Rural domestic sewage resource utilization treatment method and system

Abstract

The invention relates to the technical field of environmental protection and sewage treatment, and discloses a rural domestic sewage resource utilization treatment method and system.The method comprises the steps that sewage quality, farmland soil moisture content and crop images are collected in real time through a multi-parameter sensing module, and crop growth stages are recognized in combination with a convolutional neural network; a water quality-fertilizer demand-irrigation dynamic matching model is built based on the multi-layer sensing mechanism, and the filter material replacement period, the wetland plant harvesting opportunity and the effluent distribution proportion are intelligently regulated and controlled. The system consists of three sections of treatment units, namely a grid sedimentation tank, a biological filter tank and an artificial wetland, is equipped with an embedded simple operation and maintenance interface and a fault self-diagnosis function, and supports dialect voice prompt and remote data uploading. By means of the technical scheme, sewage nitrogen and phosphorus resources are accurately converted into farmland fertilizer, the operation and maintenance threshold is lowered, and the system stability and the recycling efficiency are improved.

Description

Technical Field

[0001] This invention belongs to the field of environmental protection and wastewater treatment technology, specifically relating to a method and system for the resource utilization and treatment of rural domestic wastewater. Background Technology

[0002] Rural domestic sewage treatment has become a crucial link in improving the living environment and realizing ecological circular agriculture. Current mainstream treatment models largely follow the urban centralized sewage treatment approach, relying on fixed process parameters and standardized discharge standards. Their core design treats sewage as a pollution source to be eliminated, rather than a usable resource. However, sewage generation in rural areas exhibits significant spatial and temporal unevenness: water volume fluctuates dramatically due to seasonal and lifestyle factors, and water quality is complex due to mixed sources such as washing, kitchen waste, and livestock farming. Simultaneously, the nitrogen and phosphorus nutrient requirements of surrounding farmland dynamically change with crop type and growth stage.

[0003] Traditional systems often lack the ability to perceive in real time the quality of influent water, soil moisture, and crop nutrient requirements. This often leads to overload failure of treatment facilities during high-load periods and idle waste during low-load periods. The nitrogen and phosphorus concentrations in the effluent are seriously mismatched with the actual needs of farmland. This can result in soil salinization and groundwater pollution due to excessively high concentrations, or farmers being forced to apply additional chemical fertilizers due to insufficient nutrients, thus deviating from the original intention of resource utilization.

[0004] Constructed wetlands and ecological filters, as low-cost treatment technologies widely used in rural areas, rely heavily on the synergistic effects of plant absorption, microbial degradation, and substrate adsorption for their effectiveness. Existing solutions generally employ a fixed-cycle filter media replacement, uniformly harvested wetland plants, and homogenized irrigation water distribution, failing to establish a dynamic response mechanism between wastewater quality, crop nutrient requirements, and irrigation strategies. When influent ammonia nitrogen or total phosphorus concentrations suddenly increase, the system cannot proactively trigger filter media regeneration or adjust plant harvesting times to enhance nitrogen and phosphorus removal capabilities. During peak fertilizer demand periods such as crop jointing and heading, the failure to proactively prioritize irrigation with nutrient-rich water results in the waste of precious nitrogen and phosphorus resources or environmental risks. This fundamental contradiction between static operation and maintenance logic and the dynamic needs of agricultural production has led many projects into a predicament of achieving construction standards but operating inefficiently and wasting resources.

[0005] While existing technologies attempt to introduce automated control modules, they generally suffer from two major drawbacks: First, the sensors are deployed in isolation, monitoring only a single aspect (such as influent COD or effluent pH), without constructing a closed-loop data chain covering the entire chain from wastewater to treatment to farmland; second, the control algorithms are detached from agricultural scenarios, simplifying complex crop physiological needs into fixed threshold comparisons, and failing to intelligently match the nutrient supply rhythm according to key growth stages such as rice tillering and fruit tree fruit expansion.

[0006] More critically, the system's interface often uses industrial-grade configuration software, requiring operators to possess professional environmental engineering knowledge. However, the actual maintenance personnel in rural areas are mostly ordinary farmers with limited education, often helpless when faced with frequent alarms and complex operations such as parameter adjustments, ultimately leading to equipment idle rates exceeding 40%. Therefore, there is an urgent need for a rural wastewater resource-based intelligent treatment system that deeply integrates multi-source sensing, agronomic models, and minimalist interaction to fundamentally overcome the three major technical bottlenecks of poor adaptability to water quality fluctuations, low resource matching accuracy, and weak operational sustainability. Summary of the Invention

[0007] This invention provides a method and system for the resource utilization and treatment of rural domestic sewage. By constructing a closed-loop control architecture integrating a multi-parameter sensing module, an intelligent control algorithm, and a simple operation and maintenance interface, it achieves dynamic response to fluctuations in the quality and quantity of rural domestic sewage, precise matching with the fertilizer and water requirements of farmland crops, and adaptability support for the operation and maintenance capabilities of non-professional farmers. Thus, without the need for external professional intervention, it can stably maintain sewage treatment efficiency, maximize the efficiency of nitrogen and phosphorus resource utilization, and ensure the long-term sustainable operation of the system.

[0008] The multi-parameter sensing module of this invention is installed at the inlet and outlet of the wastewater treatment system and in the target irrigated farmland area. It includes a water quality sensor array, a flow meter, a soil moisture sensor, and a crop growth image acquisition unit. The water quality sensor array is used to collect real-time values ​​of wastewater chemical oxygen demand, ammonia nitrogen concentration, total phosphorus concentration, and pH. The flow meter measures the amount of wastewater passing through per unit time. The soil moisture sensor is buried in the topsoil layer of the farmland to continuously monitor soil volumetric moisture content. The crop growth image acquisition unit consists of a visible light camera mounted on a fixed support in the field. It takes images once a day at a fixed time of 10:00 AM, and the acquired images are used for subsequent crop growth stage identification.

[0009] The crop growth stage identification described in this invention is based on a pre-trained convolutional neural network model. The input of this model is a sequence of farmland images over seven consecutive days, and the output is a label indicating the current growth stage of the crop. The labels include six categories: seedling stage, tillering stage, jointing stage, heading stage, grain filling stage, and maturity stage. Before deployment, the model has completed transfer learning training for five major crops: rice, leafy vegetables, solanaceous vegetables, citrus fruits, and stone fruits.

[0010] The intelligent control algorithm described in this invention constructs a dynamic matching model among wastewater quality, crop nutrient requirements, and irrigation water volume. This model uses the effluent quality parameters of the wastewater treatment system, the current crop type and growth stage of the target farmland, and the real-time soil moisture content as input variables. The output control variables are the pretreatment filter media replacement cycle, the artificial wetland plant harvesting trigger conditions, and the effluent irrigation allocation ratio. The core of the dynamic matching model is a multilayer perceptron structure with three hidden layers. Each layer has 128, 64, and 32 neurons, respectively. The activation function is a uniformly modified linear unit function, and the loss function is a mean squared error function. The training data comes from a set of correlation samples between water quality parameters, crop growth stages, and final fertilization effects in historical operation records.

[0011] The adjustment mechanism for the pretreatment filter media replacement cycle of this invention is as follows: When the chemical oxygen demand at the inlet exceeds 300 mg / L for three consecutive days, or the ammonia nitrogen concentration exceeds 25 mg / L for three consecutive days, the system automatically shortens the current filter media replacement cycle to half of the original cycle; when the above parameters are below the threshold for seven consecutive days, the system restores the original replacement cycle. The filter media is a mixture of porous ceramsite and zeolite, with a mixing ratio of 7:3 by mass and a filling height of 1.2 m. The replacement operation is performed manually by the farmer, and the system flashes a reminder on the controller interface when to replace the filter media.

[0012] The mechanism for adjusting the harvesting time of artificial wetland plants described in this invention is as follows: When the target crop enters its peak nutrient demand period, and the total phosphorus concentration in the effluent from the wastewater treatment system is higher than 80% of the upper limit of the recommended fertilization concentration for the current stage of the crop, the system triggers a harvesting command. The command includes the harvesting area number, the recommended harvesting area ratio, and the recommended harvesting tool type. The artificial wetland plants are a mixed community of reeds and cattails. The harvesting area ratio is dynamically calculated based on the crop's nutrient deficiency. The calculation formula is: the harvesting area ratio equals the crop's nutrient deficiency divided by the total nitrogen and phosphorus reserves in the wetland plants per unit area. The calculation result is rounded to one decimal place.

[0013] The water distribution ratio adjustment mechanism described in this invention is as follows: The system calculates the comprehensive water and fertilizer requirement index for each target farmland based on the crop type, current growth stage, and soil moisture content. This index is obtained by weighted summation of the crop stage water requirement coefficient, stage fertilizer requirement coefficient, and soil water shortage coefficient, with weighting coefficients of 0.4, 0.4, and 0.2, respectively. The system distributes all treated water for the day according to the proportion of the comprehensive water and fertilizer requirement index of each farmland to the total. The distribution command is executed through an electric valve assembly, and the valve opening is adjusted by a proportional-integral-derivative controller with a control cycle of 10 minutes.

[0014] The simplified operation and maintenance interface described in this invention consists of an embedded graphical controller with a seven-inch screen and a resolution of 800×480 pixels. The interface layout uses a three-row, four-column icon matrix. The first row of icons corresponds to crop type selection, including rice, leafy vegetables, solanaceous vegetables, citrus, and stone fruits. The second row of icons corresponds to system status viewing, including water quality trend graphs, soil moisture graphs, filter media lifespan bars, and plant harvest countdowns. The third row of icons corresponds to fault handling guidance, including sensor calibration, valve reset, power restart, and contacting after-sales service. After farmers select the crop type via physical buttons, the system automatically loads the corresponding crop's control parameter set. The parameter set includes a table of water and fertilizer requirement thresholds for each growth stage of the crop, a recommended harvest time window table, and a filter media replacement baseline cycle table.

[0015] The fault alarm function described in this invention includes two categories: sensor failure detection and effluent water quality exceeding standards warning. Sensor failure detection is achieved through data consistency verification between adjacent sensors. When the reading of a certain sensor deviates from the average value of other sensors in the same area by more than 30% for more than one hour, the system determines that the sensor has failed, and the interface displays that sensor number X is abnormal, suggesting cleaning the probe or replacing the battery. Effluent water quality exceeding standards warning is achieved by comparing real-time monitoring data at the water outlet with the maximum allowable fertilizer concentration for the current stage of farmland. When any water quality parameter exceeds the threshold, the system immediately closes the corresponding water outlet valve, the interface displays that nitrogen and phosphorus levels in the water exceed the standards, irrigation is suspended, wetland plant density is checked, and a backup water storage tank is activated to temporarily store the excessive effluent.

[0016] The wastewater treatment system of this invention comprises a three-stage treatment unit: a bar screen sedimentation tank, a biological filter, and an constructed wetland. The bar screen sedimentation tank is located downstream of the inlet, with a bar spacing of 10mm and a hydraulic retention time of four hours. The biological filter is filled with a porous ceramsite and zeolite mixture as filter media, with a particle size range of 3mm to 8mm and a filter layer thickness of 1.5m. Perforated aeration pipes are installed at the bottom, and the aeration rate is dynamically adjusted according to the influent ammonia nitrogen load. The constructed wetland has a horizontal subsurface flow structure with a wetland bed length-to-width ratio of 3:1. The filling substrate is a mixture of gravel and coarse sand, and the surface is planted with reeds and cattails. The wetland hydraulic load is 0.1m. 3 / m 2 ·d.

[0017] The system described in this invention performs data backup and model fine-tuning operations at 2:00 AM every day. The backup content includes all raw sensor data, control command execution records, and farmer operation logs for the day. The model fine-tuning adopts an online learning mechanism, using the deviation between the actual crop growth response and the predicted fertilization effect as a feedback signal to adjust the weight parameters of the crop fertilizer requirement calculation module in the dynamic matching model. The learning rate is set to 0.001, and the number of iterations is 50. The fine-tuning process does not affect the normal operation of the system.

[0018] The farmland irrigation distribution execution mechanism of this invention consists of an electric butterfly valve assembly, a pressure-compensating drip irrigation tape, and a flow metering module. The electric butterfly valve is installed at the beginning of each branch pipe, with a driving voltage of 24V DC and a response time of less than 5 seconds. The pressure-compensating drip irrigation tape is laid between crop rows, with a dripper spacing of 0.3m and a rated working pressure of 0.1MPa. The flow metering module uses an electromagnetic flowmeter with a measurement accuracy of ±1% and a data sampling frequency of once per minute. The measurement results are used to verify the accuracy of the distribution ratio execution.

[0019] The crop growth image acquisition unit of the present invention is equipped with an automatic cleaning mechanism, which consists of a miniature air pump, a jet nozzle, and a wiper blade. Before image acquisition each day, a cleaning procedure is performed: first, the air pump is started to spray compressed air onto the lens surface for 3 seconds to remove dust; then, the wiper blade moves back and forth twice in the horizontal direction to remove water stains and insect traces; after the cleaning procedure is completed, image acquisition is started again after a 10-second delay to ensure that the image quality meets the recognition requirements.

[0020] The system described in this invention is equipped with a remote data upload interface. The interface protocol adopts General Packet Radio Service (GPRS) technology. Every day at 8:00 AM, the system automatically uploads compressed and encrypted data from the previous day to the cloud server. The uploaded content includes average water quality monitoring data, filter media replacement records, plant harvesting status, number of fault alarms, and irrigation water allocation statistics. The cloud server deploys a data analysis module to generate monthly operation reports. These reports include a system processing efficiency trend chart, a total resource utilization statistics table, and a common fault type distribution chart. The reports are pushed to designated farmers' mobile phones via SMS.

[0021] The aeration rate adjustment mechanism of the biological filter described in this invention is as follows: the initial aeration rate is determined by referring to a table based on the influent ammonia nitrogen concentration, with a table interval of 5 mg / L; the system detects the dissolved oxygen concentration of the filter effluent every 30 minutes; if the value is below 2 mg / L for three consecutive tests, the aeration rate is increased by 20% based on the current aeration rate; if the value is above 5 mg / L for three consecutive tests, the aeration rate is decreased by 15%; the adjustment range does not exceed 25% in a single instance, and the cumulative number of adjustments per day does not exceed eight times, to prevent the biofilm from falling off due to severe disturbance.

[0022] The harvesting of artificial wetland plants described in this invention is manually completed by farmers according to system prompts. After the system triggers the harvesting command, a schematic diagram of the harvesting area is simultaneously displayed on the controller interface, indicating the suggested harvesting boundary line, recommended travel direction arrows, and safety precaution icons. A backpack brush cutter is recommended as the harvesting tool, with the blade speed set to 3,000 revolutions per minute and the harvesting height 20cm above the wetland surface. The harvested material is piled up at a designated collection point, and the system records the weight of the harvested material for subsequent nitrogen and phosphorus release calculations.

[0023] The simplified operation and maintenance interface described in this invention features a voice broadcast function. When the system detects abnormal critical parameters or requires farmers to perform operations, in addition to the screen display, it simultaneously plays pre-recorded voice prompts. The voice content includes operation step numbers, a tool preparation list, estimated time consumption, and security protection requirements. The voice speed is set to 120 words per minute, the volume can be manually adjusted, and the broadcast language is the local dialect. The system has a built-in voice library of five major dialects, and farmers can select the matching dialect type through the menu when using it for the first time.

[0024] The system described in this invention is equipped with a backup power module, which consists of a lead-acid battery pack and a solar charge controller. The battery pack has a rated capacity of 200A and a nominal voltage of 12V. The solar charge controller has a maximum input power of 300W and features overcharge protection, over-discharge protection, and short-circuit protection. When the mains power is interrupted, the system automatically switches to the backup power supply, prioritizing power to the sensor array, controller, and electric valves. The aeration equipment and image acquisition unit enter a low-power standby mode, waking up every two hours to check the mains power recovery status.

[0025] The soil moisture sensor described in this invention employs the frequency domain reflectance principle, has a probe length of 30cm, a measurement range of 5% to 50%, and a measurement accuracy of ±2%. The sensor automatically performs a self-calibration procedure every six hours. The calibration process is completed using a built-in reference capacitor. Data upload is paused during calibration. After calibration, calibration coefficients are generated and applied to correct subsequent measurements. If calibration fails, the system records an error code and displays a message indicating that the soil sensor needs to be manually reset.

[0026] The crop fertilizer requirement calculation module in the dynamic matching model of this invention employs piecewise linear interpolation, constructing a two-dimensional lookup table with crop growth stage as the horizontal axis and daily nitrogen and phosphorus requirements per unit area as the vertical axis. The lookup table data originates from crop fertilization guidelines published by agricultural research institutions and is entered into the system after field verification. When the crop is between two standard growth stages, the system interpolates the current fertilizer requirement according to the time ratio. The interpolation formula is: current fertilizer requirement equals the fertilizer requirement of the previous stage plus the increment of fertilizer requirement between stages multiplied by the time proportion within the stage.

[0027] The system described in this invention performs an initial configuration process upon first installation. The process includes: farmers selecting the climate type of their region through the controller, and the system loading the corresponding crop growth cycle baseline parameters; farmers inputting the area, soil texture type, and types of crops planted annually for each irrigated farmland; the system generating an initial control parameter set based on the input information and prompting farmers to install soil sensors and image acquisition units in each farmland; after installation, performing sensor location registration and number binding operations, with the binding information stored in non-volatile memory.

[0028] Compared with the prior art, the advantages and positive effects of the present invention are as follows: 1. This invention achieves synchronous real-time monitoring of sewage quality, farmland soil moisture, and crop growth through a multi-parameter sensing module, solving the problem of lagging regulation caused by the lack of dynamic sensing capabilities in traditional rural sewage treatment systems; 2. By constructing a dynamic matching model between wastewater quality and crop fertilizer requirements, the precise conversion of nitrogen and phosphorus resources from pollutants to fertilizers was achieved, avoiding resource waste and secondary pollution; 3. Through a simple operation and maintenance interface and fault self-diagnosis function, the complexity of system operation is reduced to a level that ordinary farmers can complete independently, solving the problem of system idleness caused by the lack of professional operation and maintenance forces in rural areas; 4. By adaptively adjusting the filter media replacement cycle and optimizing the timing of wetland plant harvesting, the service life of the core treatment unit is extended and the long-term operating cost is reduced. 5. Through precise water distribution via electric valve assembly and drip irrigation tape, irrigation water is allocated on demand, improving water resource utilization efficiency; 6. The entire system can operate stably for more than five years without the need for external professional support, with an average annual fertilizer substitution of 80 kg per mu and a sewage treatment compliance rate of over 98%. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the overall technical solution architecture of a method and system for the resource utilization and treatment of rural domestic sewage proposed in this invention; Figure 2 This is a schematic diagram of the core principle framework of the dynamic matching model of wastewater quality, crop fertilizer requirements, and irrigation water volume in this invention; Figure 3 This is a logical flow diagram of the collaborative closed-loop control of the multi-parameter sensing module and the intelligent control algorithm in this invention. Figure 4 This is a schematic diagram of the multi-level interaction relationship and data flow between the terminal farmer operation and maintenance interface and the cloud data service in this invention; Figure 5 This is a resource-based linkage control framework diagram of the sewage treatment unit and the farmland irrigation actuator in this invention; Detailed Implementation

[0030] Please refer to Figure 1-5This invention provides a method and system for the resource utilization and treatment of rural domestic sewage. By constructing a closed-loop control architecture integrating a multi-parameter sensing module, an intelligent control algorithm, and a simple operation and maintenance interface, it achieves dynamic response to fluctuations in the quality and quantity of rural domestic sewage, precise matching with the fertilizer and water requirements of farmland crops, and adaptability support for the operation and maintenance capabilities of non-professional farmers. Thus, without the need for external professional intervention, it can stably maintain sewage treatment efficiency, maximize nitrogen and phosphorus resource utilization efficiency, and ensure the long-term sustainable operation of the system.

[0031] The method includes the following steps: S1, through multi-parameter sensing modules set at the inlet and outlet of the sewage treatment system and the target irrigated farmland area, collects real-time data on sewage chemical oxygen demand, ammonia nitrogen concentration, total phosphorus concentration, pH, flow rate, soil volumetric moisture content and crop growth stage images. S2, based on a pre-trained convolutional neural network model, identifies a seven-day sequence of farmland images and outputs a label indicating the current growth stage of the crop; S3 inputs wastewater quality parameters, crop type, growth stage, and soil moisture content into the dynamic matching model to calculate the pretreatment filter media replacement cycle, artificial wetland plant harvesting trigger conditions, and effluent irrigation allocation ratio. S4, based on the calculation results, drives the actuator to complete the filter material replacement prompt, plant harvesting command issuance, and electric valve opening adjustment; S5 provides farmers with three types of operation interfaces through an embedded graphical controller: crop selection, status viewing, and fault handling. It also triggers voice broadcasts and screen alerts when an anomaly occurs. S6 performs data backup and model fine-tuning at 2 AM every day, and synchronously uploads the running data to the cloud server to generate a monthly report.

[0032] In step S1, the multi-parameter sensing module consists of a water quality sensor array, a flow meter, a soil moisture sensor, and a crop growth image acquisition unit. The water quality sensor array is deployed at the inlet and outlet, respectively, to collect the chemical oxygen demand, ammonia nitrogen concentration, total phosphorus concentration, and pH values ​​of the wastewater. The sampling frequency is once every 10 minutes, and the data is stored in the local cache after analog-to-digital conversion.

[0033] The flow meter is installed on the main water supply pipeline and uses the electromagnetic induction principle to measure the amount of sewage passing through per unit time. The measurement accuracy is ±1%, and the data is uploaded to the central processing unit after being marked with a timestamp.

[0034] The soil moisture sensor is buried 20cm below the topsoil of each target farmland. It uses the frequency domain reflectance principle to measure the volumetric water content of the soil. The probe is 30cm long and the measurement range is 5% to 50%. It automatically performs a self-calibration program every six hours. If the calibration fails, it records the error code and triggers the interface prompt.

[0035] The crop growth image acquisition unit consists of a visible light camera on a fixed bracket. The shooting frequency is once a day, and the shooting time is fixed at 10:00 am. The lens is equipped with an automatic cleaning mechanism. The cleaning procedure includes a three-second compressed air spray and two reciprocating motions of the wiper blades. After cleaning, the image acquisition is started after a 10-second delay to ensure that the image clarity meets the requirements for subsequent recognition.

[0036] In step S2, crop growth stage identification is performed based on a pre-trained convolutional neural network model. The model is input with a sequence of farmland images from seven consecutive days, each image having a resolution of at least 1280×720 pixels, which are then normalized and cropped before being input into the network. The network structure includes five convolutional layers, three max-pooling layers, and two fully connected layers. The final output layer uses a soft maximum function to generate probability distributions for six growth stages, selecting the stage with the highest probability as the current stage label. The six labels are seedling stage, tillering stage, jointing stage, heading stage, grain-filling stage, and maturity stage.

[0037] Before deployment, the model underwent transfer learning training on five major crop categories: rice, leafy vegetables, solanaceous vegetables, citrus fruits, and stone fruits. The training samples were sourced from historical image databases annotated by agricultural research institutions. The recognition results, bound to the current date and crop type, were stored in non-volatile memory and used as input parameters for subsequent dynamic matching models.

[0038] In step S3, the dynamic matching model uses the effluent quality parameters of the wastewater treatment system, the current crop type and growth stage of the target farmland, and the real-time soil moisture content as input variables, and the pretreatment filter media replacement cycle, the artificial wetland plant harvesting trigger conditions, and the effluent irrigation allocation ratio as output control variables. The core of the model is a three-layer multilayer perceptron structure, with 128, 64, and 32 neurons in the hidden layers, respectively. The activation function is uniformly a modified linear unit function, and the loss function is a mean squared error function.

[0039] The training data comes from a dataset of correlations between water quality parameters, crop growth stages, and final fertilization effects recorded in historical operation data. The dataset contains over 50,000 samples, covering operational scenarios under different seasons, climates, and soil conditions. The model input layer receives standardized parameter vectors, including normalized values ​​for chemical oxygen demand (COD), ammonia nitrogen concentration, total phosphorus concentration, pH, flow rate, soil moisture content, crop type, and growth stage. The output layer generates three control command vectors, corresponding to the filter media replacement cycle scaling factor, the proportion of harvested plant area, and the irrigation water allocation weight for each field.

[0040] The adjustment mechanism for the pretreatment filter media replacement cycle is as follows: The system continuously monitors the chemical oxygen demand (COD) and ammonia nitrogen concentration at the inlet. When the COD exceeds 300 mg / L for three consecutive days, or the ammonia nitrogen concentration exceeds 25 mg / L for three consecutive days, the central processing unit will shorten the current filter media replacement cycle to half of the original cycle. A red warning symbol will flash in the third column of the first row on the controller interface, and a voice prompt will play: "Filter media load is too high. Please replace it within seven days."

[0041] When the above parameters are below the threshold for seven consecutive days, the system resumes the original replacement cycle, and the interface warning symbol turns solid green. The filter media is a mixture of porous ceramsite and zeolite in a mass ratio of 7:3, with a filling height of 1.2m. Replacement is performed manually by the farmer, and the system records the time of each replacement and the operator's number for subsequent lifespan statistics and accountability.

[0042] The mechanism for adjusting the harvesting time of constructed wetland plants is as follows: When the target crop enters its peak nutrient demand period, and the total phosphorus concentration in the wastewater treatment system effluent exceeds 80% of the upper limit of the recommended fertilization concentration for the current stage of the crop, the system triggers a harvesting command. The command includes the harvesting area number, the recommended harvesting area ratio, and the recommended harvesting tool type.

[0043] The harvested area ratio is dynamically calculated based on the crop nutrient requirement gap. The calculation formula is: the harvested area ratio equals the crop nutrient deficiency divided by the total nitrogen and phosphorus reserves in wetland plants per unit area. The crop nutrient deficiency is output by the crop nutrient requirement calculation module in the dynamic matching model. This module uses piecewise linear interpolation, constructing a two-dimensional lookup table with the crop growth stage as the x-axis and the daily nitrogen and phosphorus requirements per unit area as the y-axis. The lookup table data comes from crop fertilization guidelines published by agricultural research institutions and is entered into the system after field verification.

[0044] When crops are between two standard growth stages, the system interpolates the current fertilizer requirement based on time proportions. The interpolation formula is: current fertilizer requirement equals the fertilizer requirement of the previous stage plus the increase in fertilizer requirement between stages multiplied by the time proportion within the stage. The total nitrogen and phosphorus reserves in wetland plants per unit area are obtained through statistical analysis of historical harvest records, with initial values ​​set at 0.8 kg of nitrogen and 0.3 kg of phosphorus per square meter. These values ​​are then dynamically updated based on the actual harvested weight and laboratory test results. The calculated results are rounded to one decimal place. If the result is less than 5%, harvesting is not triggered; if it is greater than 80%, harvesting is carried out in batches, with each batch not exceeding 40% and intervals greater than 15 days.

[0045] The water distribution ratio adjustment mechanism is as follows: The system calculates the comprehensive water and fertilizer requirement index of each target farmland based on the crop type, current growth stage, and soil moisture content. This index is obtained by weighted summation of the crop stage water requirement coefficient, stage fertilizer requirement coefficient, and soil water shortage coefficient, with weighting coefficients of 0.4, 0.4, and 0.2, respectively.

[0046] The crop stage water requirement coefficient and stage fertilizer requirement coefficient are derived from the built-in parameter table of the dynamic matching model. The soil water shortage coefficient is calculated as follows: Soil water shortage coefficient equals the lower limit of suitable crop moisture content minus the current measured moisture content, divided by the lower limit of suitable moisture content. The system allocates all treated effluent of the day according to the proportion of the comprehensive water and fertilizer requirement index of each field to the total, and the allocation command is executed through electric valve groups.

[0047] The electric butterfly valve is installed at the beginning of each branch pipe, driven by 24V DC, with a response time of less than five seconds. The opening degree is adjusted by a proportional-integral-derivative controller, with a control cycle of 10 minutes. The flow metering module uses an electromagnetic flow meter with a measurement accuracy of ±1% and a data sampling frequency of once per minute. The metering results are used to verify the accuracy of the distribution ratio execution. If the measured flow rate deviates from the commanded flow rate by more than 10% for more than 30 minutes, the system records the deviation event and triggers the valve reset procedure.

[0048] In step S4, the actuators include a filter media replacement prompt module, a plant harvesting instruction issuing module, and an electric valve adjustment module. The filter media replacement prompt module generates a flashing icon and voice announcement on the controller interface, and simultaneously sends an SMS reminder to the farmer's mobile phone. The SMS message includes the filter media type, replacement step number, a list of required tools, and estimated time. After the triggering conditions are met, the plant harvesting instruction issuing module synchronously displays a plan view of the harvesting area on the controller interface, indicating the suggested harvesting boundary line, recommended travel direction arrows, and safety precaution icons.

[0049] The recommended harvesting tool is a backpack brush cutter with a blade speed of 3,000 revolutions per minute. The harvesting height should be 20cm above the wetland surface. Harvested material should be piled at designated collection points, and the system will record the weight for subsequent nitrogen and phosphorus release calculations. The electric valve adjustment module receives the pulse width modulation signal from the proportional-integral-derivative controller, driving the motor to rotate to a specified angle. The valve position sensor provides real-time feedback on the opening value, forming a closed-loop control. If the valve fails to reach the target opening within five seconds, the system determines it is mechanically jammed, immediately closes the corresponding branch, and triggers a valve reset fault code. The icon in the third row, second column of the interface is highlighted, and a voice announcement is given: "Valve number three is jammed. Please manually reset and restart the system."

[0050] In step S5, the simplified operation and maintenance interface is composed of an embedded graphics controller. The controller screen is seven inches in size with a resolution of 800×480 pixels, and the interface layout adopts a three-row, four-column icon matrix.

[0051] The first row of icons corresponds to crop type selection, including five categories: rice, leafy vegetables, solanaceous vegetables, citrus, and stone fruits. After farmers select the crop using physical buttons, the system automatically loads the corresponding crop's control parameter set. The parameter set includes a table of water and fertilizer requirements for each growth stage of the crop, a recommended harvest time window table, and a filter media replacement baseline cycle table.

[0052] The second row of icons corresponds to the system status view, including water quality trend chart, soil moisture chart, filter media life bar, and plant harvest countdown. All charts are dynamically updated in line or bar format, and the data refresh frequency is once every 30 seconds.

[0053] The third row of icons corresponds to troubleshooting instructions, including sensor calibration, valve reset, power restart, and contacting after-sales service. Each icon is associated with a pre-recorded operation video and text steps, which will play automatically when the farmer clicks on them.

[0054] The voice broadcast function is activated simultaneously when the system detects abnormal key parameters or when farmers need to perform operations. The voice content includes operation step numbers, tool preparation list, estimated time, and safety protection requirements. The speech rate is set to 120 words per minute, the volume can be manually adjusted, and the broadcast language is the local dialect version. The system has five built-in dialect voice libraries. When farmers use it for the first time, they can select the matching dialect type through the menu.

[0055] In step S6, the system performs data backup and model fine-tuning operations at 2:00 AM every day. The backup includes all raw sensor data, control command execution records, and farmer operation logs for the day. The data is compressed and encrypted and stored on a local solid-state drive for 365 days.

[0056] The model fine-tuning employs an online learning mechanism, using the deviation between the actual crop growth response and the predicted fertilization effect as feedback signals to adjust the weight parameters of the crop nutrient requirement calculation module in the dynamic matching model. The learning rate is set to 0.001, and the number of iterations per iteration is 50. The fine-tuning process does not affect the normal operation of the system. The remote data upload interface uses general packet wireless service technology, automatically uploading the previous day's running data to the cloud server at 8:00 AM every day after compression and encryption. The uploaded content includes the average water quality monitoring data, filter media replacement records, plant harvesting execution status, number of fault alarms, and irrigation water allocation statistics.

[0057] A data analysis module is deployed on the cloud server to generate monthly operation reports. The reports include a system processing efficiency trend chart, a total resource utilization statistics table, and a common fault type distribution chart. The reports are pushed to designated farmers' mobile phones via SMS at 10:00 AM on the 5th of each month.

[0058] The system comprises a three-stage treatment unit: a bar screen sedimentation tank, a biological filter, and an constructed wetland. The bar screen sedimentation tank is located downstream of the inlet, with a bar spacing of 10 mg and a hydraulic retention time of four hours. A sludge hopper is installed at the bottom of the tank, and sludge is automatically discharged once daily, with the discharge volume dynamically adjusted according to the influent suspended solids concentration. The biological filter is filled with a porous ceramsite and zeolite mixture, with a particle size ranging from 3 mm to 8 mm and a filter layer thickness of 1.5 m. Perforated aeration pipes are installed at the bottom, and the aeration rate is dynamically adjusted according to the influent ammonia nitrogen load.

[0059] The initial aeration rate is determined by referring to a table based on the influent ammonia nitrogen concentration, with a reference interval of 5 mg / L. The system checks the dissolved oxygen concentration in the filter effluent every 30 minutes. If the value is below 2 mg / L for three consecutive times, the aeration rate is increased by 20% based on the current aeration rate. If the value is above 5 mg / L for three consecutive times, the aeration rate is decreased by 15%. The adjustment range should not exceed 25% at a time, and the cumulative number of adjustments per day should not exceed eight times to prevent the biofilm from falling off due to severe disturbance.

[0060] The constructed wetland is a horizontal subsurface flow structure with a wetland bed length-to-width ratio of 3:1. The filling substrate is a mixture of gravel and coarse sand, and the surface is planted with reeds and cattails. The wetland hydraulic loading rate is 0.1m. 3 / m 2 •d. After being irradiated with ultraviolet light in the disinfection tank, the wetland effluent enters the clear water tank. The clear water tank is equipped with a level sensor. When the level is below 0.5m, a low water level alarm is triggered. When the level is above 2.5m, the overflow valve is activated to discharge the water into the backup storage tank.

[0061] The system is equipped with a backup power module, consisting of a lead-acid battery pack and a solar charge controller. The battery pack has a rated capacity of 200A and a nominal voltage of 12V. The solar charge controller has a maximum input power of 300W and features overcharge protection, over-discharge protection, and short-circuit protection. When the mains power is interrupted, the system automatically switches to the backup power supply, prioritizing power to the sensor array, controller, and electric valves. The aeration equipment and image acquisition unit enter a low-power standby mode, waking up every two hours to check the mains power recovery status. If the backup power level drops below 20%, the system shuts down non-essential modules, maintaining only basic monitoring and alarm functions, and sends a text message to the farmer's mobile phone requesting that the power supply be depleted and to check the mains power or replace the battery.

[0062] The system performs an initialization configuration process during the first installation. The process includes: Farmers select the climate type of their region through the controller, and the system loads the corresponding crop growth cycle benchmark parameters for the region. Farmers input the area, soil texture type, and types of crops that are usually planted in each irrigated farmland. The system generates an initial set of control parameters based on the input information and prompts farmers to install soil sensors and image acquisition units in each farmland. After installation, the system performs sensor location registration and number binding operations, and the binding information is stored in non-volatile memory.

[0063] The registration process includes sensor self-test, signal strength test, and data consistency verification. If any step fails, a reinstallation prompt will be displayed. After the system completes initialization, it automatically enters trial operation mode and continuously collects baseline data for 72 hours without executing any control commands. After 72 hours, an initial operation report is generated, which is then confirmed by the farmer before the system is officially put into use.

[0064] The fault alarm function includes two categories: sensor failure detection and effluent water quality exceeding standards warning. Sensor failure detection is achieved through data consistency verification between adjacent sensors. When the reading of a certain sensor deviates from the average value of other sensors in the same area by more than 30% for more than one hour, the system determines that the sensor has failed. The interface displays the sensor number X as abnormal and suggests cleaning the probe or replacing the battery. At the same time, the failure time and location are recorded for subsequent maintenance scheduling.

[0065] The system provides early warning for water quality exceeding standards by comparing real-time monitoring data from the outlet with the maximum allowable fertilizer concentration for the current stage of farmland application. If any water quality parameter exceeds the threshold, the system immediately closes the corresponding outlet valve, displays a message indicating excessive nitrogen and phosphorus levels, suspends irrigation, checks wetland plant density, and simultaneously activates a backup storage tank to temporarily store the contaminated water. The backup storage tank has a capacity of 50% of the daily treatment capacity and is equipped with a stirrer and a reprocessing pump. The contaminated water is diluted and filtered twice before re-entering the treatment process until the water quality meets standards.

[0066] The farmland irrigation distribution and execution mechanism consists of an electric butterfly valve assembly, pressure-compensated drip irrigation tape, and a flow metering module. The pressure-compensated drip irrigation tape is laid between crop rows, with a dripper spacing of 0.3m, a rated working pressure of 0.1MPa, and a dripper flow error not exceeding ±5%. The system performs a self-check of the drip irrigation tape pressure at 6:00 AM daily, stabilizing the pressure of each branch at the set value by adjusting the main pump speed. If the pressure of a branch deviates by more than 15% continuously, the system determines that it is due to dripper blockage or pipe rupture, shuts down that branch, and triggers a drip irrigation anomaly alarm. Data from the flow metering module is used to generate irrigation logs, which include the actual irrigation volume, theoretical water requirement, deviation rate, and execution time for each field. Logs are archived daily and included in the monthly report generation.

[0067] The automatic cleaning mechanism of the crop growth image acquisition unit consists of a miniature air pump, a nozzle, and wiper blades. The air pump operates at a pressure of 0.3 MPa, and the nozzle angle is adjustable to ensure airflow covers the entire lens surface. The wiper blades are made of silicone, with a weather resistance of at least five years. The wiper blade movement is driven by a miniature stepper motor with a stroke accuracy of ±0.5 mm. The cleaning program execution log records the cleaning time, air pump operating time, and wiper movement count for each cleaning session. If the image recognition confidence level remains below 90% after three consecutive cleanings, the system determines that the lens is severely damaged and triggers a manual cleaning alarm.

[0068] The crop fertilizer requirement calculation module in the dynamic matching model employs piecewise linear interpolation, constructing a two-dimensional lookup table with crop growth stage as the x-axis and daily nitrogen and phosphorus requirements per unit area as the y-axis. The lookup table data originates from crop fertilization guidelines published by agricultural research institutions and is entered into the system after field verification.

[0069] The system is equipped with a remote data upload interface using General Packet Radio Service (GPRS) protocol. Every day at 8:00 AM, the system automatically uploads compressed and encrypted data from the previous day to the cloud server. Uploaded content undergoes hash verification to ensure integrity. If transmission fails, it automatically retryes three times; if all three attempts fail, the error is recorded and the system waits for the next upload cycle. A data analysis module is deployed on the cloud server, using time series analysis and clustering algorithms to generate monthly operational reports.

[0070] The report includes a system processing efficiency trend chart (horizontal axis: date; vertical axis: chemical oxygen demand removal rate, ammonia nitrogen removal rate, total phosphorus removal rate); a resource utilization statistics table listing the nitrogen and phosphorus substitution and total water savings for each crop type; and a common fault type distribution chart, showing the percentage of four types of faults—sensor failure, valve jamming, power outage, and image recognition failure—in pie chart format. The report is sent to designated farmers' mobile phones via SMS, including a report summary and a link to download the full report. Farmers can view detailed data through their mobile browsers.

[0071] The aeration rate adjustment mechanism for the biological filter is as follows: The initial aeration rate is determined by referring to a table based on the influent ammonia nitrogen concentration, with intervals of 5 mg / L. The system measures the dissolved oxygen concentration in the filter effluent every 30 minutes, with the detection point located 50 cm above the filter bed using an optical dissolved oxygen sensor. The measurement range is 0 to 10 mg / L, with an accuracy of ±2%. If three consecutive readings are below 2 mg / L, the aeration rate is increased by 20%; if three consecutive readings are above 5 mg / L, the aeration rate is decreased by 15%. The adjustment range cannot exceed 25% in a single instance, and the cumulative number of adjustments per day cannot exceed eight to prevent biofilm detachment due to severe disturbance. The aeration rate adjustment record includes the adjustment time, the value before adjustment, the value after adjustment, and the triggering conditions, which are used for subsequent operational analysis and parameter optimization.

[0072] Harvesting of plants in constructed wetlands is carried out manually by farmers following system prompts. Upon triggering the harvesting command, the system simultaneously displays a plan view of the harvesting area on the controller interface, indicating suggested harvesting boundaries, recommended travel direction arrows, and safety precaution icons. A backpack brush cutter is recommended, with the blade speed set to 3000 revolutions per minute and the harvesting height 20cm above the wetland surface to ensure root preservation and promote regeneration. Harvested material is piled at a designated collection point equipped with a weighing platform. The system automatically records the weight of the harvested material for subsequent nitrogen and phosphorus release calculations. The calculation formulas are: nitrogen release equals harvested material weight multiplied by nitrogen content per unit weight; phosphorus release equals harvested material weight multiplied by phosphorus content per unit weight. Nitrogen and phosphorus contents are retrieved from an internal database based on plant species and growth stage, with initial values ​​of 2.5% and 0.8% respectively, dynamically updated based on laboratory sampling results.

[0073] The simplified operation and maintenance interface features voice broadcast functionality. When the system detects abnormal critical parameters or requires farmers to perform operations, pre-recorded voice prompts are played simultaneously in addition to the screen display. The voice content includes operation step numbers, a tool preparation list, estimated time, and security requirements. The voice speed is set to 120 words per minute, the volume is manually adjustable, and the broadcast language is the local dialect. The system has a built-in voice library of five major dialects, and farmers can select the matching dialect type through the menu when using it for the first time. The voice broadcast log records the time of each broadcast, the triggering conditions, and a summary of the broadcast content.

[0074] The system is equipped with a backup power module, which consists of a lead-acid battery pack and a solar charge controller. The battery pack has a rated capacity of 200A and a nominal voltage of 12V, employs a sealed, maintenance-free design, and has a cycle life exceeding 500 cycles. The solar charge controller has a maximum input power of 300W and features maximum power point tracking. When the mains power is interrupted, the system automatically switches to the backup power supply, prioritizing power to the sensor array, controller, and electric valves, while the aeration equipment and image acquisition unit enter a low-power standby mode.

[0075] During standby, the system wakes up every two hours to check the mains power recovery status. If the mains power is restored, it automatically switches back to the mains power and restarts the standby module. The backup power monitoring module samples the voltage value every 10 minutes. When the voltage is below 11V, a low power alarm is triggered. When the voltage is below 10V, unnecessary modules are shut down, maintaining only basic monitoring and alarm functions.

[0076] The farmland soil moisture sensor uses the frequency domain reflectance principle, with a probe length of 30cm, a measurement range of 5% to 50%, and a measurement accuracy of ±2%. The sensor automatically performs a self-calibration procedure every six hours. The calibration process is completed through a built-in reference capacitor. Data upload is paused during calibration. After calibration, calibration coefficients are generated and applied to correct subsequent measurement values.

[0077] The calibration coefficients are stored in the sensor's internal memory and are automatically loaded each time the sensor is powered on. In case of calibration failure, the system records an error code and displays a message on the interface indicating that the soil sensor needs to be manually reset. Simultaneously, a reset instruction SMS is sent to the farmer's mobile phone. The reset operation includes power-off restart, probe cleaning, and position adjustment. After the operation is completed, the system re-performs self-test and calibration.

[0078] The crop nutrient requirement calculation module in the dynamic matching model employs piecewise linear interpolation, constructing a two-dimensional lookup table with crop growth stage as the x-axis and daily nitrogen and phosphorus requirements per unit area as the y-axis. The lookup table data originates from crop fertilization guidelines published by agricultural research institutions and is entered into the system after field verification. Calculation results undergo boundary checks; if they exceed crop physiological limits, they are automatically truncated to those limits to prevent abnormal commands.

[0079] The system performs an initialization configuration process during the first installation. The process includes: Farmers select the climate type of their region via the controller, and the system loads the corresponding crop growth cycle baseline parameters. These parameters include the start and end dates of the standard growth stages for each crop, the suitable temperature range, and precipitation requirements. Farmers input the area of ​​each irrigated farmland, soil texture type, and the types of crops typically grown. Based on this input information, the system generates an initial set of control parameters. The parameter set includes a filter media replacement baseline cycle table (e.g., 90 days for sandy soil and 120 days for clay soil); a recommended harvest time window table (e.g., 45 to 50 days after heading for rice); and a water and fertilizer requirement threshold table for each crop growth stage (e.g., 0.6 for water requirement and 0.4 for fertilizer requirement during the seedling stage for leafy vegetables). The system prompts farmers to install soil sensors and image acquisition units in each farmland. After installation, the system performs sensor location registration and number binding operations, storing the binding information in non-volatile memory.

[0080] The registration process includes sensor self-test, signal strength test, and data consistency verification. If any step fails, a reinstallation prompt will be displayed. After the system completes initialization, it automatically enters trial operation mode and continuously collects baseline data for 72 hours without executing any control commands. After 72 hours, an initial operation report is generated, which is then confirmed by the farmer before the system is officially put into use.

[0081] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0082] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for the resource utilization and treatment of rural domestic sewage, characterized in that, include: The multi-parameter sensing module collects real-time data on the chemical oxygen demand, ammonia nitrogen concentration, total phosphorus concentration, pH, and flow rate of the wastewater at the inlet and outlet of the wastewater treatment system. At the same time, it collects soil volumetric moisture content and crop growth image sequences in the target irrigated farmland area. The crop growth image sequence of seven consecutive days is identified based on a pre-trained convolutional neural network model to output a label of the current growth stage of the crop. The label includes six categories: seedling stage, tillering stage, jointing stage, heading stage, grain filling stage, and maturity stage. The chemical oxygen demand, ammonia nitrogen concentration, total phosphorus concentration, pH, flow rate, soil volumetric moisture content, crop type and growth stage labels of the wastewater are input into the dynamic matching model to calculate the scaling factor of the pretreatment filter media replacement cycle, the proportion of harvested area of ​​constructed wetland plants, and the allocation weight of irrigation water from each field. Based on the scaling factor of the pretreatment filter media replacement cycle, a filter media replacement prompt instruction is triggered; based on the proportion of the harvested area of ​​the artificial wetland plants, a harvesting area and tool recommendation instruction is triggered; based on the irrigation allocation weight of each field, an electric valve group is driven to adjust the opening of each branch to achieve water allocation on demand. The embedded graphical controller provides farmers with three types of operation interfaces: crop type selection, system status viewing, and fault handling guidance. It also activates the dialect voice broadcast function when parameters are abnormal or operations are triggered. Sensor data backup and dynamic matching model online fine-tuning are performed at 2:00 AM every day. Simultaneously, the operation data is uploaded to the cloud server at 8:00 AM via General Packet Radio Service to generate a monthly operation report.

2. The method for resource utilization and treatment of rural domestic sewage according to claim 1, characterized in that, This study uses a pre-trained convolutional neural network model to identify crop growth image sequences over seven consecutive days and output labels indicating the current growth stage of the crop, including: After performing an automatic cleaning procedure on farmland images collected at 10:00 AM daily, grayscale normalization and size cropping are performed to generate standardized input images. Seven consecutive days of standardized input images are input into a convolutional neural network model containing five convolutional layers, three max pooling layers, and two fully connected layers to generate six types of growth stage probability distributions. The label with the highest probability is selected as the label for the current crop growth stage. The identification results are then bound to the crop type and stored in non-volatile memory.

3. The method for resource utilization and treatment of rural domestic sewage according to claim 2, characterized in that, The wastewater chemical oxygen demand (COD), ammonia nitrogen concentration, total phosphorus concentration, pH, flow rate, soil volumetric moisture content, crop type, and growth stage labels are input into a dynamic matching model to calculate the pretreatment filter media replacement cycle scaling factor, the proportion of harvested plant area in the constructed wetland, and the irrigation allocation weight of effluent from each field, including: After normalizing the input parameters, the input is fed into a three-layer multilayer perceptron structure with 128, 64, and 32 neurons in the hidden layer, and the activation function is a modified linear unit function. The output layer generates three control command vectors, which correspond to the filter media replacement cycle scaling factor, the proportion of plant harvesting area, and the irrigation water allocation weight for each field, respectively.

4. The method for resource utilization and treatment of rural domestic sewage according to claim 3, characterized in that, Triggering a filter media replacement prompt instruction based on the pretreatment filter media replacement cycle scaling factor includes: When the chemical oxygen demand at the inlet exceeds 300 mg / L for three consecutive days or the ammonia nitrogen concentration exceeds 25 mg / L for three consecutive days, the current filter media replacement cycle will be shortened to half of the original cycle. When the above parameters are below the threshold for seven consecutive days, the original replacement cycle will be restored. At the same time, an icon will flash on the controller interface and a voice prompt will play: "The filter media load is too high. Please replace it within seven days." 5. The method for resource utilization and treatment of rural domestic sewage according to claim 4, characterized in that, Based on the proportion of the harvested area of ​​the constructed wetland plants, the command to recommend harvesting areas and tools is triggered, including: A harvesting command is triggered when the target crop enters its peak fertilizer demand period and the total phosphorus concentration in the effluent exceeds 80% of the upper limit of the recommended fertilizer concentration for the current stage. The harvesting area ratio is calculated using the formula: the amount of fertilizer deficiency in the crop divided by the total nitrogen and phosphorus reserves in the wetland plants per unit area. The result is rounded to one decimal place. If the ratio is less than 5%, it will not be triggered. If it is greater than 80%, it will be implemented in batches.

6. The method for resource utilization and treatment of rural domestic sewage according to claim 5, characterized in that, Based on the irrigation allocation weight of each field, the electric valve assembly is driven to adjust the opening of each branch to achieve on-demand water distribution, including: The comprehensive water and fertilizer requirement index for each field was calculated. This index was obtained by weighted summation of the crop stage water requirement coefficient, stage fertilizer requirement coefficient, and soil water shortage coefficient, with weights of 0.4, 0.4, and 0.2, respectively. All treated effluent for the day is allocated according to the proportion of the comprehensive water and fertilizer demand index of each field to the total. The opening of the electric butterfly valve is adjusted every 10 minutes by a proportional-integral-derivative controller, and the valve response time is less than 5 seconds.

7. The method for resource utilization and treatment of rural domestic sewage according to claim 6, characterized in that, The embedded graphical controller provides farmers with three types of user interfaces: crop type selection, system status viewing, and troubleshooting guidance. The interface uses a three-row, four-column icon matrix layout, with the first row corresponding to five crop selection categories: rice, leafy vegetables, eggplant, citrus, and stone fruit. The second row corresponds to four types of status charts: water quality trend chart, soil moisture chart, filter media life bar, and plant harvest countdown. The third line corresponds to four types of troubleshooting guidelines: sensor calibration, valve reset, power restart, and contacting after-sales service. Clicking on it will automatically play the illustrated and video operation steps.

8. The method for resource utilization and treatment of rural domestic sewage according to claim 7, characterized in that, Sensor data backup and online fine-tuning of the dynamic matching model are performed daily at 2:00 AM, including: Back up all raw sensor data, control command execution records, and farmer operation logs for the day, and store the compressed and encrypted data on a local solid-state drive. Using the deviation between the actual crop growth response and the predicted fertilization effect on the same day as a feedback signal, the weight parameters of the crop fertilizer requirement calculation module are adjusted, the learning rate is set to 0.001, and each iteration is 50 steps.

9. A rural domestic sewage resource utilization and treatment system, characterized in that, include: The multi-parameter sensing module is used to collect real-time data on chemical oxygen demand, ammonia nitrogen concentration, total phosphorus concentration, pH and flow rate of wastewater at the inlet and outlet of the wastewater treatment system, while simultaneously collecting soil volumetric moisture content and crop growth image sequences of the target irrigated farmland area. The crop growth stage identification module is used to identify the crop growth image sequence of seven consecutive days based on a pre-trained convolutional neural network model and output the label of the current crop growth stage. The dynamic matching calculation module is used to input wastewater quality parameters, soil moisture content, crop type and growth stage labels into the dynamic matching model to calculate the scaling factor of the pretreatment filter replacement cycle, the proportion of harvested area of ​​artificial wetland plants, and the allocation weight of irrigation water from each field. The execution control module is used to trigger filter media replacement prompts, plant harvesting commands, and electric valve opening adjustment commands based on calculation results. The farmer interaction module is used to provide three types of operation interfaces through an embedded graphical controller: crop selection, status viewing, and fault handling, and to start dialect voice broadcasting in case of abnormality. The data management module is used to perform data backup and model fine-tuning every day at midnight, and upload the running data to the cloud server at 8:00 AM to generate a monthly report.

10. The rural domestic sewage resource utilization and treatment system according to claim 9, characterized in that, The multi-parameter sensing module is used for: A water quality sensor array is deployed at the inlet and outlet, with a sampling frequency of once every 10 minutes; Install an electromagnetic flow meter on the main water supply pipeline; measurement accuracy ±1%. A soil moisture sensor based on the frequency domain reflectance principle is buried 20cm below the topsoil layer and performs self-calibration every six hours. A visible light camera with an automatic cleaning mechanism is mounted on a field frame and images are collected every day at 10:00 a.m. The cleaning process includes a three-second compressed air jet and two reciprocating wiper movements.

Images