Method and system for managing non-point source pollution and recycling tail water in low-hill mountainous area

Abstract

The invention provides a low-hill mountainous area non-point source pollution management and tail water reutilization method and system, and the method comprises the steps: obtaining multi-source water quality environment data corresponding to a target low-hill mountainous area based on a multi-mode intelligent environment sensing device; and inputting the multi-source water quality environment data into the low-hill mountainous area non-point source pollution management model to obtain target control parameters corresponding to each target water and fertilizer flow regulation and control device output by the low-hill mountainous area non-point source pollution management model, and performing water and fertilizer putting control on each target water and fertilizer flow regulation and control device according to the target control parameters. The target water and fertilizer flow regulation and control equipment is water and fertilizer flow regulation and control equipment arranged in each farmland area in the target low-hill mountainous area; and on the basis of each target water and fertilizer flow regulation and control device after water and fertilizer putting control, tail water of the farmland runoff key node is purified, and the purified tail water is conveyed to the farmland area of the low mountainous area. The non-point source pollution load is obviously reduced, and the water and fertilizer utilization efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring and control technology, and in particular to a method and system for non-point source pollution management and wastewater reuse in low hilly areas. Background Technology

[0002] Low hilly and mountainous areas face severe challenges from non-point source pollution due to their unique geographical features and agricultural production patterns. The rapid formation and high frequency of surface runoff lead to intensified soil erosion, and large amounts of pollutants such as nitrogen and phosphorus enter water bodies with the runoff, seriously threatening ecological and environmental security.

[0003] Existing monitoring methods largely rely on manual sampling and laboratory analysis, which have limitations such as low spatiotemporal resolution and slow response, making it difficult to meet the needs of real-time dynamic monitoring and precise management. Furthermore, traditional water and fertilizer management practices are extensive, resulting in low water and fertilizer utilization efficiency, severe nutrient loss, and further exacerbating non-point source pollution.

[0004] Therefore, there is an urgent need for a method and system for non-point source pollution management and wastewater reuse in low hilly areas to solve the above problems. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a method and system for non-point source pollution management and wastewater reuse in hilly areas.

[0006] This invention provides a method for non-point source pollution management and effluent reuse in hilly mountainous areas, comprising: Based on multimodal intelligent environmental sensing equipment, acquire multi-source water quality environmental data corresponding to the target low hilly area; The multi-source water quality environmental data is input into the low hilly area non-point source pollution management model to obtain the target control parameters corresponding to each target water and fertilizer flow regulation device output by the low hilly area non-point source pollution management model. Based on the target control parameters, water and fertilizer application control is performed on each of the target water and fertilizer flow regulation devices. The target water and fertilizer flow regulation devices are water and fertilizer flow regulation devices set up in each farmland area in the target low hilly area. Based on the target water and fertilizer flow control devices after water and fertilizer application control, the tailwater at key nodes of farmland runoff is purified and transported to the farmland area in the low hilly area.

[0007] According to the present invention, a method for non-point source pollution management and wastewater reuse in low hilly areas is provided, wherein the multimodal intelligent environmental sensing device includes a floating water quality monitoring station, a biomimetic fish for water quality monitoring, a soil sensor, and a meteorological data acquisition device; The multimodal intelligent environmental sensing device acquires multi-source water quality environmental data corresponding to the target low hilly area, including: Based on the floating water quality monitoring station, the water quality parameters in the target low hilly area are monitored to obtain the first water quality monitoring parameters, which include chemical oxygen demand, dissolved oxygen concentration, water pH value, ammonia nitrogen content, total nitrogen content, total phosphorus content, water temperature value, water conductivity, and turbidity. Based on the cruising path of the bionic fish for water quality monitoring in the target low hilly area, the second water quality monitoring parameters collected by the bionic fish for water quality monitoring are obtained, wherein the second water quality monitoring parameters include at least chemical oxygen demand, ammonia nitrogen, total phosphorus and heavy metal ion concentration. Based on the soil sensor, soil parameters in the target low hilly area are monitored to obtain soil monitoring parameters, wherein the soil monitoring parameters include at least soil moisture, soil temperature, soil pH, nitrogen, phosphorus and potassium content, organic matter content and soil electrical conductivity. Based on the meteorological data acquisition equipment, meteorological environmental data corresponding to the target low hilly area are acquired, wherein the meteorological environmental data includes at least rainfall, temperature, ambient humidity, wind speed and direction, and solar radiation value; The multi-source water quality environmental data are obtained based on the first water quality monitoring parameters, the second water quality monitoring parameters, the soil monitoring parameters, and the meteorological environmental data.

[0008] According to the present invention, a method for non-point source pollution management and wastewater reuse in hilly areas is provided. The non-point source pollution management model in hilly areas includes a pollution source tracing and diffusion prediction model, a water and fertilizer demand intelligent prediction model, and a runoff cascade coordinated management and optimization model. The pollution source tracing and diffusion prediction model is constructed based on deep learning algorithms and hydrodynamic models; the water and fertilizer demand intelligent prediction model is constructed based on reinforcement learning or deep learning algorithms; and the runoff cascade coordinated management and optimization model is constructed with the objectives of maximizing runoff pollution reduction, maximizing water and fertilizer utilization efficiency, and minimizing water resource consumption. The process involves inputting the multi-source water quality environmental data into a low-hilly mountain area non-point source pollution management model to obtain target control parameters corresponding to each target water and fertilizer flow regulation device output by the low-hilly mountain area non-point source pollution management model, including: Acquire runoff data, agricultural activity data, geographic information system data, crop types, crop growth stages, historical water and fertilizer use efficiency data, and runoff cascade distribution characteristics data of the target low hilly area; The first water quality monitoring parameter, the second water quality monitoring parameter, the runoff data, the meteorological environment data, the agricultural activity data, and the geographic information system data are input into the pollution source tracing and diffusion prediction model to obtain the pollution source location information and pollution diffusion prediction information of the target low hilly area output by the pollution source tracing and diffusion prediction model. The crop type, crop growth stage, soil monitoring parameters, meteorological environmental data, and historical water and fertilizer use efficiency data are input into the intelligent water and fertilizer demand prediction model to obtain the water and fertilizer demand prediction information of the target low hilly area output by the intelligent water and fertilizer demand prediction model. The water and fertilizer demand prediction information includes water and fertilizer demand prediction data and water and fertilizer application prediction time. The pollution source location information, the pollution diffusion prediction information, the water and fertilizer demand prediction information, and the runoff cascade distribution characteristic data are input into the runoff cascade collaborative management and optimization model to obtain the target control parameters output by the runoff cascade collaborative management and optimization model.

[0009] According to the present invention, a method for non-point source pollution management and wastewater reuse in low hilly areas includes purifying wastewater from key nodes of farmland runoff using target water and fertilizer flow control devices based on water and fertilizer application control. When the tailwater level at the key node of the farmland runoff is determined to meet the preset level height, the tailwater at the key node of the farmland runoff is purified based on a multi-level biological ecological purification module. The multi-level biological ecological purification module includes a pre-sedimentation and physical filtration unit, an anaerobic / anoxic biological reaction unit, an aerobic biological filter bed / constructed wetland unit, an ecological pond / stabilization pond, and a biofilm reactor.

[0010] According to the present invention, a method for non-point source pollution management and effluent reuse in hilly mountainous areas is provided, the method further comprising: Based on an unsupervised learning algorithm, data anomaly detection is performed on the multi-source water quality environmental data, and corresponding equipment fault alarm information or data acquisition anomaly alarm information is generated according to the data anomaly detection results.

[0011] According to the present invention, a method for non-point source pollution management and wastewater reuse in low-hilly mountainous areas, before inputting the multi-source water quality environmental data into a low-hilly mountainous area non-point source pollution management model to obtain the target control parameters corresponding to each target water and fertilizer flow regulation device output by the low-hilly mountainous area non-point source pollution management model, the method further includes: The multi-source water quality environmental data is preprocessed to obtain preprocessed multi-source water quality environmental data. The preprocessing includes data cleaning, noise reduction, missing value imputation, time series alignment, and spatial interpolation.

[0012] This invention also provides a non-point source pollution management and effluent reuse system for hilly mountainous areas, comprising: The multimodal environment sensing module is used to acquire multi-source water quality environmental data corresponding to the target low hilly area based on multimodal intelligent environment sensing equipment; The pollution control and decision-making module is used to input the multi-source water quality environmental data into the low hilly area non-point source pollution management model, obtain the target control parameters corresponding to each target water and fertilizer flow regulation device output by the low hilly area non-point source pollution management model, and control the water and fertilizer application of each target water and fertilizer flow regulation device according to the target control parameters. The target water and fertilizer flow regulation device is the water and fertilizer flow regulation device set up in each farmland area in the target low hilly area. The tailwater resource utilization module is used to purify the tailwater at key nodes of farmland runoff based on the target water and fertilizer flow regulation devices after water and fertilizer application control, and to transport the purified tailwater to the farmland area in the low hilly area.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the non-point source pollution management and wastewater reuse method in low hilly areas as described above.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the non-point source pollution management and wastewater reuse method in low hilly areas as described above.

[0015] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the non-point source pollution management and wastewater reuse methods for low hilly areas as described above.

[0016] The present invention provides a method and system for non-point source pollution management and tailwater reuse in low hilly areas. It acquires multi-source water quality environmental data of the target low hilly area through multi-modal intelligent environmental sensing equipment and inputs it into the low hilly area non-point source pollution management model to obtain the control parameters of the target water and fertilizer flow regulation equipment. Then, it controls the application of water and fertilizer and purifies the tailwater at key nodes of farmland runoff before transporting it back to the farmland area. This achieves coordinated, efficient and intelligent management of runoff cascade water and fertilizer, significantly reduces non-point source pollution load, and improves water and fertilizer utilization efficiency. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 A schematic diagram of the process for non-point source pollution management and effluent reuse in hilly areas provided by the present invention; Figure 2 A schematic diagram of the structure of the non-point source pollution management and effluent reuse system in low hilly areas provided by the present invention; Figure 3 A schematic diagram of the overall architecture of the non-point source pollution management and effluent reuse system in low hilly areas provided by the present invention; Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0020] Low hilly and mountainous areas have unique geographical and agricultural production patterns. Due to the rapid formation and frequent occurrence of surface runoff, soil erosion is exacerbated, and large amounts of pollutants such as nitrogen and phosphorus enter water bodies with the runoff, threatening ecological security. Existing monitoring methods suffer from low spatiotemporal resolution and slow response; at the same time, traditional water and fertilizer management methods are extensive, exacerbating non-point source pollution.

[0021] To address the problems existing in the prior art, this invention utilizes the Internet of Things, artificial intelligence, biomimetic sensing, and efficient biological ecological treatment technologies to achieve precise monitoring, intelligent management, and wastewater recycling of non-point source pollution, thereby constructing a sustainable agricultural ecosystem.

[0022] Figure 1 A schematic flowchart illustrating the method for non-point source pollution management and effluent reuse in low-hilly mountainous areas provided by this invention is shown below. Figure 1 As shown, the present invention provides a method for non-point source pollution management and wastewater reuse in low hilly areas, comprising: Step 101: Based on the multimodal intelligent environmental sensing device, acquire multi-source water quality environmental data corresponding to the target low hilly area.

[0023] In this invention, based on the topographic features and runoff paths of low hilly areas, floating water quality monitoring stations and intelligent soil and meteorological sensor networks are deployed at key locations such as the top, middle, and bottom of the slope, as well as runoff collection areas. These environmental sensing devices together constitute a multi-layered, high-density environmental sensing network capable of collecting environmental data from different locations and angles.

[0024] Specifically, floating water quality monitoring stations and intelligent soil and meteorological sensor networks operate continuously, collecting data on various aspects, including water quality (such as ammonia nitrogen and total phosphorus), soil (soil nutrients and humidity), and meteorological data (rainfall, wind speed, and temperature). Simultaneously, based on the water area and monitoring needs, an appropriate number of biomimetic fish for water quality monitoring are deployed. These fish can autonomously navigate underwater, collecting real-time water quality data, underwater images, and flow field information, further enriching the sources of water quality environmental data.

[0025] Furthermore, the collected data (i.e., multi-source water quality environmental data) is transmitted to the cloud management platform in real time through different communication networks. In this invention, data from floating water quality monitoring stations and smart soil and meteorological sensor networks are transmitted via LoRaWAN / NB-IoT networks; data from biomimetic fish used for water quality monitoring is transmitted via underwater acoustic / optical communication or surface-mounted via 5G / LoRaWAN. This multimodal data transmission method ensures that data reaches the cloud management platform in a timely and accurate manner, providing a foundation for subsequent analysis and decision-making.

[0026] Step 102: Input the multi-source water quality environmental data into the low hilly area non-point source pollution management model to obtain the target control parameters corresponding to each target water and fertilizer flow regulation device output by the low hilly area non-point source pollution management model, and control the water and fertilizer application of each target water and fertilizer flow regulation device according to the target control parameters. The target water and fertilizer flow regulation device is the water and fertilizer flow regulation device set up in each farmland area in the target low hilly area.

[0027] In this invention, after receiving multi-source water quality environmental data, the cloud management platform preferably first cleans, integrates and standardizes the massive heterogeneous data (i.e., multi-source water quality environmental data) to ensure the quality and consistency of the data.

[0028] Furthermore, the processed data is input into the low-hilly mountain area non-point source pollution management model. In this invention, the low-hilly mountain area non-point source pollution management model consists of the following sub-models, specifically including: Pollution source tracing and diffusion prediction model: Combining meteorological data, soil data, and water quality data, when an anomaly is detected (such as an increase in ammonia nitrogen or total phosphorus concentration in a certain area), the model is immediately activated. Combining agricultural activity records and geographic information, it accurately locates the pollution source, for example, determining that the pollution is caused by a farmer's excessive fertilization.

[0029] Intelligent water and fertilizer demand prediction model: Based on crop growth models and real-time soil nutrient data, it calculates the precise demand of crops for nutrients such as nitrogen, phosphorus, and potassium, and recommends the best fertilization time and irrigation amount in conjunction with weather forecasts.

[0030] In this invention, when the risk of runoff pollution is predicted to increase based on meteorological forecasts (such as an impending heavy rainstorm) and data displayed by soil sensors (such as high soil saturation), the runoff control strategy is dynamically adjusted.

[0031] The runoff cascade coordinated management and optimization model outputs target control parameters for each target water and fertilizer flow regulation device based on the prediction results of the aforementioned model. In this invention, these target water and fertilizer flow regulation devices include intelligent irrigation valves and intelligent fertilizer pumps installed in various farmland areas of the target low hilly mountain region. Based on these target control parameters, precise water and fertilizer application control is implemented on these devices. For example, during the initial peak of rainwater runoff, the gates are automatically closed to divert high-concentration runoff into tailwater storage facilities; simultaneously, instructions are sent to the intelligent irrigation valves and fertilizer pumps to adjust irrigation and fertilization plans, avoiding fertilization during peak runoff periods.

[0032] Step 103: Based on the target water and fertilizer flow control devices after water and fertilizer application control, the tailwater of key nodes of farmland runoff is purified and the purified tailwater is transported to the farmland area of ​​the low hilly area.

[0033] In this invention, intelligent irrigation valves, intelligent fertilizer pumps, and runoff control gates are installed in the farmland irrigation system and runoff outlets. After water and fertilizer application is controlled, agricultural wastewater is collected in wastewater storage facilities built at the end of the runoff collection point. These wastewater storage facilities serve as temporary storage and buffering facilities, providing conditions for subsequent purification treatment.

[0034] In this invention, a multi-stage biological ecological purification unit is designed and constructed to deeply treat the effluent based on its quality and treatment requirements. For example, the effluent first enters a sedimentation tank to remove large suspended solids; then it flows into an anaerobic tank for denitrification; next, it flows through an artificial wetland and an ecological pond, where it is further purified through plant absorption and microbial degradation. After this series of treatments, the effluent quality meets the standards for farmland irrigation.

[0035] After purification, the treated wastewater is transported back to farmland in low hilly areas for precision irrigation through a smart water and fertilizer recycling network. This recycling method achieves efficient use of water resources, significantly reduces the consumption of fresh water, and also reduces non-point source pollution emissions.

[0036] In this invention, indicators such as water quality, soil, crop growth, and water and fertilizer use efficiency are continuously monitored, and the data is fed back to a cloud management platform. Through AI analysis and decision-making engines, the non-point source pollution management model and control algorithms for low-hilly areas are continuously optimized based on actual operational results, achieving adaptive learning and performance improvement of the system. Furthermore, this invention provides detailed operational reports and effect evaluations through a user interface, helping managers understand the system's benefits and make necessary adjustments based on actual conditions to ensure the system is always in optimal operating condition.

[0037] The present invention provides a method for non-point source pollution management and tailwater reuse in low hilly areas. This method acquires multi-source water quality environmental data of the target low hilly area through a multi-modal intelligent environmental sensing device and inputs it into a low hilly area non-point source pollution management model to obtain control parameters for the target water and fertilizer flow regulation device. This allows for the control of water and fertilizer application, and the tailwater from key nodes of farmland runoff is then purified and transported back to the farmland area. This achieves coordinated, efficient and intelligent management of runoff cascade water and fertilizer, significantly reduces non-point source pollution load, and improves water and fertilizer utilization efficiency.

[0038] Based on the above embodiments, the multimodal intelligent environmental sensing device includes a floating water quality monitoring station, a bionic fish for water quality monitoring, a soil sensor, and a meteorological data acquisition device; The multimodal intelligent environmental sensing device acquires multi-source water quality environmental data corresponding to the target low hilly area, including: Based on the floating water quality monitoring station, the water quality parameters in the target low hilly area are monitored to obtain the first water quality monitoring parameters, which include chemical oxygen demand, dissolved oxygen concentration, water pH value, ammonia nitrogen content, total nitrogen content, total phosphorus content, water temperature value, water conductivity, and turbidity. Based on the cruising path of the bionic fish for water quality monitoring in the target low hilly area, the second water quality monitoring parameters collected by the bionic fish for water quality monitoring are obtained, wherein the second water quality monitoring parameters include at least chemical oxygen demand, ammonia nitrogen, total phosphorus and heavy metal ion concentration. Based on the soil sensor, soil parameters in the target low hilly area are monitored to obtain soil monitoring parameters, wherein the soil monitoring parameters include at least soil moisture, soil temperature, soil pH, nitrogen, phosphorus and potassium content, organic matter content and soil electrical conductivity. Based on the meteorological data acquisition equipment, meteorological environmental data corresponding to the target low hilly area are acquired, wherein the meteorological environmental data includes at least rainfall, temperature, ambient humidity, wind speed and direction, and solar radiation value; The multi-source water quality environmental data are obtained based on the first water quality monitoring parameters, the second water quality monitoring parameters, the soil monitoring parameters, and the meteorological environmental data.

[0039] In this invention, the multimodal intelligent environmental sensing device is a comprehensive sensing system composed of a floating water quality monitoring station, a biomimetic fish for water quality monitoring, a soil sensor, and a meteorological data acquisition device. These different types of devices each have unique functions and advantages, and work together to comprehensively and accurately acquire environmental data on water quality, soil, and meteorology in the target low hilly area.

[0040] Specifically, floating water quality monitoring stations are deployed in key water areas of the target low hilly and mountainous region. These stations serve as the core units for long-term fixed-point monitoring and can reliably acquire water quality information for the area.

[0041] In this invention, the floating water quality monitoring station integrates a high-precision multi-parameter sensor, enabling the monitoring of various water quality parameters to obtain the first water quality monitoring parameters. These parameters specifically include: Chemical Oxygen Demand (COD), reflecting the content of organic pollutants in the water; Dissolved Oxygen Concentration (DO), affecting the survival of aquatic organisms; pH value, affecting the acid-base balance of the water; Ammonia Nitrogen (NH3-N), Total Nitrogen (TN), and Total Phosphorus (TP), indicators closely related to eutrophication; Water Temperature, as temperature changes affect the metabolism and chemical reaction rates of aquatic organisms; Water Conductivity, reflecting the total concentration of ions in the water; and Turbidity, reflecting the clarity of the water. It should be noted that the detection accuracy of all parameters in this invention meets the relevant water quality requirements of the surface water environmental quality standards.

[0042] Preferably, the floating water quality monitoring station adopts a hybrid power supply system of solar energy, supercapacitor, and lithium battery to ensure continuous operation for more than 120 hours under continuous rainy or extreme weather conditions. Simultaneously, it is designed with an adaptive multi-point anchoring device and an anti-entanglement mechanism, enabling it to maintain stable monitoring in water bodies with flow velocities up to 2.5 m / s and resist entanglement by aquatic plants.

[0043] In this invention, LoRaWAN or NB-IoT narrowband IoT technologies can be used to achieve continuous, 24 / 7 monitoring of water quality parameters and transmit encrypted data to a cloud management platform in real time. Furthermore, the monitoring station also possesses edge computing capabilities, enabling preliminary anomaly detection and early warning, and timely identification of water quality abnormalities.

[0044] In this invention, the biomimetic fish for water quality monitoring adopts a biomimetic streamlined design and flexible biomimetic fin drive technology to achieve ultra-quiet, high-efficiency and low-energy underwater propulsion, minimizing disturbance to the aquatic ecosystem and enabling it to integrate more naturally into the aquatic environment for monitoring.

[0045] Specifically, the bionic fish for water quality monitoring has a built-in low-power AI chip, enabling it to perceive the environment, locate itself autonomously, plan its path, and avoid obstacles. It can autonomously plan the optimal patrol path based on preset tasks (such as regional grid-based patrols, tracking specific pollution sources, and exploring hidden areas). When it detects abnormal water quality, the bionic fish can immediately activate an emergency response mode, automatically adjust its patrol strategy, track the pollution spread path in real time, and accurately locate the pollution source.

[0046] In this invention, the biomimetic fish for water quality monitoring is equipped with miniaturized, highly sensitive water quality sensors (such as miniature spectrometers and ion-selective electrode arrays) to monitor key indicators such as chemical oxygen demand, ammonia nitrogen, total phosphorus, and heavy metal ions in real time, i.e., secondary water quality monitoring parameters. Simultaneously, it integrates Doppler sonar, a high-resolution underwater camera, and a lateral-line biomimetic flow field sensor (such as FlowSight technology) to achieve comprehensive perception of water flow velocity, underwater topography, aquatic organism behavior, and abnormal physical disturbances, thereby obtaining more comprehensive water quality-related information.

[0047] Preferably, in this invention, multiple biomimetic fish can form an intelligent swarm, utilizing underwater acoustic or optical communication technologies for information sharing and collaborative operations. The swarm can achieve rapid coverage monitoring of large water areas, or multi-angle, multi-level, three-dimensional monitoring of specific polluted areas, significantly improving monitoring efficiency and data comprehensiveness.

[0048] Optionally, in this invention, the biomimetic fish for water quality monitoring has the function of automatic collection and in-situ rapid analysis of trace water samples, enabling high-precision detection of specific pollutants, reducing the lag in sample transportation and laboratory analysis, and timely acquisition of accurate water quality data. Furthermore, it adopts a hybrid power supply mode combining water flow energy, solar energy (water surface charging), and wireless charging, and has an intelligent recharge function, ensuring long-term stable operation without frequent manual intervention.

[0049] In this invention, soil sensors are deployed in key areas of farmland in low hilly and mountainous areas. These areas are selected based on the principle of being sensitive to and representative of changes in the farmland soil environment, and can accurately reflect the overall condition of the farmland soil.

[0050] Specifically, soil sensors monitor soil moisture in real time to obtain soil humidity; soil temperature affects the activity of microorganisms and nutrient transformation in the soil; soil pH determines the soil's acidity and alkalinity, affecting crop growth and nutrient absorption; nitrogen, phosphorus, and potassium (N, P, K) content are the main nutrients necessary for crop growth; organic matter content reflects the soil's fertility level; and soil electrical conductivity reflects the salt content in the soil. These data together constitute soil monitoring data.

[0051] In this invention, the soil sensor uses wireless mesh network technology to achieve efficient data transmission and self-organization, ensuring that soil monitoring data is transmitted to the cloud management platform in a timely and accurate manner.

[0052] Simultaneously, high-precision small-scale meteorological stations (i.e., meteorological data acquisition equipment) are deployed at suitable locations in the target low-hilly area. These locations can represent the overall meteorological conditions of the region, avoiding interference from local environmental factors. In this invention, the meteorological data acquisition equipment acquires real-time meteorological environmental data of the target low-hilly area, specifically including: rainfall, as the amount and distribution of rainfall affect farmland irrigation and soil and water conservation; temperature, which affects the growth cycle and physiological activities of crops; environmental humidity, which is closely related to crop transpiration and the occurrence of pests and diseases; wind speed and direction, which affect heat and water vapor exchange in farmland; and solar radiation, which provides energy for crop photosynthesis and is an important environmental factor for crop growth.

[0053] Finally, the primary water quality monitoring parameters obtained through floating water quality monitoring stations, the secondary water quality monitoring parameters collected by biomimetic fish, the soil monitoring parameters obtained by soil sensors, and the meteorological environmental data acquired by meteorological data acquisition equipment are integrated to form multi-source water quality environmental data. These data are interconnected and complementary, providing comprehensive and accurate environmental parameter basis for subsequent water and fertilizer management, non-point source pollution prediction, etc., which helps to achieve precise monitoring and management of the environmental conditions in the target low hilly area.

[0054] Based on the above embodiments, the non-point source pollution management model for low hilly areas includes a pollution source tracing and diffusion prediction model, a water and fertilizer demand intelligent prediction model, and a runoff cascade coordinated management and optimization model. The pollution source tracing and diffusion prediction model is constructed based on deep learning algorithms and hydrodynamic models; the water and fertilizer demand intelligent prediction model is constructed based on reinforcement learning or deep learning algorithms; and the runoff cascade coordinated management and optimization model is constructed with the objectives of maximizing runoff pollution reduction, maximizing water and fertilizer utilization efficiency, and minimizing water resource consumption. The process involves inputting the multi-source water quality environmental data into a low-hilly mountain area non-point source pollution management model to obtain target control parameters corresponding to each target water and fertilizer flow regulation device output by the low-hilly mountain area non-point source pollution management model, including: Acquire runoff data, agricultural activity data, geographic information system data, crop types, crop growth stages, historical water and fertilizer use efficiency data, and runoff cascade distribution characteristics data of the target low hilly area; The first water quality monitoring parameter, the second water quality monitoring parameter, the runoff data, the meteorological environment data, the agricultural activity data, and the geographic information system data are input into the pollution source tracing and diffusion prediction model to obtain the pollution source location information and pollution diffusion prediction information of the target low hilly area output by the pollution source tracing and diffusion prediction model. The crop type, crop growth stage, soil monitoring parameters, meteorological environmental data, and historical water and fertilizer use efficiency data are input into the intelligent water and fertilizer demand prediction model to obtain the water and fertilizer demand prediction information of the target low hilly area output by the intelligent water and fertilizer demand prediction model. The water and fertilizer demand prediction information includes water and fertilizer demand prediction data and water and fertilizer application prediction time. The pollution source location information, the pollution diffusion prediction information, the water and fertilizer demand prediction information, and the runoff cascade distribution characteristic data are input into the runoff cascade collaborative management and optimization model to obtain the target control parameters output by the runoff cascade collaborative management and optimization model.

[0055] In this invention, the non-point source pollution management model for low hilly areas is a comprehensive model system, which includes a pollution source tracing and diffusion prediction model, a water and fertilizer demand intelligent prediction model, and a runoff cascade collaborative management and optimization model. Each model has its unique construction method and objectives.

[0056] Specifically, the pollution source tracing and diffusion prediction model is built based on deep learning algorithms and hydrodynamic models. Deep learning algorithms, such as LSTM (Long Short-Term Memory) networks and Transformer, have powerful data processing and pattern recognition capabilities, enabling them to extract potential patterns from complex data; while hydrodynamic models are used to simulate the physical movement of pollutants in water bodies.

[0057] In this invention, the data input into the pollution source tracing and diffusion prediction model includes water quality data (including first and second water quality monitoring parameters), runoff data, meteorological environmental data, agricultural activity data (such as fertilization and irrigation records), and geographic information system (GIS) data. This data comprehensively covers various factors influencing non-point source pollution. Historical water quality data reflects the historical state of pollution, runoff data reflects the dynamic changes in water flow, meteorological data affects the diffusion conditions of pollutants, agricultural activity data is closely related to the generation of pollution, and GIS data provides geospatial information.

[0058] This invention utilizes machine learning (such as random forests and support vector machines) and deep learning algorithms to construct a non-point source pollution tracing model, accurately identifying pollution sources. Simultaneously, by combining hydrodynamic models with AI prediction algorithms, it simulates the diffusion paths and concentration changes of pollutants during runoff, enabling early warning and trend prediction of pollution events.

[0059] The intelligent water and fertilizer demand prediction model is built based on reinforcement learning or deep neural network models. Reinforcement learning can continuously learn the optimal strategy through interaction with the environment, while deep neural network models can model complex nonlinear relationships. This model comprehensively considers crop type, growth stage, soil nutrient content, soil moisture, meteorological conditions (rainfall, temperature, and light), and historical water and fertilizer use efficiency data. Among these factors, crop type and growth stage determine the characteristics of crop water and fertilizer requirements, soil nutrient content and moisture reflect the soil's supply capacity, meteorological conditions affect crop growth and water evaporation, and historical water and fertilizer use efficiency data provides a reference basis.

[0060] In this invention, the intelligent water and fertilizer demand prediction model dynamically predicts the precise amount of water and nutrients required by crops and the optimal application time; moreover, the model can learn adaptively and continuously optimize the prediction accuracy based on actual results, providing a scientific basis for rational water and fertilizer management.

[0061] In this invention, the runoff cascade coordinated management and optimization model is constructed based on the objectives of maximizing runoff pollution reduction, maximizing water and fertilizer utilization efficiency, and minimizing water resource consumption. As a multi-objective optimization problem, it needs to comprehensively consider multiple factors to achieve the overall optimal effect. The input data of the runoff cascade coordinated management and optimization model includes pollution prediction (including pollution source location information and pollution diffusion prediction information), water and fertilizer demand prediction information, and runoff cascade distribution characteristics. Among them, the runoff cascade distribution characteristics take into account the hierarchical structure of water flow and the possibility of regulation.

[0062] Furthermore, the runoff cascade coordinated management and optimization model utilizes multi-objective optimization algorithms (such as genetic algorithms and particle swarm optimization) to optimize the coordinated operation of equipment such as runoff control gates, intelligent irrigation valves, and intelligent fertilizer pumps, achieving the comprehensive goals of maximizing runoff pollution reduction, maximizing water and fertilizer utilization efficiency, and minimizing water resource consumption. In this invention, the multi-objective optimization algorithm can be dynamically adjusted according to real-time environmental changes to ensure the optimal control strategy.

[0063] In this invention, the intelligent irrigation valve precisely controls the amount and timing of irrigation water based on AI-predicted crop water requirements and soil moisture, enabling on-demand irrigation and avoiding water and fertilizer loss due to over-irrigation. The intelligent fertilizer pump, on the other hand, precisely mixes and dispenses liquid fertilizer based on AI-predicted crop nutrient needs and soil nutrient content, achieving frequent, small-volume, and precise fertilization, thus improving fertilizer utilization.

[0064] Runoff control gates, as intelligent control gates installed at key nodes in runoff cascades, can dynamically adjust the interception, diversion, or discharge of runoff based on rainfall intensity, runoff flow, water quality, and the load of downstream tailwater treatment facilities, thereby achieving source reduction and process control of runoff pollution. For example, during the peak period of initial rainwater runoff pollution, the gates can be automatically closed to divert high-concentration runoff into tailwater storage facilities for treatment, avoiding direct discharge.

[0065] The integrated water and fertilizer mixing device, also known as an intelligent fertilizer pump, achieves precise mixing of water and fertilizer, ensuring the uniformity and stability of the water and fertilizer solution, and delivers water and fertilizer evenly to the crop roots through pressure-compensated drip irrigation or micro-spraying technology.

[0066] Based on the above embodiments, the target water and fertilizer flow regulation devices after water and fertilizer application control purify the tailwater at key nodes of farmland runoff, including: When the tailwater level at the key node of the farmland runoff is determined to meet the preset level height, the tailwater at the key node of the farmland runoff is purified based on a multi-level biological ecological purification module. The multi-level biological ecological purification module includes a pre-sedimentation and physical filtration unit, an anaerobic / anoxic biological reaction unit, an aerobic biological filter bed / constructed wetland unit, an ecological pond / stabilization pond, and a biofilm reactor.

[0067] In this invention, there are some key nodes in the farmland runoff system, such as the convergence point of farmland runoff and the end of drainage ditches. At these key nodes, tailwater storage facilities such as seepage-proof tailwater storage ponds or ecological ponds are constructed. The facilities are equipped with intelligent liquid level sensors that can monitor the water storage volume in real time.

[0068] When the intelligent liquid level sensor detects that the tailwater level has reached the preset height, the conditions for triggering subsequent purification treatment are met. This preset height can be determined based on factors such as peak runoff flow and storage capacity considered during the facility design, ensuring that it can effectively collect runoff tailwater and excess irrigation tailwater generated during agricultural production, and start the purification process at the appropriate time to avoid tailwater accumulation or untimely treatment.

[0069] In this invention, the multi-level biological ecological purification module specifically includes the following units: The pre-sedimentation and physical filtration unit employs high-efficiency sedimentation tanks, screens, and microfiltration technologies. Its main function is to remove large suspended solids and some organic matter from the effluent. If large suspended solids directly enter subsequent treatment units, it will increase the treatment load, affect the treatment effect, and may even clog the equipment. The presence of some organic matter can also interfere with the activity of microorganisms in subsequent biological treatment.

[0070] The specific purification process of the pre-sedimentation and physical filtration unit is as follows: the effluent first enters the pre-sedimentation and physical filtration unit. In the high-efficiency sedimentation tank, large suspended particles settle to the bottom of the tank by gravity. The screen can intercept larger debris. Microfiltration technology can further filter out smaller suspended particles and some organic matter, thereby reducing the load on the subsequent biological treatment unit and creating favorable conditions for subsequent treatment.

[0071] The anaerobic / anoxic bioreactor unit utilizes anaerobic / anoxic microorganisms to denitrify and anaerobic decompose nitrate nitrogen and some organic matter in the effluent, thereby achieving nitrogen removal.

[0072] The purification process of the anaerobic / anoxic bioreactor unit is as follows: After treatment by the pre-sedimentation and physical filtration units, the effluent enters the anaerobic / anoxic bioreactor unit, which can employ UASB (Upflow Anaerobic Sludge Blanket), A / O (Anaerobic-Aerobic) processes, or their modifications. Under anaerobic / anoxic conditions, microorganisms reduce nitrate nitrogen to nitrogen gas, releasing it into the air. Simultaneously, they anaerobicly decompose some organic matter, converting it into smaller molecules, providing easier-to-treat substrates for subsequent treatment.

[0073] The aerobic biofilter / constructed wetland unit utilizes high-efficiency biological filter media (such as bio-ceramic granules and volcanic rock) or constructs vertical / horizontal flow constructed wetlands. Plants with high adsorption, enrichment, and degradation capabilities (such as canna lilies, reeds, and cattails) are planted in the filter bed or wetland. Combined with the nitrification process of attached microorganisms, ammonia nitrogen, organic matter, and phosphorus are further removed. Ammonia nitrogen, organic matter, and phosphorus are also major contributors to water pollution; removing these substances can effectively improve water quality.

[0074] The specific purification process of the aerobic biofilter / constructed wetland unit is as follows: Wastewater from the anaerobic / anoxic bioreactor unit enters the aerobic biofilter / constructed wetland unit. Within the biofilter media or constructed wetland, plant roots can adsorb and accumulate some pollutants, while simultaneously providing an attachment site for microorganisms. Microorganisms attached to the filter media and plant roots undergo nitrification under aerobic conditions, converting ammonia nitrogen into nitrate nitrogen, further reducing the ammonia nitrogen content. In addition, microorganisms can decompose organic matter, and plants can absorb some phosphorus, thereby achieving the purpose of water purification. Enhanced microbial agents can also be introduced to improve treatment efficiency.

[0075] Ecological ponds / stabilization ponds, serving as end-of-pipe treatment and buffer units, further purify water quality through natural sedimentation, plant absorption, microbial degradation, and photosynthesis, while also providing ecological landscape functions. Simultaneously, filter-feeding fish or shellfish can be introduced to establish a food chain and enhance self-purification capabilities.

[0076] The specific purification process of the ecological pond / stabilization pond is as follows: Wastewater treated by the aerobic biological filter / constructed wetland unit enters the ecological pond / stabilization pond. In the ecological pond / stabilization pond, heavier particles naturally settle to the bottom; aquatic plants continue to absorb nutrients from the water, such as nitrogen and phosphorus; microorganisms continuously degrade organic matter; and oxygen produced by photosynthesis provides a favorable living environment for microorganisms. Filter-feeding fish or shellfish introduced can prey on plankton and organic debris in the water, further purifying the water quality and forming a complete food chain, enhancing the self-purification capacity of the entire ecosystem.

[0077] In this invention, biofilm reactors such as MBR (membrane bioreactor) or BAF (biological aerated filter) are also introduced to further improve the quality of the effluent and ensure that it meets or exceeds the standards for farmland irrigation water.

[0078] This invention, through the synergistic effect of the various units of the multi-level biological ecological purification module, can comprehensively and efficiently purify the tailwater at key nodes of farmland runoff, achieve high-standard reuse of tailwater, reduce non-point source pollution emissions, and improve the utilization efficiency of water resources.

[0079] Based on the above embodiments, the method further includes: Based on an unsupervised learning algorithm, data anomaly detection is performed on the multi-source water quality environmental data, and corresponding equipment fault alarm information or data acquisition anomaly alarm information is generated according to the data anomaly detection results.

[0080] In this invention, multi-source water quality environmental data typically encompasses information from multiple dimensions and sources. For example, data collected from water quality sensors at different locations (such as different cross-sections of rivers or different areas of lakes) may include indicators such as pH value, dissolved oxygen content, turbidity, conductivity, and heavy metal concentration. These data are characterized by large volume, high dimensionality, dynamic changes, and the potential presence of noise and outliers. Because the water quality environment is affected by various complex factors (such as changes in the natural environment and human activity disturbances), it is difficult to accurately label all possible normal and abnormal patterns in advance, necessitating the use of unsupervised learning algorithms for processing.

[0081] For example, when monitoring river water quality, if the dissolved oxygen data at a certain cross-section exceeds a threshold in an isolated forest model at a certain moment, the data may be considered anomaly. Alternatively, when monitoring multiple water quality indicators, if K-means clustering reveals that a data point is too far from the center of its cluster, it is marked as an anomaly.

[0082] In this invention, when the unsupervised learning algorithm detects data anomalies, it is necessary to further analyze the correlation between the abnormal data and equipment failure. For example, if the data collected by a water quality sensor continues to show anomalies (such as abnormally large data fluctuations, data values ​​exceeding the normal measurement range of the sensor, etc.), and this anomaly may have a causal relationship with the sensor's working state (such as sensor aging, damage, abnormal power supply, etc.), it can be preliminarily determined that the abnormal data is caused by equipment failure.

[0083] Once it's determined that the data anomaly is caused by equipment malfunction, an equipment malfunction alarm message is generated. Besides equipment malfunction, data acquisition anomalies can also be caused by other factors, such as packet loss during data transmission or environmental interference leading to inaccurate data. By comprehensively analyzing the log information and network status of the data acquisition system, it can be determined whether the data anomaly is caused by a problem in the data acquisition process. For example, if multiple sensors simultaneously show missing or incorrect data within a certain time period, and network monitoring shows data transmission interruptions during that period, it can be determined that there is a data acquisition anomaly.

[0084] This invention utilizes an unsupervised learning algorithm to detect data anomalies and generate corresponding alarm information based on the detection results. This allows for the timely detection of equipment malfunctions and data acquisition anomalies in water quality environmental monitoring, ensuring the normal operation of the monitoring system and data quality.

[0085] Based on the above embodiments, before inputting the multi-source water quality environmental data into the low-hilly mountain area non-point source pollution management model to obtain the target control parameters corresponding to each target water and fertilizer flow regulation device output by the low-hilly mountain area non-point source pollution management model, the method further includes: The multi-source water quality environmental data is preprocessed to obtain preprocessed multi-source water quality environmental data. The preprocessing includes data cleaning, noise reduction, missing value imputation, time series alignment, and spatial interpolation.

[0086] In this invention, multi-source water quality environmental data originates from various types of monitoring equipment, including floating water quality monitoring stations, bionic fish for water quality monitoring, intelligent soil sensor networks, and weather stations. The data collected by these devices is heterogeneous, meaning that the data formats, structures, and types differ. For example, floating water quality monitoring stations may primarily collect water quality indicators such as water temperature, pH value, and dissolved oxygen, storing them in a specific numerical format; bionic fish for water quality monitoring, simulating fish swimming, may collect data including water flow velocity and dynamic changes in water composition, and their data format may differ from that of the monitoring station; intelligent soil sensor networks primarily collect data such as soil moisture and nutrient content; and weather stations collect meteorological data such as air temperature, air pressure, wind speed, and precipitation. Furthermore, this data is massive and multi-source, covering multiple aspects such as water quality, soil, and meteorology, providing rich information for comprehensive analysis of the water environment.

[0087] Data cleaning primarily aims to remove erroneous, duplicate, inconsistent, or incomplete data records from multi-source datasets. Due to the wide range of data sources, various factors can interfere with the data during collection and transmission, leading to errors or anomalies. For example, floating water quality monitoring stations may collect pH values ​​outside the normal range due to sensor malfunction or signal interference; bionic fish used for water quality monitoring may collect duplicate water flow velocity data in complex water flow environments; and smart soil sensor networks may experience duplicate data storage due to communication problems.

[0088] This invention can filter and correct data by setting reasonable rules and thresholds. For numerical data, filtering is performed based on its normal value range, and data exceeding the range is marked as abnormal and corrected or deleted. For duplicate data records, deduplication is performed based on the data's timestamp or other unique identifier. For inconsistent data, such as different devices using different units for the same indicator, a unified unit conversion is required.

[0089] Noise reduction processing is used to remove noise interference from data, improving its accuracy and reliability. Noise may originate from sensor accuracy errors, environmental interference (such as electromagnetic interference, mechanical vibration, etc.), and signal attenuation during data transmission. For example, wind speed data collected by a weather station may experience high-frequency fluctuation noise due to obstruction from surrounding buildings or slight vibrations of the anemometer; soil moisture data collected by a smart soil sensor network may be affected by microbial activity in the soil or transient fluctuations in groundwater, resulting in random noise. In this invention, filtering algorithms can be used to smooth the data and reduce the impact of noise.

[0090] Missing value imputation is used to fill in missing values ​​in data, ensuring data integrity and continuity. During data acquisition, equipment failure, communication interruption, sensor maintenance, and other reasons may lead to some missing data. For example, a floating water quality monitoring station may be unable to collect data for a period of time due to insufficient battery power; when a bionic fish used for water quality monitoring swims out of the signal coverage area, data transmission may be interrupted, resulting in missing values. The presence of missing values ​​will affect subsequent data analysis and modeling, thus requiring imputation. This invention can select an appropriate imputation method based on the characteristics and distribution of the data, including mean imputation, median imputation, nearest neighbor imputation, interpolation imputation, etc.

[0091] Time series alignment is used to align time-attributed data collected from different devices, ensuring that data from the same point in time correspond and facilitating comprehensive analysis and comparison. Since different devices may have different sampling frequencies—for example, a floating water quality monitoring station might collect data every 10 minutes, while a weather station might collect data every 1 minute—this leads to inconsistencies in the timeline of the data. Without time series alignment, data time mismatch issues arise during multi-source data fusion analysis, affecting the accuracy of the analysis results. This invention uses interpolation or resampling methods based on the data's sampling frequency and timestamp to unify the data to the same time interval.

[0092] For spatially distributed data, such as soil moisture data collected by a smart soil sensor network, the limitations of sensor deployment may prevent coverage of all locations within the entire study area. Spatial interpolation can estimate data values ​​at unknown locations based on data from known locations, thus obtaining a continuous spatial data distribution across the entire study area. For example, in a farmland with only a limited number of soil moisture sensors, spatial interpolation can be used to infer the soil moisture conditions at other locations without sensors. In this invention, spatial interpolation methods include inverse distance weighted interpolation and Kriging interpolation.

[0093] After the aforementioned preprocessing operations, including data cleaning, denoising, missing value imputation, time series alignment, and spatial interpolation, the resulting multi-source water quality environmental data exhibits higher quality and consistency. Errors and noise in the data are effectively removed, missing values ​​are appropriately imputed, and the time series and spatial distribution are more orderly and continuous. This provides a reliable foundation for subsequent data analysis, model building, and decision support, helping to more accurately monitor changes in water quality, predict water quality trends, and promptly identify potential environmental problems. Furthermore, the use of distributed storage and real-time stream processing technologies supports high-concurrency data access and processing, ensuring the efficiency and real-time nature of the preprocessing process.

[0094] The non-point source pollution management and wastewater reuse system in low hilly areas provided by the present invention are described below. The non-point source pollution management and wastewater reuse system in low hilly areas described below can be referred to in correspondence with the non-point source pollution management and wastewater reuse method in low hilly areas described above.

[0095] Figure 2 A schematic diagram of the structure of the non-point source pollution management and effluent reuse system in low hilly areas provided by the present invention is shown below. Figure 2 As shown, this invention provides a system for non-point source pollution management and wastewater reuse in low hilly areas, including a multimodal environmental sensing module 201, a pollution control and decision-making module 202, and a wastewater resource utilization module 203. The multimodal environmental sensing module 201 acquires multi-source water quality environmental data corresponding to the target low hilly area based on multimodal intelligent environmental sensing devices. The pollution control and decision-making module 202 inputs the multi-source water quality environmental data into a low hilly area non-point source pollution management model to obtain target control parameters corresponding to each target water and fertilizer flow control device output by the model. Based on the target control parameters, it controls the water and fertilizer application of each target water and fertilizer flow control device. The target water and fertilizer flow control devices are water and fertilizer flow control devices set up in various farmland areas within the target low hilly area. The wastewater resource utilization module 203 purifies the wastewater from key nodes of farmland runoff based on the controlled water and fertilizer application from each target water and fertilizer flow control device and transports the purified wastewater to the farmland areas of the low hilly area.

[0096] Figure 3 The overall architecture diagram of the non-point source pollution management and effluent reuse system in low hilly areas provided by this invention can be referred to. Figure 3 As shown, in this invention, the multimodal environment sensing module 201 (sensing layer + transmission layer) constructs an integrated environmental monitoring network to achieve multi-dimensional data acquisition. The specific components of the sensing devices are as follows: Floating water quality monitoring station: It can monitor nine water quality parameters, including COD, pH, nitrogen and phosphorus, at fixed points and has edge computing capabilities.

[0097] Intelligent soil sensor network: monitors parameters such as soil moisture, soil nutrients and organic matter, and uses a mesh network for self-organized transmission.

[0098] Bionic fish for water quality monitoring: It collects underwater flow field, pollutant concentration distribution and image data through AI autonomous navigation to support pollution source tracing.

[0099] High-precision weather stations: acquire real-time meteorological data such as rainfall, temperature, humidity and radiation to support water and fertilizer demand forecasting.

[0100] This invention enables data backhaul through multi-modal networks such as LoRaWAN / NB-IoT, 5G, and satellite communication, ensuring communication reliability in complex terrains.

[0101] Pollution control and decision-making module 202 (platform layer + application layer) uses AI to achieve intelligent decision-making and precise control. The specific implementation process is as follows: Integrate heterogeneous data from multiple sources (water quality, soil, meteorology, and GIS, etc.) and perform data cleaning, spatiotemporal alignment, and standardization.

[0102] Pollution source tracing and diffusion prediction model: Combining hydrodynamic models and deep learning, it locates pollution sources and predicts diffusion paths.

[0103] Water and fertilizer demand prediction model: Based on reinforcement learning, dynamically optimize the amount and timing of water and fertilizer application to improve utilization efficiency.

[0104] Runoff Coordinated Management Model: A multi-objective optimization algorithm (genetic algorithm / particle swarm optimization) is used to generate the optimal control parameters for equipment such as gate regulation and water and fertilizer application.

[0105] Dynamic control is achieved by sending instructions to the execution layer devices (smart valves / fertilizer pumps / control gates) through the cloud management platform.

[0106] The tailwater resource utilization module 203 (treatment layer + execution layer) constructs a closed-loop system of "storage-purification-reuse", and the specific implementation process is as follows: Ecological ponds / storage tanks are built at runoff confluence points, and integrated liquid level sensors are used to achieve intelligent scheduling.

[0107] The multi-stage purification process includes pre-sedimentation and physical filtration to remove large particulate pollutants; anaerobic / anoxic biological reaction to achieve denitrification; constructed wetlands (canna lilies / reeds, etc.) combined with microbial degradation for further purification; and biofilm reactor (MBR) to ensure that the effluent meets the standards.

[0108] The system uses a smart pipeline network to deliver purified water to farmland, combined with pressure-compensated drip irrigation technology to achieve precise irrigation, and integrates flow sensors and pH regulators to ensure safe water and fertilizer delivery.

[0109] In this invention, the cloud management platform in the platform layer is based on an elastic cloud computing architecture and can provide powerful data storage, processing, analysis, and visualization services. The platform has the following functions: Data visualization dashboard: Provides intuitive, real-time, multi-dimensional data displays, including water quality parameters, soil parameters, meteorological data, water and fertilizer application rates, runoff monitoring data, and pollution load reduction, and supports custom reports and charts.

[0110] Intelligent early warning module: Based on the prediction results of the AI ​​decision engine and preset thresholds, it provides real-time early warnings for potential pollution risks, water and fertilizer imbalances, and equipment failures, and notifies users through multiple channels such as mobile APP, SMS, and email.

[0111] Decision support and task assignment: The AI ​​decision engine generates optimized water and fertilizer management strategies and pollution prevention and control suggestions, which are then issued to intelligent control equipment in the form of operable instructions, and users can manually intervene and adjust them.

[0112] Model Training and Iteration: Provides interfaces for model training and retraining, allowing the system to continuously optimize AI model performance based on new data.

[0113] Multi-user and permission management: Supports multi-level user management and fine-grained permission allocation to ensure data security and stable system operation.

[0114] In this invention, the system can also provide a user interface, including a feature-rich mobile APP and a professional-grade computer operating platform, providing convenient functions such as data viewing, system control, early warning information reception, report generation, system configuration and model parameter adjustment, to meet the needs of different users.

[0115] The non-point source pollution management and tailwater reuse system for low hilly areas provided by this invention acquires multi-source water quality environmental data of the target low hilly area through multi-modal intelligent environmental sensing equipment, and inputs it into the low hilly area non-point source pollution management model to obtain the control parameters of the target water and fertilizer flow regulation equipment, thereby controlling the application of water and fertilizer. After purifying the tailwater at key nodes of farmland runoff, it is transported back to the farmland area, thereby realizing the coordinated, efficient and intelligent management of runoff cascade water and fertilizer, significantly reducing non-point source pollution load and improving water and fertilizer utilization efficiency.

[0116] The system provided in this embodiment of the invention is used to execute the above-described method embodiments. For specific processes and details, please refer to the above embodiments, which will not be repeated here.

[0117] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 4As shown, the electronic device may include: a processor 401, a communications interface 402, a memory 403, and a communication bus 404, wherein the processor 401, the communications interface 402, and the memory 403 communicate with each other through the communication bus 404. The processor 401 can call logic instructions in the memory 403 to execute a method for non-point source pollution management and wastewater reuse in low hilly areas. This method includes: acquiring multi-source water quality environmental data corresponding to the target low hilly area based on a multi-modal intelligent environmental sensing device; inputting the multi-source water quality environmental data into a low hilly area non-point source pollution management model to obtain target control parameters corresponding to each target water and fertilizer flow control device output by the low hilly area non-point source pollution management model; and controlling the water and fertilizer application of each target water and fertilizer flow control device according to the target control parameters, wherein the target water and fertilizer flow control device is a water and fertilizer flow control device set up in each farmland area of ​​the target low hilly area; and purifying the wastewater at key nodes of farmland runoff based on the water and fertilizer application control of each target water and fertilizer flow control device, and transporting the purified wastewater to the farmland area of ​​the low hilly area.

[0118] Furthermore, the logical instructions in the aforementioned memory 403 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0119] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein when the program instructions are executed by a computer, the computer is able to execute the methods for non-point source pollution management and tailwater reuse in low hilly areas provided by the above methods, the method comprising: acquiring multi-source water quality environmental data corresponding to the target low hilly area based on a multi-modal intelligent environmental sensing device; inputting the multi-source water quality environmental data into a low hilly area non-point source pollution management model to obtain target control parameters corresponding to each target water and fertilizer flow control device output by the low hilly area non-point source pollution management model, and controlling the water and fertilizer application of each target water and fertilizer flow control device according to the target control parameters, wherein the target water and fertilizer flow control device is a water and fertilizer flow control device set up in each farmland area in the target low hilly area; purifying the tailwater at key nodes of farmland runoff based on the target water and fertilizer application control of each target water and fertilizer flow control device, and transporting the purified tailwater to the farmland area of ​​the low hilly area.

[0120] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program is implemented to perform the methods for non-point source pollution management and tailwater reuse in low-hilly mountainous areas provided in the above embodiments. The method includes: acquiring multi-source water quality environmental data corresponding to the target low-hilly mountainous area based on a multi-modal intelligent environmental sensing device; inputting the multi-source water quality environmental data into a low-hilly mountainous area non-point source pollution management model to obtain target control parameters corresponding to each target water and fertilizer flow regulation device output by the low-hilly mountainous area non-point source pollution management model; and controlling the water and fertilizer application of each target water and fertilizer flow regulation device according to the target control parameters, wherein the target water and fertilizer flow regulation device is a water and fertilizer flow regulation device set up in each farmland area in the target low-hilly mountainous area; and purifying the tailwater at key nodes of farmland runoff based on the target water and fertilizer application control devices, and transporting the purified tailwater to the farmland area of ​​the low-hilly mountainous area.

[0121] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0122] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0123] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for managing non-point source pollution and reusing effluent in low hilly mountainous areas, characterized in that, include: Based on multimodal intelligent environmental sensing equipment, acquire multi-source water quality environmental data corresponding to the target low hilly area; The multi-source water quality environmental data is input into the low hilly area non-point source pollution management model to obtain the target control parameters corresponding to each target water and fertilizer flow regulation device output by the low hilly area non-point source pollution management model. Based on the target control parameters, water and fertilizer application control is performed on each of the target water and fertilizer flow regulation devices. The target water and fertilizer flow regulation devices are water and fertilizer flow regulation devices set up in each farmland area in the target low hilly area. Based on the target water and fertilizer flow control devices after water and fertilizer application control, the tailwater at key nodes of farmland runoff is purified and transported to the farmland area in the low hilly area.

2. The method for non-point source pollution management and effluent reuse in low hilly areas according to claim 1, characterized in that, The multimodal intelligent environmental sensing equipment includes a floating water quality monitoring station, a bionic fish for water quality monitoring, a soil sensor, and a meteorological data acquisition device. The multimodal intelligent environmental sensing device acquires multi-source water quality environmental data corresponding to the target low hilly area, including: Based on the floating water quality monitoring station, the water quality parameters in the target low hilly area are monitored to obtain the first water quality monitoring parameters, which include chemical oxygen demand, dissolved oxygen concentration, water pH value, ammonia nitrogen content, total nitrogen content, total phosphorus content, water temperature value, water conductivity, and turbidity. Based on the cruising path of the bionic fish for water quality monitoring in the target low hilly area, the second water quality monitoring parameters collected by the bionic fish for water quality monitoring are obtained, wherein the second water quality monitoring parameters include at least chemical oxygen demand, ammonia nitrogen, total phosphorus and heavy metal ion concentration. Based on the soil sensor, soil parameters in the target low hilly area are monitored to obtain soil monitoring parameters, wherein the soil monitoring parameters include at least soil moisture, soil temperature, soil pH, nitrogen, phosphorus and potassium content, organic matter content and soil electrical conductivity. Based on the meteorological data acquisition equipment, meteorological environmental data corresponding to the target low hilly area are acquired, wherein the meteorological environmental data includes at least rainfall, temperature, ambient humidity, wind speed and direction, and solar radiation value; The multi-source water quality environmental data are obtained based on the first water quality monitoring parameters, the second water quality monitoring parameters, the soil monitoring parameters, and the meteorological environmental data.

3. The method for non-point source pollution management and effluent reuse in low hilly areas according to claim 2, characterized in that, The non-point source pollution management model for low hilly areas includes a pollution source tracing and diffusion prediction model, a water and fertilizer demand intelligent prediction model, and a runoff cascade coordinated management and optimization model. The pollution source tracing and diffusion prediction model is constructed based on deep learning algorithms and hydrodynamic models; the water and fertilizer demand intelligent prediction model is constructed based on reinforcement learning or deep learning algorithms; and the runoff cascade coordinated management and optimization model is constructed with the objectives of maximizing runoff pollution reduction, maximizing water and fertilizer utilization efficiency, and minimizing water resource consumption. The process involves inputting the multi-source water quality environmental data into a low-hilly mountain area non-point source pollution management model to obtain target control parameters corresponding to each target water and fertilizer flow regulation device output by the low-hilly mountain area non-point source pollution management model, including: Acquire runoff data, agricultural activity data, geographic information system data, crop types, crop growth stages, historical water and fertilizer use efficiency data, and runoff cascade distribution characteristics data of the target low hilly area; The first water quality monitoring parameter, the second water quality monitoring parameter, the runoff data, the meteorological environment data, the agricultural activity data, and the geographic information system data are input into the pollution source tracing and diffusion prediction model to obtain the pollution source location information and pollution diffusion prediction information of the target low hilly area output by the pollution source tracing and diffusion prediction model. The crop type, crop growth stage, soil monitoring parameters, meteorological environmental data, and historical water and fertilizer use efficiency data are input into the intelligent water and fertilizer demand prediction model to obtain the water and fertilizer demand prediction information of the target low hilly area output by the intelligent water and fertilizer demand prediction model. The water and fertilizer demand prediction information includes water and fertilizer demand prediction data and water and fertilizer application prediction time. The pollution source location information, the pollution diffusion prediction information, the water and fertilizer demand prediction information, and the runoff cascade distribution characteristic data are input into the runoff cascade collaborative management and optimization model to obtain the target control parameters output by the runoff cascade collaborative management and optimization model.

4. The method for non-point source pollution management and effluent reuse in low hilly areas according to claim 1, characterized in that, The aforementioned target water and fertilizer flow regulation devices, based on water and fertilizer application control, purify the tailwater at key nodes of farmland runoff, including: When the tailwater level at the key node of the farmland runoff is determined to meet the preset level height, the tailwater at the key node of the farmland runoff is purified based on a multi-level biological ecological purification module. The multi-level biological ecological purification module includes a pre-sedimentation and physical filtration unit, an anaerobic / anoxic biological reaction unit, an aerobic biological filter bed / constructed wetland unit, an ecological pond / stabilization pond, and a biofilm reactor.

5. The method for non-point source pollution management and wastewater reuse in low hilly areas according to any one of claims 1 to 4, characterized in that, The method further includes: Based on an unsupervised learning algorithm, data anomaly detection is performed on the multi-source water quality environmental data, and corresponding equipment fault alarm information or data acquisition anomaly alarm information is generated according to the data anomaly detection results.

6. The method for non-point source pollution management and wastewater reuse in low hilly areas according to any one of claims 1 to 4, characterized in that, Before inputting the multi-source water quality environmental data into the low-hilly mountain area non-point source pollution management model to obtain the target control parameters corresponding to each target water and fertilizer flow regulation device output by the low-hilly mountain area non-point source pollution management model, the method further includes: The multi-source water quality environmental data is preprocessed to obtain preprocessed multi-source water quality environmental data. The preprocessing includes data cleaning, noise reduction, missing value imputation, time series alignment, and spatial interpolation.

7. A system for managing non-point source pollution and reusing effluent in low hilly areas, characterized in that, include: The multimodal environment sensing module is used to acquire multi-source water quality environmental data corresponding to the target low hilly area based on multimodal intelligent environment sensing equipment; The pollution control and decision-making module is used to input the multi-source water quality environmental data into the low hilly area non-point source pollution management model, obtain the target control parameters corresponding to each target water and fertilizer flow regulation device output by the low hilly area non-point source pollution management model, and control the water and fertilizer application of each target water and fertilizer flow regulation device according to the target control parameters. The target water and fertilizer flow regulation device is the water and fertilizer flow regulation device set up in each farmland area in the target low hilly area. The tailwater resource utilization module is used to purify the tailwater at key nodes of farmland runoff based on the target water and fertilizer flow regulation devices after water and fertilizer application control, and to transport the purified tailwater to the farmland area in the low hilly area.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the method for non-point source pollution management and wastewater reuse in low hilly areas as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for non-point source pollution management and wastewater reuse in low hilly areas as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for non-point source pollution management and wastewater reuse in low hilly areas as described in any one of claims 1 to 6.

Images