Cascade model-based ammonia nitrogen concentration prediction control method for wastewater treatment

Abstract

The application discloses a cascade model-based ammonia nitrogen edge-cut prediction control method for sewage treatment effluent, which only requires ammonia nitrogen instruments to be installed at the positions of the inlet water and outlet water of multi-stage AO biochemical reaction pools, and does not require ammonia nitrogen instruments to be installed at intermediate links, thereby saving cost; the ammonia nitrogen set values of the first stage and intermediate stage outlet are dynamically adjusted through the inlet water ammonia nitrogen measurement value, so that each stage of the AO biochemical reaction pool can work in an effective control range, and the final outlet water ammonia nitrogen can meet environmental protection requirements; for the first stage and intermediate stage outlet water without ammonia nitrogen instruments, an extended state observer is adopted to estimate the outlet water ammonia nitrogen measurement values of the first stage and intermediate stage, so that the prediction control of each stage of AO is realized; the application strongly correlates the multi-stage AO characteristics of the AO sewage treatment process with the modeling and control of the outlet water ammonia nitrogen, fully considers multi-point inlet water and internal and external reflux, increases the interpretability of the model and the control scheme, facilitates understanding and on-site debugging, and improves implementation efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sewage treatment, in particular to a cascade model-based edge-cut predictive control method for effluent ammonia nitrogen of sewage treatment. BACKGROUND

[0002] In modern sewage treatment process, the outlet ammonia nitrogen concentration is one of the important indicators for measuring water quality. Ammonia nitrogen mainly comes from domestic sewage, industrial wastewater and agricultural runoff, etc. In order to reduce the dosage, biological nitrification and denitrification reactions are often realized by multi-stage AO anoxic-oxic process, so as to realize biological denitrification. Since the ammonia nitrogen concentration sensor is relatively expensive, there are usually only ammonia nitrogen concentration sensors installed at the inlet and outlet of the multi-stage AO, and there is no ammonia nitrogen sensor in the middle-stage oxic tank and anaerobic tank. In practice, the composition of sewage is complex, the influent load fluctuates greatly, and the nonlinear dynamic characteristics of the biological reaction process greatly increase the control difficulty of the multi-stage AO process. In addition, the multi-point influent, multi-point reflux and inlet sewage flow change of the multi-stage AO tank also bring challenges to the control of the outlet ammonia nitrogen. The traditional PID control method is difficult to adjust the PID parameters, and it is also difficult to dynamically optimize the DO (Dissolved Oxygen) set value of the anaerobic tank and the oxic tank of the multi-stage AO. In addition, due to the difficulty of system identification experiment under actual sewage treatment conditions, data that can be used for system identification can only be found under normal operating conditions, which brings difficulties to the relationship between ammonia nitrogen concentration and dissolved oxygen concentration.

[0003] In actual operation, when the aeration amount is insufficient, the effluent ammonia nitrogen concentration value is too large, and the effluent ammonia nitrogen concentration requirement cannot be met. When the aeration amount is too large, the dissolved oxygen concentration is too high, although the effluent ammonia nitrogen concentration can meet the requirement, but it is easy to over-aerate, the energy consumption of the blower is increased, and the control effect of other indicators such as phosphorus removal is affected. The effluent ammonia nitrogen concentration is affected by many factors, and it is difficult to use the mechanism model with numerous parameters and complex form for control. SUMMARY

[0004] The present application aims to overcome the shortcomings of the prior art, and proposes a cascade model-based edge-cut predictive control method for effluent ammonia nitrogen of sewage treatment. The cascade model of dissolved oxygen concentration and effluent ammonia nitrogen concentration is established by data-driven method, which is helpful to dynamically allocate the ammonia nitrogen concentration target value of each stage under the condition that the number of ammonia nitrogen concentration sensors is small, and to give the set value of each DO control loop, so as to realize the whole process ammonia nitrogen control and accurate aeration, and to ensure the aeration quality and reduce the energy consumption of the blower.

[0005] The application is achieved by the following technical scheme: the ammonia nitrogen edge prediction control method based on the cascade model of sewage treatment effluent, which is characterized in that: ammonia nitrogen instruments are installed at the water inlet of the first-stage AO biochemical reaction tank and the water outlet of the last-stage AO biochemical reaction tank in the multi-stage AO biochemical reaction tank, the ammonia nitrogen measurement value of the water inlet of the first-stage AO biochemical reaction tank and the set ammonia nitrogen value of the water outlet of the last-stage AO biochemical reaction tank are obtained, the dissolved oxygen DO measurement instrument and the dissolved oxygen DO concentration control loop are arranged in each-stage AO biochemical reaction tank, the cascade model of the dissolved oxygen DO concentration and the effluent ammonia nitrogen concentration is established, the effluent ammonia nitrogen concentration of the first-stage AO biochemical reaction tank and the intermediate-stage AO biochemical reaction tank is dynamically adjusted according to the extended state observer, and the set value of the dissolved oxygen DO concentration control loop in each-stage AO biochemical reaction tank is further calculated, so that the whole-process ammonia nitrogen control and accurate aeration are realized.

[0006] Further, the method comprises the following steps:

[0007] S1, a data-driven cascade static model of ammonia nitrogen-DO concentration of multi-stage AO is established, and the set value of the effluent ammonia nitrogen of the first-stage AO biochemical reaction tank and the intermediate-stage AO biochemical reaction tank is calculated;

[0008] S2, the parameters of each-stage model are identified, the cascade model from the first-stage AO to the intermediate-stage AO and from the intermediate-stage AO to the last-stage AO is established, and the soft measurement value of the hierarchical equivalent ammonia nitrogen is introduced into each-stage AO biochemical reaction tank according to the multi-point water inlet, internal reflux and external reflux operation mechanism of the AO process;

[0009] S3, an extended state observer taking the dissolved oxygen DO as the input and the effluent ammonia nitrogen concentration as the output is established for each-stage AO, the estimated value of the ammonia nitrogen concentration of the water outlet of the first-stage AO and the intermediate-stage AO is obtained, the estimation of the ammonia nitrogen concentration of the intermediate-stage AO is realized when there is no ammonia nitrogen sensor in the intermediate-stage AO, and the ammonia nitrogen concentration value is used for the ammonia nitrogen concentration control of the current-stage AO;

[0010] S4, according to the set value of the effluent ammonia nitrogen concentration of each-stage AO and the estimated value of the ammonia nitrogen concentration of the water inlet and the water outlet of each-stage AO, the controller and the control algorithm thereof are designed through the cascade model, the set value of the dissolved oxygen DO concentration of each-stage AO is updated in real time, and the aeration air volume is adjusted according to the difference between the set value and the measurement value of the current-stage DO concentration, so that the accurate aeration is realized.

[0011] Further, the step S1 comprises:

[0012] The multi-stage AO process employs a three-stage AO process. The wastewater to be treated sequentially passes through an anaerobic tank, a first-stage anoxic tank, a first-stage aerobic tank, a facultative zone, a second-stage anoxic tank, a second-stage aerobic tank, a third-stage anoxic tank, a third-stage aerobic tank, and a secondary sedimentation tank. The second-stage anoxic tank is an intermediate-stage anoxic tank, the second-stage aerobic tank is an intermediate-stage aerobic tank, the third-stage anoxic tank is a final-stage anoxic tank, and the third-stage aerobic tank is a final-stage aerobic tank, forming a three-stage AO treatment for the wastewater. The anaerobic tank, facultative zone, and third-stage anoxic tank all have inlets. Ammonia nitrogen meters are installed at the inlet of the first-stage anoxic tank and the outlet of the third-stage aerobic tank. Dissolved oxygen (DO) meters are installed in each aerobic tank. Blowers control the aeration of the aerobic tanks through aeration pipes and valves, forming a dissolved oxygen (DO) concentration control loop.

[0013] The ammonia nitrogen-DO concentration data of the multi-stage AO drive the cascade static model as shown in formula (1):

[0014]

[0015] In this model, the static model refers to the relationship between ammonia nitrogen concentration and dissolved oxygen (DO) when both ammonia nitrogen and DO concentrations are stable. y0 is the ammonia nitrogen measurement value at the inlet of the first-stage AO; y1 is the ammonia nitrogen measurement value at the outlet of the first-stage AO; y2 is the ammonia nitrogen measurement value at the outlet of the second-stage AO; y3 is the ammonia nitrogen measurement value at the outlet of the third-stage AO, which is the final ammonia nitrogen control target at the outlet of the entire AO process; δ1, δ2, and δ3 are the ammonia nitrogen reduction values ​​in each stage of the AO tank; β1, β2, and β3 are the distribution ratio coefficients of the upstream wastewater inflow in each stage of the AO, and β1 + β2 + β3 = 1.

[0016] The setpoint for ammonia nitrogen at the effluent outlet of the first-stage AO biochemical reactor is y. 1SV As in formula (2a):

[0017]

[0018] The setpoint for ammonia nitrogen at the effluent outlet of the second-stage AO biochemical reactor is y. 2SV As in formula (2b):

[0019]

[0020] Among them, y 3SV The target for ammonia nitrogen at the process outlet; The ammonia nitrogen levels decrease in three stages, and the values ​​are the same. This can be derived from (1). make w1>1 and w2>1 are the weighting coefficients for the decrease in ammonia nitrogen in the first two stages.

[0021] Furthermore, step S2 includes:

[0022] The models of the AO biochemical reaction tanks at different levels are considered to be consistent in structure, but the parameters of the models at different levels are different, and the parameters are obtained by identification when the data of ammonia nitrogen and chemical oxygen demand (COD) at the inlet are stable; it is assumed that the flow rates at the inlet and outlet of each AO biochemical reaction tank are the same, and the water in the AO biochemical reaction tank is fully mixed, i.e., the concentration in the tank is equal to the concentration at the outlet; according to the operation mechanism of the AO process, such as multi-point water inlet, internal reflux and external reflux, the predicted values of the soft measurement values y1 and y2 of the hierarchical equivalent ammonia nitrogen are introduced and

[0023] The equations of the ammonia nitrogen concentration variation rates of the three AO biochemical reaction tanks are formula (3a), (3b) and (3c) respectively:

[0024]

[0025]

[0026] wherein, T s is a sampling period, T s is 30 minutes; V1, V2 and V3 respectively represent the volumes of the water in the aerobic tanks at different levels, which are calculated according to the liquid level and the geometric size of the AO biochemical reaction tank; Q in represents the flow rate at the inlet, Q out represents the flow rate at the outlet of the third aerobic tank; K NH is the half-saturation constant of ammonia nitrogen, and the initial value is 1; K OA is the half-saturation constant of oxygen for autotrophic bacteria, and the initial value is 0.4, and the initial value is iteratively identified; k1, k2 and k3 are process gain coefficients, which need to be identified; d 10 , d 20 and d 30 are the time delays of the multi-point inlet water to the first, second and third levels respectively, d 21 is the time delay of the outlet water of the first level to the outlet water of the second level, and d 32 is the time delay of the outlet water of the second level to the outlet water of the third level, and the time delay parameters are calculated according to the flow rate and the geometric size of the AO biochemical reaction tank;

[0027] The DO excitation signals are input into the aerobic tanks step by step, and the values of the inlet ammonia nitrogen, the dissolved oxygen (DO) in the aerobic tank and the outlet ammonia nitrogen when the DO excitation signals are input into each level are measured, so as to obtain the model parameters of DO and ammonia nitrogen at each level; the DO excitation signals are input into the aerobic tanks step by step: the DO amplitude changes by 1 mg / L, the period T p1 of the third aerobic tank is 120 minutes, and the periods T p2360 minutes; wherein, in the parameter identification, the DO of the second stage aerobic tank remains unchanged when the third stage aerobic tank is subjected to DO excitation, and the DO of the first stage aerobic tank can change when the first stage aerobic tank is subjected to DO excitation due to the relatively long distance between the first stage aerobic tank and the third stage aerobic tank; in the parameter identification, the DO of the second stage aerobic tank remains unchanged when the third stage aerobic tank is subjected to DO excitation, and the DO of the first stage aerobic tank can change; in the parameter identification, the DO of the first stage aerobic tank remains unchanged when the second stage aerobic tank is subjected to DO excitation, and the DO of the third stage aerobic tank is increased by 1 mg / L and remains unchanged; in the parameter identification of the first stage aerobic tank, the DO of the second stage aerobic tank and the third stage aerobic tank is increased by 0.5 mg / L, so as to ensure that the output ammonia nitrogen reaches the target requirement in the identification process;

[0028] The to-be-identified parameters are The parameter model of the third stage aerobic tank is formula (4),

[0029]

[0030] In formula (5), N3 is the number of samples in the third stage test, and the optimization index of the third stage aerobic tank is formula (5)

[0031]

[0032] Through the optimization algorithm, the parameters in formula (5) are obtained, and then the estimated value of the output ammonia nitrogen of the second stage aerobic tank is inferred by measuring the output ammonia nitrogen of the third stage aerobic tank After the parameter estimation of the third stage model is completed, the first stage model and the second stage model use the same K OA and K NH , then the first stage model and the second stage model only estimate single parameters k1 and k2, and in the parameter identification of the second stage model, the estimated value of the output ammonia nitrogen of the second stage aerobic tank is inferred by measuring the output ammonia nitrogen of the third stage aerobic tank As shown in formula (6):

[0033]

[0034] In the parameter identification of the second stage model, the second stage data needs to be moved forward as a whole by d 32 T S time, as shown in formula (7):

[0035]

[0036] And the objective function of the second stage is shown in formula (8):

[0037]

[0038] In the formula, N2 is the number of samples in the third-level test. According to the structure of formula (11) and the same identification method as the third level, the samples are compared with the current time k before d. 32 The sampling data at each sampling time point, that is, the identification Following the same method as the second-stage model parameter identification, the output ammonia nitrogen of the first-stage aerobic tank is first estimated by estimating the output ammonia nitrogen of the second-stage aerobic tank, and then the first-stage model parameters are obtained through single-parameter identification.

[0039] Based on the second-level model, the estimated ammonia nitrogen value at the outlet of the first-level aerobic tank is obtained as shown in formula (9):

[0040]

[0041] From the sampling time of the third-level output, we need to move forward (d) 32 +d 21 )T S At that moment, the first-level model is identified as formula (10):

[0042]

[0043] The objective function for the first level is shown in formula (11):

[0044]

[0045] Furthermore, step S3 includes:

[0046] In real-time control, the ammonia nitrogen concentration in the intermediate-stage AO biochemical reactor without an ammonia nitrogen sensor is estimated using an extended state observer; based on the measured value y0(kd 32 -d 21 ) and y3(k), the observed value of y3(k) is obtained through the extended state observer, z3(k), and the observed value of ammonia nitrogen in the effluent of the first-stage aerobic tank is z1(kd). 32 -d 21 ), kd 32 The observed ammonia nitrogen value z2(kd) in the effluent of the second-stage aerobic tank at time [time]. 32 Based on this, the model recursive formula is used to derive kd 32 -d 21 Moment Obtain time k From kd 32 Moment up to time k Finally based on and The real-time process control of the intermediate stage AO biochemical reaction pool is realized; and the expansion state observer of the three-stage AO biochemical reaction pool is shown in formulas (12a), (12b) and (12c):

[0047]

[0048] Wherein, l1, l2 and l3 are the gains of the first stage, the second stage and the third stage observer respectively, and the values thereof are selected to ensure the stable convergence of formulas (13a) and (13b);

[0049] The first stage soft measurement recursive formula (13a) is:

[0050]

[0051] Wherein, i=-d 32 -d 21 ,-d 32 -d 21 +1,-d 32 -d 21 +2,...,-1;

[0052] The second stage soft measurement recursive formula (13b) is:

[0053]

[0054] Wherein, i=-d 32 ,-d 32 +1,s-d 32 +2,...,-1.

[0055] Further, the step S4 comprises:

[0056] The controller and the control algorithm thereof are designed through the prediction control of the cascade model, the target function and the constraint condition are set for each stage aerobic tank, the cascade model prediction control law x MPC (k) is obtained by optimization solution, the first element in the control law x MPC (k) represents the dissolved oxygen set value at the moment, the dissolved oxygen set value is updated every T s , at this time, the dissolved oxygen concentration of each stage aerobic tank is controlled to track the given dissolved oxygen set value by adjusting the air volume of the air blower and the valve opening, so that the energy-saving control of the effluent ammonia nitrogen concentration of the constrained wastewater treatment process is realized;

[0057] The target function of the first stage aerobic tank is formula (14a):

[0058]

[0059] The target function of the second stage aerobic tank is formula (14b):

[0060]

[0061] The objective function of the third-stage aerobic tank is formula (14c):

[0062]

[0063] Wherein, N p is the prediction time domain, N c is the control time domain, the recursive prediction model of the first stage is formula (13a), the recursive prediction model of the second stage is formula (13b), is the predicted value of the third-stage effluent ammonia nitrogen, and the recursive prediction model is formula (15):

[0064]

[0065] The numerical value of the dissolved oxygen change range and the effluent ammonia nitrogen is constrained, as formula (16):

[0066]

[0067]

[0068] Wherein, DO min is the lower limit of the DO set value, DO max is the upper limit of the DO set value, NH 4max is the upper limit of the effluent ammonia nitrogen set value.

[0069] A non-transitory computer readable medium storing instructions, characterized in that when the instructions are executed by a processor, the steps of the cascade model-based sideband prediction control method for effluent ammonia nitrogen of sewage treatment are performed.

[0070] A computing device comprising a processor and a memory for storing processor-executable programs, characterized in that when the processor executes the programs stored in the memory, the cascade model-based sideband prediction control method for effluent ammonia nitrogen of sewage treatment is implemented.

[0071] Compared with the prior art, the present application has the following advantages and beneficial effects:

[0072] 1、The present application strongly correlates the multi-stage AO characteristics of the AO sewage treatment process with the modeling and control of effluent ammonia nitrogen, and fully considers multi-point influent and internal and external reflux, thereby increasing the interpretability of the model and the control scheme, facilitating engineers to understand and on-site debugging, and improving the implementation efficiency.

[0073] 2. The intermediate effluent ammonia nitrogen set value adopts a dynamic adjustment method, fully considers the fluctuation of the first stage input ammonia nitrogen, realizes the indirect feedforward control of ammonia nitrogen control, improves the response time of the ammonia nitrogen controller, and dynamically adjusts the effluent ammonia nitrogen set value of the first two stages, which also fully plays the biological denitrification performance of each stage, and reduces the chemical input.

[0074] 3. The present application uses an extended state observer to realize the estimation of the effluent ammonia nitrogen of the first two stages, and uses a soft measurement recursive method to obtain the real-time estimated value of the first and second stages, which reduces the expensive ammonia nitrogen measuring instrument and reduces the engineering implementation cost. BRIEF DESCRIPTION OF DRAWINGS

[0075] Figure 1 It is a multi-stage AO sewage treatment process schematic diagram of the present application.

[0076] Figure 2 It is a multi-stage AO sewage treatment ammonia nitrogen control logic block diagram.

[0077] Figure 3 It is a parameter identification schematic diagram of the cascade model of the multi-stage AO sewage treatment.

[0078] Figure 4 It is a DO excitation waveform diagram in the parameter identification of the cascade model.

[0079] Figure 5 It is a flow chart of the intermediate pool ammonia nitrogen concentration estimation method of the multi-stage AO sewage treatment.

[0080] Figure 6 It is a DO set value model predictive control block diagram of the multi-stage AO sewage treatment. DETAILED DESCRIPTION

[0081] The present application will be further described below in combination with specific embodiments.

[0082] Example 1

[0083] The cascade model-based sewage treatment effluent ammonia nitrogen card edge predictive control method provided in the embodiment only needs to arrange ammonia nitrogen instruments at the water inlet and outlet of the multi-stage AO sewage treatment process, dynamically adjusts the ammonia nitrogen set value of the first and intermediate stages through the water inlet ammonia nitrogen value, makes each stage reaction tank run within the effective control range, and ensures that the effluent meets the standard. In view of the problem of no instrument measurement in the intermediate stage, process modeling, observer estimation of intermediate pool variables, and predictive control of each stage AO tank are used to realize the ammonia nitrogen control of the entire cascade sewage treatment.

[0084] In the actual sewage treatment process, in order to consider the removal efficiency of carbon, nitrogen and phosphorus, and control the construction cost, a three-stage AO process is usually used. Referring to Figure 1As shown, taking a three-stage AO process as an example, a schematic diagram of the three-stage AO wastewater treatment process is given. The entire process includes the wastewater to be treated entering the anaerobic tank, the first-stage anoxic tank, the first-stage aerobic tank, the facultative zone, the second-stage anoxic tank (i.e., the intermediate-stage anoxic tank), the second-stage aerobic tank (i.e., the intermediate-stage aerobic tank), the third-stage anoxic tank (i.e., the final stage of the multi-stage process, or the third stage when there are only three stages), and the third-stage aerobic tank (i.e., the final-stage aerobic tank). Finally, it enters the secondary sedimentation tank and becomes AO-treated wastewater. After subsequent disinfection and other processes, it meets the discharge standards and is discharged into the river. This multi-stage AO process features multi-point influent, with influent entering from the anaerobic tank, facultative zone, and third-stage anoxic tank. Approximately 40% of the influent volume enters the anaerobic tank, 30% from the facultative zone, and 30% from the third-stage anoxic tank. Sludge from the secondary sedimentation tank is externally returned to the anaerobic tank, while the second and third-stage aerobic tanks have internal recirculation to the facultative zone. The multi-point influent, external recirculation, and internal recirculation utilize a fixed flow rate. In the actual process, ammonia nitrogen meters are installed at the inlet of the first-stage anoxic tank and at the outlet of the third-stage aerobic tank. Dissolved oxygen (DO) meters are installed in each aerobic tank. Blowers control the aeration of the aerobic tanks and regulate DO concentration via aeration pipes and valves.

[0085] Figure 2 This is a block diagram of the ammonia nitrogen control logic for multi-stage AO wastewater treatment. Each of the three stages has an AO control system, which is a cascade control system. The outer loop is the ammonia nitrogen control loop, and the inner loop is the DO control loop. The diagram shows the DO control loop and the AO process in one block diagram. Valve adjustment is implemented in the DO control loop. Figure 2 The three cascade control systems correspond to Figure 1 The three-stage AO wastewater treatment process and cascaded control scheme make it easy for on-site engineers to understand and operate, overcoming the problem of poor interpretability when using only a single control system throughout the entire process.

[0086] To achieve Figure 2 For three-stage AO ammonia nitrogen control, the ammonia nitrogen setpoint for each stage must first be given. The ammonia nitrogen setpoint for each stage is determined by... Figure 2 The algorithm module for setting the ammonia nitrogen at the third-stage outlet is implemented. Let y 0PV The measured ammonia nitrogen value at the first inlet; y 1PV and y 1SV The measured and set values ​​of ammonia nitrogen at the first-stage AO outlet; y 2PV and y 2SV For the measured and set values ​​of ammonia nitrogen at the secondary AO outlet; y 3PV and y 3SV The measured and set values ​​of ammonia nitrogen at the third-level AO outlet are given, where y 3SVThis refers to the final ammonia nitrogen target for the entire OA process, controlling the outlet ammonia nitrogen. Let the ammonia nitrogen reduction values ​​for the three stages be δ1, δ2, and δ3, respectively; and let the distribution ratio coefficients of the influent flow rate in the three stages be β1, β2, and β3, where β1 + β2 + β3 = 1. In a stable wastewater treatment process, the effluent and influent flow rates of each tank are approximately equal at the same time, allowing us to approximate the static relationship of the cascaded ammonia nitrogen concentration. This is equivalent to the end of the dynamic control process, where the measured ammonia nitrogen value output from each stage is approximately equal to the set value.

[0087]

[0088] From (1),

[0089] y 0PV -y 3SV =β1δ1+(β1+β2)δ2+δ3. (2)

[0090] If the amount of ammonia nitrogen decreases in all three stages is the same, that is Then there is

[0091]

[0092] Since the third-stage AO biochemical reactor has the fastest response time to the output ammonia nitrogen, in order to ensure that the third stage is always in the rapid reaction phase and that the outlet ammonia nitrogen concentration is below the target value, the ammonia nitrogen concentration decrease of the first two stages is taken as greater than the average ammonia nitrogen decrease concentration.

[0093]

[0094] Where w1>1 and w2>1 are the weighting coefficients for the decrease in ammonia nitrogen in the first two stages. Taking w1 = 1.2 and w2 = 1.1, then we have

[0095]

[0096] and

[0097]

[0098] Among them, y 0PV This is the measured value of inlet ammonia nitrogen concentration, y 3SV The target ammonia nitrogen at the process outlet consists of two known process variables. Figure 2 The algorithm module for setting the ammonia nitrogen at the three-stage outlet is given. It calculates the ammonia nitrogen setting values ​​for the first and second stages based on equations (5) and (6).

[0099] Figure 2In the ammonia nitrogen control of the multi-stage AO sewage treatment shown, no expensive ammonia nitrogen instrument is installed at the first-stage and second-stage effluent outlets, and the estimation needs to be made from the ammonia nitrogen instruments at the process inlet and outlet. In order to realize the estimation of the ammonia nitrogen of the first-stage and second-stage AO biochemical reaction pool effluent, a data-driven cascade model with explainability conforming to the process requirements needs to be established.

[0100] Let y0, y1, y2 and y3 be the ammonia nitrogen measurement values of the first-stage inlet, the first-stage effluent, the second-stage effluent and the third-stage effluent respectively, which are equal to y 0PV , y 1PV , y 2PV and y 3PV in numerical value, for simplifying the formula representation. x 1PV and x 1SV , x 2PV and x 2SV , x 3PV and x 3SV are the measurement values and set values of the dissolved oxygen DO in the first-stage, second-stage and third-stage aerobic tanks respectively, and the measurement values of the dissolved oxygen DO are abbreviated as x1, x2 and x3. Since the three-stage sewage treatment adopts the same process, the three-stage AO process adopts the same model, but has different model parameters. The ammonia nitrogen concentration change rate equations of the three AO biochemical reaction pools are respectively:

[0101]

[0102] Among them, T s is the sampling period, and in the ammonia nitrogen control system, T s is taken as 30 minutes; V1, V2 and V3 respectively represent the volumes of the water in the aerobic tanks of each stage, which can be calculated according to the liquid level and the geometric size of the AO tank; Q in represents the inlet flow, Q out represents the effluent flow of the third aerobic tank; K NH is the ammonia nitrogen half-saturation constant, and the initial value is taken as 1; K OA is the half-saturation constant of the oxygen for the autotrophic bacteria, and the initial value is taken as 0.4, and iterative identification is made on the basis of the initial value; k1, k2 and k3 are process gain coefficients, which need to be identified.d 10 , d 20 and d 30 are the delays of the multi-point inlets to the first stage, the second stage and the third stage respectively, d 21 is the delay of the first-stage effluent to the second-stage effluent, and d 32 is the delay of the second-stage effluent to the third-stage effluent, and the delay parameters can be calculated according to the flow rate and the geometric size of the AO biochemical reaction tank.

[0103] k1, k2 and k3 in formula (7a), (7b) and (7c) need to be identified, refer to Figure 3 the system parameter identification process shown in the figure, Figure 4 is the DO excitation signal used in identification. Figure 4 The amplitude of the excitation signal DO of the third stage is changed by 1 mg / L, and the period T P3 of the third stage is 120 minutes, and the periods T P2 and T P1 of the second and first stages are 360 minutes; during the identification of parameters, when the third stage is excited by DO, the second stage keeps the DO unchanged, and the first stage can change the DO; when the second stage is excited by DO, the first stage keeps the DO unchanged, and the third stage increases the DO by 1 mg / L and keeps it unchanged; when the first stage is excited by DO, the second and third stages increase the DO by 0.5 mg / L, to ensure that the output ammonia nitrogen reaches the target requirement during the identification process. Figure 4 During the three-stage excitation process of T Figure 3 , the flow rate, liquid level, influent ammonia nitrogen, DO of the aerobic tank and effluent ammonia nitrogen of the process are measured, refer to , to identify the parameters of the models (7a), (7b) and (7c) in three stages.

[0104] Figure 3 Figure 3 is the system parameter identification subroutine flow chart, in the initialization of the main program, the identification state IdentState is initialized to 0, that is, no system parameter identification is performed; when the system parameter identification needs to be performed, IdentState is set to 3, to start a system parameter identification, and when IdentState is not 0, the main program periodically calls the system parameter identification subroutine of .

[0105] Figure 3In Step 31, the value of IdentState is judged, if IdentState is 3, Step 311 is entered to perform the third level system parameter identification; if IdentState is not 3, Step 32 is entered. In Step 32, whether IdentState is 2 is judged, if it is 2, Step 321 is entered to perform the second level system parameter identification; if IdentState is not 2, Step 33 is entered. In Step 33, whether IdentState is 1 is judged, if it is 1, Step 331 is entered to perform the first level system parameter identification; if IdentState is not 1, the identification subroutine returns to the main program; after each level of identification is completed, the main program is also returned. In Step 311, the third level system parameter identification is performed, DO3 is performed for three complete square wave changes, data collection and alignment are completed, model time parameters and process gain are identified, after completion, IdentState = 2, preparing for the second level parameter identification. In Step 321, the second level system parameter identification is performed, DO2 is performed for one complete square wave change, data collection and alignment are completed, model process gain is identified, after completion, IdentState = 1, preparing for the next first level parameter identification. In Step 331, the first level system parameter identification is completed, DO1 is performed for one complete square wave change, data collection and alignment are completed, process gain is identified, after completion, IdentState = 0, ending the three-level AO tank parameter identification.

[0106] For the third level aerobic tank, under the condition that the ammonia nitrogen in the influent is relatively stable and the DO concentration of the second level aerobic tank is basically unchanged, the corresponding y2 is basically unchanged, y0 can be measured, therefore, by changing the DO concentration of the third level aerobic tank and measuring the effluent ammonia nitrogen value, the parameters K NH , K OA and k3 of model (7c) are identified.

[0107]

[0108] The parameters K NH , K OA and k3 of model (7c) are identified by using the objective function

[0109]

[0110] In formula (8), N3 is the sample number in the third level test, according to the structural characteristics of formula (8), K OA and K NH are monotonic functions relative to k3, K OA and K NHsensitivity of the parameters in turn, and iteratively identify these parameters. The first sub-step iteration, K OA and K NH The initial value of K and The first estimate of k3 is obtained by the least square method using the sample data and the objective function (24) as The second sub-step iteration, and The estimate of K OA is obtained by the least square method using the objective function (24) as The third sub-step iteration, and The estimate of K NH is obtained by the least square method using the objective function (9) as The third sub-step iteration, and The estimate of k3 is obtained by the least square method using the objective function (24) as So far and are the third-level model parameters. This method makes full use of the process initial parameters and to improve the robustness of the identified parameters, and the algorithm requires less computing power and can run in real time.

[0111] After completing the estimation of the third-level model parameters, the first and second levels use the same K OA and K NH , so the first and second levels only need to estimate a single parameter k1 and k2, reducing the requirement for the change range of the sample data. After mixing the third level, the change range of the first and second levels is reduced. When performing the second-level identification test, the estimated value of the second-level output ammonia nitrogen is inferred from the third-level output ammonia nitrogen measurement data

[0112]

[0113] When performing the second-level parameter identification, the second-level measurement data needs to be moved forward by d 32 T S time,

[0114]

[0115] and the objective function of the second level

[0116]

[0117] where N2 is the sample number in the third stage test, according to the structure of formula (11), according to the same identification method of the third stage, through the sampling data of the current time k-d 32 , the output ammonia nitrogen of the second stage is estimated to obtain the output ammonia nitrogen estimation of the first stage, and then the single-parameter identification is used to obtain the first stage model parameter According to the second stage model, the first stage model parameter is obtained

[0118] According to the second stage model, the first stage model parameter is obtained

[0119]

[0120] From the sampling time of the third stage output, move forward (d 32 +d 21 )T S , the identification model of the first stage is

[0121]

[0122] and the target function of the first stage

[0123]

[0124] After completing the three-stage model identification, in real-time online control, through the three-stage model and the input ammonia nitrogen and the third stage output ammonia nitrogen and the dissolved oxygen DO value of each stage, the first stage and the second stage outlet ammonia nitrogen value can be estimated by the observer.

[0125] The ammonia nitrogen concentration estimation method of the intermediate tank is referred to Figure 5 , and the specific steps are as follows:

[0126] Step 51, calculate the effluent ammonia nitrogen observation value z3(k) of the third stage aerobic tank. Denote the current time as k, for the third stage aerobic tank, since the effluent ammonia nitrogen can be directly measured, for the measurement value at time k-1, the effluent ammonia nitrogen observation value z3(k) of the third stage aerobic tank at time k is

[0127]

[0128] Wherein, z2 is the observation value of the observer of the second stage outlet ammonia nitrogen, z3 is the predicted value of the third stage outlet ammonia nitrogen, and l3 is the gain of the third stage observer error compensation term.

[0129] Step 52, calculate the effluent ammonia nitrogen observation value z1(k-d 32 -d 21 ) of the first stage aerobic tank at the k-d 32 -d 21 th sampling time and k-d32 The observed ammonia nitrogen value z2(kd) in the effluent of the second-stage aerobic tank at time [time]. 32 ).

[0130] Using the measured value y3(k) of the effluent ammonia nitrogen from the third-stage AO biological reactor and the observed value z3(k) of the effluent ammonia nitrogen from the third-stage AO biological reactor, kd was observed. 32 -d 21 The observed value of ammonia nitrogen in the effluent of the first-stage aerobic tank at time z1(kd) 32 -d 21 )for

[0131]

[0132] Where z1 is the observed value of the first-stage outlet ammonia nitrogen, and l1 is the gain of the error compensation term of the second-stage observer.

[0133] For the second-stage aerobic tank, the observed kd 32 The observed ammonia nitrogen value z2(kd) in the effluent of the second aerobic tank at time [time]. 32 )for

[0134]

[0135] Where l1 is the gain of the first-stage observer error compensation term.

[0136] Step 53, predict the ammonia nitrogen value of the effluent from the first-stage aerobic tank at the current moment. Using model (7a), kd 32 -d 21 Substituting the relevant variables up to time k, we can recursively derive kd. 32 -d 21 From time +1 to time k The predicted ammonia nitrogen value of the effluent from the second-stage aerobic tank at the current moment. Using model (7b), kd 32 Substituting the relevant variables up to time k, we can recursively derive k. s -d 32 From time +1 to time k Used as initial values ​​for model predictive control.

[0137] For the current moment Using model (7a), kd 32 -d 21 Substituting the relevant variables up to time k, we can recursively derive kd. 32 -d 21 From time +1 to time k The initial values ​​used as model predictive control are given by the recursive formula:

[0138]

[0139] where i = -d 32 -d 21 -d 32 -d 21 -d 32 -d 21 +2,...,-1.

[0140] Step 54, for the current time Using the form of model (22b), substitute the related variables from k-d 32 to k time into, recursively derive the k s -d 32 +1 to k time Use as the initial value of model predictive control, the recursive formula is

[0141]

[0142] where i = -d 32 -d 32 +1,s-d 32 +2,...,-1.

[0143] Step 55, model predictive control is carried out for k+1 time.

[0144] The model predictive control block diagram of DO set value of multi-stage AO sewage treatment is shown in Figure 6 First, set the objective function and constraint condition for each stage of aerobic tank; take the objective function, the target is the value of dissolved oxygen and the sum of ammonia nitrogen at each stage outlet and the set value of target ammonia nitrogen is as small as possible.

[0145] For the first stage of aerobic tank, set the objective function

[0146]

[0147] where N p is the prediction time domain, N c is the control time domain, and the recursive prediction model of soft measurement is formula (17).

[0148] For the second stage of aerobic tank, the prediction model is formula, and the objective function is set as

[0149]

[0150] The recursive prediction model of soft measurement is formula (18).

[0151] For the third stage of aerobic tank, set the objective function

[0152]

[0153] wherein, is the predicted value of the third stage effluent ammonia nitrogen, and the recursive prediction model is

[0154]

[0155] With the constraints of the range of dissolved oxygen variation and the value of effluent ammonia nitrogen, there are

[0156]

[0157] wherein, DO min is the lower limit of the DO set value, DO max is the upper limit of the DO set value, NH 4max is the upper limit of the NH3 set value. The purpose of setting DO min and DO max is to make the output of the solver meet the constraints of the actual working conditions, and the purpose of setting NH 4max is to make the solver solve the DO set value that meets the effluent index requirement of ammonia nitrogen, even if the DO concentration is small. As can be seen from the performance index objective function of formula (19) and formula (21), the obtained DO set value is to make the ammonia nitrogen less than the set value but as close to the set value as possible, to realize the lower limit of the edge control, which can meet the control target and minimize the energy consumption. After rolling optimization, the set value x MPC (k) of each stage of DO is obtained. 1SV (k), x 2SV (k), x 3SV (k)] T ; the optimization algorithm fmincon function is used to solve the constraint optimization problem composed of the dissolved oxygen value of each stage at each time and the objective function and constraint condition respectively, and the model predictive control law is obtained:

[0158]

[0159] wherein, arg min represents calculating the decision variable that makes the objective function minimum.

[0160] Figure 6 After rolling optimization in the middle, the first element in the control law x MPC (k) is taken, which represents the dissolved oxygen set value at this time. The dissolved oxygen set value is updated every T s , at this time, the dissolved oxygen concentration of the aerobic tank is controlled to track the given dissolved oxygen set value by adjusting the air volume of the blower and the valve opening, and the energy-saving control of the effluent ammonia nitrogen concentration of the constrained municipal wastewater treatment process is realized.

[0161] Example 2

[0162] The embodiment discloses a non-transitory computer readable medium storing instructions, when the instructions are executed by a processor, steps of the cascade model based ammonia nitrogen outflow of sewage treatment edge prediction control method according to the embodiment 1 are executed.

[0163] The non-transitory computer readable medium in the embodiment can be a disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), a U disk, a mobile hard disk and the like.

[0164] Embodiment 3

[0165] The embodiment discloses a computing device, comprising a processor and a memory for storing a processor executable program, when the processor executes the program stored in the memory, the cascade model based ammonia nitrogen outflow of sewage treatment edge prediction control method according to the embodiment 1 is realized.

[0166] The computing device in the embodiment can be a desktop computer, a notebook computer, a smart phone, a PDA handheld terminal, a tablet computer, a programmable logic controller (PLC) or other terminal devices with processor functions.

[0167] The above-mentioned embodiments are only the preferred embodiments of the present application, and are not intended to limit the scope of the present application, and any changes made according to the shape and principle of the present application should be covered within the protection scope of the present application.

Claims

1. A method for predictive control of ammonia nitrogen in wastewater effluent based on a cascade model, characterized in that: This method involves installing ammonia nitrogen meters at the inlet of the first-stage AO biological reactor and the outlet of the final-stage AO biological reactor in a multi-stage AO biological reactor. The method acquires the influent ammonia nitrogen value of the first-stage AO biological reactor and sets the effluent ammonia nitrogen value of the final-stage AO biological reactor. Simultaneously, dissolved oxygen (DO) meters and DO concentration control loops are installed in each stage of the AO biological reactor. A cascade model of DO concentration and effluent ammonia nitrogen concentration is established. Based on the expansion state observer, the effluent ammonia nitrogen concentration of the first-stage and intermediate-stage AO biological reactors is dynamically adjusted. The setpoints of the DO concentration control loops in each stage of the AO biological reactor are further calculated, thereby achieving full-process ammonia nitrogen control and precise aeration. This method includes the following steps: S1. Establish a multi-stage AO ammonia nitrogen-DO concentration data-driven cascade static model to calculate the effluent ammonia nitrogen setpoints for the first-stage AO biological reactor and the intermediate-stage AO biological reactor. The multi-stage AO process employs a three-stage AO process. The wastewater to be treated sequentially passes through an anaerobic tank, a first-stage anoxic tank, a first-stage aerobic tank, a facultative zone, a second-stage anoxic tank, a second-stage aerobic tank, a third-stage anoxic tank, a third-stage aerobic tank, and a secondary sedimentation tank. The second-stage anoxic tank is an intermediate-stage anoxic tank, the second-stage aerobic tank is an intermediate-stage aerobic tank, the third-stage anoxic tank is a final-stage anoxic tank, and the third-stage aerobic tank is a final-stage aerobic tank, forming a three-stage AO treatment for the wastewater. The anaerobic tank, facultative zone, and third-stage anoxic tank all have inlets. Ammonia nitrogen meters are installed at the inlet of the first-stage anoxic tank and the outlet of the third-stage aerobic tank. Dissolved oxygen (DO) meters are installed in each aerobic tank. Blowers control the aeration of the aerobic tanks through aeration pipes and valves, forming a dissolved oxygen (DO) concentration control loop. The static model driven by ammonia nitrogen-DO concentration data from multi-stage AO is given by formula (1): (1) The static model refers to the relationship between ammonia nitrogen concentration and dissolved oxygen (DO) when both concentrations are stable. The measured ammonia nitrogen value at the first-stage AO inlet; The measured ammonia nitrogen value at the outlet of the first-stage AO effluent; The measured ammonia nitrogen value at the outlet of the second-stage AO effluent; The ammonia nitrogen measurement value at the outlet of the third-stage AO process is the final ammonia nitrogen control target at the outlet of the entire AO process. , and These represent the decrease in ammonia nitrogen in each AO pool; , and These are the distribution ratio coefficients of the upstream wastewater inflow in each AO level, respectively. =1; Ammonia nitrogen setpoint at the effluent of the first-stage AO biochemical reactor For formula (2a): (2a) Ammonia nitrogen setpoint at the effluent outlet of the second-stage AO biochemical reactor For formula (2b): (2b) in, The target for ammonia nitrogen output from the AO process; = = The ammonia nitrogen levels decrease in three stages, and the values ​​are the same, derived from formula (1). ,make , >1 and >1 represents the weighting coefficient for the decrease in ammonia nitrogen in the first two stages; S2. Identify the parameters of each level of the model, establish a cascade model from the first-stage AO to the intermediate-stage AO and from the intermediate-stage AO to the final-stage AO, and introduce the graded equivalent ammonia nitrogen soft measurement value in each AO biochemical reaction tank according to the multi-point water inlet, internal recirculation and external recirculation operation mechanism of the AO process. The models of each AO biological reactor are considered to have the same structure, but the parameters of each model are different. The parameters are obtained by identifying the influent ammonia nitrogen and chemical oxygen demand (COD) when they are in a stable data range. It is assumed that the influent and effluent flow rates of each AO biological reactor are the same, and that the water in the AO biological reactor is fully mixed, i.e., the concentration in the reactor is equal to the concentration in the effluent. Based on the multi-point influent, internal recirculation, and external recirculation operation mechanism of the AO process, a soft measurement value for the equivalent ammonia nitrogen in stages is introduced. Predicted value as well as Predicted value ; The equations for the rate of change of ammonia nitrogen concentration in the three AO biochemical reactors are (3a), (3b), and (3c), respectively: (3a) (3b) (3c) in, The sampling period is It lasts for 30 minutes; , , These represent the volumes of water in each aerobic tank, calculated based on the liquid level and the geometric dimensions of the AO biological reaction tank. Indicates the inflow rate. This indicates the effluent flow rate of the third aerobic tank; This is the ammonia nitrogen half-saturation constant, with an initial value of 1; It is the oxygen half-saturation constant of autotrophic bacteria, with an initial value of 0.4, and iterative identification is performed based on the initial value; , , It is the process gain coefficient, which needs to be identified; , and To account for the delays in water entering from multiple points to the first, second, and third stages, The time delay between the first stage water outlet and the second stage water outlet. The delay time from the second-stage effluent to the third-stage effluent is calculated based on the flow velocity and the geometry of the AO biochemical reactor. , and The measured values ​​of dissolved oxygen (DO) in the first, second, and third aerobic tanks are respectively. DO excitation signals were sequentially input into each aerobic tank, and the influent ammonia nitrogen, dissolved oxygen (DO) in the aerobic tank, and effluent ammonia nitrogen values ​​were measured at each stage when the DO excitation signal was introduced, thereby obtaining the model parameters for DO and ammonia nitrogen at each stage. The sequential input of DO excitation signals to the aerobic tanks was as follows: the DO amplitude varied by 1 mg / L, and the cycle of the third aerobic tank was... The cycle time is 120 minutes, and the cycle time between the first and second aerobic tanks is... The timeframe is 360 minutes. During parameter identification, when DO stimulation is applied to the third-stage aerobic tank, the DO level in the second-stage aerobic tank remains constant. If the first-stage aerobic tank is far from the third-stage aerobic tank, its DO level can change during DO stimulation. When the third-stage aerobic tank is stimulated, the second-stage aerobic tank maintains its DO level, while the first-stage aerobic tank's DO level can change. When the second-stage aerobic tank is stimulated, the first-stage aerobic tank maintains its DO level, while the third-stage aerobic tank's DO level is increased by 1 mg / L and then remains constant. When the first-stage aerobic tank is stimulated, the DO level in both the second and third-stage aerobic tanks is increased by 0.5 mg / L to ensure that the output ammonia nitrogen meets the target requirements during the identification process. The parameters to be identified are The parameter model for the third-stage aerobic tank is given by formula (4). (4) In the formula The number of samples in the third-level test is given by formula (5). (5) By optimizing the algorithm, the parameters in formula (5) are obtained. Then, by using the ammonia nitrogen output data from the third-stage aerobic tank, the estimated value of the ammonia nitrogen output from the second-stage aerobic tank is inferred. After the third-level model parameter estimation is completed, the first-level and second-level models adopt the same parameters as the third-level model. and Then, the first-level model and the second-level model only estimate single parameters. and During the second-stage model parameter identification, the estimated value of the ammonia nitrogen output from the third-stage aerobic tank was inferred using the ammonia nitrogen measurement data. For formula (6): (6) When identifying the parameters of the second-level model, the second-level measurement data needs to be shifted forward as a whole. Time, as in formula (7): (7) The objective function for the second level is given by formula (8): (8) In the formula The number of samples during the third-level test is determined by the same identification method as the third level, based on the structure of formula (11), and compared with the current time. forward The sampling data at each sampling time point, that is, the identification Following the same method as the second-stage model parameter identification, the output ammonia nitrogen of the first-stage aerobic tank is first estimated by estimating the output ammonia nitrogen of the second-stage aerobic tank, and then the first-stage model parameters are obtained through single-parameter identification. ; Based on the second-level model, the estimated value of ammonia nitrogen at the outlet of the first-level aerobic tank is obtained using formula (9): (9) The sampling time of the output from the third level needs to be moved forward. At that moment, the first-level model is identified as formula (10): (10) The objective function for the first level is given by formula (11): (11)； S3. Establish an extended state observer for each AO stage with dissolved oxygen (DO) as input and effluent ammonia nitrogen concentration as output to obtain the estimated ammonia nitrogen concentration at the outlet of the first-stage AO and the outlet of the intermediate-stage AO. This enables the estimation of ammonia nitrogen concentration in the intermediate-stage AO when there is no ammonia nitrogen sensor. The ammonia nitrogen concentration value is used for ammonia nitrogen concentration control in this stage of AO. In real-time control, the ammonia nitrogen concentration in the intermediate-stage AO biochemical reactor, which lacks an ammonia nitrogen sensor, is estimated using an extended state observer; based on the measured values... and Obtained through the extended state observer Observations The observed values ​​of ammonia nitrogen in the effluent from the first-stage aerobic tank. , Ammonia nitrogen levels in the effluent from the second-stage aerobic tank at a given time. Based on this, the model recursive formula is used to... Moment ,get Moment ;from Moment ,arrive Moment Finally based on and Real-time process control of the intermediate-stage AO biochemical reactor is achieved; the expansion state observer of the tertiary AO biochemical reactor is shown in formulas (12a), (12b), and (12c): (12a) (12b) (12c) in, , and The values ​​of the gain for the first, second, and third stage observers are respectively selected to ensure that formulas (13a) and (13b) converge stably. The first-level soft measurement recursive formula (13a) is: (13a) in, ; The second-level soft measurement recursive formula (13b) is: (13b) in, ; S4. Based on the set value of ammonia nitrogen concentration in the effluent of each AO stage and the estimated value of ammonia nitrogen concentration at the inlet and outlet of each AO stage, a controller and its control algorithm are designed through a cascade model to update the set value of dissolved oxygen (DO) concentration in each AO stage in real time. Each AO stage then adjusts the aeration air volume according to the difference between the set value and the measured value of its own DO concentration to achieve precise aeration. The predictive control controller and its control algorithm are designed based on the cascade model. Objective functions and constraints are set for each stage of the aerobic tank. The predictive control law of the cascade model is obtained by optimization. Take control law The first element in the value represents the dissolved oxygen setpoint at that moment. The dissolved oxygen setpoint is updated once. At this time, the dissolved oxygen concentration of each aerobic tank is controlled by adjusting the air volume of the blower and the valve opening to track the given dissolved oxygen setpoint, so as to achieve energy-saving control of effluent ammonia nitrogen concentration in the constrained wastewater treatment process. The objective function for the first-stage aerobic tank is given by formula (14a): (14a) The objective function for the second-stage aerobic tank is given by formula (14b): (14b) The objective function for the third-stage aerobic tank is given by formula (14c): (14c) in, It predicts the time domain. In the control time domain, the first-level recursive prediction model is formula (13a), and the second-level recursive prediction model is formula (13b). The predicted value of ammonia nitrogen in the third stage effluent is given by the recursive prediction model, which is formula (15): (15) The range of dissolved oxygen variation and the value of ammonia nitrogen in the effluent are constrained by formula (16): (16) in, It is the lower limit of the DO setting value. This is the upper limit of the DO setting. It is the upper limit of the set value for ammonia nitrogen in the effluent.

2. A non-transitory computer-readable medium storing instructions, characterized in that, When the instruction is executed by the processor, the steps of the edge prediction control method for ammonia nitrogen in wastewater treatment effluent based on a cascade model as described in claim 1 are performed.

3. A computing device, comprising a processor and a memory for storing a processor-executable program, characterized in that, When the processor executes the program stored in the memory, it implements the edge prediction and control method for ammonia nitrogen in wastewater treatment effluent based on the cascade model as described in claim 1.

Images