JP2004062450A - Method for decision support on productive activity and program therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To achieve rational decision making for evaluating risks resulting from future uncertainty, while at the same time aiming at the economical efficiency of productive activity, i.e., maximum profit. <P>SOLUTION: Data about a plurality of decision-making plans for productive activity and data D1 about a plurality of situations conceivable with respect to future uncertainty factors are input (S110). The plurality of decision-making plans are combined with the plurality of situations to create a decision-making tree D3 (S120). Based on the decision-making tree, an optimum model of productive activity under each of the plurality of situations is constructed for each of the plurality of decision-making plans and an optimum productive activity value index is calculated. Based on the optimum productive activity value index, a value variation risk index which represents the magnitude of value variation caused by the plurality of situations is calculated, and based on the optimum productive activity value index and the value variation risk index, decision making is performed over a predetermined future period during which productive activity is performed (S130). The result of the decision making is output (S140). <P>COPYRIGHT: (C)2004,JPO

Description

【００６１】
なお、事業状況としてはさらに、表１に示すような増減という大まかな分類だけでなく、「１割増加」、「１割減少」、「２割増加」、「２割減少」、…、「ｎ割増加」、「ｎ割減少」、等の、より詳細に分類した事業状況を設定したり、さらに、燃料価格や電力需要以外の不確実性要因に対する事業状況を加味したりすることも考えられる。逆に、電力需要のみに対する「電力需要増加」、「電力需要減少」、という事業状況のみを設定する等、より少ない数の事業状況を設定することも可能である。同様に、複数の投資案についても、より詳細に分類した投資案を設定したり、逆に、より少ない数の投資案のみを設定したりすることが可能である。
【００６２】
また、経費・実積データＤ２の設定に当たっては、過去の有価証券報告書に記載されているレベルのデータを読み込んで使用することができる。すなわち、経費データについては、有価証券報告書の財務諸表データから取得可能であり、また、生産活動である電源設備による電力供給の実績データについては、有価証券報告書の事業状況データや設備状況データ等から、過去の運転実績データや、既存の設備状況データ等が取得可能である。なお、これらのデータは、有価証券報告書に記載されているデータに限らず、業者が保存している各種の過去データから同様に取得可能である。
【００６３】
このように読み込んだデータは、入出力部により意思決定者であるユーザに対して画面表示され、ユーザによって、どのデータを使用するかが選択され、選択されたデータが、実際に使用するデータとして記憶部に格納される。したがって、意思決定者であるユーザは、画面表示されたデータのうち、どのデータを使用するかを選択するだけで、経費・実積データＤ２を容易に設定することができる。
【００６４】
そして、演算処理部により、格納された投資案・事業状況データＤ１ａや経費・実積データＤ２を用いて、後続の処理に必要な各種の外部パラメータＤ１１を設定し、記憶部に格納する（Ｓ４２０）。この外部パラメータの設定においては、経費・実績データＤ２を用いてユニットごとの経費や燃料使用量を算出し、初年度実績の各パラメータとして設定する。また、投資案・事業状況データＤ１ａ等を用いて、不確実性要因の変動幅等のパラメータを設定する。設定された各種のパラメータは、入出力部により意思決定者であるユーザに画面提示され、ユーザの確認を取得した時点で、記憶部に格納される。
【００６５】
次に、演算処理部により、格納された投資案・事業状況データＤ１ａを用いて、設備の追加や取替等の複数の投資案と、燃料価格等の資源コストや電力需要等の不確実性要因に対して想定される複数の事業状況とを組み合わせて意思決定木Ｄ３を生成し、記憶部に格納する（Ｓ４３０）。そして、格納された意思決定木Ｄ３と外部パラメータＤ１１を用いて、演算処理部により、将来のＴ期間に亘る意思決定として、Ｔ期間に亘る最適な電源計画の導出を行う（Ｓ４４０）。この「Ｔ期間に亘る最適電源計画の導出」の詳細は次の通りである。
【００６６】
まず、Ｔ期間の各期を示す変数ｔの値をＴとする（Ｓ４４１）。そして、（ｔ−１）期までの投資案、事業状況の組み合わせの全てについて、各組み合わせごとに、ｔ期目の最適な投資案をそれぞれ決定する（Ｓ４４２〜Ｓ４４４）。すなわち、ｔ期目の最適な意思決定を行うために、意思決定木Ｄ３を用いて、１期〜（ｔ−１）期までの各期ごとに、評価しようとする幾つかの投資案から一つの投資案をそれぞれ選択し、将来予想される幾つかの事業状況から一つの事業状況をそれぞれ選択して、各期ごとにそれぞれ選択した投資案と状況の組み合わせを（ｔ−１）期までの前提条件として設定する（Ｓ４４２）。この（ｔ−１）期までの前提条件に対して、図５に示す「ｔ期目最適投資案決定処理」を行い、意思決定木Ｄ３と外部パラメータＤ１１を用いて、ｔ期目の最適な投資案を決定する（Ｓ４４３）。
【００６７】
（ｔ−１）期までの投資案、事業状況の全ての組み合わせにおいて、ｔ期目の最適な投資案を決定していない組み合わせが残っている間（Ｓ４４４のＮＯ）は、（ｔ−１）期までの投資案、事業状況の組み合わせによる前提条件の設定（Ｓ４４２）と、ｔ期目の最適な投資案の決定（Ｓ４４３）を繰り返す。そして、（ｔ−１）期までの投資案、事業状況の全ての組み合わせに対してｔ期目の最適な投資案を決定した（Ｓ４４４のＹＥＳ）場合には、次のステップＳ４４５に進む。
【００６８】
なお、「（ｔ−１）期までの投資案、事業状況の組み合わせの全て」とは、（ｔ−１）期までの各期の投資案の選択結果として考えられる（ｔ−１）期までの全ての投資計画と、（ｔ−１）期までの各期の事業状況の変動結果として考えられる（ｔ−１）期までの全ての想定シナリオとの組み合わせの全てを意味している。例えば、投資案をＡ、事業状況をＢ、投資案の数をｎ、事業状況の数をｍとした場合に、（ｔ−１）期までの投資案、事業状況の組み合わせの総数は、ｎ^{（ｔ−１）}×ｍ^{（ｔ−１）}個となる。
【００６９】
この点について説明すると、まず、１期〜（ｔ−１）期までの全ての投資計画の各々は、各期ごとに選択された投資案からなる列Ａ１，…，Ａ_{（ｔ−１）}として表現可能であり、各期ごとに選択可能な投資案の数はｎであるため、１期〜（ｔ−１）期までに選択されうる全ての投資計画の数は、ｎ^{（ｔ−１）}個となる。また、１期〜（ｔ−１）期までの全ての想定シナリオの各々は、各期ごとに想定された事業状況からなる列Ｂ１，…，Ｂ_{（ｔ−１）}として表現可能であり、各期ごとに想定可能な事業状況の数はｍであるため、１期〜（ｔ−１）期までに想定されうる全ての想定シナリオの数は、ｍ^{（ｔ−１）}個となる。したがって、（ｔ−１）期までの投資案、事業状況の組み合わせの総数、すなわち、投資計画、想定シナリオの組み合わせの総数は、ｎ^{（ｔ−１）}×ｍ^{（ｔ−１）}個となる。
【００７０】
例えば、表１の例においては、各期ごとに選択可能な投資案の数は４であり、各期ごとに想定可能な事業状況の数は４であるため、１期〜（ｔ−１）期までに選択されうる全ての投資計画の数は、４^{（ｔ−１）}個となり、また、１期〜（ｔ−１）期までに想定されうる全ての想定シナリオの数は、４^{（ｔ−１）}個となる。したがって、表１の例において、（ｔ−１）期までの投資案、事業状況の組み合わせの総数、すなわち、投資計画、想定シナリオの組み合わせの総数は、４^{（ｔ−１）}×４^{（ｔ−１）}個となる。
【００７１】
以上のようなｔ期目の最適な意思決定を行うための一連のステップ（Ｓ４４２〜Ｓ４４４）を、ｔの値を１ずつ減算しながら、ｔ＝１となるまで繰り返す（Ｓ４４５のＮＯ、Ｓ４４６）。そして、ｔ＝１となった場合、すなわち、現在の事業状況（初期値）に対して１期目に採用するのに最適な投資案を決定した場合（Ｓ４４５のＹＥＳ）には、Ｔ期間に亘る意思決定（Ｓ４４０）を終了し、次のステップＳ４５０に進む。
【００７２】
ステップＳ４５０においては、演算処理部により、導出されたＴ期間に亘る最適電源計画の評価を行い、その評価結果を、入出力部により出力し、グラフ等の形式で画面表示する。ここで、「Ｔ期間最適電源計画の評価結果」としては、「将来の想定シナリオに応じた最適導入時期」、「様々な想定シナリオの元での最適生産活動による利益分布」、「最適生産活動モデルから導出される各設備の最適運転量」、等をグラフとして画面表示する。また、「電源が足りなくなったら追加する」という従来の電源計画による生産活動価値と、最適電源計画による生産活動価値を現在価値に換算して比較するグラフを画面表示してもよい。
【００７３】
［３−１．ｔ期目最適投資案決定処理］
ｔ期目の最適な投資案の決定においては、ｔ期目の投資案として、投資案データとして設定されている複数の投資案、例えば、前述した表１に示すような、「新規設備の追加」、「既存設備の改修」、「既存設備の廃棄」、「設備投資しない」、等の投資案の中から、最適な投資案を決定する。
【００７４】
ｔ期目にどの投資案を採用すべきかについては、その時点での事業状況に合わせて意思決定する必要があるが、図５に示すようなｔ期目最適投資案決定処理を行い、その処理中で、図６に示すようなｔ〜Ｔ期間リスク調整済期待価値算出処理を行うことにより、最適生産活動・リスク管理を考慮した、ｔ期目の最適な投資意思決定を行うことができる。
【００７５】
図５に示すｔ期目最適投資案決定処理においては、まず、記憶部に格納された意思決定木Ｄ３を用いて、ｔ期目に採用する投資案の一つの候補を選択する（Ｓ５０１）。次に、選択した投資案をｔ期目に採用した場合について、図６に示す「ｔ〜Ｔ期間リスク調整済期待価値算出処理」を行い、意思決定木Ｄ３や外部パラメータＤ１１を用いて、ｔ〜Ｔ期間のリスク調整済期待価値を算出する（Ｓ５０２）。
【００７６】
ｔ期目に採用する投資案の全ての候補において、選択されていない投資案が残っている間（Ｓ５０３のＮＯ）は、投資案の候補の選択（Ｓ５０１）と、ｔ〜Ｔ期間のリスク調整済期待価値の算出（Ｓ５０２）を繰り返す。そして、投資案の全ての候補について、ｔ〜Ｔ期間のリスク調整済期待価値を算出した場合（Ｓ５０３のＹＥＳ）には、次のステップＳ５０４に進む。
【００７７】
例えば、前述した表１の例においては、複数の投資案として、「新規設備の追加」、「既存設備の改修」、「既存設備の廃棄」、「設備投資しない」、という４つの投資案の全てについて、ｔ〜Ｔ期間のリスク調整済期待価値を算出した場合（Ｓ５０３のＹＥＳ）に、次のステップＳ５０４に進むことになる。
【００７８】
ステップＳ５０４においては、全ての投資案について算出したｔ〜Ｔ期間のリスク調整済期待価値を比較して、全ての投資案の中から、ｔ〜Ｔ期間に最大のリスク調整済期待価値を生じる投資案を、ｔ期目の最適な投資案として選択する。そして、選択したｔ期目最適投資案により得られる最大のリスク調整済期待価値を、「ｔ〜Ｔ期間の最適生産活動によるリスク調整済期待価値」Ｄ４として記憶部に格納して（Ｓ５０５）、ｔ期目最適投資案決定処理Ｓ４４３を終了し、図４のステップＳ４４４に戻る。
【００７９】
［３−２．ｔ〜Ｔ期間リスク調整済期待価値算出処理］
ｔ〜Ｔ期間のリスク調整済期待価値の算出においては、燃料価格、電力需要等の不確実性要因に対して想定され、事業状況データとして設定されている複数の事業状況の全てについて期待価値を算出し（Ｓ６０１〜Ｓ６０５）、事業状況による価値変動リスクを加味して、ｔ期目に各投資案を採用した場合（図５のＳ５０１）のｔ〜Ｔ期間のリスク調整済期待価値を算出する。
【００８０】
図６に示すｔ〜Ｔ期間リスク調整済期待価値算出処理においては、まず、記憶部に格納された意思決定木Ｄ３を用いて、ｔ期目に想定される事業状況の一つの候補を選択する（Ｓ６０１）。
【００８１】
次に、記憶部に格納された意思決定木Ｄ３と外部パラメータＤ１１を用いて、ｔ期目に選択した投資案を採用した場合（図５のＳ５０１）の、選択した事業状況におけるｔ期目の最適生産活動モデルを構築して、ｔ期の最適生産活動価値指標である「ｔ期の最大生産活動価値」と、「各設備の最適発電端電力量」を算出する（Ｓ６０２）。すなわち、最適生産活動モデルは、線形計画問題として定式化され、最適値は「ｔ期の最大生産活動価値」、最適解は「各設備の最適発電端電力量」となるが、この最適生産活動モデルの詳細については後述する。
【００８２】
選択した事業状況について算出した「ｔ期の最大生産活動価値」と、記憶部に格納されている計算済の「（ｔ＋１）〜Ｔ期間の最適生産活動によるリスク調整済期待価値」Ｄ４とを足し合わせて、選択した投資案を採用した場合（図５のＳ５０１）の、選択した事業状況におけるｔ〜Ｔ期間の価値を算出する（Ｓ６０３）。
【００８３】
ｔ期目に選択した投資案を採用した場合（図５のＳ５０１）に、想定される事業状況の全ての候補において、選択されていない事業状況が残っている間（Ｓ６０４のＮＯ）は、事業状況の候補の選択からｔ〜Ｔ期間の価値の算出までの一連のステップ（Ｓ６０１〜Ｓ６０３）を繰り返す。そして、事業状況の全ての候補について、ｔ〜Ｔ期間の価値を算出した場合（Ｓ６０４のＹＥＳ）には、次のステップＳ６０５に進む。
【００８４】
例えば、前述した表１の例においては、複数の事業状況として、「燃料価格上昇、電力需要増加」、「燃料価格低下、電力需要増加」、「燃料価格上昇、電力需要減少」、「燃料価格低下、電力需要減少」、という４つの事業状況の全てについて、ｔ〜Ｔ期間の価値を算出した場合（Ｓ６０４のＹＥＳ）に、次のステップＳ６０５に進むことになる。
【００８５】
ステップＳ６０５においては、想定される全ての事業状況によって生じるｔ〜Ｔ期間の価値から、ｔ〜Ｔ期間の期待価値を算出する。すなわち、事業状況の数をｓ、各事業状況ｉの生起確率をｐ_ｉ（ｉ＝１，２，Λ，ｓ）、各事業状況ｉにより生じるｔ〜Ｔ期間の価値をｒ_ｉ（ｉ＝１，２，Λ，ｓ）、とした場合に、ｔ〜Ｔ期間の期待価値Ｅ［ｒ］は、次の式によって算出される。
【数１】

【００８６】
続いて、想定される全ての事業状況によって生じるｔ〜Ｔ期間価値の価値変動リスク指標を算出する（Ｓ６０６）。すなわち、ｔ〜Ｔ期間価値の価値変動リスク指標Ｖ［ｒ］は、事業状況の数ｓ、事業状況ｉの生起確率ｐ_ｉとｔ〜Ｔ期間の価値ｒ_ｉ（ｉ＝１，２，Λ，ｓ）、および、ｔ〜Ｔ期間の期待価値Ｅ［ｒ］から、次の式によって算出される。
【数２】

なお、価値変動リスク指標の算出方法としては、この他にも、例えば、９５％で起こりうるＶａＲといった最大損失を価値変動リスク指標とみなす方法等を採用することが考えられる。
【００８７】
価値変動リスク指標の大きさに従って、現在価値への割引率を設定することにより、ｔ〜Ｔ期間の期待価値Ｅ［ｒ］を調整し、ｔ〜Ｔ期間のリスク調整済期待価値を算出する（Ｓ６０７）。例えば、価値変動リスク指標が大きい場合、つまり価値変動が大きい場合には、将来の不確実性が高いとみなし、ｔ〜Ｔ期間の期待価値を大きく割り引いて調整する。具体的には、株式市場におけるリスクプレミアムを参考にして、割引率を決定することもできる。
【００８８】
このような一連のｔ〜Ｔ期間リスク調整済期待価値算出処理により、ｔ期目に選択した投資案を採用した場合（図５のＳ５０１）における、想定される全ての事業状況によるｔ〜Ｔ期間のリスク調整済期待価値を求めることができる。
【００８９】
［３−３．最適生産活動モデル］
以下には、図６に示すｔ〜Ｔ期間リスク調整済期待価値算出処理において使用する最適生産活動モデルの「変数」、「目的関数」、「制約式」について説明する。なお、ここでは、設備投資案として、「新規設備の追加」と「設備投資しない」の２つを考える。第ｊ期に新規設備ｉを追加する、しないの別を、ｘ_ｉｊ＝１、または、ｘ_ｉｊ＝０、で表現する。他の設備投資案についても、変数ｘ_ｉｊを増やすことにより、同様に定式化することができる。なお、ここでは、説明を簡単にするため、制約式として、各設備の上下限運転量と電力供給量の安定性、を考える。
【００９０】
「変数」は、第ｊ期に購入した設備ｉを使った場合の、投資案ｘ、事業状態ｓのもとでの第ｔ期の発電端電力量である。ただし、ｊ≦ｔである。以下には、この第ｔ期の発電端電力量をｙ_ｉｊと表現する。
「目的関数」は、投資案ｘを採用し、事業状態ｓを想定して、第ｔ期目に生産活動を行ったときの利益を最大化することであり、この利益は次の式で表現される。
【数３】

【００９１】
「制約式」としては、各設備の上下限運転量と電力供給量の安定性等を制約とするために、次の３つの制約式を使用する。
【００９２】
「制約式１」：第ｊ期に購入した設備ｉによる電力供給量は次の式によって表現される上下限運転量制約を満たす。
【数４】

【００９３】
「制約式２」：電力総供給量（送電端電力量の総和）は電力年間需要量に等しい。これは、電力供給量の安定性制約の一つであり、次の式によって表現される。
【数５】

【００９４】
「制約式３」：瞬間需要×予備率を満たす、総設備容量を持つ。これもまた、電力供給量の安定性制約の一つであり、次の式によって表現される。
【数６】

【００９５】
したがって、投資案ｘを採用し、事業状態ｓを想定した場合における第ｔ期目の利益最大化問題、すなわち、最適生産活動モデルは、以下の線形計画問題として定式化される。
【数７】

【００９６】
なお、最適生産活動モデルを表現する上記の各式中で使用される各要素の定義は、次の表２に示す通りである。
【表２】

【００９７】
［３−４．Ｔ期間最適電源計画の評価結果の具体例］
図７〜図９は、図４に示す電源計画意思決定支援処理において、「Ｔ期間最適電源計画の評価結果」として得られた各種のグラフを示している。以下には、これらのグラフについて順次説明する。
【００９８】
まず、図７は、平成１０年までを第１期とし、３年間隔で第２期以降を設定して、平成２５年までの全６期に亘るＴ期間を設定し、複数の投資案として表１の投資案、すなわち、「新規設備の追加」、「既存設備の改修」、「既存設備の廃棄」、「設備投資しない」、という４つの投資案を想定して図４に示す電源計画意思決定支援処理を行った場合の電力需要量と最適投資時期との関係を示すグラフである。
【００９９】
この図７においては、想定状況によって決まる不確実性要因値である電力需要量を横軸とし、その状況下における最適投資時期を縦軸として、プロットの種類で投資案の別を示している。さらに、実際には、プロットの種類だけでなく、色についても投資案に応じて異なる色を使用することにより、投資案の別をより明確に表現することができる。このようなグラフを画面表示することにより、意思決定者であるユーザは、年間電力需要量等の将来の想定シナリオに応じた最適導入時期を容易に確認することができる。
【０１００】
また、図８は、複数の事業状況として、表１の事業状況、すなわち、「燃料価格上昇、電力需要増加」、「燃料価格低下、電力需要増加」、「燃料価格上昇、電力需要減少」、「燃料価格低下、電力需要減少」、という４つの事業状況を各期ごとにそれぞれ想定して図４に示す電源計画意思決定支援処理を行った場合の、様々な想定シナリオの元での最適生産活動による利益分布を示すグラフである。
【０１０１】
特に、この図８においては、「電源が足りなくなったら追加する」という従来の電源計画による生産活動価値と、図４に示す本実施形態の電源計画意思決定支援処理によって導出された最適電源計画による生産活動価値とを現在価値に換算して比較している。この図８においては、プロットの種類で本実施形態によって導出された最適電源計画による利益分布と、従来の電源計画による利益分布との別を示しているが、実際には、プロットの種類だけでなく、プロットとそれを接続してなる曲線の色についても、最適電源計画と従来の電源計画とで異なる色を使用することにより、各計画の利益分布をより明確に表現することができる。
【０１０２】
この図８において、本実施形態によって導出された最適電源計画による利益分布を示す曲線の全体形状とピークは、従来の電源計画による利益分布を示す曲線の全体形状とピークよりも、右側にずれており、本実施形態に係る最適電源計画を採用した場合に、従来の計画に比べてはるかに大きなキャッシュフローが得られることが分かる。
【０１０３】
また、図９は、燃料として、石油、ＬＮＧ、石炭を使用する設備を含む電源計画を導出した場合において、電力需要量と燃料別運転量との関係を示すグラフである。本実施形態の電源計画意思決定支援処理においては、最適生産活動モデルから、将来の想定シナリオに応じた各設備の最適運転量を導出することができるので、このようなグラフを画面表示することにより、意思決定者であるユーザは、年間電力需要量等の将来の想定シナリオに応じた各設備の最適運転量を容易に確認することができる。
【０１０４】
［４．電源計画意思決定支援処理の作用効果］
以上のような電源計画意思決定支援処理によれば、前述した生産活動意思決定支援処理の基本的な作用効果に加えて、さらに、次のような作用効果が得られる。
【０１０５】
まず、本実施形態の電源計画意思決定支援処理によれば、電力供給事業における将来的な設備投資に関し、最適な意思決定を行うことができる。本発明は、各種の生産活動を行うための意思決定に好適であるが、特に、電力供給事業における設備投資に関する意思決定においては、表１に示したように、燃料価格等の資源コストや電力需要等、重要な不確実性要因の種類やその状況の数を絞り込みやすいため、本発明を適用することにより、効率よく最適な意思決定を行うことができ、最適な電源計画を導出することができる。
【０１０６】
また、表１に示したように、投資案として、新規設備の追加、既存設備の改修、既存設備の廃棄、等を設定することにより、新規設備の最適導入時期、導入形態や既存設備の最適廃棄時期、廃棄形態等を容易に導出することができる。さらに、新規設備の追加の有無を意思決定案として設定することにより、導入を検討している複数の新規設備に対して、将来の期間における導入効果を比較することもできる。
【０１０７】
さらに、図５、図６に示すｔ期目最適投資案決定処理における最適生産活動価値指標の算出に最適生産活動モデルの問題構造を利用することで、少ない計算量での生産活動価値算出が可能となるため、実際規模の設備投資計画を立てる際に膨大な計算を行う必要がなく、実用性が高い。また、電力の供給安定性と燃料使用量制限を最適生産活動モデルの制約とすることにより、生産活動における利益最大を達成することができる。また、各設備による発電端電力量を最適生産活動モデルの決定変数にすることにより、最適生産活動モデルから最適生産活動価値を求める際に、各設備をどれだけ運転させたらよいかを示す最適運転量（最適発電端電力量）を得ることができるため、設備投資に関する最適な意思決定だけでなく、設備運用に関する最適な意思決定にも貢献できる。
【０１０８】
また、経費・実積データＤ２の設定に当たっては、必要なデータを手作業で逐一入力することも可能であるが、過去の有価証券報告書に記載されているデータ等の、業者が保存している各種の過去データを読み込んでそのままデータ初期値として使用できる。このように、既存のデータを読み込んでそのまま使用した場合には、電源計画意思決定用に特別な入力データを用意する手間が不要となり、その分だけ、電源計画を導出するための作業効率を向上することができる。
【０１０９】
一方、Ｔ期間最適電源計画の評価結果として、図７に示すように、電力需要量等の不確実性要因の想定値とＴ期間における最適な設備投資案の選択傾向との関係をグラフ表示することにより、意思決定者であるユーザは、Ｔ期間における設備投資の種類や回数、投資時期等の設備投資傾向を、意思決定結果として出力されたグラフから容易に把握することができる。また、電力需要量等の不確実性要因状況の変化が設備投資傾向に与える影響を、グラフから容易に把握することができる。
【０１１０】
また、Ｔ期間最適電源計画の評価結果として、図８に示すように、多様な想定シナリオの元で算出した最適生産活動価値の価値分布をグラフ表示することにより、多様な想定シナリオにおける最適生産活動価値が分布データとして得られるため、意思決定者であるユーザは、最大生産活動価値や最小生産活動価値だけでなく、平均や分散もグラフから容易に把握することができる。また、電源が足りなくなったら追加するという従来の電源計画手法による生産活動価値の利益分布と、本発明によって求めた最適生産活動価値の利益分布とを比較的にグラフ表示することにより、本発明の電源計画手法と従来の電源計画手法との効果の比較についても、客観的に容易に行うことができる。
【０１１１】
そしてまた、Ｔ期間最適電源計画の評価結果として、図９に示すように、最適生産活動モデルから導出される各設備の最適運転量をグラフ表示することにより、意思決定者であるユーザは、不確実性要因状況の変化が設備ごとの最適運転量に与える影響を、グラフから容易に把握することができる。
【０１１２】
［５．他の実施形態］
なお、本発明は、前述した実施形態に限定されるものではなく、本発明の範囲内で他にも多種多様な形態が実施可能である。
【０１１３】
まず、図１、図４〜図６に示した処理手順は、一例にすぎず、具体的な処理手順は自由に変更可能である。また、表１に示した投資案・事業状況データもまた、電源計画を導出する際に使用するデータの一例にすぎず、具体的に使用するデータは自由に変更可能である。さらに、本発明は、前述した通り、電源計画の導出に好適であるが、それに限らず、製造業等の、生産活動を行う各種の事業において、投資計画等の将来に亘る各種の意思決定を行う場合に幅広く適用可能であり、いずれの場合でも、同様に優れた効果が得られるものである。
【０１１４】
【発明の効果】
以上詳述したように、本発明によれば、将来の不確実性要因に対して想定される複数の状況と複数の意思決定案とを組み合わせ、各意思決定案について、各状況における最適生産活動価値指標と複数の状況により生じる価値変動リスク指標を求めることにより、将来の不確実性に起因するリスクを評価しながら同時に利益最大という生産活動の経済性を目指した合理的な意思決定を実現できる。したがって、最適な生産活動計画を導出可能な、有用な生産活動意思決定支援方法とそのためのプログラムを提供することができる。
【図面の簡単な説明】
【図１】本発明を適用した生産活動意思決定支援処理の概要を示すフローチャート。
【図２】図１に示す生産活動意思決定支援処理における最適生産活動価値指標算出の原理を説明する概念図。
【図３】図１に示す生産活動意思決定支援処理における価値変動リスク指標算出およびｔ〜Ｔ期間リスク調整済期待価値算出の原理を説明する概念図。
【図４】本発明を、電力供給事業における電源設備投資に関する意思決定に適用した場合の電源計画意思決定支援処理を示すフローチャート。
【図５】図４の処理中で行う「ｔ期目最適投資案決定処理」のアルゴリズムを示すフローチャート。
【図６】図５の処理中で行う「ｔ〜Ｔ期間リスク調整済期待価値算出処理」のアルゴリズムを示すフローチャート。
【図７】図４に示す電源計画意思決定支援処理において、「Ｔ期間最適電源計画の評価結果」として得られたグラフの一例を示す図。
【図８】図４に示す電源計画意思決定支援処理において、「Ｔ期間最適電源計画の評価結果」として得られた別のグラフの一例を示す図。
【図９】図４に示す電源計画意思決定支援処理において、「Ｔ期間最適電源計画の評価結果」として得られた別のグラフの一例を示す図。
【符号の説明】
Ｄ１．．．意思決定案・状況データ
Ｄ１ａ．．．投資案・事業状況データ
Ｄ１１．．．外部パラメータ
Ｄ２．．．経費・実績データ
Ｄ３．．．意思決定木
Ｄ４．．．リスク調整済期待価値データ[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a production activity decision support method and a program for supporting a decision for performing a production activity that produces maximum profit under an uncertain business environment, such as a decision regarding capital investment in an electric power supply business.
[0002]
[Prior art]
In various businesses that perform production activities, such as the manufacturing industry and the power supply business, it is necessary to make a capital investment plan for the future in order to carry out highly economical production activities. An investment plan will necessarily involve uncertain factors. For example, in the case of a capital investment plan in a power supply business, uncertain factors may include fossil fuel depletion, fuel price instability, future demand trends, environmental regulations, and the like. A capital investment plan under such an uncertainty situation needs to have not only economy but also flexibility and robustness against fluctuation of uncertain factors.
[0003]
Generally, investment planning problems are often dealt with in the field of financial engineering or mathematical optimization. Of these, the prior art in the field of financial engineering focuses on assessing the risks posed by future uncertainties and takes little account of the economics of production activities. Some prior art in the field of mathematical optimization considers the economics of production activities as well as uncertainties in the future, but applicable investment planning problems are limited.
[0004]
Hereinafter, a plurality of documents are listed as known examples of research related to the technology of the present invention.
[0005]
Reference 1: “BF Hobbs,“ Optimization methods for electricity resource planning ”, European Journal of Operational Research 83 (1995). It is modeled on the assumption, but does not take into account any risks arising from uncertainty.
[0006]
Reference 2: "Nishikawa / Tezuka / Kita / Nakano," Optimal Power Supply Planning under Uncertain Demand Using Scenario Integration Algorithm ", Transactions of the Institute of Electrical Engineers of Japan, Vol.112-C, No.6 (1992), pp356-pp. 363. ", the power supply planning problem is constructed by assuming several situations for uncertain demand. However, the power supply planning model is unrealistic and complicated, such as allowing a continuous amount of equipment, and requires a parallel computer, and there is still a gap in practical use.
[0007]
Reference 3: “Tanabe, Yasuda, Yokoyama,“ Robust Power Mix Decision Method Considering Uncertainty of Fixed Cost and Variable Cost of Power Source ”, Proc. Of the Institute of Electrical Engineers of Japan Vol.112-B, No.4 (1992), pp331 -338. ", The power supply planning problem is constructed by assuming several situations for uncertain power supply fixed costs and power supply variable costs. However, in order to derive the optimal power supply configuration for the target year, decision-making is performed only once.
[0008]
Reference 4: Japanese Patent Application Laid-Open No. 2001-125962 "Corporate Diagnosis and Management Decision Support System" proposes a system for performing management decision support by reflecting the future business state of a company in its present value. This technique calculates a cash flow with reference to a cash flow calculation rule module after inputting a determined simulation condition, and does not consider uncertainty.
[0009]
Reference 5: Japanese Patent Application Laid-Open No. H11-134387, "Repayment plan support system", proposes a repayment plan in which repayment of borrowing does not oppress management, a repayment plan support system that can make a capital investment plan, and a capital investment plan support system. . This technology simulates profit planning based on data on capital investment and borrowed content registered in advance and calculates the amount of repayment for each fiscal year after borrowing. Is not considered.
[0010]
Reference 6: Japanese Patent Application Laid-Open No. 2001-112171 proposes a technology for calculating a power demand expected in the future and deriving a countermeasure construction plan for a facility exceeding an available supply capacity. I have. This technology predicts future power demand based on past and present power demand data, calculates the cost of countermeasures for equipment that exceeds the available capacity, and does not consider uncertainties.
[0011]
[Problems to be solved by the invention]
As described above, in the prior art, there is no useful method for making a decision aiming at the economical efficiency of the production activity such that the profit is maximized while evaluating the risk caused by the uncertainty in the future.
[0012]
In particular, when making capital investment decisions in the power supply business, make a power supply plan that satisfies the economics of production activities and the stability of power supply, taking into account risks arising from future uncertainties. It is necessary. Here, the objectives of minimizing the risks caused by future uncertainties, aiming to achieve the maximum profit, and supplying to meet the power demand are sometimes conflicting, so that an optimal power plan is derived. In order to do so, their objectives need to be rationally adjusted. However, in the conventional capital investment method in the power supply business, there is no method capable of rationally adjusting the purpose.
[0013]
The present invention has been proposed to solve the above-mentioned problems of the prior art, and its object is to evaluate the risk caused by uncertainty in the future and at the same time to maximize the profit of production activities. It is an object of the present invention to provide a useful production activity decision support method capable of deriving an optimal production activity plan by realizing a rational decision making aiming at gender, and a program therefor.
[0014]
[Means for Solving the Problems]
The present invention combines a plurality of situations assumed for a future uncertainty factor and a plurality of decision making plans, and for each decision plan, an optimal production activity value index in each situation and a value generated by the plurality of situations. By calculating the fluctuation risk index, it is possible to evaluate the risks caused by uncertainties in the future and at the same time to realize rational decision making aiming at the economics of production activities to maximize profits. As described above, by using the optimal production activity value index and the value fluctuation risk index, it is possible to provide a method in which financial engineering and a mathematical optimization method are organically combined.
[0015]
The definitions of important terms in the present invention are as follows.
"Production activity" is a term indicating various activities for producing a certain object, but in the present invention, it means production activities performed by various companies such as a manufacturing industry and a power supply business with the aim of economic efficiency. are doing.
“Decision for conducting production activities” refers to the decision to derive some plan related to production activities, and is not limited to the decision to derive investment plans but also to various other production activity strategies. This is a concept that encompasses all decision making for deriving such a plan.
"Resources" refers to resources used for raw materials in the manufacturing business and for power generation in the power supply business.In particular, resources such as fuel that have a high economic uncertainty such as a high percentage of expenses and large price fluctuations Means
[0016]
The invention according to claim 1 uses a computer having an input / output unit, a storage unit, and an arithmetic processing unit to support a decision on a production activity for performing a production activity that produces a maximum profit under an uncertain business environment. The decision support method is characterized by having a data input step, a decision tree generation step, a decision step, and an output step.
[0017]
Here, in the data input step, the input / output unit inputs data relating to a plurality of decision making proposals for performing production activities, data relating to a plurality of situations assumed for a future uncertainty factor, and a storage unit. This is the step of storing in The decision tree generating step is a step in which the arithmetic processing unit generates a decision tree by combining a plurality of decision plans and a plurality of situations.
[0018]
The decision making step is a step of making a decision for a predetermined period in the future in which the production activity is performed by the arithmetic processing unit. In this decision step, an optimal production activity value index is calculated from the decision tree for each of a plurality of decision making plans in each of a plurality of situations, and an optimal production activity value index is calculated. Calculates a value fluctuation risk index indicating the magnitude of value fluctuation caused by a plurality of situations, and makes a decision for a predetermined period based on the optimal production activity value index and the value fluctuation risk index. Further, the output step is a step of outputting a result of the decision making by the input / output unit.
[0019]
A twelfth aspect of the present invention is obtained by grasping the first aspect of the present invention from the viewpoint of a program, and uses a computer having an input / output unit, a storage unit, and an arithmetic processing unit to maximize the profit under an uncertain business environment. In a program for supporting decision-making for performing a production activity that causes the following, the functions corresponding to each step in the invention of claim 1 include a data input function, a decision tree generation function, a decision-making function, an output function, Is realized by a computer.
[0020]
According to the invention as described above, future situations of uncertainty are assumed in a plurality of situations and combined with a plurality of decision plans, and for each decision plan, an optimal production activity value index in each situation and a plurality of By determining the value fluctuation risk index caused by the situation, it is possible to determine an optimal decision making plan from a plurality of decision making plans based on these indexes.
[0021]
In other words, by assuming the situation regarding uncertainty factors in the future and combining it with a decision-making plan, accurate decision-making takes into account not only the production activity value obtained in the current term but also the production activity value obtained in the next term and thereafter. be able to. In addition, by determining a value fluctuation risk index caused by a plurality of situations for each decision decision plan, a robust (flexible) decision can be made with respect to future fluctuations in the uncertainty factors. On the other hand, by using the optimal production activity value index obtained for each combination of the decision making plan and the situation, it is possible to make a decision aiming at achieving the maximum production activity value. Further, when obtaining the optimum production activity value index, it is also possible to obtain the optimum operation amount of each facility for achieving the optimum production activity value.
[0022]
In this way, by using the optimal production activity value index and the value fluctuation risk index, it is possible to organically combine financial engineering and mathematical optimization techniques, and to evaluate risks due to uncertainties in the future. At the same time, they can make rational decisions with the aim of maximizing profitability in production activities. Therefore, it is possible to derive an optimal production activity plan that can achieve the maximum profit while minimizing the risk due to future uncertainty.
[0023]
According to a second aspect of the present invention, in the production activity decision support method of the first aspect, when the decision making step makes a decision over a future T (T is a natural number of 1 or more) period, t = T The method is characterized in that the optimal plan determination step is repeated while the value of t is subtracted one by one from the start until t = 1. Here, in the optimal plan decision step, the evaluation when each decision plan is adopted in the t period is performed for each combination of the decision plan and the situation up to the (t-1) period, and a plurality of decision decisions are made. This is a step of determining an optimal decision-making plan in the t-th period from the plans.
[0024]
According to the present invention, when making a decision over the future T period, in each period of the T period, each decision plan is adopted for each combination of the decision plan and the situation up to the previous period. By performing each evaluation, the most appropriate decision can be made for each period. Therefore, it is possible to derive an optimal production activity plan for the production activities over the period T.
[0025]
According to a third aspect of the present invention, in the production activity decision support method of the second aspect, the optimal plan determining step includes an expected value calculating step, an optimal plan selecting step, and an expected value storing step. Here, in the expected value calculation step, when each of the plurality of decision-making plans is adopted in the t period, the optimum production activity model in each of the plurality of situations is constructed to calculate the optimal production activity value index. Based on the optimal production activity value index and the risk-adjusted expected value of the decision plan determined in the previous optimal plan decision step, a value change risk index caused by multiple situations is calculated, and each decision plan This is the step of calculating the risk-adjusted expected value in the period from t to T in the case where is adopted.
[0026]
The optimal plan selection step compares the risk-adjusted expected values in the t to T periods of the plurality of decision plans, and determines a decision plan that produces the maximum risk-adjusted expected value in the t-th period. This is the step of choosing as a plan. Further, the expected value storing step is a step of storing the maximum risk-adjusted expected value in the storage unit as the risk-adjusted expected value by the optimal production activity in the period from t to T.
[0027]
According to the present invention, in making a decision in the t-th period, assumed data such as past data such as cost data and actual data of production activities, and data relating to a situation assumed for an uncertainty factor in the future. Based on the above, when each decision plan is adopted for each combination of the decision plan and the situation up to the previous term, the optimal production activity in each situation can be mathematically modeled and the optimal production activity value index can be obtained . Then, based on the “optimum production activity value index” and the calculated “risk-adjusted expected value” in each situation, and the “value fluctuation risk index” generated by a plurality of situations, the optimal decision-making in the t-th period is performed. It can be carried out. Therefore, for the decision making over the T period, the weight of the two objectives of maximizing the value of the production activity over the T period and minimizing the risk of value fluctuation due to the uncertainty factor can be appropriately adjusted, so that the decision over the T period can be appropriately adjusted. An optimal production activity plan can be derived for a production activity.
[0028]
According to a fourth aspect of the present invention, in the production activity decision support method according to any one of the first to third aspects, the plurality of decision making plans are capital investment plans relating to capital investment in a power supply business. .
According to a fifth aspect of the present invention, in the production activity decision support method of the fourth aspect, the uncertainty factors include power demand and resource cost.
[0029]
According to the above invention, it is possible to make an optimal decision regarding a future capital investment in a power supply business. The present invention is suitable for decision making for various production activities, but in particular, in decision making relating to capital investment in a power supply business, important factors such as power demand, resource cost, power selling price, and power procurement cost are important. Since it is easy to narrow down the types of uncertainty factors and the number of situations, it is possible to efficiently and optimally make a decision by applying the present invention.
[0030]
According to a sixth aspect of the present invention, in the production activity decision support method according to the fourth or fifth aspect, the capital investment plan is to add a new facility, repair an existing facility, replace the existing facility with a new facility, and abolish a suspended facility. It is characterized by including plans selected from among the above.
[0031]
According to the present invention, by setting new equipment, renovating existing equipment, replacing existing equipment with new equipment, and disposing of suspended equipment, etc., the optimal introduction time of new equipment, the installation form and the existing equipment It is possible to easily derive an optimal disposal time, a disposal form, and the like. In addition, by setting the presence or absence of addition of a new facility as a decision-making plan, it is possible to compare the introduction effects in future periods for a plurality of new facilities whose introduction is being considered.
[0032]
According to a seventh aspect of the present invention, in the production activity decision support method according to any one of the fourth to sixth aspects, the output step includes, as a result of the decision, an estimated value of an uncertainty factor and an optimal value in the predetermined period. A step of outputting data for graphically displaying the relationship between the selected capital investment plan and the tendency of selection.
[0033]
According to the present invention, it is possible to easily grasp capital investment trends such as the type and number of capital investments and the investment period during a predetermined period from a graph output as a decision-making result. In addition, the effect of the change in the uncertainty factor situation on the capital investment tendency can be easily grasped from the graph.
[0034]
According to an eighth aspect of the present invention, in the production activity decision support method according to any one of the fourth to seventh aspects, the output step includes, as a result of the decision, an optimal production calculated based on each of a plurality of situations. The method is characterized by including a step of outputting data for displaying the activity value as a value distribution graph.
[0035]
According to the present invention, the optimal production activity value in various assumed situations of the uncertainty factor can be obtained as distribution data, so that not only the maximum production activity value and the minimum production activity value but also the average and variance can be easily grasped from the graph. can do. Also, by expressing the obtained optimal production activity value as a value distribution graph, it is also possible to compare the production activity value obtained by the conventional power supply planning method of adding power when power becomes insufficient and the optimal production activity value obtained by the present invention. It can be done objectively and easily.
[0036]
According to a ninth aspect of the present invention, in the production activity decision support method according to any one of the fourth to eighth aspects, the optimal production activity model restricts power supply stability and resource usage restrictions, and generates power by each facility. It is characterized that the end electric energy is used as a decision variable of the optimal production activity model, and the model aims at maximizing profit in the production activity.
[0037]
According to the present invention, the maximum profit in the production activity can be achieved by using the stability of the power supply and the limitation of the fuel consumption as constraints of the optimal production activity model. In addition, by using the power generated by each facility as the determining variable of the optimal production activity model and the upper and lower limits of the operation of each facility as constraints for the optimal production activity model, the optimal production activity value can be obtained from the optimal production activity model. In addition, since it is possible to obtain the optimal operation amount (optimum power generation end power amount) that indicates how much each equipment should be operated, it is possible to not only make the optimal decision on capital investment but also the optimal decision on equipment operation. Can also contribute.
[0038]
According to a tenth aspect of the present invention, in the production activity decision support method according to any one of the fourth to ninth aspects, the optimal production activity model is formulated as a linear programming problem having a special structure. I have.
According to the present invention, since the greedy solution method can be applied by using the problem structure of the optimal production activity model, it is possible to calculate the production activity value with a small amount of calculation, so that an actual-scale capital investment plan is made. When doing so, there is no need to perform enormous calculations, and the utility is high.
[0039]
According to an eleventh aspect of the present invention, in the production activity decision support method according to the ninth aspect, the output step outputs, as a result of the decision, data for graphically displaying the optimal operation amount of each facility derived from the optimal production activity model. It is characterized by including the step of performing.
According to the present invention, the influence of the change in the uncertainty factor situation on the optimal operation amount for each facility can be easily grasped from the graph.
[0040]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be specifically described with reference to the drawings. However, the embodiments described here do not limit the present invention in any way, but merely exemplify one embodiment of the present invention.
[0041]
The present invention is typically realized by controlling a computer with software. The software in this case realizes the operation and effect of the present invention by physically utilizing the hardware of the computer, and a suitable conventional technology is applied to a portion to which the conventional technology can be applied. Further, specific types and configurations of hardware and software for realizing the present invention, a range to be processed by software, and the like can be freely changed. For example, a program for realizing the present invention is one embodiment of the present invention.
[0042]
[1. Overview of production activity decision support processing]
FIG. 1 is a flowchart showing an outline of a production activity decision support process to which the present invention is applied. As shown in FIG. 1, in the production activity decision support, first, using an input / output unit and a storage unit basically provided in a computer, a decision plan or the like set by a user who is a decision maker is provided. Data is input from the input / output unit and stored in the storage unit (S110).
[0043]
Here, the data to be input and stored include data on a plurality of decision making plans for making future decisions in production activities, data on a plurality of situations assumed for future uncertainty factors, etc. In addition to data (decision-making plan / situation data D1), past data (expense / actual data D2) such as expense data and production activity actual data are included. When inputting these data, it is possible to automatically read existing data, but even in this case, the decision on which data to use as individual data used for decision making is determined by the decision maker. Is performed by the user.
[0044]
Next, using a stored decision plan / situation data D1, a plurality of decision plans and a plurality of situations assumed for the uncertainty factor are calculated by an arithmetic processing unit basically provided in the computer. To generate a decision tree D3 and store it in the storage unit (S120). Then, using the stored decision tree D3 and cost / actual data D2, the arithmetic processing unit makes an optimal decision over a future T period (T is a natural number of 1 or more) in which the production activity is performed (S130). ). The details of this decision are as follows.
[0045]
First, the value of a variable t indicating each period of the T period is set to T (S131). Then, in order to perform the optimal decision-making in the t-th period, each decision-making plan is adopted in the t-th period from the decision tree D3 for each combination of the decision-making plan and the situation up to the (t-1) -th period. In this case, the optimal production activity model in each situation is constructed using the stored expense / actual data D2, and the optimal production activity value index is calculated (S132). FIG. 2 is a conceptual diagram illustrating the principle of calculating the optimum production activity value index.
[0046]
Based on the optimal production activity value index calculated for each of the plurality of situations and the risk-adjusted expected value data D4 of the stored optimal investment plan, all situations when each decision plan is adopted in the t-th period Then, a value fluctuation risk index indicating the magnitude of the value fluctuation caused by is calculated (S133). The expected value in the period from t to T obtained from the optimal production activity value index is adjusted based on the value fluctuation risk index, and the risk-adjusted expected value in the period from t to T when each decision plan is adopted in the t period is calculated. It is calculated (S134). FIG. 3 is a conceptual diagram illustrating the principles of such value fluctuation risk index calculation and t-T period risk-adjusted expected value calculation.
[0047]
Based on the risk-adjusted expected values of the t to T periods respectively calculated for all of the plurality of decision plans, an optimal decision plan to be adopted in the t period is selected, and the t of the optimal decision plan is selected. The data relating to the risk-adjusted expected value in the period from T to T is stored in the storage unit as risk-adjusted expected value data D4 (S135).
[0048]
A series of steps (S132 to S135) for performing the above-described optimal decision making in the t-th period is repeated until the value of t is reduced by 1 until t = 1 (NO in S136, S137). . Then, when t = 1, that is, when the optimal decision-making plan to be adopted in the first period is selected, and the risk-adjusted expected value data D4 for the period from t to T is stored in the storage unit ( In S136 (YES), the decision making (S130) over the T period is ended, and the process proceeds to the next step S140.
[0049]
In step S140, the result of the optimal decision-making (S130) over the T period is output by the input / output unit and displayed on the screen in the form of a graph or the like.
[0050]
[2. Basic effects of production activity decision support processing]
According to the above-described production activity decision support process, the future production uncertainty factor is assumed to be in multiple situations, combined with multiple decision plans, and the optimal production activity in each situation is determined for each decision plan. By finding a value index and a value change risk index caused by a plurality of situations, an optimal decision-making plan can be determined from a plurality of decision-making plans based on these indices.
[0051]
In other words, by making assumptions about future uncertainty factors and combining them with decision making proposals, we will make accurate decisions taking into account not only the production activity value obtained in the current term but also the production activity value obtained in the next term and thereafter. be able to. In addition, by determining a value fluctuation risk index caused by a plurality of situations for each decision-making plan, a robust (flexible) decision-making can be performed with respect to future fluctuations in uncertainty factors.
[0052]
On the other hand, by using the optimal production activity value index obtained for each combination of the decision making plan and the situation, it is possible to make a decision aiming at achieving the maximum production activity value. In particular, by using the problem structure of the optimal production activity model to calculate the optimal production activity value index, it is possible to calculate the production activity value with a small amount of calculation, so it is enormous when planning an actual-scale capital investment plan. There is no need to perform calculations and the utility is high. Further, when obtaining the optimum production activity value index, it is also possible to obtain the optimum operation amount of each facility for achieving the optimum production activity value.
[0053]
In this way, by using the optimal production activity value index and the value fluctuation risk index, it is possible to organically combine financial engineering and mathematical optimization techniques, and to evaluate risks due to uncertainties in the future. At the same time, they can make rational decisions with the aim of maximizing profitability in production activities. Therefore, it is possible to derive an optimal production activity plan that can achieve the maximum profit while minimizing the risk due to future uncertainty.
[0054]
Further, when making a decision over the future T period, an optimal decision can be made in each period of the T period. In other words, when making a decision in the t-th period, based on past data such as cost data and actual data of production activities, and based on assumed data such as data on situations assumed for uncertainty factors in the future, When each decision plan is adopted for each combination of the decision plan and the situation up to the previous term, the optimal production activity in each situation is mathematically modeled, and the optimal production activity value index can be obtained. Then, based on the “optimum production activity value index” and the calculated “risk-adjusted expected value” in each situation, and the “value fluctuation risk index” generated by a plurality of situations, the optimal decision-making in the t-th period is performed. It can be carried out.
[0055]
As described above, since an optimal decision can be made for each period of the T period, the value of the production activity over the T period is maximized, the value fluctuation risk due to the uncertainty factor is minimized, and the decision over the T period is made. The specific gravity of the two objectives can be appropriately adjusted. Therefore, in order to carry out production activities with the aim of maximizing profits in an uncertain business environment over the period T, when and how much investment should be executed in what time in the future , And an optimal investment plan can be derived.
[0056]
[3. Power supply plan decision support processing]
FIG. 4 is a power supply plan decision support process in which the present invention is applied to a decision regarding power supply investment in a power supply business, and an optimal power supply investment plan (hereinafter abbreviated as a power supply plan) is derived. It is a flowchart which shows. FIG. 5 is a flowchart showing an algorithm of the “t-th period optimal investment plan determination process” performed in the process of FIG. 4. FIG. 6 is a flowchart showing “t-T period risk adjusted” performed in the process of FIG. It is a flowchart which shows the algorithm of "expected value calculation processing."
[0057]
In the power supply capital investment, the decision-making plan is, specifically, a capital investment plan such as addition or replacement of equipment (hereinafter abbreviated as investment plan), and the main uncertainty factors are: These are resource costs such as fuel prices and electricity demand. Therefore, in the power supply plan decision support process shown in FIG. 4, first, the input / output unit of the computer uses the input / output unit of the computer to generate investment plan data relating to a plurality of investment plans such as addition or replacement of equipment as decision plan / status data D1. , Data of a plurality of business conditions assumed for uncertainty factors such as fuel prices and power demand, etc. (investment proposal / business status data D1a), and input of expense / actual data D2. Is stored in the storage unit (S410).
[0058]
When inputting these data, as described above, it is possible to automatically read existing data, but even in that case, it is still necessary to set which data to use as individual data used for decision making Is performed by a user who is a decision maker. For the investment proposal / business status data D1a, individual investment proposals that clearly express different methods of capital investment, such as the addition or replacement of equipment, are set, and the main factors of uncertainty are fuel price and electricity demand. Set individual business situations and their probabilities of occurrence that clearly express the tendency of
[0059]
For example, as a plurality of investment proposals, investment proposals such as “addition of new equipment”, “repair of existing equipment”, “discard existing equipment”, and “do not invest in equipment” can be considered. In addition, several business conditions assumed for uncertainties such as fuel prices and power demand include “fuel price rise, power demand increase”, “fuel price drop, power demand increase”, “fuel price rise”. , A decrease in fuel demand, a decrease in fuel prices, a decrease in power demand, etc. Table 1 below shows an example of such investment proposal / business status data, in which individual occurrence probabilities are set for each business status.
[0060]
[Table 1]

[0061]
In addition, the business status is not only roughly classified as shown in Table 1, but also “increase by 10%”, “decrease by 10%”, “increase by 20%”, “decrease by 20%”,. It is also possible to set more detailed business conditions, such as "n% increase" and "n% decrease", and to consider the business conditions for uncertain factors other than fuel prices and electricity demand. Can be Conversely, it is also possible to set a smaller number of business situations, such as setting only the business situations of “power demand increase” and “power demand decrease” for only the power demand. Similarly, for a plurality of investment proposals, it is possible to set investment proposals that are classified in more detail, and conversely, it is possible to set only a smaller number of investment proposals.
[0062]
Further, in setting the expense / actual product data D2, data at the level described in the past securities report can be read and used. In other words, the expense data can be obtained from the financial statement data in the securities report, and the actual data on the power supply by the power supply equipment, which is a production activity, can be obtained from the business status data and the equipment status data in the securities report. Thus, past operation result data, existing facility status data, and the like can be obtained. These data are not limited to the data described in the securities report, but can be similarly obtained from various past data stored by the trader.
[0063]
The data read in this way is displayed on the screen by the input / output unit to the user who is the decision maker, and the user selects which data to use, and the selected data is used as the data to be actually used. Stored in the storage unit. Therefore, the user who is the decision maker can easily set the cost / actual product data D2 only by selecting which data to use from the data displayed on the screen.
[0064]
Then, the arithmetic processing unit sets various external parameters D11 necessary for the subsequent processing using the stored investment plan / business status data D1a and the cost / actual product data D2, and stores them in the storage unit (S420). ). In the setting of the external parameters, the cost and the fuel consumption for each unit are calculated using the cost / result data D2, and are set as the parameters of the first year results. In addition, parameters such as the fluctuation range of the uncertainty factor are set using the investment proposal / business status data D1a and the like. The various parameters that have been set are presented on the screen by the input / output unit to the user who is the decision maker, and are stored in the storage unit when the user's confirmation is obtained.
[0065]
Next, the arithmetic processing unit uses the stored investment proposal / business status data D1a to store a plurality of investment proposals such as addition or replacement of equipment, and uncertainties such as resource costs such as fuel prices and power demand. A decision tree D3 is generated by combining a plurality of business situations assumed for the factor, and stored in the storage unit (S430). Then, using the stored decision tree D3 and the external parameters D11, the arithmetic processing unit derives an optimal power supply plan over the T period as a decision decision over the future T period (S440). The details of the “derivation of the optimal power supply plan over the T period” are as follows.
[0066]
First, the value of a variable t indicating each period of the T period is set to T (S441). Then, for all combinations of the investment plan and the business status up to the (t-1) period, the optimal investment plan in the t period is determined for each combination (S442 to S444). That is, in order to make an optimal decision in the t-th period, a decision tree D3 is used for each period from the first period to the (t-1) period to determine one of the investment plans to be evaluated. Each investment plan, select one business situation from several business situations expected in the future, and combine the investment plan and situation combination selected for each period up to period (t-1). It is set as a precondition (S442). With respect to the preconditions up to the (t-1) period, the “t-th period optimal investment plan determination process” shown in FIG. 5 is performed, and the optimal t-th period optimal investment plan is determined using the decision tree D3 and the external parameter D11. An investment plan is determined (S443).
[0067]
In all combinations of the investment plan and the business situation up to the (t-1) period, while there remains a combination for which the optimal investment plan for the t period has not been determined (NO in S444), (t-1) The setting of the preconditions based on the combination of the investment plan up to the period and the business situation (S442) and the determination of the optimal investment plan in the t-th period (S443) are repeated. Then, when the optimal investment plan in the t-th period is determined for all combinations of the investment plan and the business status up to the (t-1) period (YES in S444), the process proceeds to the next step S445.
[0068]
Note that “all combinations of investment plans and business conditions up to period (t-1)” is considered as a result of selecting investment plans for each period up to period (t-1). Means all combinations of all investment plans and all assumed scenarios up to the period (t-1) which can be considered as a result of a change in the business situation in each period up to the period (t-1). For example, if the investment proposal is A, the business situation is B, the number of investment proposals is n, and the number of business situations is m, the total number of combinations of investment proposals and business situations up to the (t-1) period is n ^(T-1) × m ^(T-1) Individual.
[0069]
Explaining this point, first, all investment plans from the first period to the (t-1) period are each composed of columns A1,..., A including investment plans selected for each period. _(T-1) Since the number of investment plans that can be selected for each period is n, the number of all investment plans that can be selected from the first period to the (t−1) period is n. ^(T-1) Individual. Further, each of the assumed scenarios from the first period to the (t-1) period includes columns B1,..., B including the business conditions assumed for each period. _(T-1) Since the number of business situations that can be assumed for each period is m, the number of all assumed scenarios that can be assumed from the first period to the (t−1) period is m. ^(T-1) Individual. Therefore, the total number of combinations of investment plans and business conditions up to the period (t-1), that is, the total number of combinations of investment plans and assumed scenarios is n ^(T-1) × m ^(T-1) Individual.
[0070]
For example, in the example of Table 1, the number of investment proposals that can be selected for each period is four, and the number of business situations that can be assumed for each period is four, so the first period to (t−1) The number of all investment plans that can be selected by the period is 4 ^(T-1) And the number of all assumed scenarios that can be assumed from the first period to the (t-1) period is 4 ^(T-1) Individual. Therefore, in the example of Table 1, the total number of combinations of investment plans and business conditions up to the period (t-1), that is, the total number of combinations of investment plans and assumed scenarios is 4 ^(T-1) × 4 ^(T-1) Individual.
[0071]
A series of steps (S442 to S444) for making an optimal decision in the t-th period as described above is repeated until the value of t is reduced by one until t = 1 (NO in S445, S446). . Then, when t = 1, that is, when an optimal investment plan to be adopted in the first term with respect to the current business situation (initial value) is determined (YES in S445), the period T The determination (S440) is completed, and the process proceeds to the next step S450.
[0072]
In step S450, the arithmetic processing unit evaluates the derived optimal power plan for the T period, outputs the evaluation result through the input / output unit, and displays it on the screen in the form of a graph or the like. Here, the “evaluation results of the T-period optimal power supply plan” include “optimal introduction time according to future assumed scenarios”, “profit distribution by optimal production activities under various assumed scenarios”, “optimal production activities”. The optimal operation amount of each facility derived from the model "is displayed on a screen as a graph. In addition, a graph may be displayed on the screen, which converts the production activity value according to the conventional power supply plan of “add when power supply is insufficient” with the current activity value according to the optimal power supply plan and compares it.
[0073]
[3-1. t-term optimal investment proposal decision process]
In determining the optimal investment plan in the t-th period, as the investment plan in the t-th period, a plurality of investment plans set as investment plan data, for example, “addition of new equipment” as shown in Table 1 described above. ”,“ Repair of existing equipment ”,“ Disposal of existing equipment ”,“ No capital investment ”, etc.
[0074]
It is necessary to decide which investment plan should be adopted in the t-th period according to the business situation at that time. However, the optimal investment plan determination process in the t-th period as shown in FIG. Among them, by performing the risk-adjusted expected value calculation processing during the t-T period as shown in FIG. 6, it is possible to make an optimal investment decision in the t-th period in consideration of optimal production activities and risk management.
[0075]
In the t-period optimal investment plan determination process shown in FIG. 5, first, one candidate investment plan to be adopted in the t-th period is selected using the decision tree D3 stored in the storage unit (S501). Next, in the case where the selected investment plan is adopted in the t-th period, the “t-T period risk-adjusted expected value calculation process” shown in FIG. 6 is performed, and t is determined using the decision tree D3 and the external parameter D11. The risk-adjusted expected value for the period T to T is calculated (S502).
[0076]
While the unselected investment plan remains in all the investment plan candidates to be adopted in the t-th period (NO in S503), the selection of the investment plan candidate (S501) and the risk adjustment in the period from t to T are performed. The calculation of the expected expected value (S502) is repeated. When the risk-adjusted expected values for the t to T periods have been calculated for all the candidates for the investment plan (YES in S503), the process proceeds to the next step S504.
[0077]
For example, in the example of Table 1 described above, as a plurality of investment proposals, four investment proposals of “addition of new equipment”, “repair of existing equipment”, “discard existing equipment”, and “do not invest in equipment” are provided. If the risk-adjusted expected value for the period from t to T has been calculated for all (YES in S503), the process proceeds to the next step S504.
[0078]
In step S504, the risk-adjusted expected values of the t to T periods calculated for all the investment plans are compared, and among all the investment plans, the investment that produces the largest risk-adjusted expected value in the t to T period The plan is selected as the optimal investment plan in period t. Then, the maximum risk-adjusted expected value obtained by the selected optimal investment plan in the t-th period is stored in the storage unit as “Risk-adjusted expected value by the optimal production activity in the period from t to T” D4 (S505). The t-th period optimal investment plan determination process S443 is completed, and the process returns to step S444 in FIG.
[0079]
[3-2. t-T period risk adjusted expected value calculation processing]
In the calculation of the risk-adjusted expected value during the period t to T, the expected value is calculated for all of a plurality of business situations set as business situation data, which are assumed for uncertain factors such as fuel prices and power demand. Calculating (S601 to S605), taking into account the value fluctuation risk due to the business situation, and calculating the risk-adjusted expected value in the t to T periods when each investment plan is adopted in the t period (S501 in FIG. 5). .
[0080]
In the t-T period risk-adjusted expected value calculation processing shown in FIG. 6, first, one candidate of the business situation assumed in the t-th period is selected using the decision tree D3 stored in the storage unit. (S601).
[0081]
Next, using the decision tree D3 and the external parameter D11 stored in the storage unit to adopt the investment plan selected in the t-th period (S501 in FIG. 5), the t-th period in the selected business situation An optimal production activity model is constructed, and the “maximum production activity value in the t period”, which is the optimal production activity value index in the t period, and the “optimum power generation end electric energy of each facility” are calculated (S602). That is, the optimal production activity model is formulated as a linear programming problem, and the optimal value is “the maximum production activity value in period t” and the optimal solution is “the optimal power generation end electric energy of each facility”. Details of the model will be described later.
[0082]
The “maximum production activity value in period t” calculated for the selected business situation is added to the calculated “expected value after risk adjustment by optimal production activities in period (t + 1) to period T” D4 stored in the storage unit. In addition, when the selected investment plan is adopted (S501 in FIG. 5), the value of the period t to T in the selected business situation is calculated (S603).
[0083]
When the investment plan selected in the t-th period is adopted (S501 in FIG. 5), while the unselected business status remains in all of the assumed business status candidates (S604: NO), the business A series of steps (S601 to S603) from the selection of the situation candidate to the calculation of the value during the period from t to T are repeated. Then, when the value of the period from t to T is calculated for all the candidates of the business status (YES in S604), the process proceeds to the next step S605.
[0084]
For example, in the example of Table 1 described above, as a plurality of business situations, “fuel price rise, power demand increase”, “fuel price drop, power demand increase”, “fuel price rise, power demand decrease”, “fuel price rise” If the value of the period from t to T has been calculated for all of the four business situations of “decrease, decrease in power demand” (YES in S604), the process proceeds to the next step S605.
[0085]
In step S605, the expected value of the t to T period is calculated from the value of the t to T period generated by all assumed business situations. That is, the number of business situations is s, and the probability of occurrence of each business situation i is p _i (I = 1, 2, Λ, s), and the value of the period from t to T caused by each business situation i is represented by r. _i (I = 1, 2, Λ, s), the expected value E [r] in the period from t to T is calculated by the following equation.
(Equation 1)

[0086]
Subsequently, a value fluctuation risk index of the t to T period value generated by all assumed business situations is calculated (S606). That is, the value fluctuation risk index V [r] of the t to T period value is represented by the number of business situations s and the occurrence probability p of the business situation i. _i And the value r of the period from t to T _i (I = 1, 2, Λ, s) and the expected value E [r] during the period from t to T is calculated by the following equation.
(Equation 2)

In addition, as a calculation method of the value fluctuation risk index, for example, a method of regarding a maximum loss such as VaR that can occur at 95% as a value fluctuation risk index may be adopted.
[0087]
By setting a discount rate to the present value according to the magnitude of the value fluctuation risk index, the expected value E [r] in the period from t to T is adjusted, and the risk-adjusted expected value in the period from t to T is calculated ( S607). For example, when the value fluctuation risk index is large, that is, when the value fluctuation is large, it is considered that the future uncertainty is high, and the expected value in the period from t to T is greatly discounted and adjusted. Specifically, the discount rate can be determined with reference to the risk premium in the stock market.
[0088]
Through such a series of t to T period risk-adjusted expected value calculation processing, when the investment plan selected in the t period is adopted (S501 in FIG. 5), the t to T period according to all assumed business conditions. Risk-adjusted expected value of
[0089]
[3-3. Optimal production activity model]
Hereinafter, the “variable”, “objective function”, and “constraint expression” of the optimal production activity model used in the risk adjusted expected value calculation processing during the t to T period illustrated in FIG. 6 will be described. Here, two types of capital investment proposals, “addition of new equipment” and “no capital investment” are considered. Whether new equipment i is added or not in the j-th period, x _ij = 1 or x _ij = 0. The variable x _ij Can be similarly formulated by increasing. Here, in order to simplify the explanation, upper and lower limit operation amounts of each facility and stability of the power supply amount are considered as constraint expressions.
[0090]
The “variable” is the power generation end electric energy in the t-th period under the investment plan x and the business state s when the equipment i purchased in the j-th period is used. Here, j ≦ t. In the following, the power generation end electric energy in the t-th period is represented by y _ij Is expressed as
The "objective function" is to maximize the profit when the production activity is performed in the t-th period by adopting the investment plan x and assuming the business state s, and this profit is expressed by the following equation. Is done.
[Equation 3]

[0091]
As the “constraint equation”, the following three constraint equations are used in order to restrict the upper and lower limit operation amounts of each facility and the stability of the power supply amount.
[0092]
“Constraint equation 1”: The power supply amount by the equipment i purchased in the j-th period satisfies the upper and lower limit operation amount constraint expressed by the following equation.
(Equation 4)

[0093]
"Constraint equation 2": The total power supply (the sum of the power at the transmitting end) is equal to the annual power demand. This is one of the power supply stability constraints, and is expressed by the following equation.
(Equation 5)

[0094]
“Constraint formula 3”: has a total installed capacity that satisfies the instantaneous demand × reserve rate. This is also one of the power supply stability constraints, and is expressed by the following equation.
(Equation 6)

[0095]
Therefore, the profit maximization problem in the t-th period when the investment plan x is adopted and the business state s is assumed, that is, the optimal production activity model is formulated as the following linear programming problem.
(Equation 7)

[0096]
In addition, the definition of each element used in each of the above expressions expressing the optimal production activity model is as shown in Table 2 below.
[Table 2]

[0097]
[3-4. Specific example of evaluation result of T period optimal power supply plan]
FIGS. 7 to 9 show various graphs obtained as “evaluation results of the T-period optimal power supply plan” in the power supply plan decision support processing shown in FIG. Hereinafter, these graphs will be sequentially described.
[0098]
First, Figure 7 shows that the first period is up to 1998, the second period is set at three-year intervals, and the T period for all six periods up to 2013 is set. The power supply plan shown in FIG. 4 assuming the investment proposals in Table 1, that is, four investment proposals of “addition of new equipment”, “repair of existing equipment”, “discard existing equipment”, and “do not invest in equipment” It is a graph which shows the relationship between the amount of electric power required when a decision support process is performed, and the optimal investment time.
[0099]
In FIG. 7, the abscissa represents the power demand which is the uncertainty factor value determined by the assumed situation, and the ordinate represents the optimal investment time under the situation. Furthermore, in practice, not only the type of plot but also the color is different depending on the investment plan, so that the different investment plans can be expressed more clearly. By displaying such a graph on the screen, a user who is a decision maker can easily confirm the optimal introduction time according to a scenario assumed in the future such as the annual power demand.
[0100]
FIG. 8 shows, as a plurality of business conditions, the business conditions of Table 1, namely, “fuel price increase, power demand increase”, “fuel price decrease, power demand increase”, “fuel price rise, power demand decrease”, Optimal production under various assumed scenarios when the power supply plan decision support process shown in Fig. 4 is performed, assuming the four business situations of "fuel price decrease and power demand decrease" for each period. It is a graph which shows profit distribution by activity.
[0101]
In particular, in FIG. 8, the production activity value based on the conventional power supply plan of “add when the power supply is insufficient” and the optimal power supply plan derived by the power supply plan decision support process of the present embodiment shown in FIG. 4. Production activity value is converted to present value and compared. FIG. 8 shows the difference between the profit distribution based on the optimal power supply plan derived by the present embodiment and the profit distribution based on the conventional power supply plan, which is derived according to the present embodiment, based on the type of plot. Instead, the profit distribution of each plot can be more clearly expressed by using different colors for the plot and the curve connecting the plots between the optimal power plan and the conventional power plan.
[0102]
In FIG. 8, the overall shape and peak of the curve indicating the profit distribution based on the optimal power supply plan derived according to the present embodiment are shifted rightward from the entire shape and peak of the curve indicating the profit distribution according to the conventional power supply plan. Thus, it can be seen that when the optimal power supply plan according to the present embodiment is adopted, a much larger cash flow can be obtained as compared with the conventional plan.
[0103]
FIG. 9 is a graph showing the relationship between the amount of power demand and the amount of operation by fuel when a power supply plan including equipment using petroleum, LNG, and coal as fuel is derived. In the power supply plan decision support process of the present embodiment, the optimal operation amount of each facility according to the future scenario can be derived from the optimal production activity model. In addition, the user who is a decision maker can easily confirm the optimum operation amount of each facility according to a scenario assumed in the future such as the annual power demand.
[0104]
[4. Effects of Power Planning Decision Support Processing]
According to the above-described power supply plan decision support processing, the following function and effect can be obtained in addition to the basic function and effect of the production activity decision support processing described above.
[0105]
First, according to the power supply plan decision-making support processing of the present embodiment, it is possible to make an optimum decision regarding future capital investment in a power supply business. The present invention is suitable for decision making for various production activities. In particular, in decision making regarding capital investment in a power supply business, as shown in Table 1, resource costs such as fuel prices and electricity costs are reduced. Since it is easy to narrow down the types of important uncertainty factors and the number of situations, such as demand, by applying the present invention, it is possible to make optimal decisions efficiently and derive an optimal power supply plan. it can.
[0106]
In addition, as shown in Table 1, by setting new investment, renovation of existing equipment, disposal of existing equipment, etc. as investment proposals, the optimal introduction time, introduction form and existing equipment It is possible to easily derive a disposal time, a disposal form, and the like. Furthermore, by setting the presence or absence of addition of a new facility as a decision-making plan, it is possible to compare the introduction effects in future periods for a plurality of new facilities whose introduction is being considered.
[0107]
Further, by using the problem structure of the optimal production activity model for the calculation of the optimal production activity value index in the t-th optimal investment proposal determination process shown in FIGS. 5 and 6, the production activity value can be calculated with a small amount of calculation. Therefore, there is no need to perform enormous calculations when making an actual-scale capital investment plan, which is highly practical. Further, by making the power supply stability and the fuel consumption restriction as constraints of the optimal production activity model, it is possible to achieve the maximum profit in the production activity. In addition, by using the power generated by each facility as the determining variable in the optimal production activity model, the optimal operation that indicates how much each facility should be operated when determining the optimal production activity value from the optimal production activity model. Since it is possible to obtain the amount (optimum power at the generating end), it is possible to contribute not only to the optimal decision-making regarding capital investment but also to the optimal decision-making regarding equipment operation.
[0108]
In setting the expense / actual data D2, it is possible to manually input necessary data one by one. However, the data is stored and stored by a trader such as data described in a past securities report. Various past data can be read and used as data initial values as they are. In this way, when existing data is read and used as is, there is no need to prepare special input data for power plan decision-making, and the work efficiency for deriving the power plan is improved accordingly. can do.
[0109]
On the other hand, as the evaluation result of the T-period optimal power supply plan, as shown in FIG. 7, the relationship between the assumed value of the uncertainty factor such as the power demand and the selection tendency of the optimal capital investment plan in the T period is displayed in a graph. Thus, the user who is a decision maker can easily grasp the capital investment tendency such as the type and number of capital investments and the investment period in the T period from the graph output as the decision result. Further, it is possible to easily understand from a graph the influence of changes in the uncertainty factors such as the power demand on the capital investment tendency.
[0110]
In addition, as shown in FIG. 8, as the evaluation result of the T-period optimal power supply plan, the value distribution of the optimal production activity value calculated based on various assumed scenarios is displayed in a graph, so that the optimal production activities in various assumed scenarios are displayed. Since the value is obtained as distribution data, the user as a decision maker can easily grasp not only the maximum production activity value and the minimum production activity value but also the average and the variance from the graph. In addition, the profit distribution of the production activity value by the conventional power supply planning method of adding when the power supply runs short, and the profit distribution of the optimum production activity value obtained by the present invention are relatively graphically displayed, so that the present invention The comparison between the effects of the power supply planning method and the conventional power supply planning method can also be objectively and easily performed.
[0111]
Further, as an evaluation result of the T-period optimal power supply plan, as shown in FIG. 9, the optimal operation amount of each facility derived from the optimal production activity model is graphically displayed, so that the user who is a decision maker can obtain The effect of the change in the certainty factor situation on the optimal operation amount for each facility can be easily grasped from the graph.
[0112]
[5. Other Embodiments]
Note that the present invention is not limited to the above-described embodiment, and various other various embodiments can be implemented within the scope of the present invention.
[0113]
First, the processing procedures shown in FIGS. 1 and 4 to 6 are merely examples, and specific processing procedures can be freely changed. Further, the investment proposal / business status data shown in Table 1 is also merely an example of data used when deriving a power supply plan, and the data specifically used can be freely changed. Furthermore, as described above, the present invention is suitable for deriving a power supply plan, but is not limited to this. In various businesses that perform production activities, such as the manufacturing industry, it is possible to make various future decisions such as investment plans. The method can be widely applied when performing the method, and in any case, similarly excellent effects can be obtained.
[0114]
【The invention's effect】
As described in detail above, according to the present invention, a plurality of situations assumed for a future uncertainty factor and a plurality of decision plans are combined, and for each decision plan, the optimal production activity in each situation is determined. By determining the value index and the value fluctuation risk index caused by multiple situations, it is possible to evaluate the risk caused by uncertainty in the future and at the same time realize rational decision making aiming at the economic efficiency of production activities to maximize profit. . Therefore, a useful production activity decision support method capable of deriving an optimal production activity plan and a program therefor can be provided.
[Brief description of the drawings]
FIG. 1 is a flowchart showing an outline of a production activity decision support process to which the present invention is applied.
FIG. 2 is a conceptual diagram illustrating the principle of calculating an optimal production activity value index in the production activity decision support processing shown in FIG.
FIG. 3 is a conceptual diagram illustrating the principle of calculating a value fluctuation risk index and calculating a risk-adjusted expected value during a period from t to T in the production activity decision support processing shown in FIG. 1;
FIG. 4 is a flowchart showing a power supply plan decision support process when the present invention is applied to a decision regarding power supply equipment investment in a power supply business.
FIG. 5 is a flowchart showing an algorithm of “t-period optimal investment plan determination processing” performed during the processing of FIG. 4;
FIG. 6 is a flowchart showing an algorithm of “expected value calculation processing for risk adjustment during a period from t to T” performed during the processing of FIG.
FIG. 7 is a view showing an example of a graph obtained as “evaluation result of T-period optimal power supply plan” in the power supply plan decision support processing shown in FIG. 4;
8 is a diagram showing an example of another graph obtained as “evaluation result of T-period optimal power supply plan” in the power supply plan decision support process shown in FIG. 4;
9 is a view showing an example of another graph obtained as “evaluation result of T-period optimal power supply plan” in the power supply plan decision support processing shown in FIG. 4;
[Explanation of symbols]
D1. . . Decision making and situation data
D1a. . . Investment proposal / business status data
D11. . . External parameters
D2. . . Expense / actual data
D3. . . Decision tree
D4. . . Risk-adjusted expected value data

Claims (12)

Using a computer having an input / output unit, a storage unit, and an arithmetic processing unit, in a production activity decision support method for supporting a decision to perform a production activity that produces maximum profit under an uncertain business environment,
The input / output unit inputs data relating to a plurality of decision making plans for performing the production activities, data relating to a plurality of situations assumed for a future uncertainty factor, and stores the data in the storage unit. Steps and
The arithmetic processing unit, a decision tree generating step of generating a decision tree by combining the plurality of decision plans and the plurality of situations,
The arithmetic processing unit constructs an optimal production activity model in each of the plurality of situations and calculates an optimal production activity value index for each of the plurality of decision plans from the decision tree. A value fluctuation risk index indicating the magnitude of value fluctuation caused by the plurality of situations is calculated from the value index, and an intention over a predetermined period in the future to perform the production activity based on the optimal production activity value index and the value fluctuation risk index. A decision making step to make a decision;
An output step of outputting a result of the decision by the input / output unit;
A production activity decision support method comprising: The decision making step includes:
When making a decision over the future T period (T is a natural number of 1 or more), the decision-making plan up to the (t-1) period while starting from t = T and decrementing the value of t by one. An evaluation is performed when each decision plan is adopted in the t period for each combination of and the situation, and an optimal plan decision is made to determine an optimal decision plan in the t period from the plurality of decision plans. Repeating the steps sequentially until t = 1,
2. The production activity decision support method according to claim 1, wherein: The optimal plan determination step,
In a case where each of the plurality of decision making plans is adopted in the t period, an optimal production activity value index is calculated by constructing an optimal production activity model in each of the plurality of situations. Based on the risk-adjusted expected value of the decision plan determined in the optimal plan decision step, a value change risk index caused by the plurality of situations is calculated, and t An expected value calculating step of calculating a risk-adjusted expected value for the period T;
An optimal plan that compares the risk-adjusted expected values of the t to T periods in the plurality of decision plans and selects a decision plan that produces the largest risk-adjusted expected value as an optimal decision plan in the t-th period; A selection step;
An expected value storing step of storing the maximum risk-adjusted expected value in the storage unit as the risk-adjusted expected value by the optimal production activity in the period from t to T.
The production activity decision support method according to claim 2, wherein: The plurality of decision-making plans are capital investment plans related to capital investment in a power supply business,
The production activity decision support method according to any one of claims 1 to 3, wherein: The uncertainty factors include power demand and resource costs,
5. The production activity decision support method according to claim 4, wherein: The said capital investment plan includes the plan selected from addition of new equipment, renovation of existing equipment, replacement of existing equipment by new equipment, and abolition of suspension equipment, The claim 4 characterized by the above-mentioned, or claim 4 characterized by the above-mentioned. Item 6. The production activity decision support method according to Item 5. The output step includes a step of outputting, as a result of the decision, a graph display of a relationship between an assumed value of an uncertainty factor and a tendency of selecting an optimal capital investment plan in the predetermined period.
The production activity decision support method according to any one of claims 4 to 6, wherein: The output step includes a step of, as a result of the decision, outputting data that displays an optimal production activity value calculated assuming each of the plurality of situations as a value distribution graph,
The production activity decision support method according to any one of claims 4 to 7, wherein: The optimal production activity model is a model for maximizing profits in production activities, with power supply stability and resource usage restrictions as constraints, power generation end power of each facility as a determining variable of the optimal production activity model. Is,
9. The production activity decision support method according to claim 4, wherein: The optimal production activity model is formulated as a linear programming problem with a special structure,
The production activity decision support method according to any one of claims 4 to 9, wherein: The output step includes a step of outputting, as a result of the decision, data that graphically displays an optimal operation amount of each facility derived from the optimal production activity model,
The production activity decision support method according to claim 9, wherein: Using a computer having an input / output unit, a storage unit, and an arithmetic processing unit, a program for supporting a decision to perform a production activity that produces the maximum profit under an uncertain business environment,
The input / output unit inputs data relating to a plurality of decision making plans for performing the production activities, data relating to a plurality of situations assumed for a future uncertainty factor, and stores the data in the storage unit. Features and
The arithmetic processing unit, a decision tree generating function of generating a decision tree by combining the plurality of decision plans and the plurality of situations,
The arithmetic processing unit constructs an optimal production activity model in each of the plurality of situations and calculates an optimal production activity value index for each of the plurality of decision plans from the decision tree. A value fluctuation risk index indicating the magnitude of value fluctuation caused by the plurality of situations is calculated from the value index, and an intention over a predetermined period in the future to perform the production activity based on the optimal production activity value index and the value fluctuation risk index. A decision-making function for making decisions;
An output function of outputting a result of the decision by the input / output unit;
A program that causes the computer to realize the following.

Images