Whole-process cooperative control system of AO sewage treatment process

Abstract

The invention discloses a whole-process cooperative control system for an AO sewage treatment process, and the system comprises the following steps: 1, predicting the water quality of a whole process, obtaining the water quality data of inlet water through an inlet water quality monitoring instrument, and combining the observation quantity of real intermediate production; 2, performing intermediate observation quantity deduction, and based on the constructed whole-process water quality prediction model, using a model prediction control MPC algorithm to search for optimal process parameters capable of meeting an effluent water quality target under different external conditions; step 3, sub-process intelligent control: taking the deduced intermediate observation quantity as a target of each sub-process control, and obtaining the action of each sub-process equipment by using a traditional control algorithm or a reinforcement learning control algorithm; 4, the sewage treatment plant operates intelligently, the aeration fan controls the aeration amount by changing the power, the reflux ratio of internal reflux and external reflux is determined through the operation frequency of the pump set, the carbon source feeding pump and the chemical phosphorus removal pump control the dosage for cooperative control, and intelligent operation of the sewage treatment plant is achieved.

Description

Technical Field

[0001] The present invention relates to an A²O sewage treatment process control system, and particularly to a full-process collaborative control system for an A²O sewage treatment process. Background Art

[0002] Traditional sewage treatment processes are facing more and more challenges, including water quality fluctuations, high operating costs, energy waste, and insufficient system stability. In this context, the concept of a full-process collaborative system for sewage treatment has gradually emerged, aiming to achieve global optimization of the system, improve sewage treatment efficiency, and reduce resource consumption by integrating the operation control of each subprocess (such as aeration, internal reflux, external reflux, carbon source dosing, chemical phosphorus removal).

[0003] However, the main problem with the current process control methods in sewage treatment plants is that they are based on single processes for control. Although there are obvious effects on single processes, the overall treatment effect is not obvious and it cannot adapt to the fluctuations of influent water quality. Summary of the Invention

[0004] To solve the above problems, the present invention discloses a full-process collaborative control system for an A²O sewage treatment process, which can effectively solve the above two problems. By coordinating the control of each subprocess (aeration, internal reflux, external reflux, carbon source dosing, chemical phosphorus removal), it can improve the overall treatment effect, reduce resource consumption, and enhance the system's adaptability to water quality fluctuations.

[0005] A full-process collaborative control system for an A²O sewage treatment process includes the following steps: Step 1: Full-process water quality prediction. Obtain influent water quality data through influent water quality monitoring instruments, including influent COD, total phosphorus in influent, total nitrogen in influent, and ammonia nitrogen in influent; combine the observed values in actual intermediate production, including DO, NO3, MLSS in the aerobic zone, NO3 in the anoxic zone, and phosphate radical in the high-efficiency tank, to predict the effluent water quality in the next 12 hours or 24 hours. Step 2: Deduction of intermediate observed values. Based on the constructed full-process water quality prediction model, use the model predictive control (MPC) algorithm to search for the optimal process parameters that can meet the effluent water quality target under different external conditions. Step 3: Intelligent control of subprocesses. Take the deduced intermediate observed values as the control targets of each subprocess, and use traditional control algorithms or reinforcement learning control algorithms to obtain the actions of the equipment of each subprocess. Step 4: Intelligent operation of the sewage treatment plant. The aeration blower controls the aeration volume by changing the power, the internal reflux and external reflux determine the reflux ratio through the operating frequency of the pump group, and the carbon source dosing pump and chemical phosphorus removal pump control the dosing amount to achieve collaborative control and the intelligent operation of the sewage treatment plant.

[0006] Further, in step 1, the input time series data is first resampled to the hourly level, and a full - process water quality prediction model is built using the Transformer model.

[0007] Further, the model predictive control (MPC) algorithm is used to handle the global optimization of nonlinear, highly time - delayed, and multivariable systems, ensuring the stability of the effluent water quality; the MPC algorithm is used to search for intermediate observation variables that meet the effluent water quality target under different external conditions.

[0008] Further, the intelligent control of the sub - process in step 3 includes aeration, internal reflux, external reflux, carbon source dosing, and chemical phosphorus removal, and the overall optimum is achieved through coordinated control of each sub - process.

[0009] Further, in step 4, the aeration blower controls the aeration volume by changing the power, the internal reflux and external reflux determine the reflux ratio through the operating frequency of the pump group, and the carbon source dosing pump and chemical phosphorus removal pump optimize the sewage treatment process by controlling the dosing amount.

[0010] The response relationship between the power of the aeration blower and the aeration volume is that generally, the greater the power, the greater the aeration volume, and the dissolved oxygen (DO) in the biochemical tank rises. The greater the frequency of the internal and external reflux pump groups, the greater the sludge reflux pump. The greater the frequency of the carbon source dosing pump, the greater the dosing amount. The greater the frequency of the chemical phosphorus removal pump, the greater the dosing amount of the chemical agent.

[0011] Further, the system reduces manual intervention, operation complexity, and the risk of human error by real - time monitoring and dynamically optimizing operation parameters.

[0012] Further, the system can dynamically adjust operation parameters according to the fluctuations of the influent water quality, ensuring that the effluent water quality meets the standards and reducing energy consumption and chemical agent costs.

[0013] Further, the traditional control algorithms include PID, MPC, and LQR to obtain the actions of the equipment in each sub - process.

[0014] Advantages of the present invention: The present invention conducts coordinated control on each sub - process of sewage treatment. First, it enables the system to adapt to different fluctuations in influent water quality. Second, it reduces the costs of chemical dosing, electricity consumption, and labor in the sewage treatment plant. Third, it improves the overall treatment effect of the system and ensures that the effluent water quality meets the standards. Description of the Drawings

[0015] Figure 1 The full - process coordinated control system for sewage of the present invention. Detailed Implementation Modes

[0016] The present invention will be further illustrated below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. It should be noted that the terms "front", "rear", "left", "right", "upper" and "lower" used in the following description refer to the directions in the accompanying drawings, and the terms "inner" and "outer" respectively refer to the directions towards or away from the geometric center of a specific component.

[0017] As Figure 1 shown, a full-process collaborative control system for an A²O sewage treatment process in this embodiment includes the following parts: 1. Full-process water quality prediction: Obtain the influent water quality data through the influent water quality monitoring instrument, including influent COD, total phosphorus in the influent, total nitrogen in the influent, and ammonia nitrogen in the influent), and combine the observed values in the actual intermediate production (DO, NO3, MLSS in the aerobic zone, NO3 in the anoxic zone, phosphate radical in the high-efficiency tank) to predict the effluent water quality in the next 12 hours / 24 hours.

[0018] First, resample the input time series data to the hourly level, and then use the Transformer model to build a full-process water quality prediction model.

[0019] 2. Deduction of intermediate observed values Based on the constructed full-process water quality prediction model, the MPC (Model Predictive Control) can be used to search for the peak values of the optimal process parameters of the intermediate process observations that can meet the effluent water quality target under different external conditions. The MPC control algorithm is an optimization-based feedforward control method suitable for complex dynamic systems. Its logic is based on model prediction and can better handle the global optimization of nonlinear, high-time-delay, and multi-variable systems. The intermediate process variables include DO, NO3, MLSS in the aerobic zone of the biochemical tank, NO3 in the anoxic zone of the biochemical tank, and phosphate radical in the high-efficiency tank.

[0020] 3. Intelligent control of subprocesses The intelligent control of subprocesses mainly includes five aspects (aeration, internal reflux, external reflux, carbon source addition, and chemical phosphorus removal). The deduced intermediate observed values are used as the control targets of each subprocess, and traditional control algorithms (PID, MPC, LQR) or reinforcement learning control algorithms are used to obtain the actions of the equipment in each subprocess. The intelligent control of each subprocess works together to ensure that while meeting the effluent water quality target, the chemical dosage and power consumption are saved, and the overall effect of the entire sewage treatment system reaches the optimal.

[0021] 4. Intelligent operation of sewage treatment plants The aeration blower controls the aeration volume by changing the power. The internal reflux and external reflux determine the reflux ratio through the operating frequency of the pump group. The carbon source dosing pump and chemical phosphorus removal pump control the dosing amount. After these actions are applied to an actual sewage treatment plant, actual intermediate observation values and actual effluent water quality are obtained.

[0022] Specifically, the target value of the intermediate observation value can be obtained through Steps 1 and 2. With the target value, each process section can use a control algorithm to make the current observation value approach the target value.

[0023] Suppose it is known through Steps 1 and 2 that the target value of the dissolved oxygen in the biochemical tank is 2, while the current dissolved oxygen is 1. Therefore, the power of the aeration blower needs to be increased to increase the aeration volume.

[0024] Through the above steps, the sewage full-process collaborative control system basically forms a closed-loop system. This system can not only predict the future effluent, but also coordinate each sub-process (aeration, internal reflux, external reflux, carbon source dosing, chemical phosphorus removal) to operate smoothly in the most effective and economical way according to the specific effluent target.

[0025] As shown in the following table, the comparison table of the contents of each part of this embodiment: The collaborative control in this embodiment means that each process section of the sewage affects the effluent water quality, and each process section needs to be considered. For example, to make the effluent water quality meet the standard Time 1: The dissolved oxygen in the biochemical tank is 2, and the nitrate nitrogen in the anoxic tank is 3 However, due to changes in the influent or other intermediate observation values Time 2: To make the effluent water quality meet the standard, at this time, the dissolved oxygen in the biochemical tank is 1.5, and the nitrate nitrogen in the anoxic tank is 2 In this way, while ensuring that the effluent water quality meets the standard, it adapts to different influent water quality fluctuations. The aeration and dosing amount are precisely controlled, saving electricity and dosing costs. Since the overall is intelligent control, the labor cost expenditure is also reduced. This embodiment conducts collaborative control on each sub-process of sewage treatment. First, it enables the system to adapt to different influent water quality fluctuations. Second, it reduces the costs of dosing, electricity, and labor in the sewage treatment plant. Third, it improves the overall treatment effect of the system and ensures that the effluent water quality meets the standard.

[0026] The technical means disclosed in the solution of the present invention are not limited to the technical means disclosed in the above embodiments, but also include technical solutions composed of any combination of the above technical features.

Claims

1. A full-process collaborative control system for an A²O sewage treatment process, characterized in that, Including the following steps: Step 1: Full-process water quality prediction. Obtain the influent water quality data through the influent water quality monitoring instruments, including influent COD, influent total phosphorus, influent total nitrogen, and influent ammonia nitrogen. Combine the observed values in the actual intermediate production, including DO, NO3, MLSS in the aerobic zone, NO3 in the anoxic zone, and phosphate radicals in the high-efficiency tank, to predict the effluent water quality in the next 12 hours or 24 hours. Step 2: Deduction of intermediate observed values. Based on the constructed full-process water quality prediction model, use the model predictive control (MPC) algorithm to search for the optimal process parameters that can meet the effluent water quality target under different external conditions. Step 3: Intelligent control of subprocesses. Use the deduced intermediate observed values as the control targets for each subprocess, and use traditional control algorithms or reinforcement learning control algorithms to obtain the actions of each subprocess device. Step 4: Intelligent operation of the sewage treatment plant. The aeration fan controls the aeration volume by changing the power, the internal and external recirculation determine the recirculation ratio through the operating frequency of the pump group, and the carbon source dosing pump and chemical phosphorus removal pump control the dosing amount to achieve coordinated control and realize the intelligent operation of the sewage treatment plant.

2. The full-process collaborative control system of an A²O sewage treatment process according to claim 1, characterized in that: In the said Step 1, first resample the input time series data to the hourly level, and build a full-process water quality prediction model using the Transformer model.

3. The full-process collaborative control system of an A²O sewage treatment process according to claim 1, characterized in that: The said model predictive control (MPC) algorithm is used to handle the global optimization of nonlinear, high-delay, and multivariable systems to ensure the stability of the effluent water quality.

4. The full-process collaborative control system of an A²O sewage treatment process according to claim 1, wherein: The intelligent control of subprocesses in the said Step 3 includes aeration, internal recirculation, external recirculation, carbon source dosing, and chemical phosphorus removal. Each subprocess achieves overall optimization through coordinated control.

5. The full-process collaborative control system of an A²O sewage treatment process according to claim 1, characterized in that: In the said Step 4, the aeration fan controls the aeration volume by changing the power, the internal and external recirculation determine the recirculation ratio through the operating frequency of the pump group, and the carbon source dosing pump and chemical phosphorus removal pump optimize the sewage treatment process by controlling the dosing amount.

6. The full-process collaborative control system of an A²O sewage treatment process according to claim 1, characterized in that: The said system reduces manual intervention, operation complexity, and the risk of human error through real-time monitoring and dynamic optimization of operation parameters.

7. The full-process collaborative control system for an A²O sewage treatment process according to claim 1, characterized in that, The said system can dynamically adjust operation parameters according to the influent water quality fluctuations to ensure that the effluent water quality meets the standards and reduce energy consumption and chemical costs.

8. The full-process collaborative control system of an A²O sewage treatment process according to claim 1, characterized in that, The said traditional control algorithms include PID, MPC, and LQR to obtain the actions of each subprocess device.

Images