JP3751828B2 - Air conditioning heat source equipment optimum operation control device - Google Patents

Description

ここで、上式（１）乃至（４）により各熱源機器の運転状態Ｘ(k)を求める際には、熱供給条件およびプラント容量等に関して、次のような制約条件が満たされていなければならない。
【００３０】
（制約条件）
Ｖ(k)−η・Ｖ(k-1)−Ｆ・Ｘ(k)＋Ｑ_ｄ(k)＝０ …（５）
Ｘ(k)＝Ｘ(k-1)＋Ｙ(k) …（６）
｜Ｙ(k)｜＜ΔＹ …（７）
Ｐ_ｌ＜Ｐ・Ｘ(k)＜Ｐ_ｕ …（８）
Ｖ_ｌ＜Ｖ(k)＜Ｖ_ｕ …（９）
ここで、
Ｊ_ｅ：消費電力に関するペナルティ関数（なおここでは、エネルギーとしてガス等の燃料を用いない熱源機器（ターボ冷凍機やヒートポンプ等）を用いた場合を想定しているが、ガス等の燃料を用いる熱源機器があるときには、同様にして燃料消費に関するペナルティ関数も追加する。）
Ｊ_ｐ：ピークカット時間帯消費電力に関するペナルティ関数
（ピークカット時間帯；（ｅｘ）夏期；ｋ＝１３〜１５）
Ｊ_ｓ：熱源機器起動停止回数に関するペナルティ関数
Ｗ_ｅ：消費電力に関するペナルティ重み
Ｗ_ｐ：ピークカット時間帯消費電力に関するペナルティ重み
Ｗ_ｓ：熱源機器起動停止回数に関するペナルティ重み
ｋ ：離散時間
Ｌ ：１４（または２４）
Ｘ(k)＝（ｘ_１(k），ｘ_２(k)，…，ｘ_ｎ(k)）；ｘ_ｉ(k)＝０，１
：熱源機器の運転停止変数ベクトル
Ｐ＝（ｐ_１，ｐ_２，…，ｐ_ｎ）：熱源機器の消費電力ベクトル
Ｙ(k)＝（ｙ_１(k），ｙ_２(k)，…，ｙ_ｎ(k)）；ｙ_ｉ(k)＝０，１
：熱源機器の運転停止状態変化ベクトル
ここで、
ｙ_ｉ(k)＝ｘ_ｉ(k)−ｘ_ｉ(k-1) …（１０）
である。
【００３１】
Ｖ(k)＝（ｖ_ｈ(k)，ｖ_ｃ(k)）Ｔ
：蓄熱槽の蓄熱量ベクトルで、ｖ_ｈ(k)は温水蓄熱量、ｖ_ｃ(k)は冷水蓄熱量
Ｖ_ｌ＝（ｖ_ｈｌ，ｖ_ｃｌ）：蓄熱槽の最小蓄熱量ベクトル
Ｖ_ｕ＝（ｖ_ｈｕ，ｖ_ｃｕ）：蓄熱槽の最大蓄熱量ベクトル
Ｐ_ｌ＝（ｐ_１ｌ，ｐ_２ｌ，…，ｐ_ｎｌ）：熱源機器の最小消費電力ベクトル
Ｐ_ｕ＝（ｐ_１ｕ，ｐ_２ｕ，…，ｐ_ｎｕ）：熱源機器の最大消費電力ベクトル
η ：蓄熱係数（η＝１−ε；εは熱損失係数）
Ｆ＝ｆ_１ｈ，ｆ_２ｈ，…，ｆ_ｎｈ
ｆ_１ｃ，ｆ_２ｃ，…，ｆ_ｎｃ
：熱源機器の熱製造容量行列で、ｆ_１ｈ，ｆ_２ｈ，… ，ｆ_ｎｈは温水製造容量、
ｆ_１ｃ，ｆ_２ｃ，… ，ｆ_ｎｃは冷水製造容量
Ｑ_ｄ(k)＝（ｑ_ｄｈ，ｑ_ｄｃ）Ｔ
：温水需要ｑ_ｄｈ、冷水需要ｑ_ｄｃを要素とする総合需要ベクトル
（最適化問題の定式化）
ここで、熱源機器の１次運転計画を求める最適化問題は、上式（１）により表される目的関数を最小化するような各熱源機器の運転状態Ｘ(k)（ｋ＝１，２，…，Ｌ）を求める問題であり、例えば熱源機器が５台（ｎ＝５）、Ｌ＝１４の場合には、７０（＝５×１４）の０−１変数列を求める問題となる。
【００３２】
なお、上式（１）により表される目的関数において、最初の項目Ｊ_ｅは消費電力に関する項目である。Ｌ＝１４時間（８時台〜２１時台）の問題を考えた場合には、Ｊ_ｅを最小化することは、電力コストの安い夜間の消費電力、すなわち夜間の製造熱量を最大化することを意味する。
【００３３】
２番目の項目Ｊ_ｐはピークカット時間帯（電力料金の定時調整契約時間帯）における消費電力に関する項目である。ピークカットがなされた場合には受電契約に応じた月当たりの大幅割引料金が適用され、一方で、ピークカットがなされなかった場合にはペナルティが課せられる。このため、特に最小化を図るべき項目である。
【００３４】
３番目の項目Ｊ_ｓは熱源機器の起動停止回数に関する項目である。Ｊ_ｓを最小化することは空調／熱源プラント２０の安定運転を図ることを意味する。
【００３５】
一方、上式（５）〜（９）の制約条件のうち、最初の式（５）は空調／熱源プラント２０の熱フローに基づく状態を表している。２番目の式（６）は被制御変数である各熱源機器の運転状態Ｘ(k)の時間変化を表している。３番目の式（７）は、１離散時間でのＸ(k)の変化に制約をかけるためのものである。なお、最後の２式（８）（９）は時断面での消費電力および蓄熱量の制約を表している。
【００３６】
なお、目的関数（上式（１）〜（４））および制約条件（上式（５）〜（９））におけるＰ・Ｘ(k)の要素は、次式（１１）のように表すことができる。
【００３７】
【数２】

また、上式（１１）は、次式（１２）のように表すことができる。
【００３８】
Ｐ_ｉ(k)＝ａ_ｉ・Ｓ_ｉ(k)＋ｂ_ｉ …（１２）
ここで、
Ｓ_ｉ(k) ：第ｉ番目の熱源機器の熱量負荷
である。
【００３９】
このように、熱源機器における熱量（冷水熱量および温水熱量）負荷入力とエネルギー消費出力（例えば消費電力や燃料使用量等）との関係は、一次式や折れ線等の簡易近似式により表すことができる。なおここでは、熱量（冷水熱量および温水熱量）負荷入力とエネルギー消費出力（例えば消費電力や燃料使用量等）との関係を１本の直線で近似したが、２本以上の折れ線で近似することも可能である。また、上式（１２）の係数は、簡易近似式の精度を向上させるために、あらかじめ空調／熱源プラントモデルを用いてオフライン・シミュレーションを行い、それにより得られた熱源機器の出力特性に基づいてその出力特性を良く近似するａ_ｉ、ｂ_ｉ（折れ線の場合はその各係数）を算出するようにしてもよい。
【００４０】
以上のようにして定式化された最適化問題（整数計画問題）は、熱源運転１次計画部１３により、遺伝的アルゴリズム等の最適化手法を用いて解かれ、これにより熱源機器の１次運転計画が求められる。
【００４１】
なお、整数計画問題では、離散値（０，１）をとる離散変数を含んでおり、離散変数の数に依存して解くべき問題が非常に大きくなる。問題の規模は、熱源機器が５台（ｎ＝５）で、１時間ごとにＬ＝１４時間（８時台〜２１時台）分の運転計画を例えば全列挙法を用いて最適化する場合を例に挙げると、１時間ごとに２の５乗の運転状態（運転／停止）の組み合わせがあることから、１４時間分の運転計画を得るためには３２（＝２^５）の１４乗の演算および評価が必要となる。実際の空調／熱源プラント２０では１０台規模の熱源機器を有するものもあり、また数台規模の熱源機器を有する場合でも１台ごとに複雑なモード（冷水および温水別の段階運転等）を有するものもある。さらに、３０分や１５分といった単位の運転計画を得たい場合もある。このため、解くべき問題が大きくなることは避けられない。
【００４２】
熱源機器の運転計画をオンラインでかつ実時間で求めるためには、このような整数計画問題を高速で解く必要があるが、これに適しているのが遺伝的アルゴリズムである。
【００４３】
図２は、遺伝的アルゴリズムの処理手順を説明するためのフローチャートである。なお、遺伝的アルゴリズムとは、生物の進化の過程を遺伝子という情報媒体の切り口から模擬したアルゴリズムであり、最適化問題の分野で広く研究が行われている。遺伝的アルゴリズムでは、各種の遺伝子列を持つ個体に対して、制約条件と評価関数とによる淘汰のプロセスを複数回（複数世代）にわたって施しながら、交叉、突然変異および増殖の処理を行うことにより、適応度が良い解（個体）を求めるものである。なお、適応度は、上式（１）のように表される目的関数の逆数として定義することができる。
【００４４】
なおここでは、１時間単位ごとの運転状態（運転／停止）を１／０で表す遺伝子列を各熱源機器ごとに持つ個体に対して遺伝的アルゴリズムを適用する。図３はこのような遺伝子列を持つ個体の一例を示す図であり、熱源機器として５台のヒートポンプ（ＨＰ）が用いられる場合を示している。この場合、１個体の遺伝子は、７０ビット（＝５×１４）で表現される。
【００４５】
図２に示すように、まず、ランダムな遺伝子列を持つｎ個の個体を発生させ、これらを初期個体群とする（ステップ１０１）。
【００４６】
次に、ステップ１０１で生成されたｎ個の個体のうちで上式（５）乃至（９）の制約条件を満たさない個体について上式（５）乃至（９）を満たさない遺伝子列を選択するとともに、この選択された遺伝子列に含まれる制約条件を満たさない遺伝子の状態（１／０）を変更する。そして、制約条件を満たす個体がｎ個生成できた時点で各個体の適応度、およびその世代での適応度の平均値を計算する（ステップ１０２）。
【００４７】
次に、個体群中の最大の適応度に対して所定の比率よりも小さい適応度（目的関数の値が大きいもの）を有する個体を淘汰し、またこの時点で制約条件を満たさない個体が存在した場合も淘汰する（ステップ１０３）。
【００４８】
次に、ステップ１０３で淘汰された個体の数だけ適応度が最大の個体を増殖させる（ステップ１０４）。
【００４９】
次に、図４（ａ）（ｂ）に示すように、個体同士を所定の交叉確率分（全個体数に対する割合）だけランダムにペアリングし、このようにしてペアリングした組ごとにランダムに遺伝子座を選び、各個体の同一の遺伝子座同士を互いに一点交叉させる（ステップ１０５）。なお一点交叉とは、連続した一連の遺伝子（図４（ｂ）の場合には連続した２つの遺伝子）同士を互いに交叉させるこという。
【００５０】
次に、図４（ｃ）に示すように、所定の突然変異率分（全個体数に対する割合）だけランダムに個体を選び、選ばれた各個体の任意の遺伝子座の遺伝子をビット反転させる（ステップ１０６）。
【００５１】
このようにしてステップ１０１乃至１０６の処理が終了した時点で、その世代における適応度の平均値が前回および前々回の適応度の平均値と比較して所定値以下となっているか、またはあらかじめ定められた繰返し回数を越えているかを判定し（ステップ１０７）、これらの終了条件に該当するまでステップ１０１乃至１０６の処理を繰り返す。なお、上述した終了条件に該当した場合には全体の処理を終了し、その世代の最大の適応度の個体に基づいて運用計画を求める。
【００５２】
[２次運転計画の作成]
次に、熱源運転２次計画部１５は、熱負荷予測部１１により求められた熱負荷予測値と空調／熱源プラントモデル部１４に保持された空調／熱源プラントモデルとに基づいて、遺伝的アルゴリズム等を用いた最適化手法により、熱源運転１次計画部１３により求められた熱源機器の１次運転計画を初期値としてこの近傍を探索し、熱源機器の最適な運転計画を求める。
【００５３】
空調／熱源プラントモデル
ここで、空調／熱源プラントモデル部１４に保持された空調／熱源プラントモデルは、空調／熱源プラント２０に関する各種の物理モデルを組み合わせてなる空調／熱源プラントモデルを保持するものである。なお、空調／熱源プラントモデルは、(1)室内伝熱モデル（部屋や壁等の伝熱モデル）、(2)空調機の冷却コイル、加熱コイルおよび加湿器等の熱交換モデル、(3)空気線図の計算式、(4)蓄熱槽の温度分布モデル、および(5)熱源機器（吸収冷凍機、ターボ冷凍機およびヒートポンプ等）のモデル等の各種の物理モデルを組み合わせてなるものである。
【００５４】
（空調プロセス）
まず、図５により、空調／熱源プラントモデルによる空調プロセスのシミュレーションの手順について説明する。
【００５５】
まず、空調対象となる部屋の大きさ、熱容量、壁面積および壁の熱貫流率等の建物データを設定する（ステップ２０１）。また、空調機の熱負荷（冷暖房負荷）に関連する気象データとして外気温、外気湿度および日射量等を設定する（ステップ２０２）。さらに、人の発熱量や、電気機器（ＯＡ機器等）および照明等による発熱量を、人の在室率の変化に応じてシミュレーションの時間分だけあらかじめ計算しておき、室内熱負荷データとして設定する（ステップ２０３）。また、日射量に基づいて窓ガラス面を介して取得される熱量もあらかじめ計算して設定する。
【００５６】
ここで、この空調プロセスでは、制御で用いるＬＯＣ（ローカル・オブジェクト・コントローラ）の動作も模擬するので、サンプリング制御周期（シミュレータ計算周期の整数倍）の間隔でステップ２０５のＬＯＣ制御アルゴリズムを行わせる（ステップ２０４）。なお、ステップ２０５においては、主に室温設定値と室温との偏差、および湿度設定値と室内湿度との偏差に基づいて、ＰＩＤアルゴリズム（速度形）の演算を行う。
【００５７】
その後、空調機内での空気の状態点（温度、湿度およびエンタルピ等）を、冷却コイル、加熱コイルおよび加湿器の順で、熱交換器の理論計算式（熱交換モデル）および空気線図の計算式等を用いて求める（ステップ２０６）。また、後述する室内伝熱モデルを用いて、室内の空気の状態点を求める（ステップ２０７）。さらに、再循環空気と導入外気との混合計算により空調機の入口での空気の状態点を求める（ステップ２０８）。
【００５８】
なお、シミュレーション終了時刻に至るまで、上述したステップ２０４乃至２０８の処理を繰り返す（ステップ２０９および２１０）。
【００５９】
（室内伝熱モデル）
図６は空調／熱源プラントモデルにおける室内伝熱モデルを説明するための図である。
【００６０】
図６に示すように、部屋の壁を外壁（窓有り）５１、内壁５２、床５３および天井５４に分け、その壁の構造はコンクリートおよび断熱材の２つの層で近似する。この場合の状態方程式を次式（１３）〜（１５）に示す。
【００６１】
【数３】

ここで、ｉ＝１，２，３，４である。
【００６２】
また、
Ｃ_ＲＭ＝Ｃ_ａ・ρ・ＶＯＬ＋ＣＰＦ
Ｃ_ａ ：空気の比熱（Ｊ／ｋｇ・Ｋ）
ρ ：空気の密度（ｋｇ／ｍ^３）
ＶＯＬ：室容積（ｍ^３）
ＣＰＦ：家具の熱容量（Ｊ／Ｋ）
ｄｔ：計算周期（ｓ）
Ｋ_ＲＭｉ＝Ａ_ｉ・α_Ｒｉ
Ａ_ｉ ：壁面積（ｍ^２）
α_Ｒｉ：室内表面熱伝達率（Ｗ／ｍ^２・Ｋ）
Ｃ_Ａｉ，Ｃ_Ｂｉ：部位ｉの壁熱容量（Ｊ／Ｋ）
Ｋ_Ａｉ，Ｋ_Ｂｉ：部位ｉの熱貫流率×壁面積（Ｗ／Ｋ）
Ｔ_Ａｉ，Ｔ_Ｂｉ：部位ｉでの壁温（℃）
Ｔ_Ｒ ：室温（℃）
Ｔ_１ ：（相当）外気温（℃）
Ｔ_２ ：床下の温度（℃）
Ｔ_３ ：隣室の温度（℃）
Ｔ_４ ：天井の温度（℃）
Ｇ_Ｏ ：すきま風量（ｋｇ／ｓ）
Ｑ_Ｈ ：人の発熱量（Ｗ）
Ｑ_Ｅ ：電気機器の発熱量（Ｗ）
Ｑ_Ｌ ：照明の発熱量（Ｗ）
Ｑ_Ｃ ：窓ガラス面からの取得熱量（Ｗ）
Ｈ_ＥＳ：室からの除去熱量（＝Ｃ_ａ・Ｇ_Ｄ・（Ｔ_Ｒ−Ｔ_ＲＤ））（Ｗ）
Ｇ_Ｄ ：給気量（ｋｇ／ｓ）
Ｔ_ＲＤ：給気温度（℃）
である。
【００６３】
なお、室内湿度についても、同様にして、水分の物質平衡式を解くことにより求めることができる。
【００６４】
（蓄熱槽の温度分布モデル）
次に、空調／熱源プラントモデルにおける蓄熱槽の温度分布モデルについて説明する。
【００６５】
図７は空調／熱源プラントモデルを用いて模擬される熱源プラントを模式的に示す図である。
【００６６】
図７に示すように、熱源プラントとして、２台の熱源機器６１，６２を有するよこ型蓄熱槽６３による冷（温）水蓄熱システムを想定する。なお、符号６４は熱負荷機器を示している。ここで、蓄熱槽６３は、ｎ個の小区画に分割され、それぞれに１〜ｎの番号が付されている。これらの小区画は、もぐりぜき等により物理的に分けられたものでもよいし、仮想的に分割されたものでもよい。ただし、分割された各小区画は完全混合蓄熱槽となり、槽内の水温は一定になるものとする。
【００６７】
図８は空調／熱源プラントモデルにおける蓄熱槽の温度分布モデルを説明するための図であり、小区画ｉ近傍の蓄熱槽を示している。
【００６８】
なお、
ｉ ：小区画番号（ｉ＝１，…，ｎ）
ｖ ：蓄熱槽（小区画）容積［ｍ^３］
θ ：蓄熱槽（小区画内）温度［℃］
ｕｌ：（小区画内）流量［ｍ^３／ｓ］
ｔ ：時刻[ｓ]
ここで、θin、ｕｌin、θoutおよびｕｌoutは、小区画によっては複数個になる場合もある。
【００６９】
図８に示すような蓄熱槽における温度分布は次のようにして求められる。
【００７０】
（ａ）流量
小区画ｉにおける流入量および流出量により、時刻ｔでの蓄熱槽内の水の流れｕｌ（ｔ，ｉ）を次式（１６）により求める。
【００７１】
【数４】

（ｂ）水温
求められた小区画ｉ内の流量ｕｌ（ｔ，ｉ）および容積ｖ（ｉ）により、Δｔ時間［ｓｅｃ］後の小区画ｉの水温θ（ｔ＋Δｔ，ｉ）を次式（１７）により求める。
【００７２】
【数５】

ここで、ηは蓄熱効率である。
【００７３】
（熱源機器の物理モデル）
次に、空調／熱源プラントにおける熱源機器の物理モデルについて説明する。なおここでは、熱源機器として吸収冷凍機を例に挙げて説明する。
【００７４】
図９は空調／熱源プラントモデルを用いて模擬される熱源機器（吸収冷凍機）を模式的に示す図である。なお、吸収冷凍機は可動部分の少ない安定した冷凍機であり、通常、吸収剤としてＬｉＢｒ（臭化リチウム）水溶液が用いられ、冷媒として水が用いられる。吸収冷凍機は熱系の動特性に支配されるので、自己安定性に優れている。
【００７５】
図９に示すように、吸収冷凍機は、蒸発器７１、吸収器７２、発生器７３および凝縮器７４の４つの容器を有し、これらの容器内を冷媒である水が循環することにより冷凍サイクルを構成している。
【００７６】
具体的には、まず、蒸発器７１に収容された水はチューブ内の冷水から気化熱を奪って水蒸気となり、吸収器７２のＬｉＢｒ水溶液にとけ込み、ＬｉＢｒ水溶液を希釈する。吸収器７２において希釈された希薄ＬｉＢｒ水溶液は発生器７３に送られ、発生器７３で加熱されて再度水蒸気とＬｉＢｒ水溶液とに分離される。このうち、水蒸気は凝縮器７４で冷却されて凝縮して水となり、出発点の蒸発器７１に戻る。一方、凝縮された濃縮ＬｉＢｒ水溶液は再度吸収器７２に送られ、吸収剤として用いられる。
【００７７】
この冷凍サイクルを横軸を溶液および水の温度、縦軸を蒸気圧にとってＤｕｈｒｉｎｇ線図（「ｈ−ζ線図」ともいう）上に表したものが図１０である。図１０には、４個の容器（蒸発器、吸収器、発生器および凝縮器）に関して反時計回りに回転する水のサイクルと、吸収器と発生器との間を往復するＬｉＢｒ水溶液のサイクルとが存在する。蒸発器内の水は低温（６℃程度）でかつ低蒸気圧（６ｍｍＨｇ＝８００Ｐａ）である。この蒸発器内の水と吸収器内のＬｉＢｒ水溶液（４０℃）とは同じ水蒸気圧を持つ（実際には、吸収器内の蒸気圧の方が蒸発器内の蒸気圧よりもほんのわずかに低い）。同様に、発生器内のＬｉＢｒ水溶液（１５０℃）と凝縮器内の水も同じ水蒸気圧（７０ｍｍＨｇ＝９３００Ｐａ）を持つ（この場合も、凝縮器内の蒸気圧の方が発生器内の蒸気圧よりもほんのわずかに低い）。
【００７８】
ここで、このような冷凍サイクルを実現するために、それぞれの容器（蒸発器、吸収器、発生器および凝縮器）を以下のようにモデル化する。
【００７９】
ここで、記号については、
Ｔ ：温度（℃）
Ｃ ：比熱（ｋＪ／ｋｇ℃）
Ｃｏｎｃ：濃度（重量比）
Ｗ ：重量（ｋｇ）
ｒ ：蒸発潜熱（ｋＪ／ｋｇ）
ｑ_ｗ ：蒸発潜熱（水）（ｋＪ／ｋｍｏｌ）
Ｇ ：流量（ｋｇ／ｓｅｃ）
α ：熱伝達率（ｋＪ／ｈ℃）
ｈ ：エンタルピー（ｋＪ）
である。
【００８０】
また、各記号に付された添え字については、
‘ａ’ ：吸収器
‘ｒ’ ：発生器
‘ｃ’ ：凝縮器
‘ｖ’ ：蒸発器
‘ＬＢ’ ：ＬｉＢｒ
‘ｗ’ ：水
‘ｖｐｒ’：蒸気
である。
【００８１】
さらに、その他の変数については、
Ｔ_ｃｌｔ ：冷水平均温度（℃）
Ｇ_ｒ ：発生器蒸発水量（ｋｇ／ｓｅｃ）
Ｇ_ａ ：吸収器吸収蒸気量（ｋｇ／ｓｅｃ）
Ｔ_ｖｐｒ ：加熱蒸気温度（℃）
である。
【００８２】
以下、それぞれの容器（蒸発器、吸収器、発生器および凝縮器）ごとにそのモデルを説明する。
【００８３】
（１）吸収器
吸収器７２には蒸発器７１で発生した水蒸気を吸収する吸収剤としての濃縮ＬｉＢｒ水溶液が常に発生器７３から供給されている。通常の吸収器７２でのＬｉＢｒ水溶液の濃度は約６０％、温度は４０℃程度である。その水蒸気圧は約６℃の水蒸気と平衡する。この性質により、蒸発器７１の水が６℃に下がるまで吸収器７２内のＬｉＢｒ水溶液は水蒸気を吸収し続ける。一方、水蒸気を吸収することによってＬｉＢｒ水溶液に水蒸気の凝縮熱が発生し、温度が上昇する。温度が上昇すると水蒸気圧が上昇して冷却能力が低下する。それを防ぐために冷却水によりＬｉＢｒ水溶液を常に冷却する。また、水蒸気を吸収すると、ＬｉＢｒ水溶液は濃度が低下する。これも水蒸気圧の上昇と冷却能力の低下とをもたらす。そこで、ＬｉＢｒ水溶液を再度濃縮するために、発生器７３へ希薄ＬｉＢｒ水溶液が連続的に送られる。
【００８４】
（熱収支）
【数６】

（水収支）
【数７】

（ＬｉＢｒ収支）
【数８】

（凝縮熱）
【数９】

ここで、μ_ｗは水の分子量、ΔＴ_ａは吸収器７２の温度が１度上昇したときのＬｉＢｒ水溶液の蒸気圧上昇に対応する水の温度上昇率を表し、Ｄｕｈｒｉｎｇ線図から求める。
【００８５】
（２）発生器
発生器７３には吸収器７２から希薄ＬｉＢｒ水溶液が送られてくる。発生器７３ではこの希薄ＬｉＢｒ水溶液を加熱して水分を蒸発させて濃縮する。加熱源としては、高温の水蒸気を用いる。濃縮ＬｉＢｒ水溶液は吸収器７２に送られる。発生した水蒸気は凝縮器７４へと送られる。
【００８６】
（熱収支）
【数１０】

（水収支）
【数１１】

（ＬｉＢｒ収支）
【数１２】

（蒸発熱）
【数１３】

ここで、μ_ｗは水の分子量、ΔＴ_ｒは発生器７３の温度が１度上昇したときのＬｉＢｒ水溶液の蒸気圧上昇に対応する水の温度上昇率を表し、Ｄｕｈｒｉｎｇ線図から求める。
【００８７】
（３）凝縮器
凝縮器７４は、発生器７３に比べて低温に保たれ、発生器７３で蒸発した水蒸気が凝縮する。低温に保つために、冷却水により常に冷却されている。凝縮された水は再度蒸発器７１へと送られる。
【００８８】
（熱収支）
【数１４】

（水の収支）
【数１５】

（４）蒸発器
蒸発気７１では冷水がチューブ内を流れている。上部のノズルから水が噴霧され、この水がチューブの表面から蒸発することにより、気化熱をチューブから奪う。これにより、冷水を冷却する。
【００８９】
（熱収支）
【数１６】

（水の収支）
【数１７】

なお、制御器において、冷水供給温度を設定値にするようにＰＩ演算が行われ、加熱蒸気の流量が制御されている。
【００９０】
【数１８】

ここで、Ｋ_ｐ：比例ゲイン、Ｔ_ｉ：積分時間
である。
【００９１】
ここで、熱源運転２次計画部１５は、上述したような空調／熱源プラントモデルと熱負荷予測部１１により求められた熱負荷予測値とに基づいて、熱源運転１次計画部１３で用いられるのと同様の遺伝的アルゴリズム等を用いた最適化手法により、熱源運転１次計画部１３により求められた熱源機器の１次運転計画を初期値としてこの近傍を探索し、熱源機器の最適な運転計画を求める。なお、熱源運転２次計画部１５は、熱源運転１次計画部１３により求められた熱源機器の１次運転計画の近傍の運転計画に基づいて、空調／熱源プラントモデル部に保持された空調／熱源プラントモデルを用いてシミュレーションを行い、電力およびガス等の使用量、冷水および温水の製造熱量、消費熱量および温度変化等のプロセス値を精度良く計算して出力する。これにより、この出力値に基づいて目的関数の計算および制約条件のチェックが正確に行われる。
【００９２】
なおこのとき、熱源運転１次計画部１３により求められた熱源機器の１次運転計画を熱源運転２次計画部１５に対して初期値として渡す前に、空調／熱源プラントモデルを用いてシミュレーションを行い、制約条件を満足して実際に運転が可能か否かをチェックする。そして、運転計画が制約条件を満足していない場合は、１個前の運転計画、２個前の運転計画というように、制約条件を満足する運転計画があるまでシミュレーションを行い、制約条件を満足する運転計画のみを熱源運転２次計画部１５に対して初期値として渡す。
【００９３】
最後に、プラント制御部１６は、熱源運転２次計画部１５により求められた熱源機器の最適な運転計画をプロセス入出力部１７を介して空調／熱源プラント２０に送り、空調／熱源プラント２０において、図１１に示す空調／熱源プラント２０の冷凍機およびヒートポンプ（ＨＰ）等の熱源機器２２，２３，２４を制御する。
【００９４】
このように本実施の形態によれば、ビル等の空調／熱源プラント２０において、熱源機器２２，２３，２４の運転計画を最適化手法により作成する際に、目的関数および制約条件を実際のプロセスを正確に表す空調／熱源プラントモデルを用いてチェックしているので、省エネルギー、低コスト運転およびピークカット等の観点からみて適正でかつ高精度の運転計画を自動的に作成して、熱源機器２２，２３，２４の効率的な運転を実現することができる。
【００９５】
なお、上述した実施の形態においては、熱源運転１次計画部１３および熱源運転２次計画部１５において、最適化手法として遺伝的アルゴリズムを用いているが、これに限らず、熱源運転１次計画部１３および熱源運転２次計画部１５の少なくとも一方において、最適化手法として混合整数計画法を用いるようにしてもよい。ここで、上述したようにして定式化された０−１整数計画問題は一般的に分岐限定法で解くことができる（文献（伊里、今野、刀根監訳：『最適化ハンドブック』、第VI章、朝倉書店（１９９５））参照）。このため、混合整数計画法においては、分岐限定法等により定義された小問題を線形計画法で解き、これを全ての小問題に関して繰り返すことにより全体の問題を解くことができる。
【００９６】
【発明の効果】
以上説明したように本発明によれば、省エネルギー、低コスト運転およびピークカット等の観点からみて適正でかつ高精度の運転計画を自動的に作成して、熱源機器の効率的な運転を実現することができる。
【図面の簡単な説明】
【図１】本発明による空調熱源機器最適運転制御装置の一実施の形態を示す図。
【図２】遺伝的アルゴリズムの処理手順を説明するためのフローチャート。
【図３】遺伝子列を持つ個体の一例を示す図。
【図４】遺伝的アルゴリズムによる遺伝子列に対する操作を模式的に示す図。
【図５】空調／熱源プラントモデルによる空調プロセスのシミュレーションの手順を説明するためのフローチャート。
【図６】空調／熱源プラントモデルにおける室内伝熱モデルを説明するための図。
【図７】空調／熱源プラントモデルを用いて模擬される熱源プラントを模式的に示す図。
【図８】空調／熱源プラントモデルにおける蓄熱槽の温度分布モデルを説明するための図。
【図９】空調／熱源プラントモデルを用いて模擬される熱源機器（吸収冷凍機）を模式的に示す図。
【図１０】図９に示す熱源機器（吸収冷凍機）の動作をＤｕｈｒｉｎｇ線図上に表した図。
【図１１】本発明による空調熱源機器最適運転制御装置が適用される空調／熱源プラントの一例を示す図である。
【図１２】図１１に示す空調／熱源プラントの空調設備の詳細を示す図である。
【符号の説明】
１１ 熱負荷予測部
１２ プラント簡易モデル部
１３ 熱源運転１次計画部
１４ 空調／熱源プラントモデル部
１５ 熱源運転２次計画部
１６ プラント制御部
１７ プロセス入出力部
２０ 空調／熱源プラント[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an air conditioning / heat source plant such as a building, and more particularly to an air conditioning heat source equipment optimum operation control apparatus for efficiently operating a heat storage tank and a heat source equipment in an air conditioning / heat source plant such as a building.
[0002]
[Prior art]
Conventionally, an air conditioning / heat source plant including a heat load device such as an air conditioner has been used as an air conditioning plant control system for buildings and the like. In such an air conditioning / heat source plant, cold water and hot water are produced by a heat source device such as a refrigerator or a heat pump, and these are stored in a heat storage tank and supplied to a heat load device such as an air conditioner as necessary. . Here, in an air-conditioning / heat source plant in which an electric system is used as a heat source device, heat is generated at night and stored in a heat storage tank, and the heat stored in the heat storage tank is radiated for air conditioning during the day. Therefore, efficient operation is performed by using cheap nighttime electric power or performing peak cuts that shift the peak consumption of energy during the daytime to other time zones.
[0003]
By the way, in order to perform the efficient operation as described above in the air conditioning / heat source plant, a specialized operator is necessary. However, it is difficult to secure such a specialized operator, and the operator has a lot of personnel. Since it is expensive, it is desired to perform automated operation using a computer or the like.
[0004]
Conventionally, as a method for performing automated operation, an optimization method that minimizes the sum of objective functions related to power charges, gas charges, etc. while keeping the amount of heat demanded is used. In general, a method for obtaining an operation plan of a plurality of heat source devices using a simplified plant model in which is formulated.
[0005]
[Problems to be solved by the invention]
As described above, conventionally, an automatic operation of an air conditioning / heat source plant is performed by obtaining an operation plan of a plurality of heat source devices using a simple plant model using a simple approximate expression. However, in the conventional method described above, since a simple approximate expression such as a linear expression or a first-order lag expression is used in the simple plant model, it is difficult to ensure the accuracy of the operation plan obtained by the optimization method. There is a problem.
[0006]
The present invention has been made in consideration of these points, and in an air conditioning / heat source plant such as a building, an appropriate and highly accurate operation plan is automatically obtained from the viewpoint of energy saving, low cost operation, peak cut, and the like. It is an object of the present invention to provide an air-conditioning heat source equipment optimum operation control device that can be created in order to realize efficient operation of the heat source equipment.
[0007]
[Means for Solving the Problems]
The present invention relates to an air conditioning heat source equipment optimum operation control apparatus for controlling the operation of an air conditioning / heat source plant such as a building, and the heat load prediction for obtaining a heat load prediction value based on weather data and the actual heat load value of the air conditioning / heat source plant. Means, a plant simple model means for holding a plant simple model that represents the relationship between the heat load input and the energy consumption output by a simple approximate expression for each heat source device constituting the air conditioning / heat source plant, and the heat load prediction means Heat source operation primary planning means for obtaining a primary operation plan of a heat source device by an optimization method based on the obtained heat load predicted value and the plant simple model held in the plant simple model means, and air conditioning / heat source An air-conditioning / heat-source plant model means for holding an air-conditioning / heat-source plant model in which various physical models related to the plant are combined; The heat source obtained by the heat source operation primary planning means by an optimization method based on the predicted heat load obtained by the load predicting means and the air conditioning / heat source plant model held in the air conditioning / heat source plant model means. Provided is an air conditioning heat source equipment optimum operation control device characterized by comprising a heat source operation secondary planning means for searching for the vicinity of the equipment as the initial value of the equipment primary operation plan and obtaining an optimum operation plan of the heat source equipment. .
[0008]
According to the present invention, in an air conditioning / heat source plant such as a building, when an operation plan of a heat source device is created by an optimization method, an air conditioning / heat source plant model that accurately represents an actual process with objective functions and constraints is used. Therefore, it is possible to automatically create an appropriate and highly accurate operation plan from the viewpoints of energy saving, low cost operation, peak cut, etc., and realize efficient operation of the heat source equipment.
[0009]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 to FIG. 12 are diagrams for explaining an embodiment of an optimum operation control apparatus for air-conditioning heat source equipment according to the present invention.
[0010]
First, an air conditioning / heat source plant such as a building controlled by the air conditioning heat source equipment optimum operation control apparatus according to the present invention will be described with reference to FIGS.
[0011]
FIG. 11 is a diagram illustrating an example of an air conditioning / heat source plant. As shown in FIG. 11, the air conditioning / heat source plant 20 stores heat source devices 22, 23, 24 such as refrigerators and heat pumps (HP), and cold water and hot water produced by these heat source devices 22, 23, 24. The heat storage tank 21 for cooling and the cooling tower 25 are provided, and cold water, hot water, and steam can be supplied to the air conditioning equipment 30.
[0012]
FIG. 12 is a diagram showing details of the air conditioning equipment 30 of the air conditioning / heat source plant 20 shown in FIG. As shown in FIG. 12, the air conditioning facility 30 includes an air conditioner 31 as a heat load device, and the outside air sent to the air conditioner 31 is adjusted in temperature, humidity, and the like by the air conditioner 31. Then, the air is sent into the room 33 through the blower 32. Further, the air discharged from the room 33 is exhausted through the return air fan 38. Note that piping 37 is connected between the outside and the air conditioner 31, between the air conditioner 31 and the room 33, and between the room 33 and the outside. Further, a damper 39 is provided in the vicinity of the outside entrance and the exit of the pipe 37.
[0013]
Here, the air conditioner 31 is provided with a cooling coil (CC), a heating coil (HC), a humidifier (HF), and the like, and a heat storage tank 21 and heat source devices 22, 23, 24 as shown in FIG. Cold water, hot water, steam, and the like are supplied through the above. Note that the air conditioning equipment 30 is provided with a controller (DDC: Direct Digital Controller) 36, and chilled water to the air conditioner 31 based on the measured values by the thermometer 34 and the hygrometer 35 installed in the room 33, The supply amount of hot water and steam can be adjusted.
[0014]
Next, an air conditioning heat source equipment optimum operation control apparatus for controlling the operation of the air conditioning / heat source plant 20 shown in FIGS. 11 and 12 will be described with reference to FIG.
[0015]
As shown in FIG. 1, the air conditioning heat source equipment optimum operation control device controls the operation of the air conditioning / heat source plant 20 via a process input / output unit 17. The process input / output unit 17 takes in a measurement signal from the air conditioning / heat source plant 20 and outputs a control signal to the air conditioning / heat source plant 20.
[0016]
The air conditioning heat source equipment optimum operation control device includes a heat load prediction unit 11, a plant simple model unit 12, a heat source operation primary planning unit 13, an air conditioning / heat source plant model unit 14, a heat source operation secondary planning unit 15, and a plant control unit 16. I have.
[0017]
Among these, the thermal load prediction unit 11 is based on weather data such as weather forecast values and maximum / minimum outside air temperature forecast values, and actual heat load values of the air conditioning / heat source plant 20, and a predicted heat load (cooling / heating load predicted value). Is what you want. The weather data used in the heat load prediction unit 11 can be manually input at a predetermined time on the previous day. However, an external system can be used by using an online service such as the Japan Weather Association provided on the Internet or the like. It is also possible to obtain automatically from the communication means.
[0018]
The plant simple model unit 12 includes a heat input (cold water heat amount and hot water heat amount) load input and energy consumption output (for example, for each of the heat source devices 22, 23, 24 such as a refrigerator and a heat pump (HP) constituting the air conditioning / heat source plant 20. A simple plant model in which the relationship with the power consumption, fuel consumption, etc.) is expressed by a simple approximate expression such as a linear expression or a broken line is held.
[0019]
The heat source operation primary planning unit 13 generates genetic algorithms (GA) based on the predicted heat load obtained by the heat load prediction unit 11 and the plant simple model stored in the plant simple model unit 12. The primary operation plan of the heat source device is obtained by the optimization method using the above.
[0020]
The air conditioning / heat source plant model unit 14 holds an air conditioning / heat source plant model formed by combining various physical models related to the air conditioning / heat source plant 20. The air conditioning / heat source plant model is (1) indoor heat transfer model (heat transfer model for rooms and walls), (2) heat exchange model such as cooling coil, heating coil and humidifier for air conditioner, (3) It is a combination of various physical models such as airline calculation formulas, (4) temperature distribution model of heat storage tank, and (5) models of heat source equipment (absorption refrigerator, turbo refrigerator, heat pump, etc.) .
[0021]
The heat source operation secondary planning unit 15 uses a genetic algorithm or the like based on the predicted heat load obtained by the heat load predicting unit 11 and the air conditioning / heat source plant model held in the air conditioning / heat source plant model unit 14. In this optimization method, the vicinity of the heat source device is searched for using the primary operation plan of the heat source device obtained by the heat source operation primary planning unit 13 as an initial value, and the optimum operation plan of the heat source device is obtained.
[0022]
Before passing the primary operation plan of the heat source device obtained by the heat source operation primary planning unit 13 to the heat source operation secondary planning unit 15 as an initial value, a simulation is performed using the air conditioning / heat source plant model, It is checked whether or not the operation is actually possible while satisfying the constraints. If the operation plan does not satisfy the constraint conditions, simulation is performed until there is an operation plan that satisfies the constraint conditions, such as the previous operation plan and the previous operation plan, and the constraint conditions are satisfied. Only the operation plan to be performed may be passed to the heat source operation secondary planning unit 15 as an initial value.
[0023]
Based on the optimum operation plan of the heat source device obtained by the heat source operation secondary planning unit 15, the plant control unit 16 performs the refrigerating machine and the air conditioner / heat source plant 20 shown in FIG. 11 via the process input / output unit 17. The heat source devices 22, 23, 24 such as a heat pump (HP) are controlled.
[0024]
Next, the operation of the present embodiment having such a configuration will be described.
[0025]
First, the heat load prediction unit 11 calculates a heat load prediction value (cooling / heating load prediction value) based on weather data such as weather forecast values and maximum / minimum outside air temperature forecast values, and actual heat load values of the air conditioning / heat source plant 20. Ask.
[0026]
[Create primary operation plan]
Next, a genetic algorithm or the like was used based on the predicted heat load obtained by the heat load prediction unit 11 and the plant simple model held in the plant simple model unit 12 by the heat source operation primary planning unit 13. The primary operation plan of the heat source equipment is obtained by the optimization method.
[0027]
Simple plant model
Here, X (k) represents an operation state (operation / stop) at a discrete time k of the heat source devices 22, 23, and 24 such as a refrigerator and a heat pump (HP) constituting the air conditioning / heat source plant 20, and the heat load prediction unit 11 Q is the predicted heat load calculated by_dAssuming (k), the problem of obtaining the primary operation plan of the heat source equipment is the optimization problem of obtaining the operating state X (k) of each heat source equipment that minimizes the objective function expressed by the following equation (1). Formulated as (integer programming problem).
[0028]
(Objective function)
min J = J_e+ J_p+ J_s                    ... (1)
In the above equation (1), J_e, J_p, J_sAre represented by the following equations (2), (3), and (4), respectively.
[0029]
[Expression 1]

Here, when the operating state X (k) of each heat source device is obtained by the above formulas (1) to (4), the following constraints regarding the heat supply conditions and the plant capacity are not satisfied. Don't be.
[0030]
(Restrictions)
V (k) -η · V (k-1) -F · X (k) + Q_d(k) = 0 ... (5)
X (k) = X (k-1) + Y (k) (6)
| Y (k) | <ΔY (7)
P_l<P ・ X (k) <P_u                          ... (8)
V_l<V (k) <V_u                              ... (9)
here,
J_e: Penalty function for power consumption (Note that here we assume the use of heat source equipment (turbo refrigerator, heat pump, etc.) that does not use fuel such as gas as energy. (In some cases, a penalty function for fuel consumption is added in the same way.)
J_p: Penalty function for peak cut time power consumption
(Peak cut time zone; (ex) summer season; k = 13 to 15)
J_s: Penalty function related to the number of start and stop of heat source equipment
W_e: Penalty weight for power consumption
W_p: Penalty weight for peak cut time power consumption
W_s: Penalty weight on the number of start and stop of heat source equipment
k: discrete time
L: 14 (or 24)
X (k) = (x₁(k), x₂(k), ..., x_n(k)); x_i(k) = 0, 1
: Stop variable vector for heat source equipment
P = (p₁, P₂, ..., p_n): Power consumption vector of heat source equipment
Y (k) = (y₁(k), y₂(k), ..., y_n(k)); y_i(k) = 0, 1
: Operation stop state change vector of heat source equipment
here,
y_i(k) = x_i(k) −x_i(k-1) ... (10)
It is.
[0031]
V (k) = (v_h(k), v_c(k)) T
: Heat storage vector of heat storage tank, v_h(k) is warm water heat storage, v_c(k) is cold water heat storage
V_l= (V_hl, V_cl): Minimum heat storage vector of heat storage tank
V_u= (V_hu, V_cu): Maximum heat storage vector of heat storage tank
P_l= (P_{1 l}, P_2l, ..., p_nl): Minimum power consumption vector of heat source equipment
P_u= (P_1u, P_2u, ..., p_nu): Maximum power consumption vector of heat source equipment
η: heat storage coefficient (η = 1−ε; ε is a heat loss coefficient)
F = f_1h, F_2h, ..., f_nh
f_1c, F_2c, ..., f_nc
: Heat production capacity matrix of heat source equipment, f_1h, F_2h, ..., f_nhIs hot water production capacity,
f_1c, F_2c, ..., f_ncIs cold water production capacity
Q_d(k) = (q_dh, Q_dc) T
: Hot water demand q_dh, Cold water demand q_dcDemand vector with elements
(Formulation of optimization problem)
Here, the optimization problem for obtaining the primary operation plan of the heat source device is the operation state X (k) (k = 1, 2) of each heat source device that minimizes the objective function expressed by the above equation (1). ,..., L). For example, when five heat source devices (n = 5) and L = 14, 70 (= 5 × 14) 0-1 variable strings are obtained.
[0032]
In the objective function expressed by the above equation (1), the first item J_eIs an item related to power consumption. When considering the problem of L = 14 hours (from 8:00 to 21:00), J_eIs to maximize the power consumption at night when the power cost is low, that is, the amount of heat produced at night.
[0033]
Second item J_pIs an item relating to power consumption in the peak cut time zone (period of time for contracting power charges). If the peak cut is made, a large discounted monthly fee according to the power receiving contract is applied, while if the peak cut is not made, a penalty is imposed. For this reason, it is an item that should be especially minimized.
[0034]
Third item J_sIs an item related to the number of start and stop times of the heat source device. J_sMinimizing means that stable operation of the air conditioning / heat source plant 20 is achieved.
[0035]
On the other hand, among the constraints of the above formulas (5) to (9), the first formula (5) represents a state based on the heat flow of the air conditioning / heat source plant 20. The second equation (6) represents the time change of the operating state X (k) of each heat source device, which is a controlled variable. The third equation (7) is for constraining the change in X (k) in one discrete time. The last two formulas (8) and (9) represent the restrictions on the power consumption and the heat storage amount in the time section.
[0036]
The elements of P · X (k) in the objective function (the above formulas (1) to (4)) and the constraint conditions (the above formulas (5) to (9)) are expressed as the following formula (11). Can do.
[0037]
[Expression 2]

Further, the above equation (11) can be expressed as the following equation (12).
[0038]
P_i(k) = a_i・ S_i(k) + b_i                       (12)
here,
S_i(k): Calorie load of the i-th heat source device
It is.
[0039]
Thus, the relationship between the heat input (cold water heat amount and hot water heat amount) load input and energy consumption output (for example, power consumption and fuel consumption) in the heat source device can be expressed by a simple approximate expression such as a linear expression or a broken line. . Here, the relationship between the heat input (cold water heat amount and hot water heat amount) load input and energy consumption output (for example, power consumption and fuel consumption) is approximated by one straight line, but it is approximated by two or more broken lines. Is also possible. The coefficient of the above equation (12) is based on the output characteristics of the heat source equipment obtained by performing offline simulation in advance using an air conditioning / heat source plant model in order to improve the accuracy of the simple approximate expression. A that closely approximates its output characteristics_i, B_i(In the case of a broken line, each coefficient thereof) may be calculated.
[0040]
The optimization problem (integer programming problem) formulated as described above is solved by the heat source operation primary planning unit 13 using an optimization method such as a genetic algorithm, and thereby the primary operation of the heat source device. A plan is required.
[0041]
The integer programming problem includes discrete variables that take discrete values (0, 1), and the problem to be solved becomes very large depending on the number of discrete variables. The scale of the problem is when there are five heat source devices (n = 5) and the operation plan for L = 14 hours (from 8 o'clock to 21 o'clock) every hour is optimized using, for example, all enumeration methods As an example, since there are combinations of driving states of 2 to the 5th power (driving / stopping) every hour, 32 (= 2) to obtain a driving plan for 14 hours.⁵) To the 14th power and evaluation. Some of the actual air conditioning / heat source plants 20 have 10 heat source devices, and even if they have several heat source devices, each unit has a complicated mode (stage operation for each of cold water and hot water). There are also things. Furthermore, there is a case where it is desired to obtain an operation plan in units of 30 minutes or 15 minutes. For this reason, it is inevitable that the problem to be solved becomes large.
[0042]
In order to obtain an operation plan of a heat source device online and in real time, it is necessary to solve such an integer programming problem at high speed, and a genetic algorithm is suitable for this.
[0043]
FIG. 2 is a flowchart for explaining the processing procedure of the genetic algorithm. A genetic algorithm is an algorithm that simulates the evolution process of a living organism from the viewpoint of an information medium called a gene, and is widely studied in the field of optimization problems. In the genetic algorithm, individuals with various gene sequences are subjected to crossover, mutation, and multiplication processes while performing the process of wrinkling with constraints and evaluation functions multiple times (multiple generations). A solution (individual) with good fitness is obtained. The fitness can be defined as the reciprocal of the objective function expressed as the above equation (1).
[0044]
Here, a genetic algorithm is applied to an individual having a gene string for each heat source device that represents an operation state (operation / stop) for each hour unit by 1/0. FIG. 3 is a diagram showing an example of an individual having such a gene sequence, and shows a case where five heat pumps (HP) are used as heat source devices. In this case, one individual gene is represented by 70 bits (= 5 × 14).
[0045]
As shown in FIG. 2, first, n individuals having a random gene sequence are generated and set as an initial individual group (step 101).
[0046]
Next, among the n individuals generated in step 101, a gene string that does not satisfy the above equations (5) to (9) is selected for an individual that does not satisfy the constraints of the above equations (5) to (9). At the same time, the state (1/0) of the gene that does not satisfy the constraint condition included in the selected gene sequence is changed. Then, when n individuals satisfying the constraint condition are generated, the fitness of each individual and the average value of the fitness in the generation are calculated (step 102).
[0047]
Next, select individuals that have a fitness value that is smaller than the specified ratio (the value of the objective function is large) relative to the maximum fitness value in the population, and there are individuals that do not satisfy the constraints at this point. If this happens, hesitate (step 103).
[0048]
Next, the individual having the maximum fitness is multiplied by the number of individuals deceived in step 103 (step 104).
[0049]
Next, as shown in FIGS. 4 (a) and 4 (b), individuals are randomly paired by a predetermined crossover probability (ratio to the total number of individuals), and each pair thus paired is randomly selected. A gene locus is selected, and the same gene locus of each individual is crossed by one point (step 105). One-point crossing means that a series of continuous genes (two continuous genes in the case of FIG. 4B) are crossed with each other.
[0050]
Next, as shown in FIG. 4 (c), individuals are randomly selected for a predetermined mutation rate (ratio to the total number of individuals), and the genes at arbitrary loci of the selected individuals are bit-inverted ( Step 106).
[0051]
When the processing in steps 101 to 106 is completed in this way, the average fitness value in the generation is less than or equal to a predetermined value compared to the previous and previous average fitness values, or is determined in advance. It is determined whether the number of repetitions has been exceeded (step 107), and the processing of steps 101 to 106 is repeated until these end conditions are satisfied. When the above-described termination condition is met, the entire process is terminated, and an operation plan is obtained based on the individual with the maximum fitness of the generation.
[0052]
[Create secondary operation plan]
Next, the heat source operation secondary planning unit 15 uses the genetic algorithm based on the predicted heat load obtained by the heat load prediction unit 11 and the air conditioning / heat source plant model held in the air conditioning / heat source plant model unit 14. By using the optimization method using the above, the vicinity is searched using the primary operation plan of the heat source device obtained by the heat source operation primary planning unit 13 as an initial value, and the optimum operation plan of the heat source device is obtained.
[0053]
Air conditioning / heat source plant model
Here, the air conditioning / heat source plant model held in the air conditioning / heat source plant model unit 14 holds an air conditioning / heat source plant model in which various physical models related to the air conditioning / heat source plant 20 are combined. The air conditioning / heat source plant model is (1) indoor heat transfer model (heat transfer model for rooms and walls), (2) heat exchange model such as cooling coil, heating coil and humidifier for air conditioner, (3) It is a combination of various physical models such as airline calculation formulas, (4) temperature distribution model of heat storage tank, and (5) models of heat source equipment (absorption refrigerator, turbo refrigerator, heat pump, etc.) .
[0054]
(Air conditioning process)
First, the procedure of the simulation of the air conditioning process by the air conditioning / heat source plant model will be described with reference to FIG.
[0055]
First, building data such as the size of a room to be air-conditioned, heat capacity, wall area, and heat transmissivity of a wall are set (step 201). In addition, the outside air temperature, the outside air humidity, the amount of solar radiation, and the like are set as weather data related to the heat load (air conditioning load) of the air conditioner (step 202). Furthermore, the amount of heat generated by a person and the amount of heat generated by electrical equipment (OA equipment, etc.) and lighting, etc. are calculated in advance for the simulation time according to changes in the occupancy rate of the person, and set as indoor heat load data (Step 203). Also, the amount of heat acquired through the window glass surface based on the amount of solar radiation is calculated and set in advance.
[0056]
Here, in this air conditioning process, since the operation of a LOC (local object controller) used in the control is also simulated, the LOC control algorithm in step 205 is performed at intervals of a sampling control cycle (an integer multiple of the simulator calculation cycle) ( Step 204). In step 205, the PID algorithm (speed type) is calculated mainly based on the deviation between the room temperature setting value and the room temperature, and the deviation between the humidity setting value and the room humidity.
[0057]
After that, the air condition points (temperature, humidity, enthalpy, etc.) in the air conditioner are calculated in the order of the cooling coil, heating coil, and humidifier, then the theoretical calculation formula (heat exchange model) and air diagram of the heat exchanger. It is determined using an equation or the like (step 206). Further, a state point of indoor air is obtained using an indoor heat transfer model described later (step 207). Further, a state point of air at the inlet of the air conditioner is obtained by calculation of mixing of recirculated air and introduced outside air (step 208).
[0058]
It should be noted that the processing in steps 204 to 208 described above is repeated until the simulation end time is reached (steps 209 and 210).
[0059]
(Indoor heat transfer model)
FIG. 6 is a diagram for explaining an indoor heat transfer model in the air conditioning / heat source plant model.
[0060]
As shown in FIG. 6, the wall of the room is divided into an outer wall (with windows) 51, an inner wall 52, a floor 53 and a ceiling 54, and the structure of the wall is approximated by two layers of concrete and heat insulating material. The equation of state in this case is shown in the following equations (13) to (15).
[0061]
[Equation 3]

Here, i = 1, 2, 3, and 4.
[0062]
Also,
C_RM= C_a・ Ρ ・ VOL + CPF
C_a: Specific heat of air (J / kg · K)
ρ: air density (kg / m³)
VOL: chamber volume (m³)
CPF: Heat capacity of furniture (J / K)
dt: Calculation cycle (s)
K_RMi= A_i・ Α_Ri
A_i: Wall area (m²)
α_Ri: Indoor surface heat transfer coefficient (W / m²・ K)
C_Ai, C_Bi: Wall heat capacity of part i (J / K)
K_Ai, K_Bi: Thermal conductivity of part i x wall area (W / K)
T_Ai, T_Bi: Wall temperature at site i (° C)
T_R: Room temperature (℃)
T₁: (Equivalent) outside air temperature (℃)
T₂: Underfloor temperature (° C)
T₃: Adjacent room temperature (℃)
T₄: Ceiling temperature (℃)
G_O: Clearance air volume (kg / s)
Q_H: Human calorific value (W)
Q_E: Calorific value of electrical equipment (W)
Q_L: Calorific value of lighting (W)
Q_C: Acquisition amount of heat from window glass surface (W)
H_ES: Amount of heat removed from the chamber (= C_a・ G_D・ (T_R-T_RD)) (W)
G_D : Amount of air supply (kg / s)
T_RD: Supply air temperature (℃)
It is.
[0063]
Similarly, the indoor humidity can be obtained by solving the substance balance equation of moisture.
[0064]
(Temperature distribution model of heat storage tank)
Next, the temperature distribution model of the heat storage tank in the air conditioning / heat source plant model will be described.
[0065]
FIG. 7 is a diagram schematically showing a heat source plant that is simulated using an air conditioning / heat source plant model.
[0066]
As shown in FIG. 7, a cold (hot) water heat storage system using a horizontal heat storage tank 63 having two heat source devices 61 and 62 is assumed as a heat source plant. Reference numeral 64 denotes a heat load device. Here, the heat storage tank 63 is divided into n small sections, and numbers 1 to n are assigned to the respective sections. These small sections may be physically divided by grizzling or the like, or may be virtually divided. However, each divided subdivision becomes a complete mixed heat storage tank, and the water temperature in the tank is constant.
[0067]
FIG. 8 is a diagram for explaining a temperature distribution model of the heat storage tank in the air conditioning / heat source plant model, and shows the heat storage tank near the small section i.
[0068]
In addition,
i: Subdivision number (i = 1,..., n)
v: heat storage tank (small section) volume [m³]
θ: Temperature of heat storage tank (in small compartment) [° C]
ul: (in a small compartment) flow rate [m³/ S]
t: Time [s]
Here, there may be a plurality of θin, ulin, θout, and ulout depending on the small section.
[0069]
The temperature distribution in the heat storage tank as shown in FIG. 8 is obtained as follows.
[0070]
(A) Flow rate
The flow ul (t, i) of water in the heat storage tank at time t is obtained from the following equation (16) based on the inflow and outflow in the small section i.
[0071]
[Expression 4]

(B) Water temperature
The water temperature θ (t + Δt, i) of the small compartment i after Δt time [sec] is obtained by the following equation (17) from the flow rate ul (t, i) and the volume v (i) in the obtained small compartment i.
[0072]
[Equation 5]

Here, η is the heat storage efficiency.
[0073]
(Physical model of heat source equipment)
Next, a physical model of the heat source device in the air conditioning / heat source plant will be described. Here, an absorption refrigerator will be described as an example of the heat source device.
[0074]
FIG. 9 is a diagram schematically showing a heat source device (absorption refrigerator) simulated using an air conditioning / heat source plant model. The absorption refrigerator is a stable refrigerator with few moving parts. Usually, an LiBr (lithium bromide) aqueous solution is used as an absorbent and water is used as a refrigerant. Absorption refrigerators are superior in self-stability because they are dominated by the dynamic characteristics of the heat system.
[0075]
As shown in FIG. 9, the absorption refrigerator has four containers of an evaporator 71, an absorber 72, a generator 73, and a condenser 74, and refrigeration is performed by circulating water as a refrigerant in these containers. Constitutes a cycle.
[0076]
Specifically, first, the water stored in the evaporator 71 takes the heat of vaporization from the cold water in the tube to become water vapor, dissolves in the LiBr aqueous solution in the absorber 72, and dilutes the LiBr aqueous solution. The diluted LiBr aqueous solution diluted in the absorber 72 is sent to the generator 73, heated by the generator 73, and again separated into water vapor and the LiBr aqueous solution. Among these, the water vapor is cooled by the condenser 74 and condensed to become water, and returns to the evaporator 71 at the starting point. On the other hand, the condensed concentrated LiBr aqueous solution is sent again to the absorber 72 and used as an absorbent.
[0077]
FIG. 10 shows this refrigeration cycle on a Duhring diagram (also referred to as “h-ζ diagram”) where the horizontal axis represents the temperature of the solution and water and the vertical axis represents the vapor pressure. FIG. 10 shows a cycle of water rotating counterclockwise with respect to four containers (evaporator, absorber, generator and condenser), and a cycle of aqueous LiBr solution reciprocating between the absorber and generator. Exists. The water in the evaporator has a low temperature (about 6 ° C.) and a low vapor pressure (6 mmHg = 800 Pa). The water in the evaporator and the LiBr aqueous solution (40 ° C.) in the absorber have the same water vapor pressure (in fact, the vapor pressure in the absorber is only slightly lower than the vapor pressure in the evaporator). ). Similarly, the LiBr aqueous solution (150 ° C.) in the generator and the water in the condenser have the same water vapor pressure (70 mmHg = 9300 Pa) (in this case too, the vapor pressure in the condenser is higher than the vapor pressure in the generator). Just slightly lower).
[0078]
Here, in order to realize such a refrigeration cycle, each container (evaporator, absorber, generator and condenser) is modeled as follows.
[0079]
Here, about symbols,
T: temperature (° C.)
C: Specific heat (kJ / kg ° C)
Conc: concentration (weight ratio)
W: Weight (kg)
r: latent heat of vaporization (kJ / kg)
q_w     : Latent heat of evaporation (water) (kJ / kmol)
G: Flow rate (kg / sec)
α: Heat transfer coefficient (kJ / h ° C)
h: Enthalpy (kJ)
It is.
[0080]
For the subscripts attached to each symbol,
‘A’: Absorber
'R': Generator
‘C’: Condenser
‘V’: Evaporator
‘LB’: LiBr
‘W’: water
‘Vpr’: Steam
It is.
[0081]
For other variables,
T_clt   : Average temperature of cold water (℃)
G_r     : Amount of generator evaporated water (kg / sec)
G_a     : Absorber absorbed vapor volume (kg / sec)
T_vpr   : Heated steam temperature (° C)
It is.
[0082]
Hereinafter, the model will be described for each container (evaporator, absorber, generator, and condenser).
[0083]
(1) Absorber
A concentrated LiBr aqueous solution as an absorbent that absorbs water vapor generated in the evaporator 71 is always supplied from the generator 73 to the absorber 72. The concentration of the LiBr aqueous solution in the normal absorber 72 is about 60%, and the temperature is about 40 ° C. Its water vapor pressure equilibrates with water vapor at about 6 ° C. Due to this property, the LiBr aqueous solution in the absorber 72 continues to absorb water vapor until the water in the evaporator 71 falls to 6 ° C. On the other hand, by absorbing water vapor, heat of condensation of water vapor is generated in the LiBr aqueous solution, and the temperature rises. When the temperature rises, the water vapor pressure rises and the cooling capacity falls. In order to prevent this, the LiBr aqueous solution is always cooled with cooling water. Further, when water vapor is absorbed, the concentration of the LiBr aqueous solution decreases. This also results in an increase in water vapor pressure and a decrease in cooling capacity. Therefore, a dilute LiBr aqueous solution is continuously sent to the generator 73 in order to concentrate the LiBr aqueous solution again.
[0084]
(Heat balance)
[Formula 6]

(Water balance)
[Expression 7]

(LiBr balance)
[Equation 8]

(Condensation heat)
[Equation 9]

Where μ_wIs the molecular weight of water, ΔT_aRepresents the temperature increase rate of water corresponding to the increase in the vapor pressure of the LiBr aqueous solution when the temperature of the absorber 72 increases once, and is obtained from the Duhring diagram.
[0085]
(2) Generator
A dilute LiBr aqueous solution is sent from the absorber 72 to the generator 73. In the generator 73, the diluted LiBr aqueous solution is heated to evaporate the water and concentrate. As the heating source, high-temperature steam is used. The concentrated LiBr aqueous solution is sent to the absorber 72. The generated water vapor is sent to the condenser 74.
[0086]
(Heat balance)
[Expression 10]

(Water balance)
## EQU11 ##

(LiBr balance)
[Expression 12]

(Heat of evaporation)
[Formula 13]

Where μ_wIs the molecular weight of water, ΔT_rRepresents the temperature rise rate of water corresponding to the vapor pressure rise of the LiBr aqueous solution when the temperature of the generator 73 rises once, and is obtained from the Duhring diagram.
[0087]
(3) Condenser
The condenser 74 is kept at a lower temperature than the generator 73, and the water vapor evaporated in the generator 73 is condensed. In order to keep at a low temperature, it is always cooled by cooling water. The condensed water is sent to the evaporator 71 again.
[0088]
(Heat balance)
[Expression 14]

(Water balance)
[Expression 15]

(4) Evaporator
In the vapor 71, cold water flows through the tube. Water is sprayed from the upper nozzle, and this water evaporates from the surface of the tube, thereby removing heat of vaporization from the tube. Thereby, cold water is cooled.
[0089]
(Heat balance)
[Expression 16]

(Water balance)
[Expression 17]

In the controller, PI calculation is performed so that the cold water supply temperature becomes a set value, and the flow rate of the heating steam is controlled.
[0090]
[Formula 18]

Where K_p: Proportional gain, T_i: Integration time
It is.
[0091]
Here, the heat source operation secondary planning unit 15 is used in the heat source operation primary planning unit 13 based on the air conditioning / heat source plant model and the heat load prediction value obtained by the heat load prediction unit 11 as described above. By using an optimization method using a genetic algorithm similar to that described above, the primary operation plan of the heat source device obtained by the heat source operation primary planning unit 13 is used as an initial value to search for this neighborhood, and optimal operation of the heat source device is performed. Ask for a plan. The heat source operation secondary planning unit 15 is based on the operation plan in the vicinity of the primary operation plan of the heat source device obtained by the heat source operation primary planning unit 13 and is operated by the air conditioning / heat source plant model unit. A simulation is performed using a heat source plant model, and process values such as the amount of electricity and gas used, the amount of production of cold water and hot water, the amount of heat consumed and the temperature change are accurately calculated and output. Thereby, the calculation of the objective function and the check of the constraint condition are accurately performed based on the output value.
[0092]
At this time, before passing the primary operation plan of the heat source device obtained by the heat source operation primary planning unit 13 to the heat source operation secondary planning unit 15 as an initial value, a simulation is performed using the air conditioning / heat source plant model. And check whether or not the operation is actually possible while satisfying the constraints. If the operation plan does not satisfy the constraint conditions, simulation is performed until there is an operation plan that satisfies the constraint conditions, such as the previous operation plan and the previous operation plan, and the constraint conditions are satisfied. Only the operation plan to be performed is passed to the heat source operation secondary planning unit 15 as an initial value.
[0093]
Finally, the plant control unit 16 sends the optimum operation plan of the heat source equipment obtained by the heat source operation secondary planning unit 15 to the air conditioning / heat source plant 20 via the process input / output unit 17. 11 controls the heat source equipment 22, 23, 24 such as a refrigerator and a heat pump (HP) of the air conditioning / heat source plant 20 shown in FIG.
[0094]
As described above, according to the present embodiment, in the air conditioning / heat source plant 20 such as a building, when the operation plan of the heat source devices 22, 23, 24 is created by the optimization method, the objective function and the constraint condition are set in the actual process. Since the air conditioner / heat source plant model that accurately represents the air condition is checked, an appropriate and high-accuracy operation plan is automatically created from the viewpoint of energy saving, low-cost operation, peak cut, etc., and the heat source equipment 22 , 23, 24 can be realized efficiently.
[0095]
In the above-described embodiment, the heat source operation primary planning unit 13 and the heat source operation secondary planning unit 15 use a genetic algorithm as an optimization method, but the heat source operation primary planning is not limited thereto. At least one of the unit 13 and the heat source operation secondary planning unit 15 may use mixed integer programming as an optimization method. Here, the 0-1 integer programming problem formulated as described above can be generally solved by a branch and bound method (literature (translated by Iri, Konno, Tone: “Optimization Handbook”, Chapter VI, (See Asakura Shoten (1995)). For this reason, in mixed integer programming, the whole problem can be solved by solving a subproblem defined by a branch and bound method or the like by linear programming and repeating this for all the subproblems.
[0096]
【The invention's effect】
As described above, according to the present invention, an appropriate and high-accuracy operation plan is automatically created from the viewpoints of energy saving, low-cost operation, peak cut, etc., and efficient operation of the heat source equipment is realized. be able to.
[Brief description of the drawings]
FIG. 1 is a diagram showing an embodiment of an air conditioning heat source equipment optimum operation control apparatus according to the present invention.
FIG. 2 is a flowchart for explaining a processing procedure of a genetic algorithm.
FIG. 3 is a diagram showing an example of an individual having a gene string.
FIG. 4 is a diagram schematically showing an operation on a gene string by a genetic algorithm.
FIG. 5 is a flowchart for explaining a simulation procedure of an air conditioning process based on an air conditioning / heat source plant model.
FIG. 6 is a diagram for explaining an indoor heat transfer model in an air conditioning / heat source plant model.
FIG. 7 is a diagram schematically showing a heat source plant simulated using an air conditioning / heat source plant model.
FIG. 8 is a diagram for explaining a temperature distribution model of a heat storage tank in an air conditioning / heat source plant model.
FIG. 9 is a diagram schematically showing a heat source device (absorption chiller) simulated using an air conditioning / heat source plant model.
10 is a diagram showing the operation of the heat source device (absorption refrigerator) shown in FIG. 9 on a Duhring diagram.
FIG. 11 is a diagram showing an example of an air conditioning / heat source plant to which the air conditioning heat source equipment optimum operation control apparatus according to the present invention is applied.
12 is a diagram showing details of the air conditioning equipment of the air conditioning / heat source plant shown in FIG.
[Explanation of symbols]
11 Thermal load prediction unit
12 Plant simple model section
13 Heat Source Operation Primary Planning Department
14 Air conditioning / heat source plant model section
15 Heat Source Operation Secondary Planning Department
16 Plant control unit
17 Process input / output section
20 Air conditioning / heat source plant

Claims (6)

In the air conditioning heat source equipment optimal operation control device that controls the operation of air conditioning / heat source plants such as buildings,
A heat load prediction means for obtaining a heat load prediction value based on weather data and a heat load actual value of the air conditioning / heat source plant;
A simple plant model means for holding a simple plant model that represents the relationship between the heat load input and the energy consumption output by a simple approximate expression for each heat source device constituting the air conditioning / heat source plant,
A heat source operation primary plan for obtaining a primary operation plan of a heat source device by an optimization method based on the predicted heat load obtained by the heat load prediction means and the plant simple model held in the plant simple model means. Means,
An air conditioning / heat source plant model means for holding an air conditioning / heat source plant model in which various physical models relating to the air conditioning / heat source plant are combined;
Based on the heat load prediction value obtained by the heat load prediction means and the air conditioning / heat source plant model held in the air conditioning / heat source plant model means, the heat source operation primary planning means obtains by an optimization method. An air conditioning heat source equipment optimum operation control device comprising heat source operation secondary planning means for searching for the vicinity of the primary operation plan of the heat source equipment as an initial value and obtaining an optimum operation plan of the heat source equipment. The air conditioning heat source equipment optimum operation control apparatus according to claim 1, wherein at least one of the heat source operation primary planning means and the heat source operation secondary planning means uses a genetic algorithm as the optimization method. The air conditioning heat source equipment optimum operation control apparatus according to claim 1, wherein at least one of the heat source operation primary planning means and the heat source operation secondary planning means uses a mixed integer programming as the optimization method. The air conditioning heat source equipment optimum operation control apparatus according to claim 1, wherein the heat load prediction means acquires the weather data from an external system via communication means. The coefficient of the simple approximate expression of the plant simple model held in the plant simple model means is calculated based on the output characteristics of the heat source equipment obtained by performing the simulation using the air conditioning / heat source plant model. The air conditioning heat source equipment optimum operation control apparatus according to claim 1. Before passing the primary operation plan of the heat source device obtained by the heat source operation primary planning means as an initial value to the heat source operation secondary planning means, a simulation is performed using an air conditioning / heat source plant model, 2. The air conditioning heat source equipment optimum operation control device according to claim 1, wherein only an operation plan that satisfies the above is passed as an initial value to the heat source operation secondary planning means.

Images