Self-adaptive water level adjusting method, system and equipment

Abstract

The invention provides a self-adaptive water level adjusting method, system and equipment. The method comprises the steps that an adjusting system is initialized, and the initial position of the electric push rod is determined; collecting current environment parameters; the environmental parameters comprise effluent temporary storage well flow, sewage effluent temporary storage well water quality, current position weather and current position ecological parameters; according to the environmental parameters, an optimal water level control strategy is calculated based on a wetland water level intelligent regulation and control algorithm; regulating and controlling the electric push rod according to the optimal water level control strategy; according to the invention, dependence on manual operation and maintenance is thoroughly eliminated through design, full-process intelligentization of autonomous sensing, autonomous decision making and autonomous execution is realized, and the operation and maintenance efficiency is greatly improved. The workload of operation and maintenance personnel is reduced by more than 70% through a full-automatic regulation and control process, and hysteresis and subjectivity of traditional manual inspection are avoided. Dynamic adaptation of water level regulation and control and the wetland microenvironment is achieved, and the operation goals of high precision, full working conditions, low energy consumption and high efficiency are achieved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent control, and particularly relates to a self-adaptive water level adjusting method, system and device. BACKGROUND

[0002] As a low-cost and eco-friendly sewage treatment and ecological restoration technology, constructed wetland has been widely used in village sewage treatment, municipal and industrial tail water advanced treatment, and non-point source pollution control, and has become a key technical unit in water environment treatment system, due to its high removal efficiency of COD, ammonia nitrogen, total phosphorus and other pollutants. At present, mature technical specifications and research basis have been formed in the early design stage of constructed wetland in China: in terms of core functional components, there are clear design guidelines for plant optimization (such as the adaptability screening of water plants such as reed and cattail) and filler gradation (such as the combination optimization of gravel, zeolite and activated carbon) for different climate zones and pollution types; in terms of operation parameter design, there are also perfect calculation methods and engineering case supports for the determination of parameters such as hydraulic retention time and pollutant concentration threshold of influent load. However, the long-term stable operation of constructed wetland highly depends on the dynamic adaptation of operation and management strategies, and related research lags behind the development of design technology, especially in the field of fine operation and regulation technology under complex working conditions.

[0003] In practice, constructed wetland often faces challenges of four typical complex working conditions, and existing operation and management means are difficult to form an effective response: firstly, when the influent flow or water quality is too low, the water level in the wetland continues to drop, leading to water shortage and oxygen imbalance of water plant roots, and the activity of microorganisms decreases sharply due to lack of nutrients, and the pollutant degradation efficiency is greatly reduced; secondly, when the influent flow or water quality is too high, the impact load in a short time exceeds the purification capacity of the wetland, and the high water level is easy to cause the plant stems to fall down, and destroy the pore structure of the microorganism-attached filler; thirdly, under the weather of heavy rain, a large amount of rainwater flows into the wetland, causing the water level to rise sharply, which not only causes secondary pollution due to sewage overflow, but also causes damage to the functional layer of the wetland due to the scouring of the wetland substrate; fourthly, under the ice freezing mode in winter in the north and south regions, the ice on the surface of the wetland water blocks the air passage of the plant, and the metabolism of microorganisms is stalled due to low temperature, and the purification function of the wetland under the traditional operation mode is almost lost. The operation difficulties under the above working conditions directly lead to a 30%-60% reduction of the actual pollutant removal rate of the constructed wetland compared with the design value, which seriously restricts the full play of its technical value.

[0004] In summary, due to the problems of insufficient control flexibility, low adjustment precision, weak multi-working condition adaptation ability and reaction lag, etc., the existing technology cannot realize the fine and dynamic regulation of the water level of the constructed wetland, which restricts the pollutant removal efficiency and ecological stability of the constructed wetland under complex operation scenarios. SUMMARY

[0005] In view of the defects of the prior art, the present application provides a self-adaptive water level adjusting method, system and device, which can effectively control according to the water inflow and outflow, water quality, rainfall and adverse weather in the pool, and build an AI self-management and control platform based on the Internet of Things, so as to effectively deal with the problem of unable to normally operate under various complex working conditions.

[0006] In the first aspect, the present application provides a self-adaptive water level adjusting method, comprising: Initializing the adjusting system to determine the initial position of the electric push rod.

[0007] Collecting current environmental parameters; the environmental parameters include the outflow of the water temporary storage well, the water quality of the sewage outflow temporary storage well, the current location meteorology and the current location ecological parameters.

[0008] According to the environmental parameters, the optimal water level control strategy is calculated based on a wetland water level intelligent control algorithm.

[0009] According to the optimal water level control strategy, the electric push rod is regulated.

[0010] The wetland water level intelligent control algorithm includes a multi-source data sensing and preprocessing module, a prediction model module and a multi-objective rolling optimization decision module.

[0011] The multi-source data sensing and preprocessing module is used for preprocessing the collected environmental parameters.

[0012] The dynamic prediction model predicts the water quality and quantity in the future time period based on the current environmental parameters.

[0013] The multi-objective rolling optimization decision module performs optimal water level prediction based on the prediction result of the dynamic prediction model and outputs the optimal water level control strategy.

[0014] Preferably, the initial position of the electric push rod is determined, comprising: Retrieving historical operation data and screening historical scenes with an environmental feature similarity greater than 85% at the current initialization time; Extracting the initial position data of the electric push rod in the corresponding scene from the screened similar historical scenes; Calculating the mean and standard deviation of the initial position data, and if the standard deviation is less than a preset value, the mean is directly taken as the initial position reference value; otherwise, the mean is recalculated after removing outliers as the initial position reference value.

[0015] According to the environmental characteristics and historical scene data at the current initialization moment, the flow weight coefficient, the water quality weight coefficient, the ecological weight coefficient and the meteorological weight coefficient are divided, and the final electric push rod initial position is calculated based on the initial position reference value, the flow weight coefficient, the water quality weight coefficient, the ecological weight coefficient and the meteorological weight coefficient.

[0016] Preferably, the multi-source data perception and preprocessing module comprises: The spatial correlation and time dependence of the environmental parameters are captured through parallel space-time branches, and TCN feature extraction and LSTM sequence modeling are performed on the current environmental parameters; The correlation between the parameters of the TCN features and the LSTM sequences is calculated through mutual information entropy, and a joint feature is constructed; at the same time, residual connection processing is performed on the TCN features and the LSTM sequences; An attention weight network is introduced, and the weight of each feature layer is dynamically adjusted according to the historical scene data to obtain a standardized feature vector at the current time as the input of the dynamic prediction model.

[0017] Preferably, before the TCN feature extraction and LSTM sequence modeling on the current environmental parameters, it further comprises: In the current environmental parameters, a distance matrix of the low-frequency parameter time series and the reference parameter time series is constructed, and the alignment path with the minimum cumulative distance is solved through dynamic programming; Along the alignment path, the low-frequency parameters in the current environmental parameters are interpolated into time series consistent with the high-frequency parameters; The standard deviation of each type of parameter is calculated for the aligned time series data, and a Gaussian noise sequence is generated, and the Gaussian noise sequence is superimposed with the original time series data to obtain enhanced time series data; The noise-enhanced time series data is arranged in reverse order by time to generate a flip sub-sequence, and the environmental scene label of the flip sub-sequence is retained to expand the data set of the current environmental parameters.

[0018] Preferably, the dynamic prediction model comprises: A residual TCN-bidirectional LSTM-attention fusion layer backbone network is constructed, and the network is trained with the MSE completion loss and the cross-entropy weight alignment loss as the joint loss function; The standardized feature vector is received, and is reshaped according to the region-time sequence dimension to match the spatial distribution of the wetland; a depth separable convolution combined with a TCN spatial feature weight is used to extract the parameter conduction features of the inlet area, the core purification area and the outlet area; a causal convolution combined with an LSTM time sequence feature weight is used to capture short-term time sequence trends, and a space-time basic encoding feature is output; The water quality target vector and the water quantity target vector are constructed, the water quality target vector and the water quantity target vector are weighted according to the attention weight, and are spliced with the space-time basic coding features, and the target related space-time coding features are obtained through convolution compression; The spatial separation convolution and the Sigmoid gate are adopted, the TCN feature correlation degree of the preprocessing module is combined, the key region is output, and the pure space feature is output. The time separation convolution and the Tanh gate are adopted, the LSTM sequence trend capturing time sequence mutation is output, and the pure time feature is output. The time sequence step difference of the pure time feature is calculated, the difference sequence is generated, and the parameter change rate is captured. The mutual information entropy of the difference feature and the prediction target is calculated, and the self-attention weight is generated. The weighted difference feature and the pure time feature are spliced and input into the 2-layer bidirectional LSTM, and the enhanced time feature is output. The pure space feature is expanded through a full connection layer and is spliced with the enhanced time feature, and the space-time enhanced feature is obtained. The mutual information correlation degree of the space-time enhanced feature and the water quality index is calculated, the step attention weight is generated, the weighted pooling is performed, and the water quality prediction value is output through the 3-layer full connection layer LeakyReLU activation. The space-time enhanced feature is input into the 2-layer bidirectional LSTM decoding layer for decoding, the decoding output is added after matching the input feature dimension, and the water quantity prediction value is output after being processed through the full connection layer.

[0019] Preferably, the prediction result based on the dynamic prediction model is used for optimal water level prediction, and an optimal water level control strategy is output, including: A two-dimensional data table is constructed according to the water quality prediction value, the water quantity prediction value and the confidence, low-confidence prediction points are removed according to a confidence threshold, multi-time granularity data is combined according to a regulation and control response priority, and a short-term prediction data set and a medium-and-long-term prediction data set are obtained. For the short-term prediction data set, whether the water quality at each time point meets the standard is judged according to the water quality prediction value. For the medium-and-long-term prediction data set, whether the water quantity threshold is met is judged according to the water quantity prediction value. The water level adjustment range required at present is determined according to the water quality prediction value, or the water quantity prediction value, or a fixed period, and the extension and retraction amount of the electric push rod is determined according to the electric push rod energy consumption characteristic curve.

[0020] In the second aspect, based on the same inventive concept, the application further provides an adaptive water level adjustment system, which comprises an electric push rod used in cooperation with a water outlet temporary storage well, the electric push rod is connected with a control cabinet, the control cabinet is connected with a meteorological link acquisition system, an ecological parameter acquisition system, a water flow monitoring probe and a water quality monitoring probe.

[0021] The control cabinet calculates an optimal water level control strategy based on a wetland water level intelligent control algorithm according to the collected environmental parameters, and regulates the electric push rod according to the optimal water level control strategy.

[0022] The environmental parameters include water flow of the water outlet temporary storage well, water quality of the water outlet temporary storage well, current location meteorology, and current location ecological parameters.

[0023] The wetland water level intelligent control algorithm includes a multi-source data sensing and preprocessing module, a prediction model module, and a multi-objective rolling optimization decision module.

[0024] The multi-source data sensing and preprocessing module is used for preprocessing the collected environmental parameters.

[0025] The dynamic prediction model predicts the water quality and quantity in a future time period based on current environmental parameters.

[0026] The multi-objective rolling optimization decision module performs optimal water level prediction based on the prediction result of the dynamic prediction model, and outputs an optimal water level control strategy.

[0027] Preferably, the system includes that the water flow monitoring probe and the water quality monitoring probe are arranged at the water inlet pipe opening of the water outlet temporary storage well. The electric push rod is connected to a telescopic plastic hose connected to a water outlet, the control cabinet determines the required water level adjustment range according to the water quality prediction value or the water quantity prediction value or a fixed period, determines the telescopic amount of the electric push rod in combination with the energy consumption characteristic curve of the electric push rod, and when the electric push rod controls the telescopic plastic hose to be below the water surface, the water in the water outlet temporary storage well is discharged through the telescopic plastic hose.

[0028] Preferably, the system further includes: The control cabinet further switches different operation modes according to the flow load and meteorological conditions.

[0029] In a third aspect, based on the same inventive concept, the present application further provides a computer device, which comprises a processor and a memory, the memory stores at least one instruction, the at least one instruction is loaded and executed by the processor to realize the adaptive water level adjustment method according to the first aspect.

[0030] Compared with the prior art, the present application has the following advantages: The present application completely gets rid of the dependence on manual operation and maintenance, realizes the whole-process intelligentization of autonomous sensing, autonomous decision-making and autonomous execution, and greatly improves the operation and maintenance efficiency. Through the fully automatic regulation and control process, the workload of operation and maintenance personnel is reduced by more than 70%, and the lag and subjectivity of traditional manual inspection are avoided.

[0031] The innovative residual TCN-bidirectional LSTM fusion prediction model of the application completely solves the problems of time-space feature fragmentation, mutation response lag and confidence loss of traditional prediction models. For four typical complex working conditions of low load, high load, rainstorm and freezing, accurate positioning, dynamic prediction and early warning and adaptive decision control can be realized, dynamic adaptation of water level regulation and wetland microenvironment is realized, and the operation goal of high precision, full working condition, low energy consumption and high efficiency is realized. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 is the adaptive water level regulation system diagram shown in the embodiment of the application.

[0033] Figure 2 is the adaptive water level regulation method process diagram shown in the embodiment of the application.

[0034] Among them, 1 is a water outlet temporary storage well; 2 is a water inlet pipe; 3 is a water outlet pipe; 4 is a wall sleeve; 5 is an electric push rod; 6 is a telescopic plastic hose; 7 is a ring-shaped hose fixing piece; 8 is a Doppler flow monitoring system (including a probe); 9 is a multifunctional water quality monitoring system (including a probe); 10 is a meteorological environment acquisition system; 11 is an ecological parameter acquisition system; 12 is an angle steel; 13 is a monitoring and electrical integrated control cabinet; 14 is a data sending and receiving device; 15 is a solar panel; 16 is a support frame; 17 is a cover plate. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical scheme and advantages of the embodiments of the application clearer, the technical scheme will be described clearly and completely below in combination with the embodiments of the application. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0036] Embodiment one: the application proposes an adaptive water level regulation method, which is mainly applied to water level management of a water outlet temporary storage well in an artificial wetland, as shown in Figure 1 The artificial wetland mainly includes: a water outlet temporary storage well, a water inlet pipe, a water outlet pipe, a wall sleeve, an electric push rod, a telescopic plastic hose, a ring-shaped hose fixing piece, a Doppler flow monitoring system (including a probe), a multifunctional water quality monitoring system (including a probe), a meteorological environment acquisition system, an ecological parameter acquisition system, an angle steel, a monitoring and electrical integrated control cabinet, a data sending and receiving device, a solar panel and an AI control platform based on the Internet of Things.

[0037] It should be noted that: (1) Electric push rod, with telescopic characteristics, material stainless steel or aluminum alloy, stroke 10-1500mm, 24V DC power supply, speed 10-180mm / s, IP65 waterproof level. The control system and power supply of the electric push rod are located in the integrated monitoring and electrical control cabinet.

[0038] (2) Telescopic plastic hose, with the functions of extension and retraction, matching the diameter of the constructed wetland effluent pipe.

[0039] (3) Ring hose fixing piece, made of stainless steel, circular clamp type, one end is fastened and connected with the constructed wetland effluent pipe by bolts, the other end is connected with the electric push rod by a screw rod.

[0040] (4) Doppler flow monitoring system, range 0-1000m3 / d, signal output: RS485 or USB interface. The system and power supply are located in the integrated monitoring and electrical control cabinet, and the probe is placed in the water inlet adjustment tank and the water outlet temporary storage well to monitor the flow.

[0041] (5) Multifunctional water quality monitoring system, with functions of monitoring COD Cr , ammonia nitrogen, total phosphorus, total nitrogen concentration, etc. The monitoring range of COD Cr is 0-500mg / L, the monitoring range of ammonia nitrogen is 0-100mg / L, the monitoring range of total phosphorus is 0-50mg / L, and the monitoring range of total nitrogen is 0-100mg / L. The signal output is: RS485 or USB interface. The system and power supply are located in the integrated monitoring and electrical control cabinet, and the probe is placed in the water inlet adjustment tank and the water outlet temporary storage well to monitor the water quality.

[0042] (6) Weather environment acquisition system: integrated with rain gauge (resolution 0.2mm), wind speed and direction sensor, atmospheric pressure sensor and photosynthetic active radiation sensor, to build a micro-meteorological station. This unit is specially designed with a 2-hour rainfall prediction function, which can predict the arrival of heavy rain by analyzing the sudden drop in air pressure and the rising trend of humidity.

[0043] (7) Ecological parameter acquisition system: including plant growth monitoring camera (equipped with multi-spectral lens), benthic animal activity sensor and bird sound recognition microphone, to quantify the ecological state of the wetland. The image recognition algorithm is used to analyze the leaf area index and root development of the plants.

[0044] (8) Angle steel, L40*3.5, used to fix the electric push rod by bolts.

[0045] (9) Integrated monitoring and electrical control cabinet, which integrates the Doppler flow monitoring system, multifunctional water quality monitoring system, solar control system and battery.

[0046] (10) Data transmission and receiving device, located at the top of the integrated monitoring and electrical control cabinet, uses GPRS for transmission.

[0047] (11) Solar panel, single solar panel power supply voltage 12-36V, peak power 100-500W, can use multiple parallel power supply. Its control system and battery are located in the integrated monitoring and electrical control cabinet.

[0048] (12) AI control platform based on Internet of Things, the platform has online monitoring, fault management, remote control, operation record and management, AI self-diagnosis program, AI self-analysis program. The platform has PC, web and mobile APP three use ways. The platform will be used to control electric push rod, monitoring system and solar system, with manual and automatic two control conditions.

[0049] It should be noted that the connection relationship of each part is as follows: The outlet pipe at the bottom of the constructed wetland is connected to the outlet temporary storage tank through a wall bushing. The wall bushing is a finished product, and conventional waterproof measures are taken.

[0050] The retractable plastic hose is connected to the outlet pipe at the bottom of the constructed wetland at one end with a ring-shaped hose fixing piece, and the other end is connected to the electric push rod with a bolt.

[0051] One end of the electric push rod is fixed to the angle steel with a bolt, and the other end is connected to the screw rod of the ring-shaped hose fixing piece.

[0052] The angle steel is connected to the outlet temporary storage tank with expansion bolts.

[0053] One end of the Doppler flow monitoring probe and the multifunctional water quality monitoring probe is placed in the outlet temporary storage tank, and the other end is connected to the control system in the integrated monitoring and electrical control cabinet.

[0054] The meteorological environment acquisition system and the ecological parameter acquisition system are powered by solar panels and use city power as backup power supply. They can be connected in real time with the AI control platform based on Internet of Things through signal transmission.

[0055] One end of the solar panel is connected to the control system, and the other end is connected to the battery. The control system and the battery are located in the integrated monitoring and electrical control cabinet. The battery supplies power to all electrical equipment in the control cabinet, including the control system, the electric push rod, the monitoring system, etc. To ensure 24-hour normal operation of the system, city power is used as backup power supply.

[0056] The AI management and control platform based on the Internet of Things collects data collected by Doppler flow monitoring probes, multifunctional water quality monitoring probes, meteorological environment collection systems and ecological parameter collection systems, organizes, analyzes and makes decisions on the data, and issues instructions to the electric push rod control system to regulate and control it.

[0057] Before system initialization, it also includes: Electric push rod debugging: The shortest distance of the electric push rod stroke design needs to be extended is the highest water level of the constructed wetland, and the longest distance is the lowest running water level of the constructed wetland. Before the constructed wetland starts to run, the electric push rod extension rod is adjusted to the designed water level height.

[0058] Solar power supply system debugging: Connect the solar panel, control system and battery for debugging. After debugging, connect the electric push rod, monitoring probe and other electrical equipment to the battery to realize power supply. At the same time, set up a bypass to the city power supply to ensure that when the solar power supply is insufficient, the city power supply can be switched to supply power.

[0059] Flow and water quality monitoring probe debugging: During the operation and debugging of the constructed wetland, the two monitoring probes are debugged and calibrated to ensure that the probe is sensitive and the data collection is normal.

[0060] Meteorological environment collection system debugging: Test whether the rain gauge can normally record the rainfall of each rainfall. Test whether the wind speed and direction sensor, atmospheric pressure sensor and photosynthetically active radiation sensor are normally operated.

[0061] Ecological parameter collection system: Test whether the plant growth monitoring camera (equipped with a multispectral lens), benthic animal activity sensor and bird sound recognition microphone can normally operate. Test whether it can analyze the plant leaf area index and root development condition through image recognition algorithm.

[0062] AI management and control platform based on the Internet of Things debugging: After all the components are debugged normally, start debugging the platform to test whether the platform can normally receive data sent from the monitoring probe, electric push rod, solar system, meteorological environment collection system and ecological parameter collection system. At the same time, the platform is debugged for remote / hand operation to ensure normal operation of the system In the embodiment, as shown in Figure 2 The adaptive water level adjustment method specifically includes the following steps: Step 1: Initialize the adjustment system and determine the initial position of the electric push rod.

[0063] Step 2: Collect current environmental parameters; the environmental parameters include the outflow temporary storage well flow, the sewage outflow temporary storage well water quality, the current location meteorological parameters and the current location ecological parameters.

[0064] Preferably, the starting sensing unit collects environmental parameters at a preset frequency: High-frequency parameters: temporary storage well flow rate (3 minutes / time), sewage temporary storage well water quality (5 minutes / time); Low-frequency parameters: current location weather (10 minutes / time), current location ecological parameters (8 minutes / time); Preferably, the collected data is transmitted to the control unit in real time to generate the original parameter time series data set.

[0065] Step 3: According to the environmental parameters, the optimal water level control strategy is calculated based on the wetland water level intelligent control algorithm.

[0066] Step 4: According to the optimal water level control strategy, the electric push rod is regulated.

[0067] The wetland water level intelligent control algorithm includes a multi-source data sensing and preprocessing module, a prediction model module, and a multi-objective rolling optimization decision module.

[0068] The multi-source data sensing and preprocessing module is used to preprocess the collected environmental parameters.

[0069] The dynamic prediction model predicts the water quality and quantity in the future time period based on the current environmental parameters.

[0070] The multi-objective rolling optimization decision module performs optimal water level prediction based on the prediction results of the dynamic prediction model and outputs the optimal water level control strategy.

[0071] In this embodiment, the determination of the initial position of the electric push rod includes: Retrieve historical operation data and filter historical scenes with environmental feature similarity greater than 85% at the current initialization time.

[0072] For the filtered similar historical scenes, the initial position data of the electric push rod under the corresponding scene is extracted.

[0073] Calculate the mean and standard deviation of the initial position data, and if the standard deviation is less than a preset value, the mean is directly taken as the initial position reference value; otherwise, after removing outliers, the mean is recalculated as the initial position reference value.

[0074] According to the environmental features and historical scene data at the current initialization time, the flow weight coefficient, water quality weight coefficient, ecological weight coefficient, and weather weight coefficient are divided, and the final electric push rod initial position is calculated based on the initial position reference value, flow weight coefficient, water quality weight coefficient, ecological weight coefficient, and weather weight coefficient.

[0075] For example, the wetland operation data of the past 3 hydrological periods (including the 2021-2023 rainy season, dry season, and freezing period) are called, totaling 10800 samples. Based on the environmental characteristics at the current initialization time (May 10, 2024, 9:00): the outflow temporary storage well flow is 4.2 m³ / h, the sewage COD is 38 mg / L, the ammonia nitrogen is 3.6 mg / L, the TP is 0.21 mg / L, the rainfall is 0 mm, the wind speed is 1.2 m / s, the temperature is 22°C, the vegetation coverage is 88%, the soil moisture content is 35%, and the aquatic organism population density is 0.8 tail / m³, the similarity with the historical samples is calculated: Meteorological similarity: calculate the Euclidean distance of rainfall, wind speed, and temperature: ; Water quality similarity: calculate the average relative error of COD, ammonia nitrogen, and TP, and select samples with an average value of <8%. Ecological similarity: calculate the cosine similarity of vegetation coverage and soil moisture content, and select samples with a similarity of >0.9.

[0076] Finally, 210 historical scenario data with environmental characteristic similarity >85% are obtained.

[0077] Initial position reference value determination: Extract the initial position data of the electric push rod of the 210 similar scenarios, ranging from 720-750mm, calculate the average position of 735mm, and the standard deviation of 8.2mm. The preset standard deviation threshold is 10mm, and since 8.2mm <10mm, 735mm is directly taken as the initial position reference value.

[0078] Combined with the current environmental characteristics (stable flow, water quality close to design value, clear weather, and good ecological state), the weight coefficients are divided: flow weight 0.3, water quality weight 0.4, ecological weight 0.1, and meteorological weight 0.2. Retrieve the historical optimal initial positions under each weight dimension: flow optimal position 730mm, water quality optimal position 740mm, ecological optimal position 738mm, and meteorological optimal position 735mm.

[0079] ; Final initial position:

[0080] Control the electric push rod to move to the position of 735.8mm.

[0081] Preferably, the multi-source data perception and preprocessing module comprises: The spatial correlation and time dependence of the environmental parameters are captured by parallel spatio-temporal branches, and the TCN features and LSTM sequences of the current environmental parameters are extracted and modeled. The correlation between the TCN features and the LSTM sequences is calculated by mutual information entropy, and a joint feature is constructed. Meanwhile, the TCN features and the LSTM sequences are processed by residual connection. An attention weight network is introduced to dynamically adjust the weights of each feature layer based on historical scene data, and a standardized feature vector at the current time is obtained as the input of the dynamic prediction model.

[0082] Preferably, the TCN spatial feature extraction branch specifically includes: Network structure: 2 layers of residual TCN, each layer containing a combination of causal convolution, BatchNorm and ReLU, with a convolution kernel size of 3x1 (3 spatial dimensions and 1 time step), and output channel numbers of 32 and 64 respectively. Input adaptation: The aligned data is reshaped into a [batch_size, 3, 20, 10] tensor according to the region dimension (3), time dimension (20), and parameter dimension (10: flow, COD, ammonia nitrogen, TP, rainfall, wind speed, temperature, vegetation coverage, soil moisture content, and aquatic organism population density). Spatial correlation capture: The first layer of TCN extracts the correlation of multiple parameters within a single region (e.g., the concentration correlation of COD and TP in the core purification zone), and the second layer of TCN extracts the cross-region parameter transmission features (e.g., the lag correlation of flow in the inlet zone and TP in the outlet zone), finally outputting a [batch_size, 3, 20, 64] TCN spatial feature map.

[0083] Preferably, the LSTM time sequence modeling branch: Network structure: 2 layers of bidirectional LSTM, with 128 hidden layer neurons in each layer and a Dropout probability of 0.2 (to suppress overfitting), and using the Adam optimizer (initial learning rate 0.001). Input adaptation: The aligned data is reshaped into a [batch_size, 20, 10] time sequence tensor according to the time dimension (20) and the parameter dimension (10). Time dependence capture: The forward LSTM captures future time trends (e.g., the rising trend of flow from 9:00 to 9:30), and the backward LSTM captures historical time dependence (e.g., the correlation between flow at 9:30 and water quality at 9:00), and the time dimension is compressed by global average pooling, outputting a [batch_size, 256] LSTM time sequence feature vector (256=2x128, concatenating the bidirectional output).

[0084] The joint probability distribution is used to calculate the correlation between the parameters of the TCN spatial features and the LSTM time sequence features, and the specific steps are as follows: The TCN spatial feature map is compressed to [batch_size, 20, 64] through 1x1 convolution, and the LSTM time sequence feature vector (expanded to [batch_size, 20, 64] through full connection layer) is dimensionally aligned. For each pair of feature dimensions (a total of 64 pairs), the kernel density estimation method is used to calculate the joint probability distribution P(X, Y) and the marginal probability distribution P(X), P(Y) (X is the TCN feature, Y is the LSTM feature). Mutual information entropy calculation, according to the formula: ; Calculate the correlation, filter the strong correlation feature pairs with I(X, Y)>0.5, and get 28 pairs, such as the correlation between the COD feature of the core purification area of TCN and the flow time sequence trend feature of LSTM is 0.68.

[0085] First, the TCN features and LSTM features are concatenated according to the channel dimension, and then compressed to [batch_size, 20, 64] through 3x3 convolution kernel to obtain the joint feature map.

[0086] Example: After the COD spatial feature of TCN and the flow time sequence feature of LSTM are concatenated and processed by convolution, the output joint feature is 0.35, and the final joint feature fusion space correlation and time dependence.

[0087] After each layer of TCN convolution, the output feature [batch_size, 3, 20, 64] is adjusted to be consistent with the input feature [batch_size, 3, 20, 10] in dimension through 1x1 convolution, and the residual addition F(x) is performed according to the formula F(x)+x, F(x) is the convolution output, x is the input, which alleviates the gradient disappearance of deep network.

[0088] The LSTM output feature [batch_size, 256] is adjusted to [batch_size, 20, 10] through full connection layer, and is added to the input time sequence tensor to preserve the original time sequence information.

[0089] The joint feature map [batch_size, 20, 64] constructed is added to the concatenation feature of TCN spatial feature+LSTM time sequence feature to strengthen the feature expression ability.

[0090] The wetland is adopted in the past 3 hydrological period characteristic layer weight-regulation effect correlation data, 8000 samples, each sample contains TCN / LSTM / combined feature history weight corresponding scene water level regulation error label, and through 2 layers of full connection attention network, the attention network includes input layer 64 dimensions, hidden layer 32 dimensions, output layer 3 dimensions, and the output is the weight coefficient of TCN spatial feature layer w1, LSTM time sequence feature layer w2 and combined feature layer w3, that is: w1+w2+w3=1. Each dimension of the fusion feature is standardized, and the dimension of the standardized feature vector is [batch_size, 20, 64], which is completely matched with the input layer dimension of the dynamic prediction model, and can be directly transmitted into the model for space-time coding processing.

[0091] Preferably, before the TCN feature extraction and LSTM sequence modeling of the current environmental parameters, it further includes: In the current environmental parameters, a distance matrix of low-frequency parameter time series and benchmark parameter time series is constructed, and the alignment path with the minimum cumulative distance is solved by dynamic programming; Along the alignment path, the low-frequency parameters in the current environmental parameters are interpolated into time series consistent with the high-frequency parameters; The standard deviation of each type of parameter is calculated for the aligned time series data, and a Gaussian noise sequence is generated, which is superimposed with the original time series data to obtain enhanced time series data; The noise-enhanced time series data is arranged in reverse order by time to generate a flip sub-sequence, and the environmental scene label of the flip sub-sequence is retained to expand the data set of the current environmental parameters.

[0092] Preferably, taking the 3-minute / minute flow of the outlet temporary storage well as the benchmark parameter, the distance matrix is constructed for the ecological parameters 8 minutes / minute: assuming that the ecological parameter sequence X=(88, 88.2) (9:00, 9:08), the benchmark parameter sequence Y=(4.2, 4.3, 4.1) (9:00, 9:03, 9:10), the Euclidean distance matrix D is calculated, the alignment path P={(1,1),(1,2),(2,3)} with the minimum cumulative distance is solved by dynamic programming, and the cubic spline interpolation is used along the path to generate the ecological parameter 88.1% at 9:03, so that the ecological parameter and the benchmark parameter maintain the uniform frequency of 3 minutes / minute.

[0093] The standard deviation σ of the aligned flow parameter is calculated as 0.1 m3 / h, a Gaussian noise sequence N=(0.008, -0.005, 0.009) (intensity ≤10%σ) is generated, and the enhanced flow data (4.208, 4.295, 4.109) is obtained after superposition; the flow subsequence from 9:00 to 9:30 is selected, a reversed subsequence is generated in reverse order of time, and a stable water inflow scene label is marked, so that the data set is expanded to 1.8 times of the original.

[0094] Preferably, the dynamic prediction model comprises: The residual TCN-bidirectional LSTM-attention fusion layer backbone network is constructed, and the MSE completion loss and the cross-entropy weight alignment loss are used as the joint loss function for network training. The standardized feature vector is received, reshaped according to the region-time sequence dimension to match the spatial distribution of the wetland, the depth separable convolution is combined with the TCN spatial feature weight to extract the water inflow area, the core purification area and the water outflow area parameter conduction feature, the causal convolution is combined with the LSTM time sequence feature weight to capture the short-term time sequence trend, and the space-time basic coding feature is output; The water quality target vector and the water quantity target vector are constructed, the water quality target vector and the water quantity target vector are weighted according to the attention weight, and are spliced with the space-time basic coding feature, and the target related space-time coding feature is obtained through convolution compression; The spatial separable convolution and the Sigmoid gate are used, the TCN feature correlation degree of the preprocessing module is combined to filter the key area and output the pure space feature; The time separable convolution and the Tanh gate are used, the LSTM sequence trend of the preprocessing module is combined to capture the time sequence mutation and output the pure time feature; The difference between the pure time features is calculated, a difference sequence is generated, and the parameter change rate is captured; The mutual information entropy of the difference feature and the prediction target is calculated, and the self-attention weight is generated; The weighted difference feature and the pure time feature are spliced and input into the 2-layer bidirectional LSTM, and the enhanced time feature is output; The pure space feature is expanded through the full connection layer and spliced with the enhanced time feature, and the space-time enhanced feature is obtained; The mutual information correlation degree of the space-time enhanced feature and the water quality index is calculated, the step attention weight is generated, the weighted pooling is performed, and the water quality prediction value is output after the 3-layer full connection layer LeakyReLU activation; The space-time enhanced feature is input into the 2-layer bidirectional LSTM decoding layer for decoding, the decoding output is added after matching the input feature dimension, and the water quantity prediction value is output after the full connection layer processing.

[0095] Preferably, the prediction result based on the dynamic prediction model is used for optimal water level prediction, and an optimal water level control strategy is output, comprising: According to the water quality prediction value, the water quantity prediction value and the confidence, a two-dimensional data table is constructed, low confidence prediction points are removed according to a confidence threshold, multi-time granularity data is combined according to a control response priority, and short-term prediction data set and medium-long term prediction data set are obtained; For the short-term prediction data set, whether the water quality at each time point meets the standard is judged according to the water quality prediction value; For the medium-long term prediction data set, whether the water quantity threshold is met is judged according to the water quantity prediction value; According to the water quality prediction value, or the water quantity prediction value, or a fixed period, the current required water level adjustment range is determined, and the extension and retraction amount of the electric push rod is determined in combination with the electric push rod energy consumption characteristic curve.

[0096] In the embodiment, based on the above adaptive water level adjustment method, the conventional operation mode is: (1) When the constructed wetland is operated according to the design flow and the design load, the solar power supply system, the Doppler flow and the multifunctional water quality monitoring probe, the meteorological environment acquisition system and the ecological parameter acquisition system are started to start normal monitoring of the system.

[0097] (2) The electric push rod needs to be pre-adjusted to the design outlet water level height. After that, the sewage is purified by the constructed wetland and then enters the outlet water temporary storage tank.

[0098] (3) At this time, if the multifunctional water quality monitoring probe detects that the outlet water quality does not meet the standard, the platform will issue an instruction to the electric push rod to adjust the outlet pipe height to the highest water level, so as to fully play the roles of plants in absorbing pollutants, fillers in the wetland in adsorbing pollutants and microorganisms in degrading pollutants, and realize the standard discharge of sewage.

[0099] (4) If it still does not meet the standard, repeat the above operation until the highest water level line.

[0100] (5) When the wetland is normally operated for 7-30 days, the electric push rod can be used to lower it to the lowest water level at one time, so as to flush the fillers in the wetland in a short time, reduce the blockage of the fillers in the wetland, and improve the overall removal effect of the wetland on pollutants.

[0101] In the embodiment, based on the above adaptive water level adjustment method, the overload operation mode is: (1) When the water inflow of the constructed wetland exceeds the design parameter 1.0-1.2 times or the water load exceeds the design initial parameter 1.0-1.5 times, the emergency response should be started immediately to slow down the water inflow, and at the same time, the action distance of the electric push rod can be set by time period, so that it operates between the lowest water level and the highest water level.

[0102] (2) When the water level is higher than the filler, the plants, filler and microorganisms in the wetland are in an aerobic stage; when the water level is lower than the filler, it is in an anoxic stage. The removal effect of the wetland on pollutants is improved by continuously alternating between aerobic and anoxic stages.

[0103] In addition, the height of the telescopic plastic hose can also be increased by using an electric push rod to prolong the residence time of sewage in the constructed wetland, thereby achieving the discharge of sewage up to the standard. When the effluent quality does not meet the standard, the treatment method is the same as the above (1).

[0104] In this embodiment, based on the above-mentioned adaptive water level adjustment method, the low load operation mode is: When the inflow of the constructed wetland is less than 0.2-0.8 times the design parameter or the inflow load is less than 0.2-1.0 times the initial design parameter, the inflow is increased. Or, while maintaining the above-mentioned flow, the electric push rod is lowered from the design water level until the effluent just meets the standard.

[0105] In this embodiment, based on the above-mentioned adaptive water level adjustment method, the rainstorm operation mode is: (1) When the meteorological environment collection system detects that it will rain heavily in advance, an instruction is sent in advance through the platform before the rainstorm arrives, and the electric push rod is extended to a certain height from the design effluent water level height to ensure that the effluent quality meets the standard. The volume corresponding to this height is the corresponding storage capacity during the rainstorm period.

[0106] (2) During the rainstorm, the electric push rod will be retracted to a certain height to cope with the impact flow and load generated when the rainstorm arrives. At the same time, it ensures that the effluent quality meets the standard.

[0107] (3) After the rain, it gradually recovers to the design water level. At this time, the sediment flushing program is started, and the electric push rod is lowered to the lowest water level at one time to flush the filler in the wetland by lowering the water level in a short time, thereby reducing the problem of blockage caused by sudden increase of flow in the rainy season.

[0108] In this embodiment, based on the above-mentioned adaptive water level adjustment method, the ice freezing operation mode is: When the meteorological environment collection system detects that the weather changes suddenly and is in the ice freezing mode in advance, the electric push rod is moved from the design effluent water level height to the highest water level to ensure that the effluent quality meets the standard. When the weather returns to normal, it gradually recovers to the design water level.

[0109] In this embodiment, based on the above-mentioned adaptive water level adjustment method, the inspection mode is: The wetland system will start the inspection mode every 30-60 days. The inspection mode mainly monitors the growth of plants and the height of the normal water level in the wetland.

[0110] When the camera in the ecological parameter acquisition system monitors that the plants grow slowly or do not grow, the platform retrieves the water inflow and outflow and water quality, if it is found that the outflow water quality meets the standard, but the water inflow or the water inflow concentration is far lower than the initial design value, the water inflow can be increased or the outflow water level can be increased to provide more nutrients for the plants to benefit the normal growth of the plants in the later period. On the contrary, if it is monitored that the plants grow too fast, the water inflow can be reduced or the outflow water level can be reduced under the condition that the outflow water quality meets the standard.

[0111] When the camera in the ecological parameter acquisition system monitors that the plants wither, the platform can issue a work order to the operation and maintenance personnel to replant the plants.

[0112] When the camera in the ecological parameter acquisition system monitors that the plants have diseases and pests, the platform can issue a work order to the operation and maintenance personnel to spray medicine to govern the plants.

[0113] When the camera in the ecological parameter acquisition system monitors that the weeds in the wetland grow in clusters and even cover the height of the plants, the platform can issue a work order to the operation and maintenance personnel to clean the weeds.

[0114] When the camera in the ecological parameter acquisition system monitors that the plants grow to the designed height, the platform can issue a work order to the operation and maintenance personnel to harvest the plants.

[0115] When the camera in the ecological parameter acquisition system monitors that the weeds in the wetland grow in clusters and even cover the height of the plants, the platform can issue a work order to the operation and maintenance personnel to clean the weeds.

[0116] Inspection of the normal water level in the wetland: The platform automatically retrieves the recorded outflow water level data during the operation of the wetland, if it is found that the wetland is in the normal operation mode, the outflow water level is above the designed water level for a long time or the highest water level, or when the camera monitors that the wetland system surface appears gushing water, it indicates that the filler in the wetland is largely blocked. At this time, the following measures can be taken: (1) The platform issues an instruction to reduce the outflow water level to the lowest water level under the condition that the outflow water quality meets the standard, and uses the water level difference to flush the filler in the wetland to alleviate the problem of blockage.

[0117] (2) When the normal operation water level in the pool cannot be reduced after multiple flushing, the platform will issue a work order to the operation and management personnel to check the site of the wetland to determine whether the filler needs to be replaced.

[0118] In embodiment two, the application further proposes a self-adaptive water level adjusting system, the system comprises: an electric push rod matched with an outflow temporary storage well, the electric push rod is connected with a control cabinet, the control cabinet is connected with a meteorological link acquisition system, an ecological parameter acquisition system, a water flow monitoring probe and a water quality monitoring probe.​

[0119] The control cabinet calculates an optimal water level control strategy based on a wetland water level intelligent regulation algorithm according to the collected environmental parameters, and regulates the electric push rod according to the optimal water level control strategy.

[0120] The environmental parameters include water flow of the water outlet temporary storage well, water quality of the water outlet temporary storage well, current location meteorological parameters, and current location ecological parameters.

[0121] The wetland water level intelligent regulation algorithm includes a multi-source data sensing and preprocessing module, a prediction model module, and a multi-objective rolling optimization decision module.

[0122] The multi-source data sensing and preprocessing module is configured to preprocess the collected environmental parameters.

[0123] The dynamic prediction model predicts water quality and water quantity in a future time period based on current environmental parameters.

[0124] The multi-objective rolling optimization decision module performs optimal water level prediction based on the prediction result of the dynamic prediction model, and outputs an optimal water level control strategy.

[0125] Preferably, the system includes that the water flow monitoring probe and the water quality monitoring probe are arranged at the water inlet pipe opening of the water outlet temporary storage well.

[0126] The electric push rod is connected to a telescopic plastic hose, and the telescopic plastic hose is connected to a water outlet. The control cabinet determines a current required water level adjustment range according to a water quality prediction value, a water quantity prediction value, or a fixed period, determines a telescopic amount of the electric push rod in combination with an electric push rod energy consumption characteristic curve, and when the electric push rod controls the telescopic plastic hose to be below the water surface, the water in the water outlet temporary storage well is discharged through the telescopic plastic hose.

[0127] Preferably, the system further includes: The control cabinet further switches different operation modes according to flow load and meteorological conditions.

[0128] In embodiment three, the application further provides a computer device including a processor and a memory. The memory stores at least one instruction, which is loaded and executed by the processor to implement the adaptive water level adjustment method in embodiment one.

[0129] Although example embodiments have been described herein with reference to the accompanying drawings, it is to be understood that the example embodiments are only exemplary and are not intended to limit the scope of the application. Those skilled in the art can be able to make various changes and modifications without departing from the scope and spirit of the application. All such changes and modifications are intended to be included within the scope of the application as defined by the appended claims.

[0130] It should be noted that, as used in this document, the term "indicates" is merely used to distinguish one entity or operation from another, and does not necessarily require or imply any such actual relationship or order between the entities or operations. Moreover, the term "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not necessarily include only those elements in the list, but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0131] Although the present application has been described in connection with the above specific embodiments, it will be readily apparent to those skilled in the art that many substitutions, modifications and changes can be made thereto without departing from the spirit and scope of the application. Accordingly, all such substitutions, modifications and changes are intended to be included within the scope of the appended claims.

Claims

1. A method of adaptive water level regulation, characterized by, The method comprises the following steps: initializing the adjusting system to determine the initial position of the electric push rod; collecting current environmental parameters; the environmental parameters include the flow of the temporary water storage well, the water quality of the temporary water storage well, the current location meteorology, and the current location ecological parameters; calculating the optimal water level control strategy based on the wetland water level intelligent control algorithm according to the environmental parameters; controlling the electric push rod according to the optimal water level control strategy; wherein the wetland water level intelligent control algorithm comprises a multi-source data sensing and preprocessing module, a prediction model module, and a multi-objective rolling optimization decision module; the multi-source data sensing and preprocessing module is used for preprocessing the collected environmental parameters; a dynamic prediction model is used for predicting the water quality and quantity in a future time period based on the current environmental parameters; the multi-objective rolling optimization decision module is used for predicting the optimal water level based on the prediction result of the dynamic prediction model and outputting the optimal water level control strategy.

2. The self-adapting water level regulating method according to claim 1, wherein, The method for determining the initial position of the electric push rod comprises the following steps: accessing historical operation data and screening historical scenes with an environmental feature similarity greater than 85% at the current initialization time; extracting the initial position data of the electric push rod in the corresponding scene from the screened similar historical scenes; calculating the mean and standard deviation of the initial position data, and determining that the mean is the initial position reference value when the standard deviation is less than a preset value; otherwise, the mean is recalculated after removing the abnormal values as the initial position reference value; dividing the flow weight coefficient, the water quality weight coefficient, the ecological weight coefficient, and the meteorology weight coefficient according to the environmental features at the current initialization time and the historical scene data, and calculating the final initial position of the electric push rod based on the initial position reference value, the flow weight coefficient, the water quality weight coefficient, the ecological weight coefficient, and the meteorology weight coefficient.

3. The method of claim 2, wherein, The multi-source data sensing and preprocessing module comprises: performing TCN feature extraction and LSTM sequence modeling on the current environmental parameters by capturing the spatial correlation and time dependence of the environmental parameters through parallel space-time branches; calculating the correlation between the parameters of the TCN features and the LSTM sequences through mutual information entropy, constructing joint features, and simultaneously performing residual connection processing on the TCN features and the LSTM sequences; introducing an attention weight network, dynamically adjusting the weight of each feature layer according to the historical scene data, obtaining a standardized feature vector at the current time as the input of the dynamic prediction model.

4. The method of claim 3, wherein, Before performing TCN feature extraction and LSTM sequence modeling on the current environmental parameters, the following steps are further included: constructing a distance matrix of the low-frequency parameter time series and the reference parameter time series in the current environmental parameters, and solving the alignment path with the minimum cumulative distance through dynamic programming; interpolating the low-frequency parameters in the current environmental parameters along the alignment path into time series consistent with the high-frequency parameters; calculating the standard deviation of each type of parameter for the aligned time series data, generating a Gaussian noise sequence, and superimposing the Gaussian noise sequence and the original time series data to obtain enhanced time series data; arranging the noise-enhanced time series data in reverse order to generate a flip sub-sequence, retaining the environmental scene label of the flip sub-sequence, and expanding the data set of the current environmental parameters.

5. The method of claim 4, wherein, The dynamic prediction model comprises: The residual TCN-bidirectional LSTM-attention fusion layer backbone network is constructed, and the network is trained by taking the MSE completion loss and the cross-entropy weight alignment loss as the joint loss function. The standardized feature vector is received, reshaped according to the area-time sequence dimension to match the spatial distribution of the wetland, and the deep separable convolution is combined with the TCN spatial feature weight to extract the parameter conduction characteristics of the inlet area, the core purification area and the outlet area; the causal convolution is combined with the LSTM time sequence feature weight to capture the short-term time sequence trend, and the spatio-temporal basic encoding feature is output; The water quality target vector and the water quantity target vector are constructed, the water quality target vector and the water quantity target vector are weighted according to the attention weight, and are spliced with the spatio-temporal basic encoding feature, and the target related spatio-temporal encoding feature is obtained through convolution compression; The spatial separable convolution and the Sigmoid gate are used to combine the TCN feature correlation degree of the preprocessing module to filter the key area and output the pure space feature; The time separable convolution and the Tanh gate are used to combine the LSTM sequence trend of the preprocessing module to capture the time sequence mutation and output the pure time feature; The difference sequence is generated by calculating the difference between the pure time features, and the parameter change rate is captured; The mutual information entropy of the difference feature and the prediction target is calculated to generate the self-attention weight; The weighted difference feature and the pure time feature are spliced and input into the 2-layer bidirectional LSTM to output the enhanced time feature; The pure space feature is expanded through the full connection layer and spliced with the enhanced time feature to obtain the spatio-temporal enhanced feature; The mutual information correlation degree between the spatio-temporal enhanced feature and the water quality index is calculated to generate the step attention weight, which is weighted and pooled through the 3-layer full connection layer LeakyReLU activation to output the water quality prediction value; The spatio-temporal enhanced feature is input into the 2-layer bidirectional LSTM decoding layer for decoding, and the decoding output is added after matching the input feature dimension, and the water quantity prediction value is output after processing through the full connection layer.

6. The method of claim 5, wherein, The prediction result based on the dynamic prediction model is used for optimal water level prediction, and an optimal water level control strategy is output, including: According to the water quality prediction value, the water quantity prediction value and the confidence, a two-dimensional data table is constructed, low confidence prediction points are removed according to the confidence threshold, multi-time granularity data is combined according to the control response priority, and short-term prediction data set and medium-long term prediction data set are obtained; For the short-term prediction data set, whether the water quality at each time point meets the standard is judged according to the water quality prediction value; For the medium-long term prediction data set, whether the water quantity threshold is reached is judged according to the water quantity prediction value; According to the water quality prediction value, the water quantity prediction value or the fixed period, the required water level adjustment range is determined, and the extension amount of the electric push rod is determined according to the electric push rod energy consumption characteristic curve.

7. A regulating system employing an adaptive water level regulating method according to any one of claims 1 to 6, characterized in that, The system comprises an electric push rod matched with the outlet temporary storage well, an electric push rod connected control cabinet, an external meteorological link acquisition system, an ecological parameter acquisition system, a water flow monitoring probe and a water quality monitoring probe; The control cabinet calculates the optimal water level control strategy based on the wetland water level intelligent control algorithm according to the collected environmental parameters, and controls the electric push rod according to the optimal water level control strategy. The environmental parameters include water flow of the temporary water storage well, water quality of the temporary water storage well, current location meteorology, and current location ecological parameters. The wetland water level intelligent regulation algorithm comprises a multi-source data sensing and preprocessing module, a prediction model module, and a multi-objective rolling optimization decision module. The multi-source data sensing and preprocessing module is configured to preprocess the collected environmental parameters. The dynamic prediction model is configured to predict water quality and water quantity in a future time period based on current environmental parameters. The multi-objective rolling optimization decision module is configured to perform optimal water level prediction based on the prediction result of the dynamic prediction model, and output an optimal water level control strategy.

8. The regulation system of claim 7, wherein The water flow monitoring probe and the water quality monitoring probe are arranged at a water inlet pipe opening of the temporary water storage well. The electric push rod is connected to a telescopic plastic hose, the telescopic plastic hose is connected to a water outlet, a control cabinet determines a current required water level regulation range according to a water quality prediction value or a water quantity prediction value or a fixed period, determines a telescopic amount of the electric push rod in combination with an electric push rod energy consumption characteristic curve, and when the electric push rod controls the telescopic plastic hose to be below the water surface, water in the temporary water storage well is discharged through the telescopic plastic hose.

9. The conditioning system of claim 8, wherein, Further comprising: The control cabinet further switches different operation modes according to flow load and meteorological conditions.

10. A computer device, comprising: The computer device comprises a processor and a memory, the memory stores at least one instruction, the at least one instruction is loaded and executed by the processor to implement the adaptive water level regulation method of any one of claims 1 to 6.

Images