Wetland water quality purification simulation device and simulation method

Abstract

The invention discloses a wetland water quality purification simulation device and a simulation method, and mainly solves the technical problems that a traditional wetland simulation device is high in substrate replacement difficulty and cannot comprehensively analyze substance circulation in a simulated wetland system. The wetland water quality environment corresponding to the sewage tank is separated from the soil quality environment corresponding to the substrate tank, and floating plants are cultivated in the sewage tank, so that the purpose of controlling separation of water, soil and plants is achieved, and the substrate can be conveniently replaced; the problem that the whole system needs to be reconstructed again due to the fact that a traditional simulation device is difficult to replace the matrix independently after the matrix is aged is solved. And through the height difference design between the sewage tank and each matrix tank, sewage in the sewage tank can naturally flow into each matrix tank by virtue of the gravity of the sewage tank, power devices such as a water pump do not need to be additionally arranged, and the energy consumption is favorably reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of environmental engineering, and particularly relates to a wetland water purification simulation device and a simulation method. BACKGROUND

[0002] An artificial wetland is a wastewater treatment technology that simulates a natural wetland ecological system, and achieves purification and treatment of wastewater through artificial construction and supervision and control, by using plants, microorganisms, and substrates (fillers) in the wetland to realize physical, chemical, and biological synergistic effects.

[0003] An artificial wetland simulation test device is a key device for studying the removal effect of an artificial wetland system on pollutants, the operation mechanism, and the optimization design. By simulating different wetland types, operation conditions, and pollutant loads, scientific basis can be provided for actual engineering. However, the present inventors found in the process of implementing the technical solutions in the embodiments of the present application that the traditional wetland simulation device usually mixes water, soil, and plants together, which leads to difficulty in replacing the substrate alone after the service life of the substrate is reduced, and even the entire system (including the water and the plants) needs to be emptied and reconstructed, which increases the experimental cost and difficulty. The traditional wetland simulation device cannot control a single variable such as water quality, substrate, and plants, and cannot comprehensively analyze the material circulation in the wetland system.

[0004] The information disclosed in this BACKGROUND section is only for the purpose of enriching the understanding of the background of the present disclosure and should not be regarded as acknowledging or implicitly implying that it constitutes prior art known to those of ordinary skill in the art. SUMMARY

[0005] In view of at least one of the above technical problems, the present disclosure provides a wetland water purification simulation device and a simulation method, which mainly solve the technical problems of difficulty in replacing the substrate of the traditional wetland simulation device and the inability to comprehensively analyze the material circulation in the simulated wetland system.

[0006] According to one aspect of the present disclosure, a wetland water purification simulation device is provided, which comprises a sewage tank with a water outlet on one side and used for soilless culture of floating plants, at least two substrate tanks for loading corresponding substrates, each of which is independent of the other and has a bottom height lower than that of the sewage tank, a sensor array arranged in the sewage tank and each of the substrate tanks for real-time monitoring of water quality parameters, gas composition, light intensity and temperature, a controller in corresponding communication connection with the sensor array, and an adjustable light source arranged above the device and in corresponding electrical connection with the controller; the water inlets of each of the substrate tanks are in communication with the water outlet of the sewage tank through corresponding parallel water conveying pipelines provided with regulating valves; the sewage tank is arranged at a height lower than the designed liquid level height in the substrate tank; each of the substrate tanks comprises a sealing cover in corresponding sealing connection with the tank body and provided with a gas collecting port, and a gas path valve is arranged at the gas collecting port.

[0007] In some embodiments of the present disclosure, at least one overflow hole is arranged at a corresponding height position of the vertical direction of the side wall of the sewage tank and each of the substrate tanks.

[0008] In some embodiments of the present disclosure, the regulating valve is an electromagnetic valve in corresponding communication connection with the controller.

[0009] In some embodiments of the present disclosure, a liquid level meter in corresponding communication connection with the controller is arranged in the sewage tank and / or each of the substrate tanks.

[0010] In some embodiments of the present disclosure, the top cover of the sewage tank and the sealing cover are respectively transparent material pieces.

[0011] In some embodiments of the present disclosure, the wetland water purification simulation device further comprises a support for fixing the sewage tank and each of the substrate tanks, and the support is provided with a height adjusting assembly for adjusting the vertical height of the sewage tank; the water conveying pipeline comprises a hose with a length greater than the adjusting range of the height adjusting assembly.

[0012] According to another aspect of the present disclosure, a wetland water purification simulation method is provided, which is based on the above-mentioned wetland water purification simulation device and comprises the following steps: (1) injecting the sewage to be treated into the sewage tank and planting floating plants on the surface of the sewage, and filling the substrate tanks with corresponding substrate materials, and adjusting the regulating valves between the substrate tanks and the sewage tank to convey the sewage to be treated from the sewage tank to the substrate tanks; (2) acquiring water quality, gas and corresponding environmental parameter data in real time through the sensor array and inputting them into an LSTM model to correspondingly predict the pollutant concentration change trend; (3) based on the deep deterministic policy gradient algorithm, adjusting the opening degree of the regulating valve and the light intensity according to the prediction result of the LSTM model to maximize the pollutant removal rate and minimize the energy consumption. (4) Corresponding water quality, gas and environmental parameter change data are acquired.

[0013] In some embodiments of the present disclosure, in the step (2), the LSTM model makes a prediction based on time series, and the input is historical and real-time parameter data including COD, TN, TP, CH4, CO2, N2, light intensity and temperature, and the output is a predicted value of pollutant concentration in a future preset time period.

[0014] In some embodiments of the present disclosure, in the step (3), the reward function of the deep deterministic policy gradient algorithm is a weighted sum of pollutant removal rate and energy consumption rate, and the action space includes an adjustment amount of the valve opening and an adjustment amount of the light intensity.

[0015] In some embodiments of the present disclosure, in the step (3), the reward function R of the deep deterministic policy gradient algorithm satisfies: ; Wherein, C initial is an initial pollutant concentration, C current is a current concentration, E max is a maximum allowed energy consumption, E current is a current energy consumption.

[0016] One or more technical solutions provided in the embodiments of the present application have at least any of the following technical effects or advantages: 1. The water quality environment of the wetland corresponding to the sewage tank and the soil environment of the substrate tank corresponding to the substrate tank are separated, and the floating plants are cultivated in the sewage tank, so as to achieve the separation and control purpose of water, soil and plants, so that the substrate can be conveniently replaced, and the problem that the traditional simulation device needs to be reconstructed as a whole after the substrate is aged and it is difficult to replace the substrate alone.

[0017] 2. The substrate tanks are arranged in parallel and independent of each other, so that the influence of a single variable (such as substrate type) on the purification effect can be specially studied under the condition that other variables are consistent, so that the specific influencing factors of artificial wetland water purification can be accurately explored.

[0018] 3. The wastewater tank and each substrate tank are designed with a height difference, so that the wastewater in the wastewater tank can flow into each substrate tank naturally by its own gravity, without the need for additional power devices such as water pumps, which helps to reduce energy consumption. At the same time, by setting the water inlet height of the substrate tank to be lower than the liquid level, the bottom of the substrate tank is liquid-tight, thereby preventing the escape of gas generated during the wastewater treatment process. The gas can be collected directionally through the gas collection port at the sealing cover of the substrate tank, solving the problem that the existing simulation device cannot detect the conversion process of pollutants in the gas phase.

[0019] 4. The fusion of LSTM time series prediction and deep reinforcement learning algorithm can greatly improve the parameter response speed and system stability compared to traditional PID control, and speed up the simulation process. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 Figure 1 is a structural schematic diagram of a wetland water purification simulation device according to an embodiment of the present application.

[0021] Figure 2 Figure 1 is a structural schematic diagram of a wetland water purification simulation device according to an embodiment of the present application.

[0022] In the above figures, 1 is a wastewater tank, 11 is a top cover, 2 is a substrate tank, 21 is a sealing plate, and 22 is a gas collection port. DETAILED DESCRIPTION

[0023] In the description of the present application, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inner", "outer", "vertical", "horizontal", "clockwise", "counterclockwise" and the like indicate the orientation or positional relationship shown in the drawings, and are only used for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0024] The programs involved or relied on in the following embodiments are all conventional programs or simple programs in the technical field, and those skilled in the art can make conventional choices or adaptive adjustments according to specific application scenarios. The devices and the like involved in the following embodiments are all conventional commercially available products unless otherwise specified.

[0025] In order to better understand the technical solutions of the present application, the above technical solutions will be described in detail below in conjunction with the drawings and specific embodiments of the specification.

[0026] To solve the technical problems of difficulty in substrate replacement of traditional wetland simulation device and inability to comprehensively analyze the material circulation in the simulated wetland system, the wetland water purification simulation device comprises a sewage tank and a plurality of substrate tanks corresponding to and in parallel communication with the sewage tank and independent of each other.

[0027] Specifically, referring to Figure 1 In the embodiment, the sewage tank 1 is a cuboid structure box body for containing sewage to be treated. In the embodiment, a floating bed is arranged in the sewage tank 1 for soilless culture of floating plants, so that the influence of plants on the sewage treatment process can be explored. In order to facilitate the water inlet of the sewage tank 1, referring to Figure 1 , the top of the sewage tank 1 is open and a top cover 11 is arranged corresponding to the cover. After the top cover 11 is opened, sewage can be directly injected into the sewage tank through the top opening of the sewage tank 1. In other embodiments, considering that some simulation experiments require continuous sewage supplement according to the sewage treatment process during the experiment, it is more cumbersome to manually inject through the top opening and the timely supplement of sewage cannot be guaranteed. Therefore, in this embodiment, a water injection port is arranged at the side wall of the sewage tank 1, which is connected to a sewage pool through a water injection pipeline, and a water injection valve is connected in series in the water injection pipeline. Thus, by adjusting the water injection valve, the opening and closing of the water injection pipeline and the flow can be adjusted as needed, so that the sewage in the sewage tank 1 can be effectively and reliably injected and supplemented from the sewage pool. In addition, in some embodiments provided with a water injection port, in order to avoid excessive water injection flow from impacting the floating plants in the sewage tank, an L-shaped pipe is arranged in the sewage tank and connected to the water injection port, and the outlet position of the L-shaped pipe is lower than the design height of the sewage level in the sewage tank, so as to avoid the impact of injected sewage and affect the experimental effect. In another embodiment provided with a water injection port, the L-shaped pipe connected to the water injection port in the sewage tank is omitted, and the water injection port is directly arranged at a position lower than the design height of the sewage level in the sewage tank, so as to also avoid water impact.

[0028] In addition, in order to explore the influence of different substrates on water treatment, at least two independent substrate tanks are arranged. Referring to Figure 1In the embodiment, three substrate boxes 2 are provided, and the substrate boxes 2 are filled with substrates of different materials to explore the influence of different substrates on the treatment of sewage. In order to make the water in each substrate box the same, and avoid interference with the experimental results due to the difference in water quality, a water outlet is formed on one side of the sewage tank, and the water inlets of each substrate box are connected to the water outlet of the sewage tank 1 through a water conveying pipeline, so as to ensure that the water quality in each substrate box is the same. In the example, the water conveying pipeline includes a main pipe connected to the water outlet of the sewage tank, and each branch pipe having one end in communication with the main pipe and the other end in communication with the water inlet of the corresponding substrate box, that is, each branch pipe is connected in parallel, so as to realize the communication between the sewage tank and each substrate box, and thus achieve the purpose of conveying sewage. In the example, in order to control the water conveying pipeline, an adjusting valve is connected in series in the water conveying pipeline. Specifically, the adjusting valve includes a water outlet main valve connected in series in the main pipe, and a water outlet branch valve connected in series in each branch pipe. Thus, the water outlet main valve controls the opening and closing and flow rate of the flow path as a whole, and each water outlet branch valve controls the opening and closing and flow rate of the water inlet of each substrate box as needed. In order to improve the control efficiency and convenience, the water outlet main valve and each water outlet branch valve are solenoid valves in the example, and the device further includes a controller. The water outlet main valve and the water outlet branch valve are connected to the controller in communication, so as to realize reliable and convenient control of the opening and closing and flow rate of the water conveying pipeline through the controller. In addition, in the embodiment in which the sewage tank is provided with a water injection pipeline, the water injection valve connected in series in the water injection pipeline is also a solenoid valve, and is in communication with the controller, and the opening and closing and flow rate of the water injection pipeline are controlled by the controller.

[0029] In order to reduce unnecessary energy loss during conveying of the water in the sewage tank to the substrate boxes, in the embodiment, the bottom plate of the sewage tank 1 is higher than the bottom plates of the substrate boxes 2, and the bottom plates of the substrate boxes 2 are coplanar. Thus, the water in the sewage tank 1 flows to the substrate boxes 2 under the action of gravity due to the height difference between the sewage tank 1 and the substrate boxes 2, thereby avoiding the need for an additional power device. Meanwhile, in the embodiment, the water inlets of the substrate boxes are arranged at a height lower than the design liquid level of the substrate boxes, so that the water in the sewage tank can drive the water flow, and the liquid level higher than the water inlets of the substrate boxes can seal the water inlets of the substrate boxes, thereby avoiding the escape of gas generated in the substrate boxes during sewage treatment through the water inlets.

[0030] In addition, in order to effectively obtain the changes occurring in the device during the sewage treatment process, in the present embodiment, the device further comprises a sensor array, which respectively comprises water quality monitoring sensors for detecting water quality parameters (such as COD, TN, and TP) of the water in the sewage tank and each substrate tank, gas sensors provided in each substrate tank for measuring the concentration of the corresponding gas (such as CH4 and CO2), temperature sensors, and light intensity sensors, each of which is in communication connection with the controller. In addition, in order to provide the required simulated light conditions, an adjustable light source is provided on the top of the device in the present embodiment, which is electrically connected to the controller and controlled by the controller to adjust the light intensity as needed. Thus, the simulated process parameters are obtained by the sensors, and the corresponding variable parameters are adjusted by the controller, so as to explore the influencing factors of wetland water purification under different conditions and the influence degree of the corresponding factors. In other embodiments, a liquid level meter in communication connection with the controller is further provided in the sewage tank and each substrate tank, so as to obtain the liquid level of each tank in real time, and thus the water injection pipeline and the water delivery pipeline are controlled as needed.

[0031] In the present embodiment, by arranging the water inlet of each substrate tank at a position lower than the liquid level, the liquid seal of the water inlet, i.e., the bottom space of the substrate tank, is achieved. However, in order to facilitate the loading of the substrate, an opening is provided at the top of each substrate tank. In order to prevent the gas generated in the substrate tank from escaping from the top opening of the substrate tank during the sewage treatment process, in the present embodiment, a sealing plate 21 is provided at the top of each substrate tank, the bottom surface of the sealing plate 21 is provided with an annular sealing strip, and the top of the substrate tank 2 is provided with an annular shoulder for limiting the sealing plate 21. In the present embodiment, the sealing plate and the shoulder are screw-connected, so that the sealing plate is tightly contacted with the shoulder, thereby achieving the purpose of sealing the top of the substrate tank and preventing the gas from escaping. In addition, in order to ensure that the substrate tank and the sewage tank are effectively illuminated, the sealing plate 21 provided at the top of each substrate tank and the top plate of the sewage tank are made of transparent material.

[0032] Referring to Figure 1 , considering that the substrate tank is in a relatively sealed state during operation, in order to prevent the sealing state from causing the water in the sewage tank to be unable to drive the sewage to flow in under the action of gravity, and to prevent the gas generated during the sewage treatment from accumulating in the substrate tank, in the present embodiment, a gas collecting port 22 is provided at the sealing plate 21 at the top of the substrate tank, which is in communication with the internal space of the substrate tank 2. Specifically, in the present embodiment, a gas path valve is connected at the gas collecting port 22, which is an electrically controlled valve and is in communication connection with the controller, and the conduction degree of the valve body is controlled by the controller. In other embodiments, the other end of the gas path valve is connected with a gas delivery pipeline, which is connected to a gas chromatograph or a gas collection bag, so as to further analyze and study the generated gas in detail.

[0033] To avoid the water level in the sewage tank 1 and each substrate tank 2 being too high, at least one overflow hole is formed in the vertical direction on the side wall of the sewage tank and each substrate tank, and each overflow hole is connected with an overflow valve, which can be opened and closed as needed, so that when the water level exceeds the corresponding height position of the overflow hole, the water exceeding the required liquid level height can flow out from the corresponding overflow hole when the overflow valve is opened, and the overflow hole can also be used to collect water in each tank and adjust the air pressure in the substrate tank.

[0034] In other embodiments, considering that the inflow of sewage into each substrate tank is limited by the liquid level of the water in the sewage tank, in order to reliably drive the water to flow into each substrate tank under its own weight when the liquid level in the sewage tank is limited by the experiment, in this case, the wetland water purification simulation device further includes a support, and the sewage tank and each substrate tank are fixed relative to the support, wherein the support is provided with a height adjustment assembly corresponding to the position of the sewage tank to realize vertical adjustment of the height position of the sewage tank. For example, the height adjustment assembly in this case includes a plurality of vertical guide rails, a guide block fixed to the outside of the sewage tank and slidingly embedded in the guide rail, and a lifting motor connected to the bottom of the sewage tank. The sewage tank is driven to move vertically along the guide rail by the lifting motor, thereby achieving the purpose of height position adjustment, so that the water level in the sewage tank can still reliably deliver water to the substrate tank under its own weight when the water level is limited. In order to avoid the hard link of the water delivery pipeline limiting the adjustment of the position of the sewage tank, the water delivery pipeline in this case includes a hose with a length greater than the adjustment range of the height adjustment assembly.

[0035] In other embodiments, considering that the water quality purification treatment capacity of the built wetland water quality purification simulation system is limited and cannot meet the required sewage purification demand, in order to achieve effective use of sewage in the experiment, in this case, each substrate tank is provided with a water outlet, and the water outlet is connected to the sewage tank or sewage tank through a pipeline provided with a pipe valve, thereby achieving reasonable reuse of the sewage discharged from the substrate tank. In other embodiments, in order to facilitate the convenient replacement of the substrate in the substrate tank, a gate valve with a certain diameter is sealingly connected to the bottom plate of the substrate tank, and the gate valve is used for efficient cleaning and replacement of the substrate.

[0036] In addition, the example also discloses a wetland water purification simulation method based on the wetland water purification simulation device, which is used for conducting a control experiment. The method comprises the following steps: injecting sewage to be treated into the sewage tank, and planting floating plants on the surface of the sewage; filling different substrate materials into the substrate tanks, respectively; using the height difference between the sewage tank and the substrate tanks, and adjusting the flow rates of the adjusting valves between the substrate tanks and the water inlet pipelines of the substrate tanks to be consistent, so that the water in the sewage tank flows into each substrate tank at the same flow rate, and the water level in each substrate tank is controlled by adjusting the opening and closing states of the overflow holes or by monitoring the data of the liquid level meters. After the water level is constant, the overflow holes are closed, and the substrate tanks are sealed by the liquid seals at the bottoms and the sealing covers at the tops to prevent gas leakage. Water samples and gas samples are collected through the overflow holes and the gas collection ports, respectively, and specific water quality indexes and gas components are detected, respectively. Thus, the control experiment under different substrate conditions is conducted.

[0037] Specifically, in the example, simulated sewage with a TN content of 15 mg / L and a TP content of 2 mg / L is injected into the sewage tank, and Lemna with a coverage rate of 60% is planted in the sewage tank. Meanwhile, zeolite, activated carbon and sand are filled into the three substrate tanks, respectively, and the filling height is 40 cm. Thus, the system is built. In addition, in some other examples, at least one of the substrate tanks is set as a blank control and only filled with inert substrate. After the experiment starts, the adjusting valves in the water inlet pipelines are adjusted so that the water inlet flow rates of the substrate tanks are all 2.0 L / h, and the liquid level difference between the sewage tank and each substrate tank is controlled to be 20 cm. Water samples are collected through the overflow holes every day, and the removal rates of total nitrogen TN and total phosphorus TP are detected. 500 mL of gas is extracted through the gas collection ports every week, and the gas concentrations of methane CH4, carbon dioxide CO2 and nitrogen N2 are detected. The experimental results show that the TN removal rate of the substrate tank filled with activated carbon is increased by 22% compared with the substrate tank filled with sand, and the CH4 emission amount of the substrate tank filled with zeolite is reduced by 35%.

[0038] In other examples of the simulation experiment of the tidal flow of the artificial wetland, the adjusting valves in the water inlet pipelines are controlled to be periodically opened and closed and to have a conduction degree, so that the water levels in the substrate tanks experience the stages of water inlet, static and water outlet in a preset time. Specifically, in the example, the water inlet is set to be 2 h, the static is set to be 4 h and the water outlet is set to be 2 h, which are sequentially and repeatedly circulated to simulate the tidal flow. Thus, the differences in COD degradation kinetics under the continuous flow mode and the tidal flow mode can be obtained by comparison.

[0039] In addition, the example also discloses another wetland water purification simulation method based on the wetland water purification simulation device, which is used to reduce energy consumption and speed up the simulation process to explore the purification effect of the set wetland water purification condition. The method comprises the following steps: (1) Inject the sewage to be treated into the sewage tank, and plant the set floating aquatic plants on the surface of the sewage; and fill the corresponding substrate material into the substrate tank, adjust the regulating valve between the substrate tank and the sewage tank, and use the height difference between the sewage tank and the substrate tank to transport the sewage to be treated into the substrate tank.

[0040] Specifically, in the present embodiment, the wetland water purification simulation device shown in Figure 2 The wetland water purification simulation device shown in FIG. 1 is used in the present embodiment. In the present embodiment, the three substrate tanks are filled with barley straw biochar, biological ceramsite and river sand respectively, and the corresponding regulating valve opening degree and light intensity of each substrate tank are consistent under the initial condition.

[0041] (2) Real-time acquisition of water quality, gas and corresponding environmental parameter data by sensor array, and input into LSTM model to predict pollutant concentration change trend.

[0042] In the present embodiment, the sensor array detects the water quality indexes including total nitrogen TN, total phosphorus TP and chemical oxygen demand COD, the gas components including methane CH4, carbon dioxide CO2 and nitrogen N2, and the light intensity and temperature in each substrate tank. In addition, in the present embodiment, the embedded controller NVIDIA Jetson Nano integrated with GPU is used for data processing and instruction issuing. Specifically, in the present embodiment, in order to realize adaptive adjustment of the simulation device system, reduce the adverse effects of human operation error and delay on the simulation experiment, and efficiently and reliably set the sewage treatment and purification effect under the condition, an LSTM model is established in the present embodiment, and a water quality prediction model based on long short-term memory network LSTM is established to predict the change trend of the future pollutant concentration in each substrate tank, so that the environmental variables can be adaptively adjusted according to the change trend of the pollutants in the subsequent process to speed up the experimental process.

[0043] Specifically, in the present embodiment, the controller collects sensor data at a frequency of 1 Hz, performs outlier filtering and normalization processing, and starts adaptive control when the sensor array detects that the total nitrogen TN concentration fluctuation exceeds ±10% or the light intensity deviation is greater than 200 lux. Each sensor array inputs the real-time data obtained by it into the established LSTM water quality prediction model to predict the future water quality trend. If the prediction shows that the removal rate of total nitrogen TN is lower than the set threshold, the optimal adjustment amount is calculated through the deep deterministic policy gradient algorithm described in step 3, so as to improve the processing efficiency.

[0044] In this example, an LSTM model based on time series is established based on at least 100 sets of historical test data covering different substrates, plants and flow rate conditions. The input sequence is the historical parameter data of COD, TN, TP, CH4, CO2, light intensity and temperature in the past 30 minutes, and the output is the pollutant concentration (COD, TN, TP) prediction value in the next 60 minutes.

[0045] (3) Based on the deep deterministic policy gradient algorithm, the prediction results of the LSTM model are used to adjust the opening degree of the regulating valve and the light intensity, so as to maximize the pollutant removal rate and minimize the energy consumption.

[0046] In this embodiment, the state space of the control model based on the deep deterministic policy gradient algorithm DDPG is the current water quality, light and flow, the action space is the change amount of the regulating valve opening and the adjustment amount of the light intensity, and the reward function of the deep deterministic policy gradient algorithm is set as the weighted sum of the pollutant removal rate and the energy consumption rate; specifically, the reward function R of the deep deterministic policy gradient algorithm satisfies: ; Among them, C initial is the initial pollutant concentration, C current is the current concentration, E max is the maximum allowed energy consumption, E current is the current energy consumption.

[0047] Specifically, when the LSTM model prediction shows that the total nitrogen TN removal rate in the corresponding substrate tank is lower than the set threshold, the optimal adjustment amount is calculated by the DDPG model, the water inflow of the high TN concentration tank is increased (such as from 2.0 L / h to 2.5 L / h), and the hydraulic retention time is extended; and for the photosynthetic microorganism dominated cylinder, the light intensity of the substrate tank is increased (such as from 3000 lux to 4000 lux).

[0048] (4) Corresponding to obtain water quality, gas and environmental parameter change data.

[0049] The COD concentration change in each substrate tank is shown in Table 1 below, and the above data can be quickly simulated by the method. By fusing LSTM time series prediction and deep reinforcement learning algorithm, compared with the traditional PID control, the parameter corresponding speed of the wetland water purification simulation device is improved by 40% and the system temperature is improved by 35%. Compared with the fixed parameter mode, the TN removal rate is increased by 18% to 25%, the sewage treatment compliance time is shortened by 30%, and the experiment shows that under the extreme condition of sudden increase of 50% of the simulated sewage concentration, the system can automatically adjust to a stable state within 10 minutes, the water quality fluctuation standard deviation is less than 5%, in addition, by dynamically adjusting the light and flow, the total energy consumption of the experiment is less than 40% (from an average of 25W to 15W).

[0050] .

[0051] While certain preferred embodiments of the application have been described, these embodiments do not represent the full scope of applicants' invention. Rather, the disclosure herein also contemplates additional modifications and variations that can be made to the described embodiments by those skilled in the art utilizing no more than their general knowledge. Accordingly, it is intended that the appended claims cover all such additional modifications and variations as fall within the scope of applicants' invention.

[0052] It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, it is intended that the present application cover modifications and variations of this application provided they come within the scope of the appended claims and their equivalents.

Claims

1. A wetland water quality purification simulation apparatus, characterized by comprising: The device comprises a sewage tank with a water outlet for soilless cultivation of floating plants, at least two substrate tanks for loading corresponding substrates, each of which is independent of the other and has a bottom higher than that of the sewage tank, a sensor array for real-time monitoring of water quality parameters, gas composition, light intensity and temperature arranged in the sewage tank and each of the substrate tanks, a controller in corresponding communication connection with the sensor array, and an adjustable light source arranged above the device and in corresponding electrical connection with the controller; the water inlets of each of the substrate tanks are in communication with the water outlet of the sewage tank through corresponding parallel water supply pipelines provided with regulating valves; the sewage tank is arranged at a height lower than the designed liquid level in the substrate tank; each of the substrate tanks comprises a sealing cover in corresponding sealing connection with the tank body and provided with a gas collecting port, and a gas path valve is arranged at the gas collecting port.

2. The wetland water quality purification simulation apparatus according to claim 1, wherein At least one overflow hole is arranged at a corresponding height position of the side wall of the sewage tank and each of the substrate tanks in the vertical direction.

3. The wetland water quality purification simulation apparatus according to claim 1, wherein The regulating valve is an electromagnetic valve in corresponding communication connection with the controller.

4. The wetland water quality purification simulation device according to claim 1 or 3, characterized by A liquid level meter in corresponding communication connection with the controller is arranged in the sewage tank and / or each of the substrate tanks.

5. The wetland water quality purification simulation device according to claim 1, characterized by The top cover of the sewage tank and the sealing cover are respectively made of transparent material.

6. The wetland water quality purification simulation device according to claim 1, characterized by A support for fixing the sewage tank and each of the substrate tanks is further provided, and the support is provided with a height adjusting assembly for adjusting the vertical height of the sewage tank; the water supply pipeline comprises a hose with a length greater than the adjustment range of the height adjusting assembly.

7. A method for simulating water purification in a wetland, based on the wetland water purification simulation apparatus according to claim 1, characterized by, The method comprises the following steps: (1) injecting the sewage to be treated into the sewage tank and planting floating plants on the surface of the sewage, and filling the substrate tank with corresponding substrate material, and adjusting the regulating valve between the substrate tank and the sewage tank to transport the sewage to be treated from the sewage tank to the substrate tank; (2) obtaining water quality, gas and corresponding environmental parameter data in real time through the sensor array, and inputting the data into an LSTM model to correspondingly predict the pollutant concentration change trend; (3) based on the deep deterministic policy gradient algorithm, adjusting the opening degree of the regulating valve and the light intensity according to the prediction result of the LSTM model to maximize the pollutant removal rate and minimize the energy consumption; (4) obtaining water quality, gas and environmental parameter change data.

8. The method according to claim 7, wherein In the step (2), the LSTM model is based on time series prediction, and the input includes historical and real-time parameter data of COD, TN, TP, CH4, CO2, N2, light intensity and temperature, and the output is the predicted value of the pollutant concentration in a preset future period.

9. The method according to claim 7, wherein In the step (3), the reward function of the deep deterministic policy gradient algorithm is the weighted sum of the pollutant removal rate and the energy consumption rate, and the action space includes the adjusting amount of the regulating valve opening degree and the adjusting amount of the light intensity.

10. The method according to claim 9, wherein In the step (3), the reward function R of the deep deterministic policy gradient algorithm satisfies: ； wherein, C initial Ci is the initial pollutant concentration, C current Ci is the current concentration, E max Ci is the maximum allowable energy consumption, E current Ci is the current energy consumption.

Images