Efficient multifunctional river ecological filter dam and operation method thereof

Abstract

The invention belongs to the technical field of river ecological protection, and relates to an efficient multifunctional river ecological filter dam and an operation method thereof. Comprising a filter dam body, and a rough filter layer, a middle filter layer, a fine filter layer and a base layer are sequentially arranged in the filter dam body in the water flow direction; the backwashing system comprises a water collecting main pipe, a water distributing pipe, a backwashing pipe, a backwashing branch pipe and a pulse type spray head; the water collecting main pipes extend into the filter dam body from the bank side and are arranged at intervals in the length direction of the filter dam, and the water distributing pipes are evenly distributed in the width direction of the filter dam and communicate with the water collecting main pipes in a one-to-one correspondence mode; each water distribution pipe is connected with a backwashing pipe which is parallel to the fine filtering layer and is provided with a gap, the backwashing pipe is uniformly connected with a plurality of rows of backwashing branch pipes which are upwards inclined along the water flow direction, the lower ends of the backwashing branch pipes are communicated with the backwashing pipe, and the upper ends of the backwashing branch pipes are connected with pulse type spray heads. According to the invention, the problems of easy blockage, low backwashing efficiency and the like of the ecological filter dam of the river channel are solved, and efficient purification and long-acting stable operation of the water body of the river channel are realized.

Description

Technical Field

[0001] The invention belongs to the technical field of river ecological protection and relates to a high-efficiency multifunctional river ecological filter dam and an operation method thereof. Background Art

[0002] Traditional ecological filter dams face significant challenges in urban river management. The high turbidity and suspended solids content of urban rivers make filter media susceptible to clogging, and frequent manual cleaning results in high maintenance costs. Furthermore, existing filter dams generally lack effective backwashing technology, making it difficult to adjust cleaning cycles based on dynamic changes in river flow, significantly reducing purification efficiency. Furthermore, direct discharge of backwash wastewater can cause secondary pollution to the river and surrounding environment. While backwash filters at water plants can address clogging issues, they require downtime for operation and are therefore unsuitable for use in open river channels. Some filter dams have added backwash piping but lack integrated wastewater recovery systems, resulting in low overall operational efficiency. These issues have severely limited the effectiveness of river ecological filter dams in urban river management. Summary of the Invention

[0003] The purpose of the present invention is to solve the problems in the prior art of easy clogging of river ecological filter dams, low backwash efficiency, and incomplete wastewater recovery, and to provide an efficient and multifunctional river ecological filter dam and its operation method, thereby achieving efficient purification of river water and long-term stable operation.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a highly efficient and multifunctional river ecological filter dam, comprising: The filter dam body has a coarse filter layer, a medium filter layer, a fine filter layer and a base layer arranged in sequence along the water flow direction; The backwash system includes a water collecting main pipe, a water distribution pipe, a backwash pipe, a backwash branch pipe, and a pulse nozzle. The water collecting main pipe extends from the shore to the interior of the filter dam body and is arranged at intervals along the length of the filter dam. The water distribution pipes are evenly distributed along the width of the filter dam and are connected to the water collecting main pipe in a one-to-one correspondence. Each water distribution pipe is connected to a backwash pipe arranged parallel to the fine filter layer with a gap. The backwash pipes are evenly connected to multiple rows of backwash branch pipes inclined upward along the water flow direction. The lower ends of the backwash branches are connected to the backwash pipes, and the upper ends are connected to pulse nozzles extending toward the fine filter layer and arranged in the gap between the fine filter layer. The water collecting main pipe is provided with a first control valve. The wastewater recovery mechanism includes a water collection tank arranged below the fine filter layer. A wastewater collection pipe is arranged in the water collection tank. One end of the wastewater collection pipe is connected to the water collection tank, and the other end extends to the shore water inlet through a mud pump.

[0005] Preferably, the surface of the filter dam body is paved with a Reynolds mat.

[0006] Preferably, both ends of the filter dam body are provided with Binger gabions filled with pebbles.

[0007] Preferably, the coarse filter layer is made of gravel with a particle size of 30~50mm; the medium filter layer is made of 80% pebbles with a particle size of 3-8cm, 15% green zeolite with a particle size of 2~5cm, 3% iron slag with a particle size of 1~5cm and 2% activated carbon; the fine filter layer is made of sand and gravel filter material; and the base layer is made of gravel.

[0008] Preferably, a one-way valve is provided in the backwash branch pipe, and the water flow direction is limited to one-way flow from the water distribution pipe to the pulse nozzle.

[0009] Preferably, the wastewater collecting pipe is connected to a water collecting branch, and the water collecting branch extends along the width direction of the filter dam to the upstream end of the filter dam.

[0010] Preferably, it further comprises an air collecting and aeration system; the air collecting and aeration system comprises an air collecting main pipe arranged in parallel with the water collecting main pipe, and the air collecting main pipe is externally connected to an aerator; a second control valve is provided on the air collecting main pipe.

[0011] Preferably, it further comprises a water inlet sensor group arranged at the upstream end of the filter dam body and a water outlet sensor group arranged at the downstream end; each sensor group comprises a turbidity sensor, a differential pressure sensor and a flow meter.

[0012] Preferably, a PLC controller is further included, and the turbidity sensor, the differential pressure sensor and the flow meter are all connected to the PLC controller; the PLC controller is electrically connected to the first control valve and the mud pump respectively.

[0013] In a second aspect, the present invention provides a method for operating a high-efficiency multifunctional river ecological filter dam, comprising the following steps: The river water flows through the coarse filter layer, medium filter layer and fine filter layer of the filter dam body in turn for filtration; the first control valve is opened, and the backwash water passes through the water collection main pipe, water distribution pipe, backwash pipe and backwash branch pipe in turn, and is finally sprayed out by the pulse nozzle to reversely flush the fine filter layer; the flushing wastewater is collected in the water collection tank below the fine filter layer, and then transported to the shore water inlet through the wastewater collection pipe and mud pump; the first control valve is closed, the backwash is stopped, and the filtration operation state is restored.

[0014] Compared with the prior art, the present invention has the following beneficial effects: The present invention adopts a three-dimensional pipe network layout in which the main water collection pipes are arranged at intervals along the length of the filter dam and the water distribution pipes are evenly distributed along the width direction, in conjunction with the backwash pipes arranged parallel to the fine filter layer and multiple rows of inclined backwash branches evenly connected thereto, to form a backwash network with all-round coverage; by setting the gap between the pulse nozzle and the fine filter layer and designing the inclination angle of the backwash branch pipe, a strong pulse water flow is generated to efficiently clean the filter layer; at the same time, the water collection trough and wastewater collection pipe arranged below the fine filter layer constitute a complete wastewater recovery channel, ensuring the efficient collection and treatment of backwash wastewater. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 This is a structural schematic diagram of a high-efficiency multifunctional river ecological filter dam according to the present invention; Figure 2 for Figure 1 A partial enlarged view of .

[0017] Among them: 1. Coarse filter layer; 2. Medium filter layer; 3. Fine filter layer; 4. Reno pad; 5. Binger gabion; 6. Backwash branch pipe; 7. Backwash pipe; 8. Water distribution pipe; 9. One-way valve; 10. Wastewater collection pipe; 11. Base layer; 12. Water collection branch pipe; 13. Pulse nozzle; 14. Water collection main pipe; 15. Partition. DETAILED DESCRIPTION

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations.

[0019] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention as claimed, but rather merely represents selected embodiments of the present invention. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort shall fall within the scope of protection of the present invention.

[0020] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0021] In the description of the embodiments of the present invention, it should be noted that if the terms "upper," "lower," "horizontal," "inner," etc. appear, the orientation or positional relationship indicated is based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship in which the inventive product is typically placed when in use. These terms are merely for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or component referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limitations on the present invention. In addition, the terms "first," "second," etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0022] In addition, if the term "horizontal" appears, it does not mean that the component must be absolutely horizontal, but can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0023] In the description of the embodiments of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," and "connected" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0024] The present invention is described in further detail below with reference to the accompanying drawings: The first object of the present invention is to provide a highly efficient and multifunctional river ecological filter dam. Figure 1 Shown, including: The filter dam body has a coarse filter layer 1, a medium filter layer 2, a fine filter layer 3 and a base layer 11 arranged in sequence along the water flow direction; The backwash system includes a main water collection pipe 14, a water distribution pipe 8, a backwash pipe 7, a backwash branch pipe 6, and a pulse nozzle 13. The main water collection pipe 14 extends from the shore to the interior of the filter dam body and is arranged at intervals along the length of the filter dam. The water distribution pipes 8 are evenly distributed along the width of the filter dam and are connected to the main water collection pipe 14 in a one-to-one correspondence. Each water distribution pipe 8 is connected to a backwash pipe 7 arranged parallel to the fine filter layer 3 with a gap therebetween. The backwash pipes 7 are evenly connected to multiple rows of backwash branch pipes 6 that are inclined upward in the direction of water flow. The lower ends of the backwash branch pipes 6 are connected to the backwash pipe 7, and the upper ends are connected to the pulse nozzles 13 that extend toward the fine filter layer 3 and are arranged in the gap therebetween. The main water collection pipe 14 is provided with a first control valve. The wastewater recovery mechanism includes a water collection tank arranged below the fine filter layer 3. A wastewater collection pipe 10 is arranged in the water collection tank. One end of the wastewater collection pipe 10 is connected to the water collection tank, and the other end extends to the shore water inlet through a mud pump.

[0025] The high-efficiency and multifunctional river ecological filter dam provided by the present invention realizes efficient purification and long-term stable operation of river water bodies through the synergistic effect of various components. The coarse filter layer 1 in the filter dam body mainly intercepts larger particle suspended matter, the medium filter layer 2 realizes the removal and biodegradation of medium-sized particle pollutants through a specially proportioned composite filter material, the fine filter layer 3 is responsible for fine filtration and deep purification, and the base layer 11 provides stable support for the entire filter dam. The main water collection pipe 14 and the water distribution pipe 8 in the backwash system form an efficient water flow distribution network. The inclined arrangement of the backwash pipe 7 and the backwash branch pipe 6 cooperates with the pulse nozzle 13 to generate a strong pulse water flow, which can effectively remove pollutants accumulated in the filter layer. The water collection tank and the wastewater collection pipe 10 in the wastewater recovery mechanism form a complete wastewater collection channel to ensure that the backwash wastewater can be completely recovered and treated. The various components of the present invention work together to ensure a continuous and stable water purification effect. The scientific self-cleaning system design solves the problem of easy clogging of traditional filter dams, achieving a dual improvement in purification efficiency and operational stability, enabling the filter dam to maintain an efficient operating state for a long time, greatly reducing maintenance costs, and providing reliable technical support for improving river water quality.

[0026] The filter dam body is covered with a Reno mattress 4, and both ends of the dam are equipped with pebble-filled Binger gabions 5. The Reno mattresses 4, with their flexible structure and excellent water permeability, effectively stabilize the overall structure of the filter dam, preventing deformation caused by scouring, while also ensuring proper water infiltration. The Binger gabions 5 at both ends of the dam are filled with pebbles, which not only evenly disperse the water flow and prevent localized scouring, but also provide a preliminary interception of large floating objects and coarse particles in the water.

[0027] The coarse filter layer 1 is made of gravel with a particle size of 30-50 mm, effectively intercepting large suspended solids and floating impurities in the water. The medium filter layer 2 is composed of 80% pebbles with a particle size of 3-8 cm, 15% chlorite with a particle size of 2-5 cm, 3% iron slag with a particle size of 1-5 cm, and 2% activated carbon. The chlorite removes ammonium nitrogen from the water through ion exchange; the iron slag fixes phosphates through chemical precipitation; and the activated carbon enriches organic pollutants through its strong adsorption capacity. These three materials together form a porous media system, providing an ideal attachment support for microbial communities. After microorganisms form a biofilm on the filter media surface, nitrifying bacteria in the aerobic zone gradually convert ammonia nitrogen into nitrite and nitrate. In the anoxic zone, denitrifying bacteria reduce nitrate to nitrogen gas for excretion, achieving nitrogen removal. Simultaneously, phosphate-accumulating bacteria remove phosphorus from the water through a metabolic cycle of aerobic phosphorus uptake and anaerobic phosphorus release, ultimately removing phosphorus from the water in the form of excess sludge. This synergistic effect of multiple mechanisms enables the intermediate filter layer 2 to simultaneously perform physical retention, chemical precipitation, and biodegradation functions, significantly improving nitrogen and phosphorus removal efficiency. The fine filter layer 3, made of sand and gravel filter media, effectively intercepts tiny particles, ensuring clear effluent. The base layer 11, made of gravel, provides stable support and drainage for the entire filter dam.

[0028] A one-way valve 9, preferably a Tesla one-way valve, is installed within the backwash branch 6, limiting water flow from the water distribution pipe 8 to the pulse nozzle 13. Under normal filtration conditions, the one-way valve 9 completely blocks the filtered water from entering the backwash branch 6, ensuring that all treated water passes through the filter layer for purification. Simultaneously, while maintaining normal filtration function, the backwash operation can be carried out independently, achieving complete diversion of the filtered and backwash water flows.

[0029] like Figure 2 As shown, a partition 15 is provided at the bottom of the fine filter layer 3. On the one hand, the partition 15 plays a supporting role to prevent the filter material from collapsing or losing, thereby maintaining stable filtering performance; on the other hand, gaps are provided on the partition 15 to allow backwash water to pass through evenly, so that the water flow can pass through the partition 15 to flush the filter material layer, effectively remove the trapped impurities, and restore the filtering capacity of the filter layer, thereby improving the operating efficiency of the system and extending the service life of the filter material.

[0030] The wastewater collection pipe 10 is connected to a water collection branch 12, which extends along the width of the filter dam to the upstream end of the filter dam. It can guide part of the further purified backwash wastewater back to the water inlet end of the filter dam to achieve the recycling of water resources.

[0031] For example, a mobile backwash vehicle is installed on the shore. The shore water inlet is connected to the mobile backwash vehicle's inlet via a quick-connect connector, and the water collection main 14 is connected to the mobile backwash vehicle's outlet. Wastewater is processed by the mobile backwash vehicle, which is equipped with a mud compactor to dehydrate the wastewater. Solid impurities are dehydrated by the mud compactor to form a mud cake that is easy to transport and dispose of, effectively preventing secondary contamination of the river by sludge. At the same time, the purified water can be recycled for subsequent backwash operations, achieving a closed-loop utilization of water resources and significantly reducing operating costs.

[0032] In one embodiment of the present invention, the filter dam also includes an air collection and aeration system. This system includes an air collection main pipe connected in parallel with the water collection main pipe 14, and an external aerator connected to the air collection main pipe. A second control valve is provided on the air collection main pipe. By regulating the opening and closing states of the first and second control valves, the filter dam has three switchable cleaning modes: pure water flushing mode (with the first control valve open), pure air flushing mode (with the second control valve open), and air-water mixed flushing mode (with both the first and second control valves open).

[0033] In one embodiment of the present invention, the filter dam further comprises an inlet sensor group located at the upstream end of the filter dam body and an outlet sensor group located at the downstream end. Each sensor group includes a turbidity sensor, a differential pressure sensor, and a flow meter. The inlet sensor group located at the upstream end of the filter dam body monitors inlet turbidity, differential pressure, and flow rate parameters in real time, dynamically monitoring the raw water quality. The outlet sensor group located at the downstream end continuously monitors purified water quality indicators, forming a complete treatment effect evaluation system.

[0034] The turbidity sensor, differential pressure sensor, and flowmeter are all connected to the PLC controller, which is electrically connected to the first control valve, mud pump, and second control valve. All sensor data is transmitted to the PLC controller for intelligent analysis and processing, establishing a closed-loop feedback control system. The PLC controller intelligently adjusts the opening and closing states of the first and second control valves based on changes in water quality parameters, enabling automatic switching between various cleaning modes, such as pure water flushing and air-water mixed flushing. It also precisely controls the operating conditions of the mud pump, ensuring efficient and stable operation of the wastewater recovery system.

[0035] The backwash procedure performed using a PLC controller includes the following steps: S1: Data collection and preprocessing, and dynamic adjustment of collection frequency based on historical data and real-time collected data S11: Sensor placement and data acquisition A pair of turbidity sensors, flow meters and differential pressure sensors are respectively arranged near the coarse filtration layer 1 and the fine filtration layer 3 to obtain water body data before and after filtration.

[0036] The turbidity sensor monitors the turbidity of the water in and out of the filter dam in real time, and obtains information on changes in the content of suspended matter in the water.

[0037] The differential pressure sensor measures the pressure difference at both ends of the filter layer, reflecting the degree of clogging of the filter material.

[0038] The ultrasonic flow meter uses the difference in the propagation speed of ultrasonic waves in the fluid to calculate the flow rate, monitor the total amount and flow rate of water entering the filter dam in real time, and provide basic data for subsequent analysis of the purification load of the filter dam.

[0039] At the same time, temperature and humidity sensors and rain gauges are set up to collect environmental data synchronously.

[0040] S12: Dynamically adjust data collection frequency According to the operation characteristics of the filter dam and the dynamic changes of water flow, a reasonable basic data collection frequency (f base , unit Hz).

[0041] During periods of large water flow fluctuations (such as after heavy rain or during peak water usage), increase the frequency of data collection to ensure that data changes are captured in a timely manner.

[0042] During periods of relatively stable water flow, the acquisition frequency should be appropriately reduced to reduce the amount of data processing.

[0043] The formula for dynamic adjustment of data collection frequency is:

[0044] in, is the frequency of data collection; is the basic frequency, set as the benchmark acquisition frequency when the system is running stably (unit: Hz), is the rate of change of water flow velocity, which is the ratio of the difference between the current water flow velocity and the historical average water flow velocity to the historical average. is the rate of change of turbidity, based on the ratio of the difference between the current and historical averages to the historical average, is the pressure difference change rate, based on the ratio of the difference between the current and historical averages to the historical average, : Water velocity weight (0.2-0.4), : turbidity weight (0.3-0.5), : Pressure difference weight (0.3~0.5).

[0045] Dynamically adjust the calculated frequency Set the constraint threshold range ( , ), is the minimum frequency threshold, is the maximum frequency threshold. Calculated based on real-time data Then, use the constraint threshold to check: if Exceeding the constraint threshold range ( < or > ), the closest threshold endpoint value ( or ) as the actual acquisition frequency. Within the constraint threshold ( <= <= ), then directly adopt As the actual acquisition frequency.

[0046] S13: Data Preprocessing The collected raw data may contain noise, outliers or missing values ​​and need to be preprocessed.

[0047] Use filtering algorithms (such as moving average, Kalman filtering, etc.) to remove noise interference in the data.

[0048] Fill missing data through data interpolation methods (such as linear interpolation, spline interpolation, etc.).

[0049] Use statistical analysis methods (e.g., based on standard deviation, boxplots, etc.) to identify and correct outliers.

[0050] Ensure the reliability of the data used for subsequent analysis and decision-making. Given the significant temporal correlations associated with filter dam clogging (e.g., turbidity accumulation over time and a gradual increase in differential pressure), a model that can capture long-term dependencies is required (a task for which traditional feedforward neural networks are not well suited).

[0051] S2: Feature Engineering and Model Training S21: Feature Engineering The preprocessed turbidity (T), pressure difference (P) and flow (V) data are subjected to feature extraction and combination.

[0052] Raw data processing: The preprocessed data is standardized (using the Z-score method) so that different features have the same scale to facilitate model learning.

[0053] The standardization formula is:

[0054] in, To standardize data; is the original data, is the mean of the original data, is the standard deviation of the original data.

[0055] Derived feature calculation: Change rate: Calculate the change rate of turbidity, pressure difference and flow rate to reflect the dynamic change trend of the data. For example, the calculation formula for turbidity change rate is:

[0056] in, is the turbidity change rate, is the turbidity at the current moment, The turbidity is the turbidity at the previous moment. The turbidity change rate helps to promptly detect rapid changes in indicators (such as a sharp increase in suspended matter content and accelerated blockage risk).

[0057] Calculate the cumulative value of turbidity, pressure difference and flow rate within a certain period of time to reflect the change in total volume. For example, the calculation formula for the cumulative turbidity value is:

[0058] in, is the cumulative value of turbidity, n is the number of samples in the statistical time period, is the turbidity at time i; the cumulative value reflects the load borne by the filter dam over a period of time, which helps to analyze the impact of long-term operating conditions on the clogging trend.

[0059] Sliding statistical features: Use a sliding window to calculate the mean, variance and other statistical features of the data. For example, calculate the sliding mean of turbidity over k time steps:

[0060] Where, is the average value of turbidity data within the sliding window; The sliding window size determines the number of samples involved in the calculation; 、 are the starting and ending positions of the window respectively, with the current time It is a centrally symmetrical expansion.

[0061] The sliding statistics feature can smooth data fluctuations, highlight data trends, and provide more stable input information.

[0062] Among them, the turbidity change rate and the turbidity cumulative value characterize the characteristics of turbidity in different time dimensions. The former is the change ratio at adjacent moments, reflecting the short-term dynamic change speed, and timely warning of blockage risks through real-time monitoring, while the latter is the sum of turbidity in a period of time, which is the long-term total load. The total load that the filter dam bears during this period can be quantified through the counting energy, and the blockage trend under long-term operation can be analyzed.

[0063] Feature combination: Combine the original data and derived features to form a richer feature vector. For example: P ],in , , represents the sliding mean.

[0064] S22: Model training and validation Data Preparation: Collect at least one year of historical filter dam operation data covering different seasons and flow conditions as a training set to ensure data diversity and representativeness. Arrange the data chronologically and mark the time and degree of filter dam blockage.

[0065] Model architecture (example - LSTM / GRU network model): Input layer: The input dimension is the number of features (e.g., 12 features). A fixed time window (e.g., 6 hours of historical data with a sliding window step of 15 minutes) is used to form a sequence input (e.g., sequence length 24). All features are Z-score normalized.

[0066] Hidden layer: Use two layers of bidirectional LSTM or GRU units, with 128 neurons per layer. LSTM uses the tanh activation function, while GRU uses sigmoid and tanh. A dropout layer (rate = 0.2) is added after each layer to prevent overfitting.

[0067] Output layer: Single-node output, using the sigmoid activation function, outputs the probability (0-100%) of the filter dam being blocked within the next 6 hours.

[0068] Training configuration: Loss function: Mean Squared Error (MSE).

[0069] Optimizer: Adam optimizer.

[0070] Training process: Use k-fold cross-validation (usually k=5 or 10) to divide the training set into k subsets. Each time, use k-1 subsets for training and the remaining subset for validation, repeating this cycle k times.

[0071] The feature vector is fed into the model for training. The model calculates the predicted value, calculates the gradient of the loss function through backpropagation, and uses the optimizer to update the parameters.

[0072] Record the loss value of each training batch and the evaluation indicators (precision, recall, etc.) of the validation set.

[0073] Model tuning: Adjust model parameters (such as the number of network layers and neurons) based on validation results (e.g., validation loss no longer decreases or evaluation metrics stagnate). A trade-off between model complexity and the risk of overfitting is necessary.

[0074] Model Evaluation: A separate test set is used for the final evaluation of the optimized model.

[0075] Evaluation indicators include: Accuracy, Recall, F1-Score, Root Mean Square Error (RMSE), etc., which comprehensively evaluate the prediction performance of the model.

[0076] S3: Backwash cycle decision S31: Setting the threshold According to the design parameters of the filter dam (porosity, filter material particle size distribution, water flow channel size, etc.), historical operating experience and target water quality requirements, a reasonable turbidity threshold (T threshold ), pressure difference threshold (P threshold ) and flow threshold (V min ≤V threshold ≤V max ).

[0077] Based on design parameters: Calculate the theoretically allowable maximum turbidity, pressure difference, and flow range (e.g., determine the flow range through porosity, channel area, and design flow rate; estimate the maximum turbidity tolerance based on the adsorption / filtration capacity of the filter media; and use hydraulics to calculate the maximum allowable pressure difference).

[0078] Combined with operational experience: Analyze historical data (especially changes in indicators before and after backwashing), observe signs of filter dam blockage or water quality degradation under different indicator values, and fine-tune the design threshold (for example, if it is found that the filtration effect is affected when the turbidity reaches 80% of the design value, lower the turbidity threshold).

[0079] Based on water quality requirements: Refer to relevant water quality standards (such as "Surface Water Environmental Quality Standard" GB3838-2002), combine the actual treatment capacity of the filter dam, and set a threshold that meets the final effluent water quality requirements (for example, if the drinking water source has high requirements, the turbidity threshold will be more stringent).

[0080] S32: Decision Logic Real-time monitoring and analysis: Continuously monitor sensor data (turbidity, pressure difference, flow) and the blockage probability (P predict ).

[0081] Trigger backwash operation: When the prediction model shows that the filter dam is about to be blocked (for example, P predict >60%), and turbidity (T>=T threshold ), pressure difference (P>=P threshold ) or flow (V<=V min or V >= V max ) reaches or exceeds its threshold, the backwash operation is triggered immediately.

[0082] Dynamically adjust backwash parameters: Dynamically adjust the intensity (usually expressed in flushing water pressure) and duration of backwashing according to the severity and changing trend of the exceeded indicators.

[0083] Turbidity trigger: The turbidity slightly exceeds the threshold and rises slowly: low intensity (such as 0.2 MPa), short time (such as 10 minutes).

[0084] The turbidity significantly exceeds the threshold and continues to rise: high intensity (such as 0.4 MPa) and long time (such as 20 minutes).

[0085] Pressure differential trigger: The pressure difference rises slowly to the threshold: moderate intensity (such as 0.3 MPa) and moderate time (such as 15 minutes).

[0086] The pressure difference rises sharply to the threshold value: high intensity (such as 0.6 MPa) and long time (such as 30 minutes).

[0087] Traffic trigger: The flow rate is continuously lower than the lower limit (V <V min ): Stronger backwash (medium / high intensity), designed to clear blockages and restore flow.

[0088] The flow rate is continuously higher than the upper limit (V>V max ): While checking for potential causes (such as structural problems and abnormal water flow), perform backwashing with appropriate intensity to prevent the problem from worsening.

[0089] S33: Optimization Algorithms (Reinforcement Learning) Reinforcement learning algorithms (such as deep Q-network DQN) are used to continuously optimize the backwash cycle decision strategy.

[0090] State definition (s): The filter dam operation state is abstracted into a state vector, for example, s=[T,P,V,P predict ](including real-time turbidity, pressure difference, flow rate and predicted blockage probability).

[0091] Action Space Definition (A): Backwashing is an action, including whether to perform backwashing, the intensity, and duration of the backwash. Backwash intensity can be represented by the level of flushing water pressure (e.g., low water pressure 0.2 MPa, medium water pressure 0.3 MPa, high water pressure 0.4 MPa), and the duration can be set to several optional values ​​(e.g., 10 minutes, 15 minutes, 20 minutes). The action space can be represented as a discrete set A = {(0,0,0), (1,0.2,10), (1,0.3,15), (1,0.4,20), ...}, where (0,0,0) indicates no backwashing, and the remaining elements indicate backwashing with the corresponding intensity and duration settings.

[0092] The reward function is used to measure the impact of each action on the filter dam's operating performance. The design of the reward function comprehensively considers factors such as the filter dam's operating efficiency and water quality improvement effect. For example, the reward function can be defined as:

[0093] Where ΔT reduction is the turbidity reduction value after backwashing, T threshold is the turbidity threshold; ΔP reduction is the pressure difference reduction value after backwashing, P threshold is the pressure difference threshold; ΔV recovery V is the ratio of flow rate returning to normal range after backwashing. threshold is the change in the normal range of flow; E consumption is the energy consumed by backwashing, E standard is the standard backwash energy consumption; w1, w2, w3, and w4 are weight coefficients, which are set according to the importance of different factors (for example, w1=0.4, w2=0.3, w3=0.2, and w4=0.1).

[0094] At the same time, Deep Q-Network (DQN) training is used: A Deep Q-Network is used for reinforcement learning training. First, the Q-Network parameters, including the number of neural network layers and neurons, are initialized. The current state is input into the Q-Network, the Q-value of each action is calculated, and the action with the highest Q-value is selected as the current action. After executing the action, the feedback from the environment, namely the new state and reward value, is observed. The current state, action, reward value, and new state are stored in the experience replay buffer. A batch of samples is randomly drawn from the experience replay buffer at regular intervals to train the Q-Network. The Q-Network parameters are updated by minimizing a loss function (such as the mean squared error loss function), enabling the Q-Network to more accurately estimate the Q-value of each action. After repeated training, the DQN algorithm gradually learns the optimal backwashing strategy, achieving dynamic optimization of the backwashing cycle.

[0095] The backwash program, centered around a PLC controller, organically integrates deep learning, reinforcement learning, and automated control technologies to form an intelligent backwash decision-making system. The system employs a combination of dynamic prediction and real-time control. On the one hand, it uses a deep learning algorithm to intelligently predict filter media blockage; on the other hand, it uses reinforcement learning technology to autonomously optimize backwash parameters. Furthermore, by integrating a multi-source sensor network, key parameters such as water quality and flow rate are collected in real time, providing data support for intelligent decision-making. This approach not only overcomes the limitations of traditional timed backwashing but also significantly improves the stability and adaptability of filter dam operations, offering a new solution for the intelligent management of water treatment facilities.

[0096] A second object of the present invention is to provide an operating method for an efficient and multifunctional river ecological filter dam, comprising the following steps: Filtration and purification stage: The river water first undergoes preliminary filtration through the Binger gabions 5 at both ends of the filter dam to intercept large floating objects and coarse particles of impurities; the water then flows through the coarse filter layer 1, the medium filter layer 2, and the fine filter layer 3 for graded purification: the coarse filter layer 1 intercepts large suspended particles of 30-50mm; the medium filter layer 2 removes pollutants through physical retention, chemical precipitation, and biodegradation of composite filter media; the fine filter layer 3 performs fine filtration to ensure clear effluent; the purified water is discharged through the base layer 11.

[0097] Backwash stage: When the filter layer is blocked and reaches the set threshold, the backwashing process is started; Select the backwash mode according to the blockage situation: Pure water flushing mode: open the first control valve, and high-pressure water flows through the water collecting main pipe 14, the water distribution pipe 8, the backwash pipe 7 and the backwash branch pipe 6 in sequence, and is sprayed out by the pulse nozzle 13; Pure gas flushing mode: open the second control valve, and compressed air enters the system through the gas collecting main pipe; Air-water mixed flushing mode: Open the first control valve and the second control valve at the same time, and the backwash water flow / air flow will flush the filter layer from bottom to top to remove accumulated pollutants.

[0098] Wastewater recycling stage: The backwash wastewater is collected in a water collection tank and transported to the shore for treatment through a wastewater collection pipe 10 and a mud pump; part of the wastewater flows back to the water inlet of the filter dam through a water collection branch 12 for recycling.

[0099] System maintenance phase: Regularly check the condition of the filter layer and the integrity of the Reno Pad 4; adjust the backwash frequency and mode according to changes in water quality; clean up the recovered sludge and dispose of it harmlessly.

[0100] The present invention significantly improves the comprehensive benefits of river ecological filter dams by optimizing the filter material structure and precise backwash control. In terms of operational performance, the intelligent backwash system effectively reduces the risk of filter material clogging and significantly extends the filter material replacement cycle and equipment service life. In terms of ecological benefits, the multi-layer graded filtration system combined with timely backwash operations significantly improves the effluent water quality, provides a good living environment for aquatic organisms, and effectively promotes the stable recovery of the river ecosystem. In terms of economic benefits, the system significantly reduces the comprehensive operation and maintenance costs of wastewater treatment, manual cleaning, and equipment maintenance through wastewater recycling and automated operation, thereby achieving efficient recycling of resources.

[0101] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A highly efficient and multifunctional river ecological filter dam, characterized in that: include: A filter dam body, wherein a coarse filter layer (1), a medium filter layer (2), a fine filter layer (3) and a base layer (11) are sequentially arranged inside the filter dam body along the water flow direction; A backwashing system comprises a water collecting main pipe (14), a water branch pipe (8), a backwashing pipe (7), a backwashing branch pipe (6) and a pulse nozzle (13); the water collecting main pipe (14) extends from the shore to the interior of the filter dam body and is arranged at intervals along the length direction of the filter dam, and the water branch pipes (8) are evenly distributed along the width direction of the filter dam and are connected to the water collecting main pipe (14) in a one-to-one correspondence; each of the water branch pipes (8) is connected to a backwashing pipe (7) arranged parallel to the fine filter layer (3) and with a gap therebetween, and the backwashing pipe (7) is evenly connected to a plurality of rows of backwashing branch pipes (6) inclined upward along the water flow direction, the lower end of the backwashing branch pipe (6) is connected to the backwashing pipe (7), and the upper end is connected to a pulse nozzle (13) extending toward the fine filter layer (3) and arranged with a gap therebetween; the water collecting main pipe (14) is provided with a first control valve; The wastewater recovery mechanism comprises a water collecting tank provided below the fine filter layer (3), wherein a wastewater collecting pipe (10) is provided in the water collecting tank, one end of the wastewater collecting pipe (10) is connected to the water collecting tank, and the other end of the wastewater collecting pipe (10) is extended to a shore water inlet via a mud pump.

2. The high-efficiency multifunctional river ecological filter dam according to claim 1, characterized in that: The surface of the filter dam body is paved with a Reynolds pad (4).

3. The high-efficiency multifunctional river ecological filter dam according to claim 1, characterized in that: Both ends of the filter dam body are provided with Binger gabions (5) filled with pebbles.

4. The high-efficiency multifunctional river ecological filter dam according to claim 1, characterized in that: The coarse filter layer (1) is made of gravel with a particle size of 30-50 mm; the medium filter layer (2) is made of 80% pebbles with a particle size of 3-8 cm, 15% green zeolite with a particle size of 2-5 cm, 3% iron slag with a particle size of 1-5 cm, and 2% activated carbon; the fine filter layer (3) is made of sand and gravel filter material; and the base layer (11) is made of gravel.

5. The high-efficiency multifunctional river ecological filter dam according to claim 1, characterized in that: A one-way valve (9) is provided in the backwash branch pipe (6), and the water flow direction is limited to a one-way flow from the water distribution pipe (8) to the pulse nozzle (13).

6. The high-efficiency multifunctional river ecological filter dam according to claim 1, characterized in that: The wastewater collection pipe (10) is connected to a water collection branch (12), and the water collection branch (12) extends along the width direction of the filter dam to the upstream end of the filter dam.

7. The high-efficiency multifunctional river ecological filter dam according to claim 1, characterized in that: It also includes an air collection and aeration system; the air collection and aeration system includes an air collection main pipe arranged in parallel with the water collection main pipe (14), and the air collection main pipe is externally connected to an aerator; a second control valve is provided on the air collection main pipe.

8. The high-efficiency multifunctional river ecological filter dam according to claim 1, characterized in that: It also includes a water inlet sensor group arranged at the upstream end of the filter dam body and a water outlet sensor group at the downstream end; each sensor group includes a turbidity sensor, a differential pressure sensor and a flow meter.

9. The high-efficiency multifunctional river ecological filter dam according to claim 8, characterized in that: It also includes a PLC controller, and the turbidity sensor, the differential pressure sensor and the flow meter are all connected to the PLC controller; the PLC controller is electrically connected to the first control valve and the mud pump respectively.

10. The method for operating a high-efficiency multifunctional river ecological filter dam according to any one of claims 1 to 9, characterized in that: The following steps are involved: The river water flows through the coarse filter layer (1), the medium filter layer (2) and the fine filter layer (3) of the filter dam body in sequence for filtration; the first control valve is opened, and the backwash water passes through the water collecting main pipe (14), the water distribution pipe (8), the backwash pipe (7) and the backwash branch pipe (6) in sequence, and is finally sprayed out by the pulse nozzle (13) to backwash the fine filter layer (3); the flushing wastewater is collected in the water collecting tank below the fine filter layer (3) and transported to the shore water inlet through the wastewater collection pipe (10) and the mud pump; the first control valve is closed, the backwash is stopped, and the filtering operation state is restored.

Images