ѧѧ
ACTASCIENTIAECIRCUMSTANTIAE
19981µ181ڿƼڿ
෨Ķſ*
.DOŨȶзõӰ
顡ꡡӨ
(ѧѧԺ150008)

ժҪеĹڻ෨ſоĻϣ״ҪĿƲŷܽŨ(DO)ΪƱԳˮˮΪԼзΪָĻ෨ſƵо⣬ؽ˻о.ȽſҪĻ״ָ̬ķʽ.Ȼͨоͬһˮǰ¿ƲͬDOŨʱҪз.DOΪ0.9mg/Lʱз٣ӦάDOŨȴ2mg/LĴͳ۵Զ˥ʳKdDOӰļҲ.
ؼʡ෨ſƣܽ(DO)Ũ.
MULTIVARIABLE OPTIMAL CONTROL OF ACTIVATED SLUDGE PROCESS.BASIC THEORY AND EFFECT OF DO ON OPERATIONAL COST

Peng Yongzhen, Wang Baozhen,Wang Shuying
(Department Environment and Municipal Engineering,Harbin University of Architecture and Engineering,150008)

ABSTRACTThis paper originally presents the results of the study on the multivariable optimal control of activated sludge process,which takes the two most important control parameters,sludge wastage and dissolved oxygen(DO) as control variables,regards effluent quality as restriction factor and operation cost as performance index.The basic theory was studies.The fundament state equations and functional expression of performance index were established firstly. The operational costs at various DO concentrations were calculated,under the condition of the same effluent quality.The results indicate that the operational cost at DO of 0\^9mg/L is the lowest,which has great diversity compared with the conventional view that DO level should be maintained above 2mg/L.The assumption that the decay rate constant Kd is not affected by DO is not rational either. 
Keywordsactivated sludge process, optimal control, dissolved oxygen(DO) concentration.

1
ǳˮǹҵˮˮˮˮʱڱ仯ڷȶ״̬УȻͳˮĲϱ仯.ˮԶռ룬ûʵſ.ſһЩҵѵõϹ㷺Ӧ.ǹ̿вߡҪϸͼѶĿƷʽ.ܸݱϵͳ״̬̣Լָ(ҲĿ꺯ָ귺)ȣͨ㣬ṩһ()ſƱ.ſƽй̿ƣ޶ȵʹָ.ѡƱ״̬Լ״ָ̬̺ſоĻҪݣҲֱӰſƵ.
ſ60Žؽйػ෨ſƵо70ո𲽣ǽһʶоҪӦüֵ.˹ڸÿоص㣺һ漰෨ϵͳſо빫ĲࣻǶоΪƱ16ݣڵſƣŷ(ˮ)ΪƱ78ݣǼеоƽˮŨȻˮŨȲСΪָ2459ݣֻKabouris7ݵȼоԻŷܺСΪָ.Hamalainen2ݡCheruy3ݺKabouris7ݵ˵оΪƱٳˮˮʲЧܲЧ.ΪⲢЧؿеĻҺŨ(MLSS)Ӵ˶γеˮŶ.Kabouris7ݵΪŷȿЧ.
Ŀǰˮˮеĵ(л)ŨȽϸ߻򲨶зýϸձڵ⣬ҲԵо.ĽϵſƵоŷ(Qwʾ)ܽŨ(DOʾ)ΪƱ԰ʣദ빩ߵз(ܺ)֮Ϊָ꣬ԳˮˮΪԼ.Ȼ˽ˮˮˮ֮⣬QwDOǻ෨Ҫ.оָ˻෨ľ󲿷зãˮˮҪǰ£зҪĿ.ол(BOD5ơҲƵ)ŷƳˮлŨȡͬʱƳˮƽлŨȵֲͬˮŷű׼£ʹзʡſ⣬һҪойصɻ.

2ſƵĻѧģ
2.1״̵̬Ľ
ȫʽ෨ϵͳӦѧڹҵˮҲй㷺Ӧã͵Ҳɿɸȫʽش.ˣԸϵͳΪо.Grieves10Milbury11ݵ˵оʽеʵʻ״̬ӽȫ.оԶ΢ĴлˮܽΪ˱ڶŷжƣ涨ֱ࣬ϵͳͼ1ʾ.



ͼ1෨ϵͳʾͼ
Fig.1 Flow schem of activated sludge process and control
ͼ뱾еķ£
QQr,Qwֱʾˮŷŵ(m3/d)
XXrMLSSͻŨȣmg/L
So,Sˮˮ()ĵ(BOD)Ũȣmg/L;
DOܽŨȣmg/L
VЧݻm3;
R.
ȶͼ1ʾϵͳƽ⣺

(1)



(2)
ʽ(dS)/(dt)еŨȵı仯ʣmg/(L*d)
kʳd-1
Ksͳmg/L.
kS/(Ks+S)LawrenceMcCarty12ĲͬŨµĵ.ٶͼ1ϵͳ΢ƽ⣺

(3)


(4)
ʽ(dX)/(dt)΢Ũȵı仯ʣmg/(L*d)
Kd΢˥ʣd-1
ʽ(2)(4)ͼ1ʾĻϵͳĻ״̬.
DOΪƱ뿼DO״̬е.DOӦʵӰǹϵģҪڶԵȽk˥KdӰ.ǶĿغ(Switching function)ʾDOkKdӰ1314ݣʾ

(5)
ʽKoΪĿسıͳ(mg/L)ȡֵطӳDOkKdӰСһΪKoΪ0.2mg/Lң13.ʽ(5)DOӦkKdӰ󣬽ʽ(4)(2)

(6)
ʽ(6)ܽŨDOŷQwΪƱеŨS΢ŨXΪ״̬Ļ෨̵Ļ״̬.ʽеX,S,QwDOȱʱ仯ģʱtĺҲɷֱдX(t),S(t),Qw(t)DO(t).
2.2ָĽ
ϵͳзȻƱDOQwйأÿյзJcʾʣദJ1J2صĹJ3
Jc=J1+J2+J3(7)
ڷȶ״̬X,S,QwDOʱĺJ1J1J3зӦ÷ʾ.J1ŨˮɻõķãȻJ1ŷQwй.һ(t0=0d,T=1d)зJ1ʾ£
J1=(λĴA)*(ŷŵ)=TtoAQwXdt(8)
ݶͼ1ʾϵͳƽ⣬ɵ
QrXr=(Q+Qr-Qw)X(9)
(10)
ʽ(10)ɼQrQQwXXrĺʱtĸϺJ2ɱʾΪ
J2=(λB)*(Qr)(11)
к(OUR)ʽ15£
(12)
ʽRaеOURkg/d
aõλϵ
b΢Դʣһb=1.42Kd,d-1.
ԭˮDOˮлҪһЩܽܽΪQ*DOԣеܺRtΪ
Rt=Ra+Q*DO(13)
ǵDOkKdӰ죬RtΪ
(14)
Ϊͬһ豸ĻеЧʲ䣬תƵλCתЧʦǳɷ.ɱʾתroĹʽro=KLa(Ds-DO)֪һתϵKLaıŨDsǳroDs-DOҲȣDOʸصԹϵ
(15)
ʽЦ1C1ֱʾܽŨΪDO1ʱתЧʺתƵλ.ضDO1C1֪ģCDOĺϵΪ
(16)
صĹJ3ɱʾΪ
J3=(תƵλC)(ܺ)TtoCRtdt
(17)
ʽ(8)(11)(17)ʽ(7)T=1d,to0dɵõȶ״̬»෨ϵͳһյз
(18)
Ҫ÷ʾָʽҲоſ.

3DOŨȶзõӰ
ӷ(6)(18)ԿDO״̬зõı仯кܴӰ.ˮˮʵǰ£άֽϴDOɽãͬʱҲߵȽʣMLSSŨԼȷзã֮Ȼ.ͳĹ۵ΪDOӦά2mg/LΪˣ15.ôDOзõĹϵ?DOӦķΧڸ?ǱҪ۵ģҲǱоĻ.
DOӰҪ棺һǶתƵλCӰ죬ʽ(16)ظǶӦʵӰ죬ҪȡıͳKoֵ.Koȡֵ˵硣䷶ΧΪKo=0.20.4mg/LһΪKo0.2mg/L13.
ˮQˮ(BOD)ŨSoÿʱtҲ仯£
(19)
ʽУˮƽֵȡ=10000m3/d
SoˮŨȵƽֵȡSo=0.150kg/m3.
пƶභ䣬֮䲻أԣMLSSŨŨXrı仯Ȳ̫󣬽صĹͨˮعϵŨ̶ˮʸعϵXrQСQСд

(20)
1ѧ뻯ѧֵ
Table 1 Biokinetic and stoichiometrycoefficients and other parameters

ȡֵλ
k1.501/d
Ks0.046kg(BOD)/m3
Y0.58kg(MLSS)/kg(BOD)
Kd0.051/d
a0.42kg(O2)/kg(BOD)
Ko0.0002kg(O2)/m3
Ds0.0085kg(O2)/m3
DO10.0015kg(O2)/m3
r6.00kg(MLSS)/m3
A1.10Ԫ/kg(MLSS)
B0.042Ԫ/m3
C10.950Ԫ/kg(O2)
V2500m3
10000m3/d
o0.150kg(BOD)/m3

ʽrƻ׼ŨȣƽʱŨȣһΪ50008000mg/L.

ݻ෨ѧ뻯ѧԼͨȡֵΧ1ǵĵֵʱ.1кoֵʽ(19)ȷĽˮQSoı仯ͼ2ʾ.


ͼ2ˮŨʱı仯
Fig.2 Hourly variation in influentflow and substrate concentration
ΪܱȽڳˮˮͬʱͬĶDOµзãȹ涨ÿųĵΪ150kg(BOD)/d(൱ƽˮBODΪ15mg/L)ڼлʹ״̬ĳʼֵĩֵͬX(0)=X(1),S(0)=S(1)޷Ƚз.һƱQwҲֵ.ȷDOQwԱ1иֵʽ(6)(18)ü΢ַʾ״̬(6)Ȼ״̬ĽX(t)S(t)ʽ(18)õзJC.ӡı2ͼ3ֱʾDOΪ2.50.5mg/Lʱ״̬Ʊʱı仯зJCZ(t)ΪÿϵͳųBODʱtı仯зõλΪԪ.
2DOΪ2.5mg/Lʱ״̬Ʊı仯
Table 2 Changes of state variables and control variables at DO=2.5mg/L

X(t),mg/LS(t),mg/LQw(t),m3/dDO(t),mg/Lt,dZ(t),kg/d
2103.09.58283.7502.500000
2102.712.79283.7502.50000.0506.024
2106.316.42283.7502.50000.10014.987
2113.520.01283.7502.50000.15027.336
2123.822.96283.7502.50000.20042.914
2136.224.64283.7502.50000.25060.766
2149.324.60283.7502.50000.30079.243
2161.622.78283.7502.50000.35096.446
2171.919.57283.7502.50000.400110.836
2178.815.71283.7502.50000.450121.686
2181.611.92283.7502.50000.500129.138
2180.28.71283.7502.50000.550133.885
2175.06.29283.7502.50000.600136.775
2166.84.64283.7502.50000.650138.534
2156.73.66283.7502.50000.700139.677
2145.53.25283.7502.50000.750140.544
2134.13.35283.7502.50000.800141.372
2123.33.96283.7502.50000.850142.377
2114.05.15283.7502.50000.900143.842
2106.97.01283.7502.50000.950146.184
2103.09.58283.7502.50001.000149.998
J(N)=2630.0660(Ԫ/d)JC(N)=2630.0660(Ԫ/d)


ͼ3DOΪ0.5mg/Lʱ״̬Ʊı仯(JC=2491Ԫ/d)
Fig.3 Changes of state variables and control varlables at 
DO=0.5mg/L(JC=2491Yuan/d)

ӦָڼʱһŷŵĵZ(1)״̬ĩֵX(1)S(1)ֱʼX(0)S(0)(൱Ӧȥĵ)ɼСзJCǲѽ͵.ˣΪ˱Ҫдļ.⣬ͬ˿DOΪ2.01.51.21.11.00.90.80.70.60.3mg/Lʱļ.DOļСXŨΪͬʱQwͨcάֳˮˮʲ.DOԽ乩J3ҲԽʡԽߣôDOзJCĹϵΣǷʹзʡDOŨ?
ݼӡ˱ʾDOJCϵı3ͼ4.ڲͬˮദ̶뵥λദAĲϴ󣬶λBתƵλC1Ĳ첢Ϊˣ3ͼ4иAֵΪ1.12.1Ԫ/kgʱJCֵ.
3ƲͬDOŨʱзJC(Ԫ/d)
Table 3 Operation costs JC(yuan/d) at various DO concentrations
AԪ/kgDOŨȣmg/L
0.300.500.600.700.800.901.001.101.201.502.50
1.1026031883185618431837183618361843184918762022
2.1026712491246424512445244424462450245725482630

ϿɼZ(1)Ϊ150kg/dͬЧǰ£DO=0.71.2mg/LΧڣзJCϵұ仯DO=0.9mg/LʱJC.άDOŨȴڵ2mg/LʱĴͳ۵ʮݱ׼Ĺ涨15ȥԶ.DOӰԼѡʵDOŨ⣬Ϊ


ͼ4AֵΪ1.12.2Ԫ/kgʱDOŨJCϵ
Fig.4 Relation between DO and JC at A=1.1 and A=2.2yuan/kg
1.һ˵DO1.0mg/LʱͲȱ˿״.ǵ⣬ӦDO1.01.2?mg/LΪˣDOΪ2mg/Lʱ5%ҵз.ʵÿˮܽʡ1%зҲǺܿɹ.
2.ˮȾоЭ(IAWPRC)1986ġģ1Уٶ΢˥KdDOŨȵӰ13.оһ޸״̬(6)ָ(18)󣬼ƲͬDOʱзJC.˿ƵDOŨԽͣJCҲԽСƣDOΪ0.2mg/LʱJCС.DOŨȺܵҲӰKdĴС΢Դʲ£ּ輰.
3.ıͳKoԽDOӦʵӰԽʱƲͬDOJCĹϵ?оKoȡֵΧ0.4mg/Lֵ1ʽ(6)ʽ(18)㲻ͬDOµJCDO1.1mg/LʱJCСŽDO2.0mg/LϵĴͳ۵Բܴ.

4
ЧظıˮˮʣŷQw(ĴС)ܽŨDOǻ෨ҪƱ.״QwDOΪƱԳˮΪԼзΪָĻ෨ſойصĻо.
1.ŷܽŨΪƱMLSSŨȺ͵ŨΪ״̬״̬.
2.ʣദ͹з֮ΪſƵָķʽ.״ָ̬궼ſоĻ.
3.״ָ̡̬ˮзõһָ꣬ͨоͬƲͬĶDOŨʱҪзãDOΪ0.9mg/Lʱзʡ0.71.2mg/LΧڣзýϵұ仯.ǵֹͻDOŨȿܶӦиӰ(Ko=0.4mg/L)DO1.01.2mg/LҲ㹻ˣʹDO2.0mg/L.
4.оļIAWPRC13˥KdDOӰļ貢ҲԽ.

μӱоĻ㡢ܡ޺ƽͬ־.
ο
1Harris C J.Modelling and adaptive control of urban wastewater treatment plants.Environmental System Planning,Design,and Control.Proc IFAC Symp,Kyoto,Japan 1977
2 Hamalainen R P, Halme A,Gyllenberg A .A control model for activated sludge wastewater treatment process.Proc Ifac 6th World Congress,Boston,Mass, 1980
3 Cheruy A,Panzarella L,Denat J P.Multimodel simulation and adaptive stochastic control of an activated sludge process. Modelling and Control of Biotechnological Processes.Proc IFAC Symp,Helsinki,Finland,1982
4 Sincic D,Bailey E.Optimal periodic control of activated sludge processes-.Results for the base case with monod/decay kinetics.Wat Res,1978,12:4753
5S Y S Yeung,Sincic D,Bailey E.Optimal periodic control of activated sludge Processes-.Comparison with conventional control for structured sludge Kinetics.Wat Res,198014:7783
6 Chen M.Stochastic modelling and stochastic optimization of chemical engineering process ph.D.thesis,Department of Chemical Engineering,Kansas State University,Manhattan,Kansas1969
7 Kabouris J C,Georgakakos A P.Optimal control of the activated sludge Process.Wat Res,1990,24(10):11971208
8 Angelbeck D I,Shah Alam A B.Simulation studies on optimization of the activated sludge process.J WPCF,1978,50(1):31_39
9 Kabouris J C,Georgakakos A P,Camara A.Optimal control of the activated sludge process:Effect of sludge storage.Wat Res,1992,26(4):507517
10 Grieves R B,Milbury W F,Pipes W O.A Mixing model for activated sludge.J WPCF,1964,36:619
11 Milbury W F,Pipes W O,Grieves R B.Compartmentalization of aeration tank.J Sanit Eng Div,ASCE,1965,91(SA3):45
12 Lawrence A W,McCarty P L.Unifled basis for biological treatment design and operation.J Sanit Eng Div,ASCE,1970,96(SA3):757
13 Henze M et al. (IAWPRC Task Group on Mathematical Modelling for Design and Operation of Biological Wastewater Treatment) Activated sludge model No.1,IAWPRC.Scientific and Technical Reports No.1, 1986
14ĵ.ˮﴦӦ.йҵ磬1989
15Benefield L D,Randall C W.Biological Porcess design for wastewater treatment,Prentice-Hall Inc 1980
