JP4116452B2

JP4116452B2 - Transient phenomenon automatic evaluation apparatus, method and program

Info

Publication number: JP4116452B2
Application number: JP2003006698A
Authority: JP
Inventors: 誠司後藤; 英文山本
Original assignee: Toshiba Corp
Current assignee: Toshiba Corp
Priority date: 2003-01-15
Filing date: 2003-01-15
Publication date: 2008-07-09
Anticipated expiration: 2023-01-15
Also published as: JP2004220306A

Description

【０００１】
【発明の属する技術分野】
本発明は、プロセス信号の過渡変化を評価する機能を有する過渡現象自動評価装置、方法およびプログラムに関する。
【０００２】
【従来の技術】
プラントの挙動を把握することはプラントの運転にとって重要なことであり、記録したプラントのプロセス信号に基づいてその評価を行うようにしている。例えば、プロセス信号の過渡現象の評価は、過渡現象記録装置に記録されたプロセス信号のトレンドグラフを印刷出力して、そのトレンドグラフを参照しながら、信号の変化幅、オーバシュート量、応答時間等の評価項目に対して手作業により実施している。また、特定の評価項目については、過渡現象自動評価装置を用いて自動的に過渡現象の評価を行うようにしている。
【０００３】
過渡現象記録装置に記録された過渡現象の評価を行う過渡現象自動評価装置としては特願２００１−２５９２１１号に示される過渡現象自動評価装置がある。この過渡現象自動評価装置においては、プロセス信号のトレンドグラフを表示手段に表示し、表示手段に表示されたプロセス信号のトレンドグラフに対する評価点を評価点作成手段により指定すると、評価手段は指定された評価点に対して過渡現象の評価を行い評価結果を出力するようにしたものである。これにより、短時間で過渡現象の評価を得ることができるようにしている。
【０００４】
【発明が解決しようとする課題】
ところが、このような過渡現象自動評価装置においては、利用者に評価点を指定させるので、評価項目が多い場合はその作業の負担が大きくなる。また、プロセス信号に揺らぎやノイズを含む場合は、利用者によって指定する評価点が区々になり評価結果にばらつきが生じる可能性がある。
【０００５】
本発明は、揺らぎやノイズを含んだプロセス信号からそのプロセス信号本来が持つ応答量を自動で精度良く評価することができる過渡現象自動評価装置、方法およびプログラムを提供することを目的とする。
【０００６】
【課題を解決するための手段】
本発明の過渡現象自動評価装置は、プロセス信号の過渡事象における揺らぎを含む時系列デジタル信号を１次有限インパルス応答フィルタ固有の遅延時間で時間補正することにより１次有限インパルス応答フィルタ出力を算出する１次有限インパルス応答フィルタ出力算出手段と、前記過渡事象における揺らぎを含む時系列デジタル信号を２次有限インパルス応答フィルタ固有の遅延時間で時間補正することにより２次有限インパルス応答フィルタ出力を算出する２次有限インパルス応答フィルタ出力算出手段と、前記１次有限インパルス応答フィルタ出力と２次有限インパルス応答フィルタ出力とを加算平均し揺らぎ除去後の時系列デジタル信号を算出する加算平均算出手段と、前記加算平均算出手段からの過渡事象における揺らぎ除去後の時系列デジタル信号を入力し過渡現象の評価に必要な初期データを算出する初期データ算出手段と、前記加算平均算出手段からの過渡事象における揺らぎ除去後の時系列テジタル信号、前記時系列デジタル信号の信号変化率および前記初期データ算出手段で算出された初期データに基づいて過渡現象の評価に必要な評価点を算出する評価点算出手段と、前記初期データ算出手段で算出された初期データおよび前記評価点算出手段で算出された前記評価点に基づいて過渡現象の評価項目に対する応答量を算出する評価項目応答量算出手段とを備え、前記初期データ算出手段は、初期データとして、プロセス信号の過渡事象の初期値、初期値の標準偏差に所定の係数を乗算し初期値に加算した初期有意基準、過渡事象後の整定値、整定値の標準偏差に所定の係数を乗算し整定値に加算した整定値有意基準を算出し、前記評価点算出手段は、評価点として、プロセス信号が初期有意基準を超える応答開始点、プロセス信号の変化率の極性が反転するピーク点、整定値に指定回復量を加算した値を最初に下回る指定量回復点、プロセス信号の変化率が最大となる最大変化率点、初期値に任意応答量を加算した値を最初に超える任意応答点を算出し、前記評価項目応答量算出手段は、過渡現象の評価項目に対する応答量として、プロセス信号の過渡事象の解析基準時刻から評価点までの時間、評価点のうちの第１ピーク点および第３ピーク点における絶対値の差分としてのオーバーシュート量、最大変化率点における最大変化率、整定値を挟んで変動する減幅比、整定値を挟むことなく変動する減幅比を算出することを特徴とする。
【００１０】
【発明の実施の形態】
（基本構成）
図１は本発明の基本構成に係る過渡現象自動評価装置１１のブロック構成図である。過渡現象自動評価装置１１は、過渡現象の評価に必要な初期データを算出する初期データ算出手段１２と、過渡現象の評価に必要な評価点を算出する評価点算出手段１３と、過渡現象の評価項目に対する応答量を算出する評価項目応答量算出手段１４とから構成されている。
【００１１】
初期データ算出手段１２には、プロセス信号の過渡事象における揺らぎ除去後の時系列デジタル信号Ｘおよび各種の設定値が入力される。初期データ算出手段１２は、プロセス信号の過渡事象における揺らぎ除去後の時系列デジタル信号Ｘを入力し、設定値に従って過渡現象の評価に必要な初期データを算出する。
【００１２】
評価点算出手段１３には、時系列テジタル信号Ｘ、時系列デジタル信号の信号変化率Ｘ’、初期データ算出手段１２で算出された各種初期データ、指定回復量Ｘｒ、任意応答量Ｘｑが入力される。評価点算出手段１３は、これらの入力データに基づいて、過渡現象の評価に必要な評価点を算出する。
【００１３】
評価項目応答量算出手段１４には、初期データ算出手段１２で算出された初期データおよび評価点算出手段１３で算出された過渡現象の評価に必要な評価点が入力され、評価項目応答量算出手段１４は、これらの入力データに基づいて過渡現象の評価項目に対する応答量を算出する。
【００１４】
初期データ算出手段１２に設定される設定値は、過渡現象の評価を行う範囲を指定するための値であり、初期値開始時刻Ｔｓ、初期値終了時刻Ｔｓ’、整定値開始時刻Ｔｅ、整定値終了時刻Ｔｅ’、解析基準時刻Ｔａ、係数ζ等である。初期データ算出手段１２は、これらの設定値および揺らぎ除去後の時系列デジタル信号Ｘに基づいて過渡現象の評価に必要な初期データを算出する。算出される初期データは、初期値Ｘｓ、初期有意基準ζｓ、整定値Ｘｅ、整定値有意基準ζｅである。
【００１５】
図２は、プロセス信号の時系列デジタル信号Ｘに対する設定値および初期データの説明図である。図２では、初期有意基準ζｓおよび整定値有意基準ζｅの図示は省略している。初期データ算出手段１２は、初期値開始時刻Ｔｓから初期値終了時刻Ｔｓ’までの時系列デジタル信号Ｘを平均することにより初期値Ｘｓを算出すると共に、初期値Ｘｓの標準偏差に係数ζを乗算し初期値Ｘｓに加算することにより初期有意基準ζｓを算出する。同様に、整定値開始時刻Ｔｅから整定値終了時刻Ｔｅ’までの時系列デジタル信号Ｘを平均することにより整定値Ｘｅを算出すると共に、整定値Ｘｅの標準偏差に係数ζを乗算し整定値Ｘｅに加算することにより整定値有意基準ζｅを算出する。
【００１６】
次に、評価点算出手段１３は、初期データ算出手段１２で算出された各種初期データ、時系列デジタル信号の信号変化率Ｘ’、指定回復量Ｘｒおよび任意応答量Ｘｑに基づいて、過渡現象の評価に必要な各種の評価点を算出する。算出される評価点は、プロセス信号の時系列デジタル信号Ｘに対し、応答開始点Ｐ１、１０％応答点Ｐ２、９０％応答点Ｐ３、第１ピーク点Ｐ４、第２ピーク点Ｐ５、第３ピーク点Ｐ６、第４ピーク点Ｐ７、指定量回復点Ｐ８、最大変化率点Ｐ９、第１中点Ｐ１０、第２中点Ｐ１１、第３中点Ｐ１２、第４中点Ｐ１３、任意量応答点Ｐ１４である。
【００１７】
図３は、プロセス信号の時系列デジタル信号Ｘに対する設定値、初期データおよび評価点の説明図である。なお、図３では、初期有意基準ζｓ、整定値有意基準ζｅ、任意応答量Ｘｑおよび任意量応答点Ｐ１４の図示を省略している。
【００１８】
評価点算出手段１３は、まず、初期値終了時刻Ｔｓ’以降で初期有意基準ζｓを超える点から内挿する応答開始点Ｐ１を算出する。そして、応答開始点Ｐ１以降で整定値Ｘｅから初期値Ｘｓを減算した値に０．１を乗算し、初期値Ｘｓに加えた値を時系列デジタル信号Ｘが最初に超える点から内挿する１０％応答点Ｐ２を算出する。同様に、応答開始点Ｐ１以降で整定値Ｘｅから初期値Ｘｓを減算した値に０．９を乗算し初期値Ｘｓに加えた値を時系列デジタル信号Ｘが最初に超える点から内挿する９０％応答点Ｐ３を算出する。
【００１９】
次に、応答開始点Ｐ１以降で整定有意基準ζｅを超えて変化率Ｘ’が正から負となる付近で時系列デジタル信号Ｘが最大となる第１ピーク点Ｐ４、第１ピーク点Ｐ４以降で変化率Ｘ’が負から正となる付近で時系列デジタル信号Ｘが最小となる第２ピーク点Ｐ５、第２ピーク点Ｐ５以降で時系列デジタル信号Ｘが整定有意基準ζｅを超えて変化率Ｘ’が正から負となる付近で時系列デジタル信号Ｘが最大となる第３ピーク点Ｐ６、第３ピーク点Ｐ６以降で変化率Ｘ’が負から正となる付近で時系列デジタル信号Ｘが最小となる第４ピーク点Ｐ７を算出する。
【００２０】
また、第１ピーク点Ｐ４以降で時系列デジタル信号Ｘが整定値Ｘｅに指定回復量Ｘｒを加算した値を最初に下回る指定量回復点Ｐ８を算出し、応答開始点Ｐ１以降で時系列デジタル信号Ｘが整定有意基準ζｅ以下で変化率Ｘ’が最大となる最大変化率点Ｐ９を算出する。
【００２１】
さらに、応答開始点Ｐ１と第１ピーク点Ｐ４との中点を内挿する第１中点Ｐ１０、第１ピーク点Ｐ４と第２ピーク点Ｐ５との中点を内挿する第２中点Ｐ１１、第２ピーク点Ｐ５と第３ピーク点Ｐ６との中点を内挿する第３中点Ｐ１２、第３ピーク点Ｐ６と第４ピーク点Ｐ７との中点を内挿する第４中点Ｐ１３を算出する。また、図示は省略するが、応答開始点Ｐ１以降で時系列デジタル信号Ｘが初期値Ｘｓに任意応答量Ｘｑを加算した値を最初に超える点から内挿する任意応答点Ｐ１４を算出する。
【００２２】
評価項目応答量算出手段１４は、初期データ算出手段１２で算出された初期データおよび評価点算出手段１３で算出された評価点に基づいて過渡現象の評価項目に対する応答量を算出する。まず、解析基準時刻Ｔａおよび１０％応答点Ｐ２から１０％応答時間を算出し、解析基準時刻Ｔａおよび９０％応答点Ｐ３から９０％応答時間を算出する。同様に、解析基準時刻Ｔａおよび指定量回復点Ｐ８から指定量回復時間を算出し、任意量応答点Ｐ１４から任意量応答時間を算出する。
【００２３】
また、第１ピーク点Ｐ４と第３ピーク点Ｐ６との絶対値が大きい点からピーク発生時間およびその時のピーク変化量を算出し、第１ピーク点Ｐ４と第３ピーク点Ｐ６の絶対値が大きい点から整定値Ｘｅを減算することによりオーバシュート量を算出し、最大変化率点Ｐ９から最大変化率を算出する。
【００２４】
そして、減幅比１は、第３ピーク点Ｐ６から整定値Ｘｅを減算した値を第１ピーク点Ｐ４から整定値Ｘｅで減算した値で除算することにより算出する。減幅比２は、以下のようにして算出する。まず、第１ピーク点Ｐ４から時間軸に垂直に降ろした線分と、第１中点Ｐ１０と第２中点Ｐ１１とを結んだ線分とが交わるＡ点を求め、第１ピーク点Ｐ４とＡ点との距離Ｄａを求める。同様に、第３ピーク点Ｐ６から時間軸に垂直に降ろした線分と、第３中点Ｐ１２と第４中点Ｐ１３を結んだ線分とが交わるＢ点を求め、第３ピーク点Ｐ１２とＢ点との距離Ｄｂを求める。そして、Ｄａ／Ｄｂで減幅比２を算出する。
【００２５】
ここで、減幅比１は、図３に示すようにプロセス信号の時系列デジタル信号Ｘが整定値Ｘｅを挟んで変動する場合（時系列デジタル信号Ｘが初期値Ｘｓから別の値の整定値Ｘｅに整定する場合）に適用され、減幅比２は、時系列デジタル信号Ｘが整定値Ｘｅを挟むことなく変動する場合（例えば、プロセス信号に外乱が発生し時系列デジタル信号Ｘが初期値Ｘｓから初期値Ｘｓに整定する場合）に適用される。
【００２６】
以上の説明では、プロセス信号の時系列デジタル信号Ｘが初期値Ｘｓより大きな値の整定値Ｘｅに整定する場合について説明したが、初期値Ｘｓより小さな値の整定値Ｘｅに整定する場合にも同様に適用できる。この場合、「増加」を「減少」に、「加算」を「減算」に、「超える」を「下回る」に、「正」を「負」に、「負」を「正」に、「最大」を「最小」に、「最小」を「最大」に、「下回る」を「超える」に、「以下」を「以上」に読み替えることになる。
【００２７】
基本構成によれば、時系列デジタル信号Ｘとその変化率信号Ｘ’とから自動的に評価項目に対する応答量を算出することができる。従って、評価項目が多い場合であってもその作業の負担を軽減でき、利用者によって評価結果にばらつきが生じることが少なくなる。
【００２８】
（第１の実施の形態）
図４は本発明の第１の実施の形態に係る過渡現象自動評価装置１１のブロック構成図である。この第１の実施の形態は、図１に示す基本構成に対し、プロセス信号の過渡事象における揺らぎを除去するための手段として、１次有限インパルス応答フィルタ出力算出手段１５、２次有限インパルス応答フィルタ出力算出手段１６、および加算平均手段１７を追加して設けたものである。図１と同一要素には同一符号を付し重複する説明は省略する。
【００２９】
１次有限インパルス応答フィルタ出力算出手段１５は、プロセス信号の過渡事象における揺らぎを含む時系列デジタル信号を１次有限インパルス応答フィルタ固有の遅延時間で時間補正することにより１次有限インパルス応答フィルタ出力を算出するものであり、２次有限インパルス応答フィルタ出力算出手段１６は、過渡事象における揺らぎを含む時系列デジタル信号を２次有限インパルス応答フィルタ固有の遅延時間で時間補正することにより２次有限インパルス応答フィルタ出力を算出するものである。そして、加算平均算出手段１７は、１次有限インパルス応答フィルタ出力と２次有限インパルス応答フィルタ出力とを加算平均し揺らぎ除去後の時系列デジタル信号Ｘを算出し、初期データ算出手段１２に出力する。
【００３０】
１次有限インパルス応答フィルタ出力算出手段１５は揺らぎを含んだ時系列デジタル値χを１次有限インパルス応答フィルタ処理した後、フィルタ固有の遅延時間を補正した出力Ｙ１を下記式により算出する。
【００３１】
【数１】

ここで、ｎは時系列データ、Ｎはフィルタ長、△ｔはサンプリング周期、Ｔはフィルタ時間（＝Ｎ×△ｔ）、Ｎｅ^−１はフィルタ固有の遅延時間を表す。
【００３２】
２次有限インパルス応答フィルタ出力算出手段１６は揺らぎを含んだ時系列デジタル値χを２次有限インパルス応答フィルタ処理した後、フィルタ固有の遅延時間を補正した出力Ｙ２を下記式により算出する。
【００３３】
【数２】

ここで、ｎは時系列データ、Ｎはフィルタ長、△ｔはサンプリング周期、Ｔはフィルタ時間（＝Ｎ×△ｔ）、Ｎ（１＋ｅ^−１）／２はフィルタ固有の遅延時間を表す。
【００３４】
また、加算平均算出手段１７は出力Ｙ１と出力Ｙ２とを加算平均することにより、時間遅れを生ずることなく揺らぎを除去した時系列デジタル信号Ｘを算出する。図５は、揺らぎを含んだ時系列デジタル値χ、１次有限インパルス応答フィルタ出力算出手段１５の出力Ｙ１、２次有限インパルス応答フィルタ出力算出手段１６の出力Ｙ２、揺らぎ除去後の時系列デジタル信号Ｘ｛Ｘ＝（Ｙ１＋Ｙ２）／２｝の特性図であり、図５からわかるように、揺らぎを除去した時系列デジタル信号Ｘ３は、時間遅れを生ずることなく算出することができる。
【００３５】
第１の実施の形態によれば、１次有限インパルス応答フィルタ出力算出手段１５の出力Ｙ１と２次有限インパルス応答フィルタ出力算出手段１６の出力Ｙ２との平均値により、時間遅れを生ずることなく揺らぎ除去後の時系列デジタル信号Ｘを得るので、揺らぎを含んだ時系列デジタル信号χであっても、揺らぎを含んだ時系列デジタル信号χとその変化率信号Ｘ’とから自動的に精度良く評価項目に対する応答量を算出することができる。
【００３６】
（第２の実施の形態）
図６は、本発明の第２の実施の形態に係る過渡現象自動評価装置１１のブロック構成図である。この第２の実施の形態は、図２に示す基本構成に対し、プロセス信号の過渡事象における時系列デジタル信号χを所定のサンプリング数で区切った部分区間データを算出する部分区間データ算出手段１８と、部分区間データ算出手段１８で算出された部分区間データの変化率を算出し時系列デジタル信号の変化率Ｘ’として評価点算出手段１３に出力する変化率算出手段１９とを追加して設けたものである。図２と同一要素には同一符号を付し重複する説明は省略する。
【００３７】
部分区間データ算出手段１８は時系列デジタル値χを部分区間毎に区切った部分区間データに変換し、変化率算出手段１９はその部分区間データの複数個のデータを用いて下記式により変化率を求める。
【００３８】
【数３】

ここで、Ｎはデータ数でｎ＋１、Δｔはサンプリング周期、ｙｎは時点ｔｎ（ｔｎ＝ＮΔｔ）のデータ値を表す。
【００３９】
第２の実施の形態によれば、時系列デジタル信号の変化率Ｘ’を複数個のデータにより算出するので、変化率Ｘ’の精度が向上する。また、揺らぎを含んだ時系列デジタル信号χから自動的に変化率Ｘ’を算出するので、入力信号として変化率信号を不要とすることができ、精度良く評価項目に対する応答量を算出することができる。
【００４０】
以上の説明は、第１の実施の形態に対し、部分区分データ算出手段１８および変化率算出手段１９を設ける場合について説明したが、基本構成に対し、部分区分データ算出手段１８および変化率算出手段１９を設けるようにしても良い。
【００４１】
上述した各実施の形態において記載した手法は、コンピュータに実行させることのできるプログラムとして、記憶媒体に記憶し各装置に応用したり、通信媒体により伝送して各種装置に適用することも可能である。
【００４２】
本発明における記憶媒体としては、磁気ディスク、フレキシブルディスク、光ディスク（ＣＤ−ＲＯＭ、ＣＤ−Ｒ、ＤＶＤなど）、光磁気ディスク（ＭＯなど）、半導体メモリなど、プログラムを記憶でき、かつコンピュータが読み取り可能な記憶媒体であれば、その記憶形式はいずれの形態であっても良い。また、ここで記憶媒体とは、コンピュータと独立した媒体に限らず、ＬＡＮやインターネットなどにより伝送されたプログラムをダウンロードして記憶または一時記憶した記憶媒体も含まれる。
【００４３】
【発明の効果】
以上説明したように、本発明によれば、プロセス信号の時系列デジタル信号から過渡現象の評価項目に対する応答量を自動で算出できるので、プロセス信号本来が持つ応答量を精度良く評価することができる。また、揺らぎを含む時系列デジタル信号においても揺らぎを適正に除去できるため、過渡現象の評価項目に対する応答量を自動で精度良く算出できる
【図面の簡単な説明】
【図１】本発明の基本構成に係る過渡現象自動評価装置のブロック構成図。
【図２】本発明の基本構成に係る過渡現象自動評価装置でのプロセス信号の時系列デジタル信号Ｘに対する設定値および初期データの説明図。
【図３】本発明の基本構成に係る過渡現象自動評価装置でのプロセス信号の時系列デジタル信号Ｘに対する設定値、初期データおよび評価点の説明図。
【図４】本発明の第１の実施の形態に係る過渡現象自動評価装置のブロック構成図。
【図５】本発明の第１の実施の形態に係る過渡現象自動評価装置での揺らぎ除去後の時系列デジタル信号Ｘの特性図。
【図６】本発明の第２の実施の形態に係る過渡現象自動評価装置のブロック構成図。
【符号の説明】
１１…過渡現象自動評価装置、１２…初期データ算出手段、１３…評価点算出手段、１４…評価項目応答量算出手段、１５…１次有限インパルス応答フィルタ出力算出手段、１６…２次有限インパルス応答フィルタ出力算出手段、１７…加算平均手段、１８…部分区間データ算出手段、１９…変化率算出手段[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a transient phenomenon automatic evaluation apparatus, method, and program having a function of evaluating a transient change of a process signal.
[0002]
[Prior art]
Understanding the behavior of the plant is important for the operation of the plant, and the evaluation is performed based on the recorded process signal of the plant. For example, process signal transients can be evaluated by printing out a trend graph of the process signal recorded in the transient recording device, referring to the trend graph, and the signal change width, overshoot amount, response time, etc. The evaluation items are implemented manually. For specific evaluation items, transient phenomena are automatically evaluated using an automatic transient phenomenon evaluation apparatus.
[0003]
As a transient phenomenon automatic evaluation apparatus for evaluating the transient phenomenon recorded in the transient phenomenon recording apparatus, there is a transient phenomenon automatic evaluation apparatus disclosed in Japanese Patent Application No. 2001-259221. In this transient phenomenon automatic evaluation apparatus, when the trend graph of the process signal is displayed on the display means, and the evaluation point for the trend graph of the process signal displayed on the display means is designated by the evaluation point creation means, the evaluation means is designated. A transient phenomenon is evaluated for an evaluation point, and an evaluation result is output. As a result, transient phenomena can be evaluated in a short time.
[0004]
[Problems to be solved by the invention]
However, in such a transient phenomenon automatic evaluation apparatus, the user is allowed to designate an evaluation point, so that the burden of the work increases when there are many evaluation items. Further, when the process signal includes fluctuations and noise, the evaluation points designated by the user may vary, and the evaluation results may vary.
[0005]
SUMMARY OF THE INVENTION An object of the present invention is to provide a transient phenomenon automatic evaluation apparatus, method, and program capable of automatically and accurately evaluating a response amount inherent in a process signal including fluctuations and noise.
[0006]
[Means for Solving the Problems]
The transient phenomenon automatic evaluation apparatus according to the present invention calculates a first-order finite impulse response filter output by time-correcting a time-series digital signal including fluctuations in a transient event of a process signal with a delay time unique to the first-order finite impulse response filter. A first-order finite impulse response filter output calculating means; and calculating a second-order finite impulse response filter output by time-correcting a time-series digital signal including fluctuations in the transient event with a delay time unique to the second-order finite impulse response filter. A second-order finite impulse response filter output calculating means, a first-order finite impulse response filter output and a second-order finite impulse response filter output are added and averaged to calculate a time-series digital signal after fluctuation is removed; After fluctuation removal in transient events from the mean calculation means An initial data calculating means for inputting a time series digital signal and calculating initial data necessary for evaluating a transient phenomenon; a time series digital signal after removing fluctuations in a transient event from the addition average calculating means; and the time series digital signal Evaluation point calculating means for calculating an evaluation point necessary for evaluating the transient phenomenon based on the signal change rate and the initial data calculated by the initial data calculating means; the initial data calculated by the initial data calculating means; and the evaluation Evaluation item response amount calculation means for calculating a response amount for the evaluation item of the transient phenomenon based on the evaluation point calculated by the point calculation means, wherein the initial data calculation means is a transient event of a process signal as initial data Initial value, initial significance standard by multiplying the standard deviation of the initial value by a predetermined coefficient and adding to the initial value, settling value after transient event, settling value standard A settling value significance criterion is calculated by multiplying the difference by a predetermined coefficient and added to the settling value, and the evaluation point calculation means is an evaluation point of a response start point where the process signal exceeds the initial significance criterion, and the rate of change of the process signal. Peak point where polarity reverses, specified value recovery point that first falls below the value obtained by adding the specified recovery value to the set value, maximum change rate point where the change rate of the process signal is maximum, and value obtained by adding the arbitrary response amount to the initial value The evaluation item response amount calculating means calculates a response amount for the transient phenomenon evaluation item as a response amount for the transient event evaluation item, from the analysis reference time of the process signal transient event to the evaluation point, out of the evaluation points overshoot amount as the difference between the absolute value of the first peak point and the third peak point of the maximum rate of change in the maximum change rate point, the decay ratio varying across the set point, fluctuate without sandwiching the set value The reduction ratio is calculated .
[0010]
DETAILED DESCRIPTION OF THE INVENTION
( Basic configuration )
FIG. 1 is a block diagram of a transient phenomenon automatic evaluation apparatus 11 according to the basic configuration of the present invention. The transient phenomenon automatic evaluation apparatus 11 includes an initial data calculation unit 12 that calculates initial data necessary for the evaluation of the transient phenomenon, an evaluation point calculation unit 13 that calculates an evaluation point necessary for the evaluation of the transient phenomenon, and an evaluation of the transient phenomenon. The evaluation item response amount calculation means 14 calculates the response amount to the item.
[0011]
The initial data calculation means 12 receives the time-series digital signal X after removal of fluctuations in the transient event of the process signal and various set values. The initial data calculation means 12 receives the time-series digital signal X after removal of fluctuations in the transient of the process signal, and calculates initial data necessary for evaluating the transient according to the set value.
[0012]
The evaluation point calculation means 13 receives the time-series digital signal X, the signal change rate X ′ of the time-series digital signal, various initial data calculated by the initial data calculation means 12, the designated recovery amount Xr, and the arbitrary response amount Xq. The The evaluation point calculation means 13 calculates an evaluation point necessary for evaluating the transient phenomenon based on these input data.
[0013]
The evaluation item response amount calculating means 14 receives the initial data calculated by the initial data calculating means 12 and the evaluation points necessary for evaluating the transient phenomenon calculated by the evaluation point calculating means 13. 14 calculates a response amount for the evaluation item of the transient phenomenon based on these input data.
[0014]
The set value set in the initial data calculation means 12 is a value for designating a range in which the transient phenomenon is evaluated. The initial value start time Ts, the initial value end time Ts ′, the settling value start time Te, and the settling value. The end time Te ′, the analysis reference time Ta, the coefficient ζ, and the like. The initial data calculation means 12 calculates initial data necessary for evaluating the transient phenomenon based on these set values and the time-series digital signal X after the fluctuation is removed. The calculated initial data are an initial value Xs, an initial significant reference ζs, a settling value Xe, and a settling value significant reference ζe.
[0015]
FIG. 2 is an explanatory diagram of setting values and initial data for the time-series digital signal X of the process signal. In FIG. 2, the initial significant criterion ζs and the settling value significant criterion ζe are not shown. The initial data calculation means 12 calculates the initial value Xs by averaging the time series digital signal X from the initial value start time Ts to the initial value end time Ts ′, and multiplies the standard deviation of the initial value Xs by a coefficient ζ. The initial significance reference ζs is calculated by adding to the initial value Xs. Similarly, the set value Xe is calculated by averaging the time series digital signal X from the set value start time Te to the set value end time Te ′, and the standard deviation of the set value Xe is multiplied by the coefficient ζ to set the set value Xe. The settling value significance reference ζe is calculated by adding to.
[0016]
Next, the evaluation point calculation means 13 is based on the various initial data calculated by the initial data calculation means 12, the signal change rate X ′ of the time series digital signal, the specified recovery amount Xr, and the arbitrary response amount Xq. Various evaluation points necessary for evaluation are calculated. The calculated evaluation points are the response start point P1, the 10% response point P2, the 90% response point P3, the first peak point P4, the second peak point P5, and the third peak with respect to the time series digital signal X of the process signal. Point P6, fourth peak point P7, designated amount recovery point P8, maximum change rate point P9, first middle point P10, second middle point P11, third middle point P12, fourth middle point P13, arbitrary amount response point P14 It is.
[0017]
FIG. 3 is an explanatory diagram of setting values, initial data, and evaluation points for the time series digital signal X of the process signal. In FIG. 3, the initial significant criterion ζs, the settling value significant criterion ζe, the arbitrary response amount Xq, and the arbitrary amount response point P14 are omitted.
[0018]
The evaluation score calculation means 13 first calculates a response start point P1 to be interpolated from a point exceeding the initial significance criterion ζs after the initial value end time Ts ′. Then, the value obtained by subtracting the initial value Xs from the settling value Xe after the response start point P1 is multiplied by 0.1, and the value added to the initial value Xs is interpolated from the point where the time series digital signal X first exceeds 10. The% response point P2 is calculated. Similarly, after the response start point P1, the value obtained by subtracting the initial value Xs from the settling value Xe is multiplied by 0.9, and the value added to the initial value Xs is interpolated from the point where the time series digital signal X first exceeds 90. The% response point P3 is calculated.
[0019]
Next, after the response start point P1, the first significant peak value P4 at which the time-series digital signal X becomes maximum in the vicinity where the rate of change X ′ exceeds the settling significance reference ζe and becomes negative from positive to negative is obtained. The second peak point P5 at which the time series digital signal X is minimized near the rate of change X ′ from negative to positive, and the rate of change X after the second peak point P5 exceeds the settling significance reference ζe. The third peak point P6 at which the time series digital signal X is maximum near the point where 'is positive to negative, and the time series digital signal X is minimum near the third peak point P6 after which the rate of change X' is negative to positive. A fourth peak point P7 is calculated.
[0020]
Further, after the first peak point P4, the time series digital signal X calculates a designated amount recovery point P8 that first falls below the value obtained by adding the designated recovery amount Xr to the set value Xe, and after the response start point P1, the time series digital signal is calculated. A maximum change rate point P9 is calculated at which X is the settling significance reference ζe or less and the change rate X ′ is the maximum.
[0021]
Furthermore, a first midpoint P10 that interpolates the midpoint between the response start point P1 and the first peak point P4, and a second midpoint P11 that interpolates the midpoint between the first peak point P4 and the second peak point P5. The third midpoint P12 that interpolates the midpoint between the second peak point P5 and the third peak point P6, and the fourth midpoint P13 that interpolates the midpoint between the third peak point P6 and the fourth peak point P7. Is calculated. Although not shown, an arbitrary response point P14 is calculated that is interpolated from the point at which the time series digital signal X first exceeds the value obtained by adding the arbitrary response amount Xq to the initial value Xs after the response start point P1.
[0022]
The evaluation item response amount calculation unit 14 calculates a response amount for the transient phenomenon evaluation item based on the initial data calculated by the initial data calculation unit 12 and the evaluation point calculated by the evaluation point calculation unit 13. First, the 10% response time is calculated from the analysis reference time Ta and the 10% response point P2, and the 90% response time is calculated from the analysis reference time Ta and the 90% response point P3. Similarly, the specified amount recovery time is calculated from the analysis reference time Ta and the specified amount recovery point P8, and the arbitrary amount response time is calculated from the arbitrary amount response point P14.
[0023]
Further, the peak occurrence time and the peak change amount at that time are calculated from the point where the absolute values of the first peak point P4 and the third peak point P6 are large, and the absolute values of the first peak point P4 and the third peak point P6 are large. The overshoot amount is calculated by subtracting the set value Xe from the point, and the maximum change rate is calculated from the maximum change rate point P9.
[0024]
The reduction ratio 1 is calculated by dividing the value obtained by subtracting the set value Xe from the third peak point P6 by the value obtained by subtracting the set value Xe from the first peak point P4. The reduction ratio 2 is calculated as follows. First, a point A where a line segment perpendicular to the time axis from the first peak point P4 intersects with a line segment connecting the first middle point P10 and the second middle point P11 is obtained, and the first peak point P4 and A distance Da to the point A is obtained. Similarly, a B point where a line segment perpendicular to the time axis from the third peak point P6 intersects a line segment connecting the third middle point P12 and the fourth middle point P13 is obtained, and the third peak point P12 A distance Db from point B is obtained. Then, the reduction ratio 2 is calculated by Da / Db.
[0025]
Here, the reduction ratio 1 is obtained when the time series digital signal X of the process signal fluctuates with the set value Xe as shown in FIG. 3 (the time series digital signal X is set to a different set value from the initial value Xs). The reduction ratio 2 is applied when the time series digital signal X fluctuates without sandwiching the set value Xe (for example, a disturbance occurs in the process signal and the time series digital signal X is the initial value). Applied to the initial value Xs).
[0026]
In the above description, the case where the time series digital signal X of the process signal is set to the set value Xe larger than the initial value Xs has been described, but the same applies to the case where the process signal is set to the set value Xe smaller than the initial value Xs. Applicable to. In this case, “increase” is “decrease”, “addition” is “subtraction”, “exceed” is “below”, “positive” is “negative”, “negative” is “positive”, “maximum” ”As“ minimum ”,“ minimum ”as“ maximum ”,“ below ”as“ exceeding ”, and“ below ”as“ more than ”.
[0027]
According to the basic configuration , the response amount to the evaluation item can be automatically calculated from the time-series digital signal X and the change rate signal X ′. Therefore, even when there are many evaluation items, the burden of the work can be reduced, and variations in evaluation results among users are reduced.
[0028]
( First embodiment )
FIG. 4 is a block diagram of the transient phenomenon automatic evaluation apparatus 11 according to the first embodiment of the present invention. In the first embodiment , a first-order finite impulse response filter output calculating means 15 and a second-order finite impulse response filter are provided as means for removing fluctuations in a transient event of a process signal with respect to the basic configuration shown in FIG. An output calculation means 16 and an addition averaging means 17 are additionally provided. The same elements as those in FIG. 1 are denoted by the same reference numerals, and redundant description is omitted.
[0029]
The first-order finite impulse response filter output calculating means 15 time-corrects the time-series digital signal including fluctuations in the transient event of the process signal with a delay time specific to the first-order finite impulse response filter, thereby obtaining the first-order finite impulse response filter output. The second-order finite impulse response filter output calculating means 16 performs time correction on the time-series digital signal including fluctuations in the transient event with a delay time unique to the second-order finite impulse response filter, thereby obtaining a second-order finite impulse response. The filter output is calculated. Then, the addition average calculation unit 17 calculates the time series digital signal X after the fluctuation is removed by averaging the first-order finite impulse response filter output and the second-order finite impulse response filter output, and outputs them to the initial data calculation unit 12. .
[0030]
The first-order finite impulse response filter output calculation means 15 performs first-order finite impulse response filter processing on the time-series digital value χ including fluctuations, and then calculates an output Y1 in which the delay time specific to the filter is corrected by the following equation.
[0031]
[Expression 1]

Here, n represents time series data, N represents a filter length, Δt represents a sampling period, T represents a filter time (= N × Δt), and Ne ⁻¹ represents a delay time specific to the filter.
[0032]
The second-order finite impulse response filter output calculation means 16 performs a second-order finite impulse response filter process on the time-series digital value χ including fluctuations, and then calculates an output Y2 in which the delay time specific to the filter is corrected by the following equation.
[0033]
[Expression 2]

Here, n represents time series data, N represents a filter length, Δt represents a sampling period, T represents a filter time (= N × Δt), and N (1 + e ⁻¹ ) / 2 represents a delay time specific to the filter.
[0034]
Further, the addition average calculation means 17 calculates the time-series digital signal X from which fluctuation is removed without causing a time delay by averaging the output Y1 and the output Y2. FIG. 5 shows a time-series digital value χ including fluctuations, an output Y1 from the first-order finite impulse response filter output calculation means 15, an output Y2 from the second-order finite impulse response filter output calculation means 16, and a time-series digital signal after fluctuation removal. FIG. 5 is a characteristic diagram of X {X = (Y1 + Y2) / 2}. As can be seen from FIG. 5, the time-series digital signal X3 from which fluctuation has been removed can be calculated without causing a time delay.
[0035]
According to the first embodiment , the average value of the output Y1 of the first-order finite impulse response filter output calculation means 15 and the output Y2 of the second-order finite impulse response filter output calculation means 16 fluctuates without causing a time delay. Since the time-series digital signal X after removal is obtained, even the time-series digital signal χ including fluctuation is automatically and accurately evaluated from the time-series digital signal χ including fluctuation and the change rate signal X ′. The response amount to the item can be calculated.
[0036]
( Second Embodiment )
FIG. 6 is a block diagram of the transient phenomenon automatic evaluation apparatus 11 according to the second embodiment of the present invention. This second embodiment is different from the basic configuration shown in FIG. 2 in that it includes partial section data calculation means 18 for calculating partial section data obtained by dividing a time series digital signal χ in a transient event of a process signal by a predetermined sampling number. And a change rate calculation means 19 for calculating the change rate of the partial interval data calculated by the partial interval data calculation means 18 and outputting it to the evaluation point calculation means 13 as the change rate X ′ of the time series digital signal. Is. The same elements as those in FIG.
[0037]
The partial section data calculation means 18 converts the time series digital value χ into partial section data divided for each partial section, and the change rate calculation means 19 uses a plurality of data of the partial section data to calculate the change rate by the following formula. Ask.
[0038]
[Equation 3]

Here, N is the number of data, n + 1, Δt is the sampling period, and yn is the data value at time tn (tn = NΔt).
[0039]
According to the second embodiment, since the rate of change X ′ of the time-series digital signal is calculated from a plurality of data, the accuracy of the rate of change X ′ is improved. Further, since the rate of change X ′ is automatically calculated from the time-series digital signal χ including fluctuations, the rate of change signal can be eliminated as an input signal, and the response amount to the evaluation item can be calculated with high accuracy. it can.
[0040]
The above description is with respect to the first embodiment has described the case of providing the partial partitioned data calculating means 18 and the change rate calculating means 19, to the basic configuration, partial partitioned data calculating means 18 and the change rate calculating means 19 may be provided.
[0041]
The method described in each embodiment described above can be stored in a storage medium as a program that can be executed by a computer and applied to each device, or transmitted by a communication medium and applied to various devices. .
[0042]
As a storage medium in the present invention, a magnetic disk, a flexible disk, an optical disk (CD-ROM, CD-R, DVD, etc.), a magneto-optical disk (MO, etc.), a semiconductor memory, etc. can be stored and can be read by a computer. As long as it is a simple storage medium, the storage format may be any form. Here, the storage medium is not limited to a medium independent of the computer, but also includes a storage medium in which a program transmitted via a LAN or the Internet is downloaded and stored or temporarily stored.
[0043]
【The invention's effect】
As described above, according to the present invention, since the response amount to the evaluation item of the transient phenomenon can be automatically calculated from the time series digital signal of the process signal, the response amount inherent in the process signal can be accurately evaluated. . In addition, since fluctuations can be properly removed even in time-series digital signals including fluctuations, the response amount for the evaluation items of transient phenomena can be calculated automatically and accurately [Brief description of the drawings]
FIG. 1 is a block diagram of a transient phenomenon automatic evaluation apparatus according to a basic configuration of the present invention.
FIG. 2 is an explanatory diagram of set values and initial data of a process signal with respect to a time-series digital signal X in the transient phenomenon automatic evaluation apparatus according to the basic configuration of the present invention.
FIG. 3 is an explanatory diagram of set values, initial data, and evaluation points of a process signal with respect to a time-series digital signal X in the transient phenomenon automatic evaluation apparatus according to the basic configuration of the present invention.
FIG. 4 is a block configuration diagram of the transient phenomenon automatic evaluation apparatus according to the first embodiment of the present invention.
FIG. 5 is a characteristic diagram of the time-series digital signal X after fluctuation removal in the automatic transient evaluation device according to the first embodiment of the present invention.
FIG. 6 is a block diagram of a transient phenomenon automatic evaluation apparatus according to a second embodiment of the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 11 ... Transient phenomenon automatic evaluation apparatus, 12 ... Initial data calculation means, 13 ... Evaluation point calculation means, 14 ... Evaluation item response amount calculation means, 15 ... Primary finite impulse response filter output calculation means, 16 ... Secondary finite impulse response Filter output calculation means, 17 ... addition average means, 18 ... partial section data calculation means, 19 ... change rate calculation means

Claims

First-order finite impulse response filter output calculating means for calculating a first-order finite impulse response filter output by time-correcting a time-series digital signal including fluctuation in a transient event of a process signal with a delay time unique to the first-order finite impulse response filter; Second-order finite impulse response filter output calculating means for calculating a second-order finite impulse response filter output by time-correcting a time-series digital signal including fluctuations in the transient event with a delay time specific to the second-order finite impulse response filter; Addition average calculating means for calculating the time-series digital signal after the fluctuation is removed by averaging the first-order finite impulse response filter output and the second-order finite impulse response filter output, and fluctuation removal in the transient event from the addition average calculation means Input a time series digital signal later and make a transient Initial data calculation means for calculating initial data required for elephant evaluation, time-series digital signal after fluctuation removal in the transient event from the addition average calculation means, signal change rate of the time-series digital signal, and initial data calculation Evaluation point calculation means for calculating an evaluation point necessary for evaluation of the transient phenomenon based on the initial data calculated by the means, initial data calculated by the initial data calculation means, and the evaluation point calculation means Evaluation item response amount calculation means for calculating a response amount for the evaluation item of the transient phenomenon based on the evaluation point, the initial data calculation means as the initial data, the initial value of the transient of the process signal, the standard of the initial value Multiply the deviation by a predetermined coefficient and add it to the initial value. The initial significance criterion, the settling value after the transient event, and the standard deviation of the settling value are multiplied by the predetermined coefficient. The calculated settling value significance criterion is calculated, and the evaluation point calculation means designates as a response start point at which the process signal exceeds the initial significance criterion, a peak point at which the polarity of the change rate of the process signal is inverted, and a settling value as the evaluation point Calculates the specified amount recovery point that first falls below the sum of the recovery amount, the maximum change rate point at which the change rate of the process signal is maximum, and the arbitrary response point that first exceeds the initial value plus the arbitrary response amount. The evaluation item response amount calculation means, as a response amount to the evaluation item of the transient phenomenon, is the time from the analysis reference time of the transient phenomenon of the process signal to the evaluation point, the first peak point and the third peak point of the evaluation points. It is characterized by calculating the amount of overshoot as an absolute value difference, the maximum rate of change at the maximum rate of change point, the reduction ratio that fluctuates across the set value, and the reduction ratio that fluctuates without setting the set value . Automatic transient evaluation device.

A partial section data calculating means for calculating partial section data obtained by dividing a time-series digital signal in a transient event of a process signal by a predetermined sampling number; and a rate of change of the partial section data calculated by the partial section data calculating means. 2. The transient phenomenon automatic evaluation apparatus according to claim 1 , further comprising a change rate calculation unit that outputs the signal change rate of the time-series digital signal to the evaluation point calculation unit.