Artificial intelligence control-based domestic sewage coupling treatment system and method

Abstract

The application provides a kind of based on artificial intelligence control's domestic sewage coupling processing system and method, it is related to sewage treatment technical field, including: compact annular processing unit module, MFC-artificial wetland collaborative module, LSTM-MPC hybrid control module and hybrid energy supply module;Sensor acquisition network is constructed, and corresponding data are collected in real time;According to the data, and based on LSTM model, construct domestic sewage prediction processing model, then output prediction key water quality parameter set, and update the parameters of domestic sewage prediction processing model;According to the prediction key water quality parameter set, construct optimization objective function, obtain the corresponding control instruction after optimization, realize the optimization control of corresponding equipment;Full-dimensional fault model is constructed, according to the full-dimensional fault model, obtain full-dimensional fault analysis result, so that staff make maintenance preparation;Energy scheduling strategy is formulated, according to the energy scheduling strategy, realize the intelligent power supply of system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sewage treatment, in particular to a domestic sewage coupling treatment system and method based on artificial intelligence control. BACKGROUND

[0002] With the development of rural economy and the improvement of living standards, the discharge of domestic sewage in rural areas is increasing year by year, and its treatment problem is increasingly prominent. Although the existing treatment technologies such as AO process, MBR membrane bioreactor and anaerobic biogas tank have good removal effect on COD, the removal rate of total nitrogen (TN) and total phosphorus (TP) is generally insufficient under low carbon source conditions, and it is difficult to meet the requirements of the first level A standard of "Discharge Standard of Pollutants for Municipal Wastewater Treatment Plant" (GB18918-2002). The traditional process relies on external aeration or chemical agents, resulting in significant increase in energy consumption. In the prior art, microbial fuel cells (MFC) have been tried for wastewater treatment, but the independent operation design fails to deeply couple with the main process, and lacks intelligent energy management strategy.

[0003] In view of the problems of high energy consumption, low treatment efficiency and complex maintenance in the prior art, the present application provides a domestic sewage coupling treatment system and method based on artificial intelligence control. The system is designed by combining drop water oxygenation, microbial fuel cells (MFC) and artificial wetlands, and realizes efficient nitrogen and phosphorus removal, energy self-sufficiency and dynamic optimization by combining LSTM-MPC multi-algorithm control, and is suitable for decentralized rural domestic sewage treatment. SUMMARY

[0004] In order to solve the above technical problems, the purpose of the present application is to provide a domestic sewage coupling treatment system and method based on artificial intelligence control.

[0005] In order to achieve the above purpose, the present application provides the following technical scheme: a domestic sewage coupling treatment system based on artificial intelligence control, comprising: a compact annular treatment unit module, an MFC-artificial wetland collaborative module, an LSTM-MPC hybrid control module and a hybrid energy supply module.

[0006] The compact annular treatment unit module is used to realize the connection between the corresponding devices and modules in the system.

[0007] The MFC-artificial wetland collaborative module realizes the effect of nitrogen and phosphorus removal through the collaborative treatment of MFC and artificial wetland.

[0008] The LSTM-MPC hybrid control module constructs a sensor network to collect corresponding data in real time; according to the data, a domestic sewage prediction processing model is constructed based on an LSTM model, and then a set of predicted key water quality parameters is output, and the parameters of the domestic sewage prediction processing model are updated; according to the set of predicted key water quality parameters, an optimization objective function is constructed, and corresponding control instructions after optimization are obtained to realize the optimization control of corresponding equipment; a full-dimensional fault model is constructed, and a full-dimensional fault analysis result is obtained according to the full-dimensional fault model, so that the staff can make maintenance preparations;

[0009] The hybrid energy supply module is used to formulate an energy dispatching strategy, and according to the energy dispatching strategy, the power supply of the system is realized.

[0010] Further, the compact annular treatment unit module is composed of a drop-type anaerobic trickling filter (3), an anoxic tank (4), an aerobic tank (5), a sedimentation tank (6), a constructed wetland (7), an MFC (8), and a water outlet and disinfection system (10) that are sequentially connected.

[0011] Further, the MFC-constructed wetland collaborative module is composed of a constructed wetland (7) and an MFC (8), and realizes the denitrification and phosphorus removal effect according to the collaborative treatment of the MFC and the constructed wetland.

[0012] The MFC (8) is composed of an MFC anode (801), an MFC cathode (802), and a root protection net (706), the MFC anode (801) is embedded in an iron-carbon composite filler layer (703), and the surface is loaded with sulfur-oxidizing bacteria and electrogenic bacteria co-cultured biofilm; the MFC cathode (802) is arranged below the plant root layer (701), connected to a super capacitor (804) through an adjustable resistance box (803), and provided with a root protection net (706).

[0013] Further, the process of the LSTM-MPC hybrid control module constructing a sensor collection network to collect corresponding data in real time includes:

[0014] The preset data collection device is composed of a plurality of sensors, and corresponding collection time points are set, and according to the collection time points, the data of the adjustment tank, the data of the aerobic tank, the data of the secondary sedimentation tank, the data of the air blower, the data of the sludge return pump, the data of the mixed liquid return pump, and the data of the MFC are collected.

[0015] Further, the process of the LSTM-MPC hybrid control module constructing a domestic sewage prediction processing model includes:

[0016] The corresponding data are collected through the sensor collection network and standardized, and then an input vector set is formed;

[0017] According to the input vector set, and based on an LSTM model, a domestic sewage prediction processing model is constructed, and a set of predicted key water quality parameters is obtained.

[0018] Further, according to the set of predicted key water quality parameters, the process of constructing an optimization objective function includes:

[0019] A set of predicted key water quality parameters is obtained; according to the set of predicted key water quality parameters, an optimization objective function is obtained The optimization objective function is:

[0020] ;

[0021] wherein, is the predicted total nitrogen concentration of the secondary sedimentation tank at time k; is the predicted total phosphorus concentration of the secondary sedimentation tank at time k; is the predicted dissolved oxygen amount of the aerobic tank at time k; is the target function; is the output control variable; is the prediction step; , , , is an adaptive weight coefficient; , is the target total nitrogen concentration and the target total phosphorus concentration of the secondary sedimentation tank; is the target dissolved oxygen amount of the aerobic tank; is the energy consumption model of the system.

[0022] Further, the process of constructing the energy consumption model includes:

[0023] The aeration energy consumption, sludge return energy consumption, mixed liquid return energy consumption, and MFC energy consumption are calculated;

[0024] According to the aeration energy consumption, sludge return energy consumption, mixed liquid return energy consumption, and MFC energy consumption, an energy consumption model is constructed.

[0025] Further, the process of constructing the full-dimensional fault model includes:

[0026] A digital twin model is constructed by collecting data collected by a sensor collection network, denoted as a full-dimensional fault model;

[0027] The predicted value output by the full-dimensional fault model and the measured value collected by the sensor collection network in real time are analyzed for deviation, and a full-dimensional fault analysis result is obtained.

[0028] Furthermore, the hybrid energy supply module is used to provide electrical energy to the system and formulate an energy dispatch strategy; the energy dispatch strategy is as follows: the sensor acquisition network is powered by MFC; the blower, sludge return pump, and mixed liquor return pump are driven by lithium battery packs; and the lithium capacitor energy storage device... At this time, the blower, sludge return pump and mixed liquor return pump are switched to mains power.

[0029] This invention further provides a method for coupled treatment of domestic sewage based on artificial intelligence control, comprising:

[0030] Build a sensor acquisition network to collect corresponding data in real time;

[0031] Based on the data and the LSTM model, a predictive treatment model for domestic sewage is constructed, which then outputs a set of key water quality parameters for prediction and updates the parameters of the predictive treatment model for domestic sewage.

[0032] Based on the predicted set of key water quality parameters, an optimization objective function is constructed to obtain the corresponding optimized control command, thereby achieving optimized control of the corresponding equipment.

[0033] Construct a full-dimensional fault model, and obtain full-dimensional fault analysis results based on the full-dimensional fault model to facilitate maintenance preparation by staff;

[0034] Develop an energy dispatch strategy, and implement intelligent power supply for the system based on the energy dispatch strategy.

[0035] Compared with the prior art, the beneficial effects of the present invention are:

[0036] 1. Significant treatment effect: Through the synergistic effect of the compact ring treatment unit module and the MFC-constructed wetland synergistic module, it has a high removal rate of pollutants.

[0037] 2. Reduced energy consumption: The LSTM-MPC hybrid control module and hybrid power supply module effectively reduce the system's energy consumption and improve energy self-sufficiency.

[0038] 3. Convenient operation and maintenance management: The fault diagnosis module can detect and handle system faults in a timely manner, improve the accuracy of fault diagnosis, realize unattended operation, and greatly reduce operation and maintenance costs. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0040] Figure 1 This is a complete process diagram of a domestic sewage coupled treatment system based on artificial intelligence control.

[0041] Figure 2 This is a cross-sectional view of an integrated ring-shell device for a domestic sewage coupled treatment system based on artificial intelligence control.

[0042] Figure 3 This is a plan view of an integrated ring-shell device for a domestic sewage coupled treatment system based on artificial intelligence control.

[0043] Figure 4 This is a schematic diagram of the MFC-constructed wetland synergistic technology for a domestic sewage coupled treatment system based on artificial intelligence control.

[0044] Figure 5 This is a schematic diagram of a predictive treatment model for domestic sewage in a coupled domestic sewage treatment system based on artificial intelligence control.

[0045] Figure 6 This is a schematic diagram of a full-dimensional fault model for a domestic sewage coupled treatment system based on artificial intelligence control.

[0046] Figure 7 This is a block diagram of a hybrid energy supply system for a domestic sewage coupled treatment system based on artificial intelligence control.

[0047] Figure 8 This is a schematic diagram of the steps in a coupled treatment method for domestic sewage based on artificial intelligence control.

[0048] Figure 9 This is a schematic diagram of a coupled domestic sewage treatment system based on artificial intelligence control.

[0049] Figure 1The components are as follows: 1. Equalization tank; 2. Annular integrated shell; 3. Cascading anaerobic trickling filter; 4. Anoxic tank; 5. Aerobic tank; 6. Secondary sedimentation tank; 7. Constructed wetland; 8. MFC; 9. Intelligent control cabinet; 10. Effluent and disinfection system; 11. Solar energy system; 12. Equipment room; 13. Blower; 14. Sludge return pump / mixed liquor return pump; 15. Centralized monitoring cabinet; 101. Bar screen; 102. pH / T sensor; 103. Lift pump; 104. Inlet flow meter; 301. Multi-stage stepped water distribution device; 302. Porous suspended ball packing; 303. Stainless steel partition. 304. Filter plate; 501. Circular aeration system; 502. Dissolved oxygen sensor; 601. Water collection weir; 701. Plant root layer; 702. Modified sulfur limestone transition layer; 703. Iron-carbon composite packing reaction layer; 704. Gravel support layer; 705. Bottom air backwash perforated pipe; 706. Root prevention net; 708a. Water distribution channel; 708b. Water collection channel; 707. Plants; 801. MFC anode; 802. MFC cathode; 803. Adjustable resistance box; 804. Supercapacitor; 901. Sensor network; 902. Signal transmission network; 903. Execution network; Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0051] like Figure 1 As shown, the domestic sewage coupled treatment system based on artificial intelligence control according to the present invention includes: a compact ring treatment unit module, an MFC-constructed wetland collaborative module, an LSTM-MPC hybrid control module, and a hybrid energy supply module;

[0052] The compact annular treatment unit module includes: a regulating tank (1), the inlet of which is provided with a grid (101) inclined at 45° and with a grid spacing of 5mm, and a pH / T sensor (102) and a booster pump (103) are configured inside the tank; the tank is connected to the annular integrated housing (2) by the booster pump (103), and an inlet flow meter (104) is provided on the pipeline between the regulating tank (1) and the annular integrated housing (2) for monitoring the inlet flow.

[0053] like Figure 2 , 3As shown, the annular integrated shell (2) includes: a biochemical integrated unit, an artificial wetland (7), an MFC (8), and a solar energy system (11); the biochemical integrated unit is composed of a cascading anaerobic trickling filter (3), an anoxic tank (4), an aerobic tank (5), and a secondary sedimentation tank (6); the artificial wetland (7) is coupled to the MFC (8);

[0054] It should be further noted that the annular integrated shell (2) can be made of any material such as stainless steel, carbon steel, reinforced concrete, or fiberglass, and can be built together with or separately from the regulating tank (1), and can be above ground or underground. The final effluent and disinfection system (10) of the annular integrated shell (2) is integrated together. The MFC (8) is a microbial fuel cell. The solar panels of the solar energy system (11) are placed on the integrated shell, occupying 50% of the area.

[0055] The cascading anaerobic trickling filter (3) is connected to the regulating tank (1) by a lift pump (103) through a pipeline. The water outlet of the pipeline is connected to the multi-stage stepped water distribution device (301) at the top. The multi-stage stepped water distribution device (301) is connected to the anoxic tank (4) at the bottom. The side wall of the anoxic tank (4) is provided with a perforated wall to flow into the aerobic tank (5) by gravity. It is connected to the water inlet system of the secondary sedimentation tank (6) through a pipeline. The secondary sedimentation tank (6) adopts a central cylinder water distribution form and is equipped with a water collection weir (601) around the perimeter. The supernatant is collected by the water collection weir (601) and flows into the artificial wetland (7) through a pipeline.

[0056] To further explain, in order to achieve sludge and mixed liquor recirculation, centrifugal pumps and screw pumps are installed in the equipment room to extract the mixed liquor from the aerobic tank (5) and the sludge from the secondary sedimentation tank (6), and recirculate them to the anoxic tank (4). A sludge concentration meter is installed at the outlet of the recirculation pump. At the same time, a sludge discharge pump is installed, and a flow meter is installed on the sludge discharge pipe to measure the sludge discharge volume. The automatic valve is opened periodically to discharge sludge and remove phosphorus.

[0057] It should be further explained that the multi-stage stepped water distribution device (301) consists of four layers (each with a drop height ≥ 0.1m), spanning half of the anaerobic tank surface. It is fixed to the two side walls of the tank using channel steel or angle steel, suspended above the liquid surface. Each step is equipped with an overflow weir. The tank is filled with porous suspended ball packing material (302), which can be any one of porous suspended ball packing material, fluidized bed packing material, or ceramsite. The packing height is... The packing material is fixed at both ends by stainless steel mesh (303), and a filter plate (304) is installed at the bottom of the tank. The aerobic tank (5) is equipped with a ring aeration system (501) and a dissolved oxygen sensor (502). A water collection weir (504) is installed at the end of the aerobic tank (5), and the weir load is ≤1.5L / (s·m).

[0058] The constructed wetland (7) is arranged in a ring around the biochemical unit. It is divided into multiple levels and operates in series. To ensure smooth water flow, the liquid level decreases by 10cm at each level. Each level is equipped with a distribution channel (708a) and a collection channel (708b) at the front and rear ends. The distribution system and the collection system are both configured with triangular pebbles. The outlet pipe of the collection weir (601) is connected to the distribution channel (708a) of the constructed wetland, and water is distributed through the bottom of the perforated pipe.

[0059] Each level of constructed wetland is equipped with the same filler material, which, from top to bottom, consists of:

[0060] 1. Plant root layer (701), wherein the thickness of the plant root layer (701) is 20cm, and it is filled with planting soil (particle size...). Organic matter content ≥10% and coarse sand (particle size) (Evenly spread to prevent water loss), the upper part is planted with yellow iris and cattail, with a plant spacing of 15cm / plant; to prevent plant roots from growing into the transition layer, a root-proof net (706) is set up.

[0061] 2. Sulfur-modified limestone transition layer (702), the thickness of which is 15cm, is composed of sulfur-modified volcanic rock (grain size...). Composition, sulfur mass percentage It is used to buffer water flow and slow the release of sulfur.

[0062] 3. Iron-carbon composite filler reaction layer (703), the thickness of which is 30cm, is composed of iron-carbon composite filler, the iron-carbon filler being cast iron shavings (particle size...). ) and coconut shell activated carbon (particle size) iodine value Mix at a ratio of 1:2 to 1:5;

[0063] 4. Gravel support layer (704), the gravel support layer (704) is 30cm thick, composed of gravel, and the particle size is... The bottom is equipped with a bottom air backwash perforated pipe (705) for backwashing the packing and electrodes;

[0064] The water collection channel (708b) is equipped with a water discharge and disinfection system (10), which consists of an ultraviolet disinfection device and is fixed in the final wetland water collection pit. The treated water meets the standards and is directly discharged.

[0065] like Figure 4As shown, the MFC-constructed wetland synergistic module includes: an MFC anode (801) and an MFC cathode (802); the MFC anode (801) is made of carbon-based titanium mesh and has an embedded iron-carbon composite filler reaction layer (703); the MFC cathode (802) is made of carbon felt material and is arranged in layers on the plant root layer (701), and a copper root-proof mesh (706) is added to prevent root penetration;

[0066] It should be further explained that the specific process by which the MFC-constructed wetland synergistic module achieves synergistic treatment of MFC (8) and constructed wetland (7) based on the iron-carbon composite filler reaction layer (703) includes:

[0067] The MFC cathode (802) is located at the root zone (701) of the plant. In the upper layer, the oxygen reduction reaction is carried out by the secretion of oxygen from the roots: ;

[0068] Root system at distance from plant root layer (701) In the lower layer, the nitrate bioelectrochemical reduction reaction takes place as follows: ;

[0069] The surface of the MFC anode (801) is loaded with a biofilm of sulfur-oxidizing bacteria. Based on the sulfur-oxidizing bacteria, a chemical oxidation reaction occurs as follows: ;

[0070] In the solution of the iron-carbon composite packing reaction layer (703), the denitrifying bacteria metabolic reaction proceeds simultaneously based on the denitrifying bacteria: ;

[0071] By supplementing anode electrons based on the iron element in the iron-carbon composite filler reaction layer (703), the process of strengthening anode electron transfer is as follows: ;

[0072] Meanwhile, based on the iron-carbon composite filler reaction layer (703) Chemical phosphorus removal is carried out, specifically as follows:

[0073] ;

[0074] Achieve this through a cycle of redox reactions. Electro-shuttle in MFC(8).

[0075] The MFC anode (801) is connected to the adjustable resistor box (803) via a titanium wire; the MFC cathode (802) is connected to the adjustable resistor box (803) via a copper wire and connected in parallel to the lithium capacitor energy storage device (804).

[0076] It should be further noted that the diameter of the titanium wire is... The adjustable resistance range of the adjustable resistance box (803) is: ;

[0077] like Figure 5 , 6 As shown, the LSTM-MPC hybrid control module includes: a sensor acquisition network, a domestic sewage predictive treatment model, an MPC optimization control unit, and a full-dimensional fault model;

[0078] It should be further explained that the LSTM-MPC hybrid control module achieves precise control of the rural sewage treatment system through a sensor acquisition network, a domestic sewage predictive treatment model, and an MPC optimization control unit. The sensor acquisition network collects relevant data, providing data support and updates. The domestic sewage predictive treatment model uses an LSTM model and, based on real-time and historical data, obtains predicted key water quality parameters (such as predicted dissolved oxygen, total nitrogen concentration in the secondary sedimentation tank, and total phosphorus concentration in the secondary sedimentation tank). The MPC optimization control unit optimizes corresponding control commands (such as aeration rate and reflux ratio) based on the predicted key water quality parameters to ensure high efficiency in the treatment process and compliance with effluent quality standards. The optimized control commands are directly issued to the actuators (such as blowers and reflux pumps) to achieve closed-loop control.

[0079] It should be further explained that the sensor acquisition network is set with corresponding acquisition times, and the chemical oxygen demand of the regulating tank (1) is acquired according to the acquisition times. Total nitrogen concentration Total phosphorus concentration pH ,temperature and water pump to increase flow rate Dissolved oxygen content in aerobic tank (5) sludge return ratio And the mixed liquor reflux ratio R_ml, the total nitrogen concentration in the secondary sedimentation tank (6) Total phosphorus concentration Power of blower (13) The power corresponding to the sludge return pump / mixed liquor return pump (14) as well as MFC current and resistance value ;

[0080] It should be further noted that the sensor acquisition network consists of several sensors set at corresponding data acquisition points, and is always powered by MFC or supercapacitor.

[0081] It should be further explained that, in addition to the inlet flow meter (104), a centralized detection cabinet (15) is also configured. The centralized detection cabinet (15) has a built-in multi-parameter analyzer. The sampling pump switches the sampling point through a three-way valve, and the sampling pipeline of the sampling pump... slope To prevent blockage.

[0082] It should be further explained that the specific process of constructing a predictive treatment model for domestic sewage includes:

[0083] Data collected by a sensor network is standardized to eliminate the influence of different parameter units. The standardized data is then grouped into a set of input vectors, denoted as... ;

[0084] The set of input vectors for:

[0085] [ , , , , , , , , , , , , , , ];

[0086] Obtain a set of input vectors from several historical data collection times. ;

[0087] For a set of input vectors at several historical acquisition time periods Group and label them as follows: It is a natural number;

[0088] Will The set of input vectors of the group As sample data, and Less than The natural numbers, and using the sample data, the mean of the sample data is obtained, denoted as the sample set;

[0089] The remaining sets of input vectors from several historical acquisition times As a test set;

[0090] A training sample set is formed based on the aforementioned sample set and test set;

[0091] A standard prediction processing model is constructed based on the LSTM model.

[0092] The training sample set is then input into the standard prediction and treatment model to train the standard prediction and treatment model, and the trained standard prediction and treatment model is denoted as the domestic sewage prediction and treatment model.

[0093] Based on the aforementioned domestic sewage predictive treatment model, a set of key water quality parameters for prediction can be obtained. ;

[0094] The set of key water quality parameters for prediction for: ;in, ; This is an estimated value of the sludge settling ratio in the aerobic tank, representing the settling performance of the sludge in the aerobic tank at the current moment; it is obtained through manual testing.

[0095] It should be further explained that the specific process by which the MPC optimization control unit optimizes the corresponding control commands includes:

[0096] Obtain the set of key water quality parameters for prediction ;

[0097] Based on the predicted set of key water quality parameters Obtain the optimization objective function ;

[0098] The optimization objective function for:

[0099] ;

[0100] in, The objective function is... For output control variables; To predict the step size; , , , These are adaptive weighting coefficients; , The target total nitrogen concentration and target total phosphorus concentration for the secondary sedimentation tank; The target dissolved oxygen level for the aerobic tank; This is the energy consumption model for the system.

[0101] Based on the optimization objective function Output control variables Adjust the frequencies of the blower, sludge return pump, and mixed liquor return pump; optimize the objective function. Each term in the formula represents a penalty for the total nitrogen concentration in the secondary sedimentation tank effluent, a penalty for the total phosphorus concentration in the secondary sedimentation tank effluent, a penalty for the dissolved oxygen content in the aerobic tank, and a penalty for the system energy consumption.

[0102] It should be further explained that the specific process of updating the parameters of the aforementioned domestic sewage predictive treatment model includes:

[0103] When the data collected by the sensor network is combined with the set of key water quality parameters predicted three times consecutively by the domestic sewage predictive treatment model... If the error exceeds the set threshold, the parameters of the domestic sewage predictive treatment model will be automatically updated, and a calibration report will be sent.

[0104] For example, , as well as ;

[0105] It should be further explained that the adaptive weighting coefficients , , , The specific process of making adjustments includes:

[0106] Based on the different influent water quality and operating conditions, the weights of each item in the objective function are automatically adjusted to achieve more precise control;

[0107] This is the weighting factor for nitrogen removal in the secondary sedimentation tank; when the total nitrogen concentration in the influent is high, increase... The value of indicates an increase in the importance of denitrification.

[0108] This is the weighting factor for phosphorus removal in the secondary sedimentation tank; when the total phosphorus concentration in the influent is high, increase... The value of indicates an increase in the importance of phosphorus removal.

[0109] This is the weighting coefficient for dissolved oxygen removal in the aerobic tank; when the oxygen content of the influent is high, it is reduced. A value that indicates a decrease in the importance of dissolved oxygen.

[0110] This is a weighting factor for system energy consumption; when the influent flow rate is low, it is increased. A value of indicates an increase in the importance of energy conservation.

[0111] It should be further explained that the energy consumption model mentioned above... for:

[0112] ;in, Indicates aeration energy consumption; Indicates the energy consumption of sludge recirculation; Indicates the energy consumption of the mixed liquid reflux; This indicates the energy consumption of MFC.

[0113] It should be further explained that the aeration energy consumption... ;in, Indicates the first Real-time power of the blower; Indicates the time step;

[0114] The energy consumption of sludge recirculation ;in, Indicates the first Real-time power of the sludge return pump;

[0115] The energy consumption of the mixture reflux ;in, Indicates the first Real-time power of the mixture reflux pump;

[0116] The energy consumption of MFC ;in, This represents the real-time current of the MFC at time k. Indicates the first The real-time resistance value of MFC at any given time.

[0117] It should be further explained that the specific process of constructing a full-dimensional fault model includes:

[0118] A digital twin model is constructed by collecting data from a sensor network, which is denoted as a full-dimensional fault model.

[0119] A deviation analysis is performed between the predicted values ​​output by the comprehensive fault model and the measured values ​​collected in real time by the sensor acquisition network to obtain the comprehensive fault analysis results. The specific process of the deviation analysis and the comprehensive fault analysis results are shown in the table below:

[0120] Table 1:

[0121]

[0122] Based on the comprehensive fault model, the corresponding fault type can be obtained quickly and easily, so that staff can make maintenance preparations.

[0123] like Figure 7 As shown, the hybrid power supply module is used to provide power to the system and formulate an energy dispatch strategy. The energy dispatch strategy is as follows: the sensor acquisition network is powered by the MFC; when the MFC's power is insufficient, it switches to mains power or a lithium-ion battery storage device; the MFC does not participate in the power supply of high-power equipment; the blower, sludge return pump, and mixed liquor return pump are driven by lithium-ion battery packs; when the lithium-ion battery storage device... At this time, the blower, sludge return pump and mixed liquor return pump are switched to mains power.

[0124] It should be further explained that the lithium battery pack is connected to a solar energy system, through which electrical energy is stored.

[0125] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A domestic sewage coupled treatment system based on artificial intelligence control, characterized in that, The system includes: a compact ring processing unit module, an LSTM-MPC hybrid control module, and a hybrid power supply module; The compact ring-shaped treatment unit module consists of a cascading anaerobic trickling filter (3), an anoxic tank (4), an aerobic tank (5), a secondary sedimentation tank (6), an artificial wetland (7), an MFC (8), and an effluent and disinfection system (10) connected in sequence; wherein, the MFC (8) and the artificial wetland (7) form an MFC-artificial wetland synergistic module to achieve nitrogen and phosphorus removal effects; An LSTM-MPC hybrid control module constructs a sensor acquisition network to collect corresponding data in real time. Based on this data and an LSTM model, it builds a predictive treatment model for domestic sewage, outputs a set of predicted key water quality parameters, and updates the parameters of the domestic sewage predictive treatment model. Based on the set of predicted key water quality parameters, it constructs an optimization objective function, including: Obtain a set of key water quality parameters for prediction; based on the set of key water quality parameters for prediction, obtain the optimization objective function. The optimization objective function for: ； in, The predicted total nitrogen concentration in the secondary sedimentation tank at time k; The predicted total phosphorus concentration in the secondary sedimentation tank at time k; The predicted dissolved oxygen level in the aerobic tank at time k. The objective function is... The output control variable is used to adjust the frequency of the blower, sludge return pump, and mixed liquor return pump; To predict the step size; , , , These are adaptive weighting coefficients; , The target total nitrogen concentration and target total phosphorus concentration for the secondary sedimentation tank; The target dissolved oxygen level for the aerobic tank; For the system's energy consumption model; The process of building an energy consumption model includes: Calculate the energy consumption of aeration, sludge return, mixed liquor return, and MFC. An energy consumption model is constructed based on the aeration energy consumption, sludge return energy consumption, mixed liquor return energy consumption, and MFC energy consumption. Obtain the optimized control commands to achieve optimized control of the corresponding equipment; Construct a comprehensive fault model, including: A digital twin model is constructed by collecting data from a sensor network, which is denoted as a full-dimensional fault model. A deviation analysis is performed between the predicted values ​​output by the full-dimensional fault model and the measured values ​​collected in real time by the sensor acquisition network to obtain the full-dimensional fault analysis results, so as to help staff make maintenance preparations. The hybrid power supply module is used to formulate energy dispatch strategies and realize the power supply of the system according to the energy dispatch strategies.

2. The artificial intelligence-controlled coupled domestic sewage treatment system according to claim 1, characterized in that, The process of constructing a sensor acquisition network using the LSTM-MPC hybrid control module and acquiring the corresponding data in real time includes: A data acquisition device is set up, which consists of several sensors and is configured with corresponding acquisition times. Based on the acquisition times, data from the equalization tank, aerobic tank, secondary sedimentation tank, blower, sludge return pump, mixed liquor return pump, and MFC are collected.

3. The artificial intelligence-controlled coupled domestic sewage treatment system according to claim 2, characterized in that, The process of building a predictive treatment model for domestic wastewater using the LSTM-MPC hybrid control module includes: The corresponding data is collected through a sensor acquisition network and standardized to form an input vector set. Based on the input vector set and the LSTM model, a predictive treatment model for domestic sewage is constructed to obtain a set of key water quality parameters for prediction.

4. The artificial intelligence-controlled coupled domestic sewage treatment system according to claim 1, characterized in that, The hybrid power supply module is used to formulate an energy dispatch strategy; the energy dispatch strategy is as follows: the sensor acquisition network is powered by MFC; the blower, sludge return pump, and mixed liquor return pump are driven by lithium battery packs; and the lithium capacitor energy storage device... At this time, the blower, sludge return pump and mixed liquor return pump are switched to mains power.

5. A method for coupled treatment of rural domestic sewage based on artificial intelligence control, implemented using the coupled treatment system for domestic sewage based on artificial intelligence control as described in any one of claims 1 to 4, characterized in that, include: Build a sensor acquisition network to collect corresponding data in real time; Based on the data and the LSTM model, a predictive treatment model for domestic sewage is constructed, which then outputs a set of key water quality parameters for prediction and updates the parameters of the predictive treatment model for domestic sewage. Based on the predicted set of key water quality parameters, an optimization objective function is constructed to obtain the corresponding optimized control command, thereby achieving optimized control of the corresponding equipment. Construct a full-dimensional fault model, and obtain full-dimensional fault analysis results based on the full-dimensional fault model to facilitate maintenance preparation by staff; Develop an energy dispatch strategy, and implement intelligent power supply for the system based on the energy dispatch strategy.

Images