Carbon source feeding device for sewage treatment

Abstract

The utility model relates to a sewage treatment carbon source adding device which comprises a facultative tank for pretreating sewage, a main tank body MBR (membrane bioreactor) tank for treating the sewage and a sludge storage tank for storing treated sludge, the facultative tank 1 is positioned on one side of a system, and the sludge storage tank is positioned on the other side of the system. The MBR tank is located between the facultative tank and the sludge storage tank, the MBR tank further comprises a water inlet, a first lifting pump and a controller, the controller is used for controlling and predicting COD (Chemical Oxygen Demand) and ammonia nitrogen value of effluent, the water inlet is located above the facultative tank, and sewage enters the MBR tank through the water inlet; the first lifting pump is arranged at the bottom of the MBR tank, a gap is formed between the first lifting pump and the bottom of the MBR tank, and the first lifting pump is used for lifting sewage to the facultative tank, so that the sewage treatment efficiency is improved, the sewage treatment device has the advantages of high response speed, high adaptive capacity and good expandability, and the stability of effluent quality is remarkably improved.

Description

Technical Field

[0001] This utility model relates to the field of wastewater treatment, and in particular to a wastewater treatment carbon source dosing device. Background Technology

[0002] With the acceleration of industrialization and urbanization, urban wastewater treatment faces increasing challenges. Traditional biological treatment methods, especially when treating low-carbon, high-nitrogen wastewater, often suffer from low nitrogen removal efficiency due to insufficient carbon sources. To address this issue, carbon source addition technology has emerged, which improves biological nitrogen removal efficiency by supplementing the wastewater treatment system with carbon sources. However, excessive carbon source addition can lead to increased costs and potentially cause COD (chemical oxygen demand) levels in the effluent to exceed standards. Therefore, how to accurately control the amount of carbon source added while ensuring nitrogen removal efficiency has become a key research direction in the field of wastewater treatment technology.

[0003] In existing technologies, carbon source dosing devices mainly rely on traditional timed or quantitative dosing methods. The dosing strategy of these devices is relatively simple, typically based on a fixed dosage set empirically or adjusted according to a single parameter such as flow rate or pH value. The structure or operating steps of these methods generally include: firstly, monitoring the influent water quality parameters using a flow meter or sensor, and then manually or automatically controlling the carbon source dosing according to the set dosage. However, this approach has significant drawbacks:

[0004] 1. Insufficient real-time performance: Existing devices usually adopt a fixed dosing mode, which cannot be dynamically adjusted according to the real-time changes in the influent water quality, which can easily lead to insufficient or excessive dosing.

[0005] 2. Lack of precise control: Most existing carbon source dosing equipment is based on simple feedback loops, which makes it difficult to handle the coupling relationship between multiple complex parameters and to accurately control the amount of carbon source added, especially when water quality parameters fluctuate greatly.

[0006] 3. High energy consumption and cost: Due to the lack of intelligent adjustment methods, excessive addition of carbon sources will not only cause material waste, but also increase the difficulty of subsequent treatment, leading to increased energy consumption and operating costs.

[0007] 4. Response lag: Due to the system's adjustment lag and insufficient response capability to complex systems, existing technologies may struggle to achieve precise and efficient control objectives in practical applications.

[0008] In contrast, this invention aims to propose an intelligent carbon source dosing device by combining neural network technology. By introducing a deep learning control algorithm, the carbon source dosage can be adjusted in real time based on multiple parameters such as COD, ammonia nitrogen, pH, water flow rate, and dissolved oxygen of the influent, thereby achieving precise control, improving wastewater treatment efficiency, and reducing operating costs. Utility Model Content

[0009] To address the aforementioned problems, the present invention aims to provide a carbon source dosing device for wastewater treatment. By combining deep learning and automated control technologies, it can adjust the carbon source dosing amount in real time and accurately, thereby achieving intelligent control of the carbon source during wastewater treatment.

[0010] To achieve the above objectives, this utility model adopts the following technical solution: a wastewater treatment carbon source dosing device, comprising an anoxic tank for pretreatment of wastewater, an MBR tank for main wastewater treatment, and a sludge storage tank for storing treated sludge. The anoxic tank is located on one side of the system, the sludge storage tank is located on the other side of the system, and the MBR tank is located between the anoxic tank and the sludge storage tank. The device also includes an inlet, a first lift pump, and a controller. The controller is used to control and predict the effluent COD and ammonia nitrogen values. The inlet is located above the anoxic tank, and wastewater enters the MBR tank through the inlet. The first lift pump is installed at the bottom of the MBR tank with a gap from the bottom, and is used to lift the wastewater to the anoxic tank.

[0011] Preferably, the MBR tank is equipped with a blower, an air valve, and an air supply pipe. The blower is connected to the air valve and the air supply pipe in sequence, and together they are used to deliver outside air into the device.

[0012] Preferably, the system also includes an aeration head and a biofilm reaction assembly. The aeration head is located at the bottom of the air supply pipe and is used for aeration in the wastewater. The biofilm reaction assembly is located above the aeration head and is used to support the biological treatment of the wastewater.

[0013] Preferably, the system also includes an online instrument, a self-priming pump, a regulating valve, a flow meter, and a clean water pipe. The clean water pipe is connected to the biofilm reaction assembly and leads to the outside world. The online instrument, the self-priming pump, the regulating valve, and the flow meter are sequentially mounted on the clean water pipe.

[0014] Preferably, the system also includes a second lift pump, a first conveying pipe, and a second conveying pipe. The second lift pump is located at the bottom of the MBR tank for conveying sludge. The first conveying pipe is connected to the first lift pump and the anoxic tank and the MBR tank for conveying wastewater. The second conveying pipe is connected to the second lift pump and the MBR tank for conveying sludge.

[0015] Preferably, the device also includes a fluorescent probe, an excitation light source, a fluorescence spectroscopy detection module, and a fluorescence detection chamber, wherein the fluorescence detection chamber is disposed on the upper part of the wall of the sludge storage tank; the excitation light source is disposed on one side of the fluorescence detection chamber, and the fluorescence spectroscopy detection module is disposed on the other side of the fluorescence detection chamber; the fluorescent probe extends from outside the device into the interior of the fluorescence detection chamber.

[0016] Preferably, the system also includes a diversion pipe and a pressure piston. One end of the diversion pipe is connected to one end of the second conveying pipe, and the other end of the diversion pipe extends to the top of the fluorescence detection chamber. The pressure piston is disposed on the diversion pipe and is used to squeeze the water out of the conveyed sludge and discharge it into the fluorescence detection chamber.

[0017] Preferably, the system also includes the fluorescence spectroscopy detection module, the data storage module, the neural network controller, and the computer. The output terminal of the online instrument is connected to the data storage module via a signal line. The fluorescence spectroscopy detection module outputs an electrical signal to the data storage module. The data storage module is used to integrate the data from both modules and transmit it to the neural network controller via a signal line. The signal output terminal of the neural network controller is electrically connected to the computer.

[0018] Preferably, the system also includes a carbon source inlet, a carbon source storage tank, and a carbon source delivery pipe. The carbon source inlet is located above the MBR tank for injecting carbon source; the carbon source storage tank is located below the carbon source inlet for storing unused carbon source; and the carbon source delivery pipe is located on the carbon source storage tank for delivering carbon source to the MBR tank.

[0019] Compared with the prior art, the wastewater treatment carbon source dosing device provided by this utility model has the following beneficial effects:

[0020] 1. Precise dynamic control

[0021] This invention employs a neural network model, capable of handling various complex input parameters, including influent COD, ammonia nitrogen concentration, pH value, water flow rate, dissolved oxygen, and MLSS. This multi-parameter input allows the system to dynamically adjust according to real-time water quality conditions, ensuring that the carbon source dosage is always at the optimal level, avoiding the inaccuracies caused by single or fixed parameter dosages in traditional technologies.

[0022] 2. Improve wastewater treatment efficiency

[0023] By leveraging the adaptive learning capabilities of neural networks, this invention can accurately predict effluent COD and ammonia nitrogen levels under varying water quality conditions and adjust the carbon source dosing strategy accordingly. This predictive control method effectively improves nitrogen and organic matter removal rates, thereby enhancing the overall efficiency of wastewater treatment. Compared to existing technologies that rely on experience or simple feedback control, this invention is better able to handle treatment environments with significant water quality fluctuations.

[0024] 3. Save carbon sources and reduce operating costs

[0025] Existing technologies often suffer from the problem of excessive or insufficient carbon source addition. Excessive addition leads to unnecessary cost increases, while insufficient addition affects treatment efficiency. This invention, through real-time monitoring and intelligent adjustment of influent water quality, can accurately calculate the required carbon source dosage, avoiding excessive use of carbon source and thus significantly reducing material costs. Furthermore, due to the improved treatment efficiency, the energy consumption and operating costs of the entire system are also effectively controlled.

[0026] 4. Fluorescence spectroscopy technique

[0027] Fluorescent probes are introduced to detect changes in the fluorescence signal of specific pollutants in water, enabling precise determination of pollutant concentrations. To improve the accuracy and real-time performance of detecting specific pollutants in wastewater, a fluorescence spectroscopy detection module is added to the existing carbon source dosing device for wastewater treatment. By introducing fluorescent probes, changes in the fluorescence signal of specific pollutants in water are detected, achieving precise determination of pollutant concentrations.

[0028] 5. Fast response speed and adaptive capability

[0029] Traditional carbon source dosing devices often suffer from response lag. This invention, however, by introducing a nonlinear long short-time memory network (NLSTM) and an attention mechanism, enables rapid response to changes in water quality parameters. This technology improves the system's real-time response speed, allowing for faster adjustments to the dosing strategy and ensuring stable effluent quality, making it suitable for scenarios with frequent water quality changes.

[0030] 6. Improve the stability of effluent water quality

[0031] By using neural network control, this invention not only improves the system's processing efficiency but also maintains stable effluent quality. Traditional systems may experience excessive COD or ammonia nitrogen levels in the effluent due to load fluctuations or improper operation. However, this invention, through self-learning and multi-parameter linkage control, ensures that the effluent quality meets standards under complex conditions, reducing the risk of secondary pollution.

[0032] 7. Highly scalable and widely adaptable

[0033] This invention features a flexible design, allowing it to be applied to wastewater treatment systems of different scales and types by adjusting the structure and input parameters of the neural network model. Whether it's industrial wastewater treatment, municipal sewage treatment, or other treatment scenarios with specific water quality requirements, this invention can achieve optimal results through appropriate optimization.

[0034] This invention overcomes the shortcomings of existing carbon source dosing devices, such as poor real-time performance, inaccurate adjustment, and carbon source waste, by introducing neural network control technology, thus significantly improving the efficiency and stability of wastewater treatment. Furthermore, this invention possesses excellent adaptability, enabling it to flexibly cope with various complex water quality conditions and effectively reduce operating costs, demonstrating significant technical advantages and broad application prospects. Attached Figure Description

[0035] To more clearly illustrate the embodiments of this utility model or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0036] The structures, proportions, sizes, etc. illustrated in this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed herein, and are not intended to limit the implementation conditions of this utility model. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportional relationships, or adjustments to the size, without affecting the effects and purposes that this utility model can produce, should still fall within the scope of the technical content disclosed in this utility model.

[0037] Figure 1 This is a schematic diagram of the front structure of this utility model;

[0038] Figure 2 This is a schematic diagram of fluorescence detection according to the present invention;

[0039] Figure 3 This is a three-dimensional schematic diagram of the device of this utility model;

[0040] Figure 4 This is a schematic diagram of the upper-level data processing of this utility model;

[0041] Figure 5 This is a schematic diagram of the neural network control algorithm in this utility model;

[0042] In the picture:

[0043] 1. Anoxic tank; 2. MBR tank; 3. Sludge storage tank; 4. Inlet; 5. First lift pump; 6. Blower; 7. Air valve; 8. Gas supply pipe; 9. Aeration head; 10. Biofilm reactor assembly; 11. Online instrument; 12. Self-priming pump; 13. Regulating valve; 14. Flow meter; 15. Clean water pipe; 16. Second lift pump; 17. First delivery pipe; 18. Second delivery pipe; 19. Carbon source dosing port; 20. Carbon source storage tank; 21. Carbon source delivery pipe; 22. Fluorescent probe; 23. Excitation light source; 24. Fluorescence spectroscopy detection module; 25. Diverter pipe; 26. Pressure piston; 27. Fluorescence detection chamber; 28. Data storage module; 29. ​​Neural network controller; 30. Computer. Detailed Implementation

[0044] The embodiments of this utility model are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this utility model, and should not be construed as limiting this utility model.

[0045] In the description of this utility model, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "clockwise", "counterclockwise", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this utility model and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this utility model.

[0046] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this utility model, unless otherwise stated, "a plurality of" means two or more, unless otherwise expressly defined.

[0047] In this utility model, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this utility model according to the specific circumstances.

[0048] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0049] The present invention will now be described in detail with reference to preferred embodiments.

[0050] like Figure 1 , Figure 2 As shown, an embodiment of this utility model proposes a wastewater treatment carbon source dosing device, including an anoxic tank 1, an MBR tank 2, and a sludge storage tank 3; the anoxic tank 1 is located on the left side of the system and pre-treats the wastewater; the MBR tank 2 is located in the middle of the system and is the main tank for wastewater treatment; the sludge storage tank 3 is located on the right side of the MBR tank and stores the treated sludge.

[0051] The inlet 4 is located above the anoxic tank 1, and the sewage enters the MBR tank 2 through this inlet; the first lift pump 5 is set at the bottom of the MBR tank 2 and at a certain distance from the bottom of the tank, and is used to lift the sewage to the anoxic tank 1.

[0052] Blower 6 is connected in sequence to air valve 7 and air supply pipe 8, and is set up inside the MBR tank 2. Together they work to deliver outside air into the device.

[0053] Aeration head 9 is located at the bottom of the air supply pipe 8 to aerate the wastewater; biofilm reaction assembly 10 is located above the aeration head 9 to support the biological treatment of wastewater.

[0054] The clean water pipe 15 is connected to the biofilm reaction assembly 10 and leads to the outside. The online instrument 11, self-priming pump 12, regulating valve 13, and flow meter 14 are sequentially installed on the clean water pipe 15.

[0055] The second lift pump 16 is installed at the bottom of the MBR tank 2 for transporting sludge; the first transport pipe 17 is installed on the first lift pump 5 and connects the anoxic tank 1 and the MBR tank 2 for transporting wastewater; the second transport pipe 18 is installed on the second lift pump 16 and connects the MBR tank 2 and the sludge storage tank 3 for transporting sludge.

[0056] A carbon source inlet 19 is located above the MBR tank 2 for injecting carbon source; a carbon source storage tank 20 is located below the carbon source inlet 19 for storing unused carbon source; and a carbon source delivery pipe 21 is located on the carbon source storage tank 20 to deliver carbon source to the MBR tank 2.

[0057] A fluorescence detection chamber 27 is disposed on the upper part of the wall of the sludge storage tank 3; an excitation light source 23 is disposed on one side of the fluorescence detection chamber 27, and a fluorescence spectral detection module 24 is disposed on the other side of the fluorescence detection chamber 27; a fluorescence probe 22 extends from the outside of the device into the interior of the fluorescence detection chamber 27; one end of a diversion pipe 25 is connected to one end of the second delivery pipe 18, and the other end extends into the upper part of the fluorescence detection chamber 27; a pressure piston 26 is disposed on the diversion pipe 25 to squeeze out the water in the delivered sludge and discharge it into the fluorescence detection chamber 27.

[0058] In one embodiment of this utility model, such as Figure 1 , Figure 3 As shown, the cement mixture after initial sedimentation is transported to the anaerobic tank 1 by the first lift pump 5. The anaerobic microorganisms eliminate organic matter in the wastewater, and the microorganisms in the upper and lower layers work together to reduce the footprint of the device. Then, the pre-treated wastewater enters the MBR tank 2 through the inlet 4. The aeration head 9 connected to the blower 6 in the MBR tank 2 is used for aeration to provide oxygen for the biofilm reaction component 10, so as to better react and treat the wastewater. The treated sludge and clean water are respectively transported to the sludge storage tank 3 and discharged to the outside through the clean water pipe 15.

[0059] In one embodiment of this utility model, such as Figure 1As shown, fluorescence spectroscopy is used to detect changes in the fluorescence signal of specific pollutants in water, accurately determining the pollutant concentration. To improve the detection accuracy and real-time performance of specific pollutants in wastewater, the excitation light source 23, fluorescence probe 22, and fluorescence spectroscopy detection module 24 need to be calibrated to ensure normal system operation. An appropriate amount of fluorescence probe reagent is added to the fluorescence detection chamber 27 to react with the target pollutant in the wastewater, ensuring sufficient reaction time for the probe and pollutant to generate a stable fluorescence signal. The excitation light source 23 emits excitation light of a specific wavelength, illuminating the mixed liquid in the sample cell. After reacting with the target pollutant, the fluorescence probe 22 is excited by the excitation light and emits fluorescence of a specific wavelength. The fluorescence signal is transmitted to the fluorescence spectroscopy detection module 24. The fluorescence spectroscopy detection module 24 converts the received fluorescence signal into an electrical signal, which is amplified and filtered before being acquired by the data acquisition module. The data processing unit analyzes the signal, eliminates background noise and interference, calculates the real-time concentration value of the pollutant, and performs necessary corrections and calibrations to ensure the accuracy of the measurement results. The detected pollutant concentration data (such as ammonia nitrogen concentration) is transmitted to the neural network controller via a signal line or wireless communication. The neural network controller takes fluorescence spectroscopy data as input, along with data from other sensors (such as COD, pH, and dissolved oxygen), and processes it into the neural network model. The trained neural network model then predicts effluent water quality parameters in real time and calculates the optimal carbon source dosage based on the predictions.

[0060] like Figure 1 , Figure 3 , Figure 4 As shown, one embodiment of this utility model proposes a wastewater treatment carbon source dosing method based on neural network control, which is based on the wastewater treatment carbon source dosing device provided in Embodiment 1. The neural network controller is internally configured with a three-layer BP neural network with a 7-15-2 structure. The input layer is configured to input COD, ammonia nitrogen, pH value, biological tank water distribution flow rate, aeration rate, dissolved oxygen, and carbon source dosing amount. The hidden layer is configured with 15 neurons, and the output layer is configured to predict the effluent COD and ammonia nitrogen values. It can quickly solve the input-to-output mapping function and has strong self-learning and adaptive capabilities.

[0061] The aforementioned carbon source addition method for wastewater treatment based on neural network control considers seven factors closely related to the effluent quality, including influent COD, ammonia nitrogen, pH value, biological tank water distribution flow rate, aeration rate, dissolved oxygen, and carbon source addition amount. This data is transmitted to the data storage module via signal lines, and then to the neural network controller for analysis and processing. The internal control algorithm of the neural network controller adopts a 7-15-2 three-layer BP neural network to predict the effluent COD and ammonia nitrogen values. If the effluent COD and ammonia nitrogen values ​​in the water discharged from the clear water pipe 15 analyzed by the online instrument differ significantly from the output prediction value of the neural network controller, it indicates that the wastewater treatment is not ideal. The staff urgently needs to adjust the carbon source addition amount in the wastewater treatment process online to ensure that the output value matches the network prediction value. This realizes the online prediction and adjustment function of the neural network, ensuring that the effluent quality consistently meets the standards and that the wastewater treatment process operates efficiently and stably.

[0062] In one embodiment of this utility model, such as Figure 1 , Figure 3 , Figure 4 As shown, this utility model relates to a wastewater treatment control system based on a 7-15-2 three-layer BP neural network, aiming to optimize the wastewater treatment process through intelligent monitoring and adjustment, and ensure the stability and compliance of effluent quality. The following is a detailed description of the solution, including key inputs, control logic, and specific implementation methods;

[0063] This control system employs a three-layer BP neural network to predict the chemical oxygen demand (COD) and ammonia nitrogen concentrations in the effluent in real time. Its core function lies in achieving intelligent and adaptive regulation of the wastewater treatment process through comprehensive analysis of multiple key inputs.

[0064] The control system operates based on the following input parameters:

[0065] Wastewater flow rate (Q): The rate at which wastewater flows into the treatment system is monitored in real time using a flow meter to determine the treatment capacity; Influent COD concentration (COD_in): The COD in the effluent is periodically measured using online water quality monitoring instruments to assess the concentration of organic matter in the wastewater; Influent ammonia nitrogen concentration (NH3_N_in): The ammonia nitrogen concentration is also measured using online water quality monitoring instruments to determine the nitrogen source load; Carbon source dosage (C_add): The amount of carbon source added to the wastewater treatment process is directly adjusted to maintain the effectiveness of microorganisms; Temperature: Water temperature is monitored using a temperature sensor, which has a significant impact on the biochemical reaction kinetics; pH value: The acidity or alkalinity of the water is monitored using a pH meter to ensure a suitable environment for biological reactions;

[0066] Online monitoring equipment is used to continuously collect the above-mentioned input data, including wastewater flow rate, COD, ammonia nitrogen concentration, and environmental conditions (temperature, pH). This data is transmitted to the data processing unit via sensors for real-time processing and analysis.

[0067] A backpropagation neural network was trained using historical wastewater treatment datasets (including combinations of multiple variables and corresponding effluent quality results). The training data should cover different water quality fluctuations and treatment load conditions to improve the accuracy of model predictions; unnecessary noisy data was removed, effective features were selected, and the optimal network structure and hyperparameter settings were determined through cross-validation.

[0068] The real-time collected data is transmitted to a trained backpropagation (BP) neural network. The data includes the current wastewater flow rate (Q), COD concentration (COD_in), and ammonia nitrogen concentration (NH3_N_in). The network performs feedforward calculations on this input data and generates two output values: the predicted effluent COD concentration (COD_out_pred) and the predicted effluent ammonia nitrogen concentration (NH3_N_out_pred).

[0069] The actual effluent COD (COD_out_actual) and ammonia nitrogen concentration (NH3_N_out_actual) collected by online instruments are compared with the predicted values ​​(COD_out_pred and NH3_N_out_pred), respectively.

[0070] The system calculates the deviation between the actual value and the predicted value. If the difference between the two exceeds a set threshold (such as 5%), the system determines that the sewage treatment effect is not ideal.

[0071] The following is the formula for calculating deviation:

[0072] Deviation COD=|COD_out_actual-COD_out_pred|

[0073] Deviation NH3_N=|NH3_N_out_actual-NH3_N_out_pred|

[0074] Based on the magnitude and direction of the deviation, the system generates adjustment decisions. For example, if the deviation COD is positive, it means that the effluent COD is higher than the predicted value, and the system needs to increase the amount of carbon source added; if the deviation is negative, it means that the effluent COD is lower than the predicted value, and the system needs to reduce the amount of carbon source added.

[0075] The system feeds back the calculated change in the required carbon source dosage (ΔC_add) to the automated dosing device, adjusting the carbon source injection amount to ensure consistency between the predicted and actual values. After adjustment, the system re-monitors the COD and ammonia nitrogen values ​​of the effluent to ensure they meet the emission standards.

[0076] Through the implementation of this system, real-time monitoring and intelligent adjustment of water quality can be achieved during the wastewater treatment process, thereby improving the stability of effluent water quality: by online prediction and dynamic adjustment, water quality fluctuations caused by changes in wastewater composition are avoided; the system automation level is improved: manual intervention is reduced, operating efficiency is improved, and labor costs are reduced; and treatment capacity is enhanced: through refined control strategies, the treatment capacity of microorganisms is fully utilized, thereby improving the overall efficiency of wastewater treatment.

[0077] In one embodiment of this utility model, such as Figure 3 , Figure 4 The specific steps of the neural network control are shown below:

[0078] 1. Data Acquisition and Preprocessing:

[0079] Sensors in the wastewater treatment system monitor seven key parameters in real time, including influent COD, ammonia nitrogen concentration, pH value, biological tank water distribution flow rate, aeration rate, dissolved oxygen concentration, and current carbon source dosage.

[0080] The collected data is transmitted to the data storage module via a signal line. The data storage module is responsible for the initial processing of the transmitted real-time data, including noise removal, smoothing outlier data, and supplementing missing data, to ensure that the data transmitted to the neural network is complete and accurate.

[0081] 2. Input data normalization processing:

[0082] The preprocessed data is normalized before being transmitted to the neural network controller. This scaling process scales the variables to a uniform numerical range (typically between 0 and 1) to eliminate the scale effect between different physical quantities, facilitating neural network computation and training, and ensuring the model's stability and accuracy across different input ranges.

[0083] 3. Neural network structure:

[0084] The core of the neural network controller uses a three-layer BP neural network model:

[0085] Input layer: Seven input nodes receive normalized data on influent COD, ammonia nitrogen, pH value, biological tank water distribution flow rate, aeration rate, dissolved oxygen, and carbon source dosage, respectively.

[0086] Hidden layer: The hidden layer consists of 15 neurons that use activation functions (such as ReLU or Sigmoid) to perform nonlinear mapping on the input data and extract the complex relationships between the parameters.

[0087] Output layer: The output layer contains two nodes, which are used to predict the COD value and ammonia nitrogen concentration of the effluent, respectively. The output is the actual predicted value after inverse normalization.

[0088] 4. Network training and parameter optimization:

[0089] Before equipment deployment, the neural network controller is trained offline using historical data. The training data includes actual influent conditions (i.e., 7 input parameters) and corresponding effluent water quality (COD and ammonia nitrogen values). During training, gradient descent is used to continuously adjust the network weights and biases, allowing the network to gradually learn the relationships between the parameters and reduce prediction errors.

[0090] To improve the model's generalization ability, cross-validation is employed, and a regularization strategy is introduced to prevent overfitting. After training, the neural network controller can analyze the input data in real time and output prediction results.

[0091] 5. Real-time data input and prediction:

[0092] During wastewater treatment, real-time monitoring data is transmitted to a neural network controller. The controller processes the input data based on a pre-trained network model and quickly calculates the predicted values ​​of COD and ammonia nitrogen in the effluent. This process involves extremely low computational latency, ensuring rapid system response.

[0093] If the operating conditions of the wastewater treatment equipment or external conditions change during system operation, the neural network controller can adapt to the new input conditions, automatically adjust the prediction model, and ensure the accuracy of the output results.

[0094] 6. Comparison of predicted results with actual water discharge data:

[0095] Online instruments monitor the COD and ammonia nitrogen levels in the effluent in real time and compare them with the values ​​predicted by a neural network. If the actual effluent water quality parameters deviate significantly from the predicted values, it indicates an anomaly in the wastewater treatment system.

[0096] When the deviation exceeds the preset threshold, the system will trigger an alarm mechanism, prompting the operator to adjust the carbon source dosage to ensure that the effluent quality meets the standards.

[0097] 7. Feedback adjustment and automatic optimization:

[0098] After completing real-time prediction, the neural network controller will perform correlation analysis between the current carbon source addition and the prediction error, and adaptively adjust the carbon source addition strategy.

[0099] The system dynamically adjusts the carbon source dosage based on the changing trends and prediction errors of the effluent parameters, ensuring that the carbon source dosage is always at the optimal level during the wastewater treatment process.

[0100] The adjusted carbon source dosage is then input back into the control system, forming a closed-loop control structure. Through continuous feedback adjustment, the long-term stability of the effluent water quality is ensured.

[0101] 8. Online learning and model updates:

[0102] Over long-term operation, the system accumulates a large amount of historical data. The neural network controller can periodically update its model parameters through an online learning mechanism. This update process retrains some neural network layers using a new dataset to ensure that the controller can adapt to long-term changes in water quality conditions and operating environment.

[0103] 9. Fault detection and intelligent alarm:

[0104] The neural network also has a fault detection function. By analyzing the characteristics and trends of the input data, the system can predict possible faults or abnormalities (such as sensor failures, abnormal data fluctuations, etc.) in advance and issue early warning signals to prompt maintenance personnel to check or adjust, so as to avoid affecting the sewage treatment effect.

[0105] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0106] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention without departing from the principles and spirit of the present invention.

Claims

1. A wastewater treatment carbon source dosing device, comprising an anoxic tank (1) for pretreatment of wastewater, an MBR tank (2) for main wastewater treatment, and a sludge storage tank (3) for storing treated sludge, wherein the anoxic tank (1) is located on one side of the system, the sludge storage tank (3) is located on the other side of the system, and the MBR tank (2) is located between the anoxic tank (1) and the sludge storage tank (3), characterized in that, It also includes a water inlet (4) and a first lifting pump (5) and a controller for controlling the predicted water COD and ammonia nitrogen value, the water inlet is located above the facultative pond (1), sewage enters the MBR tank (2) through the water inlet; the first lifting pump (5) is arranged at the bottom of the MBR tank (2) and has a gap from the tank bottom, which is used to lift the sewage to the facultative pond (1).

2. The carbon source dosing device for sewage treatment according to claim 1, characterized in that, The MBR tank (2) is provided with a blower (6), an air valve (7) and an air pipe (8), the blower (6) is connected with the air valve (7) and the air pipe (8) in sequence, which are used to deliver external air into the device.

3. The carbon source dosing device for sewage treatment according to claim 2, characterized in that, It also includes an aeration head (9) and a biofilm reaction assembly (10), the aeration head (9) is arranged at the bottom of the air pipe (8) for aeration in the sewage; the biofilm reaction assembly (10) is arranged above the aeration head (9) for supporting the biological treatment of sewage.

4. The carbon source dosing device for sewage treatment according to claim 3, characterized in that, It also includes an online instrument (11), a self-priming pump (12), a regulating valve (13), a flowmeter (14) and a clean water pipe (15), the clean water pipe (15) is connected with the biofilm reaction assembly (10) and leads to the outside, the online instrument (11), the self-priming pump (12), the regulating valve (13) and the flowmeter (14) are arranged on the clean water pipe (15) in sequence.

5. The carbon source dosing device for sewage treatment according to claim 1, characterized in that, It also includes a second lifting pump (16), a first conveying pipe (17) and a second conveying pipe (18), the second lifting pump (16) is arranged at the bottom of the MBR tank (2) for conveying sludge; the first conveying pipe (17) is arranged on the first lifting pump (5) to connect the facultative pond (1) and the MBR tank (2) for conveying sewage; the second conveying pipe (18) is arranged on the second lifting pump (16) to connect the MBR tank (2) and the sludge storage tank (3) for conveying sludge.

6. The carbon source dosing device for sewage treatment according to claim 5, characterized in that, It also includes a fluorescent probe (22), an excitation light source (23), a fluorescence spectrum detection module (24) and a fluorescence detection chamber (27), the fluorescence detection chamber (27) is arranged on the upper part of the wall of the sludge storage tank (3); the excitation light source (23) is arranged on one side of the fluorescence detection chamber (27), and the fluorescence spectrum detection module (24) is arranged on the other side of the fluorescence detection chamber (27); the fluorescent probe (22) extends into the inside of the fluorescence detection chamber (27) outside the device.

7. The carbon source dosing device for sewage treatment according to claim 6, characterized in that, It also includes a shunt pipe (25) and a pressure piston (26), one end of the shunt pipe (25) is connected with one end of the second conveying pipe (18), the other end of the shunt pipe (25) leads into above the fluorescence detection chamber (27); the pressure piston (26) is arranged on the shunt pipe (25) for squeezing water in the conveyed sludge and discharging it into the fluorescence detection chamber (27).

8. The carbon source dosing device for sewage treatment according to claim 4, characterized in that, Also include fluorescence spectrum detection module (24), data storage module (28), neural network controller (29) and computer (30), the output end of the online instrument (11) is connected with data storage module (28) through signal line, the fluorescence spectrum detection module (24) outputs electric signal to data storage module (28), the data storage module (28) is used for integrating both data, and is transmitted to neural network controller (29) through signal line, the signal output end of neural network controller (29) is electrically connected with computer (30).

9. The carbon source dosing device for sewage treatment according to claim 1, characterized in that, Also include carbon source adding port (19), carbon source storage tank (20) and carbon source conveying pipe (21), the carbon source adding port (19) is set to the MBR pool (2) above for the injection of carbon source;The carbon source storage tank (20) is set below the carbon source adding port (19) for storing unused carbon source;The carbon source conveying pipe (21) is set on the carbon source storage tank (20), for conveying carbon source into the MBR pool (2).

Images