Method for treating ammonia-nitrogen wastewater with low carbon-nitrogen ratio

Abstract

The invention relates to the technical field of sewage treatment, and discloses a low-carbon-nitrogen-ratio ammonia-nitrogen wastewater treatment method, which comprises the following steps: measuring the carbon source concentration, ammonia-nitrogen concentration, flow and temperature of wastewater, inputting the carbon source and ammonia-nitrogen concentration into a prediction model to obtain a microbial activity predicted value, and determining the carbon source dosage based on the difference between the predicted value and an activity threshold value; converting the carbon source feeding amount into a reference feeding rate, and performing compensation correction on the denitrification rate through a temperature correction function to obtain an initial carbon source feeding rate; taking the subsequently monitored nitrogen-containing pollutant concentration of the effluent of the denitrification area as a feedback signal, comparing the feedback signal with the target concentration, and iteratively adjusting the feeding rate. According to the method, accurate matching of the carbon source feeding amount and the microbial metabolism requirement is achieved, the problems of unstable denitrification efficiency and secondary pollution caused by inaccurate carbon source feeding in a traditional process are effectively solved, and the effluent quality guarantee capacity is remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sewage treatment, in particular to a method for treating ammonia-nitrogen wastewater with low carbon-nitrogen ratio. BACKGROUND

[0002] Ammonia-nitrogen wastewater with low carbon-nitrogen ratio widely exists in landfill leachate, part of chemical wastewater and aquaculture wastewater, which is characterized by a serious shortage of organic carbon source for microorganisms to consume and a relatively high concentration of ammonia-nitrogen to be removed. The current mainstream biological denitrification process relies on two core processes of nitrification and denitrification, and denitrifying bacteria need sufficient organic carbon source as food and energy to convert nitrogen-containing pollutants into nitrogen gas and discharge the system. When treating wastewater with low carbon-nitrogen ratio, the biggest bottleneck is that the denitrification process cannot proceed smoothly due to the lack of carbon source, resulting in the total nitrogen in the effluent exceeding the standard.

[0003] To solve this problem, the most common method is to add external carbon source such as methanol or sodium acetate to the system to make up for the lack of carbon source. However, this method has derived a new core problem in practice, which is the precise control of carbon source addition. If the addition is insufficient, the denitrification effect is not ideal and the effluent still exceeds the standard; if the addition is excessive, not only the reagent resources are wasted and the operating cost is increased, but also too much organic matter is left in the effluent, causing secondary pollution and increasing the burden of the subsequent treatment unit. SUMMARY

[0004] Therefore, the purpose of the present application is to overcome the problems of unstable denitrification efficiency, carbon source waste and secondary pollution caused by difficult control of carbon source addition in the prior art, and to provide a method for treating ammonia-nitrogen wastewater with low carbon-nitrogen ratio, which keeps the dynamic balance between carbon source addition and microbial demand at all times, realizes stable and efficient denitrification efficiency, and effectively guarantees the effluent quality to meet the standard.

[0005] To solve the above technical problems, the present application provides a method for treating ammonia-nitrogen wastewater with low carbon-nitrogen ratio, comprising: determining the carbon source concentration, ammonia-nitrogen concentration, wastewater flow and wastewater temperature of the wastewater to be treated; inputting the determined carbon source concentration and ammonia-nitrogen concentration into a prediction model to obtain a microbial activity prediction value; determining the carbon source addition amount according to the difference between the microbial activity prediction value and an activity threshold value; determining a reference carbon source addition rate according to the carbon source addition amount, and correcting the reference carbon source addition rate according to the influence of the wastewater temperature on the denitrification rate to obtain an initial carbon source addition rate; adding carbon source to the denitrification zone with the initial carbon source addition rate as the control starting point, and monitoring the concentration of nitrogen-containing pollutants in the effluent of the denitrification zone; comparing the monitored concentration of nitrogen-containing pollutants with a target concentration, and iteratively adjusting the initial carbon source addition rate according to the comparison result. wherein the reference carbon source dosing rate is determined according to the carbon source dosing amount, comprising: decomposing the carbon source dosing amount into N dosing amounts according to a control duration, and dosing a first dosing amount to the denitrification zone; monitoring the change of the redox potential of the mixed liquid in the denitrification zone to obtain a redox potential drop slope; determining a correction factor of the next control period based on the redox potential drop slope, and adjusting the dosing amount according to the correction factor to obtain the dosing amount of the next control period; calculating the dosing amount of the current control period divided by the period duration to obtain the reference carbon source dosing rate of the current control period; wherein the correction factor is determined based on the following manner: ; denotes the correction factor; A denotes an activity coupling coefficient, and takes a value of 0.9-1.2; denotes the redox potential drop slope.

[0006] Preferably, the prediction model comprises: an input layer comprising two input nodes; the two input nodes respectively receive the carbon source concentration and the ammonia nitrogen concentration; a feature fusion layer for cross-combining the received carbon source concentration and ammonia nitrogen concentration to obtain an initial feature vector representing the carbon-nitrogen balance relationship; at least one hidden layer, each of the hidden layers comprising a full connection layer, the full connection layer being used for nonlinear transformation and high-order feature abstraction of the initial feature vector or the output feature of the previous hidden layer, mining the nonlinear mapping relationship between the carbon-nitrogen concentration and the microbial activity, and obtaining a high-dimensional feature vector; an attention layer connected with the last hidden layer, used for weighting and calibrating the high-dimensional feature vector; and an output layer used for mapping the calibrated high-dimensional feature vector to the microbial activity prediction value.

[0007] Preferably, the difference between the microbial activity prediction value and the activity threshold value is defined as an activity deviation Δ, and the carbon source dosing amount is determined according to the activity deviation, comprising: determining an adjustment coefficient K according to the predefined interval in which the activity deviation lies; calculating the total nitrogen mass that needs to be removed according to the measured ammonia nitrogen concentration and the target denitrification efficiency; calculating the reference carbon source dosing amount required for complete denitrification of the nitrogen mass according to the denitrification biochemical reaction metering relationship; and calculating the product of the adjustment coefficient K and the reference carbon source dosing amount as the carbon source dosing amount.

[0008] Preferably, the total nitrogen mass that needs to be removed is calculated according to the measured ammonia nitrogen concentration and the target denitrification efficiency, comprising: setting a target total nitrogen concentration of the effluent water; obtaining the total nitrogen concentration that needs to be removed by subtracting the target total nitrogen concentration from the measured ammonia nitrogen concentration; and determining the total nitrogen mass that needs to be removed according to the wastewater flow and the total nitrogen concentration that needs to be removed, in the following manner: M = C x Q x t; Wherein, M represents the total nitrogen mass to be removed; C represents the total nitrogen concentration to be removed; Q represents the wastewater flow rate; t represents the control time length.

[0009] Preferably, the adjustment coefficient K is determined according to the predefined interval in which the activity deviation is located, including: setting a negative deviation threshold Δ1 and a positive deviation threshold Δ2; if Δ ≤ Δ1, then K = K1; wherein K1 is a strengthening dosage coefficient greater than 1; if Δ1 < Δ ≤ 0, then K = K2; wherein 1 ≤ K2 < K1; if 0 < Δ ≤ Δ2, then K = K3; wherein K3 < 1; if Δ > Δ2, then K = K4; wherein K4 < K3.

[0010] Preferably, the reference carbon source dosage rate is corrected according to the influence of the wastewater temperature on the denitrification rate to obtain an initial carbon source dosage rate, including: introducing a temperature correction function, calculating a temperature correction factor corresponding to the wastewater temperature according to the temperature correction function ; calculating the product of the reference carbon source dosage rate and the temperature correction factor as the initial carbon source dosage rate; wherein the temperature correction function is: ; represents the temperature correction factor; T is the measured wastewater temperature; Tref is the reference temperature, taking a value of 30℃ ~ 35℃; θ represents the denitrification temperature coefficient, taking a value of 1.02 ~ 1.08, indicating that the rate of denitrification biochemical reaction is increased to 1.02 to 1.08 times for every 1℃ increase in wastewater temperature compared to the reference temperature.

[0011] Preferably, the monitoring of the oxidation-reduction potential drop slope includes the following enhanced steps: injecting a tracer into the denitrification zone before adding the carbon source; simultaneously monitoring the changes in the tracer concentration and the oxidation-reduction potential; correcting the interference of the hydraulic mixing effect on the oxidation-reduction potential drop slope through the tracer concentration change curve to obtain the corrected oxidation-reduction potential drop slope.

[0012] Preferably, the monitored nitrogen-containing pollutant concentration is compared with the target concentration, and the initial carbon source dosage rate is iteratively adjusted according to the comparison result, including: calculating the difference between the effluent nitrogen-containing pollutant concentration and the target concentration to obtain an instantaneous concentration deviation; analyzing the historical effluent nitrogen-containing pollutant concentration sequence to obtain an effluent nitrogen-containing pollutant concentration prediction value; summing the instantaneous concentration deviation and the effluent nitrogen-containing pollutant concentration prediction value, and inputting the sum result into a PID controller to obtain a carbon source dosage rate adjustment amount; the sum of the carbon source dosage rate adjustment amount obtained in the last control period and the initial carbon source dosage rate at the start of the current control period is taken as the actual carbon source dosage rate in the current control period.

[0013] Preferably, before adding the carbon source to the denitrification zone, the pH value and alkalinity of the wastewater to be treated are monitored; if the pH is lower than 6.8 or higher than 7.5, sodium acetate is added as the carbon source; if the pH is greater than or equal to 6.8 and less than or equal to 7.5, the alkalinity is further determined; if the alkalinity is not lower than 200 mg / L, methanol is added as the carbon source; if the alkalinity is lower than 200 mg / L, sodium acetate is added as the carbon source.

[0014] The above technical solution of the present application has the following beneficial effects compared with the prior art: The low-carbon-nitrogen-ratio ammonia-nitrogen wastewater treatment method of the present application fundamentally changes the control mode of carbon source addition, realizes the leap from empirical control to model predictive control, systematically solves the problem of precise control, and keeps the dynamic balance between carbon source addition and microbial demand at all times, realizes stable and efficient denitrification efficiency, and effectively guarantees that the effluent quality meets the standard.

[0015] Among them, the carbon source concentration and ammonia nitrogen concentration reflecting the substrate supply and microbial load core variables are input into the prediction model to obtain the prediction value representing the activity state of denitrifying microorganisms, and the theoretical carbon source demand is determined based on the difference between the prediction value and the activity threshold value, breaking through the traditional operation mode of relying on empirical addition or fixed carbon-nitrogen ratio, and instead taking the actual metabolic activity of microorganisms as the basis to quantitatively evaluate its utilization ability and demand gap for carbon source under the current water quality and operating conditions, to ensure the adaptability between carbon source addition and microbial function from the source, to avoid the problems of denitrification process being blocked due to insufficient carbon source, nitrogen removal being incomplete, or resource waste and residual organic matter in effluent due to excessive carbon source.

[0016] The method further introduces the wastewater temperature as a key environmental factor, dynamically corrects the baseline carbon source addition rate based on the significant influence of temperature on denitrification enzyme activity and microbial metabolic rate, matches the carbon source supply strength with the actual metabolic capacity of microorganisms under different temperature conditions, further improves the environmental adaptability and biological rationality of the carbon source addition strategy, and ensures that the denitrification process is carried out efficiently at an appropriate metabolic level.

[0017] Further, the dynamic feedback adjustment mechanism guided by the actual effluent quality, after completing the setting of the initial carbon source dosage rate, monitors the concentration of nitrogen-containing pollutants in the effluent of the denitrification zone, compares it with the preset target concentration, and iteratively optimizes and dynamically adjusts the initial dosage rate according to the deviation signal. This feedback control loop makes the carbon source dosage process no longer limited to static preset or single prediction, but can continuously correct the carbon source supply strategy according to the real-time feedback of the actual denitrification effect, ensuring that the carbon source dosage always keeps pace with the actual needs of microorganisms under dynamic water quality conditions, thereby maintaining efficient and stable operation of the denitrification reaction throughout the entire process cycle. Through this feedback mechanism, the system not only responds to the influence of external interference factors such as fluctuations in influent water quality and changes in environmental conditions on the metabolic state of microorganisms in a timely manner, but also self-adaptively adjusts the carbon source dosage intensity to avoid the risk of secondary pollution caused by excessive dosage or the decline in denitrification efficiency caused by insufficient dosage, achieving significant improvement in process control accuracy and anti-interference ability. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to make the content of the present application more easily understood, the following further details of the present application will be described according to specific embodiments of the present application and in conjunction with the accompanying drawings, in which: Figure 1 Flowchart of the low carbon-nitrogen ratio ammonia nitrogen wastewater treatment method in the preferred embodiment of the present application; Figure 2 Layer structure block diagram of the prediction model in the preferred embodiment of the present application; Figure 3 Flowchart of obtaining the carbon source dosage in the preferred embodiment of the present application; Figure 4 Flowchart of determining the reference carbon source dosage rate in the preferred embodiment of the present application; Figure 5 Flowchart of iteratively adjusting the initial carbon source dosage rate in the preferred embodiment of the present application. DETAILED DESCRIPTION

[0019] The present application will be further described below in conjunction with the accompanying drawings and specific embodiments, so that those skilled in the art can better understand the present application and implement it, but the embodiments are not limiting the present application.

[0020] Low carbon-nitrogen ratio ammonia nitrogen wastewater widely exists in landfill leachate, part of chemical wastewater and aquaculture wastewater, and its typical feature is that the ammonia nitrogen (NH4 + -N or NO xThe content of organic substances (such as volatile fatty acids, alcohols, sugars and the like) which are high in concentration but can be used as carbon sources and energy sources for denitrifying microorganisms is seriously insufficient. In the biological denitrification process, nitrification oxidizes ammonia nitrogen into nitrate or nitrite, and denitrification relies on denitrifying microorganisms to utilize organic carbon sources as electron donors to gradually reduce nitrate / nitrite into nitrogen under anoxic conditions, so as to achieve the final removal of nitrogen. Therefore, whether the organic carbon source is sufficient directly determines the efficiency of the denitrification process and is a key limiting factor for the denitrification of low-carbon-nitrogen-ratio wastewater.

[0021] The purpose of the embodiment of the present application is to overcome the problems in the prior art that the carbon source addition is difficult to control, resulting in unstable denitrification efficiency, waste of carbon source and secondary pollution, and to provide a low-carbon-nitrogen-ratio ammonia nitrogen wastewater treatment method, in which the carbon source addition always maintains a dynamic balance with the microbial demand, so that stable and efficient denitrification efficiency is achieved, and the effluent quality is effectively guaranteed to meet the standard.

[0022] Referring to Figure 1 The embodiment of the present application discloses a low-carbon-nitrogen-ratio ammonia nitrogen wastewater treatment method, comprising: determining the carbon source concentration, ammonia nitrogen concentration, wastewater flow and wastewater temperature of the wastewater to be treated; inputting the determined carbon source concentration and ammonia nitrogen concentration into a prediction model to obtain a microbial activity prediction value; determining the carbon source addition amount according to the difference between the microbial activity prediction value and an activity threshold value; determining a reference carbon source addition rate according to the carbon source addition amount, and correcting the reference carbon source addition rate according to the influence of the wastewater temperature on the denitrification rate to obtain an initial carbon source addition rate; taking the initial carbon source addition rate as a control starting point, adding carbon source to the denitrification zone, monitoring the concentration of nitrogen-containing pollutants in the effluent of the denitrification zone, comparing the monitored concentration of nitrogen-containing pollutants with a target concentration, and iteratively adjusting the initial carbon source addition rate according to the comparison result.

[0023] Referring to Figure 2 The prediction model used in the embodiment of the present application comprises the following structure: The input layer includes two independent input nodes, the first node receives the normalized carbon source concentration scalar value, and the second node receives the normalized ammonia nitrogen concentration scalar value; the normalization processing scales the original data to the interval [0, 1] or [-1, 1] to accelerate the convergence of the model and improve the accuracy.

[0024] The feature fusion layer includes a fully connected layer or a custom feature cross layer, which does not perform simple feature splicing, but performs explicit feature cross combination through weight matrix calculation, and the operation can be represented as: Z = φ(W * [C, N] + b); [C, N] is the input carbon source and ammonia nitrogen concentration vector, W is the weight matrix, b represents the bias vector, and φ represents the activation function. Through training, the layer automatically learns the carbon-nitrogen ratio and the complex features of carbon-nitrogen concentration affecting microbial activity, obtaining an initial feature vector that can represent the carbon-nitrogen balance relationship.

[0025] The hidden layer is stacked by at least one hidden layer, each hidden layer containing a fully connected layer, usually followed by an activation function and an optional dropout layer; each fully connected layer performs nonlinear transformation and high-order feature abstraction on the input features; wherein the first hidden layer transforms the initial feature vector to mine the primary nonlinear relationships of carbon source sufficiency but ammonia nitrogen inhibition, ammonia nitrogen saturation effect under carbon source limitation, metabolic shunt inhibition of carbon source excess, ammonia nitrogen concentration biphasic effect, and metabolic threshold effect implied by carbon source type; the subsequent hidden layers further combine and abstract on this basis to mine deeper and more complex nonlinear mapping relationships, and finally output a high-dimensional feature vector containing rich information. The dropout layer prevents model overfitting by randomly shutting down part of the neurons during training.

[0026] The attention layer is configured as a lightweight attention module to address the problem of different dominant factors affecting microbial activity under different water quality conditions. It receives the high-dimensional feature vector output by the hidden layer and calculates the importance weight of each feature dimension in the vector through a small neural network. Subsequently, the original high-dimensional feature vector is weighted and calibrated with these weights. This means that for some wastewater, the model may pay more attention to carbon source-related features; while for others, it may pay more attention to ammonia nitrogen toxicity-related features, thereby achieving dynamic feature importance allocation.

[0027] The output layer is configured as a fully connected layer with a single neuron node, which aggregates the feature vector after weighted calibration by the attention mechanism and maps it to the final microbial activity prediction value through a Sigmoid function activation function; the value is a dimensionless relative number reflecting the potential metabolic capacity of the denitrifying bacterial population under the current carbon-nitrogen conditions.

[0028] The training process of the prediction model includes: collecting a large amount of historical operation data to form a training data set. Each data should include: {carbon source concentration, ammonia nitrogen concentration} as input features, and the corresponding real microbial activity value as labels. The real value usually needs to be indirectly but reliably represented by the specific enzyme activity or specific oxygen uptake rate (SOUR) of the mixed liquid sample collected synchronously with the input data through offline determination.

[0029] Loss function definition: mean squared error is used as the loss function to measure the gap between the model prediction value and the real label.

[0030] Optimization process: using Adam optimizer, according to the error calculated by the loss function, from the output layer to the input layer, adjusting the weight and bias parameters in the network layer by layer; the goal of training is to iterate continuously until the model weight converges, so that the loss function value is minimized, that is, the model prediction result is closest to the true value.

[0031] Validation and early stopping: use an independent validation dataset to monitor the training process, when the performance on the validation set no longer improves or even starts to decline (i.e. overfitting), stop training in advance to obtain the model with the strongest generalization ability.

[0032] Integrate the trained and accuracy-qualified model into the central control system of the wastewater treatment system. After the measured carbon source concentration and ammonia nitrogen concentration are normalized in the same way as during training, they are input into the deployed model. The model follows the network structure constructed, and sequentially passes through the input layer → feature fusion layer → hidden layer → attention layer → output layer, and performs a series of forward propagation calculations. The model outputs a microbial activity prediction value. This prediction value is sent to the downstream control logic to calculate the carbon source dosage, thereby realizing precise and forward-looking intelligent control of carbon source addition.

[0033] In specific application scenarios, the key parameters of the wastewater to be treated are determined in real time: The TOC analyzer is used to measure the carbon source concentration, which usually detects easily biodegradable COD (such as BOD5, COD （sol） , or specific biodegradable carbon sources such as VFAs concentration), reflecting the level of organic carbon in wastewater that can be directly utilized by denitrifying microorganisms. The Nash reagent spectrophotometric method is used to measure the ammonia nitrogen concentration, including NH4 + -N or NO x -N, representing the total nitrogen load source that needs to be removed by the system. The electromagnetic flowmeter is used to monitor the wastewater flow, i.e. the influent quantity per unit time (unit: m 3 / h), which is used to calculate the pollutant load and the flow-based reference for carbon source addition. The PT100 temperature sensor is used to detect the wastewater temperature, which is an important environmental factor affecting microbial enzyme activity and denitrification rate. Online sensors (such as COD online instrument, ammonia nitrogen analyzer, temperature probe) can be used in combination with regular laboratory tests to ensure the real-time and accuracy of the data.

[0034] The determined carbon source concentration and ammonia nitrogen concentration are input into a neural network prediction model trained by historical data to output a microbial activity prediction value in the range of 0-1; the model realizes quantitative evaluation of the physiological state of the denitrifying bacteria group by mining the nonlinear relationship between carbon-nitrogen balance and microbial metabolism. The activity threshold is set to 0.3-0.5, and when the microbial activity prediction value is lower than the activity threshold, the compensation dosage is calculated according to the formula: carbon source dosage = reference dosage × [1 + α × (activity threshold - prediction value)], wherein α is a compensation coefficient, and the value is 0.5-2.0.

[0035] The carbon source dosage is divided by the control period (recommended 1-4 hours) to obtain the reference carbon source dosage rate, the Arrhenius correction model is used to obtain the temperature correction factor, and the reference carbon source dosage rate is multiplied by the temperature correction factor to obtain the initial carbon source dosage rate.

[0036] The initial carbon source dosage rate is used to add sodium acetate or methanol as the carbon source to the denitrification zone, an online nitrate sensor is used to monitor the effluent nitrate nitrogen concentration (control target: <5 mg / L), and the dosage rate is dynamically adjusted through the incremental PID algorithm to iteratively adjust the initial carbon source dosage rate to obtain the actual carbon source dosage rate.

[0037] The low carbon-nitrogen ratio ammonia nitrogen wastewater treatment method fundamentally changes the control mode of carbon source addition, realizes the leap from empirical control to model prediction control, systematically solves the problem of precise control, and keeps the carbon source addition in dynamic balance with the microbial demand, thereby realizing stable and efficient denitrification efficiency and effectively guaranteeing that the effluent water quality meets the standard.

[0038] The carbon source concentration and ammonia nitrogen concentration reflecting the core variables of substrate supply and microbial load are input into the prediction model to obtain the prediction value representing the activity state of denitrifying microorganisms, and the theoretical carbon source demand is determined based on the difference between the prediction value and the activity threshold, thereby breaking through the traditional operation mode relying on empirical addition or fixed carbon-nitrogen ratio and instead basing on the actual metabolic activity of microorganisms to quantitatively evaluate the utilization ability and demand gap of carbon source under the current water quality and operating conditions, thereby ensuring the adaptability between carbon source addition and microbial function from the source, avoiding the problems of blocked denitrification process and incomplete nitrogen removal due to insufficient carbon source, or resource waste and residual organic matter in the effluent due to excessive carbon source.

[0039] The method further introduces the wastewater temperature as a key environmental factor, dynamically corrects the reference carbon source dosage rate based on the significant influence of temperature on denitrifying enzyme activity and microbial metabolic rate, matches the carbon source supply strength with the actual metabolic capacity of microorganisms under different temperature conditions, further improves the environmental adaptability and biological rationality of the carbon source addition strategy, and ensures that the denitrification process is carried out efficiently at an appropriate metabolic level.

[0040] Further, the dynamic feedback adjustment mechanism guided by the actual effluent quality, after setting the initial carbon source dosage rate, monitors the concentration of nitrogen-containing pollutants in the denitrification zone effluent and compares it with the preset target concentration. Based on the deviation signal, the initial dosage rate is iteratively optimized and dynamically adjusted. This feedback control loop no longer limits the carbon source dosage process to static preset or single prediction, but continuously corrects the carbon source supply strategy based on real-time feedback of the actual denitrification effect, ensuring that the carbon source dosage always synchronizes with the actual needs of microorganisms under dynamic water quality conditions, thereby maintaining efficient and stable operation of the denitrification reaction throughout the entire process cycle. Through this feedback mechanism, the system can not only respond to the influence of external interference factors such as inflow water quality fluctuations and environmental condition changes on microbial metabolic state in a timely manner, but also adaptively adjust the carbon source dosage intensity to avoid the risk of secondary pollution caused by excessive dosage or the decline in denitrification efficiency caused by insufficient dosage, achieving significant improvement in process control accuracy and interference resistance.

[0041] When determining the carbon source dosage based on activity deviation, the conventional method usually calculates the theoretical carbon source dosage based on the fixed stoichiometric ratio according to the inflow ammonia nitrogen concentration and effluent quality requirements, such as how many kg of COD is needed for 1 kg of nitrate nitrogen in denitrification. The core assumption of this method is that microorganisms are in an ideal stable and fully active state, and can fully utilize the added carbon source for denitrification. However, in actual operation, this assumption often deviates significantly from the actual situation, mainly in the following two aspects: (1) Microbial activity is influenced by a variety of factors, including but not limited to: temperature, pH, dissolved oxygen, toxic and harmful substance concentration, sludge age, nutrient conditions, microbial population structure; the metabolic activity of microorganisms is not constant, but is in a state of continuous fluctuation, which may be high in activity at a certain time period (metabolism is vigorous, denitrification ability is strong), or may be low in activity at another time period (inhibited, aged, poisoned, etc., denitrification ability is weak or even lost), fixed stoichiometric ratio cannot perceive or respond to such activity changes, resulting in rigid carbon source dosage strategy and disconnection with the actual needs of microorganisms.

[0042] (2) If the dosage is set based on the potential maximum activity of microorganisms, when the actual activity is low, the carbon source cannot be fully utilized, causing carbon source waste, overflow to the effluent, leading to COD exceeding the standard and secondary pollution; if the dosage is set based on the current or average low activity of microorganisms, when the actual activity of microorganisms is high, the carbon source supply is insufficient, and the denitrification ability cannot be fully utilized, resulting in waste of treatment capacity, low denitrification efficiency, and total nitrogen exceeding the standard in the effluent.

[0043] To solve this problem, refer to Figure 3As shown, the embodiment scheme of the present application proposes a carbon source dosage calculation method based on microbial activity prediction, with activity deviation Δ as the core control variable, and through the adjustment coefficient K to realize dynamic feedback, specifically: define the difference between the microbial activity prediction value and the activity threshold value as the activity deviation Δ, determine the carbon source dosage according to the activity deviation, including: according to the pre-defined interval of the activity deviation, determine the adjustment coefficient K; according to the measured ammonia nitrogen concentration and the target denitrification efficiency, calculate the total nitrogen mass that needs to be removed; then according to the denitrification biochemical reaction metering relationship, calculate the reference carbon source dosage required for complete denitrification of the nitrogen mass; calculate the product of the adjustment coefficient K and the reference carbon source dosage as the carbon source dosage.

[0044] In a specific application scenario, four key intervals of activity deviation Δ and their corresponding adjustment coefficients K are defined in advance according to a large number of experiments and operation experience. First, set the negative deviation threshold Δ1 and the positive deviation threshold Δ2: Based on the setting strategy of the effluent water quality reverse correlation, analyze the data, and draw a scatter plot of activity deviation Δ and effluent total nitrogen concentration. The activity deviation Δ value when the effluent total nitrogen concentration shows a clear and persistent upward trend is set as the negative deviation threshold Δ1. This point is the warning line for system stability, and strong intervention must be carried out before this point. The activity deviation Δ value when the decrease of effluent total nitrogen concentration is very small and the denitrification efficiency is not obviously improved is set as the positive deviation threshold Δ2. If the value exceeds this point, the improvement of effluent water quality brought by inputting more carbon source is very small. In a specific embodiment, the negative deviation threshold Δ1 and the positive deviation threshold Δ2 take values of -0.2 and +0.15, respectively.

[0045] If Δ≤Δ1, it indicates a serious inhibition zone, and the microbial activity is seriously insufficient. At this time, K=K1, K1 is a strong dosage coefficient of 1.5 to 3.0. A carbon source far exceeding the theoretical value is added to provide more sufficient carbon source for microorganisms and create a more favorable metabolic environment, so as to stimulate the recovery of microbial activity, improve the metabolic rate and denitrification capacity of microorganisms, and aim to stimulate and awaken the microbial community and provide sufficient substrate to overcome possible inhibition. It is especially suitable for extreme or abnormal working conditions such as microbial inhibition by toxicity, too low temperature, unsuitable pH, sludge aging, etc.

[0046] If Δ1<Δ≤0, it indicates a mild deficiency zone, and the microbial activity is slightly lower than the ideal state. At this time, K=K2, K2 is a reference guarantee coefficient of 1.0 to 1.5; the reference guarantee strategy is adopted, which aims to provide slightly abundant carbon source supply for microorganisms on the basis of ensuring that the carbon source is sufficient to support the denitrification reaction, to help them complete the denitrification process more smoothly, and at the same time to avoid waste caused by excessive carbon source. It is suitable for working conditions where the microbial activity is affected by slight environmental fluctuations (such as slight temperature drop, small pH shift), or in the transition period of operation, which helps to maintain the stability of denitrification efficiency, while controlling the carbon source cost and effluent COD risk.

[0047] If 0 < Δ ≤ Δ2, it indicates that it is an active suitable area, and the microbial activity is good, at this time, K = K3, K3 is a preventive reduction coefficient of 0.7 to 1.0, a precise matching strategy is adopted, and the amount is added according to the theoretical value or slightly reduced, the purpose is to avoid excessive carbon source leading to excessive dependence on external carbon source of microorganisms, inhibiting endogenous metabolism, or causing carbon source overflow to effluent, resulting in COD exceeding standard; for the conventional working condition of good system running state, healthy microbial population and suitable temperature, it is helpful to optimize the carbon source use efficiency, reduce the operation cost, and prevent the increase of effluent COD at the same time.

[0048] If Δ > Δ2, it indicates that it is an active vigorous area, and the microbial activity is very high, at this time, K = K4, K4 is a deep reduction coefficient of 0.3 to 0.7, an excessive strategy is adopted, and the carbon source is actively reduced, the purpose is to provide the minimum carbon source demand to maintain high-efficiency reaction under the condition that the microorganisms have strong denitrification capacity, avoid excessive addition causing resource waste, carbon source residue and subsequent treatment burden; it is suitable for the working condition that the microbial activity is spontaneously improved due to excellent environmental conditions (such as moderate temperature, rich nutrition and no toxic interference); it is helpful to realize the operation goal of low input and high output, and is an important guarantee for long-term stable and efficient operation of the system.

[0049] In the embodiment scheme of the present application, the carbon source addition amount calculation method based on microbial activity prediction, taking activity deviation Δ as the core control variable and realizing dynamic feedback through adjustment coefficient K, can truly reflect the metabolic capacity of microorganisms and the carbon source demand, so as to add carbon source according to the actual needs of microorganisms, when the microbial activity is high, the carbon source is appropriately increased, the denitrification potential is fully tapped, and the treatment capacity and total nitrogen removal rate are improved; when the microbial activity is low, the carbon source addition is reduced, invalid addition and resource waste are avoided, and finally the high matching of carbon source addition and microbial function state is realized, the denitrification efficiency and carbon source utilization efficiency are significantly improved.

[0050] In addition, for the problem of how to determine the carbon source addition amount, the common method is to calculate the carbon source addition amount according to the concentration of influent ammonia nitrogen, however, wastewater treatment is a continuous flow process, and the actual treatment load of the system is determined by concentration and flow; the total load of a wastewater with high ammonia nitrogen concentration but small flow may be much lower than that of a wastewater with medium concentration but large flow, causing calculation deviation of carbon source addition amount.

[0051] In order to solve the problem, in the embodiment scheme of the present application, according to the measured ammonia nitrogen concentration and target denitrification efficiency, the total nitrogen mass to be removed is calculated, including: setting the target total nitrogen concentration of effluent reaching standard; subtracting the target total nitrogen concentration from the measured ammonia nitrogen concentration to obtain the total nitrogen concentration to be removed; determining the total nitrogen mass to be removed according to the wastewater flow and the total nitrogen concentration to be removed, and the calculation method is as follows: M = C x Q x t; Where M represents the total nitrogen mass to be removed; C represents the total nitrogen concentration to be removed; Q represents the wastewater flow rate; t represents the control period.

[0052] In a specific application scenario, first determine the final control target, set the target total nitrogen concentration, which is derived from the national or local emission standard; for example, 15 mg / L; in order to simplify the calculation, the measured ammonia nitrogen concentration is regarded as the influent total nitrogen concentration, which is reasonable and conservative in the denitrification pre-treatment process of low carbon-nitrogen ratio wastewater, because the organic nitrogen content of such wastewater is usually low, and the goal is to finally convert ammonia nitrogen into nitrogen gas, so the simplification ensures that the calculated carbon source amount is sufficient to deal with the worst case.

[0053] Set the control period, that is, the time period for the system to perform one feedforward calculation and adjustment, and the control period determines the response granularity of the control system. The physical meaning of the total nitrogen mass M to be removed is that in order to ensure that the effluent meets the standard in the future t period, the system must remove the total mass of nitrogen elements from the wastewater.

[0054] Through M = C x Q x t, the system can respond to the dynamic changes of the influent load (concentration x flow rate) in real time, when the flow rate Q increases or the influent concentration rises, the calculated nitrogen mass M will immediately increase, thereby driving the downstream calculated carbon source dosage to increase accordingly; this makes the carbon source dosage strictly match the real processing load of the system, and solves the problem of mismatch between the dosage and the actual demand from the main flow.

[0055] The carbon source dosage calculated by the foregoing steps is a one-time prediction based on the water quality conditions at the beginning of a control period, however, within a control period that can be as long as 10-40 hours, the microbial activity in the denitrification tank is dynamically changing (may be caused by accumulation of endogenous metabolites, minor water quality fluctuations, etc.). The traditional average allocation dosage method cannot respond to such dynamic changes within the period; on the other hand, if the open-loop execution of the dosage rate within the period, if the initial prediction is biased, or a disturbance occurs during the process, the system can only be corrected through effluent feedback after the end of the period, which has a control blind area.

[0056] In order to solve this problem, the inner loop feedback according to the dynamic response of the oxidation-reduction potential is introduced in the embodiment scheme of the present application, which makes real-time fine tuning to the feedforward main loop, referring to Figure 4As shown, the reference carbon source addition rate is determined according to the carbon source addition amount, comprising: the carbon source addition amount is evenly decomposed into N addition amounts according to the control duration, and the first addition amount is added to the denitrification zone; the change of the redox potential of the mixed liquid in the denitrification zone is monitored to obtain a redox potential drop slope; a correction factor of the next control period is determined based on the redox potential drop slope, and the addition amount is adjusted according to the correction factor to obtain the addition amount of the next control period; the reference carbon source addition rate of the current control period is obtained by dividing the addition amount of the current control period by the period duration; wherein the correction factor is determined based on the following method: ; The correction factor is represented by A; A represents an activity coupling coefficient, and the value is 0.9-1.2; The redox potential drop slope is represented by B.

[0057] In a specific application scenario, the total carbon source addition amount is evenly decomposed into N addition amounts, that is, the total carbon source addition amount is evenly divided into N parts, and is put in N times within the control duration; after the first addition of the carbon source, the redox potential of the mixed liquid is monitored, and the instantaneous drop slope of the redox potential is calculated; in an anoxic environment, the input of easily degradable carbon source will be immediately consumed by denitrifying bacteria, and the oxygen in the oxidant will be taken away during the consumption process, resulting in a sharp drop in the redox potential value. The slope of this drop directly reflects the speed and intensity of the utilization of carbon source by the microbial community and the start of denitrification, and is an extremely sensitive indicator of microbial instantaneous activity. The greater the absolute value of the redox potential drop slope, the faster the microorganism eats and the higher the activity; on the contrary, the activity is low and the reaction is slow.

[0058] The redox potential drop slope is mapped to a smooth change correction factor, and the calculated correction factor is applied to the addition amount of the next control period, and the carbon source addition amount of the next control period is the initial calculated addition amount x .

[0059] In the embodiment of the application, the feedback of the redox potential is introduced, so that the reference carbon source addition rate does not completely depend on the prediction at the beginning of the control period, but at the end of each small period, the real reaction of the microorganism is used to check and correct the carbon source input amount of the next period, so that the feedforward model has the ability of online self-correction. In addition, the correction factor is calculated by a logarithmic function, which ensures that when the redox potential drop slope is large, the correction will not increase without limit, avoiding overshoot; when the redox potential drop slope is small, the correction is sensitive enough. This nonlinear mapping relationship ensures that the response of the control system is effective, smooth and robust, preventing the dramatic change of the addition rate.

[0060] Note: Oxidation-Reduction Potential (ORP) is an index to measure the relative strength of oxidation or reduction of a solution; high positive ORP value indicates the solution is in an oxidizing environment, rich in oxygen, nitrate and other electron acceptors; low negative ORP value indicates the solution is in a reducing environment, rich in organic matter, hydrogen sulfide and other electron donors; in activated sludge process, ORP value indirectly reflects the metabolic activity of microorganisms and the chemical state of the environment they are in.

[0061] When adding easily degradable carbon source (such as sodium acetate, methanol) to the denitrification zone, a rapid biochemical chain reaction will be triggered: Step 1: Rapid consumption of residual dissolved oxygen Aerobic bacteria and denitrifying bacteria will immediately use the newly added carbon source for respiration, first consuming the residual trace amount of dissolved oxygen in the water, which will cause ORP to slowly decrease from a high level.

[0062] Step 2: Start and accelerate of denitrification When the dissolved oxygen is consumed, denitrifying bacteria become dominant; they start to use nitrate as an electron acceptor, oxidizing carbon source while reducing nitrate to nitrogen gas.

[0063] The process of reducing nitrate is a strong reducing chemical reaction, which will cause ORP value to drop sharply.

[0064] The slope of ORP drop is the key bridge connecting carbon source addition and microbial activity: if the microbial community is healthy and active, after the addition of carbon source, they will quickly use the carbon source to complete the switch from aerobic respiration to denitrification, which will cause ORP to drop sharply with a large absolute value. If the microorganisms are low in activity due to poisoning, inhibition or adaptation, even if carbon source is added, their metabolic reaction will still be slow. ORP will drop slowly, with a small absolute value of the slope.

[0065] After adding carbon source, ORP value drops, indicating that the carbon source has been accepted and a denitrification environment has been established; the slope of ORP drop quantitatively reflects the instantaneous activity and metabolic intensity of microorganisms, a large slope indicates high activity, which can increase subsequent carbon source addition; a small slope indicates low activity, which should reduce or maintain subsequent carbon source addition to avoid waste.

[0066] In the aforementioned scheme, the reference carbon source addition rate is calculated based on mass balance, which assumes that the denitrification rate of denitrifying bacteria is constant; however, denitrification is a biological enzyme-catalyzed process, whose rate strongly depends on temperature: at low temperature, microbial metabolism is slow, if the original rate is still added, the carbon source will accumulate in the system due to the inability to be consumed in time, leading to excessive COD in effluent and resource waste. Conversely, at high temperature, microbial metabolism is fast, and the original addition rate may not meet its needs, limiting the denitrification efficiency.

[0067] To solve the problem, the embodiment of the application introduces a temperature correction function based on microbial dynamics to dynamically calibrate the static reference carbon source dosage rate to match the current wastewater temperature. Specifically, the reference carbon source dosage rate is corrected according to the influence of the wastewater temperature on the denitrification rate to obtain an initial carbon source dosage rate, including: introducing a temperature correction function, calculating a temperature correction factor corresponding to the wastewater temperature according to the temperature correction function ; calculating the product of the reference carbon source dosage rate and the temperature correction factor as the initial carbon source dosage rate; wherein the temperature correction function is: ; η represents the temperature correction factor; T is the measured wastewater temperature; Tref is the reference temperature, which is 30℃-35℃; θ represents the denitrification temperature coefficient, which is 1.02-1.08, indicating that the rate of denitrification biochemical reaction is increased by 1.02 to 1.08 times for every 1℃ increase in wastewater temperature compared to the reference temperature.

[0068] In a specific application scenario, Tref is the reference temperature, which is 30℃-35℃, which is the optimal temperature range for denitrifying bacteria, and at this temperature, the metabolic activity of microorganisms is the highest. Therefore, this point is set as the correction reference point (i.e. at this time, the temperature correction factor η=1), which means that the system control goal is to make the carbon source dosage at different temperatures achieve the same treatment effect as at the optimal temperature.

[0069] When the wastewater temperature is less than the reference temperature, η is less than 1, the reference carbon source dosage rate is adjusted lower to match the slower denitrification rate; when the wastewater temperature is equal to the reference temperature, η is equal to 1, the reference carbon source dosage rate is maintained; when the wastewater temperature is greater than the reference temperature, η is greater than 1, the reference carbon source dosage rate is adjusted higher to meet the faster denitrification requirement.

[0070] The embodiment of the application breaks the rigid thinking that the dosage rate must be equal to the reaction rate, predicts the influence of the wastewater temperature on the biochemical rate in advance by introducing the temperature correction factor η, and actively adjusts the dosage rate to match, which enables the physical dosage process and the biological reaction process to realize intelligent cooperation, solving the disconnection problem caused by different kinetic characteristics.

[0071] In the foregoing scheme, the reference carbon source addition rate is corrected based on the redox potential drop slope, which is caused by the superposition of two processes: the physical mixing and diffusion process of the carbon source and the biochemical consumption process of the microorganisms; and it is inaccurate to directly attribute the monitored redox potential drop to biochemical action. In a reactor with poor mixing conditions (dead zones, short flow), the carbon source mixing is slow, and even if the microbial activity is high, the overall redox potential drop will be slowed down, and the redox potential drop slope will be underestimated, thereby causing the system to misjudge that the microbial activity is low. The target of the control system is to respond to the physiological state of the microorganisms, rather than the physical structure of the reactor; if the mixing efficiency is low due to poor stirring effect or pool design problems, the control system will incorrectly increase the carbon source addition, which not only cannot solve the problem, but can also cause other problems due to local carbon source excess, which causes the control system to respond to non-target interference, resulting in distorted control instructions.

[0072] To solve this problem, in the embodiment scheme of the present application, the monitoring of the redox potential drop slope includes the following enhanced steps: injecting a tracer into the denitrification zone before adding the carbon source; synchronously monitoring the changes in the tracer concentration and the redox potential; and correcting the interference of the hydraulic mixing effect on the redox potential drop slope through the tracer concentration change curve to obtain the corrected redox potential drop slope.

[0073] By synchronously injecting an inert tracer and monitoring the concentration change curve thereof, a transfer function of the hydraulic mixing characteristics of the reactor is obtained, and this physical transfer model provides a reference frame for the redox potential signal analysis: the actual monitored redox potential change curve is differentially processed with the ideal curve predicted based on the pure mixing model, so as to effectively separate the redox potential change component affected only by the biochemical reaction of the microorganisms; this signal decoupling process essentially creates a virtual biological reactor in which the interference of the physical transmission process is excluded, so that the corrected redox potential slope obtained in the subsequent calculation is a biological signal that only represents the metabolic state of the microorganisms.

[0074] The effluent monitoring point is usually located at the end of the denitrification zone, and from the carbon source addition to the effect thereof on the effluent monitoring value, there is an inherent large time lag related to the hydraulic retention time; if the adjustment is only based on the current instantaneous concentration deviation, the system is always processing past problems; when the effluent exceeds the standard is monitored, the working condition that causes the exceeding may have lasted for a period of time, and the adjustment action is seriously delayed.

[0075] To solve this problem, refer to Figure 5As shown, in the embodiment scheme of the present application, the monitored nitrogen-containing pollutant concentration is compared with the target concentration, and the initial carbon source dosing rate is iteratively adjusted according to the comparison result, including: calculating the difference between the effluent nitrogen-containing pollutant concentration and the target concentration to obtain an instantaneous concentration deviation; analyzing the historical effluent nitrogen-containing pollutant concentration sequence to obtain an effluent nitrogen-containing pollutant concentration prediction value; summing the instantaneous concentration deviation and the effluent nitrogen-containing pollutant concentration prediction value, and inputting the sum into a PID controller to obtain a carbon source dosing rate adjustment amount; and the sum of the carbon source dosing rate adjustment amount obtained in the last control period and the initial carbon source dosing rate at the beginning of the current control period is taken as the actual carbon source dosing rate in the current control period.

[0076] By trend analysis on the historical data sequence, the real change trend of the effluent concentration and the random measurement noise are identified, so that the signal-to-noise ratio of the input to the PID controller is significantly improved, thereby avoiding frequent action of the actuator caused by small fluctuations of the sensor. In addition, the design of applying the adjustment amount of the last period to the current period builds a buffer link in the digital control system, further filters high-frequency interference, and ensures the smoothness of the carbon source dosing rate instruction, thereby providing a stable growth environment for the denitrifying bacteria group.

[0077] On the basis of the above embodiment, before the carbon source is added to the denitrification zone, the pH value and alkalinity of the wastewater to be treated are monitored; if the pH is lower than 6.8 or higher than 7.5, sodium acetate is added as the carbon source; if the pH is greater than or equal to 6.8 and less than or equal to 7.5, the alkalinity is further determined: if the alkalinity is not lower than 200 mg / L, methanol is added as the carbon source; if the alkalinity is lower than 200 mg / L, sodium acetate is added as the carbon source.

[0078] In the embodiment scheme of the present application, a two-level decision mechanism is established to solve the technical problem of coordination between carbon source selection and system stability: the first level takes the pH value range of 6.8-7.5 as the safety boundary of the optimal activity range of denitrifying bacteria, and when the measured pH value exceeds this range, sodium acetate with alkalinity adjustment ability is preferentially selected to actively correct the pH deviation of the system by using its metabolic alkaline production characteristics, which essentially gives the carbon source dosing system a feedforward control function of acid-base balance. The second level decision is activated when the pH value is in the safety range, and whether the alkalinity supplement function of acetate needs to be enabled is determined by judging whether the alkalinity is lower than 200 mg / L, which is a critical buffer capacity threshold: when the alkalinity is sufficient, the system can safely select methanol which is lower in cost; when the alkalinity is close to exhaustion, sodium acetate is switched to prevent potential pH collapse risk.

[0079] Those skilled in the art will appreciate that embodiments of the application can be readily used as software, hardware, or a combination of software and hardware. In a software embodiment, the methods can be tangibly embodied in a machine-readable storage medium having stored thereon instructions that can be used to program a processing system to perform the methods. The program can be embedded in a computer program product which can be a storage medium or a memory, for example, a semiconductor memory, a floppy diskette, a CD-ROM, or an optical storage device, for example, a DVD or a Blu-ray Disc. The storage medium can be loaded into the processing system, for example, a computer, to cause the processing system to perform the methods.

[0080] The present application is described in reference to the drawings, which are as follows: Figure 1 Figure 1 The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks.

[0081] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the functions specified in the flowchart block or blocks. Figure 1 Figure 1 The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks.

[0082] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 Figure 1 The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks.

[0083] Obviously, the embodiments described above are only examples for clarity and are not intended to limit the implementation. For those skilled in the art, other different forms of changes or variations can be made on the basis of the above description. Here, all the embodiments are not required to be exhausted, and the obvious changes or variations derived therefrom are still within the protection scope of the present application.​​​

Claims

1. A method for treating ammonia-nitrogen wastewater with a low carbon-nitrogen ratio, characterized by, The method comprises the following steps: determining the carbon source concentration, ammonia nitrogen concentration, wastewater flow rate and wastewater temperature of the wastewater to be treated; inputting the determined carbon source concentration and ammonia nitrogen concentration into a prediction model to obtain a microbial activity prediction value; determining a carbon source dosage based on the difference between the microbial activity prediction value and an activity threshold value; determining a reference carbon source dosage rate based on the carbon source dosage, and correcting the reference carbon source dosage rate based on the influence of the wastewater temperature on the denitrification rate to obtain an initial carbon source dosage rate; adding carbon source to the denitrification zone at the initial carbon source dosage rate as a control starting point, monitoring the concentration of nitrogen-containing pollutants in the effluent of the denitrification zone, comparing the monitored concentration of nitrogen-containing pollutants with a target concentration, and iteratively adjusting the initial carbon source dosage rate based on the comparison result; wherein determining a reference carbon source dosage rate based on the carbon source dosage comprises: dividing the carbon source dosage into N dosages according to a control time length, and adding the first dosage to the denitrification zone; monitoring the change of the oxidation-reduction potential of the mixed liquor in the denitrification zone to obtain an oxidation-reduction potential drop slope; determining a correction factor for the next control period based on the oxidation-reduction potential drop slope, adjusting the dosage based on the correction factor to obtain the dosage for the next control period; and calculating the reference carbon source dosage rate for the current control period by dividing the dosage for the current control period by the period length; wherein the correction factor is determined based on the following method: ； represents a correction factor; A represents an activity coupling coefficient, and has a value of 0.9 to 1.2; represents a slope of the reduction in oxidation-reduction potential.

2. The low C / N ratio ammonia nitrogen wastewater treatment method according to claim 1, characterized by, The prediction model comprises: an input layer comprising two input nodes that respectively receive the carbon source concentration and the ammonia nitrogen concentration; a feature fusion layer for cross-combining the received carbon source concentration and ammonia nitrogen concentration to obtain an initial feature vector representing the carbon-nitrogen balance relationship; at least one hidden layer, each of which comprises a fully connected layer for performing nonlinear transformation and high-order feature abstraction on the initial feature vector or the output feature of the previous hidden layer to mine the nonlinear mapping relationship between the carbon-nitrogen concentration and the microbial activity, and obtaining a high-dimensional feature vector; an attention layer connected to the last hidden layer for weighting and calibrating the high-dimensional feature vector; an output layer for mapping the calibrated high-dimensional feature vector to the microbial activity prediction value.

3. The low C / N ratio ammonia nitrogen wastewater treatment method according to claim 1 or 2, characterized by, The difference between the microbial activity prediction value and the activity threshold value is defined as an activity deviation Δ, and the carbon source dosage is determined based on the activity deviation, which comprises: determining an adjustment coefficient K based on the predefined interval in which the activity deviation lies; calculating the total nitrogen mass to be removed based on the determined ammonia nitrogen concentration and the target denitrification efficiency, and calculating the reference carbon source dosage required for complete denitrification of the nitrogen mass based on the denitrification biochemical reaction metering relationship; calculating the product of the adjustment coefficient K and the reference carbon source dosage as the carbon source dosage.

4. The low C / N ratio ammonia nitrogen wastewater treatment method according to claim 3, characterized by, The calculation of the total nitrogen mass to be removed based on the determined ammonia nitrogen concentration and the target denitrification efficiency comprises: setting a target total nitrogen concentration for the effluent to meet the standard; subtracting the target total nitrogen concentration from the determined ammonia nitrogen concentration to obtain the total nitrogen concentration to be removed; The total nitrogen mass to be removed is determined according to the wastewater flow and the total nitrogen concentration to be removed, and is calculated as follows: M = C * Q * t Wherein, M represents the total nitrogen mass to be removed; C represents the total nitrogen concentration to be removed; Q represents the wastewater flow; and t represents the control duration.

5. The low C / N ratio ammonia nitrogen wastewater treatment method according to claim 3, characterized by, According to the predefined interval in which the activity deviation is located, a regulation coefficient K is determined, including: Setting a negative deviation threshold Δ1 and a positive deviation threshold Δ2; If Δ ≤ Δ1, then K = K1; wherein K1 is a reinforced dosage coefficient greater than 1; If Δ1 < Δ ≤ 0, then K = K2; wherein 1 ≤ K2 < K1; If 0 < Δ ≤ Δ2, then K = K3; wherein K3 < 1; If Δ > Δ2, then K = K4; wherein K4 < K3.

6. The low C / N ratio ammonia nitrogen wastewater treatment method according to claim 1, characterized in that, According to the influence of the wastewater temperature on the denitrification rate, the reference carbon source dosage rate is corrected to obtain an initial carbon source dosage rate, including: introducing a temperature correction function, and calculating a temperature correction factor corresponding to the temperature of the wastewater according to the temperature correction function ; Calculating the product of the reference carbon source dosage rate and the temperature correction factor as the initial carbon source dosage rate; Wherein, the temperature correction function is: ； represents a temperature correction factor; T is the measured temperature of the wastewater; Tref is a reference temperature, which is 30-35℃; θ represents a denitrification temperature coefficient, which is 1.02-1.08, indicating that the rate of denitrification biochemical reaction is increased to 1.02-1.08 times for each 1℃ increase in wastewater temperature compared with the reference temperature.

7. The low C / N ratio ammonia nitrogen wastewater treatment method according to claim 1, characterized by, The monitoring of the redox potential drop slope includes the following enhancement steps: Before adding the carbon source, a tracer is injected into the denitrification zone; The change of the tracer concentration and the change of the redox potential are monitored synchronously; The interference of the hydraulic mixing effect on the redox potential drop slope is corrected through the tracer concentration change curve to obtain a corrected redox potential drop slope.

8. The low C / N ratio ammonia nitrogen wastewater treatment method according to claim 1, characterized by, The monitored nitrogen-containing pollutant concentration is compared with the target concentration, and the initial carbon source dosage rate is iteratively adjusted according to the comparison result, including: Calculating the difference between the effluent nitrogen-containing pollutant concentration and the target concentration to obtain an instantaneous concentration deviation; Analyzing the historical effluent nitrogen-containing pollutant concentration sequence to obtain an effluent nitrogen-containing pollutant concentration prediction value; Summing the instantaneous concentration deviation and the effluent nitrogen-containing pollutant concentration prediction value, and inputting the summation result into a PID controller to obtain a carbon source dosage rate adjustment amount; The sum of the carbon source dosage rate adjustment amount obtained in the last control cycle and the initial carbon source dosage rate at the start of the current control cycle is taken as the actual carbon source dosage rate in the current control cycle.

9. The low C / N ratio ammonia nitrogen wastewater treatment method according to claim 1, characterized by, Before adding the carbon source to the denitrification zone, the pH value and alkalinity of the wastewater to be treated are monitored; If the pH is lower than 6.8 or higher than 7.5, sodium acetate is added as the carbon source; If the pH is greater than or equal to 6.8 and less than or equal to 7.5, the alkalinity is further determined: If the alkalinity is not lower than 200 mg / L, methanol is added as the carbon source; If the alkalinity is lower than 200 mg / L, sodium acetate is added as the carbon source.

Images