JPH06335694A - Controller of device for biologically treating waste water - Google Patents
Controller of device for biologically treating waste waterInfo
- Publication number
- JPH06335694A JPH06335694A JP5128582A JP12858293A JPH06335694A JP H06335694 A JPH06335694 A JP H06335694A JP 5128582 A JP5128582 A JP 5128582A JP 12858293 A JP12858293 A JP 12858293A JP H06335694 A JPH06335694 A JP H06335694A
- Authority
- JP
- Japan
- Prior art keywords
- sludge
- control
- amount
- neural network
- excess
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02W—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO WASTEWATER TREATMENT OR WASTE MANAGEMENT
- Y02W10/00—Technologies for wastewater treatment
- Y02W10/10—Biological treatment of water, waste water, or sewage
Landscapes
- Activated Sludge Processes (AREA)
- Feedback Control In General (AREA)
Abstract
Description
【０００１】[0001]
【産業上の利用分野】本発明は排水の水質制御を行う活
性汚泥処理システムにおける排水の生物学的処理装置の
制御装置に関する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for a biological treatment device for wastewater in an activated sludge treatment system for controlling the quality of wastewater.
【０００２】[0002]
【従来の技術】一般に排水は活性汚泥微生物により好気
的（脱窒，脱リンは必要な場合等、一部嫌気的プロセス
を含む）に処理される。このとき、微生物は増殖するた
め処理系を安定（定常）に維持するためには、活性汚泥
の一部は余剰汚泥として処理系外に排出するが、一部は
返送汚泥として処理系に返送される。余剰汚泥の排出量
の制御としては現在では以下のような手段がある。2. Description of the Related Art Generally, wastewater is aerobically treated by activated sludge microorganisms (including denitrification and dephosphorization, if necessary, including some anaerobic processes). At this time, the microorganisms proliferate, so in order to maintain the treatment system stable (steady), some of the activated sludge is discharged outside the treatment system as excess sludge, but some is returned to the treatment system as return sludge. It Currently, the following measures are available to control the amount of excess sludge discharged.
【０００３】（１）一日当りの目標引き抜き汚泥量を設
定し、余剰の積算流量がその目標値になるまで引き抜く
方法（定量引き抜き制御） （２）曝気槽内汚泥流量の一定割合を毎日引き抜く方法
（汚泥日令制御、ＳＡ制御） （３）系内汚泥量の一定割合を毎日引き抜く方法（平均
汚泥滞留時間制御、ＳＲＴ制御） 等が提案され、一部実用化されているが、現在、余剰汚
泥量制御として最も良好な管理方法はＳＲＴ制御であ
る。(1) A method of setting a target draw-out sludge amount per day and drawing out until the surplus integrated flow rate reaches its target value (quantitative draw-out control) (2) A method of drawing out a fixed ratio of the sludge flow rate in the aeration tank every day (Sludge age control, SA control) (3) A method of extracting a certain proportion of the amount of sludge in the system every day (average sludge retention time control, SRT control), etc. has been proposed and partially put into practical use, but currently there is a surplus. The best management method for sludge amount control is SRT control.
【０００４】また、返送汚泥量制御手段としては、定量
返送手段、流入水比例制御手段（返送率一定制御）、Ｍ
ＬＳＳ演算制御手段および界面一定制御手段等がある。Further, as the returned sludge amount control means, a fixed amount return means, an inflow water proportional control means (return rate constant control), M
There are LSS arithmetic control means and constant interface control means.
【０００５】ここでは余剰汚泥量制御のうちのＳＲＴ制
御を例として述べる。図５はＳＲＴ制御を行うための概
略構成図を示したもので、図５において、曝気槽１の出
力付近に配置された汚泥容量（ＳＶ）計２、並びに活性
汚泥浮遊量（ＭＬＳＳ）計３の検出値をそれぞれ汚泥容
量指標（ＳＶＩ）計４に導入して、汚泥容量指標ＳＶＩ
を算出する。この算出された汚泥容量指標ＳＶＩの値、
流入水量（Ｑｓ）計５による流入水量Ｑｓ、並びに返送
（余剰）汚泥濃度（ＣR）計６１により求められる返送
汚泥量などから、演算処理（ＷＳ）部７によって、最終
沈殿池１０内における汚泥量を演算する。またこの演算
された最終沈殿池内汚泥量ＭＦ、余剰汚泥量（Ｑｗ）計
１１からの余剰汚泥量Ｑｗ、並びに上記の返送（余剰）
濃度ＣRをＳＲＴ制御部８に入力する。そしてこれらの
入力に基づいてＳＲＴ制御部８により余剰汚泥量を求
め、この余剰汚泥量に応じて余剰汚泥引き抜き用の余剰
汚泥ポンプ９をＯＮ／ＯＦＦ制御して所定量の余剰汚泥
を排出することで、処理系を安定に維持している。Here, SRT control of the excess sludge amount control will be described as an example. FIG. 5 shows a schematic configuration diagram for performing SRT control. In FIG. 5, sludge volume (SV) meter 2 and activated sludge floating amount (MLSS) meter 3 arranged near the output of aeration tank 1 are shown. The detected values of SVI are introduced into the sludge volume index (SVI) meter 4 respectively, and the sludge volume index SVI
To calculate. The value of the calculated sludge capacity index SVI,
Inflow water amount Qs by inflow water amount (Qs) five, and the like return (excess) sludge concentration (C R) return sludge weight determined by meter 61, by the processing (WS) unit 7, the sludge in the settling tank 10 Calculate the quantity. Further, the calculated sludge amount MF in the final settling basin, the excess sludge amount Qw from the excess sludge amount (Qw) total 11, and the above-mentioned return (excess)
The concentration C R is input to the SRT controller 8. Then, based on these inputs, the SRT control unit 8 calculates the excess sludge amount, and ON / OFF-controls the excess sludge pump 9 for extracting the excess sludge in accordance with this excess sludge amount to discharge a predetermined amount of excess sludge. Therefore, the processing system is kept stable.
【０００６】[0006]
【発明が解決しようとする課題】上述した汚泥量制御シ
ステムにおいては、活性汚泥の沈降特性を考慮していな
いため、汚泥の沈降特性が変化すると、制御が目標値に
維持できなくなる。また、返送汚泥量制御と余剰汚泥量
制御が別々に制御されるため、相互干渉により相互に悪
影響を及ぼすおそれがある。In the sludge amount control system described above, since the sedimentation characteristics of activated sludge are not taken into consideration, if the sedimentation characteristics of sludge change, the control cannot be maintained at the target value. Further, since the control of the amount of sludge to be returned and the control of the amount of excess sludge are separately controlled, there is a risk that they may adversely affect each other due to mutual interference.
【０００７】本発明は上記の事情に鑑みてなされたもの
で、汚泥の沈降特性が変化した場合も安定した制御がで
きるとともに、返送、余剰汚泥量制御を同時に行って相
互干渉を防止するようにした排水の生物学的処理装置の
制御装置を提供することを目的とする。The present invention has been made in view of the above circumstances. It is possible to perform stable control even when the sedimentation characteristics of sludge are changed, and to prevent mutual interference by simultaneously performing return and excess sludge amount control. An object of the present invention is to provide a control device for a biological treatment device for waste water.
【０００８】[0008]
【課題を解決するための手段】本発明は上記の目的を達
成するために、排水の生物学的処理装置において、この
処理装置におけるＳＲＴ設定値、汚泥容量指標（ＳＶ
Ｉ）および流入水量をニューラルネットモデルへの入力
変数とし、このニューラルネットモデルで前記変数を処
理して出力に返送汚泥量と余剰汚泥量とを出力変数とし
て送出すると共に、前記ニューラルネットモデルは階層
形モデルであることを特徴とするものである。In order to achieve the above-mentioned object, the present invention provides a biological treatment apparatus for wastewater, wherein an SRT set value and a sludge volume index (SV) in this treatment apparatus are used.
I) and the amount of inflow water are used as input variables to the neural network model, the variable is processed by this neural network model, and the amount of returned sludge and the amount of excess sludge are sent to the output as output variables. It is characterized by being a shape model.
【０００９】[0009]
【作用】定常解析プログラムにより返送汚泥量制御およ
び余剰汚泥量制御方式からそれぞれ一方式づつ選択し、
選択した制御方式の各設定値（目標値）の組み合わせを
変更し、その収束解を求めて教師データを生成する。こ
の教師データとニューロモデル出力値との誤差の二乗和
を評価関数としてこの値が最小となるようにニューロモ
デルに含まれる重み係数やオフセット値の同定（学習）
を行う。[Operation] Select one of the returned sludge amount control method and the surplus sludge amount control method by the steady analysis program,
The combination of each set value (target value) of the selected control method is changed, the convergent solution is obtained, and the teacher data is generated. Identification of the weighting factors and offset values included in the neuro model so that the sum of squares of the error between the teacher data and the output value of the neuro model is used as an evaluation function (minimum)
I do.
【００１０】[0010]
【実施例】以下本発明の実施例を図面に基づいて説明す
るに、図５に示した従来の制御システムとの主な違い
は、演算処理部７とＳＲＴ制御部８に代えて、ニューラ
ルネット制御部２０を設けたことである。DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. The main difference from the conventional control system shown in FIG. 5 is that a neural network is used instead of the arithmetic processing unit 7 and the SRT control unit 8. That is, the control unit 20 is provided.
【００１１】ＳＲＴ制御における操作量は余剰汚泥量
（通常は１日当りの余剰汚泥量）であるから、ニューラ
ルネット制御の入力項目（入力変数）はＳＲＴ制御下で
この余剰汚泥量に影響を及ぼす因子が選定され、具体的
には、汚泥容量ＳＶ30と活性汚泥浮遊量ＭＬＳＳ濃度よ
り演算される汚泥容量指標ＳＶＩ、平均汚泥滞留時間Ｓ
ＲＴの設定値（ＳＲＴｓｅｔ）および流入水量Ｑｓの３
項目を用いる。また、ニューラルネット制御における操
作量である出力項目（出力変数）は、返送汚泥量ＱRと
余剰汚泥量Ｑｓである。Since the manipulated variable in the SRT control is the surplus sludge amount (usually the surplus sludge amount per day), the input item (input variable) of the neural network control is a factor that affects this surplus sludge amount under the SRT control. Is selected, specifically, the sludge volume index SVI calculated from the sludge volume SV 30 and the activated sludge floating amount MLSS concentration, the average sludge retention time S
3 of RT set value (SRTset) and inflow amount Qs
Use items. Also, the output items (output variable) is the operation amount of the neural net control is returned sludge quantity Q R and excess sludge amount Qs.
【００１２】なお、入力項目としては上記の場合、３項
目としたが、次のような項目を必要に応じて選択する。In the above case, the input items are three items, but the following items are selected as necessary.
【００１３】（１）ＳＶＩ、（２）Ｑｓ、（３）流入Ｂ
ＯＤ濃度（Ｌｓ）、（４）返送汚泥量制御目標値（例え
ば、流入水比例制御の場合、返送比率）、（５）余剰汚
泥量制御目標値（例えば、ＳＲＴ制御の場合はＳＲＴ設
定値、（６）流入ＳＳ濃度、（７）流入ＮＨ4−Ｎ濃
度。(1) SVI, (2) Qs, (3) inflow B
OD concentration (Ls), (4) Return sludge amount control target value (for example, return ratio in case of inflow water proportional control), (5) Surplus sludge amount control target value (for example, SRT set value in case of SRT control, (6) Inflow SS concentration, (7) Inflow NH 4 —N concentration.
【００１４】上記入力項から教師データを生成するには
まず定常解析により上記した返送汚泥量制御および余剰
汚泥量制御方式より、それぞれ一方式づつ選択し、選択
した制御方式の各設定値（目標値）の組み合わせを変更
し、その収束解を求める。その収束解は返送汚泥量ＱR
と余剰汚泥量Ｑｗである。In order to generate the teacher data from the input items, first, one method is selected from the above-mentioned return sludge amount control method and surplus sludge amount control method by steady state analysis, and each set value (target value) of the selected control method is selected. ) Is changed and the convergent solution is obtained. The convergent solution is the amount of returned sludge Q R
And the excess sludge amount Qw.
【００１５】図２にニューラルネットワークの１例を示
す。図２において、入力層２４にはＳＶＩ、ＳＲＴ、Ｑ
ｓの３つのデータが入力される。これらデータは中間層
２５で処理されて出力層２６にＱｗ、Ｑｓの２つの出力
を送出する。この場合、返送汚泥量制御として界面一定
制御、また、余剰汚泥量制御としてはＳＲＴ制御をそれ
ぞれ選択した。ここではＳＶＩを７種類、ＳＲＴも７種
類合計７×７＝４９種類の入力データに対する出力（Ｑ
ｗ、ＱR値）を定常解析により求め、教師データとし
た。FIG. 2 shows an example of the neural network. In FIG. 2, the input layer 24 includes SVI, SRT, and Q.
Three data of s are input. These data are processed by the intermediate layer 25, and two outputs Qw and Qs are sent to the output layer 26. In this case, constant interface control was selected as the amount of sludge to be returned, and SRT control was selected as the amount of excess sludge. Here, there are 7 types of SVI, and 7 types of SRT in total, 7 × 7 = 49 types of output (Q
determined by the steady-state analysis w, the Q R value), was a teacher data.
【００１６】次に、上記教師データとニューラルネット
ワークモデル出力値との誤差の二乗和を評価関数とし
て、この値が最小となるように、ニューラルネットワー
クモデルに含まれる重み係数やオフセットの同定（学
習）を行った。約１万回学習させた場合のＱｗとＱRの
結果を図３と図４に示した。図３と図４から教師データ
とニューラルネット出力値が良く一致したことからニュ
ーラルネットを利用したニューロ制御が可能であるが示
唆された。Next, the sum of squares of the error between the teacher data and the output value of the neural network model is used as an evaluation function, and the weighting coefficient and the offset included in the neural network model are identified (learning) so that these values are minimized. I went. The results of Qw and Q R when was learned 10,000 times shown in FIG. 3 and FIG. 4. It was suggested from FIGS. 3 and 4 that the teaching data and the output value of the neural network were in good agreement, and thus neural control using a neural network was possible.
【００１７】上記のように得られたＱｗは図１に示すよ
うにニューラルネット制御部２０からオンオフ制御器２
１に与えられ、余剰汚泥ポンプ９が制御される。また、
ＱRもニューラルネット制御部２０から送出され、これ
により制御弁２２が制御されてポンプ２３からの流量が
制御される。The Qw obtained as described above is supplied from the neural network control unit 20 to the on / off controller 2 as shown in FIG.
1 to control the excess sludge pump 9. Also,
Q R is also sent from the neural network controller 20, thereby the flow rate of the control valve 22 is controlled pump 23 is controlled.
【００１８】[0018]
【発明の効果】以上述べたように、本発明によれば、生
物学的処理装置の制御装置にニューラルネットワークを
用いることにより、多変量制御（ＱR、Ｑｗの２項目の
ようなもの）が簡単に構築でき、また、学習機能によ
り、ニューロモデルに含まれる重み係数等のパラメータ
を簡単に同定することができる。さらに入力項目にＳＶ
Ｉを含んでいるため、汚泥の沈降特性が変化した場合も
安定した制御が継続でき、しかも返送、余剰汚泥量制御
を同時に行っても相互干渉が生じないようにできる等の
利点がある。As described above, according to the present invention, according to the present invention, by using a neural network to control the biological treatment device, a multivariate control (Q R, like two items Qw) is It can be easily constructed, and the learning function can easily identify parameters such as weighting factors included in the neuro model. In addition, SV in the input item
Since it contains I, there is an advantage that stable control can be continued even if the sedimentation characteristics of sludge change, and mutual interference does not occur even if return and excess sludge amount control are simultaneously performed.
【図１】本発明の実施例を示すニューラルネット制御シ
ステムの説明図である。FIG. 1 is an explanatory diagram of a neural network control system showing an embodiment of the present invention.
【図２】階層形ニューラルネットモデル図である。FIG. 2 is a hierarchical neural network model diagram.
【図３】余剰汚泥量Ｑｗの学習結果を示す説明図であ
る。FIG. 3 is an explanatory diagram showing a learning result of a surplus sludge amount Qw.
【図４】返送汚泥量ＱRの学習結果を示す説明図であ
る。FIG. 4 is an explanatory diagram showing a learning result of an amount of returned sludge Q R.
【図５】従来のＳＲＴ制御装置の構成図である。FIG. 5 is a configuration diagram of a conventional SRT control device.
１…曝気槽 ２…汚泥容量（ＳＶ）計 ３…活性汚泥浮遊量（ＭＬＳＳ）計 ４…汚泥容量指標（ＳＶＩ）計 ５…流入水量（Ｑｓ）計 ９…余剰汚泥ポンプ １０…最終沈殿池 ２０…ニューラルネット制御部 ２１…オンオフ制御部 ２２…制御弁 ２３…ポンプ 1 ... Aeration tank 2 ... Sludge volume (SV) total 3 ... Activated sludge floating amount (MLSS) total 4 ... Sludge volume index (SVI) total 5 ... Inflow water amount (Qs) total 9 ... Excess sludge pump 10 ... Final sedimentation tank 20 ... Neural network control unit 21 ... On-off control unit 22 ... Control valve 23 ... Pump
Claims (1)
処理装置におけるＳＲＴ設定値、汚泥容量指標（ＳＶ
Ｉ）および流入水量をニューラルネットモデルへの入力
変数とし、このニューラルネットモデルで前記変数を処
理して出力に返送汚泥量と余剰汚泥量とを出力変数とし
て送出すると共に、前記ニューラルネットモデルは階層
形モデルであることを特徴とする排水の生物学的処理装
置の制御装置。1. A biological treatment device for wastewater, wherein an SRT set value and a sludge volume index (SV) in this treatment device are used.
I) and the amount of inflow water are used as input variables to the neural network model, the variable is processed by this neural network model, and the amount of returned sludge and the amount of excess sludge are sent to the output as output variables. A controller for a biological treatment device for wastewater, which is a shape model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP5128582A JPH06335694A (en) | 1993-05-31 | 1993-05-31 | Controller of device for biologically treating waste water |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP5128582A JPH06335694A (en) | 1993-05-31 | 1993-05-31 | Controller of device for biologically treating waste water |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPH06335694A true JPH06335694A (en) | 1994-12-06 |
Family
ID=14988321
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP5128582A Pending JPH06335694A (en) | 1993-05-31 | 1993-05-31 | Controller of device for biologically treating waste water |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPH06335694A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10151480A (en) * | 1996-11-25 | 1998-06-09 | Maezawa Ind Inc | Wastewater treatment apparatus and operation method |
| FR2784093A1 (en) * | 1998-10-06 | 2000-04-07 | Suez Lyonnaise Des Eaux | IMPROVEMENTS IN THE TREATMENT OF WASTE WATER BY ACTIVATED SLUDGE METHODS |
| EP1376276A1 (en) * | 2002-06-21 | 2004-01-02 | H2L Co., Ltd | An AI based control system and method for treating sewage/waste water by means of a neural network and a back-propagation algorithm |
| WO2017033160A1 (en) * | 2015-08-27 | 2017-03-02 | Suez International | Method for treating waste water comprising a quick static decanter and associated facility |
| WO2020183576A1 (en) * | 2019-03-11 | 2020-09-17 | 株式会社 ゴーダ水処理技研 | Operation management system for wastewater treatment facility |
-
1993
- 1993-05-31 JP JP5128582A patent/JPH06335694A/en active Pending
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10151480A (en) * | 1996-11-25 | 1998-06-09 | Maezawa Ind Inc | Wastewater treatment apparatus and operation method |
| FR2784093A1 (en) * | 1998-10-06 | 2000-04-07 | Suez Lyonnaise Des Eaux | IMPROVEMENTS IN THE TREATMENT OF WASTE WATER BY ACTIVATED SLUDGE METHODS |
| WO2000020344A1 (en) * | 1998-10-06 | 2000-04-13 | Suez Lyonnaise Des Eaux | Improvements to waste water treatment using activated sludge process |
| KR100642974B1 (en) * | 1998-10-06 | 2006-11-10 | 수에즈 리오네즈 데 조 | Improvements to waste water treatment using activated sludge process |
| EP1376276A1 (en) * | 2002-06-21 | 2004-01-02 | H2L Co., Ltd | An AI based control system and method for treating sewage/waste water by means of a neural network and a back-propagation algorithm |
| WO2017033160A1 (en) * | 2015-08-27 | 2017-03-02 | Suez International | Method for treating waste water comprising a quick static decanter and associated facility |
| FR3040388A1 (en) * | 2015-08-27 | 2017-03-03 | Degremont | PROCESS FOR TREATING WASTEWATER WITH RAPID STATIC DECANTER AND ASSOCIATED INSTALLATION |
| WO2020183576A1 (en) * | 2019-03-11 | 2020-09-17 | 株式会社 ゴーダ水処理技研 | Operation management system for wastewater treatment facility |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Schmitt et al. | Development of artificial neural networks to predict membrane fouling in an anoxic-aerobic membrane bioreactor treating domestic wastewater | |
| Lindberg | Control and estimation strategies applied to the activated sludge process | |
| AU2021101438A4 (en) | Adaptive control method and system for aeration process | |
| FI114019B (en) | A method for controlling wastewater treatment plants, whereby several control functions are utilized | |
| Ingildsen et al. | Dissolved oxygen controller based on on-line measurements of ammonium combining feed-forward and feedback | |
| Dalmau et al. | Comparison of a deterministic and a data driven model to describe MBR fouling | |
| US20030234218A1 (en) | System and method for AI controlling waste-water treatment by neural network and back-propagation algorithm | |
| Schraa et al. | Ammonia-based aeration control with optimal SRT control: improved performance and lower energy consumption | |
| JP4334317B2 (en) | Sewage treatment system | |
| CN114380378B (en) | Intelligent phosphorus control drug feeding method and device and storage medium | |
| Flores-Alsina et al. | Including the effects of filamentous bulking sludge during the simulation of wastewater treatment plants using a risk assessment model | |
| EP0643672B1 (en) | Method of controlling wastewater purification plants using quality evaluation of measuring data | |
| JPH06335694A (en) | Controller of device for biologically treating waste water | |
| Nam et al. | An autonomous operational trajectory searching system for an economic and environmental membrane bioreactor plant using deep reinforcement learning | |
| Lindberg | Multivariable modeling and control of an activated sludge process | |
| JPH07115025B2 (en) | Bulking control method using SVI | |
| Shen et al. | Nitrate control strategies in an activated sludge wastewater treatment process | |
| JP3123167B2 (en) | Nitrification and denitrification process simulator | |
| JP3460211B2 (en) | Sewage treatment control device | |
| JPH06328091A (en) | Sludge capacity index estimating method in control system for biological treatment device | |
| JP7494669B2 (en) | Method and device for predicting parameter values of wastewater treatment simulator, and method and device for controlling sewage treatment plant | |
| JP2002062927A (en) | Waste water treatment process simulator | |
| JPH06335693A (en) | Neurocontroller of activated-sludge process | |
| JPH081189A (en) | Anaerobic and aerobic activated sludge method | |
| Ladiges et al. | Applied off-line expert system for effluent, operational and technical problems of waste water treatment plants |