JP4158243B2 - Control device for electric motor for electric power steering - Google Patents

Description

【００３４】
ここで、ｄ（ｋ）は第ｋ周期のＰＷＭデューティ値、ｅ（ｋ）は第ｋ周期の電流偏差であり、ａ（ｋ），ｂ（ｋ）は予め決められた定数である。
その後、上記演算によって得たＰＷＭデューティ値ｄ（ｎ）に上下限リミッタ処理を施して、最終的なＰＷＭデューティ値とする。step１９０では、その最終的なＰＷＭデューティ値を駆動回路３６に出力してモータ電流の制御を行う。
【００３５】
このように、本第２実施例の場合には、電流偏差に基づいてフィードバック駆動指令値を演算し、そのフィードバック駆動指令値に基づいて電動モータ２３を通電駆動しているため、高精度のモータ制御を実現できる。特に電動パワーステアリングに用いることが前提であるため、熱等の環境変化や経年変化を考慮してフィードバック制御をすることが好ましいと考えられる。また、大量生産する場合は、出荷前に精密な調整作業を施しておくことは現実的に難しいため、特に大量生産を前提とした車両に搭載する電動パワーステアリング用という観点からは、フィードバック制御によって対応する方が好ましいと言える。
【００３６】
［第３実施例］
ところで、マイクロコンピュータ３３にて実モータ電流を所定の電流値に一致させる制御を行うには、演算の負荷量から考えて、制御周期は１ｍｓ程度以上であることが望ましい。これに対して、標準的に行われるデジタル制御器の設計手法は、所望の性能を持つアナログフィードバック制御器の伝達関数を双一次変換して、デジタル制御則を導出し、それをマイクロコンピューターにて演算させる手法である。この手法により、実モータ電流を所定の電流値に一致させる制御を行うには、制御周期は２００マイクロ秒（μｓ）程度以下である必要がある。これより遅い制御周期で制御を行うと、実モータ電流が振動的になり、操舵フィーリングが悪化するという不具合がある。これは、電動パワーステアリングに用いられる電動モータの時定数（＝インダクタンス（Ｌ）／抵抗値（Ｒ））が通常は数ｍｓ程度であるのに対し、制御周期が１ｍｓに近づいてくると、ゼロオーダーホールドの影響が無視できなくなるためである。
【００３７】
本第３実施例は、ゼロオーダーホールドの影響を小さくし、非振動的な電流制御系を実現するための実施例である。
［第３実施例の第１態様］
図７に示す制御演算ブロックは、上述した第２実施例の説明のために用いた図５の制御演算ブロック中の「モータ電流フィードバックループ」に相当する部分を取り出したものである。図７に示すＣｄ（ｚ）は、上記［式１］に対応するパルス伝達関数であり、ｚは遅延演算子を示す。また、ＺＯＨはゼロオーダーホールドであり、Ｐ（ｓ）は電動モータ２３と駆動回路部分（モータ駆動回路３６，パワートランジスタ３７）を表す連続時間の伝達関数で、ｓはラプラス演算子である。
【００３８】
本第１態様では、図７の制御演算ブロック中のモータ電流フィードバック制御器が、カットオフ周波数（＝抵抗値（Ｒ）／インダクタンス（Ｌ））よりも高域の周波数成分を除去するフィルタ処理機能を有することを特徴としている。これは、制御周期（サンプリング時間）が、電動モータ２３の時定数（Ｌ／Ｒ）に比べ無視できない大きさになった場合に特に効果がある。
【００３９】
そのモータ電流フィードバック制御器の構成に関して、パルス伝達関数Ｃｄ（ｚ）の構成につき、実例を上げて具体的に説明する。
電動モータ２３として、インダクタンス（Ｌ）＝８３．４５μＨ，抵抗（Ｒ）＝５３．９５ｍΩのＤＣモータを想定し、それに対して、図７に示すようなモータ電流フィードバック制御系を構成した場合を考える。なお、制御周期を１ｍｓとする。この場合のモータ時定数はＬ／Ｒ＝１．５ｍｓとなり、制御周期１ｍｓが無視できない大きさとなる。
【００４０】
このように制御対象の時定数に比べて制御周期が無視できない場合に、ゼロオーダーホールドの影響を考慮する手法として、例えば文献「デジタル制御理論入門」（荒木光彦著，朝倉書店，Ｐ１３７〜Ｐ１３９）に記載の公知技術がある。以下、その要点を述べる。
【００４１】
図８において、Ｆｄ（ｚ）はフィードバック制御器のパルス伝達関数、ＺＯＨはゼロオーダーホールド、Ｇｃ（ｓ）は制御対象の連続時間伝達関数を示す。この図８に示すようなデジタル制御器において、連続時間の制御系を制御する場合の安定性を論じるのに、まず、下記の［式２］で示すＧｄ（ｚ）を考え、
【００４２】
【数２】

【００４３】
さらに、下記の［式３］で示すｗを導入して、
【００４４】
【数３】

【００４５】
連続時間伝達関数Ｇｃ（ｗ）を下記の［式４］のように表す。
【００４６】
【数４】

【００４７】
この連続時間伝達関数Ｇｃ（ｗ）で表わされた制御対象に対して、図９に示すような制御系を構成する。連続時間伝達関数Ｆｃ（ｗ）で表わされた制御器によって実現される安定余裕は、下記［式５］のように定義される連続時間伝達関数Ｆｄ（ｚ）にて図８の制御系を構成したときの安定余裕と一致する。これが上記文献に記載された公知技術の内容である。
【００４８】
【数５】

【００４９】
すなわち、図８に示すゼロオーダーホールド（ＺＯＨ）を含む制御系に対し、ゼロオーダーホールドを考慮して所望の安定の安定余裕を持つデジタル制御器を得るには、まず図９に示す制御系において所望の安定の安定余裕をもつように連続時間伝達関数Ｆｃ（ｗ）を構成し、その後、上記［式５］にしたがって連続時間伝達関数Ｆｄ（ｚ）を構成すれば良い。
【００５０】
ただし、特定の制御対象に対して適切な連続時間伝達関数Ｆｃ（ｗ）を如何に構成するかは、公知の技術ではない。本第１態様では、このような公知技術を前提としながら、特定の制御対象として電動パワーステアリング用の電動モータ２３を考え、この電動モータ２３に対する適切な連続時間伝達関数Ｆｃ（ｗ）を構成したものである。
【００５１】
実例として取り上げたＤＣモータ（Ｌ＝８３．４５μＨ，Ｒ＝５３．９５ｍΩ）の、等価的な印加電圧からモータ電流までの伝達関数Ｐ（ｓ）は、下記の［式６］に示すようになる。
Ｐ（ｓ）＝１／（Ｌ＊ｓ＋Ｒ） …［式６］
上記［式２］，［式３］，［式４］にしたがって、伝達関数Ｐ（ｓ）に対し、図９に示したＧｃ（ｗ）に相当する連続時間伝達関数Ｐｃ（ｗ）を求めると、下記の［式７］に示すようになる。
【００５２】
【数６】

【００５３】
この連続時間伝達関数Ｐｃ（ｗ）のボード線図を図１０中に（Ａ）で示す。図１０より、１次遅れであったＰ（ｓ）がゼロオーダーホールドを考慮すると、厳密にはＰｃ（ｗ）についてのゲイン交差周波数は定義できないが、位相の変化を見た場合、高周波側で位相が１８０度遅れて位相余裕の少ない系になっていることが判る。
【００５４】
本第１態様では、この制御対象に対し、図９のＦｃ（ｗ）にあたる制御器として下記の［式８］，［式９］，［式１０］で示す伝達関数を用いている。
【００５５】
【数７】

【００５６】
そのボード線図を図１０中に（Ｂ）で示す。この内、伝達関数Ｆｃ１（ｗ）は、モータ電流フィードバック制御器がカットオフ周波数（Ｒ／Ｌ）よりも高域の周波数成分を除去する要素として１次遅れを用いている。
図１０中に（ｃ）で示した一巡伝達関数Ｆｃ（ｗ）＊Ｐｃ（ｗ）のボード線図から判るように、Ｆｃ１（ｗ）の効果で、一巡伝達関数Ｆｃ（ｗ）＊Ｐｃ（ｗ）のカットオフ周波数が低域に移動し、安定余裕が増していることが判る。また、Ｆｃ２（ｗ）はＦｃ１（ｗ）による位相遅れを減らす効果がある。
【００５７】
ここで、上記［式５］にしたがって、図７におけるモータ電流フィードバック制御器のパルス伝達関数Ｃｄ（ｚ）を求めると、下記［式１１］に示すようになる。
【００５８】
【数８】

【００５９】
このパルス伝達関数Ｃｄ（ｚ）を用いて、制御周期１ｍｓにて上述したＤＣモータ（Ｌ＝８３．４５μＨ，Ｒ＝５３．９５ｍΩ）の電流フィードバック制御を行なった例として、目標電流をＯＡから１０Ａにステップ状に変化させた時の応答波形を図１１中に（Ａ）で示す。なお、図１１中の（Ｂ）は、高域周波数成分の除去要素なしで単純な比例制御を行った場合の応答波形である。図１１中の（Ａ）が（Ｂ）に比べて、応答が非振動的で且つ定常偏差も小さいことが判る。
【００６０】
［第３実施例の第２態様］
上述の第１態様では遅れ要素を用いたが、この第２態様では積分要素を用いる。つまり、図９中に示した連続時間伝達関数Ｆｃ（ｗ）に代えて、下記の[式１２］に示すものを用いる。
【００６１】
【数９】

【００６２】
その場合のボード線図を図１２中に（Ｂ）で示す。この［式１２］で示される連続時間伝達関数Ｆｃ（ｗ）は、モータ電流フィードバック制御器がカットオフ周波数（Ｌ／Ｒ）より高域の周波数成分を除去する要素として、積分要素が用いられている。図１２中に（Ｃ）で示した一巡伝達関数Ｆｃ（ｗ）＊Ｐｃ（ｗ）のボード線図から判るように、連続時間伝達関数Ｆｃ（ｗ）の効果で、一巡伝達関数Ｆｃ（ｗ）＊Ｐｃ（ｗ）のカットオフ周波数が低域に移動し、安定余裕が増していることが判る。
【００６３】
また、この［式１２］で示される連続時間伝達関数Ｆｃ（ｗ）を用いて上記［式５］で定義される連続時間伝達関数Ｆｄ（ｚ）を求め、それにしたがって、図７におけるモータ電流フィードバック制御器のパルス伝達関数Ｃｄ（ｚ）を求めると、下記［式１３］に示すようになる。
【００６４】
【数１０】

【００６５】
このパルス伝達関数Ｃｄ（ｚ）を用いて、制御周期１ｍｓにて上述したＤＣモータ（Ｌ＝８３．４５μＨ，Ｒ＝５３．９５ｍΩ）の電流フィードバック制御を行なった例として、目標電流をＯＡから１０Ａにステップ状に変化させた時の応答波形を図１３中に（Ａ）で示す。
【００６６】
この場合も、図１１中に（Ａ）で示した「遅れ要素を用いた場合」と同様に、非振動的な応答であることが判る。そして、この場合には、積分要素の効果によって、定常偏差が０となっていることも判る。
なお、上述した遅れ要素を用いた第１態様（式８に示す伝達関数参照）及び積分要素を用いた第２態様（式１２に示す伝達関数参照）いずれの場合も、遅いサンプリング時間で、非振動的な電流応答を実現するために、電動モータ２３のカットオフ周波数より高域の周波数成分を除去する要素を有していることが本質的に寄与している。
【００６７】
［第２態様の場合の工夫］
ところで、この第２態様のように積分要素を用いる場合には、次の点を考慮すると、演算されたフィードバック駆動指令値が最大であっても、実モータ電流が目標モータ電流に達しない場合には、偏差積分を増大させる方向の積分動作を停止することが好ましい。これは次の理由からである。
【００６８】
電動パワーステアリングにおいて急操舵をするような状況で偏差積分動作を停止しない場合には、次のような不都合が生じる、つまり、ＤＣモータが使用された電動モータ２３のモータ回転数に比例した逆起電力が発生し、急操舵するような場合には、逆起電力は大きな値となるので、ＰＷＭデューティが１００％であっても、なお実モータ電流が目標モータ電流よりも低い状態が存在する（図１４中のＡ−Ｂ間参照）。ＰＷＭデューティ値は、偏差の積分値と偏差の線形結合値として与えられるが、図１４中のＤ−Ｅ間では１００％にリミットされる。このような場合に、目標モータ電流と実モータ電流の偏差を積分する機能が動作していると、偏差の積分値図１４中のＤ→Ｅの経路を辿ることとなる。図１４中のＢ点においては逆起電力の影響が減り、実モータ電流が目標モータ電流に追いついていても、偏差積分は図１４中のＥ→Ｆの経路を辿るので、ＰＷＭデューティは１００％のままである。この結果、図１４中のＢ−Ｃ間では実モータ電流のオーバーシュートが発生するため、操舵感が悪化する。
【００６９】
そこで、図１５のフローチャートに示すような制御を実行して、必要な場合には偏差積分動作を停止する。
図１５のフローチャートの説明の前に、その処理に関連する説明をしておく。上記［式１３］に示したモータ電流フィードバック制御器のパルス伝達関数Ｃｄ（ｚ）は、下記［式１４］に示すように変形できる。
【００７０】
【数１１】

【００７１】
すなわち、偏差比例に対応する部分［０．０５］と、偏差積分に対応する部分［０．０５／（１−Ｚ^-1）］とに分けられる。この偏差比例対応部分と偏差積分対応部分それぞれについての演算が図１５においてなされる。
図１５のフローチャートにおけるstep１１０〜１８０及びstep１９０は、第２実施例の場合の図６に示した内容と同じであり、図６におけるstep１８５に代えて、step１８１〜tep１８４を実行する点が異なるだけである。そこで、主にその部分を説明する。
【００７２】
step１８０で目標モータ電流と実モータ電流の差（電流偏差）を演算した後、続くstep１８１では、前周期のＰＷＭデューティ値が最大値となっているか否かがチェック（飽和チェック）される。そのチェック結果に応じて、step１８２でのＰＷＭデューティ値ｄ（ｎ）の偏差積分演算は変わってくる。
【００７３】
すなわち、step１８２では、前周期のＰＷＭデューティ値が最大値となっている場合は、積分値が増加する方向の積分演算（上記［式１４］に示した偏差積分対応部分［０．０５／（１−Ｚ^-1）］）は行わない。なお、積分値が減少する方向の積分演算は行う。一方、前周期のＰＷＭデューティ値が最大値となっていない場合には、そのまま積分演算（［式１４］中の偏差積分対応部分）を行う。
【００７４】
続くstep１８３では、［式１４］に示した偏差比例対応部分［０．０５］の演算を実行する。
そして、step１８４では、上記step１８２とstep１８３の結果を加算し、必要なガード処理を施して、最終的なＰＷＭデューティ値とする。
【００７５】
このような処理を実行することで、ＰＷＭデューティが１００％に達した図１４中のＤ点で、偏差積分を増大させる方向の積分動作が停止することとなる。そのため、偏差積分値は図１４中のＤ→Ｇの経路を辿る。したがって、図１４中のＢ点では、逆起電力の影響が減り、実モータ電流が目標モータ電流に追いつくと、偏差積分は、図１４中のＧ→Ｈの経路を辿るので、ＰＷＭデューティは直ちに減少し始め、実モータ電流のオーバーシュートが発生せず、良好な操舵感が得られる。
【００７６】
［第４実施例］
第４実施例は、上述した第２あるいは第３実施例におけるモータ電流フィードバック制御器と並行に、目標電流に基づいてモータ駆動回路３６に対する指令値をフィードフォワード項として演算する、モータ電流フィードフォワード制御器を設けたものである。本第４実施例の電動パワーステアリング制御装置の制御演算ブロック（図１６）及び制御演算ブロックに対応する制御フローチャート（図１７）を参照して、動作説明を行う。なお、図１５のフローチャートに示す演算は、所定の制御周期ごとにマイクロコンピュータ３３内で実行されるものである。また、上述した第１実施例の場合の電動パワーステアリング制御装置を中心に示す車両構成（図１）及びＥＣＵ３０の内部構成を示す処理ブロック図（図２）は、この第２実施例の場合も同様であるので重複説明はしない。
【００７７】
図１７のフローチャートにおけるstep２１０〜２７０及びstep２９０は、第１実施例の場合の図４に示したstep１０〜７０及びstep９０と同じ内容であるので説明は省略する。
step２８０では、上述した第２実施例の場合のstep１８０と同様に、目標モータ電流と実モータ電流の差（電流偏差）を演算する。そして、続くstep２８５では、下記の［式１５］に示す漸化式に従って、ＰＷＭデューティフィードバック値ｄｆｂ（ｎ）を演算する。
【００７８】
【数１２】

【００７９】
ここで、ｄｆｂ（ｋ）は第ｋ周期のＰＷＭデューティフィードバック値、ｅ（ｋ）は第ｋ周期の電流偏差であり、ａ（ｋ），ｂ（ｋ）は予め決められた定数である。
続くstep２８６では、下記の［式１６］に示す漸化式に従って、ＰＷＭデューティフィードフォワード値ｄｆｆ（ｎ）を演算する。
【００８０】
ｄｆｆ（ｎ）＝ｆ（Ｒ＊Ｉｒｅｆ（ｎ）） …［式１６］
ここで、ＲはＤＣモータの抵抗値、ｌｒｅｆ（ｎ）は第ｎ周期の目標モータ電流、ｆ（……）は、予め設定されたマップである。
そしうて、step２８７では、下記の［式１７］に示すようにＰＷＭデューティ最終値ｄ（ｎ）を演算する。
【００８１】
ｄ（ｎ）＝ｄｆｆ（ｎ）＋ｄｆｂ（ｎ） …［式１７］
その後、上記演算によって得たＰＷＭデューティ値ｄ（ｎ）に上下限リミッタ処理を施して、最終的なＰＷＭデューティ値とする。step２９０では、その最終的なＰＷＭデューティ値を駆動回路３６に出力してモータ電流の制御を行う。
【００８２】
このように、本第４実施例の場合には、ＰＷＭデューティフィードバック値ｄｆｂ（ｎ）とＰＷＭデューティフィードフォワード値ｄｆｆ（ｎ）とに基づいてＰＷＭデューティ最終値ｄ（ｎ）を演算し、その値に基づいてモータ電流の制御を行っているため、モータ電流の定常偏差の低減、電流立ち上がりの応答改善を図ることができ、操舵フィーリングをさらに向上させることとなる。
【００８３】
［その他］
（１）上述した実施例では、電動モータ２３が減速機２２を介してシャフトを回動させることで補助操舵する構成を前提としたが、それ以外にも、例えばラック１３を電磁力等によって直接駆動させる「ラック同軸式」の構成を前提としても同様に適用可能である。
【００８４】
（２）上述した第３実施例の第１態様では、１次遅れ要素を持つ伝達関数を用いたが、１次遅れ要素に限られるものではなく、２次以上の遅れ要素を持つ伝達関数を用いても同様に実現できる。
【図面の簡単な説明】
【図１】 本発明が適用された実施例の電動パワーステアリング制御装置を中心に示す車両構成図である。
【図２】 実施例のＥＣＵの内部構成を示す処理ブロック図である。
【図３】 第１実施例の電動パワーステアリング制御装置の制御演算ブロック図である。
【図４】 第１実施例の制御演算ブロックに対応する制御フローチャートである。
【図５】 第２実施例の電動パワーステアリング制御装置の制御演算ブロック図である。
【図６】 第２実施例の制御演算ブロックに対応する制御フローチャートである。
【図７】 図５の制御演算ブロック中の「モータ電流フィードバックループ」に相当する部分を取り出したものである。
【図８】 第３実施例の第１態様を示すための制御演算ブロック図である。
【図９】 第３実施例の第１態様を示すための制御演算ブロック図である。
【図１０】 第３実施例の第１態様における伝達関数のボード線図である。
【図１１】 第３実施例の第１態様におけるパルス伝達関数を用いてＤＣモータの電流Ｆ／Ｂ制御を行なった場合の、目標電流をステップ状に変化させた時の応答波形の説明図である。
【図１２】 第３実施例の第２態様における伝達関数のボード線図である。
【図１３】 第３実施例の第２態様におけるパルス伝達関数を用いてＤＣモータの電流Ｆ／Ｂ制御を行なった場合の、目標電流をステップ状に変化させた時の応答波形の説明図である。
【図１４】 第３実施例の第２態様に関して、電動パワーステアリングにおいて急操舵をする状況で偏差積分動作を停止しない場合と停止した場合との比較等を示す説明図である。
【図１５】 第３実施例の第２態様に関して、偏差積分動作を停止する場合の制御を示すフローチャートである。
【図１６】 第４実施例の電動パワーステアリング制御装置の制御演算ブロック図である。
【図１７】 第４実施例の制御演算ブロックに対応する制御フローチャートである。
【符号の説明】
１１…ハンドル １２…シャフト
１３…ラック １４…ピニオンギヤ
１５…タイロッド １６…夕イヤ
２１…操舵トルクセンサ ２２…減速機
２３…電動モータ ２４…操舵角センサ
２５…モータ回転角センサ ２７…モータ電流センサ
３０…マイクロコンピュータ ３１，３２…ＬＰＦ
３４…Ａ／Ｄ変換器 ３５…ＣＰＵ
３６…モータ駆動回路 ３７…パワートランジスタ
３８，３９，４０…波形整形部 ５０…車速センサ[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a control device for an electric motor used for electric power steering of a vehicle that contributes to reduction of steering force of a driver and improvement of steering feeling.
[0002]
[Prior art]
In electric power steering, the direction of the motor and the rotational torque of the electric motor are controlled according to the assist signal calculated based on the output signal of the torsion torque sensor that detects the torsion torque of the steering system. Steering load is reduced. For example, in the electric power steering disclosed in Japanese Examined Patent Publication No. 6-24942, in order to generate a necessary rotational torque, control is performed so that the actual motor current matches a predetermined current value, and an electric power steering excellent in responsiveness is configured. If so.
[0003]
[Problems to be solved by the invention]
However, feedback control by an analog circuit using an operational amplifier or the like is used as a device for performing control for making the actual motor current coincide with a predetermined current value, and the circuit is complicated and expensive.
[0004]
In view of the above circumstances, an object of the present invention is to provide an electric motor control device capable of exhibiting sufficient performance as an electric power steering without using a complicated and expensive analog circuit.
[0005]
[Means for Solving the Problems and Effects of the Invention]
The electric power steering electric motor control apparatus according to claim 1, wherein the command value calculation means calculates the drive command value based on the steering torque in the steering mechanism detected by the steering torque detection means. Then, based on the calculated drive command value, the motor drive means energizes and drives the electric motor for assisting the steering mechanism.
[0006]
Here, the motor current detection means detects the current flowing through the electric motor, and the detected value is converted into digital data that can be processed by the microcomputer by the AD conversion means. On the other hand, the target motor current calculation means calculates the target motor current for the electric motor as digital data based on the steering torque detected by the steering torque detection means. Then, the calculation means converts the drive command value for the motor drive means into the digital data by the AD conversion means and the target value so that the current (actual motor current) flowing through the electric motor matches the target motor current. Based on the motor current value, calculation is performed by a microcomputer at predetermined time intervals.
[0007]
Therefore, according to this electric motor control device, the electric motor can be controlled without using a complicated and expensive analog circuit. The fact that this analog circuit is not used leads to the following effect. In other words, when configuring an analog circuit, for example, since there is a need for parts such as capacitors that vary among individuals, the performance of the device as a finished product may also vary. It is troublesome. If configured like the control device of the present invention, such a variation does not occur, and an adjustment operation becomes unnecessary. When an analog circuit is used, it is necessary to set and manufacture an analog circuit corresponding to the type of object on which the electric power steering is mounted. For example, an electric motor for electric power steering for a light vehicle and a large vehicle needs to cope with a difference in current used. If configured like the control device of the present invention, it is only necessary to change the program in accordance with the type of the mounting target, and the hardware configuration can be made common.
[0008]
  When controlling the electric motor, ActingIt is conceivable that the calculating means calculates a deviation between the target motor current and the actual motor current, and calculates a feedback drive command value for the motor driving means based on the calculated current deviation. Then, based on the calculated feedback drive command value, the motor drive means drives the electric motor to be energized. In this way, highly accurate motor control can be realized. In particular, since it is premised on use for electric power steering, it is considered preferable to perform feedback control in consideration of environmental changes such as heat and aging. In addition, in mass production, it is practically difficult to carry out precise adjustment work before shipment, so feedback control is particularly effective from the viewpoint of electric power steering installed in vehicles that are premised on mass production. It can be said that it is preferable to respond.
[0009]
By the way, in order to control the actual motor current to a predetermined current value by the microcomputer, it is desirable that the control cycle (sampling time) is about 1 ms or more in view of the calculation load. In contrast, the standard digital controller design method derives a digital control law by bilinearly transforming the transfer function of an analog feedback controller having a desired performance, and this is performed by a microcomputer. This is a method of calculating. It has been confirmed by the present applicant that the control cycle needs to be about 200 microseconds (μs) or less in order to perform control for matching the actual motor current to a predetermined current value by this method. If the control is performed at a slower control cycle, the actual motor current becomes oscillating and the steering feeling is deteriorated. This is due to the following reason.
[0010]
In the digital control system, the command value for the motor drive circuit takes a constant value during each sampling. This is so-called zero order hold (zero order hold). The actual motor current becomes oscillating as the control cycle is delayed. Usually, the time constant (= inductance (L) / resistance (R)) of an electric motor used for electric power steering is about several ms. On the other hand, when the control period approaches 1 ms, the influence of this zero order hold cannot be ignored.
[0011]
  ThereforeTheIn order to reduce the influence of low-order hold and realize a non-vibrating current control system, the following configuration was adopted. That is, the cutoff frequency (= resistance value (R) / inductance / (L)) when the calculation means for calculating the feedback drive command value approximates the response of the current to the applied voltage of the electric motor as a first order delay. The filter processing for removing the higher frequency components is executed. In this way, by removing frequency components higher than the cutoff frequency, the control band of the feedback controller can be narrowed to the lower frequency side than the band of the analog feedback controller. Thereby, the influence of zero order hold can be reduced and a non-vibrating current control system can be realized.
[0012]
With such a configuration, the response of the actual motor current is generally slower than that of the analog feedback controller, but this influence does not cause a problem in the motor current control of the electric power steering for the following reason. That is, (1) the control band of the electric motor control of the electric power steering for which the present invention is an object is practically sufficient if it is about 100 Hz, unlike a servo system such as a machine tool, and (2) electric power steering. In general, the cutoff frequency of a motor that is generally used is several hundred Hz. Therefore, even if a high frequency band higher than the cutoff frequency of the motor is removed, a practically sufficient control band for electric motor control can be obtained. Based on the applicant's knowledge.
[0013]
  Note that, as described above, when configuring the calculation means for executing the filter processing for removing the high frequency component, the following can be performed. That is, the claim2As shown in FIG. 2, the feedback drive command value is calculated using a transfer function having an element that removes a frequency component higher than the cutoff frequency. In this case, the claim3Or a transfer function having a delay element (first-order delay element, second-order delay element, etc.) as shown in FIG.4As shown in FIG. 5, the transfer function having an element for integrating the deviation between the target motor current and the actual motor current may be used.
[0014]
  Among these, if the calculation means is configured using a transfer function having an integral element, the steady-state deviation when the target motor current changes stepwise can be eliminated.5As shown in FIG. 5, even if the calculated feedback drive command value is the maximum, if the actual motor current does not reach the target motor current, it is preferable to stop the integration operation in the direction in which the deviation integration is increased. This is for the following reason.
[0015]
For example, consider a case where sudden steering is performed in electric power steering. When the deviation integration operation is not stopped, the following inconvenience occurs. That is, when a DC motor is used as the electric motor, a back electromotive force proportional to the motor rotational speed is generated. When the steering is suddenly performed, the back electromotive force becomes a large value. For example, when the drive command value is output by PWM, the actual motor current is still the target even if the PWM duty is 100%. There is a state lower than the motor current (see between A and B in FIG. 14). The PWM duty value is given as an integral value of deviation and a linear combination value of deviation, but is limited to 100% between DE in FIG. In such a case, if the function of integrating the deviation between the target motor current and the actual motor current is operating, the integrated value of the deviation follows the path D → E in FIG. At point B in FIG. 14, the influence of the counter electromotive force decreases, and even if the actual motor current catches up with the target motor current, the deviation integral follows the path E → F in FIG. Remains. As a result, an overshoot of the actual motor current occurs between B and C in FIG.
[0016]
  In contrast, the claims5As shown in FIG. 14, when the deviation integration operation is stopped, the integration operation in the direction of increasing the deviation integration is stopped at the point D in FIG. 14 where the PWM duty has reached 100%. Follow the middle DG path. Therefore, at the point B in FIG. 14, the influence of the counter electromotive force is reduced, and when the actual motor current catches up with the target motor current, the deviation integration follows the G → H path in FIG. The motor starts to decrease, the actual motor current does not overshoot, and a good steering feeling can be obtained.
[0017]
  According to a sixth aspect of the present invention, in the control device for an electric motor for an electric power steering according to any one of the first to fifth aspects, the calculation means is a feedforward drive command value for the motor drive means based on the target motor current. May be calculated and added to the feedback drive command value to obtain the final drive command value for the motor drive means. In this way, it is possible to reduce the steady deviation of the motor current and improve the response of the current rise, and further improve the steering feeling.
  Next, in the control device for an electric motor for electric power steering according to claim 7, the calculating means reduces the phase delay due to the delay element as an element for removing frequency components higher than the cutoff frequency, and the phase delay due to the delay element. A transfer function having elements is used.
  Furthermore, an electric power steering electric motor control device according to an eighth aspect of the present invention provides the above [formula a], [formula b] as a transfer function having the delay element and an element for reducing a phase delay due to the delay element. ], The transfer function shown by [Formula c] is used.
[0018]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments to which the present invention is applied will be described below with reference to the drawings. Needless to say, the embodiments of the present invention are not limited to the following examples, and can take various forms as long as they belong to the technical scope of the present invention.
[0019]
[First embodiment]
FIG. 1 is a vehicle configuration diagram mainly showing an electric power steering control device according to an embodiment to which the above-described invention is applied.
A shaft (steering shaft) 12 is connected to a handle (steering wheel) 11 that is steered by a vehicle driver, and a rack 13 and a pinion gear 14 corresponding to a “steering mechanism” are connected to the shaft 12. . When the shaft 12 rotates according to the steering of the handle 11, the rotation angle of the shaft 12 becomes the amount of movement of the rack 13. Tie rods 15 are provided at both ends of the rack 13, and the evening ears 16 are steered left and right by the tie rods 15.
[0020]
The shaft 12 is provided with a steering torque sensor 21 and a steering angle sensor 24 corresponding to “steering torque detection means”. The steering torque sensor 21 responds to torque generated when the vehicle driver steers the steering wheel. A signal (steering torque signal) is output, and a steering angle generated when the steering wheel is steered from the steering angle sensor 24 is output.
[0021]
An electric motor 23 is attached between the steering torque sensor 21 and the steering angle sensor 24, the rack 13 and the pinion gear 14 via a speed reducer 22. The speed reducer 22 is a well-known one composed of a worm and a worm wheel. If the electric motor 23 is energized and driven, the force when the handle 11 is rotated is reduced.
[0022]
In addition, the electric motor 23 is provided with a motor rotation angle sensor 25 and a motor current sensor 27. The ECU 30 that performs energization control on the electric motor 23 is configured to control the steering torque signal from the steering torque sensor 21 and the steering. Based on the steering angle from the angle sensor 24, the motor rotation angle from the motor rotation angle sensor 25, the motor current from the motor current sensor 27, and the vehicle speed signal from the vehicle speed sensor 50 that detects the vehicle speed, Control the current to be energized.
[0023]
Next, the internal configuration of the ECU 30 will be described with reference to the process block diagram of FIG.
In order to digitally calculate the electric motor control based on the signal output from the steering torque sensor 21, the ECU 30 first passes the detection signal through an analog circuit low pass filter (LPF) 31, and then the microcomputer 33. A / D conversion is performed by an A / D converter 34 provided therein, and the data is taken into the CPU 35. Here, the LPF 31 is configured as a fill filter having a cutoff frequency equal to or lower than the Nyquist frequency (1/2 of the sampling frequency) in order to remove aliasing noise during A / D conversion.
[0024]
Similarly, the signal output from the motor current sensor 27 is passed through the LPF 32, converted into digital data by the A / D converter 34 in the microcomputer 33, and then taken into the CPU 35.
On the other hand, the signal from the vehicle speed sensor 50 is generally a signal whose pulse frequency changes according to the vehicle speed. This signal is waveform-shaped by a waveform shaping unit 38 inside the ECU 30 and input to a port of the CPU 35. The CPU 35 obtains a signal corresponding to the vehicle speed by measuring the cycle of the vehicle speed pulse. Similarly, signals from the motor rotation angle sensor 25 and the steering angle sensor 24 are also input to the port of the CPU 35 via the waveform shaping sections 39 and 40 provided correspondingly.
[0025]
  The microcomputer 33 outputs a PWM (pulse width modulation) duty to the drive circuit 36. The drive circuit 36 uses this PWMBased on the signal, the power transistor 37 is switched so that the current of the electric motor 23 follows. Next, the operation will be described with reference to a control calculation block (FIG. 3) and a control flowchart (FIG. 4) corresponding to the control calculation block of the electric power steering control apparatus of the first embodiment. Note that the calculation shown in the flowchart of FIG. 4 is executed in the microcomputer 33 every predetermined control cycle.
[0026]
In the first step 10, the detection signal (steering torque signal) from the steering torque sensor 21 is converted into digital data by the A / D converter 34 and is taken into the CPU 35.
The subsequent steps 20 to 50 correspond to the target motor torque calculation in the control calculation block of FIG. First, in step 20, the basic assist torque is calculated based on the steering torque data. In the subsequent step 30, the basic assist torque is corrected so that the assist torque decreases as the vehicle speed increases.
[0027]
In step 40, based on the motor rotation angle and the steering angle, the assist torque is corrected to improve the return of the steering wheel. In step 50, the assist torque is corrected based on the motor rotation angular velocity and the steering angular velocity in order to improve the convergence when the handle is returned. What is obtained in this way becomes the target motor torque.
[0028]
In subsequent step 60, a motor current corresponding to the target motor torque calculated in steps 20 to 50 described above is calculated. In step 70, the motor current signal is A / D converted, and is converted into digital data and taken into the CPU 35.
In the subsequent step 80, a command value (PWM duty value) for the drive circuit is calculated based on the actual motor current value and the target motor current value. In the final step 90, the calculated PWM duty value is output to the drive circuit 36, and the motor current is controlled.
[0029]
As described above, the drive circuit 36 performs switching of the power transistor 37 based on the PWM signal so that the current of the electric motor 23 follows.
As described above, according to the electric power steering control apparatus of the first embodiment, the actual motor current detected by the motor current sensor 27 is converted into digital data by the AD converter 34 and is taken into the CPU 35. Based on the motor current and the target current value, the CPU 35 calculates a command value for the motor drive circuit 36. As a result, the motor current can be controlled without using a complicated and expensive analog circuit.
[0030]
The fact that the analog circuit is not used in this way leads to the following effect. In other words, when configuring an analog circuit, for example, since there is a need for parts such as capacitors that vary among individuals, the performance of the device as a finished product may also vary. It is troublesome. If configured as in the present embodiment, such variations do not occur and adjustment work becomes unnecessary. When an analog circuit is used, it is necessary to set and manufacture an analog circuit corresponding to the type of object on which the electric power steering is mounted. For example, in the electric motor 23 for electric power steering for light vehicles and large vehicles, it is necessary to cope with the difference in current used. If configured as in the present embodiment, it is only necessary to change the program in accordance with the type of the mounting target, and the hardware configuration can be made common.
[0031]
  [Second Embodiment] In the second embodiment, a feedback drive command value is calculated as a drive command value for the electric motor 23.didIs. The operation will be described with reference to a control calculation block (FIG. 5) and a control flowchart (FIG. 6) corresponding to the control calculation block of the electric power steering control apparatus of the second embodiment. The calculation shown in the flowchart of FIG. 6 is executed in the microcomputer 33 every predetermined control period. Further, the vehicle configuration (FIG. 1) mainly showing the electric power steering control device in the case of the first embodiment described above and the processing block diagram (FIG. 2) showing the internal configuration of the ECU 30 are also the case of this second embodiment. Since it is the same, redundant description will not be given.
[0032]
Steps 110 to 170 and step 190 in the flowchart of FIG. 6 have the same contents as steps 10 to 70 and step 90 shown in FIG. 4 in the case of the first embodiment, and thus description thereof will be omitted, and only step 180 and step 185 will be described.
In step 180, the difference (current deviation) between the target motor current and the actual motor current is calculated. In the subsequent step 185, the PWM duty value d (n) is calculated according to the recurrence formula shown in [Formula 1] below.
[0033]
[Expression 1]

[0034]
Here, d (k) is the PWM duty value of the kth cycle, e (k) is the current deviation of the kth cycle, and a (k) and b (k) are predetermined constants.
Thereafter, the PWM duty value d (n) obtained by the above calculation is subjected to upper / lower limiter processing to obtain a final PWM duty value. In step 190, the final PWM duty value is output to the drive circuit 36 to control the motor current.
[0035]
As described above, in the case of the second embodiment, the feedback drive command value is calculated based on the current deviation, and the electric motor 23 is energized and driven based on the feedback drive command value. Control can be realized. In particular, since it is premised on use for electric power steering, it is considered preferable to perform feedback control in consideration of environmental changes such as heat and aging. In addition, in mass production, it is practically difficult to carry out precise adjustment work before shipment, so feedback control is particularly effective from the viewpoint of electric power steering installed in vehicles that are premised on mass production. It can be said that it is preferable to respond.
[0036]
[Third embodiment]
By the way, in order to control the actual motor current to be equal to a predetermined current value by the microcomputer 33, it is desirable that the control cycle is about 1 ms or more in view of the calculation load. In contrast, the standard digital controller design method derives a digital control law by bilinearly transforming the transfer function of an analog feedback controller having a desired performance, and this is performed by a microcomputer. This is a method of calculating. In order to perform control to match the actual motor current to a predetermined current value by this method, the control cycle needs to be about 200 microseconds (μs) or less. If the control is performed at a slower control cycle, the actual motor current becomes oscillating and the steering feeling is deteriorated. This is because the time constant (= inductance (L) / resistance value (R)) of an electric motor used for electric power steering is usually about several ms, but zero when the control period approaches 1 ms. This is because the influence of order hold cannot be ignored.
[0037]
The third embodiment is an embodiment for reducing the influence of zero order hold and realizing a non-vibrating current control system.
[First Aspect of Third Embodiment]
The control calculation block shown in FIG. 7 is obtained by extracting a portion corresponding to the “motor current feedback loop” in the control calculation block of FIG. 5 used for explaining the second embodiment described above. Cd (z) shown in FIG. 7 is a pulse transfer function corresponding to the above [Expression 1], and z represents a delay operator. ZOH is zero order hold, P (s) is a continuous time transfer function representing the electric motor 23 and the drive circuit portion (motor drive circuit 36, power transistor 37), and s is a Laplace operator.
[0038]
In the first embodiment, the motor current feedback controller in the control calculation block of FIG. 7 performs a filter processing function for removing frequency components in a higher frequency range than the cutoff frequency (= resistance value (R) / inductance (L)). It is characterized by having. This is particularly effective when the control cycle (sampling time) becomes a value that cannot be ignored compared to the time constant (L / R) of the electric motor 23.
[0039]
Regarding the configuration of the motor current feedback controller, the configuration of the pulse transfer function Cd (z) will be specifically described with reference to an actual example.
Assume that a DC motor having an inductance (L) = 83.45 μH and a resistance (R) = 53.95 mΩ is assumed as the electric motor 23, and a motor current feedback control system as shown in FIG. 7 is configured. . The control cycle is 1 ms. In this case, the motor time constant is L / R = 1.5 ms, and the control period of 1 ms is not negligible.
[0040]
In this way, when the control cycle cannot be ignored compared with the time constant of the controlled object, for example, the document “Introduction to Digital Control Theory” (written by Mitsuhiko Araki, Asakura Shoten, P137-P139) is considered as a method that considers the influence of zero order hold. There is a known technique described in. The main points are described below.
[0041]
In FIG. 8, Fd (z) represents a pulse transfer function of the feedback controller, ZOH represents zero order hold, and Gc (s) represents a continuous time transfer function to be controlled. In order to discuss the stability in the case of controlling a continuous time control system in the digital controller as shown in FIG. 8, first consider Gd (z) represented by the following [Equation 2],
[0042]
[Expression 2]

[0043]
Furthermore, w shown by the following [Formula 3] is introduced,
[0044]
[Equation 3]

[0045]
The continuous time transfer function Gc (w) is expressed as in [Equation 4] below.
[0046]
[Expression 4]

[0047]
A control system as shown in FIG. 9 is configured for the controlled object represented by the continuous time transfer function Gc (w). The stability margin realized by the controller represented by the continuous time transfer function Fc (w) is obtained by changing the control system of FIG. 8 by the continuous time transfer function Fd (z) defined as [Equation 5] below. This is consistent with the stability margin when configured. This is the content of the known technique described in the above-mentioned document.
[0048]
[Equation 5]

[0049]
That is, in order to obtain a digital controller having a desired stability margin for the control system including zero order hold (ZOH) shown in FIG. 8 in consideration of zero order hold, first, in the control system shown in FIG. The continuous time transfer function Fc (w) may be configured to have a desired stability margin, and then the continuous time transfer function Fd (z) may be configured according to the above [Equation 5].
[0050]
However, how to configure an appropriate continuous time transfer function Fc (w) for a specific control target is not a known technique. In the first aspect, assuming such a known technique, an electric motor 23 for electric power steering is considered as a specific control target, and an appropriate continuous time transfer function Fc (w) for the electric motor 23 is configured. Is.
[0051]
The transfer function P (s) from the equivalent applied voltage to the motor current of the DC motor (L = 83.45 μH, R = 53.95 mΩ) taken as an example is as shown in [Formula 6] below. .
P (s) = 1 / (L * s + R) [Formula 6]
When the continuous time transfer function Pc (w) corresponding to Gc (w) shown in FIG. 9 is obtained for the transfer function P (s) according to the above [Expression 2], [Expression 3], and [Expression 4]. As shown in [Formula 7] below.
[0052]
[Formula 6]

[0053]
A Bode diagram of the continuous time transfer function Pc (w) is shown in FIG. From FIG. 10, it is not possible to define the gain crossover frequency for Pc (w) strictly when considering the zero order hold for P (s), which was the first-order lag, but when looking at the phase change, It can be seen that the phase is 180 degrees behind and the system has a small phase margin.
[0054]
In this first mode, the transfer functions shown in the following [Expression 8], [Expression 9], and [Expression 10] are used as the controller corresponding to Fc (w) in FIG.
[0055]
[Expression 7]

[0056]
The Bode diagram is shown by (B) in FIG. Among these, the transfer function Fc1 (w) uses a first-order lag as an element for the motor current feedback controller to remove a frequency component higher than the cutoff frequency (R / L).
As can be seen from the Bode diagram of the round transfer function Fc (w) * Pc (w) indicated by (c) in FIG. 10, the round trip transfer function Fc (w) * Pc (w It can be seen that the cut-off frequency of () has moved to a low frequency range and the stability margin has increased. Further, Fc2 (w) has an effect of reducing the phase delay due to Fc1 (w).
[0057]
Here, when the pulse transfer function Cd (z) of the motor current feedback controller in FIG. 7 is obtained according to the above [Equation 5], the following [Equation 11] is obtained.
[0058]
[Equation 8]

[0059]
As an example of performing the current feedback control of the above-described DC motor (L = 83.45 μH, R = 53.95 mΩ) using the pulse transfer function Cd (z) at a control period of 1 ms, the target current is changed from OA to 10 A. FIG. 11A shows a response waveform when changing to stepwise. Note that (B) in FIG. 11 is a response waveform when a simple proportional control is performed without a high frequency component removal element. It can be seen that (A) in FIG. 11 has a non-oscillating response and a smaller steady-state deviation than (B).
[0060]
[Second Aspect of Third Embodiment]
In the first aspect described above, a delay element is used. In the second aspect, an integral element is used. That is, instead of the continuous time transfer function Fc (w) shown in FIG. 9, the one shown in the following [Equation 12] is used.
[0061]
[Equation 9]

[0062]
A Bode diagram in that case is shown by (B) in FIG. In the continuous time transfer function Fc (w) shown in [Equation 12], an integral element is used as an element by which the motor current feedback controller removes frequency components higher than the cutoff frequency (L / R). Yes. As can be seen from the Bode diagram of the round transfer function Fc (w) * Pc (w) indicated by (C) in FIG. 12, the round trip transfer function Fc (w) is obtained by the effect of the continuous time transfer function Fc (w). * It can be seen that the cutoff frequency of Pc (w) has moved to the low band, and the stability margin has increased.
[0063]
Further, the continuous time transfer function Fd (z) defined by the above [Expression 5] is obtained using the continuous time transfer function Fc (w) shown by the [Expression 12], and the motor current feedback in FIG. The pulse transfer function Cd (z) of the controller is obtained as shown in [Equation 13] below.
[0064]
[Expression 10]

[0065]
As an example of performing the current feedback control of the above-described DC motor (L = 83.45 μH, R = 53.95 mΩ) using the pulse transfer function Cd (z) at a control period of 1 ms, the target current is changed from OA to 10 A. FIG. 13A shows a response waveform when changing to a step shape.
[0066]
In this case as well, it can be seen that the response is non-vibrating as in the case of using a delay element shown in FIG. In this case, it can also be seen that the steady-state deviation is 0 due to the effect of the integral element.
Note that in both the first mode using the delay element (see the transfer function shown in Expression 8) and the second mode using the integral element (see the transfer function shown in Expression 12), the sampling time is low. In order to realize an oscillating current response, having an element that removes a frequency component higher than the cutoff frequency of the electric motor 23 contributes essentially.
[0067]
[Ingenuity for the second mode]
By the way, when the integration element is used as in the second mode, the following points are taken into consideration when the actual motor current does not reach the target motor current even if the calculated feedback drive command value is maximum. It is preferable to stop the integration operation in the direction of increasing the deviation integration. This is for the following reason.
[0068]
If the deviation integration operation is not stopped in a situation where the electric power steering is suddenly steered, the following inconvenience occurs, that is, the back electromotive force proportional to the motor speed of the electric motor 23 using the DC motor. When electric power is generated and sudden steering is performed, the back electromotive force becomes a large value, so even if the PWM duty is 100%, there is still a state where the actual motor current is lower than the target motor current ( (See between A and B in FIG. 14). The PWM duty value is given as an integral value of deviation and a linear combination value of deviation, but is limited to 100% between DE in FIG. In such a case, if the function of integrating the deviation between the target motor current and the actual motor current is operating, the integrated value of the deviation follows the path D → E in FIG. At point B in FIG. 14, the influence of the counter electromotive force decreases, and even if the actual motor current catches up with the target motor current, the deviation integral follows the path E → F in FIG. Remains. As a result, an overshoot of the actual motor current occurs between B and C in FIG.
[0069]
Therefore, control as shown in the flowchart of FIG. 15 is executed, and the deviation integration operation is stopped if necessary.
Prior to the description of the flowchart of FIG. 15, a description related to the processing will be given. The pulse transfer function Cd (z) of the motor current feedback controller shown in [Equation 13] can be modified as shown in [Equation 14] below.
[0070]
[Expression 11]

[0071]
That is, the part corresponding to the deviation proportional [0.05] and the part corresponding to the deviation integral [0.05 / (1-Z^-1)]]. The calculation for each of the deviation proportional correspondence portion and the deviation integration correspondence portion is performed in FIG.
Steps 110 to 180 and step 190 in the flowchart of FIG. 15 are the same as the contents shown in FIG. 6 in the second embodiment, except that step 181 to tep 184 are executed instead of step 185 in FIG. . Then, the part is mainly demonstrated.
[0072]
After calculating the difference (current deviation) between the target motor current and the actual motor current in step 180, it is checked in step 181 whether or not the PWM duty value of the previous cycle is the maximum value (saturation check). Depending on the check result, the deviation integral calculation of the PWM duty value d (n) in step 182 changes.
[0073]
That is, in step 182, when the PWM duty value of the previous cycle is the maximum value, the integral calculation in the direction in which the integral value increases (the deviation integration corresponding part [0.05 / (1 -Z^-1)]) Is not performed. Note that the integral calculation is performed in the direction in which the integral value decreases. On the other hand, if the PWM duty value of the previous cycle is not the maximum value, the integral calculation (the deviation integration corresponding part in [Expression 14]) is performed as it is.
[0074]
In the subsequent step 183, the calculation of the deviation proportional corresponding part [0.05] shown in [Expression 14] is executed.
In step 184, the results of step 182 and step 183 are added, and necessary guard processing is performed to obtain a final PWM duty value.
[0075]
By executing such processing, the integration operation in the direction of increasing the deviation integration is stopped at the point D in FIG. 14 where the PWM duty has reached 100%. Therefore, the deviation integral value follows the path D → G in FIG. Therefore, at point B in FIG. 14, when the influence of the counter electromotive force is reduced and the actual motor current catches up with the target motor current, the deviation integral follows the G → H path in FIG. The motor starts to decrease, the actual motor current does not overshoot, and a good steering feeling can be obtained.
[0076]
[Fourth embodiment]
In the fourth embodiment, in parallel with the motor current feedback controller in the second or third embodiment described above, a motor current feedforward control that calculates a command value for the motor drive circuit 36 as a feedforward term based on the target current. A vessel is provided. The operation will be described with reference to the control calculation block (FIG. 16) and the control flowchart (FIG. 17) corresponding to the control calculation block of the electric power steering control apparatus of the fourth embodiment. The calculation shown in the flowchart of FIG. 15 is executed in the microcomputer 33 every predetermined control cycle. Further, the vehicle configuration (FIG. 1) mainly showing the electric power steering control device in the case of the first embodiment described above and the processing block diagram (FIG. 2) showing the internal configuration of the ECU 30 are also the case of this second embodiment. Since it is the same, redundant description will not be given.
[0077]
Steps 210 to 270 and step 290 in the flowchart of FIG. 17 have the same contents as steps 10 to 70 and step 90 shown in FIG. 4 in the case of the first embodiment, and thus description thereof is omitted.
In step 280, the difference (current deviation) between the target motor current and the actual motor current is calculated in the same manner as in step 180 in the case of the second embodiment described above. In the subsequent step 285, the PWM duty feedback value dfb (n) is calculated according to the recurrence formula shown in [Formula 15] below.
[0078]
[Expression 12]

[0079]
Here, dfb (k) is the PWM duty feedback value in the kth cycle, e (k) is the current deviation in the kth cycle, and a (k) and b (k) are predetermined constants.
In the subsequent step 286, the PWM duty feedforward value dff (n) is calculated according to the recurrence formula shown in [Expression 16] below.
[0080]
dff (n) = f (R * Iref (n)) [Formula 16]
Here, R is the resistance value of the DC motor, lref (n) is the target motor current in the nth period, and f (...) Is a preset map.
Therefore, in step 287, the PWM duty final value d (n) is calculated as shown in [Equation 17] below.
[0081]
d (n) = dff (n) + dfb (n) (Equation 17)
Thereafter, the PWM duty value d (n) obtained by the above calculation is subjected to upper / lower limiter processing to obtain a final PWM duty value. In step 290, the final PWM duty value is output to the drive circuit 36 to control the motor current.
[0082]
As described above, in the case of the fourth embodiment, the PWM duty final value d (n) is calculated based on the PWM duty feedback value dfb (n) and the PWM duty feedforward value dff (n), and the value is calculated. Since the motor current is controlled based on the above, the steady deviation of the motor current can be reduced and the response of the current rising can be improved, and the steering feeling can be further improved.
[0083]
[Others]
(1) In the above-described embodiment, it is assumed that the electric motor 23 performs auxiliary steering by rotating the shaft via the speed reducer 22, but other than that, for example, the rack 13 is directly driven by electromagnetic force or the like. The same applies to the premise of a “rack coaxial” configuration to be driven.
[0084]
(2) In the first mode of the third embodiment described above, a transfer function having a first-order lag element is used. However, the transfer function is not limited to the first-order lag element, and a transfer function having a second-order or more delay element is used. Even if it uses, it is realizable similarly.
[Brief description of the drawings]
FIG. 1 is a vehicle configuration diagram mainly showing an electric power steering control device of an embodiment to which the present invention is applied.
FIG. 2 is a processing block diagram illustrating an internal configuration of an ECU according to an embodiment.
FIG. 3 is a control calculation block diagram of the electric power steering control device of the first embodiment.
FIG. 4 is a control flowchart corresponding to the control calculation block of the first embodiment.
FIG. 5 is a control calculation block diagram of an electric power steering control device according to a second embodiment.
FIG. 6 is a control flowchart corresponding to a control calculation block of the second embodiment.
7 shows a portion corresponding to the “motor current feedback loop” in the control calculation block of FIG.
FIG. 8 is a control calculation block diagram for illustrating a first mode of a third embodiment.
FIG. 9 is a control calculation block diagram for illustrating a first mode of a third embodiment;
FIG. 10 is a Bode diagram of a transfer function in the first mode of the third embodiment.
FIG. 11 is an explanatory diagram of a response waveform when the target current is changed stepwise when the current F / B control of the DC motor is performed using the pulse transfer function in the first mode of the third embodiment. is there.
FIG. 12 is a Bode diagram of a transfer function in the second mode of the third embodiment.
FIG. 13 is an explanatory diagram of a response waveform when the target current is changed stepwise when the current F / B control of the DC motor is performed using the pulse transfer function in the second mode of the third embodiment. is there.
FIG. 14 is an explanatory diagram showing a comparison between a case where the deviation integration operation is not stopped and a case where it is stopped in a situation where the electric power steering is suddenly steered, with respect to the second mode of the third embodiment.
FIG. 15 is a flowchart showing control in the case where the deviation integration operation is stopped with respect to the second mode of the third embodiment.
FIG. 16 is a control calculation block diagram of an electric power steering control device according to a fourth embodiment;
FIG. 17 is a control flowchart corresponding to the control calculation block of the fourth embodiment.
[Explanation of symbols]
11 ... Handle 12 ... Shaft
13 ... Rack 14 ... Pinion gear
15 ... Tie rod 16 ... Evening
21 ... Steering torque sensor 22 ... Reducer
23 ... Electric motor 24 ... Steering angle sensor
25 ... Motor rotation angle sensor 27 ... Motor current sensor
30 ... Microcomputer 31,32 ... LPF
34 ... A / D converter 35 ... CPU
36 ... Motor drive circuit 37 ... Power transistor
38, 39, 40 ... Waveform shaping unit 50 ... Vehicle speed sensor

Claims (8)

An electric motor for auxiliary steering of the steering mechanism;
Motor driving means for energizing and driving the electric motor;
Steering torque detection means for detecting steering torque in the steering mechanism;
Command value calculating means for calculating a drive command value for the motor driving means based on the steering torque detected by the steering torque detecting means;
An electric power steering electric motor control device comprising:
Furthermore, a motor current detecting means for detecting a current flowing through the electric motor is provided,
The command value calculation means includes
Target motor current calculation means for calculating a target motor current for the electric motor as a digital value based on the steering torque detected by the steering torque detection means;
AD conversion means for converting the value detected by the motor current detection means into digital data that can be processed by a microcomputer;
The detected current value obtained by converting the drive command value for the motor drive means into digital data by the AD conversion means and the target motor so that the current (actual motor current) flowing through the electric motor matches the target motor current. Calculation means for calculating by a microcomputer at predetermined time intervals based on the current value,
The calculating means calculates a deviation between the target motor current and the actual motor current, calculates a feedback drive command value for the motor driving means based on the calculated current deviation, and applies an applied voltage of the electric motor. A control device for an electric motor for an electric power steering, wherein digital filter processing for removing a frequency component higher than the cutoff frequency when the response of the current to is approximated as a first order lag is executed. In the control apparatus of the electric motor for electric power steering according to claim 1,
The electric filter characterized in that the calculation means realizes the digital filter processing by calculating the feedback drive command value using a transfer function having an element that removes a frequency component higher than the cutoff frequency. Control device for power steering electric motor. In the control apparatus of the electric motor for electric power steering according to claim 2,
A control device for an electric motor for an electric power steering, wherein the calculation means uses a transfer function having a delay element as an element for removing a frequency component higher than the cutoff frequency. In the control apparatus of the electric motor for electric power steering according to claim 2,
The electric power characterized in that the calculation means uses a transfer function having an element for integrating a deviation between the target motor current and the actual motor current as an element for removing a frequency component higher than the cutoff frequency. Control device for steering electric motor. In the control apparatus of the electric motor for electric power steering according to claim 4,
The calculation means stops the integration operation in a direction to increase the deviation integration when the actual motor current does not reach the target motor current even if the calculated feedback drive command value is maximum. A control device for an electric motor for electric power steering. In the control apparatus of the electric motor for electric power steering according to any one of claims 1 to 5,
The calculation means calculates a feedforward drive command value for the motor drive means based on the target motor current, and adds the calculated value to the feedback drive command value to obtain a final drive command value for the motor drive means. A control device for an electric motor for electric power steering. In the control apparatus of the electric motor for electric power steering according to claim 3,
The electric power steering characterized in that the arithmetic means uses a transfer function having a delay element and an element for reducing a phase delay due to the delay element as an element for removing a frequency component higher than the cutoff frequency. Electric motor control device. In the control device of the electric motor for electric power steering according to claim 7,
The arithmetic means uses a transfer function represented by the following [Expression a], [Expression b], and [Expression c].

Images