. Then, when a predetermined time has elapsed from the initial processing, the process proceeds to step 103, where the output voltage of the output shaft sensor 14 is read, and whether or not the current range is the P range is determined by whether or not this output voltage is equal to or less than the range determination value. It is determined whether the range is the NotP range. If the range is the P range, the process proceeds to step 104, where the range determination flag Xnp is set to “0” meaning the P range. If the range is the NotP range, the process proceeds to step 105 and the range determination flag Xnp Is set to “1” meaning NotP range.
[0066]
Thereafter, the process proceeds to step 106, where it is determined whether or not the range determination flag Xnp = 0 (P range). If the range determination flag Xnp = 0 (P range), the process proceeds to step 107, where P The range initial drive routine is executed, and if the range determination flag Xnp = 1 (NotP range), the routine proceeds to step 108, where the NotP range initial drive routine of FIG. 11 is executed.
[0067]
When the P range initial drive routine of FIG. 10 is started in step 107, in steps 201 to 206, it is determined whether the excitation number counter CASE for counting the number of excitations during the initial drive is 0 to 5. . The excitation number counter CASE has an initial value of 0 set in the initial processing, and is incremented by one each time excitation is performed once (step 114 in FIG. 9). Then, the energizing phase and the energizing time T are set as follows according to the determination result of the excitation number counter CASE.
[0068]
If CASE = 0 (first excitation), the process proceeds to step 207, where W-phase energization is selected, and the energization time T is set to T1 (for example, 10 ms).
If CASE = 1 (second excitation), the process proceeds to step 208, where UW-phase energization is selected, and the energization time T is set to T2 (for example, 100 ms).
[0069]
If CASE = 2 (third excitation), the process proceeds to step 209 to select U-phase energization and set the energization time T to T1 (for example, 10 ms).
If CASE = 3 (fourth excitation), the process proceeds to step 210, where UV phase energization is selected, and the energization time T is set to T2 (for example, 100 ms).
[0070]
If CASE = 4 (fifth excitation), the process proceeds to step 211, where V-phase energization is selected, and the energization time T is set to T1 (for example, 10 ms).
If CASE = 5 (sixth excitation), the process proceeds to step 212, where VW-phase energization is selected, and the energization time T is set to T2 (for example, 100 ms).
[0071]
Thus, when the initial drive is performed in the P range, the energized phases are switched in the order of W-phase energization → UW-phase energization → U-phase energization → UV-phase energization → V-phase energization → VW-phase energization, and the rotor 32 is rotated. Driving is performed in the forward rotation direction (P range → NotP range rotation direction). At this time, the one-phase energization time T1 is set shorter than the two-phase energization time T2.
[0072]
On the other hand, when the NotP range initial drive routine of FIG. 11 is started in step 108, it is determined in steps 221 to 226 whether the excitation number counter CASE is 0 to 5 and the energization is performed according to the determination result. The phase and the energizing time T are set as follows.
[0073]
If CASE = 0 (first excitation), the process proceeds to step 227, where V-phase energization is selected, and the energization time T is set to T1 (for example, 10 ms).
If CASE = 1 (second excitation), the process proceeds to step 228, where UV phase energization is selected, and the energization time T is set to T2 (for example, 100 ms).
[0074]
If CASE = 2 (third excitation), the process proceeds to step 229, where U-phase energization is selected, and the energization time T is set to T1 (for example, 10 ms).
If CASE = 3 (fourth excitation), the process proceeds to step 230, where UW-phase energization is selected, and the energization time T is set to T2 (for example, 100 ms).
[0075]
If CASE = 4 (fifth excitation), the process proceeds to step 231, where W-phase energization is selected, and the energization time T is set to T1 (for example, 10 ms).
If CASE = 5 (sixth excitation), the process proceeds to step 232, where VW-phase energization is selected, and the energization time T is set to T2 (for example, 100 ms).
[0076]
Thus, when the initial drive is performed in the NotP range, the switching of the energized phases is performed in the order of V phase energization → UV phase energization → U phase energization → UW phase energization → W phase energization → VW phase energization, and the rotor 32 Drive is performed in the reverse rotation direction (the rotation direction from the NotP range to the P range). Also in this case, the one-phase energization time T1 is set shorter than the two-phase energization time T2.
[0077]
After executing the P range initial drive routine of FIG. 10 or the NotP range initial drive routine of FIG. 11 as described above, the process proceeds to step 109 of FIG. 8, and the range switching operation (P range switch 43 or NotP It is determined whether or not the range switch 44 has been operated). If the range switch operation has been performed during the initial drive, the process proceeds to step 110, where the range switch operation flag Xchg is set to ON, and the range switch operation is performed. If not, the routine proceeds to step 111, where the range switching operation flag Xchg is set to OFF.
[0078]
Thereafter, the process proceeds to step 112 in FIG. 9, in which the energizing time counter CT for counting the energizing time of the current energizing phase is counted up. It is determined whether or not the energization time T (= T1 or T2) set in the routine 11 has been exceeded, and if not, this routine is terminated without performing the subsequent processing. Thereby, the energization to the current energized phase is continued until the energized time CT of the current energized phase exceeds the energized time T (= T1 or T2) set in the routine of FIG. 10 or FIG.
[0079]
Thereafter, when the energization time CT of the current energization phase exceeds the energization time T (= T1 or T2) set in the routine of FIG. 10 or FIG. 11, the routine proceeds to step 114, where the excitation number counter CASE is counted up by one. Then, the energized phase is switched to the next energized phase. Then, in the next step 115, after resetting the energization time counter CT, the routine proceeds to step 116, where it is determined whether or not the excitation number counter CASE has reached "6" indicating the end of the initial drive. If the number counter CASE has reached "6", the routine proceeds to step 118, where the initial drive end flag Xend is set to "ON" indicating the end of the initial drive.
[0080]
If the number-of-excitations counter CASE has not reached "6", that is, if the initial drive is being performed, the process proceeds to step 117, and it is determined whether the condition for ending the initial drive is satisfied. The condition for terminating the initial drive halfway is determined by the following three conditions (1) to (3).
(1) The range determination flag Xnp is 0 (P range)
{Circle around (2)} The excitation number counter CASE is 2 or 4, that is, the end of two-phase energization.
(3) The range switching operation flag Xchg is ON, that is, the range switching operation has been performed during the initial drive.
[0081]
If at least one of these three conditions (1) to (3) is not satisfied, the condition for ending the initial drive is not satisfied, and the initial drive is continued. On the other hand, if all of the three conditions (1) to (3) are satisfied, the condition for terminating the middle of the initial drive is satisfied, the process proceeds to step 118, and the initial drive end flag Xend indicates the end of the initial drive. Set to “ON”.
[0082]
Thereafter, the process proceeds to step 119, where it is determined whether or not the range determination flag Xnp = 1 (whether or not the initial drive was performed in the NotP range). If the range determination flag Xnp = 1, the process proceeds to step 120, The encoder count value Ncnt at the end of the initial drive is stored as an initial position deviation learning value Gcnt. Then, in the next step 121, the encoder count value Ncnt is corrected to a value based on the NotP range by the following equation.
Ncnt = Ncnt + 288
[0083]
In the present embodiment, the encoder count value Ncnt is counted up by setting the holding position of the P range to the zero point position, and when the rotor 32 rotates to the holding position of the NotP range, the encoder count value Ncnt becomes, for example, 288. ing. Therefore, when the initial drive is performed in the NotP range, the encoder count value Ncnt is corrected to a value based on the NotP range by adding 288 to the encoder count value Ncnt at the end of the initial drive.
[0084]
On the other hand, if it is determined in step 119 that the range determination flag Xnp = 0 (initial driving in the P range), the process proceeds to step 122, where the initial position deviation learning value Gcnt is calculated using the encoder count value Ncnt at the end of the initial driving. It is calculated by the following equation.
Gcnt = Ncnt + 2 × (6-CASE)
[0085]
In this case, if the initial drive is performed to the end, CASE = 6 by the processing of step 114, so that the encoder count value Ncnt at the end of the initial drive becomes the initial position deviation learning value Gcnt as it is, but the initial drive in the P range is performed. Is performed, when the range switching operation is performed during the initial driving, the next two-phase energization (or when a range switching request occurs during the execution of the two-phase energization, the two-phase energization) ends. To the normal drive, the initial drive is assumed to have been executed to the end, the encoder count value Ncnt at the end of the initial drive is estimated, and the estimated value is used as the initial position deviation learning value Gcnt. ing. 2 × (6-CASE) is a count-up value (hereinafter referred to as “Ncnt correction amount”) corresponding to the rotation angle for the number of unfinished excitations.
[0086]
FIG. 14 is a time chart for explaining the relationship among the excitation number counter CASE, the Ncnt correction amount, the energized phase, the A-phase signal, the B-phase signal, and the encoder count value Ncnt during the initial drive. For example, if the initial drive is to be terminated at the end of the UW-phase energization (when the number-of-excitations counter CASE has changed from 1 to 2) during the initial drive, the Ncnt correction amount = 2 × (6-CASE) = 2 × ( 6-2) = 8, and when ending the initial drive at the end of the UV-phase energization (when the number-of-excitations counter CASE changes from 3 to 4), Ncnt correction amount = 2 × (6-CASE) = 2 × (6-4) = 4.
[0087]
[Encoder counter]
Next, the processing content of the encoder counter routine shown in FIG. 15 will be described. This routine is started in synchronization with both the rising / falling edges of the A-phase signal and the B-phase signal by the AB-phase interrupt processing, and the rising / falling edges of the A-phase signal and the B-phase signal are set next. Count as follows. When this routine is started, first, in step 301, the values A (i) and B (i) of the A-phase signal and the B-phase signal are read, and in the next step 302, the count-up value ΔN calculation map of FIG. A search is performed to calculate a current value A (i), B (i) of the A-phase signal and the B-phase signal and a count-up value ΔN corresponding to the previous value A (i−1), B (i−1). .
[0088]
Here, the reason why the present values A (i) and B (i) of the A-phase signal and the B-phase signal and the previous values A (i-1) and B (i-1) are used is that the A-phase signal and the B-phase signal are used. This is because the rotation direction of the rotor 32 is determined based on the order in which the signals are generated. As shown in FIG. Is counted up, and in the reverse rotation (the rotation direction from the NotP range to the P range), the count-up value ΔN is set to a negative value, and the encoder count value Ncnt is counted down.
[0089]
After calculating the count-up value ΔN, the process proceeds to step 303, where the count-up value ΔN calculated in step 302 is added to the previous encoder count value Ncnt to determine the current encoder count value Ncnt. Thereafter, the process proceeds to step 304, where the current values A (i) and B (i) of the A-phase signal and the B-phase signal are changed to A (i-1) and B (i-1) for the next count processing. And the routine ends.
[0090]
[Motor F / B control]
Next, the processing content of the motor F / B control routine shown in FIG. 18 will be described. This routine is executed by the AB-phase interrupt processing and serves as the F / B control means referred to in the claims. When the motor F / B control execution condition is satisfied after the end of the initial driving, the rotor 32 is controlled. The energized phase is switched based on the encoder count value Ncnt and the initial position deviation learning value Gcnt until the rotational position (encoder count value Ncnt) of the target position Acnt reaches, for example, within 0.5 ° from the target position Acnt, and the rotor 32 is moved to the target position. Rotate in the direction of Acnt. Here, the motor F / B control execution condition is satisfied, for example, when all of the following conditions (1) to (4) are satisfied.
[0091]
(1) After the end of the initial drive (initial drive end flag Xend = ON)
(2) No fail-safe processing is performed (fail-safe processing execution flag Xfsop = OFF)
{Circle around (3)} Recovery processing is not being performed (recovery processing execution flag Xrcv = OFF)
(4) A predetermined stop holding time (for example, 10 ms) has elapsed since the end of the initial drive, or a predetermined stop holding time (for example, 10 ms) has elapsed since the start of motor energization.
[0092]
The stop holding time (4) is provided to determine the position of the rotor 32 at the start of the F / B control.
When all of the above conditions (1) to (4) are satisfied, the motor F / B control execution condition is satisfied, and the F / B permission flag Xfb is set to ON.
[0093]
When the motor F / B control routine of FIG. 18 is started, first, in step 311, whether the F / B permission flag Xfb is set to ON (whether the motor F / B control execution condition is satisfied or not) ), And if the F / B permission flag Xfb is OFF (the F / B control execution condition is not satisfied), the routine ends without performing the subsequent processing.
[0094]
On the other hand, if the F / B permission flag Xfb is set to ON, the routine proceeds to step 312, where the energized phase setting routine shown in FIG. 19 described later is executed, and the current encoder count value Ncnt and the initial position deviation learning value are set. The energization phase is set based on Gcnt, and in the next step 313, an energization processing routine (not shown) is executed to energize the energization phase set in step 312.
[0095]
On the other hand, when the energization phase setting routine of FIG. 19 is started in step 312, the rotation direction instruction value D is "1" meaning forward rotation (rotation direction from P range to NotP range) in step 321. It is determined whether or not. The rotation direction instruction value D is set based on the magnitude relationship between the current encoder count value Ncnt and the target value Acnt.
[0096]
If the rotation direction instruction value D = 1 (forward rotation), the process proceeds to step 322, where it is determined whether the rotation direction has been reversed in reverse to the rotation direction instruction (whether the encoder count value Ncnt has decreased). If not, the process proceeds to step 323, and the energized phase determination value Mptn is calculated using the current encoder count value Ncnt, the initial position deviation learning value Gcnt, the forward rotation direction phase lead amount K1, and the speed correction amount Ks by the following equation. Update.
Mptn = Ncnt−Gcnt + K1 + Ks
[0097]
Here, the forward rotation direction phase lead amount K1 is the phase lead amount of the conduction phase necessary for rotating the rotor 32 in the forward direction (the phase lead amount of the conduction phase with respect to the current position of the rotor 32). For example, K1 = 4 Is set.
[0098]
The speed correction amount Ks is a phase lead correction amount set according to the rotation speed of the rotor 32. In the low speed range, the speed correction amount Ks is set to 0, and as the speed increases, the speed correction amount Ks is increased to, for example, 1 or 2. Thus, the energized phase determination value Mptn is corrected so that the energized phase is suitable for the rotation speed of the rotor 32.
[0099]
On the other hand, if it is determined in step 322 that the rotation direction has been reversed against the rotation direction instruction, the energized phase determination value Mptn is not updated to prevent reverse rotation. In this case, electricity is supplied to the energized phase immediately before the reverse rotation (previous energized phase), and a braking torque is generated in a direction to suppress the reverse rotation of the rotor 32.
[0100]
If it is determined in step 321 that the rotation direction instruction value D = 0 (reverse rotation), that is, the rotation direction is from NotP range to P range, the process proceeds to step 324, where the rotation direction is reversed against the rotation direction instruction. It is determined whether or not the encoder count value Ncnt has increased. If the rotation has not been reversed, the process proceeds to step 325, where the current encoder count value Ncnt, the initial displacement learning value Gcnt, and the phase in the reverse rotation direction are advanced. The energized phase determination value Mptn is updated by the following equation using the amount K2 and the speed correction amount Ks.
Mptn = Ncnt-Gcnt-K2-Ks
[0101]
Here, the amount of phase advance K2 in the reverse rotation direction is the amount of phase advance of the energized phase required to rotate the rotor 32 in the reverse direction (the amount of phase advance of the energized phase with respect to the current position of the rotor 32). Is set. The speed correction amount Ks is the same as in the case of the forward rotation.
[0102]
On the other hand, when it is determined in step 324 that the rotation direction has been reversed against the rotation direction instruction, the energized phase determination value Mptn is not updated to prevent the rotation from being reversed. In this case, electricity is supplied to the energized phase immediately before the reverse rotation (previous energized phase), and a braking torque is generated in a direction to suppress the reverse rotation of the rotor 32.
[0103]
After the current energized phase determination value Mptn is determined as described above, the process proceeds to step 326, where the energized phase determination value Mptn is divided by "12" to obtain the remainder Mptn% 12. Here, “12” corresponds to the increase / decrease amount of the encoder count value Ncnt during one cycle of the energized phase.
[0104]
After the calculation of Mptn% 12, the process proceeds to step 327, where the conversion table of FIG. 20 is searched, and the energized phase corresponding to Mptn% 12 is selected and set as the current energized phase.
[0105]
FIG. 21 is a time chart for explaining a phase that is first energized when rotation is started from the U phase. In this case, since the speed correction amount Ks = 0, when the forward rotation (rotation in the direction from the P range to the NotP range) is started, the energized phase determination value Mptn is calculated by the following equation.
Mptn = Ncnt−Gcnt + K1 = Ncnt−Gcnt + 4
When starting the normal rotation from the U phase, the remainder of (Ncnt−Gcnt) / 12 is 6, so that Mptn% 12 = 6 + 4 = 10, and the first energized phase is the V phase.
[0106]
On the other hand, when the reverse rotation (rotation from the NotP range to the P range direction) is started from the U phase, the energized phase determination value Mptn is calculated by the following equation.
Mptn = Ncnt-Gcnt-K2 = Ncnt-Gcnt-3
When the reverse rotation is started from the U phase, Mptn% 12 = 6-3 = 3, and the first energized phase is the W phase.
[0107]
In this way, by setting the forward rotation direction phase lead amount K1 and the reverse rotation direction phase lead amount K2 to 4 and 3, respectively, the switching pattern of the energized phase in the forward rotation direction and the reverse rotation direction can be made symmetric. In both the forward rotation direction and the reverse rotation direction, the phase at a position shifted by two steps from the current position of the rotor 32 can be first excited to start rotation.
[0108]
[Z phase correction]
By the way, while the SR motor 12 is rotating, as shown in FIG. 23, the A-phase signal and the B-phase signal are alternately output from the encoder 46 in synchronization with the rotation of the rotor 32, and the switching of the energized phase goes through one cycle. Each time the rotor 32 rotates 45 °, a Z-phase signal is output. Since the reference rotational position of the rotor 32 can be accurately detected from the Z-phase signal, the energized phase (encoder count value) when the Z-phase signal is output changes to the energized phase (encoder count value) corresponding to the reference rotational position of the rotor 32. ), It is possible to confirm whether or not the correspondence between the rotational position of the rotor 32 and the energized phase (encoder count value) is not shifted. By performing the Z-phase correction for correcting the deviation, highly reliable motor control can be performed.
[0109]
This Z-phase correction is executed by a Z-phase correction routine shown in FIG. This routine is started in synchronization with both rising and falling edges of the A-phase signal by the A-phase interrupt processing. When this routine is started, first, at step 401, it is determined whether or not the following (1) to (3) Z-phase correction execution conditions are satisfied.
(1) The value Z of the Z-phase signal is 1 (high level)
{Circle around (2)} The encoder failure flag Xfailenc is set to OFF, meaning that there is no failure of the encoder 46.
{Circle around (3)} The Z-phase correction prohibition flag Xzng is set to OFF, meaning that Z-phase correction is permitted.
[0110]
Here, the encoder failure flag Xfailenc is determined to be a failure of the encoder 46 when the number of times of the Z-phase correction exceeds the determination value, and Xfailenc = ON is set in step 410 described later. Further, the Z-phase correction prohibition flag Xzng is set to Xzng = ON when an abnormality of the Z-phase signal of the encoder 46 is detected in the routine of FIG.
[0111]
If at least one of the above conditions (1) to (3) is not satisfied, the Z-phase correction execution condition is not satisfied, and the routine ends without performing the subsequent processing.
[0112]
On the other hand, if all of the above conditions (1) to (3) are satisfied, the condition for executing the Z-phase correction is satisfied, and the routine proceeds to step 402, where the Z-phase signal is output (the value Z of the Z-phase signal is inverted to 1). ) Is stored. The position detection count value Nzone is obtained by correcting the encoder count value Ncnt at the time of outputting the Z-phase signal with the initial position deviation learning value Gcnt. The reference rotation position of the rotor 32 (when the Z-phase signal is Output position) is detected.
Nzone = Ncnt−Gcnt
[0113]
After that, the process proceeds to step 403, and whether the rotation direction of the rotor 32 is the forward rotation direction (the rotation direction from the P range to the NotP range) is determined by whether or not the current activation timing of this routine is the rising of the A-phase signal. Determine whether or not.
[0114]
As shown in FIG. 23, since the A-phase signal and the Z-phase signal are output with a phase difference of 90 ° in electrical angle, the Z-phase signal output period in the forward rotation direction (the rotation direction from the P range to the NotP range). During the Z-phase signal output period, the A-phase signal falls in the reverse rotation direction (NotP range → P range rotation direction). Therefore, whether the A-phase signal rises or falls during the Z-phase signal output period can determine whether the rotation direction is the normal rotation direction or the reverse rotation direction.
[0115]
If it is determined in step 403 that the activation timing of this routine is the rising timing of the A-phase signal, that is, the rotation direction of the rotor 32 is the forward rotation direction (the rotation direction from the P range to the NotP range), Proceeding to 404, a deviation gz between the design value and the actual value of the energized phase when the Z-phase signal is output is calculated.
gz = K1-Nzon% 12
[0116]
Here, K1 is a forward rotation direction lead amount of an energized phase required for rotating the rotor 32 in the forward direction, and is set to, for example, K1 = 4. Nzone% 12 is the remainder when the position detection count value Nzon at the time of outputting the Z-phase signal is divided by “12”.
[0117]
In the present embodiment, as shown in FIG. 23, Nzon% 12 is designed to be 4 in the forward rotation direction (P range → NotP range rotation direction), so that the control system operates normally. If so, gz = K1−Nzon% 12 = 0.
[0118]
On the other hand, when it is determined in step 403 that the current activation timing of this routine is the falling timing of the A-phase signal, that is, the rotation direction of the rotor 32 is the reverse rotation direction (the rotation direction from the NotP range to the P range). Proceeds to step 405, and calculates a deviation gz between the design value and the actual value of the energized phase when the Z-phase signal is output.
gz = K2-Nzon% 12
Here, K2 is the amount of phase advance in the reverse rotation direction of the energization phase required to rotate the rotor 32 in the reverse direction, for example, K2 = 3.
[0119]
In the present embodiment, as shown in FIG. 23, Nzon% 12 is designed to be 3 in the reverse rotation direction (the rotation direction from the NotP range to the P range), so that the control system operates normally. If so, gz = K2−Nzon% 12 = 0.
[0120]
After calculating the deviation gz between the design value and the actual value, the process proceeds to step 406, and it is determined whether the deviation gz between the design value and the actual value of the energized phase at the time of outputting the Z-phase signal is 0, and If the deviation gz from the actual value is 0, the control system is operating normally, and this routine ends without performing the subsequent processing such as Z-phase correction.
[0121]
On the other hand, if the deviation gz between the design value and the actual value of the energized phase at the time of outputting the Z-phase signal is not 0, it is determined that the Z-phase correction is necessary, and the process proceeds to step 407, where the number of times of the Z-phase correction is The Z-phase correction number counter Cgz to be counted is counted up, and the routine proceeds to step 408, where the deviation of the initial position deviation learning value Gcnt is corrected by the deviation gz between the design value and the actual value.
Gcnt = Gcnt−gz
[0122]
Thereafter, it is determined whether or not the value of the Z-phase correction number counter Cgz (the number of Z-phase corrections) has exceeded a determination value. If the value is equal to or less than the determination value, it is not determined that the encoder 46 has failed yet. If the value of the number-of-corrections counter Cgz (the number of times of Z-phase correction) exceeds the determination value, it is determined that the encoder 46 has failed, and the routine proceeds to step 410, where the encoder failure flag Xfailenc is set to ON indicating that the encoder 46 has a failure. And the routine ends.
[0123]
When the encoder failure flag Xfailenc is set to ON, the routine is always determined as “No” in step 401 and the routine is forcibly terminated even if the routine is started by the A-phase interrupt processing thereafter. , Z-phase correction is not performed.
[0124]
[Recovery processing]
The recovery processing routine shown in FIG. 28 is started at a predetermined cycle (for example, 1 ms cycle), and executes the recovery processing as follows when the F / B control state becomes abnormal. When this routine is started, first, in step 801, it is determined whether or not the recovery processing execution flag Xrcv is OFF. If Xrcv = OFF (recovery processing is not being executed), the process proceeds to step 802, and the next step is performed. It is determined whether the recovery processing execution conditions (1) to (4) are satisfied.
[0125]
(1) The A-phase signal, B-phase signal of the encoder 46, and the output signal of the output shaft sensor 14 are all normal.
(2) All abnormality detection flags Xfnsw, Xfcnt, and Xfrcv set in a system abnormality diagnosis routine of FIG.
(3) F / B permission flag Xfb is ON (F / B control is being performed)
(4) One of three abnormality detection flags Xfab, Xfta, and Xftb relating to the A-phase signal and the B-phase signal set in the AB-phase signal abnormality detection routine during F / B control in FIG. Or the reverse rotation detection flag Xr set in a reverse rotation determination routine of FIG. 32 described later is ON (there is reverse rotation).
[0126]
If at least one of the four conditions (1) to (4) is not satisfied, the recovery processing execution condition is not satisfied, and the routine ends without performing the subsequent processing.
[0127]
On the other hand, if all of the above conditions (1) to (4) are satisfied, the recovery process execution condition is satisfied, the process proceeds from step 802 to step 803, the recovery process execution flag Xrcv is set to ON, and the next step 804 Then, based on the relationship shown in FIG. 29, the current position count value Ncntop is calculated by the following equation using the current output signal Vnsw of the output shaft sensor 14.
Ncntop = INT [｛(Vnsw−Vp) / ΔV
× Kop × 61 / 3.75-Kofp｝ × 1/2] × 2
ΔV = Vnp−Vp
[0128]
Here, INT is a function for converting a numerical value in [] into an integer, Vp is a P range output learning value of the output shaft sensor 14, Vnp is a NotP range output learning value of the output shaft sensor 14, and Kop is a detent lever 15 The maximum rotation angle and Kofp are offset values on the P range side. The above equation is set so that the position count value Ncntop is a multiple of two.
[0129]
The maximum rotation angle Kop of the detent lever 15 is the rotation angle of the detent lever 15 from the side wall of the P range holding recess 24 of the detent lever 15 to the side wall of the NotP range holding recess 25, and the design value ΔNd of the movable range of the detent lever 15 And the play amount ΔGp on the P range side. In the present embodiment, ΔNd = 20.5 (deg) and ΔGp = 3 (deg).

[0130]
In the above formula for calculating the position count value Ncntop, Kop × 61 / 3.75 is a value obtained by converting the maximum rotation angle Kop of the detent lever 15 into the number of motor drive steps (the number of excitations). “61” is the reduction ratio of the reduction mechanism 26, and “3.75” is the rotation angle of the rotor 32 in one step (per excitation). This position count value Ncntop is used as an initial value of the position count value Ncntop at the start of the open loop control.
[0131]
After calculating the position count value Ncntop, the process proceeds to step 805, where a recovery process execution number counter Crcv for counting the number of times of execution of the recovery process is counted up, and in the next step 806, the open loop control execution flag Xopen is set to ON. The open loop control is executed by an open loop control routine shown in FIG.
[0132]
During execution of the recovery process (open loop control), each time this routine is started, “No” is determined in step 801, and the process proceeds to step 807, where the open loop control completion flag Xopcmp indicates that the open loop control is completed. Then, it is determined whether or not Xopcmp = OFF, and if Xopcmp = OFF, this routine is terminated as it is, and the open loop control is continued.
[0133]
After that, when the open loop control is completed, the process proceeds to step 808, where the initial drive is re-executed, and the initial drive end flag Xend = OFF (initial drive is not completed), the range switching operation flag Xchg = OFF (range switching operation) None), the energization time counter CT = 0, the number of excitation counter CASE = 0, the encoder count value Ncnt = 0, and the initial position deviation learning value Gcnt = 0. Then, in the next step 811, after setting the open loop control execution flag Xopen = OFF and ending the open loop control (recovery processing), the process proceeds to step 812, where the recovery processing execution flag Xrcv = OFF, the A-phase signal, B A phase signal abnormality detection flag Xfab = OFF, Xfta = OFF, Xftb = OFF, and a third abnormality detection flag Xfrcv = OFF described later are set.
[0134]
[AB phase signal error detection during F / B control]
The F / B control AB phase signal abnormality detection routine shown in FIG. 30 is started at a predetermined cycle (for example, 1 ms cycle), and determines whether there is an abnormality in the A phase signal and the B phase signal of the encoder 46 during the F / B control. Is determined as follows. When this routine is started, first, in step 821, it is determined whether or not the F / B permission flag Xfb = ON (F / B control is being performed). If the F / B permission flag Xfb = OFF, the process proceeds to step 821. Proceeding to step 822, the A-phase signal counter Na = 0, the B-phase signal counter Nb = 0, the A-phase signal edge interval counter ΔTa = 0, and the B-phase signal edge interval counter ΔTb = 0, and this routine ends.
[0135]
On the other hand, if it is determined in step 821 that the F / B permission flag Xfb = ON (F / B control is being performed), the AB phase signal abnormality determination execution condition is satisfied, and the process proceeds to step 823, where the A phase signal is determined. A deviation ΔNab between the count value Na and the B-phase signal count value Nb is calculated.
ΔNab = | Na-Nb |
Here, the A-phase signal count value Na and the B-phase signal count value Nb are values obtained by separately counting the edges of the A-phase signal and the edges of the B-phase signal.
[0136]
Then, in the next step 824, the AB-phase signal count value deviation ΔNab is compared with the abnormality determination value Kab (for example, 3), and if the AB-phase signal count value deviation ΔNab is equal to or greater than the abnormality determination value Kab, the A-phase signal It is determined that one of the count value Na and the B-phase signal count value Nb is abnormal, and the process proceeds to step 825, where the AB-phase signal abnormality detection flag Xfab is set to ON (abnormal), and the process proceeds to the next step 826. move on.
[0137]
On the other hand, if the AB phase signal count value deviation ΔNab is smaller than the abnormality determination value Kab, the AB phase signal abnormality detection flag Xfab is kept OFF (no abnormality), and the routine proceeds to step 826.
[0138]
If the A-phase signal and the B-phase signal are normally output during the F / B control, the deviation ΔNab between the A-phase signal count value Na and the B-phase signal count value Nb becomes 0 or 1, so If the signal count value deviation ΔNab is 2 or more, it is possible that one of the signals is missing or a noise pulse is mixed in (see FIG. 25). Therefore, whether or not the A-phase signal and the B-phase signal are abnormal can be determined based on whether or not the AB-phase signal count value deviation ΔNab is equal to or larger than the abnormality determination value Kab. In the present embodiment, the abnormality determination value Kab is set to 3 with a margin, but may be set to 2 or 4 or more.
[0139]
After performing the abnormality determination process based on the AB-phase signal count value deviation ΔNab as described above, the process proceeds to step 826, where the counters that count the edge interval ΔTa of the A-phase signal and the edge interval ΔTb of the B-phase signal are counted up. After that, the process proceeds to step 827, where the A-phase signal edge interval ΔTa is compared with the abnormality determination value Kta (for example, a value that is 10 times the average value of ΔTa), and if the A-phase signal edge interval ΔTa is equal to or greater than the abnormality determination value Kta. If so, it is determined that the A-phase signal is abnormal, and the routine proceeds to step 828, where the A-phase signal abnormality detection flag Xfta is set to ON (abnormal), and the routine proceeds to the next step 829.
[0140]
On the other hand, if the A-phase signal edge interval ΔTa is less than the abnormality determination value Kta, the A-phase signal abnormality detection flag Xfta is kept OFF (no abnormality), and the routine proceeds to step 829. Then, in step 829, the B-phase signal edge interval ΔTb is compared with an abnormality determination value Ktb (for example, a value ten times the average value of ΔTb), and if the B-phase signal edge interval ΔTb is greater than or equal to the abnormality determination value Ktb. For example, it is determined that the B-phase signal is abnormal, the process proceeds to step 830, the B-phase signal abnormality detection flag Xftb is set to ON (abnormal), and the routine ends.
[0141]
On the other hand, if the B-phase signal edge interval ΔTb is smaller than the abnormality determination value Ktb, the B-phase signal abnormality detection flag Xftb is kept OFF (no abnormality), and this routine ends.
[0142]
Since the edge intervals ΔTa and ΔTb of each signal change according to the rotor rotation speed, if the edge intervals ΔTa and ΔTb of each signal become too long during the F / B control, the rotor 32 stops abnormally and abnormally stops. Since a drop and a missing pulse are considered (see FIG. 25), A is determined by whether or not the edge intervals ΔTa and ΔTb of each signal are equal to or larger than the abnormality determination values Kta and Kta (for example, 10 times the average value of ΔTa and ΔTb). It is possible to determine whether or not the phase signal and the B-phase signal are abnormal.
[0143]
If the edge intervals ΔTa and ΔTb of each signal become too short, the noise pulse may be erroneously recognized as a legitimate pulse signal. Therefore, the edge intervals ΔTa and ΔTb of each signal are equal to or less than a predetermined abnormality determination value. May be used to determine whether there is an abnormality in the A-phase signal and the B-phase signal.
[0144]
Further, only one of the abnormality determination processing based on the AB phase signal count value deviation ΔNab (steps 823 to 825) and the abnormality determination processing based on the edge intervals ΔTa and ΔTb of each signal (steps 826 to 830) is performed, and the other is performed. The processing may be omitted.
[0145]
[Reverse counter]
The reverse rotation counter routine shown in FIG. 31 is executed by the AB phase interrupt processing, and counts the number of times the rotation direction of the rotor 32 continuously reverses against the rotation direction instruction as follows. When this routine is started, first, in step 831, the rotation direction instruction value D is any one of 1 (the rotation direction from the P range to the NotP range), 0 (stop), and -1 (the rotation direction from the NotP range to the P range). Is determined, and if the rotation direction instruction value D is 0 (stop), this routine ends without performing the subsequent processing.
[0146]
On the other hand, if the rotation direction instruction value D is 1 (the rotation direction from the P range to the NotP range), the process proceeds to step 832, where the previous encoder count value Ncnt (i-1) and the current encoder count value Ncnt (i) are obtained. It is determined whether or not the rotation direction of the rotor 32 has been reversed (whether or not the encoder count value Ncnt has decreased) against the rotation direction instruction by determining whether or not the difference is equal to or greater than 1. The process proceeds to step 834 to count up the reverse rotation counter Cr for counting the number of reverse rotations. If the reverse rotation has not been performed, the process proceeds to step 835 to reset the reverse rotation counter Cr to zero.
[0147]
If the rotation direction instruction value D is -1 (the rotation direction from the NotP range to the P range), the process proceeds to step 833, and the previous encoder count value Ncnt (i-1) and the current encoder count value Ncnt (i) It is determined whether or not the rotation direction of the rotor 32 has been reversed (whether or not the encoder count value Ncnt has increased) against the rotation direction instruction, based on whether or not the difference from the rotation direction is less than or equal to −1. If so, the flow proceeds to step 836, where the reverse rotation counter Cr is counted up. If the rotation is not reversed, the flow proceeds to step 835 to reset the reverse rotation counter Cr to zero.
[0148]
[Reverse judgment]
The reverse rotation determination routine shown in FIG. 32 is started at a predetermined cycle (for example, 1 ms cycle). When this routine is started, first, in step 841, it is determined whether or not the F / B permission flag Xfb = ON (F / B control is being performed). Proceeding to 844, the reverse rotation detection flag Xr is set to OFF, and this routine ends.
[0149]
On the other hand, if it is determined in step 841 that the F / B permission flag Xfb = ON (F / B control is being performed), the reverse rotation determination execution condition is satisfied, and the routine proceeds to step 842, where the reverse rotation counter routine of FIG. It is determined whether or not the rotor 32 continues to rotate reversely against the rotation direction instruction by determining whether or not the reverse rotation count value Cr obtained in the above is equal to or greater than the abnormality determination value Kr (for example, 4). If it is equal to or greater than the determination value Kr, the process proceeds to step 843, where the reverse rotation detection flag Xr is set to ON to determine the determination of the reverse rotation. If the reverse rotation count value Cr is less than the abnormality determination value Kr, the process proceeds to step 844 to perform the reverse rotation. The detection flag Xr is kept off.
[0150]
[System error diagnosis]
The system abnormality diagnosis routine shown in FIG. 33 is started at a predetermined cycle (for example, a cycle of 8 ms). When this routine is started, first, in step 851, it is determined whether or not the following abnormality diagnosis execution conditions (1) and (2) are satisfied.
(1) A predetermined time (for example, 1 s) has elapsed after the power supply to the ECU 41 is turned on by operating an ignition switch (not shown).
(2) Recovery processing is not being executed (recovery processing execution flag Xrcv = OFF)
[0151]
If any one of these two conditions (1) and (2) is not satisfied, the abnormality diagnosis execution condition is not satisfied, and the routine proceeds to step 852, where the provisional abnormal state continuation time counters Cfnsw and Cfcnt are described. This routine is reset to 0 and the routine ends.
[0152]
On the other hand, if both of the above two conditions (1) and (2) are satisfied, the abnormality diagnosis execution condition is satisfied, and the routine proceeds from step 851 to step 854, where the output shaft sensor 14 is normal and the command shift range And the range detected from the output signal Vnsw of the output shaft sensor 14 is determined. If the determination result is “Yes”, the process proceeds to step 856 and the first temporary abnormal state continuation time Reset the counter Cfnsw to 0.
[0153]
On the other hand, if the decision result in the step 854 is “No”, the process proceeds to a step 855 to count up the first temporary abnormal state continuation time counter Cfnsw, and to output the command shift range and the output signal of the output shaft sensor 14. The duration of the state where Vnsw does not match is measured, and in the next step 857, it is determined whether or not the value of the first temporary abnormal state duration counter Cfnsw is larger than the abnormality determination value Kfnsw. Here, the abnormality determination value Kfnsw is set to a time (for example, 10 s) sufficiently longer than the longest time required for a normal range switching operation. Therefore, if the value of the first temporary abnormal state duration time counter Cfnsw is larger than the abnormality determination value Kfnsw, the range switching operation is not performed normally, and it is determined that there is an abnormality in the system. Proceeding to 858, the first abnormality detection flag Xfnsw is set to ON. If the determination result of step 857 is “No”, the first abnormality detection flag Xfnsw is kept OFF (no abnormality).
[0154]
As described above, after performing the first abnormality diagnosis using the output signal Vnsw of the output shaft sensor 14, the process proceeds to step 859 in FIG. 34, where the deviation | Ncnt− between the encoder count value Ncnt and the target count value Acnt− It is determined whether Actnt | is greater than a predetermined value Kcnt (e.g., a value corresponding to 5 [deg.]). If this count value deviation | Ncnt-Acnt | is equal to or smaller than the predetermined value Kcnt, the process proceeds to step 861, and the second temporary The abnormal state duration time counter Cfcnt is reset to 0.
[0155]
On the other hand, if the decision result in the above step 859 is "Yes", the process proceeds to a step 860 to count up the second provisional abnormal state duration time counter Cfcnt, and the count value deviation | Ncnt-Acnt | The duration of the state greater than Kcnt is measured, and in the next step 862, it is determined whether or not the value of the second temporary abnormal state duration time counter Cfcnt is greater than the abnormality determination value Kfcnt. Here, the abnormality determination value Kfcnt is set to a time (for example, 10 s) sufficiently longer than the maximum time required for a normal range switching operation. Therefore, if the value of the second temporary abnormal state continuation time counter Cfcnt is larger than the abnormality determination value Kfcnt, the range switching operation is not performed normally, and it is determined that there is an abnormality in the system. Proceeding to 863, the second abnormality detection flag Xfcnt is set to ON. If the determination result in step 862 is “No”, the second abnormality detection flag Xfcnt is kept OFF (no abnormality).
[0156]
As described above, after performing the second abnormality diagnosis using the count value deviation | Ncnt−Acnt |, the process proceeds to step 864, and the number of times of execution of the recovery process Crcv counted in step 805 of the recovery process routine of FIG. Is greater than or equal to the abnormality determination value Kcrv (for example, three times). If the number of times of execution of the recovery process Crcv is greater than or equal to the abnormality determination value Kcrv, the process proceeds to step 865, where it is determined that there is a system abnormality. Proceeding to 865, the third abnormality detection flag Xfrcv is set to ON, and this routine ends. That is, if the recovery process is continuously repeated for an abnormality determination value Kcrv (for example, three times or more), the F / B control system is not functioning properly, and it is determined that there is a system abnormality. is there.
[0157]
On the other hand, if the recovery processing execution number Crcv is less than the abnormality determination value Kcrv, the third abnormality detection flag Xfrcv is kept OFF (no abnormality), and this routine ends.
[0158]
[Fail-safe processing]
The fail-safe processing routine shown in FIG. 35 is started at a predetermined cycle (for example, a cycle of 8 ms). When this routine is started, first, in step 901, it is determined whether or not the following conditions (1) and (2) for performing fail-safe processing are satisfied.
(1) The range can be switched by open loop control (system failure flag Xfailoff = OFF)
{Circle around (2)} At least one of the three abnormality detection flags Xfnsw, Xfcnt, and Xfrcv set in the system abnormality diagnosis routine of FIG. 33 is ON (abnormal), or an AB phase disconnection detection of FIG. The AB phase disconnection detection flag Xfailab set in the routine is ON (AB phase disconnected)
[0159]
If any one of these two conditions (1) and (2) is not satisfied, the fail-safe processing execution condition is not satisfied, and the routine is terminated without performing the subsequent processing.
[0160]
On the other hand, if the above two conditions (1) and (2) are both satisfied, the fail-safe process execution condition is satisfied, the process proceeds from step 901 to step 902, and the fail-safe process execution flag Xfsop is set to ON. In the next step 903, a warning lamp is turned on, or a warning is displayed on an information display section or the like of the meter panel to warn the driver.
[0161]
Thereafter, the process proceeds to step 904, where it is determined whether or not the value of the fail-safe processing interval counter Cfsop for measuring the elapsed time after the end of the previous fail-safe processing is less than a predetermined value Kfsop. Here, the predetermined value Kfsop is such that the engaging portion 23a of the detent spring 23 slides down to the bottom thereof along the inclined side walls of the holding concave portions 24 and 25 of the detent lever 15 by the elastic force of the detent spring 23 after the end of the previous fail-safe processing. The time is set to be somewhat longer (for example, 500 ms) than the time required for the operation.
[0162]
If the value of the fail-safe processing interval counter Cfsop (the elapsed time after the end of the previous fail-safe processing) is less than the predetermined value Kfsop, the process proceeds to step 905, where the fail-safe processing interval counter Cfsop is counted up and the previous fail-safe processing is performed. The elapsed time after the end is counted. Thereafter, the process proceeds to step 906, where the energization of all phases is turned off, and in the next step 907, the energization flag Xon is set to OFF meaning energization off, and the routine ends. As a result, the energization of all phases is kept off until the elapsed time after the end of the previous fail-safe processing exceeds the predetermined value Kfsop.
[0163]
Thereafter, when the value of the fail-safe processing interval counter Cfsop (elapsed time after the end of the previous fail-safe processing) becomes equal to or greater than a predetermined value Kfsop, the process proceeds from step 904 to step 908, and the open-loop control execution flag Xopen is turned off. It is determined whether or not there is, and if it is OFF, the process proceeds to step 909, and it is determined whether or not the command shift range sft has been switched. As a result, if it is determined that the command shift range sft has not been switched, the process proceeds to step 910, and the state in which energization of all phases is turned off is continued.
[0164]
On the other hand, if it is determined in step 909 that the command shift range sft has been switched, the process proceeds to step 911, and the current output shaft sensor 14 is output in the same manner as in step 804 of the recovery processing routine of FIG. The current position count value Ncntop is calculated using the output signal Vnsw. This position count value Ncntop is used as an initial value of the position count value Ncntop at the start of the open loop control.
[0165]
Thereafter, the process proceeds to step 912, in which the open-loop control execution flag Xopen is set to ON, and the open-loop control is executed by an open-loop control routine shown in FIG. 37, which will be described later. , The fail-safe process execution number counter Cfs is counted up, and this routine ends.
[0166]
If it is determined in step 908 that the open loop control execution flag Xopen is ON, the process proceeds to step 914 in FIG. 36, where the open loop control completion flag Xopcmp is set to ON indicating open loop control completion. It is determined whether or not Xopcmp = OFF, and this routine ends without performing the subsequent processing.
[0167]
On the other hand, if it is determined in step 914 that the open loop control completion flag Xopcmp is set to ON, the process proceeds to step 915, where the command shift range and the range detected from the output signal Vnsw of the output shaft sensor 14 are set. Is determined, and if they match, the routine proceeds to step 916, where the fail-safe process execution counter Cfs is reset to zero. Then, in the next step 918, the open loop control completion flag Xopcmp is set to OFF, and in the next step 919, the open loop control execution flag Xopen is set to OFF, and this routine ends.
[0168]
If it is determined in step 915 that the command shift range does not match the output signal Vnsw of the output shaft sensor 14, the process proceeds to step 917, where the value of the fail-safe process execution counter Cfs is set to the failure determination value. (For example, three times) or more, and if the value of the fail-safe processing execution counter Cfs is less than the failure determination value, it is not determined that a failure has occurred yet, and the open-loop control completion flag Xopcmp = OFF. Then, the open loop control execution flag Xopen = OFF is set (steps 918 and 919), and this routine ends.
[0169]
Thereafter, when the value of the fail-safe processing execution number counter Cfs becomes equal to or more than a failure determination value (for example, three times), it is determined at that time that the system has failed (a state in which range switching is impossible even with open loop control). Proceeding from step 917 to step 920, the system failure flag Xfailoff is set to ON, and in the next step 921, energization of all phases is turned off, and this routine ends.
[0170]
[Open loop control]
The open loop control routine shown in FIG. 37 to FIG. 39 is started at a predetermined cycle (for example, 1 ms cycle) and plays a role as an open loop control means described in the claims. When this routine is started, it is first determined in step 931 whether or not the open loop control execution flag Xopen is set to ON. If Xopen = OFF, this routine is executed without performing the subsequent processing. finish.
[0171]
On the other hand, if the open loop control execution flag Xopen is set to ON, the process proceeds from step 931 to step 932, where the time counter Top for measuring time is counted up. At the next step 933, the value of the time counter Top is Is determined to be equal to or longer than the excitation time tm of the current energized phase, and if less than the excitation time tm, this routine is terminated without performing the subsequent processing. Thus, the processing after step 934 is executed for each excitation time tm.
[0172]
Thereafter, when the value of the time counter Top becomes equal to or longer than the excitation time tm of the current energized phase, the process proceeds from step 933 to step 934, where the target count value Acnt is compared with the current position count value Ncntop, and the rotor 32 is compared. Is determined. As a result, when the current position count value Ncntop is smaller than the target count value Actnt, the driving direction of the rotor 32 is determined to be the rotation direction (forward rotation direction) from the P range to the NotP range, and the process proceeds to step 935. Then, the energized phase determination value Mptnop is counted up by one, and in the next step 936, the position count value Ncntop is counted up by two. This is because the rotor 32 rotates by two counts (3.75 °) by one excitation.
[0173]
On the other hand, if the current position count value Ncntop is larger than the target count value Acnt, it is determined that the driving direction of the rotor 32 is the rotation direction from the NotP range side to the P range side (reverse rotation direction), and step 937 is performed. Then, the energized phase determination value Mptnop is counted down by one, and in the next step 938, the position count value Ncntop is counted down by two.
[0174]
Thereafter, the routine proceeds to step 939, where a remainder Mptnop% 6 when the energized phase determination value Mptnop is divided by "6" is calculated. Here, “6” corresponds to the amount of increase or decrease of the energized phase determination value Mptnop during one cycle of the energized phase in the open loop control.
[0175]
After calculating Mptnop% 6, the routine proceeds to step 940, where the energized phase is selected according to Mptnop% 6 from the table of FIG. 40, and the energized phase is excited. In the present embodiment, during the open loop control, the energized phase is selected so as to be driven by the 1-2 phase excitation method, and the first excitation when the open loop control is started is two-phase energized (WU phase energized or UV phase energization or VW phase energization), that is, the initial value of the energized phase determination value Mptnop is set so that Mptnop% 6 = 1, 3 or 5.
[0176]
Thereafter, the process proceeds to step 941 in FIG. 38, and the position deviation dev is calculated by the following equation.
dev = | Acnt-Ncntop | / 2
[0177]
As is apparent from the above equation, the position deviation dev is の of the deviation (absolute value) between the target count value Acnt and the current position count value Ncntop.
[0178]
Then, in the next step 942, the position deviation dev is compared with the rotation speed parameter mspd. The rotation speed parameter mspd is a parameter that is increased or decreased according to the position deviation dev. By setting the excitation time tm from the table in FIG. 42 according to the rotation speed parameter mspd, the rotation speed, acceleration, Control deceleration. In this case, as the target rotation speed increases, the rotation speed parameter mspd increases, and the excitation time tm is set to a shorter time. The initial value of the rotation speed parameter mspd is 0.
[0179]
If it is determined in step 942 that the position deviation dev is larger than the rotation speed parameter mspd, the process proceeds to step 943, where the rotation speed parameter mspd is counted up by one, and in the next step 944, the minimum rotation speed parameter mspdmin is set. Set to 0. This minimum rotation speed parameter mspdmin is a parameter (lower limit guard value) for limiting the minimum value of the rotation speed parameter mspd. As described above, when the position deviation dev is larger than the rotation speed parameter mspd, by setting the minimum rotation speed parameter mspdmin to 0, the excitation time tm can be reduced from the start of the open loop control to the start of rotation of the rotor 32. The maximum time (50 ms) can be set.
[0180]
On the other hand, if it is determined in step 942 that the position deviation dev is smaller than the rotation speed parameter mspd, the process proceeds to step 945, where the rotation speed parameter mspd is counted down by 1, and in the next step 946, the minimum rotation speed parameter mspdmin is set. Is set to 4. Thus, after the rotation of the rotor 32 rises, the excitation time tm can be limited to, for example, 5 ms or less.
[0181]
If it is determined in step 942 that the position deviation dev matches the rotation speed parameter mspd, the process proceeds to step 947 without changing the rotation speed parameter mspd, and the minimum rotation speed parameter mspdmin is set to 0.
[0182]
After the rotation speed parameter mspd and the minimum rotation speed parameter mspdmin are set as described above, the process proceeds to step 948, and the maximum rotation speed parameter mspdmax is set according to the battery voltage (power supply voltage) using the table of FIG. . The maximum rotation speed parameter mspdmax is a parameter (upper limit guard value) for limiting the maximum value of the rotation speed parameter mspd. As the battery voltage increases, the maximum rotation speed parameter mspdmax increases to increase the rotation speed of the rotor 32. I can do it.
[0183]
Then, in the next step 949, guard processing is performed so that the rotation speed parameter mspd falls within the range between the minimum rotation speed parameter mspdmin and the maximum rotation speed parameter mspdmax. For example, if the rotation speed parameter mspd updated in step 943 or 945 is smaller than the minimum rotation speed parameter mspdmin, then mspd = mspdmin, and if the rotation speed parameter mspd is larger than the maximum rotation speed parameter mspdmax, mspd = mspdmax, and when mspdmin ≦ mspd ≦ mspdmax, the rotation speed parameter mspd is used as it is.
[0184]
After that, the routine proceeds to step 950, where the excitation time tm is set according to the rotation speed parameter mspd using the table of FIG. In the example of FIG. 42, when the rotation speed parameter mspd is in the range of 0 to 3, the excitation time tm is set to the longest time (for example, 50 ms), and synchronization between the rotation phase of the rotor 32 and the energized phase is ensured at startup. Have to be able to. When the rotation speed parameter mspd is in the range of 4 to 9, as the rotation speed parameter mspd increases, the excitation time tm is reduced in three stages from, for example, 5 ms to 3 ms, and the rotation speed of the rotor 32 is adjusted in three stages. I can do it. Needless to say, the switching of the excitation time tm may be performed in two or less steps or four or more steps.
[0185]
After setting the excitation time tm, the routine proceeds to step 951, where the energization flag Xon is set to ON, and in the next step 952, the time counter Top is cleared, and this routine ends.
[0186]
If it is determined in step 934 in FIG. 37 that the current position count value Ncntop matches the target count value Actnt, the process proceeds to step 955 in FIG. 39, where the time counter Top measures the predetermined time Khld (for example, 500 ms). ) It is determined whether or not the above is achieved. As a result, if it is determined that the time measured by the time counter Top is shorter than the predetermined time Khld, the process proceeds to step 956, where the energization of all phases is turned off. ).
[0187]
Thereafter, the process proceeds to step 952 in FIG. 38, where the time counter Top is cleared and the present routine ends. By such processing, after the end of the open loop control, the processing after step 934 in FIG. 37 is not performed until a predetermined time Khld (for example, 500 ms) elapses, so that the next open loop control is not started. This is to wait until the engaging portion 23a of the detent spring 23 slides down to the bottom along the inclined side walls of the holding concave portions 24 and 25 of the detent lever 15 by the elastic force of the detent spring 23 after the end of the open loop control.
[0188]
If it is determined in step 955 of FIG. 39 that the value of the time counter Top is equal to or greater than the predetermined value Khld, the process proceeds to step 958, in which the energized phase determination value Mptnop = 0 and the rotation speed parameter mspd = 0 are set. In the next step 959, the open loop control completion flag Xopcmp is set to ON. Thereafter, the routine proceeds to step 960, where the excitation time Tm is set to 0. At the next step 961, the energizing flag Xon is set to OFF. Then, the routine proceeds to step 952 in FIG. 38, where the time counter Top is cleared and this routine is executed. To end.
[0189]
In this routine described above, the excitation method during the open loop control is set to the 1-2 phase excitation method, the first excitation at the start of the open loop control is performed by two-phase energization, and as shown in FIG. For example, for three steps from the start of the loop control, the excitation time tm of each energized phase is set to the longest time (for example, 50 ms), and the rotor 32 is surely gripped and the rotational phase of the rotor 32 and the energized phase are reliably synchronized. . Thereafter, the excitation time tm is greatly reduced, but the excitation time tm is gradually shortened as the rotation speed of the rotor 32 increases, so that the rotation speed of the rotor 32 is gradually increased.
[0190]
In order to estimate the rotational position of the rotor 32 during the open loop control, the initial position of the rotor 32 at the start of the open loop control (the initial value of the position count value Ncntop) is calculated based on the output signal of the output shaft sensor 14. Thereafter, every time the energized phase is switched, the position count value Ncntop is counted up (or down) by two, the rotational position of the rotor 32 is estimated based on the position count value Ncntop, and the position count value Ncntop and the target count are counted. At a stage where the deviation from the value Actnt has decreased to some extent, it is determined that the vehicle has entered the deceleration region, the excitation time tm is gradually increased, the rotation speed of the rotor 32 is gradually reduced, and the position count value Ncntop reaches the target value. At the time when the count value Acnt has been reached, it is determined that the rotor 32 has reached the target position, To end the Punrupu control.
[0191]
In this routine, the excitation time tm is set in accordance with the rotation speed parameter mspd. However, at least one of the rotation amount after the start of the drive, the elapsed time after the start of the drive, the rotation amount to the target position, and the load torque. The excitation time tm may be set based on the above.
[0192]
In this routine, the first excitation at the start of the open-loop control is two-phase energization. However, the first excitation may be one-phase energization. The method is not limited to the method, and a one-phase excitation method or a two-phase excitation method may be employed.
[0193]
[Suppress engine torque during open loop control]
The engine torque suppression routine during open loop control shown in FIG. 43 is started at a predetermined cycle (for example, 1 ms cycle). When this routine is started, first, in step 971, it is determined whether or not the open loop control execution flag Xopen = ON (open loop control is being executed). If Xopen = ON, the process proceeds to step 972 and the engine is started. The torque suppression flag Xtrq is set to ON, and if Xopen = OFF, the routine proceeds to step 973, where the engine torque suppression flag Xtrq is set to OFF. The information on the engine torque suppression flag Xtrq is transmitted from the ECU 41 for range switching control to an ECU (not shown) for engine control.
[0194]
When the ECU for engine control receives the information of the open-loop control execution flag Xopen = ON, the ECU controls the engine control parameters in a direction to suppress the engine torque, and performs the control by the open-loop control in a state where the engine torque is suppressed. Execute the range switching operation.
[0195]
At this time, at least one of an air system, a fuel system, and an ignition system may be used as an engine control parameter for controlling the engine torque in a direction to suppress the engine torque. The driving may be prohibited to keep the throttle opening at the current level, or the throttle opening may be closed within a range that does not hinder driving. In the case of the fuel system, the fuel injection amount may be limited within a range that does not hinder the operation, and in the case of the ignition system, the ignition timing is delayed within a range that does not hinder the operation. It is good to make it corner.
[0196]
[AB phase disconnection detection]
The AB phase disconnection detection routine shown in FIG. 44 is started by AB phase interrupt processing. When this routine is started, first, in step 501, it is determined whether or not the open loop control execution flag Xopen = ON (open loop control is being executed), and the open loop control execution flag Xopen = ON (open loop control execution). If it is (medium), the process proceeds to step 502, and it is determined whether or not the open loop control is started (immediately after the open loop control execution flag Xopen is switched from OFF to ON). As a result, if it is determined that the open loop control is to be started, the process proceeds to step 503, and the difference between the encoder count value Ncnt at the start of the open loop control and the position count value Ncntop is set as the count value deviation amount ΔNinit at the start of the open loop control. Ask.
ΔNinit = Ncnt−Ncntop
[0197]
On the other hand, if it is after the start of the open loop control (during the open loop control), "No" is determined in step 502, and the process proceeds to step 504, where the encoder count value Ncnt and the position count value during the open loop control are determined. A deviation amount ΔNerr from Ncntop is calculated.
ΔNerr = Ncnt−Ncntop
[0198]
Thereafter, the process proceeds to step 505, where it is determined whether or not the open loop control has ended (immediately after the open loop control execution flag Xopen has been switched from ON to OFF). This routine is terminated without performing the processing. If the open loop control is to be terminated, the process proceeds to step 506, where the count deviation ΔNerr at the end of the open loop control and the count deviation ΔNinit at the start of the open loop control. The absolute value of the deviation | ΔNerr−ΔNinit | is compared with the determination value to determine whether the A-phase and the B-phase of the encoder 46 are disconnected.
[0199]
For example, when the rotor 32 is normally driven by the open loop control and the A phase and the B phase of the encoder 46 are not disconnected, the count value deviation ΔNerr at the end of the open loop control and the count at the start of the open loop control The deviation | ΔNerr−ΔNinit | from the value deviation amount ΔNinit | becomes zero, but | ΔNerr−ΔNinit | does not become zero if one or both of the A phase and the B phase are disconnected. Therefore, if | ΔNerr−ΔNinit | is equal to or larger than the determination value, it is determined that the A-phase / B-phase disconnection of the encoder 46 is present, and the process proceeds to step 507, where the AB-phase disconnection detection flag Xfailab is disconnected from the A-phase / B-phase. It is set to ON, which means that there is, and this routine ends.
[0200]
On the other hand, if | ΔNerr−ΔNinit | is smaller than the determination value, it is determined that there is no disconnection of the A-phase and B-phase of the encoder 46, and the process proceeds to step 507, where the AB-phase disconnection detection flag Xfailab is disconnected from the A-phase and B-phase. In step 509, the AB phase normality determination flag Xnormab is set to ON, which means normal, and the routine ends.
[0201]
In this routine, the absolute value | ΔNerr−ΔNinit | of the deviation between the count value deviation ΔNerr at the end of the open loop control and the count value deviation ΔNinit at the start of the open loop control is calculated. The amount of change in the encoder count value Ncnt and the amount of change in the position count value Ncntop from the start to the end of the control are calculated, and the A phase of the encoder 46 is determined based on whether the absolute value of the difference between the two is greater than or equal to the determination value. It may be determined whether or not the B phase is disconnected. This is the absolute value of the difference between the change amount of the encoder count value Ncnt and the change amount of the position count value Ncntop from the start to the end of the open loop control | Ncnt change amount−Ncntop change amount | is | ΔNerr−ΔNinit | This is because
[0202]
It should be noted that the A-phase and B-phase disconnection detection of the encoder 46 can be performed only during open loop control. In other words, during the F / B control, if the A-phase and B-phase signals are not input due to the disconnection, the motor 12 is stopped to hold the energization (to output a drive pulse at the edge of the A-phase and B-phase signals). Can be determined, but it cannot be determined whether this is caused by a failure of the encoder 46.
[0203]
[Z-phase disconnection detection]
The Z-phase disconnection detection routine shown in FIG. 45 is started by the AB-phase interrupt processing, and determines whether the encoder 46 has a Z-phase disconnection as follows. When this routine is started, first, in step 511, it is determined whether or not the open loop control execution flag Xopen = ON (open loop control is being executed). If the open loop control execution flag Xopen = OFF, the process proceeds to step 511. Proceeding to 512, the Z-phase disconnection provisional detection counter Czng is cleared, and proceeding to step 518, the Z-phase signal output interval counter Czon is cleared.
[0204]
If the open-loop control execution flag Xopen = ON (open-loop control is being executed), the process proceeds to step 513, where it is determined whether or not the value of the Z-phase signal is 1 (high level). If the value is 1, the routine proceeds to step 514, in which the Z-phase normality determination flag Xnormz is set to ON meaning normal, the routine proceeds to step 518, the Z-phase signal output interval counter Czon is cleared, and the routine proceeds to step 519.
[0205]
On the other hand, if the value of the Z-phase signal is 0 (low level), the process proceeds to step 515, and the Z-phase signal output interval counter Czon for measuring the output interval of the Z-phase signal is counted up. Since this routine is started by the AB-phase interrupt processing, the Z-phase signal output interval counter Czon, like the encoder count value Ncnt, is synchronized with both rising and falling edges of the A-phase signal and the B-phase signal. Counted up.
[0206]
In the present embodiment, the output interval of the Z-phase signal (the interval between the reference rotation positions of the rotor 32) is 45 ° in the rotor rotation angle, and the edge interval between the A-phase signal and the B-phase signal is 3.75 °. When the output interval of the Z-phase signal is converted into the count value (encoder count value Ncnt) of the Z-phase signal output interval counter Czone, 45 ° ÷ 3.75 ° = 12. Therefore, if the Z-phase signal is not output even if the count value of the Z-phase signal output interval counter Czon exceeds 12, it means that the Z-phase signal has been lost.
[0207]
Therefore, in this routine, in step 516, it is determined whether or not the count value of the Z-phase signal output interval counter Czon has exceeded 24, thereby determining whether or not the Z-phase signal has been output continuously two or more times. If it is determined that the Z-phase signal is not output continuously two or more times, the process proceeds to step 517, where the Z-phase abnormality temporary detection counter Czng is counted up, and the process proceeds to step 518. The Z-phase signal output interval counter Czone is cleared, and the routine proceeds to step 519.
If it is determined in step 516 that the count value of the Z-phase signal output interval counter Czone does not exceed 24, the process proceeds to step 519.
[0208]
In step 519, it is determined whether there is a possibility of the Z-phase disconnection of the encoder 46 by determining whether the count value of the Z-phase abnormality temporary detection counter Czng is equal to or more than a determination value (for example, 3). If the count value of the phase abnormality temporary detection counter Czng is smaller than the determination value, it is determined that the Z-phase disconnection diagnosis is still being performed, and this routine ends.
[0209]
On the other hand, if the count value of the Z-phase abnormality temporary detection counter Czng is equal to or greater than the determination value, it is determined that either the Z-phase disconnection of the encoder 46 or the A-phase / B-phase signal is abnormal, and step 520 is performed. Then, it is determined whether the A phase and the B phase of the encoder 46 are normal based on whether the AB phase normality determination flag Xnormab is ON. As a result, if it is determined that the AB phase normality determination flag Xnormab = ON (the A phase and the B phase are normal), it is determined that the Z phase is broken, and the routine proceeds to step 521, where the Z phase abnormality detection flag Xfailz is set to Z. The flag is set to ON, meaning phase break (abnormal), and the routine proceeds to step 522, where the Z-phase correction inhibition flag Xzng is set to ON, and this routine ends.
[0210]
When the Z-phase correction prohibition flag Xzng is set to ON, Z-phase correction by the Z-phase correction routine of FIG. 22 is prohibited.
[0211]
[Z-phase signal error detection]
The Z-phase signal abnormality detection routine shown in FIG. 46 is started by the Z-phase interrupt processing, and determines whether there is an abnormality in the Z-phase signal of the encoder 46 as follows. When this routine is started, first, in step 531, a Z-phase signal output interval ΔNz from the previous Z-phase signal to the current Z-phase signal is calculated by the following equation.
ΔNz = | Ncnt−Ncntold |
In the above equation, Ncnt is the encoder count value when the current Z-phase signal is output, and Ncntold is the encoder count value when the previous Z-phase signal is output.
[0212]
Then, in the next step 532, it is determined whether or not the Z-phase signal output interval ΔNz is 0 or 12. If the Z-phase signal output interval ΔNz is 0, it means that the A-phase and B-phase are disconnected. If the Z-phase signal output interval ΔNz is 12, the Z-phase signal output interval ΔNz is normal. Means
[0213]
Therefore, if it is determined in step 532 that the Z-phase signal output interval ΔNz is 0 or 12, the process proceeds to step 533, in which the Z-phase normality determination flag Xnormz is set to ON indicating normality, and the process proceeds to step 538. Then, in order to calculate the next Z-phase signal output interval ΔNz, the encoder count value Ncnt at the time of the current Z-phase signal output is stored as the previous value Ncntold, and this routine ends.
[0214]
On the other hand, if the Z-phase signal output interval ΔNz is not 0 or 12, there is a possibility that the Z-phase signal is abnormal, so the process proceeds from step 532 to step 534, and the Z-phase abnormality temporary detection counter Czong is counted up. Then, in the next step 535, it is determined whether or not there is a possibility that the Z-phase signal of the encoder 46 is abnormal, based on whether or not the count value of the Z-phase abnormality temporary detection counter Czonng is equal to or more than a determination value (for example, 3). If it is determined that the count value of the Z-phase abnormality temporary detection counter Czonng is less than the determination value, it is determined that the Z-phase signal abnormality diagnosis is still being performed, and the process proceeds to step 538, where the current Z-phase signal output time is determined. The encoder count value Ncnt is stored as the previous value Ncntold, and this routine ends.
[0215]
On the other hand, if the count value of the Z-phase abnormality provisional detection counter Czonng is equal to or greater than the determination value, it is determined that the Z-phase signal of the encoder 46 is abnormal, and the process proceeds to step 536 to set the Z-phase abnormality detection flag Xfailz to abnormal. Is set to ON, and the routine proceeds to step 537, where the Z-phase correction inhibition flag Xzng is set to ON, and the Z-phase correction by the Z-phase correction routine of FIG. 22 is prohibited. Then, the process proceeds to a step 538, where the encoder count value Ncnt at the time of outputting the current Z-phase signal is stored as the previous value Ncntold, followed by terminating the present routine.
Note that each routine in FIGS. 44 to 46 plays a role as abnormality diagnosis means in claims.
[0216]
According to the present embodiment described above, the encoder count value Ncnt (the count value of the A-phase signal / B-phase signal of the encoder 46) and the position count value Ncntop (during open-loop control (fail-safe processing and recovery processing)). (Count value of the motor drive signal) to determine the presence or absence of the A-phase / B-phase disconnection of the encoder 46. Therefore, the A-phase / B-phase disconnection of the encoder 46 is detected during the open loop control. In this case, the return from the open loop control to the F / B control can be prohibited, so that unnecessary F / B control is not performed, and the rotor 32 can be rotated to the target position by the open loop control. it can.
[0219]
Furthermore, in the present embodiment, the presence or absence of the abnormality of the Z-phase signal of the encoder 46 is determined based on the Z-phase signal output interval during the open loop control. Phase correction can be prohibited, and erroneous Z-phase correction can be avoided.
[0218]
The encoder used in the present invention is not limited to the magnetic encoder 46, and may be, for example, an optical encoder or a brush encoder.
The motor used in the present invention is not limited to the SR motor 12, but may be any brushless motor that detects the rotational position of the rotor based on the count value of the output signal of the encoder and sequentially switches the energized phase of the motor. A brushless motor other than the SR motor may be used.
[0219]
The range switching device of the present embodiment is configured to switch between the two ranges of the P range and the NotP range. For example, the range switching valve and the manual valve of the automatic transmission are interlocked with the turning operation of the detent lever 15. The present invention can also be applied to a range switching device that switches the ranges P, R, N, D,.
[0220]
In addition, it goes without saying that the present invention is not limited to the range switching device but can be applied to various devices using a brushless type motor such as an SR motor as a drive source.
[Brief description of the drawings]
FIG. 1 is a perspective view of a range switching device according to an embodiment of the present invention.
FIG. 2 is a diagram illustrating a configuration of an SR motor.
FIG. 3 is a circuit diagram showing a circuit configuration for driving an SR motor.
FIG. 4 is a diagram schematically showing a configuration of an entire control system of the range switching device.
FIG. 5 is a plan view illustrating a configuration of a rotary magnet of the encoder.
FIG. 6 is a side view of the encoder.
FIG. 7A is a time chart showing an output waveform of an encoder, and FIG. 7B is a time chart showing an energized phase switching pattern;
FIG. 8 is a flowchart showing a flow of processing of an initial drive routine (part 1);
FIG. 9 is a flowchart (part 2) showing a flow of processing of an initial drive routine.
FIG. 10 is a flowchart showing the flow of processing of a P range initial drive routine.
FIG. 11 is a flowchart showing the flow of processing of a NotP range initial drive routine.
FIG. 12 is a time chart showing a control example when initial driving is performed in a P range.
FIG. 13 is a time chart showing a control example when a range switching operation is performed during initial driving;
FIG. 14 is a time chart for explaining an initial position deviation learning method when a range switching operation is performed during initial driving;
FIG. 15 is a flowchart showing the flow of processing of an encoder counter routine;
FIG. 16 is a diagram showing an example of a count-up value ΔN calculation map.
FIG. 17 is a time chart showing a relationship among a command range shift, an A-phase signal, a B-phase signal, and an encoder count value.
FIG. 18 is a flowchart showing the flow of processing of a motor F / B control routine;
FIG. 19 is a flowchart illustrating the flow of a process of an energized phase setting routine.
FIG. 20 is a diagram showing an example of a conversion table from Mptn% 12 to an energized phase.
FIG. 21 is a time chart illustrating an energization process.
FIG. 22 is a flowchart showing the flow of processing of a Z-phase correction routine.
FIG. 23 is a time chart illustrating Z-phase correction.
FIG. 24 is a time chart illustrating an example of a recovery process.
FIG. 25 is a view for explaining a method for detecting an abnormality in the A-phase signal and the B-phase signal of the encoder.
FIG. 26 is a view for explaining a method of detecting reverse rotation of the rotor during F / B control;
FIG. 27 is a time chart illustrating an example of open loop control.
FIG. 28 is a flowchart showing the flow of a recovery processing routine;
FIG. 29 is a view for explaining a method of calculating an initial value of the position count value Ncnt at the start of open loop control.
FIG. 30 is a flowchart showing the processing flow of an AB phase signal abnormality detection routine during F / B control;
FIG. 31 is a flowchart showing the flow of processing of a reverse rotation counter routine;
FIG. 32 is a flowchart showing the flow of a reverse rotation determination routine process;
FIG. 33 is a flowchart showing a processing flow of a system abnormality diagnosis routine (part 1);
FIG. 34 is a flowchart (part 2) illustrating a processing flow of a system abnormality diagnosis routine;
FIG. 35 is a flowchart (part 1) showing a flow of processing of a fail-safe processing routine;
FIG. 36 is a flowchart (part 2) illustrating a flow of processing of a fail-safe processing routine;
FIG. 37 is a flowchart (part 1) showing a flow of processing of an open loop control routine;
FIG. 38 is a flowchart (part 2) showing the flow of processing of the open loop control routine;
FIG. 39 is a flowchart (part 3) illustrating the processing flow of the open loop control routine;
FIG. 40 is a diagram showing an example of a conversion table from Mptnop% 6 to a conducting phase.
FIG. 41 is a diagram showing an example of a table for obtaining a maximum rotation speed parameter mspdmax from a battery voltage.
FIG. 42 is a diagram showing an example of a table for obtaining an excitation time tm from a rotation speed parameter mspd.
FIG. 43 is a flowchart showing the flow of processing of an engine torque suppression routine during open loop control;
FIG. 44 is a flowchart showing the flow of processing of an AB phase disconnection detection routine;
FIG. 45 is a flowchart showing the flow of processing of a Z-phase disconnection detection routine.
FIG. 46 is a flowchart showing the flow of processing of a Z-phase signal abnormality detection routine;
[Explanation of symbols]
11 Range switch mechanism, 12 SR motor, 14 Output shaft sensor, 15 Detent lever, 18 Parking rod, 20 Parking gear, 21 Lock lever, 23 Detent spring, 23a Engagement part, 24 P range holding recess, 25 NotP range holding recess, 26 ... speed reduction mechanism, 27 ... automatic transmission, 31 ... stator, 32 ... rotor, 33, 34 ... winding, 35, 36 ... motor exciting parts, 37, 38 ... Motor driver (drive circuit), 41: ECU (F / B control means, open loop control means, abnormality diagnosis means), 43: P range switch, 44: NotP range switch, 46: encoder, 47: rotary magnet, 48 ... A magnetic detection element for A-phase signal, 49 ... magnetic detection element for B-phase signal, 50 ... magnetic detection element for Z-phase signal.

Claims (7)

An encoder that outputs a pulse signal in synchronization with the rotation of a rotor of a motor that rotationally drives a control target; and detects a rotational position of the rotor based on a count value of the pulse signal of the encoder (hereinafter, referred to as an “encoder count value”). An F / B control means for executing feedback control (hereinafter referred to as "F / B control") for rotating the rotor to a target position by sequentially switching the energized phase of the motor. At
When the condition for executing the F / B control is not satisfied, the control is switched from the F / B control to the open-loop control, and a drive signal is output to the drive circuit of the motor without feeding back the information of the encoder count value to supply a current to the motor. Open-loop control means for sequentially switching between, and counting the drive signal and rotationally driving the rotor to the target position based on the count value;
A motor control device comprising: abnormality diagnosis means for comparing the encoder count value counted during the open-loop control with the count value of the drive signal to determine whether the encoder is abnormal. 2. The motor control device according to claim 1, wherein the abnormality diagnosis unit prohibits a return from the open loop control to the F / B control when detecting an abnormality of the encoder. 3. The encoder outputs an A-phase signal and a B-phase signal having a predetermined phase difference in synchronization with the rotation of the rotor, and outputs a Z-phase signal at a reference rotation position of the rotor,
The F / B control means counts the edges of the A-phase signal and the B-phase signal during the F / B control, sequentially switches the energizing phase based on the encoder count value, and changes the Z-phase signal. The presence or absence of a difference between the rotational position of the rotor and the energized phase is determined based on the difference. If there is a difference, the difference is corrected (hereinafter, this correction is referred to as “Z-phase correction”).
The abnormality diagnosis means compares the encoder count value counted during the open loop control with the count value of the drive signal to determine whether there is an abnormality in the A-phase signal and the B-phase signal of the encoder. The motor control device according to claim 1 or 2, wherein The abnormality diagnosis means determines whether or not there is an abnormality in the Z-phase signal by determining an output interval of the Z-phase signal based on the encoder count value when it determines that the A-phase signal and the B-phase signal of the encoder are normal. The motor control device according to claim 3, wherein is determined. The motor control device according to claim 4, wherein the abnormality diagnosis unit inhibits the Z-phase correction when detecting an abnormality in the Z-phase signal. The motor control device according to claim 1, wherein the motor is a switched reluctance motor. 7. The motor control device according to claim 1, wherein the control target is a range switching device that switches between a parking range of the vehicle and another range.

Images