Event-driven soil pollution flux monitoring system and method

Abstract

The embodiment of the present application provides a kind of based on event-driven soil pollution flux monitoring system and method, it is related to the technical field of soil pollution monitoring technique.The method includes: the physical parameter in the target soil region is acquired in real time, and based on the change of physical parameter, the occurrence of environmental event is determined;In response to the determination of environmental event, robot platform is navigated to the region where environmental event occurs;Through robot platform, in the region where environmental event occurs, continuous interception extraction is carried out to in-situ soil pore water, to obtain environmental sample flow;Flow injection analysis is carried out to environmental sample flow, to determine the real pollutant concentration of the environmental sample flow;Based on real pollutant concentration, the flow velocity of environmental sample flow and the preset probe effective capture section area, the instantaneous mass flux of pollutant in the target soil region is determined.Through the present application, the problem of low soil pollution detection precision is solved, and the effect of improving soil pollution monitoring precision is achieved.

Description

Technical Field

[0001] The present invention relates to the field of soil pollution monitoring technology, and more specifically, to an event-driven soil pollution flux monitoring system and method. Background Technology

[0002] Soil, as a crucial component of the ecosystem, directly impacts agricultural safety, groundwater quality, and the human living environment. However, industrial activities, agricultural practices, and accidental leaks frequently expose soil to pollutants such as heavy metals, pesticides, and organic matter. Therefore, effective monitoring of soil pollution is a fundamental prerequisite for risk assessment, pollution source tracing, and remediation. Currently, land consolidation assessments lack precise evaluation methods based on high-throughput analytical testing and quantitative models.

[0003] Current soil pollution monitoring methods primarily rely on timed, location-based, discrete manual sampling, sending collected soil samples to laboratories for analysis. This method has extremely low temporal resolution, while the migration and release of pollutants in soil are often closely coupled with sudden environmental events such as rainfall, irrigation, and abrupt temperature changes. For example, a rainstorm can leach a large amount of surface-adsorbed pollutants within hours, causing them to rapidly migrate to lower soil layers or groundwater, forming a high-intensity, short-duration pollution flux peak. Traditional weekly or monthly sampling monitoring models are highly likely to completely miss such critical instantaneous processes, leading to a severe underestimation of pollutant migration risks. This fails to provide timely and effective data support for precise pollution containment and emergency remediation, easily creating time blind spots for instantaneous pollution processes. Summary of the Invention

[0004] This invention provides an event-driven soil pollution flux monitoring system and method to at least address the problem of low accuracy in soil pollution flux monitoring in related technologies.

[0005] According to an embodiment of the present invention, an event-driven method for monitoring soil pollution fluxes is provided, comprising: The physical parameters within the target soil area are acquired in real time, and the occurrence of an environmental event is determined based on the changes in the physical parameters; the physical parameters include at least one of soil volumetric water content, soil electrical conductivity, and soil temperature. In response to the determination of the environmental event, the robot platform is dispatched to navigate to the area where the environmental event occurred; Using the robotic platform, in-situ soil pore water is continuously intercepted and extracted within the area where the environmental event occurred to obtain environmental sample streams. Flow injection analysis is performed on the environmental sample stream to determine the true pollutant concentration of the environmental sample stream. The flow injection analysis includes dynamic volume correction of the measured apparent pollutant concentration based on the monitored flow rate of the environmental sample stream and the flow rate of the simultaneously injected analytical reagent. Based on the actual pollutant concentration, the flow rate of the environmental sample stream, and the preset effective capture cross-sectional area of ​​the probe, the instantaneous mass flux of pollutants in the target soil area is determined.

[0006] In one exemplary embodiment, determining the occurrence of an environmental event based on the change in the physical parameters includes: Collect the physical parameters and construct an instantaneous state vector; The Mahalanobis distance of the instantaneous state vector is obtained based on the historical baseline state vector mean and historical state covariance matrix. The Mahalanobis distance is compared with a preset trigger threshold. If the Mahalanobis distance is greater than the trigger threshold, the occurrence of the environmental event is determined.

[0007] In one exemplary embodiment, the method further includes: Based on peak flux, decay time constant, and delay time of environmental event occurrence, a pollution event feature tensor is constructed; Based on the pollution event feature tensor, the pollution release behavior of the target soil area is classified.

[0008] In one exemplary embodiment, determining the instantaneous mass flux of pollutants within the target soil region further includes: Obtain the total mixing volume flow rate; The equivalent volumetric flow velocity of the directional electroosmotic flow is calculated based on the electric field strength of the applied constant DC electric field, the preset effective capture cross-sectional area of ​​the probe, and the electroosmotic permeability of the target soil area. The actual soil hydrological seepage velocity is determined based on the total mixing volume velocity and the equivalent volume velocity. The instantaneous mass flux is determined based on the actual soil hydrological seepage velocity.

[0009] According to another embodiment of the present invention, an event-driven soil pollution flux monitoring system is provided, comprising: A node sensor network is used to monitor preset physical parameters within a target soil area in real time, and to identify the occurrence of an environmental event based on changes in the physical parameters. A scheduling and navigation gateway is used to schedule a robot platform in response to the identification of the environmental event, so that it can autonomously navigate to the area where the environmental event occurred; A robotic platform is used to continuously intercept and extract pore water in the soil within the area where the environmental event occurs to obtain a continuously flowing environmental sample stream; flow injection analysis is performed on the environmental sample stream to determine the true pollutant concentration of the environmental sample stream, wherein the flow injection analysis includes dynamic volume correction of the measured apparent pollutant concentration based on the monitored flow rate of the environmental sample stream and the flow rate of the synchronously injected analytical reagent. The flux analysis platform is used to determine the instantaneous mass flux of pollutants in the target soil area based on the actual pollutant concentration, the flow rate of the environmental sample stream, and the preset effective capture cross-sectional area of ​​the probe.

[0010] In one exemplary embodiment, the node sensor network determines the occurrence of an environmental event based on changes in the physical parameters by including: Collect the physical parameters and construct an instantaneous state vector; The Mahalanobis distance of the instantaneous state vector is obtained based on the historical baseline state vector mean and historical state covariance matrix. The Mahalanobis distance is compared with a preset trigger threshold. If the Mahalanobis distance is greater than the trigger threshold, the occurrence of the environmental event is determined.

[0011] In one exemplary embodiment, the system further includes: The tensor construction module is used to construct pollution event feature tensors based on flux peak, decay time constant, and the delay time of environmental event occurrence. The classification module is used to classify the pollution release behavior of the target soil area based on the pollution event feature tensor.

[0012] In one exemplary embodiment, the flux resolution platform is further configured to: Obtain the total mixing volume flow rate; The equivalent volumetric flow velocity of the directional electroosmotic flow is calculated based on the electric field strength of the applied constant DC electric field, the preset effective capture cross-sectional area of ​​the probe, and the electroosmotic permeability of the target soil area. The actual soil hydrological seepage velocity is determined based on the total mixing volume velocity and the equivalent volume velocity. The instantaneous mass flux is determined based on the actual soil hydrological seepage velocity.

[0013] According to yet another embodiment of the present invention, a computer-readable storage medium is also provided, wherein a computer program is stored therein, wherein the computer program is configured to perform the steps in any of the above method embodiments when executed.

[0014] According to yet another embodiment of the present invention, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0015] This invention quantifies the peak mass flux of real pollutants and its decay process during open-environment hydrological scouring by using flow injection analysis and dynamic volume correction mechanism. It solves the problem of low accuracy in soil pollution monitoring by addressing the shortcomings of traditional time-based sampling methods in terms of temporal resolution. Therefore, it can solve the problem of low accuracy in soil pollution monitoring and achieve the effect of improving the accuracy of soil pollution monitoring. Attached Figure Description

[0016] Figure 1 This is a structural block diagram of an event-driven soil pollution flux monitoring system according to an embodiment of the present invention. Figure 2 This is a flowchart of an event-driven soil pollution flux monitoring method according to an embodiment of the present invention; Figure 3 This is a diagram illustrating the effect of changes in mass flux according to an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0018] In the following description, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0019] Furthermore, in this application, directional terms such as "upper," "lower," "left," and "right" may be defined relative to the orientation of the components shown in the accompanying drawings. It should be understood that these directional terms can be relative concepts, used for relative description and clarification, and may change accordingly depending on the orientation of the components in the accompanying drawings.

[0020] In this application, unless otherwise expressly specified and limited, the term "connection" should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral part; it can be a direct connection or an indirect connection through an intermediate medium. Furthermore, the term "coupled" can refer to an electrical connection that enables signal transmission.

[0021] As used herein, “about,” “approximately,” or “approximately” includes the stated value and the average value within an acceptable range of deviation from the given value, wherein the acceptable range of deviation is determined by a person skilled in the art taking into account the measurement under discussion and the error associated with the measurement of the given quantity (i.e., the limitations of the measurement system).

[0022] Example 1

[0023] This embodiment provides an event-driven method for monitoring soil pollution fluxes. Figure 1 This is a flowchart of an event-driven soil pollution flux monitoring method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: This embodiment provides an event-driven soil pollution flux monitoring system. Through the collaboration of a node sensor network, a scheduling and navigation gateway, a robot platform, and a flux analysis platform, this system can achieve real-time capture and quantitative analysis of the pollutant migration process caused by sudden environmental events in the target soil area. It solves the technical problem in the prior art that the peak value of instantaneous pollution flux is missed due to the use of low-frequency discrete sampling, and achieves the beneficial effect of improving the accuracy of monitoring dynamic processes of soil pollution.

[0024] Reference Figure 1 The event-driven soil pollution flux monitoring system includes: a node sensor network 100, a scheduling and navigation gateway 200, a robot platform 300, and a flux analysis platform 400.

[0025] Among them, the node sensor network 100 is used to monitor preset physical parameters within the target soil area in real time, and to identify the occurrence of environmental events based on changes in physical parameters.

[0026] The node sensor network 100 consists of a large number of independent sentinel nodes deployed in the target soil area. Each sentinel node 110 can be an integrated device in the form of a nail, with its upper part exposed on the ground surface, integrating a small monocrystalline silicon solar panel for energy harvesting and a LoRa antenna for wireless communication; its lower part is a probe inserted into the soil, which encapsulates sensing units for monitoring physical parameters. Preferably, the set of physical parameters constitutes a three-dimensional state vector, which includes soil volumetric water content, soil conductivity, and soil temperature. Soil volumetric water content can be measured using a composite probe that can invert the dielectric constant by measuring the change in the propagation constant of electromagnetic waves in the soil medium based on the principle of frequency domain reflection, and then convert it into water content. Soil conductivity is measured using two or four ring electrodes on the same probe. Soil temperature can be measured using a high-precision platinum resistance thermometer (e.g., PT1000).

[0027] Each sentry node 110 also integrates an ultra-low-power microcontroller unit (MCU), such as the STM32L4 series microcontroller based on the ARM Cortex-M4 core. The MCU performs edge computing tasks, specifically analyzing the acquired physical parameter sequences locally on the node to identify environmental events. To achieve extreme power control, the MCU remains in deep sleep mode most of the time, only being periodically woken up by the real-time clock (RTC) to perform one data acquisition and analysis cycle. The preset heartbeat period for this periodic wake-up can be set to 10 minutes.

[0028] The node sensor network 100 and the dispatch and navigation gateway 200 communicate via long-range low-power wide area network (LPWAN) technology. Preferably, this communication technology is LoRaWAN. Sentinel node 110, acting as a terminal device, initiates an uplink data transmission as a Class A device after detecting an environmental event, sending an alarm packet containing event information to the LoRaWAN gateway, which then forwards it to the dispatch and navigation gateway 200 via the IP network.

[0029] The dispatching and navigation gateway 200 can be a high-performance edge computing server deployed at the edge of the target area (e.g., a server room or monitoring station next to farmland). This server receives alarm packets from the node sensor network 100 and, in response to the identification of the environmental event, dispatches the robot platform 300. The dispatching process includes parsing the alarm packets, extracting the unique identifier of the sentinel node that triggered the event, and its pre-calibrated GPS geographic coordinates. Subsequently, the dispatching and navigation gateway 200 uses these coordinates as the navigation target point and runs a path planning algorithm to generate a driving path for the robot platform 300. The path planning algorithm can be an A* search algorithm or a Dijkstra algorithm, which is based on a pre-built high-precision digital terrain model (DSM) containing information such as the terrain undulations, ditches, ecological buffer zones, and other remediation facilities or fixed obstacles in the target area.

[0030] The chassis of the robot platform 300 can be a four-wheel independent drive differential chassis or a tracked chassis adapted to complex terrain, and it is controlled at the underlying level based on the robot operating system ROS. The platform receives navigation commands from the scheduling and navigation gateway 200 and utilizes its integrated navigation system (e.g., a fusion of GPS / RTK and IMU) to achieve high-precision autonomous navigation, ultimately reaching the area where the event occurred. The platform then continuously intercepts and extracts pore water from the in-situ soil and performs online flow injection analysis on the acquired continuously flowing environmental sample stream.

[0031] The flux analysis platform 400 can be a cloud server or a local data center. This platform receives high-frequency analysis data, specifically the instantaneous mass flux time series of pollutants, uploaded from the robotic platform 300 via a wireless network (e.g., 5G or Wi-Fi). The platform is used for further mathematical modeling and data mining of this time series. For example, it can classify pollution release behavior in different regions using clustering algorithms, or fit the half-life of pollutants using decay models. Ultimately, it generates a visualized spatiotemporal heat map of pollution flux and an analysis report, providing support for subsequent ecological technology applicability assessments and investment decisions.

[0032] Example 2 Reference Figure 2 The event-driven soil pollution flux monitoring method provided in this embodiment includes the following steps: S100: Through the node sensor network 100, preset physical parameters within the target soil area are monitored in real time, and the occurrence of environmental events is determined based on changes in the physical parameters.

[0033] In the initial stage of system deployment, system initialization and baseline calibration must be performed first. During this time, all sentinel nodes 110 continuously run for a preset time period—for example, 14 calendar days—during which time they continuously record the soil volumetric water content at their location. Soil electrical conductivity and soil temperature The data includes diurnal periodic fluctuations. This data is used in the local memory of each node to learn and build a multivariate statistical model describing the normal or background state of that location. Preferably, this statistical model can be a three-dimensional Gaussian distribution model, whose parameters include a three-dimensional state vector mean. and covariance matrix .

[0034] Simultaneously, it is also necessary to calibrate the effective capture cross-sectional area of ​​the intercepting probe mounted on the robot platform 300 under different soil textures. The calibration method here is as follows: Construct a known size (e.g., A standard sandbox filled with the target area whose porosity needs to be measured. The original soil was used as a sandbox, and a full-area leachate collection tray was installed at the bottom of the sandbox to measure the theoretical global downward seepage velocity. A stable seepage field was established by simulating uniform rainfall of known intensity (e.g., 50 mm / h) applied to the surface of the sandbox. A robotic interceptor probe was inserted to the target depth at the center of the sandbox, and a constant suction negative pressure (e.g., -20 kPa) was applied, identical to that used in actual operation. The steady flow rate obtained by continuous suction from the probe was measured using a high-precision flow meter. By combining the known overall hydraulic gradient of the sandbox and the measured global seepage velocity, and using Darcy's law or the HYDRUS-1D numerical model, the equivalent cross-sectional area of ​​soil moisture actually intercepted by the probe under a specific negative pressure can be calculated in reverse. This equivalent cross-sectional area is the effective capture cross-sectional area. .

[0035] During the routine monitoring phase, the microcontroller unit of each sentinel node 110 wakes up from deep sleep at each preset heartbeat cycle (e.g., 10 minutes) to perform edge event recognition and alarm triggering.

[0036] Specifically, the microcontroller unit first collects the physical parameters at the current moment to form an instantaneous state vector. Subsequently, to quantify the degree to which the current state deviates from the normal baseline and effectively suppress false alarms caused by drift from a single sensor or isolated noise points, the microcontroller unit employs an edge event-triggered model. This model calculates the current state vector... To the center of the reference Gaussian distribution The square of the Mahalanobis distance To achieve quantification and suppress false alarms:

[0037] The distance calculation takes into account the correlation between different physical parameters, which is determined by the covariance matrix. Description. For example, soil moisture content during a normal rainfall event. An increase in conductivity is usually accompanied by an increase in electrical conductivity. The increase and temperature The slight decrease in variance, this multivariate collaborative change pattern is encoded in the covariance matrix. Therefore, Mahalanobis distance is more robust in identifying genuine environmental anomalies than simple Euclidean distance.

[0038] The square of the Mahalanobis distance is calculated. Then, the microcontroller unit compares it with a preset trigger threshold. A comparison is then made. To ensure that the threshold is set statistically significant, rather than relying solely on experience, it is defined as a chi-square distribution with 3 degrees of freedom. The 99th percentile, where 3 degrees of freedom correspond to the three dimensions of the state vector. Consulting statistical tables, the specific value of this threshold is 11.34. This threshold is applied if and only if the calculated... When the distance is greater than 11.34, the system can statistically consider that the current state has more than 99% confidence that it does not belong to normal background fluctuations, that is, a significant environmental event has occurred. In order to further filter out transient interference, a sliding window logic can be set, that is, only when the Mahalanobis distance exceeds the trigger threshold for three consecutive heartbeat cycles will the occurrence of the event be finally confirmed, and an alarm packet will be generated and sent to the scheduling and navigation gateway 200 through the LoRa network.

[0039] For example, suppose that the background mean vector learned by a sentinel node N_203 during the initialization phase is... And the corresponding covariance matrix was obtained. At 2:00 PM one afternoon, a sudden downpour caused the instantaneous state vector measured at this node to be... The node MCU substitutes this vector into the Mahalanobis distance formula for calculation. Because both water content and conductivity have shown significant positive co-movement, while temperature has shown a slight negative co-movement, this pattern of change is significantly different from historical normal fluctuations, leading to errors in the calculated values. The value was 35.6, which is significantly higher than the trigger threshold of 11.34. Furthermore, between 14:10 and 14:20, the node's measurement remained high, and the calculated Mahalanobis distance consistently exceeded the threshold. Therefore, at 14:20, the node officially confirmed the environmental event and sent an alarm packet to the gateway.

[0040] S200: In response to the recognition of environmental events, it dispatches the robot platform to enable it to autonomously navigate to the area where the environmental event occurred.

[0041] When the dispatch navigation gateway 200 receives an alarm packet from the sentinel node 110 through the LoRaWAN network, it immediately starts the dispatch procedure.

[0042] First, the dispatch navigation gateway 200 parses the alarm packet and extracts two key pieces of information: the unique identifier of the sentry node that triggered the alarm, and the instantaneous state vector carried in the packet. By querying the node information table pre-installed in the local database, the dispatching and navigation gateway 200 can uniquely determine the geographical location of the event, i.e., the GPS coordinates of the node, based on the node ID. ).

[0043] After obtaining the coordinates of the target navigation point, the core task of the scheduling navigation gateway 200 is to plan a route for the robot platform 300 from its current location (e.g., the location of the charging base station). The optimal, collision-free driving path to the target point is determined. To this end, the scheduling and navigation gateway 200 invokes its internal path planning module. This module performs scheduling calculations based on a pre-constructed high-precision digital terrain model (DSM). This DSM not only contains detailed elevation information of the target area but also marks all known static obstacles, such as ditches, field ridges, buildings, and trees. Preferably, the path planning algorithm employs the A* search algorithm, which is a heuristic search algorithm that evaluates a function... To determine the next node to visit, where This refers to the actual costs, such as the distance traveled from the starting point to the current node. This is a heuristic estimate of the cost from the current node to the destination, such as the Euclidean distance or Manhattan distance. By running the A* search algorithm on a grid map constructed by the DSM, an optimal path that comprehensively considers distance, terrain slope, and obstacle avoidance can be found efficiently.

[0044] After path planning is completed, the scheduling and navigation gateway 200 generates a waypoint sequence consisting of a series of consecutive GPS waypoints. This sequence is encapsulated into a navigation command and sent to the robot platform 300 via high-bandwidth wireless communication methods such as Wi-Fi or 5G. Upon receiving the command, the robot platform 300's underlying ROS motion control nodes parse the waypoints one by one and, combined with the real-time pose information provided by its own GPS / RTK and IMU integrated navigation system, precisely follow the path to the target point by controlling the differential speed of the left and right wheels or tracks.

[0045] For example, the scheduling and navigation gateway 200 receives an alarm from node N_203 (coordinates (30.65°N, 104.06°E)), indicating that the robot is currently docked at a base station (coordinates (30.64°N, 104.05°E)). The scheduling and navigation gateway 200's A* search algorithm searches the DSM map and finds an insurmountable irrigation canal on the straight path between the two points. The algorithm automatically detours, planning a slightly longer but safe and reachable path consisting of 15 GPS waypoints. The scheduling and navigation gateway 200 then sends the coordinate sequence of these 15 waypoints to the robot. After the robot starts, its navigation system continuously compares its real-time position with the first waypoint, adjusting the motor speed through a PID controller to reduce positional errors. After reaching the first waypoint, it uses the second waypoint as the target, and so on, until it finally accurately reaches the event area where node N_203 is located.

[0046] S300: Using a robotic platform, it continuously intercepts and extracts pore water in the soil within the area where an environmental event occurs, in order to obtain a continuous flow of environmental sample streams.

[0047] This method replaces the inefficient, non-in-situ method of manually excavating soil samples with online continuous high-frequency sampling to directly capture and measure the flux of pollutants that are migrating in the soil driven by environmental events. Specifically, it can be further decomposed into three closely coupled sub-steps: S310, S320, and S330.

[0048] S310: In-situ leachate interception and extraction.

[0049] Upon reaching the target area, the robot platform 300 uses its onboard robotic arm to vertically insert the in-situ pore water interception probe to a preset target depth. The probe's outer shell is preferably made of inert materials such as polytetrafluoroethylene, and its end is wrapped with a layer of porous sintered metal or hydrophilic ceramic material with a specific pore size of 0.45 micrometers. The porous material acts as a physical filter, allowing environmental samples such as pore water in the soil to pass through while preventing soil solid particles from entering, thereby avoiding blockage of subsequent analysis pipelines.

[0050] Once the probe is in place, a miniature vacuum diaphragm pump inside the robot starts. This pump is connected to the probe via a flexible tube, and during operation, it creates a stable and controllable negative pressure of -20 kPa inside the probe. Under this negative pressure, pore water in the soil around the probe—especially saturated or unsaturated gravity flows caused by events such as rainfall—is continuously drawn into the probe and transported through the tube to the robot's onboard analysis chamber, thus forming a continuous flow of environmental samples.

[0051] To achieve accurate flux calculation, a high-precision miniature liquid flow meter is connected in series in the suction line. This flow meter can be based on the thermal principle or the Coriolis principle, and it can output the instantaneous volumetric flow rate of the environmental sample flow with a response time of seconds or even sub-seconds. Its unit is liters per second (L / s).

[0052] For example, such as Figure 3 As shown, in the early stages of a rainstorm, the soil is not yet fully saturated, and the seepage rate is slow. The real-time environmental sample flow rate measured by the flow meter of the robot probe may be... As rainfall continues, soil saturation increases, and the seepage velocity driven by gravitational potential energy accelerates. The flow meter reading may increase linearly or non-linearly, reaching its peak at the height of rainfall. .

[0053] The robot platform 300 then employs an open flux measurement model to perform online volume correction, addressing the issue of misinterpreting chemical leaching flux within a closed container with hydrological flushing flux in an open environment. This model restores the apparent concentration measured in the robot's internal electrochemical flow cell, which is affected by analytical reagent dilution, to the "real environmental pore water concentration" entering the robot's probe. Combined with the actual intercepted environmental sample flow rate, the model ultimately calculates the true environmental mass flux per unit area.

[0054] Specifically, firstly, in sub-step S310, the robot's in-situ interception probe is inserted into the soil. An internal miniature vacuum diaphragm pump operates to generate a constant negative pressure, continuously drawing saturated or unsaturated pore water from the surrounding soil, generated by events such as rainfall, into the pipeline, thus forming a continuous flow of environmental sample. Subsequently, a miniature liquid flow meter outputs the actual environmental sample flow rate in real time. The unit is .

[0055] In sub-step S320, because the original ionic strength, pH value, etc. of the environmental sample stream may not be suitable for direct electrochemical measurement, the environmental sample stream will not directly enter the electrochemical analysis unit. Therefore, another high-precision micro-injection pump driven by a stepper motor will inject the sample at a preset constant flow rate. An analytical reagent or supporting electrolyte buffer is simultaneously injected into the sample stream, and the mixed liquid then enters an electrochemical flow cell. This flow cell integrates a multi-channel working electrode, a counter electrode, and a reference electrode. A potentiostat applies an optimized potential scan waveform to the flow cell and records the current signal at the dissolution peak. This signal is then converted into the apparent contaminant concentration of the mixture using a pre-calibrated working curve. The unit is .

[0056] In substep S330, the robot's onboard edge computing unit receives three real-time inputs: , and known According to the law of conservation of mass, at the mixing point, the mass of pollutants flowing in per unit time must be equal to the mass of pollutants flowing out, that is:

[0057] This equation can be used to solve for the concentration of undiluted, real-world pore water:

[0058] Once the true pore water concentration is obtained, the true environmental mass flux can be calculated, which is the amount of substance passing through a unit area per unit time.

[0059] in, It is the effective capture cross-sectional area of ​​the probe, pre-calibrated through a standard sandbox experiment, and its unit is... The aforementioned Substituting, we get:

[0060] For example, in the aforementioned rainstorm event, when the rainfall reached its peak, the flow meter of the robot platform 300 measured the actual soil seepage velocity. Meanwhile, the infusion pump operates at a constant rate After injecting buffer solution, the apparent cadmium ion concentration of the mixture measured by the electrochemical flow cell was [value missing]. The robot's probe has a pre-calibrated effective capture cross-sectional area of ​​[area value missing]. The onboard computing unit then substitutes these values ​​into the equation to obtain the instantaneous mass flux at that moment. After conversion, it is approximately equal to This means that at the peak of the rainstorm, each square meter of soil was losing cadmium pollutants to the lower layers at a rate of 4 milligrams per minute. If the monitoring target is new pollutants such as perfluorinated compounds specified in the project plan, the flux unit can be converted accordingly. The magnitude and calculation principles are exactly the same, so I will not repeat them here.

[0061] S400: Based on time series data of instantaneous mass flux, perform flux decay modeling and pollution behavior classification.

[0062] The throughput analysis platform 400 performs in-depth spatiotemporal analysis and data mining on the large amount of high-frequency throughput data uploaded by the robot platform 300 throughout the entire event monitoring period (which may last for several hours). Specifically: S410: Flux Decay Modeling and Feature Extraction.

[0063] The flux analysis platform 400 first receives and integrates instantaneous mass flux time series data from the robot platform. In order to extract key parameters that characterize the core dynamic features of the pollution event from this complex curve, the platform uses mathematical modeling.

[0064] First, the platform automatically identifies the peak points in the flux curve and records the peak flux. And the moment of reaching the peak Then determine the moment when the event was first identified by the sentinel node. and The time difference is the event delay time. .

[0065] Subsequently, the platform extracts the data from the descending segment of the flux curve starting from the peak point.

[0066] Since many natural decay processes (such as gravity-driven solute leaching) follow an exponential decay law, this application assumes that the leaching decay process of pollutants also approximates this. Therefore, the platform uses the Levenberg-Marquardt algorithm to fit the descent data into a standard single exponential decay model:

[0067] in, It is the background baseline value after the event ends, when flux gradually recovers and tends to stabilize.

[0068] By iteratively fitting and minimizing the sum of squared residuals between the measured values ​​and the model predictions, the decay time constant can be solved. This constant represents the rate at which flux decays from its peak to its minimum. (Approximately 36.8%) the required time span, The larger the value, the more firmly the pollutants are adsorbed and intercepted by the soil or ecological ditches, buffer zones, and the slower they are leached away. Conversely, the smaller the value, the stronger the pollutant's mobility.

[0069] S420: Clustering of pollution behavior based on feature tensors.

[0070] For each pollution event monitored by the robot, a three-dimensional feature tensor can be extracted. ,in It represents the absolute intensity of pollution loss. This represents the response speed of the soil hydrological system to the release of pollutants, while This represents the system's physicochemical retention capacity for pollution. When the system detects multiple events in different areas (e.g., plots where different remediation technologies have been implemented), the flux analysis platform 400 collects the feature tensors of all events to form a comprehensive dataset to be analyzed.

[0071] To automatically classify soil pollution behavior patterns in different regions and thus assess the applicability of ecological technologies, the platform employs the K-Means++ clustering algorithm. The algorithm's execution process is as follows: First, select features in the feature space according to the K-Means++ strategy. We first generate an initial cluster center, and then iteratively assign the feature tensor of each event to the cluster to which the nearest center belongs by calculating the Euclidean distance. We then take the mean of all tensors in each cluster to recalculate the center of that cluster. This process is repeated until the offset of all center positions is below a very small convergence threshold.

[0072] After clustering is completed, each cluster represents a specific type of pollution release behavior.

[0073] For example, one cluster might exhibit high peak value, short delay, and short decay, typically corresponding to highly mobile pollutant events in topsoil sandy soil without effective intervention, and could be labeled as a high-velocity penetration zone. Another cluster might exhibit medium peak value, long delay, and long decay, potentially corresponding to highly adsorbed pollutant events in soil with organic fertilizer amendments or heavy metal passivation measures, and could be labeled as an effective pollution retention zone. By mapping these clustering labels and their corresponding quantitative characteristic indicators back to the GIS geospatial system, the platform can ultimately generate a high-resolution "Event-Driven Spatiotemporal Thermodynamics and Effectiveness Assessment Report of Soil Pollution Flux," visually demonstrating the remediation effectiveness in different areas and providing scientific data support for evaluating the applicability of measures such as comprehensive farmland remediation and ecological ditch construction.

[0074] For example, the platform collected data on 100 rainfall events occurring within a comprehensive land consolidation project area. The K-Means++ algorithm was then run (with settings...). Afterwards, the data was divided into three clusters. Cluster A, with average characteristics of [high, short, short], corresponds to bare areas without plant buffer zones and is marked in red, indicating a high risk of water runoff. Cluster B, with average characteristics of [medium, medium, long], corresponds to areas with deep tillage but without passivation agents and is marked in yellow, indicating a medium risk. Cluster C, with average characteristics of [low, long, extremely long], corresponds to demonstration areas with ecological ditches and high-intensity plant buffer zones. This area shows significant interception effects on pollutants and is marked in green, indicating a safe containment zone.

[0075] Example 3 In arid or high-organic-matter remediation sites, early, weak pollution fluxes cannot be effectively captured by electrochemical sensors due to low flow rates and adsorption masking. Therefore, unlike Example 2, this example uses an in-situ electro-acoustic co-intercepting probe on the robot platform 300, along with a corresponding electroosmotic fluid dynamics decoupling algorithm. This allows the system to overcome the strong physicochemical adsorption of new pollutants such as perfluorinated compounds by organic matter in the soil under conditions of extremely low soil moisture content and weak natural gravity seepage at the beginning of rainfall. By using artificial electroosmosis to obtain a sufficient volume of pore water sample, the system can lower the detection limit and provide early warning of pollution release trends.

[0076] Once the robot platform 300 arrives at the event area, the following hierarchical steps are executed: S311: High-frequency acoustic wave desorption.

[0077] The robot's main control unit acquires the current soil temperature and baseline moisture content, and uses this information to call the pre-set acoustic attenuation compensation table in memory to generate ultrasonic excitation parameters. Subsequently, the robotic arm inserts the in-situ electro-acoustic co-intercepting probe into the target soil depth. The probe includes a central titanium alloy capillary, an annular piezoelectric transducer array surrounding it, and an outer conductive metal mesh on the outermost layer. The central titanium alloy capillary serves as the extraction channel and electric field cathode, while the outer conductive metal mesh serves as the electric field anode. The onboard signal generator outputs a high-frequency alternating voltage signal to the annular piezoelectric transducer array.

[0078] For example, the signal frequency is set to 40kHz, the power to 20W, and the duration of excitation to 15 seconds. 40kHz is a low-frequency ultrasonic wave, which generates larger cavitation bubbles and produces the strongest mechanical shear force upon collapse, making it ideal for breaking the physical encapsulation and weak chemical bonds between macromolecular soil organic matter / humic acid and the target material. The short 15-second pulse prevents the probe from evaporating and scaling due to continuous heating. The high-frequency vibration is transmitted through the probe housing to the surrounding solid-liquid interface, generating a strong ultrasonic cavitation effect in the microscale water body within the soil capillary pores. The instantaneous generation and collapse of local microbubbles produce extremely high local temperatures and micro-jet shear forces, thereby forcibly disrupting the hydrogen bonds and van der Waals forces formed between the large molecules of commercial organic fertilizer or humic substances applied extensively in comprehensive land remediation and new perfluorinated pollutants. Subsequently, the bound pollutants, originally firmly adsorbed on the surface of solid particles, are forcibly released into the free liquid phase, generating high-concentration transient pore water microregions, thus eliminating the pollution signal masking effect against a high organic matter background.

[0079] S312: Directional electroosmotic flow artificial pumping and interception.

[0080] The system receives energy from micro-regions of pore water in a desorbed state, as well as from the robot's onboard programmable DC power module. The robot's control node then switches circuits to apply a constant DC electric field between the external conductive metal mesh (anode) and the central titanium alloy capillary (cathode) of the probe. The voltage intensity of this field is set to 2 volts per centimeter. Driven by the DC electric field, due to the generally negatively charged surface of soil particles, hydrated cations in the pore water carry water molecules and move directionally towards the cathode, forming electroosmotic flow. Simultaneously, a micro-vacuum diaphragm pump is activated at a set low negative pressure of -5 kPa. This then provides a stable total mixing volumetric flow rate to the microfluidic mixing chip. The sample flow ensures sufficient injection volume for subsequent electrochemical analysis.

[0081] Under the aforementioned forced extraction mechanism, the flow meter measures... The seepage velocity, which is not a true reflection of the natural environment, would lead to a significant overestimation of the flux if used directly in flux calculations. Therefore, it is essential to remove artificially introduced velocity increments.

[0082] The total mixing volume velocity measured by the airborne microfluidic flowmeter here is:

[0083] in, Real soil hydrological seepage velocities driven by environmental rainfall events; The equivalent volumetric velocity generated by an artificially applied DC electric field.

[0084] The macroscopic electroosmotic flow derived from the Helmholtz-Smoluchowski equation is as follows:

[0085] in, The electroosmotic permeability of the target soil, i.e., the seepage velocity produced per unit electric field strength, is expressed in units of... or , Represents electric field strength; This represents the effective capture cross-sectional area of ​​the probe.

[0086] By solving the inverse equations simultaneously, we can obtain the true natural flow velocity used for flux calculation:

[0087] Airborne edge computing units acquire real-time data. Then, continue to substitute it. And the final flux formula In this way, the true flux can be obtained.

[0088] For example, during a light initial shower, the probe was preset... The robot pre-marked the electroosmotic permeability of the site. Among them, electroosmotic permeability It is an empirical property parameter, and the electroosmotic permeability of most cohesive and silty soils is stable at [value missing]. arrive Within a certain range, it hardly changes with drastic changes in the hydraulic permeability coefficient, therefore, it is set... It can have high engineering representativeness; the applied electric field strength In in-situ electroosmosis experiments, excessively high voltage can lead to severe electrolysis of soil water, producing hydrogen and oxygen. These gases impede fluid flow and alter pH levels, affecting electrochemical detection. Conversely, excessively low voltage has no significant driving effect. This is the empirically recognized window range for balancing electroosmotic efficiency and suppressing electrolysis side reactions, therefore, we take... For example, substituting the values ​​yields the artificially generated electroosmotic flow. If the total flow velocity measured by the airborne high-precision flow meter is The system automatically decouples and calculates the actual natural soil seepage velocity to be only [amount missing]. The system then uses This real, weak natural hydrological parameter is multiplied by the ultra-high concentration of pore water forcibly desorbed by sound waves, thus avoiding being masked by organic matter, and finally yields the potential real pollution release flux of the area under long-term rainwater immersion.

[0089] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, 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 is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0090] It should be noted that the above modules can be implemented by software or hardware. For the latter, they can be implemented in the following ways, but are not limited to: all the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0091] Embodiments of the present invention also provide a computer-readable storage medium storing a computer program, wherein the computer program is configured to perform the steps in any of the above method embodiments when executed.

[0092] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0093] Embodiments of the present invention also provide an electronic device including a memory and a processor, the memory storing a computer program and the processor being configured to run the computer program to perform the steps in any of the above method embodiments.

[0094] In one exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0095] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0096] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another apparatus, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0097] The units described as separate components may or may not be physically separate. A component shown as a unit can be one or more physical units; that is, it can be located in one place or distributed in multiple different locations. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0098] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0099] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. 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.

[0100] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An event-driven method for monitoring soil pollution fluxes, characterized in that, include: The physical parameters within the target soil area are acquired in real time, and the occurrence of an environmental event is determined based on the changes in the physical parameters; the physical parameters include at least one of soil volumetric water content, soil electrical conductivity, and soil temperature. In response to the determination of the environmental event, the robot platform is dispatched to navigate to the area where the environmental event occurred; Using the robotic platform, in-situ soil pore water is continuously intercepted and extracted within the area where the environmental event occurred to obtain environmental sample streams. Flow injection analysis is performed on the environmental sample stream to determine the true pollutant concentration of the environmental sample stream. The flow injection analysis includes dynamic volume correction of the measured apparent pollutant concentration based on the monitored flow rate of the environmental sample stream and the flow rate of the simultaneously injected analytical reagent. Based on the actual pollutant concentration, the flow rate of the environmental sample stream, and the preset effective capture cross-sectional area of ​​the probe, the instantaneous mass flux of pollutants in the target soil area is determined.

2. The method according to claim 1, characterized in that, The determination of the occurrence of an environmental event based on the changes in the physical parameters includes: Collect the physical parameters and construct an instantaneous state vector; The Mahalanobis distance of the instantaneous state vector is obtained based on the historical baseline state vector mean and historical state covariance matrix. The Mahalanobis distance is compared with a preset trigger threshold. If the Mahalanobis distance is greater than the trigger threshold, the occurrence of the environmental event is determined.

3. The method according to claim 1, characterized in that, The method further includes: Based on peak flux, decay time constant, and delay time of environmental event occurrence, a pollution event feature tensor is constructed; Based on the pollution event feature tensor, the pollution release behavior of the target soil area is classified.

4. The method according to claim 1, characterized in that, Determining the instantaneous mass flux of pollutants within the target soil region also includes: Obtain the total mixing volume flow rate; The equivalent volumetric flow velocity of the directional electroosmotic flow is calculated based on the electric field strength of the applied constant DC electric field, the preset effective capture cross-sectional area of ​​the probe, and the electroosmotic permeability of the target soil area. The actual soil hydrological seepage velocity is determined based on the total mixing volume velocity and the equivalent volume velocity. The instantaneous mass flux is determined based on the actual soil hydrological seepage velocity.

5. An event-driven soil pollution flux monitoring system, characterized in that, include: A node sensor network is used to acquire physical parameters within a target soil area in real time, and to determine the occurrence of environmental events based on changes in the physical parameters; the physical parameters include at least one of soil volumetric water content, soil electrical conductivity, and soil temperature; A scheduling and navigation gateway is used to schedule the robot platform to navigate to the area where the environmental event occurred in response to the determination of the environmental event; A robotic platform is used to continuously intercept and extract in-situ soil pore water in the area where the environmental event occurs, in order to obtain an environmental sample stream. Flow injection analysis is performed on the environmental sample stream to determine the true pollutant concentration of the environmental sample stream. The flow injection analysis includes dynamic volume correction of the measured apparent pollutant concentration based on the monitored flow rate of the environmental sample stream and the flow rate of the simultaneously injected analytical reagent. The flux analysis platform is used to determine the instantaneous mass flux of pollutants in the target soil area based on the actual pollutant concentration, the flow rate of the environmental sample stream, and the preset effective capture cross-sectional area of ​​the probe.

6. The system according to claim 5, characterized in that, The node sensor network determines the occurrence of environmental events based on changes in the physical parameters, including: Collect the physical parameters and construct an instantaneous state vector; The Mahalanobis distance of the instantaneous state vector is obtained based on the historical baseline state vector mean and historical state covariance matrix. The Mahalanobis distance is compared with a preset trigger threshold. If the Mahalanobis distance is greater than the trigger threshold, the occurrence of the environmental event is determined.

7. The system according to claim 5, characterized in that, The system also includes: The tensor construction module is used to construct pollution event feature tensors based on flux peak, decay time constant, and the delay time of environmental event occurrence. The classification module is used to classify the pollution release behavior of the target soil area based on the pollution event feature tensor.

8. The system according to claim 5, characterized in that, The flux analysis platform is also used for: Obtain the total mixing volume flow rate; The equivalent volumetric flow velocity of the directional electroosmotic flow is calculated based on the electric field strength of the applied constant DC electric field, the preset effective capture cross-sectional area of ​​the probe, and the electroosmotic permeability of the target soil area. The actual soil hydrological seepage velocity is determined based on the total mixing volume velocity and the equivalent volume velocity. The instantaneous mass flux is determined based on the actual soil hydrological seepage velocity.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the computer program is configured to perform the method described in any one of claims 1 to 4 when executed.

10. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the method described in any one of claims 1 to 4.

Images