Study of Composite Fuzzy Control of Dissolved Oxygen in a ...
Proceedings of the 2012 IEEE International Conference on Robotics and Biomimetics December 11-14, 2012, Guangzhou, China
Study of Composite Fuzzy Control of Dissolved Oxygen in a Sequencing Batch Reactor Pilot Process of Synthetic Papermaking Wastewater Wenhao Shen, Erpan Tao, Li Ning and Tianlong Liu
Abstract — The studies on the automatic control strategy of papermaking wastewater treatment process have always been sparse for a long time. During the process of treatment, the oxygen concentration in the wastewater is a key substrate in animal cell metabolism and its consumption is a parameter of great interest for the monitoring. This paper introduces an effective and robust control strategy to control the dissolved oxygen concentration in a pilot sequencing batch reactor of wastewater treatment process. For comparison, four control strategies are tested and evaluated: ON/OFF, PID, fuzzy and composite fuzzy control strategies. The control results with composite fuzzy control strategy show that, the controlled dissolved oxygen concentration is around the set point with 0.05 standard deviation and 3.5 minutes rising time, but the good controlling performance is obtained at the price of relatively high aeration volume for the heavier pollution load. It takes less time than that of the other strategies to arrive the ammonia valley, which shows its good ability of chemical oxygen demand removal. From the compared results of these four control strategies, we can conclude that the proposed composite fuzzy control strategy is proved to be a robust and effective controller for the DO concentration of synthetic papermaking wastewater.
I. INTRODUCTION The automatic control technology in sewage like urban sewage treatment process has been studied and applied widely. But the corresponding researches in the papermaking wastewater treatment process are scarce [1]. The papermaking industry consumes large quantities of water, and the generated wastewater has its own characteristics, such as the vast flow rate and the heavy pollution load [2]. In order to satisfy the new severe discharge standard of effluent for papermaking industry, the development of advanced control strategies in papermaking wastewater treatment process is necessary and important. It is well known that the sequencing batch reactor (SBR) has been greatly focused and applied in decades for its Manuscript received July 9, 2012. This work was supported by the financial support of National Natural Science Foundation of China (61074109). W.H. Shen is with State Key Lab of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640, P.R. China. 008620-87110961; E-mail: [email protected] E.P. Tao is with State Key Lab of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640, P.R. China. E- mail: [email protected] L. Ning is with State Key Lab of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640, P.R. China. E-mail: [email protected], [email protected] T.L. Liu is with State Key Lab of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640, P.R. China. E- mail: [email protected]
978-1-4673-2126-6/12/$31.00 © 2012 IEEE
control of dissolved oxygen (DO) concentration has become one of the most important parameters in the process for its impact on the biological processes and the energy saving related to aeration [4]. The constant of DO concentration around 2 mg/L satisfies the need of the growth of bacterium in wastewater [5-7]. Because of the inherent nonlinearity and complexity of DO concentration in wastewater, the traditional control strategies like ON/OFF and proportional-integral-derivative (PID) are not competent for it, while intelligent techniques like fuzzy control, neural networks, expert and process identification can obtain better control effects [8-11], and can offer integrating expert knowledge and control rules of the process [12]. This paper is an early exploration before the application in the actual papermaking wastewater treatment process. It concerns the DO control of wastewater in a laboratorial SBR pilot process. Treating the synthetic papermaking wastewater instead of the wastewater from the paper mill, and only considering the important aeration phase during the process, four control strategies which include ON/OFF, PID, fuzzy and composite fuzzy control strategy, are implemented to test the feasibility of obtaining the best DO control result for the synthetic papermaking wastewater treatment process. II. MATERIALS AND METHODS A. SBR pilot process The preparation method of synthetic papermaking wastewater refers to Wang [13]. The volume of synthetic papermaking wastewater in SBR pilot process is about 32 L. The aim of the SBR pilot process mainly concerns the removal of biological nutrient in the wastewater. An unchangeable cycle, 6 hours, which includes eight steps: static feeding, blending, aerating feeding, aeration, blending, settling, sludge discharge and effluent discharge (Fig.1), is provided by the engineers in the wastewater treatment plant of paper mill. In order to maximize the removal rate of COD, the aeration phases accounts for 57 % of the whole cycle. The scheme of the SBR pilot process is shown in Fig.2. The biological reactions of the treating process take place in a cylinder-shaped reactor (40 L capacity). The monitoring and control system is based on the sensors, programmable logic controller (PLC) and interfaces developed with KingView 6.53 (WellinTech). The process is equipped with DO (Clean- DO5200), pH (ENTEX PC-300 EGC-148/PH), oxidation-reduction potential (ORP) (ENTEX PC-300 EGC-149/ORP) and level (HS-PTH601) sensors. The signals are captured by S7-200 (Siemens). The control is con-
1262
ducted with analog and digital output boards of S7-200. A user friendly interface can repeat a previously defined operation cycle by controlling the switch ON/OFF, adjust the filling and draining of the peristaltic pumps. The blender S tatic feed ing
0
S lud ge d isch arge
B lend ing A eration feed in g
10
15
and air supply are controlled by a compressor whose frequency is calculated by PLC according to the measurement of the DO concentration. The sampling time of DO, pH, ORP and Level is one second.
A eration
B len d in g
40
220
S ettling
230
E fflu en t d isch arge
295 300 305
360
T im e (m inu tes)
Fig. 1. Operational cycle periods of the SBR pilot process. Six hours cycle with eight steps.
B. Dissolved oxygen parameter The DO is an important ecological parameter in the SBR process. The actual DO concentration of papermaking wastewater treatment plants is influenced by many factors, such as temperature of the wastewater, hydraulic retention time, sludge age, and the oxygen demand of various floras. Therefore, its control is of extremely importance in the field. However, the SBR is such a complicated physical and biochemical process that it is difficult even impossible to develop an accurate mathematical model to describe it [10]. Fortunately, the skilful operators who work in the SBR site can manually regulate the DO concentration well; this stimulated us to design a fuzzy controller with their experiences. DO, pH, ORP and Level
1600
A - Aeration (L/h)
1400
Peristaltic pump
Flow meter
FI
Air
Compressor
Peristaltic pump
A
B
Peristaltic pump Frequency converter
Fig. 2. Schematic of the SBR pilot process.
C. Control strategy The objective of the control in the study is to develop an on-line control action to make the DO concentration in the SBR be stable and be as near as possible to 2 mg/L. The following specialties are considered during the development of an efficient DO controller for the synthetic papermaking wastewater:
800 600 400
9 12 15 18 21 24 27 30 33 36 39 42 45 48 51
F - Frequency (Hz)
Effluent Sludge discharge
1000
Fig. 3. Relationship of input frequency and output aeration of f requency converter.
Sensors Influent
1200
0
blender
Aerator
A =-2479.7+305.9 F-11.1 F 2 +0.2F 3
200
PLC
IPC SBR reactor
Thirdly, the relationship of input frequency and output aeration of the frequency converter is not linear. The aeration will not start until the input frequency is greater than 13 Hz. The relationship of input frequency and the output aeration satisfies the nonlinear equation in Fig. 3.
Firstly, the DO evolution in the reactor strongly depends on both the microorganism’s behaviour and the wastewater pollution level. Secondly, the inherent 60 seconds delay of DO probe greatly affects the precision of real-time control algorithms.
These specialties demonstrate a strong nonlinear process, hence, four control strategies including ON/OFF, PID, Fuzzy and composite fuzzy controllers have been considered to deal with the nonlinearity of the SBR process. In the case of an ON/OFF control strategy, the controller instantaneously commands the system. As long as the DO is lower than the set point, the command variable is set to ON position; conversely, the command variable is set to OFF position. The PID is the most common control strategy in the process control. It is a continuous feedback loop that keeps the process running normally by taking corrective action whenever there is any deviation from the set point of a process variable. An error occurs when an operator manually changes the set point or when an event or a disturbance changes the load, thus causes a change in the process variable. Fuzzy logic is the emulation of human reasoning on computers. Since it was introduced by Zadeh in 1965 [14], it has become popular in various applications, ranging from space shuttle control to industrial process control. The key concepts in fuzzy logic are the linguistic variable and membership function. A linguistic variable is a variable whose value is not a number but a word, described in terms of numerals by a membership function. To perform fuzzy
1263
2.8
1.2 0.8 0.4 0.0 -20
0
20
7.8
40
60
80
Time(min)
100 120 140
7.7
ORP (mV)
7.5
pH
60
80
7.4 7.3
AV
7.2 7.1 0
20
40
60
80
Time(min)
100 120 140
0
20
40
60
80
Time(min)
100 120 140
Fig. 4. Controlling results with ON/OFF control strategy in the aeration phase of synthetic wastewater treatment process.
Frequency (Hz)
1.6 1.2 0.8 0.4 0
20
40
60
80
Time(min)
100 120 140
55
0
60
30 20 10 0 -20
0
20
0
20
40
60
80
100 120 140
40
60
80
100 120 140
Time(min)
240
7.4 7.3
DO (mg/L)
7.0 -20
40
300
AV
7.1 0
20
40
60
80
100 120 140
Time(min)
180 120 60 0 -60 -20
Time(min)
Fig. 6. Controlling results with PID control strategy in the aeration phase of synthetic wastewater treatment process.
120
0
50
T ime (min)
50
7.2
180
-60 -20
45
7.5
100 120 140
60
40
7.6
240
7.6
7.0 -20
40
5
7.7
0
Time(min)
10
7.8
10
20
15
2.0
-20
20
0
20
0.0
30
-20
25
2.4
40
300
30
2.8
DO (mg/L)
Frequency (Hz)
1.6
35
2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5
S etpoint DO A era tion
40
45
50 45 40 35 30 25 20 15 10 5 0
Frequency (Hz)
DO (mg/L)
2.0
40
The application of the PID control strategy was very hard since we did not have a process model. Consequently, the traditional techniques for controller parameters determination were delicate to realize. Our optimal adjusted parameters of PID controller are as follows: proportional gain Kc is 0.65, integration time constant Ti is 0.01 minutes and derivative time constant Td is 270 minutes. As revealed in Fig.6 and Fig.7, the amplitude fluctuations of DO concentration is violent during the whole transient process, and the DO concentration presents a significant overshoot even larger than that of the ON/OFF control strategy.
50
2.4
45
ORP (mV)
The DO control of the SBR pilot process with the simple ON/OFF algorithm was carried out basing on two given limiting frequencies: 42 Hz and 0 Hz. When the DO concentration was lower than the set point (2mg/L), the output of frequency converter was set at 42 Hz, which meant that the maximum aeration performed; whereas the output of frequency converter was set at 0 Hz when the DO concentration exceeded the set point, which implied that no aeration existed. It is worth noting that the value of 42 Hz is an empirical parameter which depends on the specific wastewater treatment process. As shown in Figs.4 and 5, the controlling result of DO presents frequent fluctuation and significant overshoot around the set point. The sustaining oscillations were due to the system inertia and the controller structure. This phenomenon was more or less amplified and the algorithm was proved to be sensitive to the DO probe delay, the water pollution level and the bacteriological activity. The absence of the intermediate states between ON and OFF made the control action be not very efficient, especially with small errors due to the delay of sensors.
50
B. PID control
DO (mg/L)
A. On/off control
S etpoint DO A eration
Fig. 5. Enlarged part of the controlling results of DO concentration with ON/OFF control strategy.
pH
III. RESULTS AND DISCUSSION
2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5
Frequence (Hz)
logic reasoning, a set of IF–THEN rules that emulate human reasoning needs to be defined. An IF–THEN rule is a simple statement that expresses cause and effect and offers a final decision. This operation is called inferences. Fuzzy inference systems have been proven to be powerful tools for solving complex problems. A detailed description of fuzzy logic applied to DO control can refer to [15, 16]. Expert system, artificial neural networks, fuzzy logic technology and other intelligent control methods are still emerging disciplines at the initial stage of development and have not yet formed themselves complete theoretical systems. Taking advantage of integrating two or three intelligent control methods or integrating intelligent and traditional control methods together and overcoming their flaws, are to achieve efficient and stable operations of wastewater treatment systems [1,17]. The integration of fuzzy, ON/OFF, and PID control to get a composite fuzzy control of DO is used in the study.
50 55 60 Time (min) Fig. 7. Enlarged part of the controlling results of DO concentration with PID control strategy.
C. Fuzzy control A two inputs - one output fuzzy controller was designed for the SBR wastewater treatment system. The first input is
1264
TABLE 1 CONTROL RULE TABLE OF DO FUZZY CONTROL STRATEGY U
NB PB PB PB PB PM PS Z
NB NM NS Z PS PM PB
e
NM PB PB PB PM PS Z NS
de Z PB PB PM Z NS NM NB
NS PB PB PM PS Z NS NM
PS PB PM PS NS NS NB NB
PM PM PS Z NM NM NB NB
PB PS Z NS NB NB NB NB
Figs. 8 and 9 present the results of the fuzzy DO control for aeration phase. Comparing with the obtained results by ON/OFF and PID methods, the results are better but also not very satisfactory. In fact, in order to approach the set point of DO, the variation range of output frequency is large; and the lowest frequency of converter reached 15.72 Hz, which was not good for the system to be stable. However, the overshoot of the response was tiny because its largest output frequency was not high. 2.8
Frequency (Hz)
DO (mg/L)
2.0 1.6 1.2 0.8 0.4 0.0 -20
0
20
7.8
40
60
80 100 120 140
Time(min)
7.7
DO (mg/L)
40 35 30 25 20 15 10 5
40
45
50
55
In order to overcome the shortcoming of the fuzzy control, the composite fuzzy control was proposed. As demonstrates in Fig. 10, when the DO concentration is in the two ranges, [0, 1.1] or [2.2, ∞], the ON/OFF control strategy will be used; when the DO concentration is in the two ranges, [1.1, 1.9] or [2.1, 2.2], the PID control strategy (whose controller parameters are shown in Fig.10) will be used; when the DO concentration is between [1.9, 2.1], the fuzzy control strategy will be used. The designed core thoughts of composite fuzzy control is as follows: the usages of ON/OFF control strategy in the low and high DO values are to bring the DO close to the set point quickly and not to overshoot high; making the process goes towards the range of fuzzy control gently with PID control strategy; in the narrow range around the set point (red circle sign in Fig.10), the fuzzy control strategy is applied. The point that needs to be made about the fuzzy control strategy is that, the bounds for the continuous sets of e and de are defined as [-0.1, 0.1] and [-0.05, 0.05] respectively according to the narrow range of DO concentration. Furthermore, when the error e locates in the different universes, the reference frequency of the frequency converter output and the variation of the output frequency are defined as follows. U 30 Hz 3 Hz
E 6, 5, 4, 3, 2, 1
U 27.5 Hz 2.5 Hz
E0
U 25 Hz 5 Hz
30
E 1, 2, 3, 4, 5, 6
20 10 0 -20
0
20
40
60
80 100 120 140
Time(min)
ORP (mV)
120
7.4 7.3
AV
7.2 7.1 0
20
40
60
80 100 120 140
Time(min)
f1.9~2.0 (30 3)Hz
60
f 2.0 (27.5 2.5)Hz f 2.0~2.1 (25 5) Hz
0 -60 -20
0
20
40
60
0
D. Composite fuzzy control
40
300
60
Fig. 9. Enlarged part of the controlling results of DO concentration with fuzzy control strategy.
180
7.5
pH
45
240
7.6
7.0 -20
50 Setpoint DO Aeration
Time (min)
50
2.4
2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5
Frequency (Hz)
the error between the required value and the measured value of DO, named e. The second input is the change rate of DO error, named de. The output was the output frequency of the frequency converter, named U. The e, de and U were calculated every two seconds. The bounds for continuous sets of e, de and U are defined as [-2, 1], [-0.2, 0.2] and [-9, 9] respectively according to the pilot SBR process. Simultaneously, the universes of e, de, and U were defined as {-6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6}. The fuzzy control rule was showed in Table 1, and the other information refers to Liu [16]. According to the operating experiences, the reference frequency of the frequency converter output and the variation of the output frequency were defined as 25.72 Hz and 15.72 Hz separately, consequently, the whole range of output was [10, 41.44]. The fuzzy control query table contains 169 output data which can be calculated quickly with the proposed method by Liu [16], and the further calculation of the corresponding control signals are performed and put into the data block of PLC to be called in real time.
80 100 120 140
Time(min)
Fig. 8. Controlling results with fuzzy control strategy in the aeration phase of synthetic wastewater treatment process. Fig. 10. Description of composite fuzzy control strategy.
1265
Consequently, the whole ranges of the corresponding outputs of the frequency converter are as follows. U 27, 33
E 6, 5, 4, 3, 2, 1
U 25, 30
E 0
U 20, 30
E 1, 2, 3, 4, 5, 6
Figs.11 and 12 present the controlling results with the composite fuzzy DO control strategy. Comparing with Figs.5, 7 and 9 by ON/OFF, PID and fuzzy control methods, the result of composite fuzzy control is optimal. It makes the DO concentration fluctuate slowly around the set point with tiny overshoot. The reason for the good performance of composite fuzzy control strategy could be easily got from the comparisons of the manipulated frequency variation ranges with different controllers; the details are included in Table 2. In Table 2, the frequency variation range of composite fuzzy control is the smallest one in the four control strategies，that is the reason why the fluctuations of the DO concentration with it are the slowest one among the four control strategies, and its robustness and stability is the optimal. 2.8
50
Frequency (Hz)
DO (mg/L)
2.4 2.0 1.6 1.2 0.8 0.4 0.0 -20
0
20
7.8
40
60
80 100 120 140
20 10 0 0
20
0
20
300
40
60
80 100 120 140
40
60
80 100 120 140
Time(min)
240
ORP (mV)
7.6 7.5
pH
30
-20
Time(min)
7.7
7.4 7.3 7.2
AV
7.1 7.0 -20
40
0
20
40
60
80 100 120 140
Time(min)
180 120 60 0 -60 -20
Time(min)
TABLE 2 COMPARISON OF THE MANIPULATED FREQUENCY VARIATION RANGES WITH DIFFERENT CONTROLLERS Control strategy
ON/OFF control
PID control
Fuzzy control
Composite Fuzzy control
Frequency variation (Hz)
0, 42
0-50
14-33
21-33
As demonstrated in Table 3, the standard deviation of composite fuzzy control is 0.05, which is the smallest one in all of the four control strategies. The maximum DO is 2.14 mg/L which is the smallest one in four, and the minimum DO is 1.87 mg/L which is the largest one in four. These data indicates that the composite control strategy is the best one to control the DO concentration in SBR pilot process. Except for the controlling performance, the operational performance should also be considered. Although the controlling performance of composite fuzzy control is optimal, its consumed aeration volume is 1840 L, which is the largest one in four and implies that it is not an energy saving strategy. What is the reason for this result? We reckon that is the result of the increasing polluted load of treated wastewater. The experiments of control strategies were carried out in the order of ON/OFF, PID, fuzzy and composite fuzzy control strategy. Since the bacteria in the SBR needed a period of accustoming to the wastewater, more synthetic wastewater was added into the reactor prior to the implement of each control strategy. Therefore, with more and more wastewater adding into the reactor, the pollution level of the wastewater increased accordingly. It is evidenced that the initial ORP values of wastewater in four batches with four control strategies decrease gradually. As demonstrated in Figs. 4, 6, 8 and 11, the corresponding initial ORP values of wastewater with ON/OFF, PID, fuzzy and composite fuzzy control strategies are about 140 mV, 70 mV, 30 mV and -60 mV individually, which indicates the pollution level of wastewater was rising gradually. That is the main element to explain the above phenomenon.
2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5
TABLE 3 STATISTICAL RESULTS FOR DO CONTROL
50 Setpoint DO Aeration
45
Control result
40 35 30 25 20 15 10
Frequency (Hz)
DO (mg/L)
Fig. 11. Controlling results with composite fuzzy control strategy in the aeration phase of synthetic wastewater treatment process.
Rising time (min)
Items
5 40
45
50
Time (min)
55
60
0
Fig. 12. Enlarged part of the controlling results of DO concentration with composite fuzzy control strategy.
Control algorithms Composite ON/OFF PID Fuzzy fuzzy 4.2 3.3 4.9 3.5
Overshoot (%)
14
23
1
5.5
Mean (mg/L)
2.10
2.11
2.00
2.01
Max (mg/L)
2.67
2.54
2.74
2.14
Min (mg/L)
1.59
1.69
1.62
1.87
Std
0.24
0.22
0.20
0.05
Aeration volume (L)
1181
1194 1697
1840
According to Li [18], when the ammonia depletion is achieved, a minimum point will appear in the pH profile, the point is called ammonia valley (AV). Watching the AV
1266
signs in Figs. 4, 6, 8 and 11, the locations of AV with ON/OFF, PID and fuzzy control strategies are almost the same, about 70 minutes; while that with the composite fuzzy control is about 50 minutes, which is much earlier than the other control strategies. In other words, the ability of removal COD with composite fuzzy control is much better than that with other three kinds of control strategies. IV. CONCLUSIONS This paper presents an effective and robust control strategy to control the DO concentration in a pilot SBR wastewater treatment process. Constructing with Siemens PLC S7-200, necessary sensors and KingView 6.53 software, the data sampling and processing system is set up; developing with Matlab, a new composite fuzzy control strategy is studied. For the purpose of comparison, four control strategies are tested and evaluated: ON/OFF, PID, fuzzy and composite fuzzy control strategies. Their controlling results demonstrate that the new developed composite fuzzy control strategy is proved to be robust and effective. The controlled DO concentration was around the set point with 0.05 standard deviation and 3.5 minutes rising time, but the good controlling performance is obtained at the price of 1840 L aeration volume which is relatively high for the heavy pollution load of wastewater. It takes only 50 minutes which was less than that of the other strategies to arrive the ammonia valley; this shows its great ability of COD removal. The algorithm was not affected by the probe delay or the process strong nonlinearity. The obtained results showed the effectiveness of composite fuzzy controller and its superiority compared with the traditional regulation techniques and only with fuzzy control. It is the best control strategy in the provided four to satisfy the pilot synthetic papermaking wastewater treatment process.
[9] T.R. Girija and C. Mahanta, Fuzzy logic and artificial neural network approaches for dissolved oxygen prediction. International Journal of Environment and Waste Management, vol. 6, 2010, pp. 237 - 254. [10] M.Z. Jia, B.B. Gao, and B. Jia, The expert controller for dissolved oxygen control system in wastewater treatment. Communications in Computer and Information Science, vol. 237, 2011, pp. 351-356. [11] C. Yoo and M.H. Kim, Industrial experience of process identification and set-point decision algorithm in a full-scale treatment plant. Journal of Environmental Management, vol. 90 , 2009, pp. 28232830. [12] S. Grieu, A. Traoré, M. Polit, and J. Colprim, Prediction of parameters characterizing the state of a pollution removal biologic process. Eng. Appl. Artif. Intell, vol. 18, 2005, pp. 559–573. [13] C.L. Wang, Reducing the environmental impact of a paper mill’s water effluent. China Pulp & Paper, vol. 30 (S), 2011, pp. 156-160. [14] L.A. Zadeh, Fuzzy Sets. Information and Control, vol. 8, 1965, pp. 338-353. [15] A. Traoré, S. Grieu, S. Puig, L. Corominas, F. Thiery, M. Polit, and J. Colprim, Fuzzy control of dissolved oxygen in a sequencing batch reactor pilot plant. Chemical Engineering Journal, vol. 111, 2005, pp. 13–19. [16] T.L. Liu, L. Ning, and W. H. Shen, A quick way to get fuzzy control query table and its application in papermaking wastewater treatment process. Paper Science & Technology, vol. 30, 2011, pp. 78-82. [17] S.B. Gu, S.Y. Wang, Q. Yang, and Y.Z. Peng, Research progress on intelligent control strategies of SBR process. Industrial Water & Wastewater, vol. 40, 2009, pp. 1-6. [18] L.Y. Li, Y.Z. Peng, L. Li, and S.Y. Wang, On-line determination and variations of specific oxygen uptake rate in activated sludge system. Journal of Chemical Industry and Engineering (China), vol.61, 2010, pp. 995-1000.
REFERENCES [1] W. H. Shen and L. Ning, The study of the development situation of indstrial wastewater treatment automatic motoring technology. Paper Science & Technology, vol. 30, 2011, pp. 83-91. [2] W. Liu, Progress and trends in treatment technologies of pulp and paper mill effluents. China Pulp & Paper, vol. 30 (S), 2011, pp. 156160. [3] S.Y. Wang, S.B. Gu, Q. Yang, and Y.Z. Peng, Research progress on real-time control strategies for sequencing batch reactors. Acta Scientiae Circumstantiae, vol. 29, 2009, pp. 1121-1130. [4] S. Puig, L. Corominas, A. Traoré, J. Colomer, M.D. Balaguer, and J. Colgrim, An on-line optimization of a SBR cycle for carbon and nitrogen removal based on on-line pH and OUR: the dissolved oxygen control role. Proceedings of the Second IWA Conference on Instrumentation, Control and Automation for Water and Wastewater Treatment and Transport System. 2005. [5] Z.W. Wang, Y.Z. Peng, S.Y. Wang, J.H. Guo, H.J Huang, and Y.N. Zhen, Feasibility study for limited filamentous bulking under low dissolved oxygen at different operation regimes. Environmental Science, vol. 32, 2011, pp. 2347-2352. [6] Y. Ma, Y.Z. Peng, S.Y. Wang, and Y. Bai, DO control strategies and its fuzzy controller design in A/O process. Modern Chemical Industry, vol. 24 (S1): 2004, pp. 154-157. [7] W.J. Li and Y.W Ma,. Study on designing fuzzy controller for DO in A/O Process. Measurement & Control, vol. 26, 2010, pp. 54-56. [8] L.P. Fan and Boshnakov. Kosta, Fuzzy logic based dissolved oxygen control for SBR wastewater treatment process. Proceedings of the 8th World Congress on Intelligent Control and Automation, Jinan, China, July. 2010.
1267
Study of Composite Fuzzy Control of Dissolved Oxygen in a Sequencing Batch Reactor Pilot Process of Synthetic Papermaking Wastewater Wenhao Shen, Erpan Tao, Li Ning and Tianlong Liu
Abstract — The studies on the automatic control strategy of papermaking wastewater treatment process have always been sparse for a long time. During the process of treatment, the oxygen concentration in the wastewater is a key substrate in animal cell metabolism and its consumption is a parameter of great interest for the monitoring. This paper introduces an effective and robust control strategy to control the dissolved oxygen concentration in a pilot sequencing batch reactor of wastewater treatment process. For comparison, four control strategies are tested and evaluated: ON/OFF, PID, fuzzy and composite fuzzy control strategies. The control results with composite fuzzy control strategy show that, the controlled dissolved oxygen concentration is around the set point with 0.05 standard deviation and 3.5 minutes rising time, but the good controlling performance is obtained at the price of relatively high aeration volume for the heavier pollution load. It takes less time than that of the other strategies to arrive the ammonia valley, which shows its good ability of chemical oxygen demand removal. From the compared results of these four control strategies, we can conclude that the proposed composite fuzzy control strategy is proved to be a robust and effective controller for the DO concentration of synthetic papermaking wastewater.
I. INTRODUCTION The automatic control technology in sewage like urban sewage treatment process has been studied and applied widely. But the corresponding researches in the papermaking wastewater treatment process are scarce [1]. The papermaking industry consumes large quantities of water, and the generated wastewater has its own characteristics, such as the vast flow rate and the heavy pollution load [2]. In order to satisfy the new severe discharge standard of effluent for papermaking industry, the development of advanced control strategies in papermaking wastewater treatment process is necessary and important. It is well known that the sequencing batch reactor (SBR) has been greatly focused and applied in decades for its Manuscript received July 9, 2012. This work was supported by the financial support of National Natural Science Foundation of China (61074109). W.H. Shen is with State Key Lab of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640, P.R. China. 008620-87110961; E-mail: [email protected] E.P. Tao is with State Key Lab of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640, P.R. China. E- mail: [email protected] L. Ning is with State Key Lab of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640, P.R. China. E-mail: [email protected], [email protected] T.L. Liu is with State Key Lab of Pulp and Paper Engineering, South China University of Technology, Guangzhou, 510640, P.R. China. E- mail: [email protected]
978-1-4673-2126-6/12/$31.00 © 2012 IEEE
control of dissolved oxygen (DO) concentration has become one of the most important parameters in the process for its impact on the biological processes and the energy saving related to aeration [4]. The constant of DO concentration around 2 mg/L satisfies the need of the growth of bacterium in wastewater [5-7]. Because of the inherent nonlinearity and complexity of DO concentration in wastewater, the traditional control strategies like ON/OFF and proportional-integral-derivative (PID) are not competent for it, while intelligent techniques like fuzzy control, neural networks, expert and process identification can obtain better control effects [8-11], and can offer integrating expert knowledge and control rules of the process [12]. This paper is an early exploration before the application in the actual papermaking wastewater treatment process. It concerns the DO control of wastewater in a laboratorial SBR pilot process. Treating the synthetic papermaking wastewater instead of the wastewater from the paper mill, and only considering the important aeration phase during the process, four control strategies which include ON/OFF, PID, fuzzy and composite fuzzy control strategy, are implemented to test the feasibility of obtaining the best DO control result for the synthetic papermaking wastewater treatment process. II. MATERIALS AND METHODS A. SBR pilot process The preparation method of synthetic papermaking wastewater refers to Wang [13]. The volume of synthetic papermaking wastewater in SBR pilot process is about 32 L. The aim of the SBR pilot process mainly concerns the removal of biological nutrient in the wastewater. An unchangeable cycle, 6 hours, which includes eight steps: static feeding, blending, aerating feeding, aeration, blending, settling, sludge discharge and effluent discharge (Fig.1), is provided by the engineers in the wastewater treatment plant of paper mill. In order to maximize the removal rate of COD, the aeration phases accounts for 57 % of the whole cycle. The scheme of the SBR pilot process is shown in Fig.2. The biological reactions of the treating process take place in a cylinder-shaped reactor (40 L capacity). The monitoring and control system is based on the sensors, programmable logic controller (PLC) and interfaces developed with KingView 6.53 (WellinTech). The process is equipped with DO (Clean- DO5200), pH (ENTEX PC-300 EGC-148/PH), oxidation-reduction potential (ORP) (ENTEX PC-300 EGC-149/ORP) and level (HS-PTH601) sensors. The signals are captured by S7-200 (Siemens). The control is con-
1262
ducted with analog and digital output boards of S7-200. A user friendly interface can repeat a previously defined operation cycle by controlling the switch ON/OFF, adjust the filling and draining of the peristaltic pumps. The blender S tatic feed ing
0
S lud ge d isch arge
B lend ing A eration feed in g
10
15
and air supply are controlled by a compressor whose frequency is calculated by PLC according to the measurement of the DO concentration. The sampling time of DO, pH, ORP and Level is one second.
A eration
B len d in g
40
220
S ettling
230
E fflu en t d isch arge
295 300 305
360
T im e (m inu tes)
Fig. 1. Operational cycle periods of the SBR pilot process. Six hours cycle with eight steps.
B. Dissolved oxygen parameter The DO is an important ecological parameter in the SBR process. The actual DO concentration of papermaking wastewater treatment plants is influenced by many factors, such as temperature of the wastewater, hydraulic retention time, sludge age, and the oxygen demand of various floras. Therefore, its control is of extremely importance in the field. However, the SBR is such a complicated physical and biochemical process that it is difficult even impossible to develop an accurate mathematical model to describe it [10]. Fortunately, the skilful operators who work in the SBR site can manually regulate the DO concentration well; this stimulated us to design a fuzzy controller with their experiences. DO, pH, ORP and Level
1600
A - Aeration (L/h)
1400
Peristaltic pump
Flow meter
FI
Air
Compressor
Peristaltic pump
A
B
Peristaltic pump Frequency converter
Fig. 2. Schematic of the SBR pilot process.
C. Control strategy The objective of the control in the study is to develop an on-line control action to make the DO concentration in the SBR be stable and be as near as possible to 2 mg/L. The following specialties are considered during the development of an efficient DO controller for the synthetic papermaking wastewater:
800 600 400
9 12 15 18 21 24 27 30 33 36 39 42 45 48 51
F - Frequency (Hz)
Effluent Sludge discharge
1000
Fig. 3. Relationship of input frequency and output aeration of f requency converter.
Sensors Influent
1200
0
blender
Aerator
A =-2479.7+305.9 F-11.1 F 2 +0.2F 3
200
PLC
IPC SBR reactor
Thirdly, the relationship of input frequency and output aeration of the frequency converter is not linear. The aeration will not start until the input frequency is greater than 13 Hz. The relationship of input frequency and the output aeration satisfies the nonlinear equation in Fig. 3.
Firstly, the DO evolution in the reactor strongly depends on both the microorganism’s behaviour and the wastewater pollution level. Secondly, the inherent 60 seconds delay of DO probe greatly affects the precision of real-time control algorithms.
These specialties demonstrate a strong nonlinear process, hence, four control strategies including ON/OFF, PID, Fuzzy and composite fuzzy controllers have been considered to deal with the nonlinearity of the SBR process. In the case of an ON/OFF control strategy, the controller instantaneously commands the system. As long as the DO is lower than the set point, the command variable is set to ON position; conversely, the command variable is set to OFF position. The PID is the most common control strategy in the process control. It is a continuous feedback loop that keeps the process running normally by taking corrective action whenever there is any deviation from the set point of a process variable. An error occurs when an operator manually changes the set point or when an event or a disturbance changes the load, thus causes a change in the process variable. Fuzzy logic is the emulation of human reasoning on computers. Since it was introduced by Zadeh in 1965 [14], it has become popular in various applications, ranging from space shuttle control to industrial process control. The key concepts in fuzzy logic are the linguistic variable and membership function. A linguistic variable is a variable whose value is not a number but a word, described in terms of numerals by a membership function. To perform fuzzy
1263
2.8
1.2 0.8 0.4 0.0 -20
0
20
7.8
40
60
80
Time(min)
100 120 140
7.7
ORP (mV)
7.5
pH
60
80
7.4 7.3
AV
7.2 7.1 0
20
40
60
80
Time(min)
100 120 140
0
20
40
60
80
Time(min)
100 120 140
Fig. 4. Controlling results with ON/OFF control strategy in the aeration phase of synthetic wastewater treatment process.
Frequency (Hz)
1.6 1.2 0.8 0.4 0
20
40
60
80
Time(min)
100 120 140
55
0
60
30 20 10 0 -20
0
20
0
20
40
60
80
100 120 140
40
60
80
100 120 140
Time(min)
240
7.4 7.3
DO (mg/L)
7.0 -20
40
300
AV
7.1 0
20
40
60
80
100 120 140
Time(min)
180 120 60 0 -60 -20
Time(min)
Fig. 6. Controlling results with PID control strategy in the aeration phase of synthetic wastewater treatment process.
120
0
50
T ime (min)
50
7.2
180
-60 -20
45
7.5
100 120 140
60
40
7.6
240
7.6
7.0 -20
40
5
7.7
0
Time(min)
10
7.8
10
20
15
2.0
-20
20
0
20
0.0
30
-20
25
2.4
40
300
30
2.8
DO (mg/L)
Frequency (Hz)
1.6
35
2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5
S etpoint DO A era tion
40
45
50 45 40 35 30 25 20 15 10 5 0
Frequency (Hz)
DO (mg/L)
2.0
40
The application of the PID control strategy was very hard since we did not have a process model. Consequently, the traditional techniques for controller parameters determination were delicate to realize. Our optimal adjusted parameters of PID controller are as follows: proportional gain Kc is 0.65, integration time constant Ti is 0.01 minutes and derivative time constant Td is 270 minutes. As revealed in Fig.6 and Fig.7, the amplitude fluctuations of DO concentration is violent during the whole transient process, and the DO concentration presents a significant overshoot even larger than that of the ON/OFF control strategy.
50
2.4
45
ORP (mV)
The DO control of the SBR pilot process with the simple ON/OFF algorithm was carried out basing on two given limiting frequencies: 42 Hz and 0 Hz. When the DO concentration was lower than the set point (2mg/L), the output of frequency converter was set at 42 Hz, which meant that the maximum aeration performed; whereas the output of frequency converter was set at 0 Hz when the DO concentration exceeded the set point, which implied that no aeration existed. It is worth noting that the value of 42 Hz is an empirical parameter which depends on the specific wastewater treatment process. As shown in Figs.4 and 5, the controlling result of DO presents frequent fluctuation and significant overshoot around the set point. The sustaining oscillations were due to the system inertia and the controller structure. This phenomenon was more or less amplified and the algorithm was proved to be sensitive to the DO probe delay, the water pollution level and the bacteriological activity. The absence of the intermediate states between ON and OFF made the control action be not very efficient, especially with small errors due to the delay of sensors.
50
B. PID control
DO (mg/L)
A. On/off control
S etpoint DO A eration
Fig. 5. Enlarged part of the controlling results of DO concentration with ON/OFF control strategy.
pH
III. RESULTS AND DISCUSSION
2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5
Frequence (Hz)
logic reasoning, a set of IF–THEN rules that emulate human reasoning needs to be defined. An IF–THEN rule is a simple statement that expresses cause and effect and offers a final decision. This operation is called inferences. Fuzzy inference systems have been proven to be powerful tools for solving complex problems. A detailed description of fuzzy logic applied to DO control can refer to [15, 16]. Expert system, artificial neural networks, fuzzy logic technology and other intelligent control methods are still emerging disciplines at the initial stage of development and have not yet formed themselves complete theoretical systems. Taking advantage of integrating two or three intelligent control methods or integrating intelligent and traditional control methods together and overcoming their flaws, are to achieve efficient and stable operations of wastewater treatment systems [1,17]. The integration of fuzzy, ON/OFF, and PID control to get a composite fuzzy control of DO is used in the study.
50 55 60 Time (min) Fig. 7. Enlarged part of the controlling results of DO concentration with PID control strategy.
C. Fuzzy control A two inputs - one output fuzzy controller was designed for the SBR wastewater treatment system. The first input is
1264
TABLE 1 CONTROL RULE TABLE OF DO FUZZY CONTROL STRATEGY U
NB PB PB PB PB PM PS Z
NB NM NS Z PS PM PB
e
NM PB PB PB PM PS Z NS
de Z PB PB PM Z NS NM NB
NS PB PB PM PS Z NS NM
PS PB PM PS NS NS NB NB
PM PM PS Z NM NM NB NB
PB PS Z NS NB NB NB NB
Figs. 8 and 9 present the results of the fuzzy DO control for aeration phase. Comparing with the obtained results by ON/OFF and PID methods, the results are better but also not very satisfactory. In fact, in order to approach the set point of DO, the variation range of output frequency is large; and the lowest frequency of converter reached 15.72 Hz, which was not good for the system to be stable. However, the overshoot of the response was tiny because its largest output frequency was not high. 2.8
Frequency (Hz)
DO (mg/L)
2.0 1.6 1.2 0.8 0.4 0.0 -20
0
20
7.8
40
60
80 100 120 140
Time(min)
7.7
DO (mg/L)
40 35 30 25 20 15 10 5
40
45
50
55
In order to overcome the shortcoming of the fuzzy control, the composite fuzzy control was proposed. As demonstrates in Fig. 10, when the DO concentration is in the two ranges, [0, 1.1] or [2.2, ∞], the ON/OFF control strategy will be used; when the DO concentration is in the two ranges, [1.1, 1.9] or [2.1, 2.2], the PID control strategy (whose controller parameters are shown in Fig.10) will be used; when the DO concentration is between [1.9, 2.1], the fuzzy control strategy will be used. The designed core thoughts of composite fuzzy control is as follows: the usages of ON/OFF control strategy in the low and high DO values are to bring the DO close to the set point quickly and not to overshoot high; making the process goes towards the range of fuzzy control gently with PID control strategy; in the narrow range around the set point (red circle sign in Fig.10), the fuzzy control strategy is applied. The point that needs to be made about the fuzzy control strategy is that, the bounds for the continuous sets of e and de are defined as [-0.1, 0.1] and [-0.05, 0.05] respectively according to the narrow range of DO concentration. Furthermore, when the error e locates in the different universes, the reference frequency of the frequency converter output and the variation of the output frequency are defined as follows. U 30 Hz 3 Hz
E 6, 5, 4, 3, 2, 1
U 27.5 Hz 2.5 Hz
E0
U 25 Hz 5 Hz
30
E 1, 2, 3, 4, 5, 6
20 10 0 -20
0
20
40
60
80 100 120 140
Time(min)
ORP (mV)
120
7.4 7.3
AV
7.2 7.1 0
20
40
60
80 100 120 140
Time(min)
f1.9~2.0 (30 3)Hz
60
f 2.0 (27.5 2.5)Hz f 2.0~2.1 (25 5) Hz
0 -60 -20
0
20
40
60
0
D. Composite fuzzy control
40
300
60
Fig. 9. Enlarged part of the controlling results of DO concentration with fuzzy control strategy.
180
7.5
pH
45
240
7.6
7.0 -20
50 Setpoint DO Aeration
Time (min)
50
2.4
2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5
Frequency (Hz)
the error between the required value and the measured value of DO, named e. The second input is the change rate of DO error, named de. The output was the output frequency of the frequency converter, named U. The e, de and U were calculated every two seconds. The bounds for continuous sets of e, de and U are defined as [-2, 1], [-0.2, 0.2] and [-9, 9] respectively according to the pilot SBR process. Simultaneously, the universes of e, de, and U were defined as {-6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6}. The fuzzy control rule was showed in Table 1, and the other information refers to Liu [16]. According to the operating experiences, the reference frequency of the frequency converter output and the variation of the output frequency were defined as 25.72 Hz and 15.72 Hz separately, consequently, the whole range of output was [10, 41.44]. The fuzzy control query table contains 169 output data which can be calculated quickly with the proposed method by Liu [16], and the further calculation of the corresponding control signals are performed and put into the data block of PLC to be called in real time.
80 100 120 140
Time(min)
Fig. 8. Controlling results with fuzzy control strategy in the aeration phase of synthetic wastewater treatment process. Fig. 10. Description of composite fuzzy control strategy.
1265
Consequently, the whole ranges of the corresponding outputs of the frequency converter are as follows. U 27, 33
E 6, 5, 4, 3, 2, 1
U 25, 30
E 0
U 20, 30
E 1, 2, 3, 4, 5, 6
Figs.11 and 12 present the controlling results with the composite fuzzy DO control strategy. Comparing with Figs.5, 7 and 9 by ON/OFF, PID and fuzzy control methods, the result of composite fuzzy control is optimal. It makes the DO concentration fluctuate slowly around the set point with tiny overshoot. The reason for the good performance of composite fuzzy control strategy could be easily got from the comparisons of the manipulated frequency variation ranges with different controllers; the details are included in Table 2. In Table 2, the frequency variation range of composite fuzzy control is the smallest one in the four control strategies，that is the reason why the fluctuations of the DO concentration with it are the slowest one among the four control strategies, and its robustness and stability is the optimal. 2.8
50
Frequency (Hz)
DO (mg/L)
2.4 2.0 1.6 1.2 0.8 0.4 0.0 -20
0
20
7.8
40
60
80 100 120 140
20 10 0 0
20
0
20
300
40
60
80 100 120 140
40
60
80 100 120 140
Time(min)
240
ORP (mV)
7.6 7.5
pH
30
-20
Time(min)
7.7
7.4 7.3 7.2
AV
7.1 7.0 -20
40
0
20
40
60
80 100 120 140
Time(min)
180 120 60 0 -60 -20
Time(min)
TABLE 2 COMPARISON OF THE MANIPULATED FREQUENCY VARIATION RANGES WITH DIFFERENT CONTROLLERS Control strategy
ON/OFF control
PID control
Fuzzy control
Composite Fuzzy control
Frequency variation (Hz)
0, 42
0-50
14-33
21-33
As demonstrated in Table 3, the standard deviation of composite fuzzy control is 0.05, which is the smallest one in all of the four control strategies. The maximum DO is 2.14 mg/L which is the smallest one in four, and the minimum DO is 1.87 mg/L which is the largest one in four. These data indicates that the composite control strategy is the best one to control the DO concentration in SBR pilot process. Except for the controlling performance, the operational performance should also be considered. Although the controlling performance of composite fuzzy control is optimal, its consumed aeration volume is 1840 L, which is the largest one in four and implies that it is not an energy saving strategy. What is the reason for this result? We reckon that is the result of the increasing polluted load of treated wastewater. The experiments of control strategies were carried out in the order of ON/OFF, PID, fuzzy and composite fuzzy control strategy. Since the bacteria in the SBR needed a period of accustoming to the wastewater, more synthetic wastewater was added into the reactor prior to the implement of each control strategy. Therefore, with more and more wastewater adding into the reactor, the pollution level of the wastewater increased accordingly. It is evidenced that the initial ORP values of wastewater in four batches with four control strategies decrease gradually. As demonstrated in Figs. 4, 6, 8 and 11, the corresponding initial ORP values of wastewater with ON/OFF, PID, fuzzy and composite fuzzy control strategies are about 140 mV, 70 mV, 30 mV and -60 mV individually, which indicates the pollution level of wastewater was rising gradually. That is the main element to explain the above phenomenon.
2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5
TABLE 3 STATISTICAL RESULTS FOR DO CONTROL
50 Setpoint DO Aeration
45
Control result
40 35 30 25 20 15 10
Frequency (Hz)
DO (mg/L)
Fig. 11. Controlling results with composite fuzzy control strategy in the aeration phase of synthetic wastewater treatment process.
Rising time (min)
Items
5 40
45
50
Time (min)
55
60
0
Fig. 12. Enlarged part of the controlling results of DO concentration with composite fuzzy control strategy.
Control algorithms Composite ON/OFF PID Fuzzy fuzzy 4.2 3.3 4.9 3.5
Overshoot (%)
14
23
1
5.5
Mean (mg/L)
2.10
2.11
2.00
2.01
Max (mg/L)
2.67
2.54
2.74
2.14
Min (mg/L)
1.59
1.69
1.62
1.87
Std
0.24
0.22
0.20
0.05
Aeration volume (L)
1181
1194 1697
1840
According to Li [18], when the ammonia depletion is achieved, a minimum point will appear in the pH profile, the point is called ammonia valley (AV). Watching the AV
1266
signs in Figs. 4, 6, 8 and 11, the locations of AV with ON/OFF, PID and fuzzy control strategies are almost the same, about 70 minutes; while that with the composite fuzzy control is about 50 minutes, which is much earlier than the other control strategies. In other words, the ability of removal COD with composite fuzzy control is much better than that with other three kinds of control strategies. IV. CONCLUSIONS This paper presents an effective and robust control strategy to control the DO concentration in a pilot SBR wastewater treatment process. Constructing with Siemens PLC S7-200, necessary sensors and KingView 6.53 software, the data sampling and processing system is set up; developing with Matlab, a new composite fuzzy control strategy is studied. For the purpose of comparison, four control strategies are tested and evaluated: ON/OFF, PID, fuzzy and composite fuzzy control strategies. Their controlling results demonstrate that the new developed composite fuzzy control strategy is proved to be robust and effective. The controlled DO concentration was around the set point with 0.05 standard deviation and 3.5 minutes rising time, but the good controlling performance is obtained at the price of 1840 L aeration volume which is relatively high for the heavy pollution load of wastewater. It takes only 50 minutes which was less than that of the other strategies to arrive the ammonia valley; this shows its great ability of COD removal. The algorithm was not affected by the probe delay or the process strong nonlinearity. The obtained results showed the effectiveness of composite fuzzy controller and its superiority compared with the traditional regulation techniques and only with fuzzy control. It is the best control strategy in the provided four to satisfy the pilot synthetic papermaking wastewater treatment process.
[9] T.R. Girija and C. Mahanta, Fuzzy logic and artificial neural network approaches for dissolved oxygen prediction. International Journal of Environment and Waste Management, vol. 6, 2010, pp. 237 - 254. [10] M.Z. Jia, B.B. Gao, and B. Jia, The expert controller for dissolved oxygen control system in wastewater treatment. Communications in Computer and Information Science, vol. 237, 2011, pp. 351-356. [11] C. Yoo and M.H. Kim, Industrial experience of process identification and set-point decision algorithm in a full-scale treatment plant. Journal of Environmental Management, vol. 90 , 2009, pp. 28232830. [12] S. Grieu, A. Traoré, M. Polit, and J. Colprim, Prediction of parameters characterizing the state of a pollution removal biologic process. Eng. Appl. Artif. Intell, vol. 18, 2005, pp. 559–573. [13] C.L. Wang, Reducing the environmental impact of a paper mill’s water effluent. China Pulp & Paper, vol. 30 (S), 2011, pp. 156-160. [14] L.A. Zadeh, Fuzzy Sets. Information and Control, vol. 8, 1965, pp. 338-353. [15] A. Traoré, S. Grieu, S. Puig, L. Corominas, F. Thiery, M. Polit, and J. Colprim, Fuzzy control of dissolved oxygen in a sequencing batch reactor pilot plant. Chemical Engineering Journal, vol. 111, 2005, pp. 13–19. [16] T.L. Liu, L. Ning, and W. H. Shen, A quick way to get fuzzy control query table and its application in papermaking wastewater treatment process. Paper Science & Technology, vol. 30, 2011, pp. 78-82. [17] S.B. Gu, S.Y. Wang, Q. Yang, and Y.Z. Peng, Research progress on intelligent control strategies of SBR process. Industrial Water & Wastewater, vol. 40, 2009, pp. 1-6. [18] L.Y. Li, Y.Z. Peng, L. Li, and S.Y. Wang, On-line determination and variations of specific oxygen uptake rate in activated sludge system. Journal of Chemical Industry and Engineering (China), vol.61, 2010, pp. 995-1000.
REFERENCES [1] W. H. Shen and L. Ning, The study of the development situation of indstrial wastewater treatment automatic motoring technology. Paper Science & Technology, vol. 30, 2011, pp. 83-91. [2] W. Liu, Progress and trends in treatment technologies of pulp and paper mill effluents. China Pulp & Paper, vol. 30 (S), 2011, pp. 156160. [3] S.Y. Wang, S.B. Gu, Q. Yang, and Y.Z. Peng, Research progress on real-time control strategies for sequencing batch reactors. Acta Scientiae Circumstantiae, vol. 29, 2009, pp. 1121-1130. [4] S. Puig, L. Corominas, A. Traoré, J. Colomer, M.D. Balaguer, and J. Colgrim, An on-line optimization of a SBR cycle for carbon and nitrogen removal based on on-line pH and OUR: the dissolved oxygen control role. Proceedings of the Second IWA Conference on Instrumentation, Control and Automation for Water and Wastewater Treatment and Transport System. 2005. [5] Z.W. Wang, Y.Z. Peng, S.Y. Wang, J.H. Guo, H.J Huang, and Y.N. Zhen, Feasibility study for limited filamentous bulking under low dissolved oxygen at different operation regimes. Environmental Science, vol. 32, 2011, pp. 2347-2352. [6] Y. Ma, Y.Z. Peng, S.Y. Wang, and Y. Bai, DO control strategies and its fuzzy controller design in A/O process. Modern Chemical Industry, vol. 24 (S1): 2004, pp. 154-157. [7] W.J. Li and Y.W Ma,. Study on designing fuzzy controller for DO in A/O Process. Measurement & Control, vol. 26, 2010, pp. 54-56. [8] L.P. Fan and Boshnakov. Kosta, Fuzzy logic based dissolved oxygen control for SBR wastewater treatment process. Proceedings of the 8th World Congress on Intelligent Control and Automation, Jinan, China, July. 2010.
1267
