Abstract: Based on the SBR method, this paper studies the adaptive fuzzy control of dissolved oxygen (DO) concentration in wastewater treatment. Based on the control experience obtained by manually adjusting the aeration rate to control DO under actual operating conditions, fuzzy rules are summarized. Fuzzy control algorithms are embedded in the host computer to achieve fuzzy control of DO. Experimental data show that the adaptive fuzzy control system has ideal DO control effect and strong anti-disturbance ability, thus improving the efficiency of wastewater treatment. Keywords: Adaptive fuzzy control; Dissolved oxygen concentration (DO); SBR The activated sludge process is one of the main processes for urban wastewater treatment. Its mechanism is that aeration allows activated sludge to obtain sufficient oxygen in full contact with wastewater. Soluble organic pollutants in the water are adsorbed by the activated sludge and decomposed by microorganisms living on the activated sludge, thus purifying the wastewater. Because the internal mechanism of wastewater treatment is very complex and cannot be described by a precise mathematical model, traditional control strategies (such as typical PID control) are difficult to achieve satisfactory control results. As an important branch of intelligent control, fuzzy control does not rely on the precise mathematical model of the controlled object. It can control the controlled variable according to the error and the change of the error. It has strong robustness. The change of the controlled object parameter has little effect on fuzzy control. It can be used for the control of nonlinear, time-varying and time-delay systems. The control has good real-time performance. The control mechanism is in line with people's intuitive description and thinking logic of process control [1]. At the same time, the use of adjustable fuzzy control rules in the design of fuzzy controller can significantly improve the stability and adaptability of fuzzy controller. Peng Yongzhen et al. [2] proposed to use DO as the fuzzy control parameter of SBR method to realize the control of aeration volume. Wang Xianlu et al. [3] designed a fuzzy controller with the effluent COD deviation and deviation change as input and the pump opening as output to realize the fuzzy control of COD. However, there is a major problem in the sewage treatment process using fuzzy control system, which is the stability and adaptability of the control system. Due to the PD control effect of fuzzy control and the continuous change of water environment, the control effect of most systems is obvious, and the stability and adaptability are not good. This paper, based on the Sequencing Batch Reactor (SBR) process, investigates the impact of dissolved oxygen (DO) concentration variations on the wastewater treatment process. An adaptive fuzzy control system for DO, incorporating an adjustment factor, is designed. By adjusting the blower frequency, the DO is stably controlled at an ideal level, improving wastewater treatment efficiency and demonstrating good stability and adaptability. 1. Introduction to SBR and Control Strategy The Sequencing Batch Reactor (SBR) is an intermittent biological wastewater treatment process that has experienced rapid development since the 1980s and is well-suited for small and medium-sized wastewater treatment plants. Its operation includes five stages: influent, reaction (aeration), sedimentation, decanting, and idle. Compared to the activated sludge process, it eliminates the need for primary and secondary sedimentation tanks; all five stages occur within the same reaction tank. Currently, the primary control method for SBRs is time-programmed control. The five stages of the treatment process proceed sequentially according to a pre-set time sequence. Automatic control can be easily achieved using a programmable logic controller (PLC), which is one of the reasons for its widespread application. In the SBR wastewater treatment process, the biological oxidation stage (aeration stage) is the core part. In this stage, microorganisms, mainly aerobic bacteria, treat organic pollutants in wastewater through biochemical reactions. One of the key factors determining the treatment effect is the dissolved oxygen concentration (DO) in the biological tank [4]. Since the raw water quality is often constantly changing and may fluctuate drastically under certain conditions, the traditional time-program control method has significant drawbacks. Long aeration time or large aeration volume will result in a large waste of energy, while short aeration time or small aeration volume may cause large fluctuations in effluent quality or even failure to meet standards. Therefore, using fixed time and fixed air volume for aeration is out of sync with the actual process of wastewater reaction. According to the research of domestic scholars, the treatment effect of activated sludge is most ideal when DO is maintained at around 2 mg/L [5]. Moreover, using DO value as the fuzzy control parameter of SBR method can save as much operating cost as possible while ensuring effluent quality, and can avoid sludge bulking caused by insufficient aeration volume or excessive reaction time. Therefore, how to control DO under ideal conditions becomes the key to improving treatment efficiency. 2. DO Adaptive Fuzzy Control System The fan speed in this system is regulated by a frequency converter. Its control principle is as follows: First, the setpoint and the detected value are compared to obtain precise quantities E and EC. These are then fuzzified into fuzzy quantities. Next, based on a fuzzy knowledge base derived from extensive experimental data and expert experience, the fuzzy input quantities are used for fuzzy inference to obtain the corresponding fuzzy control quantities. Finally, the fuzzy control quantities are converted into precise control outputs through fuzzy decision-making, thereby controlling the aeration rate and adjusting the DO concentration in the tank. The control flowchart is shown in Figure 1. Figure 1 DO Fuzzy Control System Flowchart 2.1 Fuzzy Controller Structure Selection The so-called fuzzy controller structure selection refers to determining the input and output variables of the fuzzy controller. The structure of the fuzzy controller has a significant impact on the performance of the entire system and must be rationally selected based on the specific circumstances of the controlled object. The main structures of fuzzy controllers are single-input single-output (SISO) and multiple-input multiple-output (MIMO) structures. Based on the actual wastewater treatment process, this system adopts a typical dual-input single-output two-dimensional fuzzy controller. The input variables are the DO deviation E and the deviation change rate EC, respectively, and the output variable U is the frequency VRI of the variable frequency fan. 2.2 Determination and Fuzzification of Fuzzy Linguistic Variables and Domain of Discourse Fuzzy rules are fuzzy conditional statements composed of several linguistic variables, reflecting certain human thought processes. When determining fuzzy variables, their basic linguistic values ​​are first determined, and then several linguistic sub-values ​​are generated as needed. Generally, the more linguistic values ​​a linguistic variable has, the more accurate the description of things, and the better the control effect may be. However, overly fine divisions can complicate the control rules and make implementation more difficult; therefore, the determination should be based on specific circumstances. In this system, the fuzzy subsets of E, EC, and U are defined as follows: E = EC = U = {Negative Large (NB), Negative Medium (NM), Negative Small (NS), Zero (ZO), Positive Small (PS), Positive Medium (OM), Positive Large (PB)}. The basic universe of discourse for E is (-0.6, 0.6), and the linguistic variables are {-6, -5, -4, -3, -2, -1, -0, 0, 1, 2, 3, 4, 5, 6}. The basic universe of discourse for EC is (-0.15, 0.15), and the linguistic variables are {-6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6}. The basic universe of discourse for U is (-4, 4), and the linguistic variables are {-7, -6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6, 7}. 2.3 Determination of Fuzzy Control Rules and Establishment of Fuzzy Control Table In a fuzzy control system, the performance of the fuzzy controller greatly affects the system's control characteristics, and the performance of the fuzzy controller largely depends on the establishment and adjustability of the fuzzy control rules. We adopted an empirical inductive method, summarizing forty-nine rules based on the experience of manually adjusting the fan frequency to change the DO, in the form of "IF E= ┅ and IF EC= ┅ then U=┅". All control rule bases are shown in Table 1. Table 1 Fuzzy Control Rule Table Based on the fuzzy control rule table, the input variables were defuzzified using the maximum membership method. The fuzzy controller output for each state was obtained through offline calculation, resulting in the fuzzy control table shown in Table 2. Table 2 Fuzzy Control Table Conventional Two-Dimensional Fuzzy Control In this conventional two-dimensional fuzzy control, the output variable value is determined by the input quantities E and EC, and their weight coefficients are each 0.5. Once the design is completed, the control rules are determined. However, in the water treatment process, water quality is constantly changing, which is obviously detrimental to the stability of the control system. Therefore, we introduced an adjustment factor to adjust the control rules to enable them to adapt to changing water environments. The control rule with adjustment factors can be expressed as: u = -[αE + (1 - α)C], 0 < α < 1. α in the formula is the adjustment factor, also called the weighting factor. We adjust the value of α based on the COD value measured by the online monitoring instrument. The range of α (0, 1) corresponds linearly to the COD range (0, 1000) (mg/L). For every 100 mg/L change in COD, α changes by 0.1. Thus, the system adaptively changes the weighting degree of error E and error change EC. 3. Fuzzy Control System Operation Results 3.1 Experimental System Apparatus This experimental system is a small-scale SBR wastewater treatment system, divided into upper-level and lower-level computer parts. The lower-level computer uses PLC control, while the upper-level computer is an industrial control computer and KingSCADA, with a good human-machine interface. The adaptive fuzzy control algorithm is embedded in the script program of KingSCADA. The SBR tank of this treatment system is 1.8M deep, 0.8M wide, and 1.5M long, with an effective volume of 2M3. Aeration was achieved using blowers, with the air distribution system employing Φ70*500 microporous aeration heads. Oxygen utilization was 18-28%; air resistance was ≤150 mm water column; dissolved oxygen could be controlled within the range of 0.5 mg/L to 10.0 mg/L. The system operated for 2 hours per cycle: 10 minutes of influent, 1 hour of aeration, 30 minutes of sedimentation, 20 minutes of decanting, and 30 minutes of idle time. Each cycle discharged 1 m³, closely mimicking the actual engineering environment. 3.2 Experimental Results and Analysis: Wastewater from the university's residential area was used in the experiment. A certain amount of glucose was added to adjust the required COD concentration. The influent COD was 860 mg/L, BOD was 620 mg/L, MLSS sludge concentration was 4200 mg/L, DO was set at 3 mg/L, and the blower frequency was adjusted every 60 seconds. To test the adaptive capability of our designed fuzzy control system, at the 40-minute mark of operation, we added glucose equivalent to 300 mg/L COD to the tank to simulate a sudden change in water quality. The results are shown in Figure 2: the effluent COD was 37 mg/L, BOD was 29 mg/L, the COD removal rate reached 95.7%, and the BOD removal rate reached 95.3%. From the figure, we can see that under load changes, the control system can adapt well to water quality changes, the process is stable, and it has good adaptability and shock resistance. Figure 2: Adaptive Fuzzy Control for DO. 4. Conclusion The use of intelligent control methods, including fuzzy control, for wastewater treatment processes has become an industry consensus. However, to date, there are still few intelligent control technologies truly applied in actual production. The main reason is that the huge changes in the wastewater environment place high demands on the stability and adaptability of the control system, and many factors affect control performance, making it difficult to comprehensively consider all factors. This paper addresses this situation and conducts an in-depth study on how to improve the stability and adaptability of the control system. It proposes an adaptive fuzzy control system with an adjustment factor that integrates various influencing factors. After experimental verification, the system has good stability and adaptability and is suitable for practical engineering applications. In addition, the establishment of the DO adaptive fuzzy control system should be targeted. The establishment of the rule table and the setting of time parameters are not completely the same for different working conditions. It is related to the concentration of organic matter in the influent, the concentration of sludge, the hydraulic load and the size of the reaction tank. The relationship between the determination of the adjustment factor and various influencing factors needs further research. References: [1] Feng Dongqing, Xie Songhe. Fuzzy Intelligent Control [M]. Beijing: Chemical Industry Press, 2000. 75-93. [2] Peng Yongzhen et al. Research, application and development of intelligent control for sewage treatment [J]. China Water & Wastewater, 2002, 6 [3] Wang Xianlu et al. Research on fuzzy control of COD in sewage treatment [J]. Technology Review, 2002, 2 [4] Zhang Zijie, Lin Rongchen, Jin Rulin. Biological treatment of wastewater [M], Drainage Engineering, 2000. [5] Zeng Wei, Peng Yongzhen, Zhang Dongli, et al. Fuzzy control of aeration volume in SBR method [J], Journal of Harbin University of Architecture, 2002, 35(1), 53-57.