JP3806872B2 - Motor driving method and apparatus - Google Patents

Description

【００８０】
前記固定座標モータ逆モデル部５ｅは、αβ電流ベクトルに相抵抗を乗算して電圧降下分を算出する抵抗乗算部５ｅ１と、αβ電圧ベクトルから電圧降下分を減算する第１減算部５ｅ２と、第１減算部５ｅ２からの出力を積分する積分部５ｅ３と、αβ電流ベクトルにｑ軸インダクタンスを乗算するインダクタンス乗算部５ｅ４と、積分結果からｑ軸インダクタンス乗算結果を減算する第２減算部５ｅ５とを有している。
【００８１】
前記位置算出部５ｆは、固定座標モータ逆モデル部５ｅからの出力に対してｔａｎ−１処理を行ってロータ回転位置を出力する。
【００８２】
なお、以下においては、両３相→２相変換部５ａ、５ｂ、固定座標モータ逆モデル部５ｅ、および位置算出部５ｆを位置検出部５と総称する。
【００８３】
このモータ駆動装置に含まれる固定座標モータ逆モデル部５ａにおいて設定される相抵抗およびｑ軸インダクタンスが正確であり、かつ電流検出部５ａおよび電圧検出部５ｂが正確であれば、ロータ回転位置を正確に推定することができ、推定されたロータ回転位置を用いてモータ４を安定にかつ正確に運転することができる。
【００８４】
しかし、一般的には、ばらつき、温度変化などが原因となって、電流検出部５ａ、電圧検出部５ｂが誤差を有し、および／または固定座標モータ逆モデル部５ｅに設定されるモータ機器定数が誤差を有しているので、ロータ回転位置の推定結果が不正確になり、モータ４を安定にかつ正確に運転することができなくなってしまう。
【００８５】
このような不都合を解消するために、この発明は、負荷の大きさに対してモータ電流過大、かつモータ電流が遅れ位相の状態において、電流検出部５ａ、電圧検出部５ｂの定数（例えば、ゲイン）、モータ機器定数のうち、少なくとも１つの誤差を削減する誤差削減部６を設けている。
【００８６】
さらに説明する。
【００８７】
電流位相に対するモータトルクの一例を図２に示す。なお、この例は、永久磁石を内部に埋め込んでなるロータを有するブラシレスＤＣモータに基づくものである。
【００８８】
図２において、電流位相０°はｑ軸方向にモータ電流が流れている場合（誘起電圧と同位相である場合）であり、プラス側が位相進みを示している。
【００８９】
負荷に対してモータ電流が大きく、しかも位相遅れの場合には、モータ電流に対してトルクの出にくい電流位相（図２中の丸を参照）で運転される。
【００９０】
また、負荷の大きさに対してモータ電流過大、かつモータ電流が遅れ位相の状態におけるベクトルシミュレーションの一例を図３に示す。
【００９１】
ｄ軸は磁石の磁束の方向であり、ｑ軸はそれから電気的に９０度進んだ位相（つまり誘起電圧と同位相）に取られている。Ｉはモータ電流ベクトルであり、ｑ軸に対して遅れている。Ｅは磁石磁束による誘起電圧、ωは回転角周波数（電気角）、Ｌｑ、Ｌｄはｑ軸、ｄ軸インダクタンス、Ｉｄ、Ｉｑはモータ電流Ｉのｄ軸、ｑ軸成分、Ｒは巻線抵抗、Ｖはモータ電圧、φは巻線鎖交磁束を示している。なお、添え字ｍで表した文字はモータの値であって、制御での値と区別している。
【００９２】
そして、図１に示すモータ駆動装置では、モータに供給されるモータ電流、モータ電圧と巻線抵抗Ｒ、ｑ軸インダクタンスＬｑに基づいてｄ軸方向の長さがφｍ＋（Ｌｄ−Ｌｑ）Ｉｄのベクトルを算出し、そのベクトルの角度によりロータ位置を検出している。
【００９３】
図４に電流位相に対するφｍ＋（Ｌｄ−Ｌｑ）Ｉｄの振幅の関係を示す。なお、この関係は、Ｌｄ＝１１．０ｍＨ、Ｌｑ＝２３．９ｍＨ、Ｒ＝０．７９４５Ω、φ＝０．１４９ｗｂに設定して得たものである。
【００９４】
また、極対数をＰｎとしたときモータトルクは
｛φｍ＋（Ｌｄ−Ｌｑ）Ｉｄ｝Ｉｄ＊Ｐｎ＝φＩｓｉｎθφ−ＩＰｎ
ここでθφ−Ｉは巻線鎖交磁束φとモータ電流Ｉとの成す角で、トルクに対して電流が大きい場合には図３に示したようにＩｑ≠０のためφｍ＋（Ｌｄ−Ｌｑ）Ｉｄが極めて小さくなる。このため、負荷の大きさに対して電流過大の場合には、モータの機器定数や電流検出部５ａ、電圧検出部５ｂの定数の誤差により極めて大きな角度誤差を生じることになり、このような条件では安定な位置検出を行うことが極めて困難となる。しかし、逆に、誤差が大きな角度偏差として現れるため、この運転条件を用いれば誤差削減部６によって、極めて容易にモータの機器定数、電流検出部５ａ、電圧検出部５ｂの定数の誤差削減を実現することができる。
【００９５】
この場合において、誤差の削減はロータ位置にかかわらずステータに強制的に回転磁界を加えて行うことが好ましい。
【００９６】
さらに説明する。
【００９７】
前述したように負荷の大きさに対して電流過大であり、かつ電流が遅れ位相である運転条件では位置検出部５の出力は極めて不安定であって、位置検出部５の出力によりモータを運転すると不安定になって、最悪発散停止してしまう。そこで、位置検出部の出力を用いずに強制的にステータに回転磁界を加えてモータを回転させた状態で誤差の削減を行うことで、不安定を解消できる。また、通常は、強制的な回転磁界による運転を行うと、トルク変動などによって脱調してしまうことが懸念されるが、この場合には負荷に比べて十分大きな電流を流しているため脱出トルクに十分余裕があり、したがって、脱調の問題も発生しない。
【００９８】
特に強制的にステータに回転磁界を加える場合、図２から分かるように進め位相側（略９０゜のポイント）でトルクに対して電流の大きな部分が出現するが、進め位相側はロータが回転することにより位相が遅れ側に動きトルクが出てさらに位相が遅れるという不安定なポイントであるので、結局遅れ位相側でしか安定することがない。このため、位相制御せず強制的に回転磁界を加えれば、自動的に遅れ位相で運転される。この結果、不安定を解消でき、しかも誤差を削減することができる。
【００９９】
さらに、位置検出部５の出力を用いずに強制的にステータに回転磁界を加えてモータを回転させて起動するようにするとともに、モータの機器定数、電流検出部５ａ、電圧検出部５ｂの定数の誤差の削減を起動時に行うことが好ましい。 すなわち、強制的にステータに回転磁界を加えてモータを回転させて起動する場合には、起動時の負荷トルクが脱出トルク以下でかつ起動時には位置検出部５を用いずに駆動を行っているので、前述の何れかの条件を容易に実現することができ、誤差の削減に好適である。
【０１００】
さらにまた、誤差の削減を、モータコイルの鎖交磁束の位相が位置検出部５の推定結果から見たｄ軸を基準として回転方向略０°〜９０°の範囲に入るように行うことが好ましい。
【０１０１】
さらに説明する。
【０１０２】
磁石磁束φｍはｄ軸方向を向いたベクトルであり、巻線鎖交磁束φのｄ、ｑ軸成分（φｄ、φｑ）は（φｍ＋ＬｄＩｄ、ＬｑＩｑ）である。
【０１０３】
図３に示したように負荷に対して電流が大きくしかも位相遅れの場合には、Ｉｄは正（遅れ位相）、Ｉｑは正（力行）であるので、巻線鎖交磁束のｄ軸を基準とした位相は０°〜９０°の範囲に必ず存在する。
【０１０４】
したがって、負荷の大きさに対して電流過大でかつ遅れ位相時には推定磁束位相がｄ軸を基準とした回転方向略０°〜９０°に有るとしてそれを外れた場合に位置検出部５におけるモータの機器定数、電流検出部５ａ、電圧検出部５ｂの定数を調節すれば、誤差を削減できる。
【０１０５】
上記の何れかの場合において、モータコイルの鎖交磁束の位相が位置検出部５の推定結果から見たｄ軸を基準としてマイナス方向にある場合には抵抗を削減し、略９０°〜１８０°の範囲にある場合にはｑ軸インダクタンスＬｑを削減することが好ましい。
【０１０６】
図５のフローチャートを参照してさらに説明する。
【０１０７】
ステップＳＰ１において、モータ電流およびモータ電圧を検出し、ステップＳＰ２において、モータ電流、モータ電圧、およびモータの機器定数を用いてロータ回転位置（ｄ軸方向の位置）の検出（推定）を行い、ステップＳＰ３において、ロータ回転位置検出結果および算出されたモータ磁束ベクトルの方向とから巻線鎖交磁束の位相を算出し、ステップＳＰ４において、巻線鎖交磁束の位相が０°よりも小さいか否かを判定する。
【０１０８】
そして、ステップＳＰ４において巻線鎖交磁束の位相が０°よりも小さいと判定された場合には、ステップＳＰ５において、位置検出部５に設定されたモータの機器定数のうち巻線抵抗Ｒを減少させ、再びステップＳＰ１の処理を行う。
【０１０９】
ステップＳＰ４において巻線鎖交磁束の位相が０°よりも小さくないと判定された場合には、ステップＳＰ６において、巻線鎖交磁束の位相が９０°よりも大きいか否かを判定する。
【０１１０】
ステップＳＰ６において巻線鎖交磁束の位相が９０°よりも大きいと判定された場合には、ステップＳＰ７において、位置検出部５に設定されたモータの機器定数のうちｑ軸インダクタンスＬｑを減少させ、再びステップＳＰ１の処理を行う。
【０１１１】
ステップＳＰ６において巻線鎖交磁束の位相が９０°よりも大きくないと判定された場合には、そのまま一連の処理を終了する。
【０１１２】
さらに説明する。
【０１１３】
図６中ａに図１のモータ駆動装置の位置検出部５が位置検出を行った場合の巻線抵抗Ｒの変化に対する推定されたｄ軸からみた磁束の位相のシミュレーション結果を、図６中ｂにｑ軸インダクタンスＬｑの変化に対する推定されたｄ軸からみた磁束の位相のシミュレーション結果をそれぞれ示す。
【０１１４】
図６中ａより、モータの巻線抵抗が略−１５％よりも小さい場合に磁束の位相がマイナスになるため、誤差があることを検出できる。また、図６中ｂより、モータのｑ軸インダクタンスＬｑが略−５％よりも小さい場合に位相が９０°〜１８０゜になりモータのｑ軸インダクタンスＬｑが小さいことが検出できる。したがって、モータコイルの鎖交磁束の位相が位置検出部５の推定結果から見たｄ軸を基準としてマイナス方向にある場合には抵抗を削減し、略９０°〜１８０°の範囲にある場合にはｑ軸インダクタンスＬｑを削減することでモータと位置検出部５の機器定数との誤差を削減できる。
【０１１５】
さらに精度を高めようとする場合には位置検出部における推定誤差が大きく現れる状態で前記誤差の削減を行えばよいことから、さらにモータ電流を負荷に対して過大に流せばよいことが分かる。例えば図６のシミュレーションで電流をさらに３０％増せば、モータの巻線抵抗が略−１０％よりも小さい場合に磁束の位相がマイナスになり誤差を検出可能となる。
【０１１６】
また、ここでは機器定数に誤差があるものとして議論したが、同様に電流検出部５ａ、電圧検出部５ｂの定数の誤差についても削減可能である。電流検出部５ａ、電圧検出部５ｂの定数と機器定数が同時に誤差をもった場合にはそれぞれを分離することはできないが、組み合わせられた状態で誤差を削減し、位置検出部５の推定誤差を削減することができる。
【０１１７】
さらにまた、誤差の削減を、位置検出部５による推定電圧位相が平均的にｄ軸を基準とした略９０°〜１８０°の範囲に入るように行うことが好ましい。
【０１１８】
すなわち、図３に示されるように、電圧位相は磁束の微分に巻線抵抗による電圧降下を加えたベクトルとなる。ここで、巻線抵抗による電圧降下が小さいとすれば、電圧位相は巻線鎖交磁束から９０゜進んだ、ｄ軸から略９０°〜１８０°の範囲に位置することになる。したがって、前述と同様に電圧位相を用いて位置検出部５におけるモータの機器定数、電流検出部５ａ、電圧検出部５ｂの定数の誤差削減を実現できる。
【０１１９】
また、誤差の削減を、推定電流位相がｄ軸を基準とした回転方向略０°〜９０°の範囲に入るように行うことが好ましい。
【０１２０】
さらに説明する。
【０１２１】
前述したようにモータトルクはφｘＩｘｓｉｎθ φ−ＩＰｎで表される。ここではモータ電流がトルクに対して十分大きいとしているので、Ｉベクトル、φベクトルの長さは長く、Ｉベクトルとφベクトルの成す角が小さいことになる。そして、φベクトルが０°〜９０゜の範囲に位置するわけであるから、Ｉベクトルも略０°〜９０゜の範囲に位置するとし、前述と同様にして電流位相を用いて位置検出部５におけるモータの機器定数、電流検出部５ａ、電圧検出部５ｂの定数の誤差削減を実現できる。
【０１２２】
上記各実施態様において、位置検出部５に設定されるｑ軸インダクタンスＬｑ、巻線抵抗Ｒの値をモータバラツキの中心値よりも大きく設定し、または電流検出部５ａ、電圧検出部５ｂの定数をそれと等価な方向に偏差をもって設定することが好ましい。
【０１２３】
さらに説明する。
【０１２４】
前述したように、上記各実施態様においてはモータの巻線抵抗Ｒ、ｑ軸インダクタンスＬｑが小さいことを検知し、位置検出部５における設定値を補正することができる。しかし、逆にモータの巻線抵抗Ｒ、ｑ軸インダクタンスＬｑが大きい場合には誤差を削減することが難しい。このため、予め位置検出部５に設定されるｑ軸インダクタンスＬｑ、巻線抵抗Ｒの値を大きく設定することで、より確実に誤差の大きさを狭めることができる。また、電流検出部５ａ、電圧検出部５ｂの定数により調整する場合には例えば図１のモータ駆動装置においては、電流を実際よりも大きく設定するか、または電圧を小さく設定することで等価な方向に偏差を作り出すことができる。
【０１２５】
そして、これらの処理を行うことによって、誤差削減を簡単に実現することができる。
【０１５１】
図７はこの発明のモータ駆動方法のさらに他の実施態様の要部を説明するフローチャートである。
【０１５２】
ステップＳＰ１において、長さ検出処理を行い、ステップＳＰ２において、ｑ軸インダクタンスＬｑを増加させ、ステップＳＰ３において、長さ検出処理を行い、ステップＳＰ４において、今回の長さが前回の長さよりも長いか否かを判定する。
【０１５３】
そして、今回の長さが前回の長さよりも長くないと判定された場合には、再びステップＳＰ２の処理を行う。
【０１５４】
逆に、ステップＳＰ４において今回の長さが前回の長さよりも長いと判定された場合には、ステップＳＰ５において、ｑ軸インダクタンスＬｑを減少させ、ステップＳＰ６において、長さ検出処理を行い、ステップＳＰ７において、今回の長さが前回の長さよりも長いか否かを判定する。
【０１５５】
そして、今回の長さが前回の長さよりも長くないと判定された場合には、再びステップＳＰ５の処理を行う。
【０１５６】
逆に、ステップＳＰ７において今回の長さが前回の長さよりも長いと判定された場合には、そのまま一連の処理を終了する。
【０１５７】
図８は前記長さ検出処理を説明するフローチャートである。
【０１５８】
ステップＳＰ１において、モータ電流およびモータ電圧を検出し、ステップＳＰ２において、モータの回転子の磁極位置を検出し、ステップＳＰ３において、磁石鎖交磁束φｍに対してｄ軸インダクタンスＬｄとｑ軸インダクタンスＬｑとの差にｄ軸電流ｉｄを乗算した値を加算した結果φｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さ（大きさ）を検出し、そのまま一連の処理を終了する。
【０１５９】
さらに説明する。
【０１６０】
前述したように、トルクに対して電流が大きい場合には図４に示したようにφｍ＋(Ｌｄ-Ｌｑ）ｉｄが極めて小さくなるので、逆に、十分大きな電流を流している場合に、φｍ＋(Ｌｄ-Ｌｑ）ｉｄが最小になるようにモータ機器定数またはセンサーのゲイン等の定数（センサー定数）を変化させることで、位置検出誤差を削減することが可能となる。
【０１６１】
図９に負荷に比して十分大きな電流が流れている場合（最大トルクを発生する電流位相に合わせた場合に比して１０倍以上の電流を流している）のモータ機器定数誤差に対して、センサレス制御により算出されるφｍ＋(Ｌｄ-Ｌｑ）ｉｄの大きさを示す。図９から分かるように、Ｌｑの誤差がほぼ０のポイントで算出されるφｍ＋(Ｌｄ-Ｌｑ）ｉｄの大きさが最小になる。
【０１６２】
したがって、逆に、算出されるφｍ＋(Ｌｄ-Ｌｑ）ｉｄの大きさが最小になるようにＬｑを調節することで、正しいＬｑを求めることができることが分る。また、φｍ＋(Ｌｄ-Ｌｑ）ｉｄに対応する量はセンサレス制御（位置センサを用いることなく位置検出を行ってインバータを制御するもの）では∫（Ｖ−Ｒｉ）ｄt− Ｌｑｉで求めることができる。
【０１６３】
したがって、図７、図８のフローチャートの処理を行うことによって、ｑ軸インダクタンスＬｑに含まれる誤差を低減することができる。
【０１６４】
また、定数調整後に調整結果に対して所定の修正を行うことが好ましい。
【０１６５】
負荷が若干大きな場合（最大トルクを発生する電流位相に合わせた場合に比して５倍程度の電流を流している）の、算出されるφｍ＋(Ｌｄ-Ｌｑ）ｉｄの大きさを図１０に示す。
【０１６６】
Ｌｑの誤差に対して算出されるφｍ＋(Ｌｄ-Ｌｑ）ｉｄの大きさの最小値が若干マイナス側にシフトしていることが分る。
【０１６７】
Ｒについては図９，１０ともにマイナス側にシフトしているが、図１０の方がより大きくシフトしていることが分る。
【０１６８】
また、圧縮機において負荷（差圧）を変化させながら機器定数を推定させた結果を図１１に示す。
【０１６９】
図１１から、負荷が高い場合ほど推定されるＬｑの値が大きくなっているのが分る。
【０１７０】
このことから、起動時の負荷が想定される場合には、負荷の状態に応じて前述の方法で算出された機器定数に補正をかけることで正しい機器定数を得られることが分る。
【０１７１】
図１２はこの発明のモータ駆動方法のさらに他の実施態様の要部を説明するフローチャートである。
【０１７２】
ステップＳＰ１において、Ｌｑを小さく設定し、ステップＳＰ２において、長さ位相検出処理を行い、ステップＳＰ３において、検出された位相が所定位相より小さいか否かを判定する。
【０１７３】
そして、検出された位相が所定位相より小さいと判定されれば、ステップＳＰ４において、Ｒを減少させる。
【０１７４】
逆に、ステップＳＰ３において、検出された位相が所定位相より小さくないと判定された場合には、ステップＳＰ５において、検出された位相が所定位相より大きいか否かを判定する。
【０１７５】
そして、検出された位相が所定位相より大きいと判定されれば、ステップＳＰ６において、Ｒを増加させる。
【０１７６】
逆に、ステップＳＰ５において、検出された位相が所定位相より大きくないと判定された場合には、ステップＳＰ７において、φｍ＋（Ｌｄ-Ｌｑ）ｉｄが所定値より短いか否かを判定する。
【０１７７】
そして、φｍ＋（Ｌｄ-Ｌｑ）ｉｄが所定値より短いと判定されれば、ステップＳＰ８において、Ｌｑを減少させる。
【０１７８】
逆に、ステップＳＰ７において、φｍ＋（Ｌｄ-Ｌｑ）ｉｄが所定値より短くないと判定された場合には、ステップＳＰ９において、φｍ＋（Ｌｄ-Ｌｑ）ｉｄが所定値より長いか否かを判定する。
【０１７９】
そして、φｍ＋（Ｌｄ-Ｌｑ）ｉｄが所定値より長いと判定されれば、ステップＳＰ１０において、Ｌｑを増加させる。
【０１８０】
また、ステップＳＰ４の処理が行われた場合、ステップＳＰ６の処理が行われた場合、ステップＳＰ８の処理が行われた場合、ステップＳＰ１０の処理が行われた場合には、再びステップＳＰ１の処理を行う。
【０１８１】
逆に、ステップＳＰ９において、φｍ＋（Ｌｄ-Ｌｑ）ｉｄが所定値より長くないと判定された場合には、そのまま一連の処理を終了する。
【０１８２】
図１３は長さ位相検出処理を説明するフローチャートである。
【０１８３】
ステップＳＰ１において、モータ電流およびモータ電圧を検出し、ステップＳＰ２において、モータの回転子の磁極位置を検出し、ステップＳＰ３において、磁束の位相を検出し、ステップＳＰ４において、磁石鎖交磁束φｍに対してｄ軸インダクタンスＬｄとｑ軸インダクタンスＬｑとの差にｄ軸電流Ｉｄを乗算した値を加算した結果φｍ＋（Ｌｄ-Ｌｑ）Ｉｄの長さ（大きさ）を検出し、そのまま一連の処理を終了する。
【０１８４】
図１２、図１３のフローチャートの処理を行うことによって、モータコイルの鎖交磁束の位相が位置センサレスの推定したｄ軸を基準として予め設定された値になるように、かつ、推定されたφｍ＋（Ｌｄ−Ｌｑ）ｉｄに対応する量の大きさがあらかじめ設定された所定の値になるように誤差の削減を行うことができる。
【０１８５】
同じ条件での推定された鎖交磁束角度及びφｍ＋(Ｌｄ-Ｌｑ）ｉｄの大きさを示す図６、および図７を参照してさらに説明する。
【０１８６】
先ず、制御内でＬｑを小さく設定し（図６、図９はモータ定数にずれが生じた場合を想定して表現しているため、制御内の定数の変化を符号が逆であることに注意する。つまり、制御内で機器定数を小さく設定したことはモータの機器定数が大きくなったことに相当する）、φｍ＋（Ｌｄ−Ｌｑ）ｉｄの長さとモータコイルの鎖交磁束の位相を算出する。そして、所定位相よりも小さいことに応答してＲを減少、大きいことに応答してＲを増加することで位相を所定の値に調整することができる。これは例えば、所定値を位相６０度、長さ０.０１５とし、Ｌｑを小さくしたことにより図６中（ａ）のＬｑ１０％、Ｒ１０％の位置に乗ったとすれば、位相が小さいことに応答して制御上のＲを減少させる（つまりモータの定数でかかれている図６中（ａ）ではＲを増加させる）ことで、Ｌｑ１０%のラインを右にたどり６０度の位置まで来ることに相当する。
【０１８７】
そして、Ｒによる位相調整を行いながら、φｍ＋(Ｌｄ-Ｌｑ）iｄが所定値よりも短いことに応答してＬｑを減少させ、φｍ＋(Ｌｄ-Ｌｑ）iｄが所定値よりも長いことに応答してＬｑを増加させることで、所望の位相、長さに制御できる。
【０１８８】
これは図９中（ｂ）においてＲ１０%、Ｌｑ１０%の位置にあるとき、制御内でのＬｑを増加させる（図９中（ｂ）上では減少させる）ことでφｍ＋(Ｌｄ-Ｌｑ）iｄを小さくし、これによる位相の増大をＲを増大（図６中（a）上では減少）させることで制御することに相当し、結局所望の位相、φｍ＋(Ｌｄ-Ｌｑ)iｄ長さに制御して、正しい機器定数を得ることが可能になる。
【０１８９】
図１４はこの発明のモータ駆動方法のさらに他の実施態様の要部を説明するフローチャートである。
【０１９０】
ステップＳＰ１において、モータ電流およびモータ電圧を検出し、ステップＳＰ２において、モータの回転子の磁極位置を検出し、ステップＳＰ３において、磁束の位相を検出し、ステップＳＰ４において、磁石鎖交磁束φｍに対してｄ軸インダクタンスＬｄとｑ軸インダクタンスＬｑとの差にｄ軸電流ｉｄを乗算した値を加算した結果φｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さ（大きさ）を検出し、ステップＳＰ５において、磁束の位相が所定位相であり、かつφｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さが所定長さであるか否かを判定する。
【０１９１】
そして、ステップＳＰ５において磁束の位相が所定位相でないか、またはφｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さが所定長さでないと判定された場合には、ステップＳＰ６において、磁束の位相が±９０°以外か否かを判定し、磁束の位相が±９０°以外であると判定された場合には、ステップＳＰ７において、ｑ軸インダクタンスＬｑを減少させる。
【０１９２】
ステップＳＰ６において磁束の位相が±９０°以外でないと判定された場合には、ステップＳＰ８において、φｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さが所定長さより短いか否かを判定し、φｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さが所定長さより短いと判定された場合には、ステップＳＰ９において、ｑ軸インダクタンスＬｑを減少させる。
【０１９３】
逆にステップＳＰ８において、φｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さが所定長さより短くないと判定された場合、またはステップＳＰ９の処理が行われた場合には、ステップＳＰ１０において、磁束の位相が所定位相より大きいか否かを判定し、磁束の位相が所定位相より大きいと判定された場合には、ステップＳＰ１１において、巻線抵抗Ｒを増大させる。
【０１９４】
逆に、ステップＳＰ１０において、磁束の位相が所定位相より大きくないと判定された場合には、ステップＳＰ１２において、巻線抵抗Ｒを増大させ、ステップＳＰ１３において、磁束の位相が所定位相であり、かつφｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さが所定値よりも長いか否かを判定し、磁束の位相が所定位相であり、かつφｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さが所定値よりも長いと判定された場合には、ステップＳＰ１４において、ｑ軸インダクタンスＬｑを増大させる。
【０１９５】
そして、ステップＳＰ７の処理が行われた場合、ステップＳＰ９の処理が行われた場合、ステップＳＰ１１の処理が行われた場合、ステップＳＰ１４の処理が行われた場合、またはステップＳＰ１３において、磁束の位相が所定位相でないか、またはφｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さが所定値よりも長くないと判定された場合には、再びステップＳＰ１の処理を行う。
【０１９６】
そして、最終的に、ステップＳＰ５において磁束の位相が所定位相であり、かつφｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さが所定長さであると判定された場合には、一連の誤差削減処理を終了する。
【０１９７】
さらに説明する。
【０１９８】
トルクに対してモータ電流が大きい場合には、φｍ＋（Ｌｄ-Ｌｑ）ｉｄが極めて小さくなるので、逆に、十分に大きな電流を流している場合には、φｍ＋（Ｌｄ-Ｌｑ）ｉｄが小さくなるようにモータの機器定数またはセンサ・ゲインなどの定数（センサ定数）を変化させることで、位置検出誤差を削減することが可能となる。
【０１９９】
もちろん、正確な機器定数、センサ定数が設定された場合にとる値に調整することも可能である。
【０２００】
また、Ｒの変化に対する磁束長さを示す図９中（ａ）、およびＬｑの変化に対する磁束長さを示す図９中（ｂ）から分かるように、５％程小さな値に調整することで実際のインダクタンス値と等しくすることができる。
【０２０１】
さらに、±９０度以外の時はインダクタンスＬを小さくし、±９０度以内の時は、磁束の長さが短いとインダクタンスＬを小さく、長いと大きくし、±９０度以内で、磁束の位相が所定位相よりも大きい場合にはＲを大きくすることで０〜９０度の所定位相、所定磁束長さに制御することができる。
【０２０２】
図１５はこの発明のモータ駆動方法のさらに他の実施態様の要部を説明するフローチャートである。
【０２０３】
ステップＳＰ１において、従来公知の方法でモータトルクを算出し、ステップＳＰ２において、モータトルクに合わせた補正係数を参照し、ステップＳＰ３において、φｍ＋（Ｌｄ-Ｌｑ）ｉｄの長さによる機器定数の推定（例えば、図７の処理を参照）を行い、ステップＳＰ３において、補正係数による推定結果の補正を行い、そのまま一連の処理を終了する。
【０２０４】
なお、前記補正係数については、次のように設定することが励磁できる。
【０２０５】
図１１に示すように、差圧により負荷０の時の値からずれていることが分かる。
【０２０６】
また、先の議論から負荷０の時の値２１４０×１０μＨが真値であると考えられるので、例えば、差圧３ｋｇ/ｃｍ²の時のＬｑの誤差は３０×１０μＨで、約１．４％となる。
【０２０７】
したがって、例えば、１−０．００４７×差圧ｋｇ/ｃｍ²を補正係数（つまり、１ｋｇ/ｃｍ²で０．９９５３）とし、差圧により補正することによって正確な値を算出することができる。
【０２０８】
もちろん、差圧に代えて負荷トルクに基づく補正を行う補正係数を採用することも可能である。
【０２０９】
このフローチャートの処理を採用すれば、モータトルクに応じて設定値または修正量を変化させることができる。
【０２１０】
図１０、図１６に示したように、機器定数推定時のモータトルクによって機器定数推定値に若干の誤差がでる。そこで、モータトルクを検出し、この値によって機器定数を修正することにより、より正確な機器定数推定を可能にすることができる。
【０２１１】
モータトルクは前述したようにφ×I×ｓｉｎθφ_-I×Pnで与えられるが、φは電圧検出値の積分（抵抗によるドロップはＲｉとして補正）として、ｉは電流検出値として、φとｉの成す角θφ_-Iもφとｉから求められるので、機器定数やセンサレスの精度に依存しない。
このためモータトルクを機器定数推定時に利用することが可能である。
【０２１２】
【発明の効果】
請求項１の発明は、モータの機器定数、センサ定数などのロータ回転位置推定のための定数の誤差削減をきわめて容易に達成することができるという特有の効果を奏する。
【０２１３】
請求項２の発明は、不安定になり、最悪発散停止してしまうという不都合の発生を防止して、請求項１と同様の効果を奏する。
【０２１４】
請求項３の発明は、請求項１または請求項２の条件を容易に実現でき、ひいては請求項１または請求項２と同様の効果を奏する。
【０２１５】
請求項４の発明は、処理を簡単化することができ、請求項１から請求項３の何れかと同様の効果を奏する。
【０２１６】
請求項５の発明は、処理を簡単化することができ、請求項１から請求項４の何れかと同様の効果を奏する。
【０２１７】
請求項６の発明は、電圧位相を用いて請求項１から請求項３の何れかと同様の効果を奏する。
【０２１８】
請求項７の発明は、電流位相を用いて請求項１から請求項３の何れかと同様の効果を奏する。
【０２１９】
請求項８の発明は、確実に誤差の大きさを狭めることができ、ひいては請求項１から請求項７の何れかと同様の効果を奏する。
【０２２３】
請求項９の発明は、モータの機器定数、センサ定数などのロータ回転位置推定のための定数の誤差削減をきわめて容易に達成することができるという特有の効果を奏する。
【０２２４】
請求項１０の発明は、不安定になり、最悪発散停止してしまうという不都合の発生を防止して、請求項９と同様の効果を奏する。
【０２２５】
請求項１１の発明は、請求項９または請求項１０の条件を容易に実現でき、ひいては請求項９または請求項１０と同様の効果を奏する。
【０２２６】
請求項１２の発明は、処理を簡単化することができ、請求項９から請求項１１の何れかと同様の効果を奏する。
【０２２７】
請求項１３の発明は、処理を簡単化することができ、請求項９から請求項１２の何れかと同様の効果を奏する。
【０２２８】
請求項１４の発明は、電圧位相を用いて請求項９から請求項１１の何れかと同様の効果を奏する。
【０２２９】
請求項１５の発明は、電流位相を用いて請求項９から請求項１１の何れかと同様の効果を奏する。
【０２３０】
請求項１６の発明は、確実に誤差の大きさを狭めることができ、ひいては請求項９から請求項１５の何れかと同様の効果を奏する。
【０２３４】
請求項１７の発明は、十分に大きな電流を流しているような場合に、請求項１から請求項３の何れかと同様の効果を奏する。
【０２３５】
請求項１８の発明は、定数を正確にすることができるほか、請求項１７と同様の効果を奏する。
【０２３６】
請求項１９の発明は、十分に大きな電流を流しているような場合に、請求項１から請求項３の何れかと同様の効果を奏する。
【０２３７】
請求項２０の発明は、請求項１７から請求項１９の何れかと同様の効果を奏する。
請求項２１の発明は、請求項１７から請求項１９の何れかと同様の効果を奏する。
【０２３８】
請求項２２の発明は、十分に大きな電流を流しているような場合に、請求項９から請求項１１の何れかと同様の効果を奏する。
【０２３９】
請求項２３の発明は、定数を正確にすることができるほか、請求項２２と同様の効果を奏する。
【０２４０】
請求項２４の発明は、十分に大きな電流を流しているような場合に、請求項９から請求項１１の何れかと同様の効果を奏する。
【０２４１】
請求項２５の発明は、誤差をより低減することができ、しかも請求項２３と同様の効果を奏する。
請求項２６の発明は、誤差をより低減することができ、しかも請求項２４と同様の効果を奏する。
【図面の簡単な説明】
【図１】固定座標モータ逆モデルを用い、ロータ回転位置を検出するセンサを用いることなくモータを駆動する装置の一例を示すブロック図である。
【図２】電流位相に対するモータトルクの一例を示す図である。
【図３】負荷の大きさに対してモータ電流過大、かつモータ電流が遅れ位相の状態におけるベクトルシミュレーションの一例を示す図である。
【図４】電流位相に対するφｍ＋（Ｌｄ−Ｌｑ）Ｉｄの振幅の関係を示す図である。
【図５】この発明のモータ駆動方法の一実施態様の要部を説明するフローチャートである。
【図６】図１のモータ駆動装置の位置検出部５が位置検出を行った場合の巻線抵抗Ｒの変化に対する推定されたｄ軸からみた磁束の位相のシミュレーション結果、およびｑ軸インダクタンスＬｑの変化に対する推定されたｄ軸からみた磁束の位相のシミュレーション結果を示す図である。
【図７】この発明のモータ駆動方法のさらに他の実施態様の要部を説明するフローチャートである。
【図８】長さ検出処理を説明するフローチャートである。
【図９】Ｒの変化に対する磁束長さ、およびＬｑの変化に対する磁束長さを示す図である。
【図１０】負荷が大きい時における、Ｒの変化に対する磁束長さ、およびＬｑの変化に対する磁束長さを示す図である。
【図１１】圧縮機における差圧に対するＬｑ推定値の関係およびＬｑの実測値を示す図である。
【図１２】この発明のモータ駆動方法のさらに他の実施態様の要部を説明するフローチャートである。
【図１３】長さ位相検出処理を説明するフローチャートである。
【図１４】この発明のモータ駆動方法のさらに他の実施態様の要部を説明するフローチャートである。
【図１５】この発明のモータ駆動方法のさらに他の実施態様の要部を説明するフローチャートである。
【図１６】負荷が大きい時における、Ｒの変化に対する角度誤差、およびＬｑの変化に対する角度誤差を示す図である。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a motor driving method and an apparatus therefor, and more specifically, a rotor rotational position is estimated using a current and a voltage supplied to the motor and a motor device constant, and the current is determined based on the estimated rotor rotational position. Alternatively, the present invention relates to a motor driving method and apparatus for driving a motor by controlling the voltage and supplying the controlled current or voltage to the motor.
[0002]
[Prior art]
Conventionally, as a device suitable for driving a motor in harsh environments such as high temperature and high pressure, the rotor rotational position is estimated by using the current and voltage supplied to the motor and the motor's device constants. There has been proposed a motor drive device that controls the current or voltage based on the rotor rotational position and supplies the controlled current or voltage to the motor to drive the motor.
[0003]
In such a motor drive device, the design constant of the motor is adopted as it is, or the value adjusted at the time of production is adopted.
[0004]
[Problems to be solved by the invention]
Since the conventional motor driving device does not have a function for identifying the device constant of the motor at all, if the set device constant includes an error, the estimation accuracy of the rotor rotational position decreases. In addition, the desired motor drive (for example, motor drive with a desired efficiency) cannot be performed, and inconveniences such as the operable range become narrow.
[0005]
In order to eliminate the inconvenience that the operable range is narrowed, for example, a motor with a higher output may be adopted, but this disadvantageously causes an increase in cost.
[0006]
OBJECT OF THE INVENTION
The present invention has been made in view of the above problems, and an object of the present invention is to provide a motor driving method and apparatus capable of reducing errors included in constants for rotor rotational position estimation.
[0007]
[Means for Solving the Problems]
  The motor driving method according to claim 1 is controlled by estimating the rotor rotational position using the current and voltage supplied to the motor and the motor device constant, and controlling the current or voltage based on the estimated rotor rotational position. When driving the motor by supplying the current or voltage to the motor,
  Reduce the error of the constant for estimating the rotor rotational position with an excessive current flowing to the load size and the current in a lagging phase.Is the method.
[0008]
The motor driving method according to claim 2 is a method for reducing an error by forcibly applying a rotating magnetic field to the stator regardless of the rotor rotational position.
[0009]
According to a third aspect of the present invention, the motor driving method forcibly applies a rotating magnetic field to the stator regardless of the rotor rotational position, starts the motor, and reduces the constant error for estimating the rotor rotational position at the time of starting. .
[0010]
The motor driving method according to claim 4 reduces the error so that the phase of the interlinkage magnetic flux of the motor coil falls within a range of about 0 to 90 degrees in the rotation direction with reference to the d-axis viewed from the estimation result of the rotor rotational position. Is the method.
[0011]
The motor driving method according to claim 5 reduces the error by reducing the resistance when the phase of the interlinkage magnetic flux of the motor coil is in the minus direction with reference to the d-axis viewed from the estimation result of the rotor rotational position. When the angle is in the range of approximately 90 to 180 degrees, the error is reduced by reducing the q-axis inductance.
[0012]
The motor driving method according to claim 6 is a method for reducing an error so that a voltage phase obtained by estimating the rotor rotational position falls within a range of approximately 90 to 180 degrees on the basis of the d axis on average.
[0013]
The motor driving method according to claim 7 is a method for reducing an error so that a current phase obtained by estimating the rotor rotational position falls within a range of approximately 0 to 90 degrees in the rotational direction with respect to the d axis.
[0014]
  The motor driving method according to claim 8 is set for estimating the rotor rotational position.q-axis inductance Lq, phase resistance R Set a value larger than the center value of the variation due to the motor of these values.OrConstants for current and voltage detection, q-axis inductance Lq, phase resistance R Is set with a deviation in the direction equivalent to setting these values to be larger than the center value of the variation due to the motor.Is the method.
[0018]
  Claim9The motor driving device estimates the rotor rotational position using the current and voltage supplied to the motor and the motor device constant, controls the current or voltage based on the estimated rotor rotational position, and controls the controlled current or In what drives the motor by supplying voltage to the motor,
    Reduce the error of the constant for estimating the rotor rotational position with an excessive current flowing to the load size and the current in a lagging phase.It includes constant error reduction means.
[0019]
  Claim10In this motor driving device, the constant error reducing means employs a device that forcibly applies a rotating magnetic field to the stator regardless of the rotor rotational position to reduce the error.
[0020]
  Claim11As the constant error reduction means, the motor driving device forcibly applies a rotating magnetic field to the stator regardless of the rotor rotational position to start the motor, and reduces the constant error for estimating the rotor rotational position at the time of startup. Adopt what you do.
[0021]
  Claim12As the constant error reduction means, the motor drive device of the motor has an error so that the phase of the flux linkage of the motor coil falls within a range of about 0 to 90 degrees in the rotation direction with reference to the d axis as seen from the estimation result of the rotor rotational position. The one that reduces is used.
[0022]
  Claim13As the constant error reducing means, when the phase of the interlinkage magnetic flux of the motor coil is in the minus direction with respect to the d axis viewed from the estimation result of the rotor rotational position, the error is reduced by reducing the resistance. If the angle is in the range of approximately 90 to 180 degrees, the error is reduced by reducing the q-axis inductance.
[0023]
  Claim14As the constant error reducing means, the motor driving apparatus reduces the error so that the voltage phase obtained by estimating the rotor rotational position falls within a range of approximately 90 to 180 degrees on the basis of the d axis on average. Is adopted.
[0024]
  Claim15In the motor driving apparatus, the constant error reducing means reduces the error so that the current phase obtained by estimating the rotor rotational position falls within a range of approximately 0 to 90 degrees in the rotational direction with respect to the d axis. Adopted.
[0025]
  Claim16The motor driving apparatus is set as the constant error reducing means for estimating the rotor rotational position.q-axis inductance Lq, phase resistance R Set a value larger than the center value of the variation due to the motor of these values.OrConstants for current and voltage detection are set with a deviation in a direction equivalent to setting the values of the q-axis inductance Lq and the phase resistance R to be larger than the center value of the variation of these values due to the motor.The thing is adopted.
[0029]
  Claim17In this motor driving method, the motor device constant or the motor constant is reduced so that the estimated amount corresponding to the result obtained by adding the difference between the d-axis inductance and the q-axis inductance to the magnet flux linkage and the d-axis current is reduced. This is a method of reducing errors by adjusting sensor constants.
[0030]
  Claim18After the constant adjustment,Correction is performed using the magnitude of the load that is assumed in advance for the adjusted device constant or detected by other means.Is the method.
[0031]
  Claim19The motor driving method is a value set in advance with reference to the d-axis as viewed from the rotor rotational position where the phase of the winding flux linkage is estimated using the current and voltage supplied to the motor and the motor device constants. And reducing the error so that the estimated amount corresponding to the result of adding the value obtained by multiplying the difference between the d-axis inductance and the q-axis inductance by the d-axis current to the magnetic flux linkage becomes a preset value. It is a method to do.
[0032]
  Claim20The motor drive method depends on the motor torqueCorrection amountIt is a method to change.
  A motor driving method according to a twenty-first aspect is a method of changing a set value in accordance with a motor torque.
[0033]
  Claim22In the motor drive apparatus, the estimated error corresponding to the result obtained by adding the value obtained by multiplying the difference between the d-axis inductance and the q-axis inductance to the difference between the d-axis inductance and the d-axis current is reduced as the constant error reduction means. As described above, a device that reduces the error by adjusting the device constant or sensor constant of the motor is adopted.
[0034]
  Claim23After the constant adjustment, as the constant error reduction means,Make adjustments using the magnitude of the load assumed in advance or detected by other means for the adjusted device constants.Adopt a fish.
[0035]
  Claim24As the constant error reduction means, the motor drive apparatus of No. 1 has a d-axis as viewed from the rotor rotational position in which the phase of the winding flux linkage is estimated using the current and voltage supplied to the motor and the motor device constant. A value set in advance as a reference, and an estimated amount corresponding to a result obtained by adding a value obtained by multiplying the magnetic flux linkage by the difference between the d-axis inductance and the q-axis inductance by the d-axis current is set in advance The one that reduces the error is adopted.
[0036]
  Claim25According to the motor torque, the motor drive device ofCorrection amountThe one to change is adopted.
  The motor driving apparatus according to claim 26 employs, as the constant error reducing means, one that changes a set value in accordance with motor torque..
[0037]
[Action]
  The motor driving method according to claim 1 is controlled by estimating the rotor rotational position using the current and voltage supplied to the motor and the motor device constant, and controlling the current or voltage based on the estimated rotor rotational position. When driving the motor by supplying the current or voltage to the motor,Since an excessive current with respect to the size of the load is passed and the current is in a lagging phase, the constant error for estimating the rotor rotational position is reduced.Error reduction of constants for estimating the rotor rotational position such as motor constants and sensor constants can be achieved very easily.
[0038]
In other words, paying attention to the fact that the error looks larger as the current is passed, the error is made larger by flowing a current larger than the current during normal operation such as optimum efficiency and maximum torque constant. Reduction can be easily achieved. In addition, the accuracy of error reduction can be increased.
[0039]
According to the motor driving method of claim 2, since the error is reduced by forcibly applying a rotating magnetic field to the stator regardless of the rotor rotational position, it becomes unstable and the divergence is stopped in the worst case. Generation | occurrence | production can be prevented and the effect | action similar to Claim 1 can be achieved.
[0040]
According to the motor driving method of claim 3, the motor is started by forcibly applying a rotating magnetic field to the stator regardless of the rotor rotational position, and at the time of starting, a constant error for estimating the rotor rotational position is reduced. Therefore, the conditions of claim 1 or claim 2 can be easily realized, and as a result, the same effect as that of claim 1 or claim 2 can be achieved.
[0041]
According to the motor driving method of claim 4, the error is reduced so that the phase of the interlinkage magnetic flux of the motor coil falls within a range of about 0 to 90 degrees in the rotation direction with reference to the d-axis viewed from the estimation result of the rotor rotational position. Therefore, the processing can be simplified and the same operation as any one of claims 1 to 3 can be achieved.
[0042]
According to the motor driving method of claim 5, when the phase of the interlinkage magnetic flux of the motor coil is in the minus direction with reference to the d axis viewed from the estimation result of the rotor rotational position, the error is reduced by reducing the resistance. When the angle is in the range of about 90 to 180 degrees, the error can be reduced by reducing the q-axis inductance, so that the processing can be simplified. A similar effect can be achieved.
[0043]
According to the motor driving method of the sixth aspect, the error is reduced so that the voltage phase obtained by estimating the rotor rotational position falls within a range of approximately 90 to 180 degrees on the basis of the d axis on average. Using the voltage phase, it is possible to achieve the same effect as any one of claims 1 to 3.
[0044]
According to the motor driving method of claim 7, the error is reduced so that the current phase obtained by estimating the rotor rotational position falls within a range of approximately 0 to 90 degrees in the rotational direction with respect to the d axis. Using the current phase, it is possible to achieve the same effect as any one of the first to third aspects.
[0045]
  If it is the motor drive method of Claim 8, it will set for estimation of a rotor rotational position.Set the values of q-axis inductance Lq and phase resistance R to be larger than the center value of the variation of these values due to the motor.OrConstants for current and voltage detection are set with a deviation in a direction equivalent to setting the values of the q-axis inductance Lq and the phase resistance R to be larger than the center value of the variation of these values due to the motor.Therefore, the magnitude of the error can be surely reduced, and as a result, the same operation as any one of claims 1 to 7 can be achieved.
[0049]
  Claim9If the motor driving device of the motor is controlled, the rotor rotational position is estimated using the current and voltage supplied to the motor and the motor device constant, and the current or voltage is controlled based on the estimated rotor rotational position. In driving the motor by supplying current or voltage to the motor,
  Reduce the error of the constant for estimating the rotor rotational position with an excessive current flowing to the load size and the current in a lagging phase.be able to.
[0050]
Therefore, it is very easy to reduce errors in constants for estimating the rotor rotational position such as motor device constants and sensor constants.
[0051]
  Claim10In this motor drive device, the constant error reducing means that employs a device that forcibly applies a rotating magnetic field to the stator regardless of the rotor rotational position to reduce the error becomes unstable and worst. To prevent the inconvenience of stopping the divergence,9The same effect can be achieved.
[0052]
  Claim11If the motor driving device is a constant error reduction means, the motor is started by forcibly applying a rotating magnetic field to the stator regardless of the rotor rotational position, and the constant error for estimating the rotor rotational position is Since the one to reduce is adopted, the claim9Or claims10Can easily be realized, and as a result the claims9Or claims10The same effect can be achieved.
[0053]
  Claim12As the constant error reducing means, the phase of the interlinkage magnetic flux of the motor coil is in the range of about 0 to 90 degrees in the rotation direction with reference to the d axis as seen from the estimation result of the rotor rotational position. Since the one that reduces the error is adopted, the processing can be simplified and the claims can be made.9Claims from11The same action as any of the above can be achieved.
[0054]
  Claim13If the motor driving device is a constant error reducing means, the resistance is reduced when the phase of the flux linkage of the motor coil is in the minus direction with reference to the d-axis viewed from the estimation result of the rotor rotational position. In this case, the error can be reduced, and if it is in the range of approximately 90 to 180 degrees, the error can be reduced by reducing Lq, so that the processing can be simplified.9Claims from12The same action as any of the above can be achieved.
[0055]
  Claim14In the case of the motor drive apparatus, the constant error reducing means reduces the error so that the voltage phase obtained by estimating the rotor rotational position falls within a range of approximately 90 to 180 degrees on the basis of the d axis on average. Since what is to be adopted is used, the voltage phase is used to claim9Claims from11The same action as any of the above can be achieved.
[0056]
  Claim15In the motor driving apparatus, the constant error reducing means reduces the error so that the current phase obtained by estimating the rotor rotational position falls within the range of approximately 0 to 90 degrees in the rotational direction with respect to the d axis. So that the current phase is used for the claim.9Claims from11The same action as any of the above can be achieved.
[0057]
  Claim16In this case, the constant error reducing means is set for estimating the rotor rotational position.Set the values of q-axis inductance Lq and phase resistance R to be larger than the center value of the variation of these values due to the motor.OrConstants for current and voltage detection are set with a deviation in the direction equivalent to setting the values of q-axis inductance Lq and phase resistance R to be larger than the center value of the variation of these values due to the motor.So that the size of the error can be reliably reduced, and as a result the claims9Claims from15The same action as any of the above can be achieved.
[0063]
  Claim17In this motor driving method, the motor device is such that the estimated amount corresponding to the result obtained by adding the value obtained by multiplying the difference between the d-axis inductance and the q-axis inductance by the d-axis current to the magnetic flux linkage is reduced. Since the error is reduced by adjusting the constant or the sensor constant, the same action as any one of claims 1 to 3 can be achieved when a sufficiently large current is flowing. .
[0064]
  Claim18If the motor drive method ofCorrection is performed using the magnitude of the load that is assumed in advance for the adjusted device constant or detected by other means.In addition to making the constants accurate, the claims17The same effect can be achieved.
[0065]
  Claim19In this motor driving method, the phase of the winding flux linkage is preset based on the d-axis viewed from the rotor rotational position estimated using the current and voltage supplied to the motor and the motor device constants. So that the estimated amount corresponding to the result of adding the value obtained by multiplying the difference between the d-axis inductance and the q-axis inductance by the d-axis current to the magnetic flux linkage becomes a preset value. Therefore, when a sufficiently large current is flowing, the same operation as that of any one of claims 1 to 3 can be achieved.
[0066]
  Claim20The motor drive method depends on the motor torqueCorrection amountThe error can be further reduced, and the claim17Claims from19The same action as any of the above can be achieved.
  Since the motor drive method of claim 21 changes the set value in accordance with the motor torque, the error can be further reduced, and the same effect as any of claims 17 to 19 can be achieved. Can do.
[0067]
  Claim22In the case of the motor drive apparatus, the constant error reduction means has an estimated amount corresponding to a result obtained by adding a value obtained by multiplying the difference between the d-axis inductance and the q-axis inductance by the d-axis current to the magnetic flux linkage. Since a device that reduces the error by adjusting the device constant or the sensor constant of the motor so as to be reduced is adopted, when a sufficiently large current is flowing, the claims 9 to11The same action as any of the above can be achieved.
[0068]
  Claim23If the motor drive device is, as the constant error reduction means,Make adjustments using the magnitude of the load assumed in advance or detected by other means for the adjusted device constants.Because it adopts the other, the constant can be made accurate and the claims22The same effect can be achieved.
[0069]
  Claim24As the constant error reduction means, the phase of the winding flux linkage is estimated from the rotor rotational position estimated using the current and voltage supplied to the motor and the motor device constant as d. An estimated amount corresponding to a result obtained by adding a value obtained by multiplying the magnetic flux linkage by the difference between the d-axis inductance and the q-axis inductance to the difference between the d-axis inductance and the d-axis current is preset. In order to reduce the error so that it becomes the same value, it is adopted.9Claims from11The same action as any of the above can be achieved.
[0070]
In the motor driving device according to claim 25, since the constant error reducing means is a device that changes the correction amount according to the motor torque, the error can be further reduced, and23The same effect can be achieved.
According to the motor driving device of claim 26, since the constant error reducing means that changes the set value according to the motor torque is adopted, the error can be further reduced, and24The same effect can be achieved.
[0071]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of a motor driving method and apparatus according to the present invention will be described below in detail with reference to the accompanying drawings.
[0072]
FIG. 1 is a block diagram showing an example of an apparatus for driving a motor using a fixed coordinate motor inverse model without using a sensor for detecting a rotor rotational position.
[0073]
In this motor drive device, an AC power supply 1 is supplied to a rectifier circuit 2 to be DC, smoothed by a smoothing capacitor 2 a, and then converted into a three-phase AC by a three-phase inverter 3 and supplied to a three-phase motor 4.
[0074]
Then, the current detection unit 5a detects each phase motor current, and the voltage detection unit 5b detects each phase motor voltage, and supplies the motor current for three phases to the first three-phase → two-phase conversion unit 5c. The motor current for two phases is converted to a motor voltage for three phases, and the motor voltage for three phases is supplied to the second three-phase to two-phase converter 5d to convert the motor voltage for two phases.
[0075]
Then, the motor current for two phases and the motor voltage for two phases are supplied to the fixed coordinate motor inverse model unit 5e, and the output from the fixed coordinate motor inverse model unit 5e is supplied to the position calculation unit 5f, thereby rotating the rotor position. Is output.
[0076]
The rotor rotational position is supplied to the differentiating unit 5g to calculate the speed, the calculated speed and the speed command are supplied to the speed control unit 5h, the speed control calculation is performed, and the current command is output.
[0077]
The current command and the phase command are supplied to the phase control unit 5i, the phase control calculation is performed and the final current command is output, and the final current command, the motor current for three phases and the rotor rotation position are supplied to the current control unit 5j. Then, a current control calculation is performed to output a voltage command and supplied to the three-phase inverter 3.
[0078]
The three-phase to two-phase converters 5c and 5d multiply the three-phase motor current and the three-phase motor voltage by Equation 1 to obtain the αβ current vector and αβ voltage vector in the two-phase orthogonal stator coordinates. Output.
[0079]
[Expression 1]

[0080]
The fixed coordinate motor inverse model unit 5e includes a resistance multiplication unit 5e1 that multiplies the αβ current vector by a phase resistance to calculate a voltage drop, a first subtraction unit 5e2 that subtracts the voltage drop from the αβ voltage vector, 1 has an integration unit 5e3 that integrates the output from the subtraction unit 5e2, an inductance multiplication unit 5e4 that multiplies the αβ current vector by the q-axis inductance, and a second subtraction unit 5e5 that subtracts the q-axis inductance multiplication result from the integration result. is doing.
[0081]
The position calculation unit 5f performs tan-1 processing on the output from the fixed coordinate motor inverse model unit 5e and outputs the rotor rotation position.
[0082]
In the following description, the three-phase to two-phase conversion units 5a and 5b, the fixed coordinate motor inverse model unit 5e, and the position calculation unit 5f are collectively referred to as a position detection unit 5.
[0083]
If the phase resistance and the q-axis inductance set in the fixed coordinate motor inverse model unit 5a included in the motor drive device are accurate and the current detection unit 5a and the voltage detection unit 5b are accurate, the rotor rotational position is accurate. The motor 4 can be operated stably and accurately using the estimated rotor rotational position.
[0084]
However, in general, the current detection unit 5a and the voltage detection unit 5b have errors due to variations, temperature changes, and / or the like, and / or motor device constants set in the fixed coordinate motor inverse model unit 5e. Therefore, the estimation result of the rotor rotational position becomes inaccurate, and the motor 4 cannot be operated stably and accurately.
[0085]
In order to eliminate such an inconvenience, the present invention provides constants (for example, gains) of the current detector 5a and the voltage detector 5b when the motor current is excessive with respect to the load and the motor current is in a delayed phase. ), An error reduction unit 6 for reducing at least one of the motor device constants is provided.
[0086]
Further explanation will be given.
[0087]
An example of the motor torque with respect to the current phase is shown in FIG. This example is based on a brushless DC motor having a rotor in which a permanent magnet is embedded.
[0088]
In FIG. 2, the current phase 0 ° is when the motor current is flowing in the q-axis direction (when it is in phase with the induced voltage), and the plus side indicates the phase advance.
[0089]
When the motor current is large with respect to the load and the phase is delayed, the motor is operated at a current phase (see a circle in FIG. 2) in which torque is hardly generated with respect to the motor current.
[0090]
FIG. 3 shows an example of a vector simulation in a state where the motor current is excessive with respect to the load and the motor current is in a delayed phase.
[0091]
The d-axis is the direction of the magnetic flux of the magnet, and the q-axis is taken in a phase that is electrically advanced by 90 degrees (that is, the same phase as the induced voltage). I is a motor current vector, which is delayed with respect to the q axis. E is an induced voltage due to magnet magnetic flux, ω is a rotation angular frequency (electrical angle), Lq and Ld are q-axis and d-axis inductance, Id and Iq are d-axis and q-axis components of motor current I, R is winding resistance, V represents a motor voltage, and φ represents a winding flux linkage. The character represented by the subscript m is a motor value and is distinguished from the value in the control.
[0092]
In the motor drive device shown in FIG. 1, the length in the d-axis direction is φm + (Ld−Lq) Id based on the motor current, motor voltage, winding resistance R, and q-axis inductance Lq supplied to the motor. And the rotor position is detected from the angle of the vector.
[0093]
FIG. 4 shows the relationship of the amplitude of φm + (Ld−Lq) Id with respect to the current phase. This relationship was obtained by setting Ld = 11.0 mH, Lq = 23.9 mH, R = 0.7945Ω, and φ = 0.149 wb.
[0094]
When the number of pole pairs is Pn, the motor torque is
{Φm + (Ld−Lq) Id} Id * Pn = φIsin θφ−IPn
Here, θφ−I is an angle formed by the winding linkage magnetic flux φ and the motor current I. When the current is large with respect to the torque, as shown in FIG. 3, Iq ≠ 0, so φm + (Ld−Lq) Id becomes extremely small. For this reason, when the current is excessive with respect to the size of the load, an extremely large angle error is generated due to an error in the motor device constants and constants of the current detection unit 5a and the voltage detection unit 5b. Then, it is extremely difficult to perform stable position detection. However, on the contrary, since the error appears as a large angular deviation, the error reduction unit 6 makes it very easy to reduce the error in the motor device constants, the current detection unit 5a, and the voltage detection unit 5b using this operating condition. can do.
[0095]
In this case, it is preferable to reduce the error by forcibly applying a rotating magnetic field to the stator regardless of the rotor position.
[0096]
Further explanation will be given.
[0097]
As described above, the output of the position detection unit 5 is extremely unstable under the operating conditions in which the current is excessive with respect to the size of the load and the current is in a delayed phase, and the motor is operated by the output of the position detection unit 5. Then, it becomes unstable and stops the worst divergence. Therefore, instability can be eliminated by reducing the error while forcibly applying a rotating magnetic field to the stator and rotating the motor without using the output of the position detector. In addition, normally, there is a concern that if the operation is performed with a forced rotating magnetic field, the motor will step out due to torque fluctuations, etc., but in this case, a sufficiently large current flows compared to the load, so the escape torque Therefore, the problem of step-out does not occur.
[0098]
In particular, when a rotating magnetic field is forcibly applied to the stator, as shown in FIG. 2, a large current portion appears with respect to torque on the advance phase side (approximately 90 ° point), but the rotor rotates on the advance phase side. This is an unstable point in which the phase is moved to the lag side and torque is generated and the phase is further delayed. Therefore, the phase is eventually stabilized only on the lag phase side. For this reason, if a rotating magnetic field is forcibly applied without phase control, operation is automatically performed with a delayed phase. As a result, instability can be eliminated and errors can be reduced.
[0099]
Further, the motor is started by forcibly applying a rotating magnetic field to the stator without using the output of the position detector 5, and the motor device constants, the current detector 5a, and the voltage detector 5b are constants. It is preferable to reduce the error at startup. That is, when starting by forcibly applying a rotating magnetic field to the stator and rotating the motor, the load torque at the time of starting is equal to or less than the escape torque, and driving is performed without using the position detection unit 5 at the time of starting. Any of the above-mentioned conditions can be easily realized, which is suitable for reducing errors.
[0100]
Furthermore, it is preferable to reduce the error so that the phase of the interlinkage magnetic flux of the motor coil falls within the range of approximately 0 ° to 90 ° in the rotation direction with reference to the d axis viewed from the estimation result of the position detection unit 5. .
[0101]
Further explanation will be given.
[0102]
The magnet magnetic flux φm is a vector oriented in the d-axis direction, and the d and q-axis components (φd, φq) of the winding linkage magnetic flux φ are (φm + LdId, LqIq).
[0103]
As shown in FIG. 3, when the current is large and the phase is delayed with respect to the load, since Id is positive (lagging phase) and Iq is positive (powering), the d-axis of the winding linkage flux is used as a reference. The above phase always exists in the range of 0 ° to 90 °.
[0104]
Accordingly, when the current is excessive with respect to the magnitude of the load and the phase is delayed, the estimated magnetic flux phase is approximately 0 ° to 90 ° in the rotation direction with respect to the d axis, and the motor of the position detection unit 5 is moved when it deviates. The error can be reduced by adjusting the constants of the device constant, the current detector 5a, and the voltage detector 5b.
[0105]
In any of the above cases, when the phase of the interlinkage magnetic flux of the motor coil is in the minus direction with reference to the d-axis viewed from the estimation result of the position detection unit 5, the resistance is reduced, and approximately 90 ° to 180 °. If it is within the range, it is preferable to reduce the q-axis inductance Lq.
[0106]
This will be further described with reference to the flowchart of FIG.
[0107]
In step SP1, the motor current and the motor voltage are detected. In step SP2, the rotor rotational position (position in the d-axis direction) is detected (estimated) using the motor current, the motor voltage, and the motor device constants. In SP3, the phase of the winding linkage magnetic flux is calculated from the rotor rotational position detection result and the calculated direction of the motor magnetic flux vector. In step SP4, whether or not the phase of the winding linkage magnetic flux is smaller than 0 °. Determine.
[0108]
If it is determined in step SP4 that the phase of the winding linkage magnetic flux is smaller than 0 °, the winding resistance R of the motor device constants set in the position detection unit 5 is decreased in step SP5. Step SP1 is performed again.
[0109]
If it is determined in step SP4 that the phase of the winding linkage magnetic flux is not smaller than 0 °, it is determined in step SP6 whether or not the phase of the winding linkage magnetic flux is larger than 90 °.
[0110]
If it is determined in step SP6 that the phase of the winding linkage magnetic flux is greater than 90 °, in step SP7, the q-axis inductance Lq is reduced among the device constants of the motor set in the position detector 5, The process of step SP1 is performed again.
[0111]
If it is determined in step SP6 that the phase of the winding linkage magnetic flux is not greater than 90 °, the series of processing is terminated as it is.
[0112]
Further explanation will be given.
[0113]
FIG. 6a shows the simulation result of the phase of the magnetic flux as seen from the d-axis with respect to the change in the winding resistance R when the position detection unit 5 of the motor drive device of FIG. Shows the simulation results of the phase of the magnetic flux as seen from the estimated d-axis with respect to the change in the q-axis inductance Lq.
[0114]
From FIG. 6a, it can be detected that there is an error because the phase of the magnetic flux becomes negative when the winding resistance of the motor is smaller than about -15%. Further, from FIG. 6b, it can be detected that when the q-axis inductance Lq of the motor is smaller than about −5%, the phase becomes 90 ° to 180 ° and the q-axis inductance Lq of the motor is small. Therefore, the resistance is reduced when the phase of the interlinkage magnetic flux of the motor coil is in the minus direction with reference to the d-axis viewed from the estimation result of the position detector 5, and when the phase is in the range of approximately 90 ° to 180 °. By reducing the q-axis inductance Lq, the error between the motor and the device constants of the position detector 5 can be reduced.
[0115]
In order to further improve the accuracy, it is only necessary to reduce the error in a state where the estimation error appears in the position detection unit. Therefore, it is understood that the motor current may be excessively applied to the load. For example, if the current is further increased by 30% in the simulation of FIG. 6, the phase of the magnetic flux becomes negative and the error can be detected when the winding resistance of the motor is smaller than about −10%.
[0116]
Although the discussion has been made here assuming that there is an error in the device constant, the constant error in the current detection unit 5a and the voltage detection unit 5b can also be reduced. If the constants of the current detection unit 5a and the voltage detection unit 5b and the device constants have errors at the same time, they cannot be separated, but the error is reduced in the combined state, and the estimation error of the position detection unit 5 is reduced. Can be reduced.
[0117]
Furthermore, it is preferable to reduce the error so that the estimated voltage phase by the position detection unit 5 falls within a range of approximately 90 ° to 180 ° on the basis of the d axis on average.
[0118]
That is, as shown in FIG. 3, the voltage phase is a vector obtained by adding the voltage drop due to the winding resistance to the derivative of the magnetic flux. Here, if the voltage drop due to the winding resistance is small, the voltage phase is 90 ° from the winding linkage magnetic flux and is located in the range of about 90 ° to 180 ° from the d-axis. Therefore, the error reduction of the motor device constant in the position detection unit 5, the current detection unit 5a, and the voltage detection unit 5b can be realized using the voltage phase in the same manner as described above.
[0119]
Further, it is preferable to reduce the error so that the estimated current phase falls within a range of approximately 0 ° to 90 ° in the rotation direction with respect to the d-axis.
[0120]
Further explanation will be given.
[0121]
As described above, the motor torque is represented by φxIxsin θ φ-IPn. Here, since the motor current is sufficiently large with respect to the torque, the lengths of the I vector and the φ vector are long, and the angle formed by the I vector and the φ vector is small. Since the φ vector is located in the range of 0 ° to 90 °, the I vector is also located in the range of about 0 ° to 90 °, and the position detector 5 is used using the current phase in the same manner as described above. It is possible to realize reduction of errors in the motor device constants, constants of the current detection unit 5a, and the voltage detection unit 5b.
[0122]
In each of the above embodiments, the q-axis inductance Lq and winding resistance R set in the position detection unit 5 are set to be larger than the center value of the motor variation, or the constants of the current detection unit 5a and the voltage detection unit 5b are set. It is preferable to set a deviation in the equivalent direction.
[0123]
Further explanation will be given.
[0124]
As described above, in each of the above embodiments, it is possible to detect that the winding resistance R and q-axis inductance Lq of the motor are small, and to correct the set values in the position detection unit 5. However, if the motor winding resistance R and q-axis inductance Lq are large, it is difficult to reduce the error. For this reason, the magnitude | size of an error can be narrowed more reliably by setting large the value of q-axis inductance Lq and winding resistance R which are preset to the position detection part 5. FIG. Further, when adjusting by the constants of the current detection unit 5a and the voltage detection unit 5b, for example, in the motor drive device of FIG. 1, an equivalent direction is set by setting the current larger than the actual value or setting the voltage smaller. Can produce deviations.
[0125]
Then, by performing these processes, error reduction can be easily realized.
[0151]
  Figure7These are the flowcharts explaining the principal part of other embodiment of the motor drive method of this invention.
[0152]
In step SP1, length detection processing is performed. In step SP2, the q-axis inductance Lq is increased. In step SP3, length detection processing is performed. In step SP4, is the current length longer than the previous length? Determine whether or not.
[0153]
If it is determined that the current length is not longer than the previous length, step SP2 is performed again.
[0154]
Conversely, if it is determined in step SP4 that the current length is longer than the previous length, the q-axis inductance Lq is decreased in step SP5, a length detection process is performed in step SP6, and step SP7. It is determined whether or not the current length is longer than the previous length.
[0155]
If it is determined that the current length is not longer than the previous length, step SP5 is performed again.
[0156]
Conversely, if it is determined in step SP7 that the current length is longer than the previous length, the series of processing ends.
[0157]
  Figure8FIG. 6 is a flowchart illustrating the length detection process.
[0158]
In step SP1, the motor current and the motor voltage are detected. In step SP2, the magnetic pole position of the rotor of the motor is detected. In step SP3, the d-axis inductance Ld and the q-axis inductance Lq with respect to the magnetic flux linkage φm. As a result of adding a value obtained by multiplying the difference between the two by the d-axis current id, the length (size) of φm + (Ld−Lq) id is detected, and the series of processing ends.
[0159]
Further explanation will be given.
[0160]
As described above, when the current is large with respect to the torque, φm + (Ld−Lq) id becomes extremely small as shown in FIG. 4. Conversely, when a sufficiently large current is flowing, φm + ( By changing a constant (sensor constant) such as a motor device constant or a sensor gain so that (Ld−Lq) id is minimized, a position detection error can be reduced.
[0161]
  Figure9Sensorless error against motor device constant error when a sufficiently large current is flowing in comparison with the load (current flowing 10 times or more compared with the current phase that generates the maximum torque) The size of φm + (Ld−Lq) id calculated by the control is shown. Figure9As can be seen, the magnitude of φm + (Ld−Lq) id calculated at the point where the error of Lq is almost zero is minimized.
[0162]
Therefore, conversely, it can be seen that the correct Lq can be obtained by adjusting Lq so that the calculated φm + (Ld−Lq) id is minimized. In addition, the amount corresponding to φm + (Ld−Lq) id can be obtained by ∫ (V−Ri) dt−Lqi in sensorless control (which controls the inverter by detecting the position without using the position sensor).
[0163]
  Therefore, figure7The figure8By performing the processing of the flowchart of (2), errors included in the q-axis inductance Lq can be reduced.
[0164]
Further, it is preferable to perform a predetermined correction on the adjustment result after the constant adjustment.
[0165]
  FIG. 10 shows the magnitude of φm + (Ld−Lq) id calculated when the load is slightly large (a current that is about five times that of the current phase that generates the maximum torque is passed). Show.
[0166]
It can be seen that the minimum value of φm + (Ld−Lq) id calculated for the error of Lq is slightly shifted to the minus side.
[0167]
  Figure for R9, 10Both have shifted to the negative side, but10It can be seen that there is a greater shift.
[0168]
  Also, the result of estimating the equipment constant while changing the load (differential pressure) in the compressor11Shown in
[0169]
  Figure11From this, it can be seen that the higher the load, the larger the estimated value of Lq.
[0170]
From this, it is understood that when a load at the time of activation is assumed, a correct device constant can be obtained by correcting the device constant calculated by the above-described method according to the state of the load.
[0171]
  Figure12These are the flowcharts explaining the principal part of other embodiment of the motor drive method of this invention.
[0172]
In step SP1, Lq is set small. In step SP2, length phase detection processing is performed. In step SP3, it is determined whether or not the detected phase is smaller than a predetermined phase.
[0173]
If it is determined that the detected phase is smaller than the predetermined phase, R is decreased in step SP4.
[0174]
Conversely, when it is determined in step SP3 that the detected phase is not smaller than the predetermined phase, it is determined in step SP5 whether or not the detected phase is larger than the predetermined phase.
[0175]
If it is determined that the detected phase is greater than the predetermined phase, R is increased in step SP6.
[0176]
Conversely, when it is determined in step SP5 that the detected phase is not greater than the predetermined phase, it is determined in step SP7 whether or not φm + (Ld−Lq) id is shorter than a predetermined value.
[0177]
If it is determined that φm + (Ld−Lq) id is shorter than a predetermined value, Lq is decreased in step SP8.
[0178]
Conversely, if it is determined in step SP7 that φm + (Ld−Lq) id is not shorter than the predetermined value, it is determined in step SP9 whether φm + (Ld−Lq) id is longer than the predetermined value. .
[0179]
If it is determined that φm + (Ld−Lq) id is longer than a predetermined value, Lq is increased in step SP10.
[0180]
In addition, when the process of step SP4 is performed, when the process of step SP6 is performed, when the process of step SP8 is performed, when the process of step SP10 is performed, the process of step SP1 is performed again. Do.
[0181]
Conversely, if it is determined in step SP9 that φm + (Ld−Lq) id is not longer than the predetermined value, the series of processing is terminated as it is.
[0182]
  Figure13FIG. 6 is a flowchart for explaining length phase detection processing;
[0183]
In step SP1, the motor current and the motor voltage are detected. In step SP2, the magnetic pole position of the rotor of the motor is detected. In step SP3, the phase of the magnetic flux is detected. In step SP4, the magnetic flux linkage φm is detected. As a result of adding the value obtained by multiplying the difference between the d-axis inductance Ld and the q-axis inductance Lq by the d-axis current Id, the length (size) of φm + (Ld−Lq) Id is detected, and the series of processing is finished as it is. To do.
[0184]
  Figure12The figure13By performing the processing of the flowchart, the phase difference of the interlinkage magnetic flux of the motor coil is set to a value set in advance with reference to the position sensorless estimated d-axis, and the estimated φm + (Ld−Lq) id It is possible to reduce the error so that the amount corresponding to 1 becomes a predetermined value set in advance.
[0185]
  FIG. 6 shows the estimated flux linkage angle and the magnitude of φm + (Ld−Lq) id under the same conditions, and FIG.7This will be further described with reference to FIG.
[0186]
  First, Lq is set to a small value in the control (FIG. 6, FIG.9Is expressed assuming that a deviation occurs in the motor constants, it should be noted that the sign of the change in the constants in the control is reversed. That is, setting the device constant small in the control corresponds to increasing the device constant of the motor), the length of φm + (Ld−Lq) id and the phase of the flux linkage of the motor coil are calculated. The phase can be adjusted to a predetermined value by decreasing R in response to being smaller than the predetermined phase and increasing R in response to being larger. For example, the predetermined value is set to 60 degrees in phase and 0.015 in length, and Lq is reduced.6If the position of Lq 10% and R 10% in the middle (a) is reached, the control R is decreased in response to the small phase (that is, R in FIG. 6 (a) written by the motor constant). This is equivalent to tracing the Lq 10% line to the right and reaching the position of 60 degrees.
[0187]
Then, while performing phase adjustment by R, Lq is decreased in response to φm + (Ld−Lq) id being shorter than the predetermined value, and in response to φm + (Ld−Lq) id being longer than the predetermined value. By increasing Lq, the desired phase and length can be controlled.
[0188]
  This is a figure9When it is in the position of R10% and Lq10% in middle (b), Lq is increased in the control (Fig.9It is equivalent to controlling φ by increasing φ (decreasing on (a) in FIG. 6) by reducing φm + (Ld−Lq) id by decreasing (middle (b)). Eventually, it is possible to obtain a correct device constant by controlling the desired phase, φm + (Ld−Lq) id length.
[0189]
  Figure14These are the flowcharts explaining the principal part of other embodiment of the motor drive method of this invention.
[0190]
In step SP1, the motor current and the motor voltage are detected. In step SP2, the magnetic pole position of the rotor of the motor is detected. In step SP3, the phase of the magnetic flux is detected. In step SP4, the magnetic flux linkage φm is detected. As a result of adding the value obtained by multiplying the difference between the d-axis inductance Ld and the q-axis inductance Lq by the d-axis current id, the length (size) of φm + (Ld−Lq) id is detected. It is determined whether or not the phase is a predetermined phase and the length of φm + (Ld−Lq) id is a predetermined length.
[0191]
If it is determined in step SP5 that the phase of the magnetic flux is not a predetermined phase or the length of φm + (Ld−Lq) id is not a predetermined length, the phase of the magnetic flux is other than ± 90 ° in step SP6. If it is determined that the phase of the magnetic flux is other than ± 90 °, the q-axis inductance Lq is decreased in step SP7.
[0192]
If it is determined in step SP6 that the phase of the magnetic flux is not other than ± 90 °, it is determined in step SP8 whether the length of φm + (Ld−Lq) id is shorter than a predetermined length, and φm + (Ld− Lq) If it is determined that the length of id is shorter than the predetermined length, the q-axis inductance Lq is decreased in step SP9.
[0193]
Conversely, if it is determined in step SP8 that the length of φm + (Ld−Lq) id is not shorter than the predetermined length, or if the processing in step SP9 is performed, the phase of the magnetic flux is determined in step SP10. It is determined whether or not the phase is greater than the phase. If it is determined that the phase of the magnetic flux is greater than the predetermined phase, the winding resistance R is increased in step SP11.
[0194]
Conversely, if it is determined in step SP10 that the phase of the magnetic flux is not greater than the predetermined phase, the winding resistance R is increased in step SP12, and the phase of the magnetic flux is the predetermined phase in step SP13, and It is determined whether or not the length of φm + (Ld−Lq) id is longer than a predetermined value, the phase of the magnetic flux is a predetermined phase, and the length of φm + (Ld−Lq) id is longer than the predetermined value If it is determined, the q-axis inductance Lq is increased in step SP14.
[0195]
When the process of step SP7 is performed, the process of step SP9 is performed, the process of step SP11 is performed, the process of step SP14 is performed, or in step SP13, the phase of the magnetic flux Is not a predetermined phase, or when it is determined that the length of φm + (Ld−Lq) id is not longer than a predetermined value, the process of step SP1 is performed again.
[0196]
When it is finally determined in step SP5 that the phase of the magnetic flux is a predetermined phase and the length of φm + (Ld−Lq) id is a predetermined length, a series of error reduction processing is completed. To do.
[0197]
Further explanation will be given.
[0198]
When the motor current is large relative to the torque, φm + (Ld−Lq) id is extremely small. Conversely, when a sufficiently large current is flowing, φm + (Ld−Lq) id is small. As described above, it is possible to reduce a position detection error by changing a constant (sensor constant) such as a motor device constant or a sensor gain.
[0199]
Of course, it is also possible to adjust the values to take when accurate device constants and sensor constants are set.
[0200]
  Moreover, the figure which shows the magnetic flux length with respect to the change of R9The figure which shows the magnetic flux length with respect to the inside (a) and the change of Lq9As can be seen from the middle (b), it can be made equal to the actual inductance value by adjusting to a value as small as 5%.
[0201]
In addition, when the angle is other than ± 90 degrees, the inductance L is decreased, and when it is within ± 90 degrees, the inductance L is decreased when the magnetic flux length is short, and is increased when the magnetic flux length is long, and the phase of the magnetic flux is within ± 90 degrees. When it is larger than the predetermined phase, it is possible to control to a predetermined phase of 0 to 90 degrees and a predetermined magnetic flux length by increasing R.
[0202]
    Figure15These are the flowcharts explaining the principal part of other embodiment of the motor drive method of this invention.
[0203]
In step SP1, the motor torque is calculated by a conventionally known method. In step SP2, a correction coefficient according to the motor torque is referred to. In step SP3, the device constant is estimated based on the length of φm + (Ld−Lq) id ( For example, the figure7In step SP3, the estimation result is corrected using the correction coefficient, and the series of processes ends.
[0204]
In addition, about the said correction coefficient, setting as follows can be excited.
[0205]
  As shown in FIG. 11, it can be seen that there is a deviation from the value at the time of zero load due to the differential pressure.
[0206]
Moreover, since the value 2140 × 10 μH when the load is 0 is considered to be a true value from the previous discussion, for example, the differential pressure is 3 kg / cm.²In this case, the error of Lq is 30 × 10 μH, which is about 1.4%.
[0207]
Therefore, for example, 1-0.0047 × differential pressure kg / cm²The correction factor (ie 1 kg / cm²0.9953), and an accurate value can be calculated by correcting with the differential pressure.
[0208]
Of course, it is also possible to employ a correction coefficient for performing correction based on load torque instead of differential pressure.
[0209]
If the processing of this flowchart is employed, the set value or the correction amount can be changed according to the motor torque.
[0210]
    Figure10The figure16As shown in Fig. 5, there is a slight error in the device constant estimated value due to the motor torque at the time of device constant estimation. Therefore, by detecting the motor torque and correcting the device constant based on this value, it is possible to estimate the device constant more accurately.
[0211]
The motor torque is φ × I × sin θφ as described above._-IXPn, where φ is the integral of the voltage detection value (drop due to resistance is corrected as Ri), i is the current detection value, and the angle θφ formed by φ and i_-IIs also obtained from φ and i, and does not depend on device constants or sensorless accuracy.
For this reason, it is possible to use motor torque at the time of apparatus constant estimation.
[0212]
【The invention's effect】
The invention of claim 1 has a specific effect that it is very easy to reduce errors in constants for estimating the rotor rotational position, such as motor device constants and sensor constants.
[0213]
According to the second aspect of the present invention, the inconvenience of the instability and the worst divergence stop is prevented, and the same effect as in the first aspect is obtained.
[0214]
The invention of claim 3 can easily realize the conditions of claim 1 or claim 2, and as a result, has the same effect as that of claim 1 or claim 2.
[0215]
The invention of claim 4 can simplify the process, and has the same effect as any one of claims 1 to 3.
[0216]
The invention according to claim 5 can simplify the processing, and has the same effect as any one of claims 1 to 4.
[0217]
The invention of claim 6 has the same effect as any of claims 1 to 3 by using the voltage phase.
[0218]
The invention of claim 7 has the same effect as any of claims 1 to 3 using the current phase.
[0219]
The invention of claim 8 can surely reduce the size of the error, and as a result, has the same effect as any one of claims 1 to 7.
[0223]
  Claim9The invention of the present invention has a specific effect that it is possible to very easily achieve error reduction of constants for estimating the rotor rotational position such as motor device constants and sensor constants.
[0224]
  Claim10The present invention prevents the occurrence of inconvenience that it becomes unstable and stops the worst divergence.9Has the same effect as.
[0225]
  Claim11The invention of claim9Or claims10Can easily be realized, and as a result the claims9Or claims10Has the same effect as.
[0226]
  Claim12The invention can simplify the process and claim9Claims from11The same effect as any of the above.
[0227]
  Claim13The invention can simplify the process and claim9Claims from12The same effect as any of the above.
[0228]
  Claim14The invention of claim 1 uses voltage phase.9Claims from11The same effect as any of the above.
[0229]
    Claim15The invention of claim 1 uses current phase.9Claims from11The same effect as any of the above.
[0230]
  Claim16According to the invention, the magnitude of the error can be surely reduced, and as a result, the claims 9 to 9 are applied.15The same effect as any of the above.
[0234]
  Claim17The present invention has the same effect as any one of claims 1 to 3 when a sufficiently large current flows.
[0235]
  Claim18In addition to being able to make the constant accurate,17Has the same effect as.
[0236]
  Claim19The present invention has the same effect as any one of claims 1 to 3 when a sufficiently large current flows.
[0237]
  Claim20The invention of claim17Claims from19The same effect as any of the above.
  Claim21The invention of claim17Claims from19The same effect as any of the above.
[0238]
  Claim22In the case where a sufficiently large current flows, the present invention claims9Claims from11The same effect as any of the above.
[0239]
  Claim23In addition to being able to make the constant accurate,22Has the same effect as.
[0240]
  Claim24In the case where a sufficiently large current flows, the present invention claims9Claims from11The same effect as any of the above.
[0241]
  In the invention of claim 25, the error can be further reduced.23Has the same effect as.
  The invention of claim 26 can further reduce the error, and24Has the same effect as.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating an example of an apparatus that drives a motor using a fixed coordinate motor inverse model without using a sensor that detects a rotor rotational position.
FIG. 2 is a diagram illustrating an example of a motor torque with respect to a current phase.
FIG. 3 is a diagram showing an example of a vector simulation in a state where the motor current is excessive with respect to the load and the motor current is in a delayed phase.
FIG. 4 is a diagram showing the relationship of the amplitude of φm + (Ld−Lq) Id with respect to the current phase.
FIG. 5 is a flowchart illustrating a main part of an embodiment of the motor driving method according to the present invention.
6 shows a simulation result of the phase of magnetic flux viewed from the d-axis with respect to a change in winding resistance R when the position detection unit 5 of the motor drive device of FIG. 1 performs position detection, and the q-axis inductance Lq. It is a figure which shows the simulation result of the phase of the magnetic flux seen from the estimated d-axis with respect to a change.
[Figure7It is a flow chart for explaining a main part of still another embodiment of the motor driving method of the present invention.
[Figure8It is a flow chart for explaining a length detection process.
[Figure9FIG. 11 is a diagram showing a magnetic flux length with respect to a change in R and a magnetic flux length with respect to a change in Lq.
[Figure10FIG. 10 is a diagram showing a magnetic flux length with respect to a change in R and a magnetic flux length with respect to a change in Lq when the load is large.
[Figure11FIG. 10 is a diagram showing the relationship of the estimated Lq value to the differential pressure in the compressor and the measured value of Lq.
[Figure12It is a flow chart for explaining a main part of still another embodiment of the motor driving method of the present invention.
[Figure13It is a flowchart for explaining a length phase detection process.
[Figure14It is a flow chart for explaining a main part of still another embodiment of the motor driving method of the present invention.
[Figure15It is a flow chart for explaining a main part of still another embodiment of the motor driving method of the present invention.
FIG. 16 is a diagram showing an angular error with respect to a change in R and an angular error with respect to a change in Lq when the load is large.

Claims (26)

The rotor rotational position is estimated using the current and voltage supplied to the motor (4) and the device constants of the motor (4), and the current or voltage is controlled based on the estimated rotor rotational position. In the motor driving method of driving the motor (4) by supplying voltage to the motor (4),
A motor driving method characterized by reducing a constant error for estimating a rotor rotational position in a state in which an excessive current is passed with respect to a load size and the current is in a delayed phase .   The motor driving method according to claim 1, wherein the error is reduced by forcibly applying a rotating magnetic field to the stator regardless of the rotor rotational position.   The constant error for estimating the rotor rotational position is reduced at the time of starting the motor (4) by forcibly applying a rotating magnetic field to the stator regardless of the rotor rotational position. Motor drive method.   4. The error is reduced so that the phase of the interlinkage magnetic flux of the motor coil falls within a range of about 0 to 90 degrees in the rotational direction with reference to the d-axis viewed from the estimation result of the rotor rotational position. A motor driving method according to claim 1.   When the phase of the interlinkage magnetic flux of the motor coil is in the negative direction with reference to the d-axis viewed from the estimation result of the rotor rotational position, the error is reduced by reducing the resistance, and the range is approximately 90 to 180 degrees. 5. The motor driving method according to claim 1, wherein in some cases, the error is reduced by reducing q-axis inductance.   The motor according to any one of claims 1 to 3, wherein the error is reduced so that a voltage phase obtained by estimating the rotor rotational position falls within a range of approximately 90 to 180 degrees on the basis of the d axis on average. Driving method.   The motor drive according to any one of claims 1 to 3, wherein the error is reduced so that the current phase obtained by estimating the rotor rotational position falls within a range of approximately 0 to 90 degrees in the rotational direction with respect to the d axis. Method. The values of q-axis inductance Lq and phase resistance R set for estimating the rotor rotational position are set to be larger than the center value of the variation due to the motor of these values , or constants for detecting current and voltage are set . 8. The q-axis inductance Lq and the phase resistance R are set with a deviation in a direction equivalent to setting the values of these values to be larger than the center value of the variation due to the motor . Motor drive method . The rotor rotational position is estimated using the current and voltage supplied to the motor (4) and the device constants of the motor (4), and the current or voltage is controlled based on the estimated rotor rotational position. In the motor drive device that supplies the voltage to the motor (4) to drive the motor (4),
And a constant error reduction means (6) for reducing a constant error for estimating the rotor rotational position in a state where an excessive current is passed with respect to the size of the load and the current is in a lagging phase. Motor drive device. 10. The motor driving apparatus according to claim 9 , wherein the constant error reducing means (6) performs error reduction in a state where a rotating magnetic field is forcibly applied to the stator regardless of the rotor rotational position. The constant error reducing means (6) forcibly applies a rotating magnetic field to the stator regardless of the rotor rotational position to start the motor (4), and reduces the constant error for estimating the rotor rotational position at the time of starting. The motor driving device according to claim 9 or 10 , wherein the motor driving device is performed. The constant error reducing means (6) reduces the error so that the phase of the interlinkage magnetic flux of the motor coil falls within a range of about 0 to 90 degrees in the rotation direction with reference to the d axis viewed from the estimation result of the rotor rotational position. the motor drive apparatus according to claim 11 claim 9 is performed. The constant error reducing means (6) reduces the error by reducing the resistance when the phase of the interlinkage magnetic flux of the motor coil is in the minus direction with respect to the d axis as seen from the estimation result of the rotor rotational position. The motor drive device according to any one of claims 9 to 12 , wherein the error is reduced by reducing q-axis inductance when the angle is in a range of approximately 90 to 180 degrees. The constant error reduction means (6) reduces the error so that the voltage phase obtained by estimating the rotor rotational position falls within a range of approximately 90 to 180 degrees on the basis of the d axis on average. Item 12. The motor drive device according to any one of Items 9 to 11 . The constant error reducing means (6) reduces the error so that the current phase obtained by estimating the rotor rotational position falls within a range of about 0 to 90 degrees in the rotational direction with respect to the d axis. The motor drive device according to any one of claims 9 to 11 . The constant error reducing means (6) sets the values of the q-axis inductance Lq and the phase resistance R set for estimating the rotor rotational position to be larger than the center value of the variation due to the motor of these values , or Constants for current and voltage detection are set with a deviation in a direction equivalent to setting the values of the q-axis inductance Lq and the phase resistance R to be larger than the center value of the variation of these values due to the motor. The motor drive device according to any one of claims 9 to 15 .   Adjusting the motor device constant or sensor constant so that the estimated value corresponding to the result of adding the value obtained by multiplying the difference between the d-axis inductance and the q-axis inductance by the d-axis current to the magnetic flux linkage is reduced. The motor driving method according to claim 1, wherein the error is reduced by the method. The motor driving method according to claim 17 , wherein after the constant adjustment, correction is performed using a magnitude of a load that is preliminarily assumed for the adjusted device constant or detected by other means .   A value preset based on the d-axis viewed from the rotor rotational position in which the phase of the winding flux linkage is estimated using the current and voltage supplied to the motor (4) and the device constants of the motor (4). And reducing the error so that the estimated amount corresponding to the result of adding the value obtained by multiplying the difference between the d-axis inductance and the q-axis inductance by the d-axis current to the magnetic flux linkage becomes a preset value. The motor driving method according to claim 1, wherein the motor driving method is performed. The motor driving method according to claim 18 , wherein the correction amount is changed according to the motor torque. The motor driving method according to claim 19, wherein the set value is changed in accordance with the motor torque. The constant error reducing means (6) is configured to reduce the estimated amount corresponding to the result obtained by adding the value obtained by multiplying the difference between the d-axis inductance and the q-axis inductance by the d-axis current to the magnetic flux linkage. the motor drive apparatus according to claim 11 claim 9 is performed to reduce errors by adjusting the equipment constant or sensor constants. The constant error reduction means (6) performs correction using a magnitude of a load assumed in advance or detected by other means after the constant adjustment, for the adjusted equipment constant. The motor drive device according to 22 . The constant error reducing means (6) is viewed from the rotor rotational position where the phase of the winding flux linkage is estimated using the current and voltage supplied to the motor (4) and the device constant of the motor (4). An estimated amount corresponding to a result obtained by adding a value obtained by multiplying the difference between the d-axis inductance and the q-axis inductance to the difference between the d-axis inductance and the d-axis current to the magnet linkage flux is set in advance. the motor drive apparatus according to claim 11 claim 9 is performed to reduce errors so that value. 24. The motor driving device according to claim 23 , wherein the constant error reducing means (6) changes a correction amount in accordance with a motor torque. The motor driving device according to claim 24, wherein the constant error reducing means (6) changes a set value in accordance with a motor torque.

Images