A groundwater pollution dissolved oxygen response monitoring and early warning system and method

Abstract

The application provides a groundwater pollution dissolved oxygen response monitoring and early warning system and method, and belongs to the technical field of groundwater environment monitoring. The internal dissolved oxygen concentration three-dimensional distribution, medium oxygen consumption flux field and seepage flow field of a groundwater area are obtained; a dynamic oxygen demand threshold is determined, and then an oxygen deficit voxel set is determined; a priority operation path and a stop point sequence with the incentive oxygen deficit voxel set as a target are planned, and a target incentive horizon and a controllable gas injection intensity are determined according to the vertical gradient distribution and the medium oxygen consumption flux; a mobile controllable disturbance platform is driven to inject disturbance, real-time dissolved oxygen response data is collected, and the internal dissolved oxygen concentration three-dimensional distribution is obtained again after disturbance; a contribution degree matrix of each stop point to the global dissolved oxygen is constructed, the contribution degree matrix is input into an ecological environment large model, and groundwater pollution risk early warning information is obtained by inversion. The scheme of the application can improve the energy efficiency of the mobile disturbance and reduce invalid disturbance.

Description

Technical Field

[0001] This application relates to the field of groundwater environmental monitoring technology, and more specifically, to a groundwater pollution dissolved oxygen response monitoring and early warning system and method. Background Technology

[0002] Groundwater pollution monitoring refers to the systematic observation and analysis of the types, concentrations, and spatiotemporal distribution of pollutants in groundwater through methods such as deploying monitoring wells, collecting water samples, or using in-situ sensors. This provides data support for pollution risk assessment, remediation decisions, and regulatory early warning. Dissolved oxygen is an important indicator characterizing the redox state of groundwater, and its concentration distribution is closely related to the natural decay of organic pollutants and the migration and transformation of reducing pollutants.

[0003] In existing groundwater pollution monitoring and early warning processes, traditional methods rely heavily on periodic sampling from fixed monitoring wells and laboratory analysis. These methods lack a refined understanding of the three-dimensional dissolved oxygen concentration distribution within groundwater and a real-time correlation with the dynamic oxygen demand thresholds of downstream protected objects. This makes it difficult to effectively identify dynamic, hidden hypoxic spaces formed by pollutant migration, aquifer heterogeneity, or uneven biological activity. For example, traditional methods may repeatedly monitor non-hypoxic or slightly hypoxic areas based on preset fixed thresholds. However, deep or distant hypoxic areas formed by aquifer stratification or excessive distance may fail to receive effective response signals due to oxygen diffusion compensation. Furthermore, changes in local dissolved oxygen concentration around monitoring points can hinder the oxygen gradient from penetrating heterogeneous media to reach the target hypoxic area, ultimately creating oxygen-deficient islands that are mismatched with ecological needs during monitoring. This leads to spatial distortion in the accuracy of groundwater pollution monitoring and early warning. Therefore, improving the efficiency of mobile disturbances and reducing ineffective disturbances has become a challenge for the industry. Summary of the Invention

[0004] This application provides a groundwater pollution dissolved oxygen response monitoring and early warning system and method, which improves the energy efficiency of mobile disturbances and reduces ineffective disturbances.

[0005] In a first aspect, this application provides a method for monitoring and early warning of dissolved oxygen response in groundwater pollution, comprising the following steps: To obtain the three-dimensional distribution of dissolved oxygen concentration, oxygen consumption flux field, and seepage flow field within the groundwater area; Determine the dynamic oxygen demand threshold at the downstream protected object, and identify the connected voxels in the three-dimensional distribution of dissolved oxygen concentration that have a dissolved oxygen concentration lower than the dynamic oxygen demand threshold as oxygen-deficient voxel sets. Based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, a priority operation path and stopping point sequence are planned with the goal of stimulating the oxygen-deficient voxel set. Based on the vertical gradient distribution of dissolved oxygen concentration and the oxygen consumption flux of the medium at each planned stopping point, the target excitation layer and controllable gas injection intensity are determined. The mobile controllable disturbance platform is driven to carry out controllable gas injection disturbance according to the priority operation path, target excitation layer and controllable gas injection intensity. During the disturbance, dissolved oxygen response data at different depths around the stopping point are collected in real time. After the disturbance is completed, the three-dimensional distribution of internal dissolved oxygen concentration in the groundwater area is obtained again. Based on the difference in the three-dimensional distribution of dissolved oxygen concentration before and after the disturbance, a contribution matrix is ​​constructed to quantify the impact of controllable gas injection disturbance at each station on the spatial distribution of dissolved oxygen in the whole region. The contribution matrix is ​​then input into a pre-constructed ecological environment model to obtain groundwater pollution risk early warning information.

[0006] In some embodiments, determining the dynamic oxygen demand threshold at the downstream protected object specifically includes: Acquire real-time water quality and seepage parameters related to the functions of downstream protected objects; The theoretical oxygen consumption rate of the downstream protected object is determined based on the real-time water quality and seepage parameters and the preset oxygen consumption kinetic model. The dynamic oxygen demand threshold at the downstream protected object is determined based on the theoretical oxygen consumption rate and the dissolved oxygen safety margin.

[0007] In some embodiments, identifying the oxygen-deficient voxel set in the three-dimensional distribution of dissolved oxygen concentration where the dissolved oxygen concentration is lower than the dynamic oxygen demand threshold specifically includes: The dissolved oxygen concentration of each voxel in the three-dimensional distribution of dissolved oxygen concentration is compared with the dynamic oxygen demand threshold, thereby identifying hypoxic voxels whose dissolved oxygen concentration is lower than the dynamic oxygen demand threshold. Three-dimensional spatial clustering analysis was performed on all hypoxic voxels to obtain multiple sets of oxygen-deficient voxels.

[0008] In some embodiments, based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, planning a priority operation path and stopping point sequence aimed at stimulating the oxygen-deficient voxel set specifically includes: The mobility and endurance constraints of the mobile controllable disturbance platform are quantified as input parameters for the path planning model; Based on the oxygen diffusion influence domain of each candidate stopping point, the excitation coverage and excitation efficiency of each candidate stopping point are calculated, and a constrained path optimization model is established with the goal of exciting the oxygen-deficient voxel set. Solve the constrained path optimization model to output the preferred operation path and stopping point sequence of the mobile controllable disturbance platform.

[0009] In some embodiments, determining the target excitation layer and the controllable gas injection intensity based on the vertical gradient distribution of dissolved oxygen concentration and the oxygen consumption flux of the medium at each planned stopping point specifically includes: Extract the characteristic data of the vertical gradient distribution of dissolved oxygen concentration at each planned stopping point; The target incentive level is determined based on the feature data and the preset incentive level selection rules; The controllable gas injection intensity of the mobile controllable disturbance platform is determined based on the degree of oxygen deficiency and the oxygen flux of the medium at the target excitation layer.

[0010] In some embodiments, driving the mobile controllable perturbation platform to perform controllable gas injection perturbation according to the priority operation path, target excitation level, and controllable gas injection intensity specifically includes: The preferred operation path, target excitation level, and controllable gas injection intensity are converted into an executable control instruction set for the mobile controllable disturbance platform; The mobile controllable disturbance platform is driven to sequentially complete movement, positioning, and controllable gas injection disturbance according to the executable control instruction set.

[0011] In some embodiments, constructing a contribution matrix for quantifying the impact of controllable gas injection disturbances at each dwell point on the spatial distribution of dissolved oxygen across the entire domain specifically includes: The difference between the three-dimensional distribution of dissolved oxygen concentration after the disturbance and the three-dimensional distribution of dissolved oxygen concentration before the disturbance is calculated to obtain the global dissolved oxygen increment field. The contribution ratio of the controllable gas injection disturbance at each stopping point to the increase in dissolved oxygen of each voxel in the whole domain is calculated based on the oxygen diffusion influence domain of each candidate stopping point. Based on the contribution ratio and the global dissolved oxygen increment field, the total contribution of each stopping point is calculated, and a contribution matrix is ​​constructed.

[0012] In some embodiments, inputting the contribution matrix into a pre-constructed large-scale ecological environment model to obtain groundwater pollution risk early warning information specifically includes: The contribution matrix, the characteristic parameters of the seepage flow field, and the characteristic parameters of the medium oxygen consumption flux field are fused together to form the input feature data of the ecological environment large model. The input feature data is input into the ecological environment big model, and the ecological environment big model performs inversion analysis on the input feature data based on the pollution response law it has pre-learned internally to obtain the pollution risk level of each oxygen-deficient voxel set; Based on the pollution risk level and spatial coordinates of each oxygen-deficient voxel set, groundwater pollution risk early warning information containing risk level, spatial location and confidence level is generated.

[0013] In some embodiments, the downstream protected object refers to a sensitive target or ecological purification structure located within the groundwater area that has functional requirements for dissolved oxygen concentration.

[0014] Secondly, this application provides a groundwater pollution dissolved oxygen response monitoring and early warning system, comprising: The acquisition module is used to acquire the three-dimensional distribution of dissolved oxygen concentration, oxygen consumption flux field, and seepage flow field within the groundwater area. The processing module is used to determine the dynamic oxygen demand threshold at the downstream protected object and to identify the clusters of connected voxels in the three-dimensional distribution of dissolved oxygen concentration that are lower than the dynamic oxygen demand threshold as oxygen-deficient voxel sets. The processing module is also used to plan a priority operation path and a sequence of stopping points with the goal of stimulating the oxygen-deficient voxel set based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, and to determine the target excitation layer and controllable gas injection intensity based on the vertical gradient distribution of dissolved oxygen concentration and the oxygen consumption flux of the medium at each planned stopping point. The processing module is also used to drive the mobile controllable disturbance platform to implement controllable gas injection disturbance according to the priority operation path, target excitation layer and controllable gas injection intensity. During the disturbance, dissolved oxygen response data at different depths around the stopping point are collected in real time. After the disturbance is completed, the three-dimensional distribution of internal dissolved oxygen concentration in the groundwater area is obtained again. The execution module is used to construct a contribution matrix that quantifies the impact of controllable gas injection disturbance at each station on the spatial distribution of dissolved oxygen in the whole region based on the difference in the three-dimensional distribution of dissolved oxygen concentration before and after the disturbance. The contribution matrix is ​​then input into a pre-constructed ecological environment model to obtain groundwater pollution risk early warning information.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: The groundwater pollution dissolved oxygen response monitoring and early warning system and method provided in this application firstly transforms the groundwater hypoxia state into a quantifiable spatial distribution characterization by determining a dynamic oxygen demand threshold and identifying oxygen-deficient voxel sets with dissolved oxygen concentrations below that threshold. This provides a data foundation for identifying ineffective disturbances and hypoxic spaces from a physical perspective through a clear comparison of dissolved oxygen concentration with the threshold. Subsequently, three-dimensional spatial clustering analysis is performed based on all hypoxic voxels to accurately identify isolated hypoxic regions formed by pollutant migration, aquifer heterogeneity, or uneven biological activity within the three-dimensional distribution of dissolved oxygen concentration. This enables automatic location and spatial delineation of oxygen-deficient voxel sets, transforming abstract data problems into stimulus tasks to be analyzed. This provides a crucial basis for this application from data characterization to operational target location. Furthermore, targeting the oxygen-deficient voxel sets, and considering the mobility and endurance constraints of the mobile controllable disturbance platform, a priority operational path and stop sequence are planned. This process can stimulate hypoxic areas. The two key engineering constraints of energy conservation are transformed from external empirical requirements into the core objective of driving path planning. Precise incentive interventions are applied to the identified oxygen-deficient voxel sets through the oxygen diffusion influence domain of each candidate stopping point, while path optimization reduces ineffective movement during operations. Finally, a contribution matrix is ​​constructed to quantify the contribution of each stopping point to the spatial distribution of dissolved oxygen across the entire region, and this matrix is ​​input into a large-scale ecological environment model to obtain groundwater pollution risk early warning information. This process allows incentive efficiency parameters verified by actual responses to be applied to the dynamic adjustment of pollution risk early warning and subsequent path planning strategies, achieving efficient allocation of incentive resources. This transforms disturbance operations from a crude process that merely pursues maximum coverage area into a scientific monitoring approach where the overall scheme both meets actual hypoxia response needs and strictly adheres to energy consumption constraints, effectively suppressing the coexistence of energy waste and insufficient response caused by blind disturbances. In summary, this scheme can improve the energy efficiency of mobile disturbances and reduce ineffective disturbances. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a groundwater pollution dissolved oxygen response monitoring and early warning method according to some embodiments of this application; Figure 2 This is a schematic flowchart illustrating the process of determining the dynamic oxygen demand threshold according to some embodiments of this application; Figure 3 This is a schematic flowchart illustrating the process of determining the controllable gas injection intensity according to some embodiments of this application; Figure 4 This is a schematic diagram of the structure of a groundwater pollution dissolved oxygen response monitoring and early warning system according to some embodiments of this application; Figure 5This is an internal structural diagram of a computer device for implementing a groundwater pollution dissolved oxygen response monitoring and early warning method, as shown in some embodiments of this application. Detailed Implementation

[0017] To better understand the technical solutions in this embodiment, the technical solutions in this embodiment will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0018] refer to Figure 1 The figure is a flowchart illustrating a groundwater pollution dissolved oxygen response monitoring and early warning method according to some embodiments of this application. The groundwater pollution dissolved oxygen response monitoring and early warning method mainly includes the following steps: In step 101, the three-dimensional distribution of dissolved oxygen concentration, oxygen consumption flux field, and seepage flow field within the groundwater area are obtained; In practice, firstly, a wireless sensor network consisting of several dissolved oxygen sensor nodes is deployed three-dimensionally within the groundwater area to be monitored, according to a preset spatial layout rule. These dissolved oxygen sensor nodes can be fixed in monitoring wells, anchored cables, or on mobile monitoring platforms to simultaneously collect raw data on dissolved oxygen concentrations at different horizontal positions and vertical depths. Each sensor node, through its integrated low-power wide-area IoT communication module, sends dissolved oxygen concentration data with precise spatial coordinates and timestamps to a central data processing unit deployed on shore-based or in the cloud, either periodically or via event triggering. In the central data processing unit, the received spatiotemporally discrete data stream is cleaned, aligned, and time-synchronized to form a set of dissolved oxygen concentration sampling points under a unified spatiotemporal reference. Subsequently, based on this sampling point set, a preset spatial interpolation algorithm is used to estimate the dissolved oxygen concentration at groundwater locations not directly covered by the sensors. A digital model of dissolved oxygen concentration distribution is generated, continuously covering the entire three-dimensional space of the groundwater area within a specified time section or time window, i.e., the three-dimensional distribution of dissolved oxygen concentration. Preferably, in this embodiment, the medium oxygen consumption flux field can be obtained in the following way: in the groundwater area to be monitored, medium sampling points are arranged in a 50m×50m grid, and the medium oxygen consumption rate of each sampling point is measured by in-situ incubation method; the medium oxygen consumption rate data of each sampling point is input into the central data processing unit, and the medium oxygen consumption flux field continuously covering the bottom of the entire groundwater area is generated by Kriging interpolation algorithm; in addition, when obtaining the seepage flow field, the groundwater seepage model can be used, and the aquifer topographic data, real-time water level data, recharge flow data and discharge flow data can be input to calculate the seepage velocity and flow direction distribution of the entire groundwater area, i.e., the seepage flow field; the time step of the seepage model is set to 10 minutes, and the spatial resolution is set to 10m×10m.

[0019] It should be noted that, in this application, the groundwater area refers to the spatial area of ​​groundwater within the target monitoring range that includes the saturated water medium in the lower part of the vadose zone and the unconfined aquifer, and its range is determined by a preset monitoring boundary; as a preferred embodiment, the groundwater area is a shallow unconfined aquifer contaminated site with a medium permeability coefficient of not less than 10⁻ 4 The dissolved oxygen concentration is on the order of m / s and does not contain free-phase organic matter. The controllable gas injection disturbance is not applicable to confined aquifers and low-permeability strata. The method described in this application is used for shallow unconfined aquifer areas that meet the above permeability conditions. The three-dimensional distribution of dissolved oxygen concentration refers to a data structure that digitizes and models the spatial distribution and temporal changes of dissolved oxygen concentration within a groundwater area. Its physical significance lies in providing a high-resolution global information base for subsequent identification of anoxic areas, assessment of the degree of anoxic conditions, and dynamic change trends. The medium described in this application refers to the aquifer skeleton and vadose zone soil medium. The sensor array refers to a collection of multiple dissolved oxygen sensors deployed in a certain topological structure, such as a grid or cross-sectional structure, to meet the requirements of three-dimensional spatial coverage and real-time monitoring. The deployment density and position determine the spatial resolution and accuracy of the generated map. The spatial interpolation algorithm is a mathematical method that predicts attribute values ​​at unsampled points based on the measured values ​​of known spatial sampling points. Its core lies in modeling spatial autocorrelation. This application does not limit the specific algorithm variants. The key is to use this method to transform discrete point data into continuous three-dimensional field data.

[0020] In step 102, the dynamic oxygen demand threshold at the downstream protected object is determined, and then the oxygen-deficient voxel set in the three-dimensional distribution of dissolved oxygen concentration that has a dissolved oxygen concentration lower than the dynamic oxygen demand threshold is identified.

[0021] In some embodiments, reference Figure 2 As shown in the figure, this is a schematic flowchart illustrating the process of determining the dynamic oxygen demand threshold in some embodiments of this application. The determination of the dynamic oxygen demand threshold at the downstream protected object can be achieved by the following steps: First, in step 1021, real-time water quality and seepage parameters related to the function of the downstream protected object are obtained; Then, in step 1022, the theoretical oxygen consumption rate of the downstream protected object is determined based on the real-time water quality and seepage parameters and the preset oxygen consumption kinetic model. Finally, in step 1023, the dynamic oxygen demand threshold at the downstream protected object is determined based on the theoretical oxygen consumption rate and the dissolved oxygen safety margin.

[0022] In specific implementation, obtaining real-time water quality and seepage parameters related to the function of the downstream protected object can be achieved in the following way: For example, in the groundwater upstream of the inlet of the downstream protected object, online water quality sensors and flow meters matching the preset function of the object are deployed to synchronously and continuously collect raw monitoring data. The raw monitoring data is then processed for format unification and time alignment and used as the real-time water quality and seepage parameters. Specifically, if the functional target of the downstream protected object is the nitrification removal of ammonia nitrogen, then the water quality sensor is an online ammonia nitrogen analyzer, and the collected ammonia nitrogen concentration time series data is the key water quality parameter. At the same time, the instantaneous flow time series data collected by the flow meter is the key seepage parameter. The online monitoring equipment uploads the raw data to the central processing unit in real time through the industrial Internet of Things network. In the central processing unit, outlier filtering and missing time interpolation repair are performed on the received data to ensure the continuity and reliability of the parameter data stream, thereby providing a consistent input for subsequent model calculations.

[0023] It should be noted that the real-time water quality and seepage parameters mentioned in this application refer to the set of instantaneous monitoring data collected synchronously from the downstream protected object and directly related to the preset functions of the object, used to characterize the pollutant load and hydrodynamic conditions currently entering the object.

[0024] In specific implementation, the theoretical oxygen consumption rate of the downstream protected object can be determined based on the real-time water quality and seepage parameters and the preset oxygen consumption kinetic model in the following way: For example, the real-time water quality and seepage parameters are used as input variables, and the oxygen consumption kinetic model preset in the central processing unit is called for calculation. The instantaneous oxygen consumption intensity value output by the model is used as the theoretical oxygen consumption rate. The oxygen consumption kinetic model is a mathematical relationship based on the principles of pollutant mass conservation and microbial degradation reaction kinetics, which expresses the amount of dissolved oxygen required per unit time to achieve the preset pollutant removal target of the downstream protected object. As a specific embodiment, for protected objects undertaking nitrification functions, the core correlation of the model is... The theoretical oxygen consumption rate is directly proportional to the influent ammonia nitrogen load and corrected by the water temperature factor, wherein the influent ammonia nitrogen load is obtained by multiplying the real-time ammonia nitrogen concentration by the instantaneous flow rate. The kinetic parameters in the model are obtained through reverse identification of historical operating data of the specific protected object and stored as known constants in the model. The central processing unit automatically executes the calculation of this mathematical relationship, thereby dynamically outputting the theoretical oxygen consumption rate. In a preferred embodiment, the reverse identification of historical operating data can use the least squares method to fit the relationship curve between actual oxygen consumption and influent load to determine the model proportionality coefficient. In other embodiments, if the functional objective of the downstream protected object is aerobic degradation of chemical oxygen demand (COD), then a preset COD-based parameter is invoked. An oxygen consumption model for COD (Chemical Oxygen Demand) degradation kinetics is used. The core correlation of this model is that the theoretical oxygen consumption rate is directly proportional to the influent COD load and is synergistically corrected by water temperature and pH factors. The influent COD load is obtained by multiplying the real-time COD concentration by the instantaneous flow rate. The oxygen consumption coefficient in the model can be obtained through small-scale experiments or fitting long-term operational data of the object. In another embodiment, if the protected object is a multifunctional type (e.g., simultaneously undertaking nitrification and COD degradation), then the theoretical oxygen consumption rate is the sum of the nitrification oxygen consumption rate based on ammonia nitrogen load and the carbon oxidation oxygen consumption rate based on COD load. The oxygen consumption rate is summed and weighted according to the weight of the target pollutant removal ratio; the weight can be preset according to the target concentration of influent and effluent or the area ratio of functional zones within the protected object; as a preferred embodiment, for seasonal groundwater that is significantly affected by water temperature, an activity correction coefficient that has a non-linear relationship with water temperature can be introduced into the model to more accurately reflect the impact of seasonal changes in microbial activity on the oxygen consumption rate; in other embodiments, online microbial sensor data can also be integrated to correct the biomass parameters in the model in real time and realize the adaptive calibration of the model, which is not limited in this application.

[0025] It should be noted that the downstream protected object mentioned in this application refers to a sensitive target or ecological purification structure located within the groundwater area that has functional requirements for dissolved oxygen concentration. The dissolved oxygen requirement at its inlet is defined as a dynamic target. In addition, the theoretical oxygen consumption rate refers to the amount of dissolved oxygen consumed by the downstream protected object per unit time to achieve its target function, calculated based on real-time influent conditions and a preset reaction kinetic model. It is used to quantify the instantaneous oxygen demand intensity of the object, and thus serves as a core intermediate variable for estimating the dissolved oxygen concentration requirement at its inlet.

[0026] In practice, determining the dynamic oxygen demand threshold at the downstream protected object based on the theoretical oxygen consumption rate and dissolved oxygen safety margin can be achieved in the following way: For example, the theoretical oxygen consumption rate is combined with the hydraulic retention characteristics or volumetric characteristics of the downstream protected object to convert it into a reference value of the minimum dissolved oxygen concentration required to maintain its internal aerobic biochemical reactions. Then, a preset dissolved oxygen safety margin is added to this reference value, and the resulting dissolved oxygen concentration value is used as the dynamic oxygen demand threshold. The physical principle underlying this conversion process is that, under the steady-state assumption, the change in dissolved oxygen concentration in groundwater is jointly determined by the oxygen mass transfer replenishment rate and the biochemical reaction consumption rate. This is achieved by dividing the theoretical oxygen consumption rate by a comprehensive coefficient related to the water volume and oxygen mass transfer capacity of the protected object. The corresponding dissolved oxygen concentration reference value can be obtained; the dissolved oxygen safety margin is a positive constant preset according to functional stability requirements and system uncertainty, used to buffer the risks caused by load fluctuations and model prediction errors; the finally calculated dynamic oxygen demand threshold is a scalar that changes with time, which is directly used as a benchmark for comparison with the measured dissolved oxygen concentration at each spatial location in the three-dimensional distribution of dissolved oxygen concentration; as a preferred embodiment, the comprehensive coefficient can be determined by multiplying the water volume of the downstream protected object by an empirical constant characterizing the total oxygen transfer coefficient; in other embodiments, the dissolved oxygen safety margin can be adaptively adjusted according to the range of load fluctuations, for example, increasing the margin when the load fluctuation is large, which is not limited in this application.

[0027] It should be noted that the dynamic oxygen demand threshold mentioned in this application refers to a scalar value of dissolved oxygen concentration that changes over time. It serves as a dynamic benchmark for determining whether each spatial location in the three-dimensional distribution of dissolved oxygen concentration is in a state of hypoxia, thereby transforming the functional oxygen demand of downstream protected objects into a unified and variable scale for assessing the dissolved oxygen concentration across the entire region.

[0028] In some embodiments, identifying oxygen-deficient voxel sets in the three-dimensional distribution of dissolved oxygen concentration where the dissolved oxygen concentration is below the dynamic oxygen demand threshold can be achieved using the following steps: The dissolved oxygen concentration of each voxel in the three-dimensional distribution of dissolved oxygen concentration is compared with the dynamic oxygen demand threshold, thereby identifying hypoxic voxels whose dissolved oxygen concentration is lower than the dynamic oxygen demand threshold. Three-dimensional spatial clustering analysis was performed on all hypoxic voxels to obtain multiple sets of oxygen-deficient voxels.

[0029] In specific implementation, comparing the dissolved oxygen concentration of each voxel in the three-dimensional distribution of dissolved oxygen concentration with the dynamic oxygen demand threshold, and then marking hypoxic voxels whose dissolved oxygen concentration is lower than the dynamic oxygen demand threshold, can be achieved in the following way: For example, traversing each voxel in the three-dimensional distribution of dissolved oxygen concentration, extracting its dissolved oxygen concentration value, comparing the concentration value with the dynamic oxygen demand threshold, and determining whether the concentration value is lower than the threshold; if the dissolved oxygen concentration of a voxel is lower than the dynamic oxygen demand threshold, then the voxel is marked as a hypoxic voxel, and its three-dimensional grid coordinates are recorded; as a preferred embodiment, a minimum duration threshold can be set. If the dissolved oxygen concentration of a voxel is lower than the threshold for three consecutive sampling periods, then it is marked as a hypoxic voxel to eliminate false judgments caused by sensor noise.

[0030] It should be noted that the hypoxic voxel mentioned in this application refers to a single three-dimensional grid cell in the three-dimensional distribution of dissolved oxygen concentration whose dissolved oxygen concentration is determined to be lower than the dynamic oxygen demand threshold, and is the basic element constituting a continuous hypoxic space.

[0031] In practical implementation, performing three-dimensional spatial clustering analysis on all hypoxic voxels to obtain multiple oxygen-deficient voxel sets can be achieved in the following way: For example, using all hypoxic voxels and their three-dimensional coordinates as input, a three-dimensional connected component analysis algorithm is used for spatial clustering. The three-dimensional connected component analysis is based on the spatial adjacency of voxels in the three-dimensional mesh, merging adjacent hypoxic voxels into the same connected component, each of which is defined as an oxygen-deficient voxel set. Spatial adjacency is typically defined as two voxels sharing a face, an edge, or a corner in the three-dimensional mesh. In practical implementation, the six-neighbor rule can be used to simplify calculations. The algorithm execution process is as follows: Unvisited hypoxic voxels are... Initially, all spatially adjacent hypoxic voxels are recursively searched and marked until no new adjacent voxels are added, thus forming a complete set of oxygen-deficient voxels. This process is repeated until all hypoxic voxels have been visited, ultimately resulting in multiple spatially disconnected sets of oxygen-deficient voxels. As a preferred embodiment, after clustering is completed, a minimum voxel count threshold can be set to filter out small sets of oxygen-deficient voxels containing fewer voxels than the threshold, thereby eliminating interference from sporadic hypoxic points caused by sensor noise or interpolation errors on subsequent path planning. In other embodiments, a clustering algorithm based on three-dimensional spatial density can also be used, and this application does not limit this approach.

[0032] It should be noted that the oxygen-deficient voxel set mentioned in this application refers to a continuous three-dimensional spatially connected domain formed by clustering multiple spatially adjacent oxygen-deficient voxels, which is a digital object representing an independent and contiguous oxygen-deficient spatial entity in groundwater.

[0033] In step 103, based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, a priority operation path and stopping point sequence are planned with the goal of stimulating the oxygen-deficient voxel set. Based on the vertical gradient distribution of dissolved oxygen concentration and the oxygen consumption flux of the medium at each planned stopping point, the target excitation layer and the controllable gas injection intensity are determined.

[0034] In some embodiments, based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, planning a priority operation path and stopping point sequence aimed at stimulating the oxygen-deficient voxel set can be achieved through the following steps: The mobility and endurance constraints of the mobile controllable disturbance platform are quantified as input parameters for the path planning model; Based on the oxygen diffusion influence domain of each candidate stopping point, the excitation coverage and excitation efficiency of each candidate stopping point are calculated, and a constrained path optimization model is established with the goal of exciting the oxygen-deficient voxel set. Solve the constrained path optimization model to output the preferred operation path and stopping point sequence of the mobile controllable disturbance platform.

[0035] In specific implementation, the mobility and endurance constraints of the mobile controllable disturbance platform can be quantified as input parameters of the path planning model in the following ways: for example, extract the maximum speed and the maximum working distance after a single energy replenishment from the factory specifications or real-time status monitoring data of the mobile controllable disturbance platform. Use the maximum speed as a rate parameter characterizing mobility and the maximum working distance as a distance parameter characterizing endurance constraints, which together constitute the input parameters of the path planning model. As a preferred embodiment, if the platform's movement is significantly affected by seepage, the rate parameter can be corrected to the effective ground speed relative to static groundwater. In other embodiments, if the platform has different travel gears or energy consumption modes, multiple speed-energy consumption pairs can be extracted as more refined inputs, which is not limited in this application.

[0036] In this embodiment, the oxygen diffusion influence domain can be determined in the following way: Based on the two-film theory and the convection-diffusion equation, a mathematical model of oxygen diffusion at the gas injection point is established; the model input parameters include controllable gas injection intensity, injection depth, seepage velocity, medium turbulence intensity, and water temperature; the model output is the dissolved oxygen concentration increment at different locations and times around the injection point; through this model, the range and shape of the oxygen diffusion influence domain of a certain stationary point under a specified controllable gas injection intensity and injection time, as well as the dissolved oxygen increment of each voxel within the influence domain, can be predicted.

[0037] It should be noted that the mobile controllable disturbance platform described in this application refers to an unmanned operating platform that can move autonomously within a target area and is equipped with a controllable gas injection function unit. Its function is to flexibly reach a designated spatial location according to central instructions and perform gas injection operations of a fixed depth and intensity, thereby realizing the stimulation intervention of dispersed oxygen-deficient areas. In addition, the path planning model refers to a mathematical abstraction that takes multiple oxygen-deficient areas as nodes to be visited and takes the mobility and endurance of the mobile controllable disturbance platform as constraints.

[0038] In specific implementation, the mobile controllable disturbance platform can be implemented in the following two physical forms: Implementation Method 1: The mobile controllable disturbance platform is a vehicle-mounted wellhead operation system, including a mobile carrier, a gas injection device, a dissolved oxygen sensor, and a control device; the mobile carrier is a wheeled or tracked vehicle chassis; the gas injection device is fixed to the mobile carrier and includes a gas source, a gas pipeline, and an injection pipe, the output end of which extends into the target underground depth through the wellhead of a pre-set injection well; the dissolved oxygen sensor includes at least one dissolved oxygen sensor probe, which extends into the target underground depth through the wellhead of a monitoring well; the control device is integrated with the mobile carrier. The gas injection device and the dissolved oxygen sensor are communicatively connected; the preferred operating path is the transfer route of the mobile carrier between the preset injection wellheads; the mobility is characterized by the maximum speed of the mobile carrier, and the endurance constraint is characterized by the gas storage capacity of the gas source and the energy capacity of the mobile carrier; during operation, the mobile carrier stops at each preset injection wellhead in sequence according to the preferred operating path, and lowers the injection pipe and the dissolved oxygen sensor probe to the corresponding depth to implement controllable gas injection disturbance and collect dissolved oxygen response data.

[0039] Implementation Method Two: The mobile controllable disturbance platform is a direct-push in-situ operation system, including a direct-push drive device, a gas injection probe, and a dissolved oxygen sensing probe. The direct-push drive device directly pushes the gas injection probe and the dissolved oxygen sensing probe into the underground target excitation layer, implements controllable gas injection disturbance at the predetermined stopping point, and collects dissolved oxygen response data. The preferred operation path is the movement route of the direct-push drive device between various stopping points within the target area. The mobility is characterized by the travel speed of the direct-push drive device, and the endurance constraint is characterized by the gas source capacity of the gas injection probe and the energy capacity of the direct-push drive device.

[0040] In a preferred embodiment, the vehicle-mounted wellhead operation system may also be equipped with a satellite positioning module. The wellhead coordinates of each preset injection well are pre-stored in the control device, which controls the mobile carrier to travel to the target wellhead location based on the priority operation path and real-time positioning information. In other embodiments, the mobile controllable disturbance platform may also be an amphibious mobile carrier, and this application does not limit it to this.

[0041] In specific implementation, the excitation coverage and excitation efficiency of each candidate stopping point are calculated based on the oxygen diffusion influence domain of each candidate stopping point. A constrained path optimization model aimed at exciting the oxygen-deficient voxel set can be established in the following way: For example, within the groundwater area to be monitored, a set of candidate stopping points is generated using a 20m×20m grid; for each candidate stopping point, the oxygen diffusion influence domain determination method described above is called to calculate its oxygen diffusion influence domain after 30 minutes of injection at the standard controllable gas injection intensity, as well as the number of oxygen-deficient voxels and the total oxygen-deficient excitation amount within this influence domain; the total oxygen-deficient excitation amount of each candidate stopping point is divided by the sum of its controllable gas injection energy consumption and the energy consumption for moving to that point, to obtain the... The incentive efficiency of the stopping points; taking the starting point of the mobile controllable disturbance platform as the path start and end point, the rate parameter and distance parameter are used to construct model constraints; the global oxygen deficit incentive rate is defined as the ratio of the total oxygen deficit voxel volume covered by the oxygen diffusion influence domain of all stopping points in this operation to the total oxygen deficit voxel volume of the entire region; the total energy consumption is defined as the sum of the controllable gas injection energy consumption of all stopping points and the sum of the movement energy consumption; the goal of the constrained path optimization model is to find a sequence that starts from the starting point, visits several candidate stopping points, and returns to the end point, so as to maximize the ratio of global oxygen deficit incentive rate to total energy consumption, while satisfying the constraint that the total operation path length does not exceed the distance parameter; As a specific embodiment, the model can be constructed as a maximum coverage path problem model with capacity constraints; in other embodiments, the objective function can also be defined as the ratio of total oxygen deficit stimulus to total energy consumption, which is not limited in this application; in addition, this embodiment can also be implemented by the following method, for example: the objective function can be constructed as a weighted sum of multiple sub-objectives to simultaneously optimize multiple performance indicators; the first sub-objective is to maximize the global oxygen deficit stimulus rate to prioritize the processing of the most oxygen-deficient areas; the second sub-objective is to minimize the total operation path length to reduce ineffective movement energy consumption; the third sub-objective is to maximize the continuity of the spatial distribution of access stopping points, that is, to prioritize accessing candidate stopping points that are clustered in space. The point cluster is used to reduce frequent platform turns and long-distance movements. This sub-objective can be quantified by calculating the negative value of the spatial dispersion of the visited stopping point set. The objective of the constrained path optimization model is to find a visiting sequence that maximizes the weighted multi-objective function value while satisfying the constraint that the total path length does not exceed the range. The weight of each sub-objective can be dynamically configured according to the focus of this operation. As a preferred embodiment, a penalty term related to the excitation efficiency of stopping points can be introduced into the objective function. For candidate stopping points with low excitation efficiency, even if they cover a lot of oxygen-deficient voxels, their access priority can be appropriately reduced to avoid energy waste. This application does not limit this.

[0042] It should be noted that the constrained path optimization model described in this application refers to a specific instance of the path planning model. It explicitly sets a mathematical model with the ratio of global oxygen deficit incentive rate to total energy consumption as the core optimization objective and the platform's range as a hard constraint. This model is used to solve for a theoretically better sequence of operation paths under complex multi-objective trade-offs and resource constraints.

[0043] In specific implementation, solving the constrained path optimization model and outputting the preferred operation path and stopping point sequence of the mobile controllable disturbance platform can be achieved in the following ways, for example: using a heuristic optimization algorithm to solve the constrained path optimization model, such as a genetic algorithm, simulated annealing algorithm, or ant colony algorithm; taking a genetic algorithm as an example, firstly, a possible access sequence is encoded as a chromosome, and an initial population containing multiple chromosomes is randomly generated; then, using the reciprocal of the model objective function as the fitness evaluation criterion, selection, crossover, and mutation operations are iteratively performed. The population evolves gradually. In each generation, the path is obtained by decoding the chromosome and its total length is calculated to verify whether it meets the distance parameter constraint. Individuals that do not meet the constraint are penalized or eliminated. When the evolution reaches the preset number of iterations or fitness convergence threshold, the individual with the best fitness is selected from the final population, and the access order and position of the stopping point obtained after decoding are output as the priority operation path and stopping point sequence, respectively. In other embodiments, an accurate algorithm or a commercial solver can also be used to solve the simplified model, and this application does not limit this.

[0044] It should be noted that the priority operation path and stop point sequence mentioned in this application refer to the solution output of the constrained path optimization model, specifically a continuous trajectory that starts from the starting point, visits several planned stop points in sequence, and finally returns, along with a list of spatial coordinates of the corresponding stop points.

[0045] In some embodiments, determining the target excitation layer and controllable gas injection intensity based on the vertical gradient distribution of dissolved oxygen concentration and the oxygen consumption flux of the medium at each planned stopping point can be achieved through the following steps: Extract the characteristic data of the vertical gradient distribution of dissolved oxygen concentration at each planned stopping point; The target incentive level is determined based on the feature data and the preset incentive level selection rules; The controllable gas injection intensity of the mobile controllable disturbance platform is determined based on the degree of oxygen deficiency and the oxygen flux of the medium at the target excitation layer.

[0046] In specific implementation, the feature data of the vertical gradient distribution of dissolved oxygen concentration at each planned stopping point can be extracted in the following way: for example, based on the spatial coordinates of each planned stopping point, the dissolved oxygen concentration values ​​corresponding to a series of discrete depth layers along the vertical direction at the coordinate point are extracted from the three-dimensional distribution of dissolved oxygen concentration to form the vertical profile data of dissolved oxygen concentration at the stopping point; a first-order difference calculation is performed on the vertical profile data to obtain the dissolved oxygen concentration change values ​​between adjacent depth layers, thereby obtaining a vertical gradient sequence characterizing the rate of concentration change with depth; the dissolved oxygen concentration values ​​of each depth layer in the vertical profile data and the gradient values ​​of each depth interval in the vertical gradient sequence are used together as the feature data; as a preferred embodiment, the original vertical profile data can be smoothed and filtered before calculating the difference to suppress the interference of small-scale fluctuations on the gradient calculation; in other embodiments, the feature data may also include the depth of the thermocline layer or the depth of the dissolved oxygen concentration abrupt change layer, which is not limited in this application.

[0047] It should be noted that the characteristic data of vertical gradient distribution mentioned in this application refers to the quantitative description of the change of dissolved oxygen concentration with depth from the groundwater surface to the deep layer at any planned stopping point, which is used to reveal the stratification structure and degree of change of dissolved oxygen in the vertical direction at that point.

[0048] In specific implementation, determining the target excitation layer based on the feature data and the preset excitation layer selection rule can be achieved in the following way: for example, the feature data is input into the preset excitation layer selection rule for judgment, and the depth range output by the rule is taken as the target excitation layer; the preset excitation layer selection rule is a logical judgment criterion used to select a specific layer that most needs gas injection excitation from multiple candidate layers in the vertical direction; as a specific embodiment, the rule is: starting from the shallow part, traverse downwards to find the first depth layer whose dissolved oxygen concentration value is continuously lower than the dynamic oxygen demand threshold, and include the depth layer containing this depth layer. The continuous depth range within the vertical gradient sequence, up to and including the concentration rising above the threshold or reaching the groundwater boundary, is determined as the target excitation layer. In another embodiment, the rule can also be defined as selecting the layer corresponding to the depth range with the most drastic gradient change in the vertical gradient sequence, i.e., the fastest concentration change. As a preferred embodiment, if multiple candidate layers are identified according to the rule, the layer with the lowest average dissolved oxygen concentration or the largest thickness can be selected as the final target. In other embodiments, the rule can also be combined with water temperature vertical distribution data to preferentially select anoxic layers where thermal stratification hinders oxygen exchange between upper and lower layers; this application does not limit this.

[0049] It should be noted that the target excitation layer mentioned in this application refers to a specific vertical depth range determined based on the vertical gradient distribution characteristic data. It is the range within which controllable gas injection energy is vertically focused, aiming to improve the hypoxia condition in this layer with the highest efficiency and promote the vertical mixing of the water.

[0050] For specific implementation, refer to Figure 3 As shown in the figure, this is a flowchart illustrating the process of determining the controllable gas injection intensity in some embodiments of this application. The controllable gas injection intensity of the mobile controllable disturbance platform can be determined based on the degree of oxygen deficiency and the oxygen flux of the medium at the target excitation layer. This can be achieved in the following ways: for example, calculating the difference between the average dissolved oxygen concentration at the target excitation layer and the dynamic oxygen demand threshold, using this difference as the core indicator for quantifying the degree of oxygen deficiency; subsequently, determining the corresponding control command based on the degree of oxygen deficiency indicator and the oxygen flux of the medium through a preset mapping relationship, and using this command as the controllable gas injection intensity; the preset mapping relationship defines different ranges of degree of oxygen deficiency indicator values ​​and medium oxygen flux values ​​with different levels of controllable gas injection intensity. The correspondence between degree control commands; the controllable gas injection intensity control command will directly regulate the output power or gas flow rate of the injection unit of the mobile controllable disturbance platform; as a preferred embodiment, the mapping relationship can be designed as a piecewise linear relationship, that is, the larger the hypoxia degree index value and the larger the medium oxygen consumption flux value, the higher the corresponding controllable gas injection intensity control command level; in another embodiment, the thickness of the target excitation layer can also be introduced as an auxiliary correction factor to perform weighted adjustment on the controllable gas injection intensity initially determined according to the hypoxia degree and medium oxygen consumption flux, the thicker the layer, the higher the adjusted intensity; in other embodiments, more accurate calculations can also be performed based on the oxygen transfer efficiency model of the platform, which is not limited in this application.

[0051] It should be noted that the controllable gas injection intensity mentioned in this application refers to the graded or continuous control command for the output power or gas volume of the injection unit on the mobile controllable disturbance platform, which is determined according to the degree of hypoxia and the oxygen consumption flux of the medium at the target excitation layer. This is used to match the energy input of controllable gas injection with the severity of local hypoxia and the oxygen consumption demand of the medium.

[0052] In step 104, the mobile controllable disturbance platform is driven to perform controllable gas injection disturbance according to the priority operation path, target excitation layer and controllable gas injection intensity. During the disturbance, dissolved oxygen response data at different depths around the stopping point are collected in real time. After the disturbance is completed, the three-dimensional distribution of internal dissolved oxygen concentration in the groundwater area is obtained again.

[0053] In some embodiments, driving the mobile controllable perturbation platform to implement controllable gas injection perturbation according to the priority operation path, target excitation level, and controllable gas injection intensity can be achieved by the following steps: The preferred operation path, target excitation level, and controllable gas injection intensity are converted into an executable control instruction set for the mobile controllable disturbance platform; The mobile controllable disturbance platform is driven to sequentially complete movement, positioning, and controllable gas injection disturbance according to the executable control instruction set.

[0054] In specific implementation, the conversion of the priority operation path, target excitation layer, and controllable gas injection intensity into an executable control instruction set for the mobile controllable disturbance platform can be achieved in the following ways: For example, in the central processing unit, the priority operation path is parsed into a series of waypoints arranged in sequence, containing latitude and longitude coordinates, and based on the navigation positioning accuracy and control characteristics of the mobile controllable disturbance platform, corresponding heading and speed control instructions are generated for each waypoint; simultaneously, the target excitation layer is parsed into the target operation depth that the injection disk or injection head on the mobile controllable disturbance platform needs to reach, and corresponding depth adjustment or hovering control instructions are generated; furthermore, the controllable gas injection intensity is parsed into an adjustment instruction for the output power or gas flow rate of the injection unit, which is usually expressed in the form of duty cycle, frequency, or percentage; finally, all the above control instructions are sorted, encapsulated, and verified according to the time logic and spatial logic of the operation path to form a complete executable control instruction set that can be directly recognized and executed by the control system of the mobile controllable disturbance platform, and then sent to the platform through a wireless communication link.

[0055] It should be noted that the executable control instruction set mentioned in this application refers to a series of low-level drive instructions that are translated and encapsulated into abstract priority operation paths, target excitation levels and controllable gas injection intensity, and can be directly recognized and executed by the low-level control system of the mobile controllable disturbance platform.

[0056] In specific implementation, driving the mobile controllable disturbance platform to sequentially complete movement, positioning, and controllable gas injection disturbance according to the executable control command set can be achieved in the following way: For example, the control system of the mobile controllable disturbance platform receives and parses the executable control command set, first driving its movement and heading control system to autonomously navigate to the first planned stopping point according to the waypoint sequence, heading, and speed control commands; after arriving at the stopping point, according to the depth adjustment command, the onboard injection unit is precisely lowered or positioned to the target excitation layer depth through a lifting device or positioning mechanism; subsequently, the control system starts and regulates the injection unit to begin controllable gas injection disturbance at a specified power or flow rate according to the adjustment command; during the disturbance process, the dissolved oxygen sensor on the platform collects data every 5 minutes. Dissolved oxygen concentration data at depths of 0.5m, 1m, 2m, and 3m around the stopping point are collected and uploaded to the central processing unit in real time. When the dissolved oxygen concentration at a certain depth reaches the preset stopping threshold (e.g., dynamic oxygen demand threshold + 0.5mg / L) for three consecutive measurements, the central processing unit sends a stop injection command to the platform. The platform prematurely terminates the controllable gas injection disturbance at that stopping point and proceeds to the next stopping point. After completing the disturbance at the current point, the platform automatically follows the command sequence to the next stopping point, repeating the positioning and controllable gas injection disturbance process until all planned stopping points are disturbed in sequence and the platform finally returns. Throughout the disturbance process, the platform's navigation and positioning module, depth sensor, and attitude sensor continuously provide feedback, and the control system performs closed-loop control based on the feedback to ensure operational accuracy.

[0057] In this embodiment, during the disturbance process, dissolved oxygen response data at different depths around the stopping point are collected at preset time intervals. When the dissolved oxygen concentration reaches a preset stopping threshold, the controllable gas injection disturbance at the stopping point is terminated in advance. In addition, in a specific implementation, after the disturbance is completed, the three-dimensional distribution of dissolved oxygen concentration in the groundwater area can be obtained again in the following way: after the mobile controllable disturbance platform completes the planned disturbance task and returns, the central control system automatically issues a data acquisition command to trigger each dissolved oxygen sensor node in the wireless sensor network to start a new round of synchronous monitoring. Each node collects dissolved oxygen concentration data at each spatial location in the groundwater after the disturbance with the same or shorter sampling period as in the previous steps, and sends the data with coordinates and timestamps to the central data processing unit through the Internet of Things communication module. The central data processing unit receives this new spatiotemporal discrete data stream and performs the same cleaning, alignment, time synchronization and spatial interpolation processing flow as in step 101, thereby generating a new digital model of dissolved oxygen concentration distribution within a specified time section or window after the disturbance is completed, that is, the three-dimensional distribution of dissolved oxygen concentration obtained again.

[0058] It should be noted that the technical essence of the re-acquisition operation described in this application lies in providing the latest and most direct data feedback for evaluating the effect of the current controlled gas injection disturbance and initiating the next round of optimization decisions. This operation is automatically triggered by the disturbance completion event, ensuring the timeliness of data in the evaluation and optimization stages. The generated three-dimensional distribution of dissolved oxygen concentration is consistent with the three-dimensional distribution of dissolved oxygen concentration generated in the aforementioned steps in terms of data structure, spatial coverage, and temporal resolution, thereby ensuring the comparability of the data before and after in terms of spatiotemporal reference and laying the foundation for accurate calculation of contribution. The core physical significance of the re-acquired three-dimensional distribution of dissolved oxygen concentration lies in depicting the latest three-dimensional spatial state of dissolved oxygen concentration inside groundwater after the directional controlled gas injection disturbance.

[0059] In step 105, a contribution matrix is ​​constructed to quantify the impact of controllable gas injection disturbances at each stopping point on the spatial distribution of dissolved oxygen across the entire region. The contribution matrix is ​​then input into a pre-constructed large-scale ecological environment model to obtain groundwater pollution risk early warning information.

[0060] In some embodiments, constructing a contribution matrix that quantifies the impact of controllable gas injection perturbations at each dwell point on the spatial distribution of dissolved oxygen across the entire domain can be achieved through the following steps: The difference between the three-dimensional distribution of dissolved oxygen concentration after the disturbance and the three-dimensional distribution of dissolved oxygen concentration before the disturbance is calculated to obtain the global dissolved oxygen increment field. The contribution ratio of the controllable gas injection disturbance at each stopping point to the increase in dissolved oxygen of each voxel in the whole domain is calculated based on the oxygen diffusion influence domain of each candidate stopping point. Based on the contribution ratio and the global dissolved oxygen increment field, the total contribution of each stopping point is calculated, and a contribution matrix is ​​constructed.

[0061] In practice, the difference between the three-dimensional distribution of dissolved oxygen concentration after perturbation and the three-dimensional distribution of dissolved oxygen concentration before perturbation is calculated to obtain the global dissolved oxygen increment field. This can be achieved in the following way: For example, traverse all voxels in the entire domain, subtract the dissolved oxygen concentration value of the corresponding voxel before perturbation from the dissolved oxygen concentration value of each voxel after perturbation, and obtain the dissolved oxygen increment value of that voxel; the dissolved oxygen increment values ​​of all voxels constitute the global dissolved oxygen increment field; for voxels with negative dissolved oxygen increment values, it indicates that the dissolved oxygen concentration of that voxel has decreased after perturbation, which may be due to pollutants carried by upstream water or oxygen consumption by the medium, and it is regarded as 0 when calculating the total contribution.

[0062] In practice, the contribution ratio of the controllable gas injection disturbance at each candidate stopping point to the dissolved oxygen increment of each voxel in the entire domain can be calculated based on the oxygen diffusion influence domain of each stopping point. This can be achieved in the following way: For each stopping point in this disturbance, the oxygen diffusion influence domain determination method described above is called, and the actual controllable gas injection intensity, actual injection time, seepage velocity, and water temperature data of the stopping point are input. The theoretical dissolved oxygen increment of the controllable gas injection disturbance at the stopping point to each voxel in the entire domain is calculated. For each voxel, the theoretical dissolved oxygen increments of all stopping points to that voxel are added together to obtain the total theoretical dissolved oxygen increment of that voxel. The theoretical dissolved oxygen increment of each stopping point to that voxel is divided by the total theoretical dissolved oxygen increment to obtain the contribution ratio of the stopping point to the dissolved oxygen increment of that voxel. If the total theoretical dissolved oxygen increment is 0, then all contribution ratios of that voxel are 0.

[0063] In specific implementation, the total contribution of each stationing point is calculated based on the contribution ratio and the global dissolved oxygen increment field. The contribution matrix can be constructed in the following way: for each stationing point, all voxels in the global domain are traversed, the contribution ratio of the stationing point to each voxel is multiplied by the actual dissolved oxygen increment value of the voxel, and then all products are added together to obtain the total contribution of the stationing point; the total contributions of all stationing points are arranged according to the stationing point number to construct the contribution matrix; the contribution matrix is ​​a one-dimensional matrix, and each element of the matrix corresponds to the total contribution of a stationing point.

[0064] It should be noted that the contribution matrix mentioned in this application refers to a digital matrix used to quantify the total contribution of each stationing point in a single controlled gas injection disturbance to the overall spatial distribution of dissolved oxygen. Its physical meaning is to characterize the actual dissolved oxygen response effect of the controlled gas injection disturbance at each stationing point during this monitoring process.

[0065] In some embodiments, the contribution matrix can be input into a pre-constructed large-scale ecological environment model to obtain groundwater pollution risk early warning information, which can be achieved by the following steps: The contribution matrix, the characteristic parameters of the seepage flow field, and the characteristic parameters of the medium oxygen consumption flux field are fused together to form the input feature data of the ecological environment large model. The input feature data is input into the ecological environment big model, and the ecological environment big model performs inversion analysis on the input feature data based on the pollution response law it has pre-learned internally to obtain the pollution risk level of each oxygen-deficient voxel set; Based on the pollution risk level and spatial coordinates of each oxygen-deficient voxel set, groundwater pollution risk early warning information containing risk level, spatial location and confidence level is generated.

[0066] In specific implementation, the contribution matrix, the feature parameters of the seepage flow field, and the feature parameters of the medium oxygen consumption flux field are fused to form the input feature data of the ecological environment large model. This can be achieved in the following way: In the central processing unit, the contribution matrix is ​​arranged according to the stationary point number to form a first feature vector; the seepage velocity and direction data at each planned stationary point are extracted from the seepage flow field to form a second feature vector; the medium oxygen consumption flux data at each planned stationary point are extracted from the medium oxygen consumption flux field to form a third feature vector; the first feature vector, the second feature vector, and the third feature vector are concatenated according to a preset dimension, and the concatenated vector is normalized to obtain the input feature data. As a preferred embodiment, historical monitoring data and pollution source distribution data can also be integrated into the input feature data to enrich the contextual information of the ecological environment large model. This application does not limit this.

[0067] In specific implementation, the input feature data is input into the ecological environment big data model. The ecological environment big data model, based on its pre-learned pollution response patterns, performs inversion analysis on the input feature data to obtain the pollution risk level of each oxygen-deficient voxel set. This can be achieved in the following way: the ecological environment big data model is a deep learning model pre-trained based on historical pollution event data and groundwater monitoring data. The deep learning model includes an input layer, an encoder, a decoder, and an output layer. The encoder adopts a multi-layer Transformer structure to extract the spatiotemporal correlation features in the input feature data. The decoder, based on the spatiotemporal correlation features, inverts the pollution response intensity of each oxygen-deficient voxel set voxel by voxel. The output layer maps the pollution response intensity to a preset risk level label. The risk level label includes three levels: low risk, medium risk, and high risk. The threshold for each level is preset according to the environmental quality judgment standard.

[0068] The training process of the ecological environment large model is as follows: At least three years of historical pollution event records, corresponding contribution matrices, hydrogeological parameters, and known pollution risk level labels are obtained from the watershed where the target area is located, forming a training sample set. The contribution matrix and hydrogeological parameters in the training sample set are used as model inputs, and the corresponding known pollution risk level labels are used as supervision signals. The model parameters are iteratively optimized through a backpropagation algorithm until the model's classification accuracy on the validation set reaches a preset threshold, thus obtaining the trained ecological environment large model. As a preferred embodiment, the encoder can also adopt a graph neural network structure, constructing a graph structure based on the spatial adjacency relationships between voxels; this application does not limit this approach.

[0069] In specific implementation, the generation of groundwater pollution risk early warning information containing risk level, spatial location, and confidence level can be achieved by the following method based on the pollution risk level and spatial coordinates of each oxygen-deficient voxel set. For example, the pollution risk level and central spatial coordinates of each oxygen-deficient voxel set output by the ecological environment model are encapsulated according to a preset early warning information format to generate structured early warning data. The structured early warning data includes a risk level field, a spatial location field, and a confidence level field, wherein the confidence level is calculated by the output layer of the ecological environment model using the Softmax function. The structured early warning data is pushed to the monitoring terminal through an Internet of Things communication module, and different risk levels are identified with different colors in the three-dimensional geographic information system to achieve a visual presentation of the groundwater pollution risk early warning information. As a preferred embodiment, the early warning information also includes suggested control measures, which are determined according to a risk level and a preset control measure mapping table. This application does not limit this aspect.

[0070] It should be noted that the inversion described in this application refers to the process of inversely inferring the spatial distribution characteristics of pollution risk using the contribution matrix as observation data and the ecological environment big data model. The difference between this and conventional forward modeling is that it uses measured response data as constraints, and the output results are the pollution risk level and spatial location of each oxygen-deficient voxel set. The groundwater pollution risk early warning information refers to structured data containing risk level, spatial location and confidence level, which is used to provide quantitative basis for regulatory decision-making.

[0071] In another aspect, in some embodiments, this application provides a groundwater pollution dissolved oxygen response monitoring and early warning system, with reference to... Figure 4 The figure is a schematic diagram of a groundwater pollution dissolved oxygen response monitoring and early warning system according to some embodiments of this application. The groundwater pollution dissolved oxygen response monitoring and early warning system 200 includes: an acquisition module 201, a processing module 202, and an execution module 203, which are described below: The acquisition module 201 in this application is mainly used to acquire the three-dimensional distribution of dissolved oxygen concentration, oxygen consumption flux field and seepage flow field inside the groundwater area; Processing module 202, in this application, is mainly used to determine the dynamic oxygen demand threshold at the downstream protected object, and to identify the connected voxels in the three-dimensional distribution of dissolved oxygen concentration that have a dissolved oxygen concentration lower than the dynamic oxygen demand threshold as oxygen-deficient voxel sets. In addition, the processing module 202 in this application is also used to plan a priority operation path and a sequence of stopping points with the goal of stimulating the oxygen-deficient voxel set based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, and to determine the target excitation layer and controllable gas injection intensity based on the vertical gradient distribution of dissolved oxygen concentration and the medium oxygen consumption flux at each planned stopping point. In addition, the processing module 202 in this application is also used to drive the mobile controllable disturbance platform to implement controllable gas injection disturbance according to the priority operation path, target excitation layer and controllable gas injection intensity. During the disturbance, dissolved oxygen response data at different depths around the stopping point are collected in real time. After the disturbance is completed, the three-dimensional distribution of internal dissolved oxygen concentration in the groundwater area is obtained again. The execution module 203 in this application is mainly used to construct a contribution matrix of the controllable gas injection disturbance at each stationing point on the spatial distribution of dissolved oxygen in the whole region based on the difference in the three-dimensional distribution of dissolved oxygen concentration before and after the disturbance, and input the contribution matrix into the pre-constructed ecological environment model to obtain groundwater pollution risk early warning information.

[0072] In addition, this application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to acquire the code and execute the above-described groundwater pollution dissolved oxygen response monitoring and early warning method.

[0073] In some embodiments, reference Figure 5 This figure is an internal structural diagram of a computer device for implementing a groundwater pollution dissolved oxygen response monitoring and early warning method according to some embodiments of this application. The groundwater pollution dissolved oxygen response monitoring and early warning method in the above embodiments can be implemented through... Figure 5 The computer device shown is used to implement this, and the computer device 300 includes at least one processor 301, a communication bus 302, a memory 303, and at least one communication interface 304.

[0074] The processor 301 may be a general-purpose central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more devices used to control the execution of the groundwater pollution dissolved oxygen response monitoring and early warning method in this application.

[0075] The communication bus 302 is used to transmit information between the aforementioned components.

[0076] Memory 303 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. Memory 303 may exist independently and be connected to processor 301 via communication bus 302. Memory 303 may also be integrated with processor 301.

[0077] The memory 303 stores program code for executing the scheme of this application, and its execution is controlled by the processor 301. The processor 301 executes the program code stored in the memory 303. The program code may include one or more software modules. In the above embodiments, the groundwater pollution dissolved oxygen response monitoring and early warning method can be implemented by the processor 301 and one or more software modules in the program code in the memory 303.

[0078] Communication interface 304 uses any transceiver-like device to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0079] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0080] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device may be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0081] In addition, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described method for monitoring and early warning of dissolved oxygen response in groundwater pollution.

[0082] In summary, the groundwater pollution dissolved oxygen response monitoring and early warning system and method disclosed in this application obtains the three-dimensional distribution of dissolved oxygen concentration, the medium oxygen consumption flux field, and the seepage flow field within the groundwater area; determines the dynamic oxygen demand threshold at the downstream protected object; clusters connected voxels with dissolved oxygen concentrations below the dynamic oxygen demand threshold in the three-dimensional distribution of dissolved oxygen concentration as oxygen-deficient voxel sets; based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, plans a priority operation path and stopping point sequence aimed at stimulating the oxygen-deficient voxel sets; and determines the target based on the vertical gradient distribution of dissolved oxygen concentration and the medium oxygen consumption flux at each planned stopping point. The system defines the excitation layer and the controllable gas injection intensity; drives the mobile controllable disturbance platform to perform controllable gas injection disturbance according to the priority operation path, target excitation layer, and controllable gas injection intensity. During the disturbance, dissolved oxygen response data at different depths around the stopping point are collected in real time. After the disturbance is completed, the three-dimensional distribution of dissolved oxygen concentration in the groundwater area is acquired again. Based on the difference in the three-dimensional distribution of dissolved oxygen concentration before and after the disturbance, a contribution matrix is ​​constructed to quantify the impact of controllable gas injection disturbance at each stopping point on the spatial distribution of dissolved oxygen in the entire area. The contribution matrix is ​​then input into a pre-constructed ecological environment model to obtain groundwater pollution risk early warning information. This improves the energy efficiency of mobile disturbance and reduces ineffective disturbances.

[0083] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0084] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for monitoring and early warning of dissolved oxygen response in groundwater pollution, characterized in that, Includes the following steps: To obtain the three-dimensional distribution of dissolved oxygen concentration, oxygen consumption flux field, and seepage flow field within the groundwater area; Determine the dynamic oxygen demand threshold at the downstream protected object, and identify the connected voxels in the three-dimensional distribution of dissolved oxygen concentration that have a dissolved oxygen concentration lower than the dynamic oxygen demand threshold as oxygen-deficient voxel sets. Based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, a priority operation path and stopping point sequence are planned with the goal of stimulating the oxygen-deficient voxel set. Based on the vertical gradient distribution of dissolved oxygen concentration and the oxygen consumption flux of the medium at each planned stopping point, the target excitation layer and controllable gas injection intensity are determined. The mobile controllable disturbance platform is driven to carry out controllable gas injection disturbance according to the priority operation path, target excitation layer and controllable gas injection intensity. During the disturbance, dissolved oxygen response data at different depths around the stopping point are collected in real time. After the disturbance is completed, the three-dimensional distribution of internal dissolved oxygen concentration in the groundwater area is obtained again. Based on the difference in the three-dimensional distribution of dissolved oxygen concentration before and after the disturbance, a contribution matrix is ​​constructed to quantify the impact of controllable gas injection disturbance at each station on the spatial distribution of dissolved oxygen in the whole region. The contribution matrix is ​​then input into a pre-constructed ecological environment model to obtain groundwater pollution risk early warning information.

2. The method as described in claim 1, characterized in that, Determining the dynamic oxygen demand threshold at the downstream protected object specifically includes: Acquire real-time water quality and seepage parameters related to the functions of downstream protected objects; The theoretical oxygen consumption rate of the downstream protected object is determined based on the real-time water quality and seepage parameters and the preset oxygen consumption kinetic model. The dynamic oxygen demand threshold at the downstream protected object is determined based on the theoretical oxygen consumption rate and the dissolved oxygen safety margin.

3. The method as described in claim 1, characterized in that, Identifying oxygen-deficient voxel sets in the three-dimensional distribution of dissolved oxygen concentration where the dissolved oxygen concentration is below the dynamic oxygen demand threshold specifically includes: The dissolved oxygen concentration of each voxel in the three-dimensional distribution of dissolved oxygen concentration is compared with the dynamic oxygen demand threshold, thereby identifying hypoxic voxels whose dissolved oxygen concentration is lower than the dynamic oxygen demand threshold. Three-dimensional spatial clustering analysis was performed on all hypoxic voxels to obtain multiple sets of oxygen-deficient voxels.

4. The method as described in claim 1, characterized in that, Based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, the planned priority operation path and stopping point sequence aimed at stimulating the oxygen-deficient voxel set specifically include: The mobility and endurance constraints of the mobile controllable disturbance platform are quantified as input parameters for the path planning model; Based on the oxygen diffusion influence domain of each candidate stopping point, the excitation coverage and excitation efficiency of each candidate stopping point are calculated, and a constrained path optimization model is established with the goal of exciting the oxygen-deficient voxel set. Solve the constrained path optimization model to output the preferred operation path and stopping point sequence of the mobile controllable disturbance platform.

5. The method as described in claim 1, characterized in that, Based on the vertical gradient distribution of dissolved oxygen concentration and the oxygen flux of the medium at each planned stopping point, the target excitation layer and the controllable gas injection intensity are determined, specifically including: Extract the characteristic data of the vertical gradient distribution of dissolved oxygen concentration at each planned stopping point; The target incentive level is determined based on the feature data and the preset incentive level selection rules; The controllable gas injection intensity of the mobile controllable disturbance platform is determined based on the degree of oxygen deficiency and the oxygen flux of the medium at the target excitation layer.

6. The method as described in claim 1, characterized in that, Driving the mobile controllable disturbance platform to perform controllable gas injection disturbance according to the priority operation path, target excitation level, and controllable gas injection intensity specifically includes: The preferred operation path, target excitation level, and controllable gas injection intensity are converted into an executable control instruction set for the mobile controllable disturbance platform; The mobile controllable disturbance platform is driven to sequentially complete movement, positioning, and controllable gas injection disturbance according to the executable control instruction set.

7. The method as described in claim 1, characterized in that, The contribution matrix for quantifying the impact of controllable gas injection disturbances at each stopping point on the spatial distribution of dissolved oxygen across the entire region specifically includes: The difference between the three-dimensional distribution of dissolved oxygen concentration after the disturbance and the three-dimensional distribution of dissolved oxygen concentration before the disturbance is calculated to obtain the global dissolved oxygen increment field. The contribution ratio of the controllable gas injection disturbance at each stopping point to the increase in dissolved oxygen of each voxel in the whole domain is calculated based on the oxygen diffusion influence domain of each candidate stopping point. Based on the contribution ratio and the global dissolved oxygen increment field, the total contribution of each stopping point is calculated, and a contribution matrix is ​​constructed.

8. The method as described in claim 1, characterized in that, The contribution matrix is ​​input into a pre-constructed large-scale ecological environment model to obtain groundwater pollution risk early warning information, specifically including: The contribution matrix, the characteristic parameters of the seepage flow field, and the characteristic parameters of the medium oxygen consumption flux field are fused together to form the input feature data of the ecological environment large model. The input feature data is input into the ecological environment big model, and the ecological environment big model performs inversion analysis on the input feature data based on the pollution response law it has pre-learned internally to obtain the pollution risk level of each oxygen-deficient voxel set; Based on the pollution risk level and spatial coordinates of each oxygen-deficient voxel set, groundwater pollution risk early warning information containing risk level, spatial location and confidence level is generated.

9. The method as described in claim 1, characterized in that, The downstream protected objects refer to sensitive targets or ecological purification structures located within the groundwater area that have functional requirements for dissolved oxygen concentration.

10. A groundwater pollution dissolved oxygen response monitoring and early warning system, characterized in that, include: The acquisition module is used to acquire the three-dimensional distribution of dissolved oxygen concentration, oxygen consumption flux field, and seepage flow field within the groundwater area. The processing module is used to determine the dynamic oxygen demand threshold at the downstream protected object and to identify the clusters of connected voxels in the three-dimensional distribution of dissolved oxygen concentration that are lower than the dynamic oxygen demand threshold as oxygen-deficient voxel sets. The processing module is also used to plan a priority operation path and a sequence of stopping points with the goal of stimulating the oxygen-deficient voxel set based on the oxygen diffusion influence domain of each candidate stopping point and the mobility and endurance constraints of the mobile controllable disturbance platform, and to determine the target excitation layer and controllable gas injection intensity based on the vertical gradient distribution of dissolved oxygen concentration and the oxygen consumption flux of the medium at each planned stopping point. The processing module is also used to drive the mobile controllable disturbance platform to implement controllable gas injection disturbance according to the priority operation path, target excitation layer and controllable gas injection intensity. During the disturbance, dissolved oxygen response data at different depths around the stopping point are collected in real time. After the disturbance is completed, the three-dimensional distribution of internal dissolved oxygen concentration in the groundwater area is obtained again. The execution module is used to construct a contribution matrix that quantifies the impact of controllable gas injection disturbance at each station on the spatial distribution of dissolved oxygen in the whole region based on the difference in the three-dimensional distribution of dissolved oxygen concentration before and after the disturbance. The contribution matrix is ​​then input into a pre-constructed ecological environment model to obtain groundwater pollution risk early warning information.

Images