Mine slope turf lattice and microorganism synergistic remediation method and system

Abstract

The application discloses a mine slope turf lattice and microbial synergistic remediation method and system, establishes a quantitative correction relationship of heavy metal pollution concentration on slope body mechanical parameters, determines and calculates the inhibition release threshold value of arbuscular mycorrhizal fungi, identifies and matches the lattice layout position of the colonization diffusion throat node, and constructs a colonization grade partition set; based on the slope runoff directed graph, the failure conduction path of the colonization unit is deduced, the key cut point unit is identified, the construction sequence and construction execution configuration oriented to failure blockage are formed; according to the root layer water potential residence time length, the gradient of the fungicide ratio is planned, the gradient ratio sequence and the regulation scheme set are formed; the gradient reseeding and fungicide critical value is determined through the precursor identification of the cover degree-infection rate deviation trend, the slope synergistic remediation regulation instruction is output, and the accurate space configuration of the remediation resources and the active early warning response of the performance degradation are realized.

Description

Technical Field

[0001] This invention relates to the field of mine ecological restoration technology, and in particular to a method and system for the synergistic restoration of mine slopes using turf and microorganisms. Background Technology

[0002] Abandoned mine slopes generally face the combined pressure of heavy metal pollution accumulation and slope mechanical degradation. Heavy metals disrupt the electrochemical bonds between particles by displacing soil colloidal cations, leading to a continuous decline in cohesion as pollution intensifies. Existing remediation projects conduct pollution assessments and mechanical stability assessments independently, and the selection of lattice framework sites relies on topographic slope and engineering experience, failing to reflect the spatial distribution of pollution-mechanical coupling degradation, thus limiting the targeted allocation of lattice support resources.

[0003] The combined inoculation of arbuscular mycorrhizal fungi and growth-promoting bacteria provides a way to improve grass seed planting conditions under heavy metal stress. However, the determination of the inoculation ratio relies on static soil type maps, which fail to reflect the constraints of spatiotemporal fluctuations in root water potential driven by rainfall and evaporation on the activity of the inoculants. The inoculation effect varies across different slope locations due to differences in water conditions. Monitoring of remediation effects mainly relies on periodic surveys of vegetation cover, lacking a mechanism to identify the underground pre-decline signal of mycorrhizal infection rate decline preceding vegetation cover degradation. Reseeding and inoculant interventions are usually triggered only after visible degradation occurs, resulting in insufficient long-term stability of the remediation. Summary of the Invention

[0004] This invention discloses a method and system for the co-remediation of mine slopes using turf grids and microorganisms. It aims to transform the coupling effect of heavy metal pollution distribution and slope mechanical degradation into a basis for the precise colonization and configuration of grid nodes. Through topological analysis of failure cascade paths, it generates a failure-blocking-oriented construction sequence and execution configuration. Furthermore, it utilizes root water potential-driven inoculation gradient planning and identification of early signs of underground degradation to achieve full-cycle coordinated control from construction configuration to long-term maintenance. This provides a methodological support for mine slope turf grid and microbial co-remediation projects that balances precise spatial configuration and dynamic response capabilities.

[0005] The first aspect of this invention proposes a method for the synergistic restoration of mine slopes using turf lattice and microorganisms, comprising the following steps: Collect heavy metal pollution detection data and slope mechanical parameters, and construct a coupled feature set by performing pollution-slope stability coupling analysis on the heavy metal pollution detection data and the slope mechanical parameters. Based on the coupled feature set, a colonization self-enhancing analysis is performed to generate a colonization potential map. The colonization core node is identified in the colonization potential map to generate a colonization node set. Based on the colonization node set, the colonization diffusion range is calibrated to construct a colonization level partition set. A failure propagation analysis of the colonization unit is performed on the colonization level partition set to determine the failure blocking succession table. The construction priority is sorted according to the failure blocking succession table to generate a construction sequence. The key lattice node parameters are parsed based on the construction sequence to form a construction execution configuration. A root layer water potential fluctuation analysis is performed on the construction execution configuration to generate a water potential feature map. Based on the water potential feature map, inoculation gradient planning is implemented to obtain a tiered mix ratio sequence. Based on the tiered mix ratio sequence, an enhanced section and a conventional section are identified to construct a set of control schemes. A decay trend map is generated by analyzing the coverage-infection rate effectiveness decay trend of each colonization unit corresponding to the application sequence. Based on the decay trend map, the characteristics of the decay phase transition precursor are identified to determine the critical value of tiered reseeding and inoculum replenishment. Based on the control scheme set and the critical value of tiered reseeding and inoculum replenishment, a multi-threshold hierarchical decision-making process for reseeding and inoculum replenishment is performed to output the slope collaborative restoration control command.

[0006] The second aspect of this invention proposes a synergistic restoration system for turf and microorganisms on mine slopes, comprising: The data parsing module is used to collect heavy metal pollution detection data and slope mechanical parameters, and to perform pollution-slope stability coupling analysis on the heavy metal pollution detection data and the slope mechanical parameters to construct a coupling feature set; The partitioning construction module is used to generate a colonization potential map by performing colonization self-enhancing analysis based on the coupled feature set, identify colonization core nodes in the colonization potential map to generate a colonization node set, and construct a colonization level partitioning set by calibrating the colonization diffusion range based on the colonization node set. The sequence planning module is used to perform colonization unit failure propagation analysis on the colonization level partition set to determine the failure blocking succession table, sort the implementation priority according to the failure blocking succession table to generate an implementation sequence, and parse the key lattice node parameters based on the implementation sequence to form a construction execution configuration. The mix proportion control module is used to perform root layer water potential fluctuation analysis on the construction execution configuration to generate a water potential feature map, implement inoculation gradient planning based on the water potential feature map to obtain a tiered mix proportion sequence, and identify the enhanced section and the conventional section according to the tiered mix proportion sequence to construct a set of control schemes; The instruction output module is used to perform coverage-infection rate effectiveness decay trend analysis on each colonization unit corresponding to the application sequence to generate a decay trend map, identify the characteristics of the decay phase transition precursor based on the decay trend map to determine the critical value of tiered reseeding and inoculation, and perform multi-threshold hierarchical decision-making for reseeding and inoculation based on the control scheme set and the critical value of tiered reseeding and inoculation to output slope collaborative restoration control instructions.

[0007] The beneficial effects of this invention are reflected in the following points: 1. The quantitative weakening of soil cohesion and the correction relationship of heavy metal concentration through electrochemical pathways are incorporated into the coupled analysis, thereby driving the spatial determination of the inhibition and release threshold of arbuscular mycorrhizal fungi. This allows for joint analysis of pollution patterns and mycorrhizal colonization domains within the same coordinate framework, unifying the two types of degradation signals from independent evaluations into a unified coupled feature input. In the selection of grid nodes, the concept of diffusion connectivity gain is introduced, jointly quantifying the independent coverage contribution of candidate nodes to the slope blank area and their infiltration distance to the core of the diffusion blank area. This identifies the colonization throat node that covers the largest blank area with the fewest nodes, changing the previous problem of grid site selection being disconnected from ecological restoration needs. 2. Traditional construction sequence planning is guided by slope gradient zoning or construction convenience. This invention treats the failure of colonization units as a cascading event extending along the runoff conduction path. By using a cut point identification algorithm to extract key blocking nodes from the failure conduction path network, the construction sequence is transformed from terrain-driven to failure network topology-driven. Simultaneously, a replacement unit with a non-common path is assigned to each critical cut point unit, so that the blocking function when the main unit is missing is continued by the replacement unit in the adjacent time sequence, forming an application priority system with network redundancy and fault tolerance, extending the risk resistance of the restoration project from single-point blocking to the network cascading level. 3. The annual residence time of root layer water potential within the appropriate range is incorporated into the ratio planning. The ratio gradient of Bacillus and arbuscular mycorrhizal fungi is differentiated according to the water dynamics of each lattice node. The inoculation sequence matches the slope water potential change pattern, and the limitation of using static soil texture type as the basis for ratio is avoided. The divergence between the decline in mycorrhizal infection rate and vegetation cover degradation is extracted as the intervention trigger signal. The time window of decoupling of underground mycorrhizal symbiotic structure before the appearance of visible symptoms on the ground is captured, and the reseeding and inoculum intervention is moved forward to the early stage when the underground mycorrhizal system can still be intervened for restoration. Attached Figure Description

[0008] The accompanying drawings illustrate specific examples of the technical solutions described in this invention and, together with the detailed embodiments, form part of the specification, serving to explain the technical solutions, principles, and effects of this invention.

[0009] Figure 1 This is a schematic diagram of the process of a method for the synergistic restoration of mine slopes using turf and microorganisms, according to the present invention.

[0010] Figure 2 This is a structural block diagram of a mine slope turf grid and microbial synergistic restoration system according to the present invention. Detailed Implementation

[0011] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0012] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0013] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0014] The technical solutions of the embodiments of this application will be described below.

[0015] like Figure 1 As shown, this embodiment of the invention provides a method for the synergistic restoration of mine slopes using turf lattice and microorganisms, including the following steps S110-S150: Step S110: Collect heavy metal pollution detection data and slope mechanical parameters, and construct a coupled feature set by performing pollution-slope stability coupling analysis on the heavy metal pollution detection data and slope mechanical parameters.

[0016] Specifically, heavy metal pollution detection data and slope mechanical parameters were collected. Grid spacing was set according to the slope area and the density of pollution sources, with denser grids in areas with concentrated pollution sources to capture local pollution peaks. Each grid node simultaneously recorded its planar coordinates and ground elevation. The ground elevation was obtained from on-site topographic surveys or a digital elevation model and used for subsequent slope runoff path calculations and aspect analysis. Each sampling point was sampled in two layers: 0-20 cm and 20-40 cm. This layered sampling ensured the representativeness of heavy metal pollution detection data in the vertical direction. ICP-MS analysis of samples from each sampling point yielded the content values ​​of target heavy metal elements such as copper, lead, zinc, and cadmium. pH and organic matter content were simultaneously measured. pH affects the speciation and bioavailability of heavy metals in the soil, while organic matter content affects the degree of complexation and fixation of heavy metals with soil particles. These two auxiliary indicators were incorporated into the records of each sampling point along with the heavy metal pollution detection data, carrying coordinates and stratigraphic labels. The determination of slope mechanical parameters employed a dual-track approach, combining in-situ direct shear tests and indoor triaxial compression tests. In-situ direct shear tests obtained the internal friction angle φ and cohesion c of the undisturbed soil, while triaxial compression tests supplemented the strength envelope data under different confining pressure conditions. The calibrated confining pressure strength envelope was used for back-calculation of anchorage depth. Unit weight γ and natural moisture content ω were obtained by in-situ sampling using the ring sampler method, followed by drying and weighing. The porosity n at each sampling point was calculated based on the soil particle specific gravity determined by the unit weight and hydrometer bottle method. These four slope mechanical parameters, along with porosity and sampling point coordinates, constituted the mechanical record for each location. Heavy metal pollution detection data and slope mechanical parameters were aligned with the slope grid node coordinates. Paired data with coordinate deviations exceeding half a grid spacing were labeled with positional deviations. These labeled entries were weighted down during the coupled analysis stage for spatial interpolation to avoid systematic pairing errors introduced by coordinate offsets.

[0017] A coupled feature set was constructed by analyzing heavy metal pollution detection data and slope mechanical parameters to construct a pollution-slope stability coupling analysis. The lead, zinc, and cadmium contents (from the 0 to 20 cm root zone) of each grid node were subjected to single-factor regression with the cohesion *c* of the corresponding node. Nodes with larger absolute values ​​of regression slopes showed a significant inhibitory effect on cohesion. Stepwise regression was then used to screen and determine the element combinations for inclusion in the model. The heavy metal pollution detection data of each node were substituted into the correction equation to calculate the corrected cohesion point by point. Nodes with correction values ​​exceeding 30% of the original cohesion were designated as high-correction-intensity nodes and given priority inclusion in stability-sensitive areas in subsequent colonization zoning. The regression slopes of lead content and cohesion in the wastewater discharge slope section were typically higher than those of zinc and cadmium, because lead ions preferentially replace calcium and magnesium ions on the soil colloid surface and disrupt the electrochemical bonds between particles, resulting in stronger inhibition of cohesion at the same concentration increment. The internal friction angle φ is corrected by using the product of moisture content ω and pH as a joint predictor variable. The decay of φ accelerates when low pH and high moisture content coexist. The correction coefficient is higher in the waterlogged area at the slope toe than in the dry area at the slope top, characterizing the spatial non-uniformity of the mechanical degradation response to pollution. The corrected cohesion and corrected internal friction angle are extended to unsampled locations across the entire slope using Kriging interpolation. Areas where the interpolation variance exceeds a set accuracy threshold are marked as interpolation uncertainty areas, which are replaced by the mean of the corrected parameters from adjacent deterministic nodes. Replacement is only enabled when the continuous range of the uncertainty area does not exceed two grid cells; otherwise, the corresponding area is marked with low confidence. Heavy metal pollution detection data, organic matter, porosity, ground elevation, corrected cohesion, corrected internal friction angle, and high-correction-intensity and low-confidence labels for each grid node are summarized by coordinate and incorporated into the coupling feature set as a unified representation of the pollution-slope stability coupling analysis, for subsequent self-reinforcing colonization analysis, slope runoff path calculation, and root layer water-holding capacity measurement.

[0018] Step S120: Based on the coupling feature set, perform colonization self-enhancing analysis to generate a colonization potential map, identify the colonization core nodes in the colonization potential map to generate a colonization node set, and construct a colonization level partition set based on the colonization node set to calibrate the colonization diffusion range.

[0019] In some embodiments, the step of generating a colonization potential map based on the colonization self-enhancing analysis of the coupled feature set includes: determining the inhibition release threshold of arbuscular mycorrhizal fungi on the coupled feature set to generate an inhibition release threshold map; calculating the tillering diffusion radius of grass seeds based on the inhibition release threshold map to obtain a grass seed diffusion radius map; and performing overlay analysis on the inhibition release threshold map and the grass seed diffusion radius map to form a colonization potential map.

[0020] The inhibition release threshold of arbuscular mycorrhizal fungi was determined using the coupled feature set to generate an inhibition release threshold map. The concentrations of copper, lead, zinc, and cadmium at each grid node in the coupled feature set were compared node-by-node with the single-element inhibition critical concentration. The inhibition critical concentrations were typically between 150 and 200 mg / kg for copper, 300 to 400 mg / kg for lead, 5 to 10 mg / kg for cadmium, and 200 to 300 mg / kg for zinc. Nodes with an exceedance rate exceeding 1.0 had triggered individual inhibition for that element, while nodes with an exceedance rate below 0.6 had essentially eliminated the inhibitory effect. The combined inhibition effect of multiple elements was calculated using the toxicity unit summation method: TU = Σ(Ci / EC). 50i ), where Ci is the measured concentration of the i-th heavy metal at this node in the coupling feature set, and EC 50i The critical concentration (TU) for this element to inhibit the colonization rate of arbuscular mycorrhizal fungi by 50% is defined as follows: TU below 0.8 indicates the joint inhibition is lifted, and arbuscular mycorrhizal fungi regain normal colonization ability; TU between 0.8 and 1.5 indicates partial inhibition, with colonization ability decreasing to 30% to 70% of normal levels, the specific percentage determined by linear interpolation of TU within this range; TU above 1.5 indicates complete inhibition, with colonization ability returning to zero. Nodes in the coupling feature set carrying high-correction-strength annotations (with corrected cohesion decreasing by more than 30% compared to the original cohesion) are considered in the TU calculation for the EC50 of cadmium and lead. 50 The parameters were lowered by 10% because the combined damage of heavy metals to soil structure further reduced the tolerance of arbuscular mycorrhizal fungi to the same heavy metal concentrations. The TU values ​​and corresponding inhibition states of each grid node were plotted on the slope grid to form an inhibition relief threshold map. In the tailings dam leakage area, the lead and cadmium concentrations in the coupled feature set were synchronously high, and the TU values ​​formed a high-value cluster in this area. The boundary between the completely inhibited area and the partially inhibited area in the inhibition relief threshold map matched the contour lines of the composite pollution in the coupled feature set. The dominant inhibiting element was recorded synchronously at each node. The dominant inhibiting element was the one with the largest exceedance ratio among all elements and was retained as the main control factor of composite pollution at that node along with the inhibition relief threshold map.

[0021] The grass seed diffusion radius map is obtained by calculating the grass seed tiller diffusion radius based on the inhibition release threshold map. In the suppression and release threshold diagram, the TU values ​​of each node and the cohesion after the coupling feature set are respectively introduced with chemical stress correction coefficient f_chem and physical stability correction coefficient f_mech. The grass seed diffusion radius R = R0 × min(f_chem, f_mech), where R0 is the baseline tillering radius of the target grass seed on the uncontaminated reference slope. For bermudagrass, R0 is usually 25 to 40 cm, and for tall fescue, it is 15 to 25 cm. f_chem decreases linearly with the increase of TU. When TU is 0, it is taken as 1.0. When TU is 1.5, it is taken as 0.3. When TU exceeds 1.5, the grass seed diffusion radius is forcibly reduced to zero. f_mech decreases linearly with the decrease of the corrected cohesion. When the cohesion is equal to the reference point level, f_mech is taken as the upper limit of 1.0. When the cohesion drops to less than 10% of the reference point, f_mech is taken as the lowest value of 0.4. The smaller value of the two correction coefficients is taken so that the grass seed diffusion radius simultaneously meets the dual constraints of chemical and physical stress. The spatial distribution of R values ​​at each node in the grass seed diffusion radius map characterizes the combined constraint effect of the inhibition release threshold map and the cohesion after correction by the coupling feature set. Within the same TU classification zone of the inhibition release threshold map, spatial variations still exist in the R values. The source of these variations is the uneven distribution of the corrected cohesion within that classification zone. The spatial deviation between the TU isosurface and the R isosurface is a direct mapping of the asynchronous mechanical degradation and pollution distribution areas within the coupling feature set. In the middle of the slope, the TU values ​​in the inhibition release threshold map are in the partially inhibited range. The corrected cohesion decreases significantly due to lead contamination, with both f_chem and f_mech less than 1.0. In this section, the R values ​​in the grass seed diffusion radius map are compressed to 40% to 60% of the baseline value, showing a significant difference from the R values ​​in the upper part of the slope where TU is low and cohesion is close to normal. In the grass seed diffusion radius map, the area where the R value drops sharply to zero for more than two consecutive grids is called the diffusion fault zone. The diffusion fault zone divides the slope into several connected and independent diffusion zones. The mean distribution of R values ​​in each zone serves as a representative indicator of the colonization and diffusion potential of that zone. The grass seed diffusion radius map and the inhibition release threshold map share the same grid coordinate system, and the spatial pairing of the overlay analysis is thus aligned.

[0022] An overlay analysis of the inhibition relief threshold map and the grass seed diffusion radius map was performed to form a colonization potential map. The arbuscular mycorrhizal fungi colonization capacity coefficient f_AMF for each node in the inhibition relief threshold map was converted from the TU value. When TU was below 0.8, f_AMF was set to 1.0; linear interpolation was performed between 0.8 and 1.5; and 0 was set when TU was above 1.5. The basic overlay value was obtained by multiplying the R value of each node in the grass seed diffusion radius map with f_AMF. This basic overlay value combines the constraints of the inhibition relief threshold map on the arbuscular mycorrhizal fungi colonization capacity with the constraints of the grass seed diffusion radius map on the grass seed spatial expansion capacity. A larger product indicates a higher potential for grass-fungi synergistic coverage at that node. The base superposition value is superimposed with a self-reinforcing coefficient. The self-reinforcing coefficient of each node is determined by the weighted average of f_AMF of adjacent eight-directional nodes. When the weighted average exceeds 0.5, the self-reinforcing coefficient is superimposed with a gain of 10% to 20% on the base superposition value. The gain magnitude is linearly positively correlated with the weighted average to simulate the promoting effect of arbuscular mycorrhizal fungal hyphae network on new colonization points after forming connecting bridges between adjacent colonization units. Nodes with high R values ​​in the grass seed diffusion radius map have high root density, and adjacent nodes with high f_AMF contribute more significantly to their self-reinforcing effect. After the superposition values ​​of each node are calculated across the entire slope, they are pieced together in order of numerical magnitude to form a colonization potential map. High-value areas fall on slope locations with low inhibition release threshold (TU), large grass seed diffusion radius (R), and strong f_AMF of adjacent nodes. Low-value areas correspond to completely inhibited areas and areas covered by diffusion fault zones. Several transitional nodes in the middle of the pollution gradient zone, with TU values ​​just below 0.8, have full f_AMF values ​​and high R values. After the self-reinforcing coefficient is superimposed, the superimposed value of these nodes is significantly higher than that of isolated low-pollution nodes, forming a high-potential ridge line extending along the pollution gradient zone in the colonization potential map. The ridge line follows the tangent direction of the TU isosurface in the inhibition release threshold map. Continuous areas in the colonization potential map with potential values ​​below the colonization threshold are delineated as diffusion blank areas based on their spatial connectivity, and are independently identified as connected components, with each connected component independently included in the spatial classification.

[0023] In some embodiments, the step of identifying colonization core nodes and generating a colonization node set from the colonization potential map includes: detecting potential throat points in the colonization potential map to generate a potential throat point set; performing lattice node position matching based on the potential throat point set to obtain a candidate node set; evaluating the connectivity gain of the candidate node set in the diffusion blank area to generate connectivity gain parameters; and selecting the optimal connectivity throat node based on the connectivity gain parameters to determine the colonization node set.

[0024] Potential throat points are generated by detecting potential throat points in the colonization potential map. The potential gradient of each node in the colonization potential map is calculated by taking the absolute value of the difference between the potential values ​​of that node and its four neighboring nodes and then averaging the values. Nodes whose gradient mean exceeds 0.8 times the global potential standard deviation of the colonization potential map are identified as gradient abrupt change points. Among these gradient abrupt change points, nodes whose local potential mean is higher than the mean within two grid ranges of their surrounding area are selected as candidate throat points. Candidate throat points are located at the transitional position between high and low potential value regions in the colonization potential map. Candidate throat points must also satisfy the condition that their removal reduces the potential connectivity between adjacent high-potential and low-potential regions by more than 15%. Connectivity is calculated for each candidate point using the maximum flow algorithm for eight connected regions. The maximum flow is taken as the flow capacity of each node in the colonization potential map. A sharp drop in the maximum flow value after the removal of a candidate throat point indicates that the node has an irreplaceable relay role in connecting the regions on both sides. Candidate points that do not meet the connectivity reduction condition are removed from the potential throat point set. The potential throat point set records the dominant direction of gradient abrupt changes at each throat point. The dominant direction is determined by the direction with the largest absolute value among the four gradient directions. The dominant direction labeling depicts the diffusion blocking direction of each throat point in the colonization potential map, helping to determine the consistency between the grid beam orientation and the potential diffusion direction. In the area where the pollution gradient zone and the remediation grid intersect, the gradient abrupt change points in the colonization potential map are densely distributed. After the removal of candidate throat points in this area, the maximum flow values ​​in both high and low potential zones drop sharply by more than the threshold, and all of them are included in the potential throat point set. The spatial density of the potential throat point set in this area is much higher than in other sections of the slope. This intersection area is the core control zone for the colonization diffusion connectivity of the entire slope.

[0025] Candidate node sets are obtained by performing lattice node position matching based on the potential throat point set. Candidate positions for lattice nodes are generated using the coordinates of each throat point in the potential throat point set as seeds. The candidate position is the coordinate of the throat point itself. Adjacent throat points whose Euclidean distance is less than the aggregation radius are merged into one candidate position. The coordinates of the merged position are the weighted centroid of the potential values ​​of all participating throat points. The aggregation radius is set according to a certain proportion of the median nearest neighbor distance of all throat points in the potential throat point set. If the aggregation radius is too large, adjacent throat points will be excessively merged, resulting in sparse candidate positions; if the aggregation radius is too small, merging will be insufficient, resulting in redundant candidate positions. The proportion is chosen to ensure that the aggregated candidate positions are neither sparsely distributed, omitting throat control areas, nor redundantly overlapping on the slope. The comprehensive throat strength of each candidate position is the weighted average of the potential values ​​of all participating throat points in the colony potential map. The weight is the reciprocal of the distance between each throat point and the centroid of the candidate position. Throat points with closer distances contribute more to the comprehensive strength calculation, and candidate positions with higher comprehensive strength have higher priority in the candidate node set. When multiple throat points are merged into the same candidate location, the dominant direction is taken as the vector synthesis result of the dominant directions of each throat point. If the dispersion of the directional angle of the vector synthesis result exceeds a set threshold, the merged group is split into two candidate locations to avoid the dominant directions canceling each other out and causing the candidate location to lose its diffusion blocking direction. Each candidate location uses the potential value of the participating throat points in the colonization potential map and the dominant direction labeling information to form a candidate node set. When the potential value distribution of the candidate node set is uneven, it indicates that the throat control density is limited in some areas of the slope, and the overall throat strength of the candidate nodes in the corresponding area is low.

[0026] For example, the step of evaluating the connectivity gain of the candidate node set in the diffusion blank area to generate connectivity gain parameters includes: performing a two-way penetration path deduction of grass and bacteria in the surrounding blank area of ​​the candidate node set to generate a penetration reachability map; removing redundant segments with overlapping coverage from the penetration reachability map to obtain a non-redundant coverage increment set; correcting the reachability boundary distance of the non-redundant coverage increment set to obtain a corrected gain set; and selecting the node with the best connectivity contribution based on the corrected gain set to determine the connectivity gain parameters.

[0027] A bidirectional infiltration path for grass and fungi in the surrounding blank areas of the candidate node set is generated to create an infiltration reachability map. The grass seed infiltration path starts from each node in the candidate node set and extends gradually along the direction of minimum slope resistance. The extension distance at each step is the R value of the corresponding position in the grass seed diffusion radius map. The grass seed infiltration path terminates when it encounters a grid node in the inhibition relief threshold map where the TU exceeds the complete inhibition threshold. During the extension process, the coordinates of the nodes reached at each step and the corresponding R value are recorded simultaneously. A decreasing R value indicates an area where infiltration conditions are gradually deteriorating. The arbuscular mycorrhizal fungi infiltration path extends towards high porosity areas along the soil porosity gradient direction. The porosity gradient is calculated by the difference between the porosity of each node in the coupled feature set and the porosity of its four neighboring nodes. When encountering a node where the TU exceeds the complete inhibition threshold, the path deflects and attempts to bypass the suboptimal porosity gradient direction. The arbuscular mycorrhizal fungi infiltration path terminates when all four directions exceed the inhibition threshold. The coordinates of the nodes reached at each step of the grass seed infiltration path and the arbuscular mycorrhizal fungus infiltration path are combined. Nodes reached by both types of paths are denoted as bidirectional cooperative infiltration nodes, and those reached only in one direction are denoted as unidirectional infiltration nodes. Together, they form the infiltration reachability domain of the candidate node. The infiltration reachability domains of all nodes in the candidate node set are grouped and collected into an infiltration reachability domain map according to the node identifier. The spatial assignment of bidirectional and unidirectional infiltration nodes is stored in the map using each candidate node as an index.

[0028] The non-redundant coverage increment set is obtained by removing overlapping redundant sections from the penetration reachability map. The reachability node sets of each candidate node in the penetration reachability map are compared. Nodes belonging to the reachability domains of two or more candidate nodes are identified as redundant nodes. Redundant sections are removed in descending order of overall throat strength. Candidate nodes with high overall throat strength retain their coverage sovereignty in the penetration reachability map, while candidate nodes with low overall throat strength relinquish their coverage sovereignty at redundant nodes overlapping with high-strength nodes. The relinquishment of coverage sovereignty is performed pairwise for each candidate node sequentially. After each pair is executed, the attribution label of the redundant node in the penetration reachability map is updated. The updated penetration reachability map then proceeds to the next pair of candidate nodes for redundancy removal, ensuring the entire removal process is based on the real-time attribution status of the penetration reachability map. When three or more candidate nodes simultaneously cover the same redundant node, their assignment is determined sequentially from highest to lowest comprehensive throat strength, ultimately assigning the node to the candidate with the highest strength. The non-overlapping nodes remaining after deducting abandoned sections in the permeability reachability map constitute the non-redundant coverage increment for each candidate node. The total number of non-redundant coverage increment nodes for each candidate node is the non-redundant coverage increment set. In densely populated areas of slope grid nodes, when the differences in comprehensive throat strength among candidate nodes are small, the number of redundant removal rounds increases significantly. The sensitivity of the removal results to the differences in comprehensive throat strength between adjacent candidate nodes increases as the strength difference narrows. In sparsely populated areas of slope grid nodes, the spatial intervals between candidate nodes are large, the degree of overlap in permeability reachability is low, and the number of redundant nodes is small; therefore, the non-redundant coverage increment set is not significantly different from the original reachability. The redundant node removal process does not change the geometry of the original coverage area in the permeability reachability map of each candidate node; only the assignment label of the redundant nodes is adjusted. The original reachability geometry information is retained as an independent field for subsequent reachability boundary distance correction stages, and the non-redundant coverage increment set serves as the calculation basis for this stage.

[0029] The corrected gain set is obtained by correcting the reachability domain boundary distance of the non-redundant coverage increment set. The number of non-redundant coverage increment nodes for each candidate node in the non-redundant coverage increment set is the net coverage. The slope distance between the boundary node closest to the centroid of the diffusion blank area in the non-redundant coverage area of ​​each candidate node in the non-redundant coverage increment set and the centroid of the diffusion blank area is taken as the correction benchmark. The boundary distance correction coefficient is determined by dividing this distance by the maximum infiltration path length of the candidate node and taking the inverse function. The smaller the ratio, the larger the correction coefficient. The corrected gain value is obtained by multiplying the net coverage of each candidate node in the non-redundant coverage increment set by the correction coefficient. When there are multiple independent diffusion blank areas on the slope, the boundary distance correction for each candidate node for different blank areas is calculated separately and then summed according to the area of ​​each blank area. Blank areas with larger areas have higher weights to prevent candidate nodes from obtaining excessively high overall corrected gain values ​​by only covering small blank areas. Candidate nodes with zero net coverage in the non-redundant coverage increment set have their correction gain values ​​set to zero. Zero-value entries are retained in the correction gain set with their original node identifier. A distinction is made between zeroing due to redundancy removal and zeroing due to the complete inhibition of the infiltration path by the suppression threshold map. The correction gain values ​​of each candidate node are arranged from highest to lowest to form a correction gain set. For example, a candidate node in the middle section of the slope may have a low net coverage in the non-redundant coverage increment, but its coverage area boundary node is close to the centroid of the diffusion blank area, and its boundary distance correction coefficient is high. Therefore, the correction gain value of this node in the correction gain set is higher than that of candidate nodes with larger net coverage but more remote coverage locations. The difference between the ranking of net coverage and the ranking in the correction gain set is recorded in the ranking offset field, which is only for manual verification.

[0030] The connectivity gain parameter is determined by selecting the node with the best connectivity contribution based on the correction gain set. Candidate nodes with correction gain values ​​below 30% of the global mean in the correction gain set are considered low-contribution nodes and are removed. Low-contribution nodes typically correspond to candidate nodes in the permeability reachability map where the coverage is zero or the net coverage is extremely low due to blockage in the suppressed area. The removal process is a quality screening stage and does not involve spatial redistribution. After quality screening, the remaining candidate nodes are sorted from high to low according to their correction gain values ​​to form an ordered candidate node queue. Each node in the queue carries its reachability map coverage coordinate index for subsequent connectivity gain evaluation. When the difference in correction gain values ​​between adjacent candidate nodes in the correction gain set is less than 5%, they are included in the node with the best connectivity contribution. When the difference exceeds 5%, they are strictly truncated according to their ranking. The 5% tolerance threshold for tying is appropriately narrowed as the total area of ​​the slope diffusion blank area increases to avoid excessive expansion of the number of optimal nodes when the blank area is large. When the number of actually available candidate nodes in the correction gain set is less than the number of target nodes, the connectivity gain parameter is capped at including all actually available candidate nodes. This insufficiency typically occurs when a large number of candidate nodes in the reachable region map are blocked and reduced to zero. The connectivity gain parameter consists of the correction gain value of the node with the best connectivity contribution and the corresponding candidate node identifier. Significant faults are marked between adjacent ranked node pairs whose absolute difference in correction gain value exceeds two standard deviations. The fault location is the natural boundary where coverage contribution switches from the dominant node to the supplementary node. The difference in coverage quality between the nodes on both sides can be directly read from the comparison of their respective boundary distance correction coefficients.

[0031] The optimal connectivity bottleneck node is selected based on the connectivity gain parameter to determine the set of colonized nodes. Candidate nodes in the connectivity gain parameter are ranked from highest to lowest according to their corrected gain value. In the first round, the candidate node with the highest corrected gain value is selected, and the diffusion blank area is updated. The overlapping portion between the node's diffusion coverage and the diffusion blank area is removed. After removal, the diffusion blank area shrinks, and the centroid coordinates are updated accordingly. The corrected gain values ​​of the remaining candidate nodes are recalculated using the reachable domain coverage coordinate index carried by each candidate node based on the updated diffusion blank area. The updated connectivity gain parameter serves as the basis for the next round of selection. This process is iterated until the diffusion blank area is less than 5% of the total slope area or all candidate nodes have been selected. The centroid coordinates of the diffusion blank area dynamically shift with each round of removal operations. After the shift, the ranking of the corrected gain values ​​of the remaining candidate nodes may be rearranged. Previously ranked lower candidate nodes may have their corrected gain values ​​surpass those of previously ranked higher nodes due to the shortened distance from the centroid. This dynamic update mechanism ensures that the connectivity gain parameter always corresponds to the real-time spatial state of the diffusion blank area in each round. If, after a certain round of updates, the corrected gain values ​​of all remaining candidate nodes are less than 10% of the average initial connectivity gain parameter, the marginal gain for continued selection is deemed insufficient, and the iteration is terminated. The termination condition of 5% residual ratio is determined according to the upper limit of the gap that can be filled by natural diffusion in mine slope restoration projects. The currently selected nodes are summarized and determined as the colony node set. The ranking of the corrected gain values ​​of each node in the colony node set records its selection order in each round of iterative selection. The nodes with higher selection order have a concentrated coverage contribution, while the coverage contribution of nodes with lower selection order belongs to the marginal supplementation of the residual gap area.

[0032] Based on the set of established nodes, a set of established planting levels and zoning areas is constructed to define the planting and diffusion range. Each node in the established node set, centered on its own coordinates, extends outwards according to the R value of the corresponding position on the grass seed diffusion radius map, defining its diffusion coverage range. When the diffusion coverage ranges of adjacent nodes overlap, the boundary is determined by the combined throat strength of the two nodes. The entire slope is divided into non-overlapping planting units by the diffusion coverage ranges of each established node. Residual blank areas not covered by any established node are merged into the nearest adjacent planting unit. The strength level of each planting unit is jointly determined by the mean potential value of the planting potential map within the unit and the cohesion after correction by the coupling feature set. Units with high mean potential value and normal cohesion are classified as high strength level; units with medium mean potential value or low cohesion are classified as medium strength level; and units with low mean potential value or stability-sensitive units carrying high-corrected strength labels are classified as low strength level. The colonization nodes selected in the order of inclusion have concentrated coverage contributions within their respective colonization units, typically achieving higher intensity levels. Outer throat points act as auxiliary colonization points, incorporating nearby colonization units to supplement the colonization function outside the lattice coverage area. The colonization level zoning set consists of the boundary coordinates, intensity level, and associated colonization node identifiers of each colonization unit. The spatial distribution of intensity levels within the zoning set corresponds to the mechanical degradation pattern of the coupling feature set and the orientation of high-value ridge lines on the colonization potential map. Low-intensity level units are concentrated in heavily polluted slope sections with low cohesion, providing a basis for subsequent slope runoff path calculations and failure transmission analysis.

[0033] Step S130: Perform colonization unit failure propagation analysis on the colonization level partition set to determine the failure blocking succession table, sort the construction priority according to the failure blocking succession table to generate the construction sequence, and parse the key lattice node parameters based on the construction sequence to form the construction execution configuration.

[0034] In some embodiments, performing colonization unit failure propagation analysis on the colonization level partition set to determine the failure blocking succession table includes: calculating the slope runoff path of the colonization level partition set to generate a runoff directed graph; identifying the dependency relationship between upstream and downstream colonization units based on the runoff directed graph to construct a dependency relationship matrix; performing failure cascade path tracing on the dependency relationship matrix to obtain a failure propagation path set; and extracting key cut point units based on the failure propagation path set to determine the failure blocking succession table.

[0035] A directed runoff graph is generated by calculating slope runoff paths for the colony-level zoning set. The D8 algorithm determines the flow direction from the current grid to the direction with the largest slope in the eight neighboring areas. The slope is calculated by dividing the difference between the current grid elevation and the elevation of the neighboring grids by the grid spacing. The flow directions of the grid nodes of each colony unit in the colony-level zoning set are summarized to form the runoff inflow direction of that unit. Adjacent colony units with runoff inflow relationships form directed edges, with the direction of the directed edges pointing from the upstream unit to the downstream unit. The directed runoff graph is constructed with all colony units in the colony-level zoning set as nodes. Colony units with larger catchment areas have higher in-degree values, receive larger upstream runoff volumes, and are more sensitive to hydraulic disturbances as the in-degree increases. The flow direction determination of colony units located in depressions or flat areas may exhibit multi-directional equilibrium. The direction of the maximum slope in the neighboring areas is forcibly specified. When the slope difference is less than 0.5%, a low-confidence flow direction is marked. The directed edges marked with this low confidence value participate in the dependency strength calculation with a weight of 0.7 during the dependency matrix construction stage. At the nodes of the lattice beams, local depressions are created due to construction disturbances in the micro-topography. When the slope difference is low, the frequency of low-confidence flow direction annotations increases. The overall local topological confidence of the runoff directed graph in such densely populated lattice node areas is low. Therefore, when constructing the dependency matrix, the dependency strength results of concentrated low-confidence flow direction annotation areas need to be uniformly annotated with low precision. When the main runoff line on the slope passes through multiple colony units, the directed edge sequence along the main runoff line forms a high-flux conduction chain in the runoff directed graph. The in-degree values ​​of each colony unit on the chain are accumulated sequentially according to the slope direction. Such high-flux conduction chains are most dense in gully-developed areas. The in-degree value of the terminal converging node is the highest in the entire slope runoff directed graph. The conduction direction of the main runoff chain is the spatial direction of the strongest dependency conduction direction in the dependency matrix. The number and length distribution of the high-flux conduction chain paths directly determine the spatial density of strong dependency pairs in the subsequent dependency matrix.

[0036] For example, the step of identifying the dependency relationships between upstream and downstream colonization units and constructing a dependency relationship matrix based on the runoff directed graph includes: extracting the runoff intensity classification features of each colonization unit in the runoff directed graph to generate a unit runoff net flux distribution map; annotating the runoff connection relationships between upstream and downstream units based on the unit runoff net flux distribution map to obtain a connection relationship map; prioritizing cross-unit failure transmission paths based on the connection relationship map to obtain a transmission priority set; and constructing a dependency relationship matrix by weighting the failure probability of the transmission priority set.

[0037] The runoff intensity classification characteristics of each settlement unit in the runoff directed graph are extracted to generate a net flux distribution map of the unit runoff. The weighted in-degree is equal to the sum of the runoff fluxes of all upstream directed edges of the settlement unit in the runoff directed graph. The runoff flux is estimated by the product of the catchment area of ​​the upstream settlement unit and the slope runoff coefficient and is used as a relative flux index in terms of area. The runoff coefficient is determined by looking up a table according to the settlement level of each unit. The runoff coefficient is lower in high-intensity level zones due to high vegetation cover, and higher in low-intensity level zones. The effect of slope difference on the runoff coefficient within the same level zone is separately superimposed by the slope correction coefficient. The slope correction coefficient increases with the slope. When the slope exceeds 35°, the upper limit value is taken to reflect the amplification effect of the shortened rainfall infiltration time on runoff under steep slope conditions. The difference between the weighted in-degree and the downstream output flux in the runoff directed graph is the net flux of that unit. A positive net flux indicates a runoff accumulation zone, while a negative net flux indicates a transit zone. The net flux of all settled units is plotted according to coordinates to form a unit runoff net flux distribution map. Positive value areas are concentrated at the slope catchment lines and gullies, while negative value areas are concentrated at the slope watershed ridges and high-speed transit zones. Slope micro-topographic undulations cause local water catchment, resulting in positive value islands in the unit runoff net flux distribution map. The net flux at the center of the island is higher than that of the surrounding units, corresponding to micro-topographic depressions. Local depressions formed by the disturbance of the lattice beam construction also produce artificially induced positive value islands. The peak net flux of these islands is usually lower than that of natural catchment line islands. The two types of positive value islands are distinguished by source labeling. The labeling of natural catchment islands and artificially induced islands triggers different connection strength correction rules during the connection relationship map construction stage.

[0038] A runoff connection diagram is obtained by annotating the runoff connection relationships between upstream and downstream units based on the unit runoff net flux distribution map. When the net flux is positive, the colonizing unit establishes a runoff connection relationship with the upstream unit. The connection strength is determined by the proportion of the upstream unit's output runoff flux to the weighted in-degree of the downstream unit. A higher proportion indicates a concentrated contribution of the upstream unit to the downstream unit's net flux. The connection strength at artificially induced positive islands in the unit runoff net flux distribution map is calculated using a correction factor of 0.85. This is because the convergence effect of local depressions in lattice nodes is partly induced by structural topography, and their sensitivity to upstream natural failures is lower than that of natural runoff islands. Transiting units with negative net flux also establish a connection relationship with upstream units but are marked as transiting. The transit label reduces the path priority weight of the corresponding directed edge by 0.2 in the conduction priority ranking. After receiving the runoff, the transiting unit quickly transfers it downstream, but its own net flux accumulation is limited, and its buffering capacity against upstream failures is weaker than that of accumulation zone units. The transfer relationship diagram is constructed by adding transfer intensity values ​​to each directed edge of the runoff directed graph. The spatial pattern of transfer intensity distribution is directly affected by the net flux gradient in the unit runoff net flux distribution diagram. The direction with a high net flux gradient corresponds to the concentrated direction of the high transfer intensity area in the transfer relationship diagram. The colonizing unit with net flux close to zero in the middle of the slope acts as both the receiver of the upstream unit and the supplier of the downstream unit. The transfer intensity of the directed edges on both the upstream and downstream sides is similar, showing a bidirectional balanced relay transit topology, which contrasts with the unilateral strong transfer characteristic of the positive net flux area. The two types of characteristics are quantified by the node topological symmetry field. Symmetry is the degree to which the ratio of upstream and downstream transfer intensity is close to 1. Nodes with high topological symmetry are identified as relay nodes in the transmission priority ranking stage. The path priority contribution of relay nodes is different from that of source and convergence nodes. The product contribution weight of the directed edge where it is located is corrected by 0.9 times to avoid overestimating the priority of relay paths.

[0039] Based on the connection relationship diagram, a priority set is obtained by prioritizing cross-unit failure propagation paths. Directed edges with a connection strength exceeding 0.3 in the connection relationship diagram are taken as backbone propagation edges. The backbone propagation edge set is arranged from high to low connection strength and then each path is traced sequentially. Each backbone propagation edge is traced upwards to the source unit and downwards to the leaf node unit. A complete path forms a cross-unit failure propagation path. Directed edges with transit markings participate in the path priority product calculation with a coefficient of 0.8, while directed edges containing relay nodes participate with a coefficient of 0.9. The priority of each cross-unit failure propagation path is determined by the product of the connection strength values ​​of all directed edges on the path. A larger product indicates stronger continuous runoff propagation capacity from upstream to downstream. The more nodes in the path, the greater the natural attenuation of the product. To avoid systematically low priority due to a large number of nodes in long paths, the product value is normalized by the square root of the path hop count. After normalization, the propagation priorities of different path lengths are comparable. When multiple directed paths connect the same starting point and ending point in the connection diagram, the product of the priorities of each path is calculated separately, and the maximum value is taken as the representative transmission priority of the starting point-ending point pair. All cross-unit failure transmission paths are arranged in descending order of priority to form a transmission priority set. The bearing strength of the directed edges along the slope gully is continuously high, and the corresponding path ranks high in the transmission priority set after normalization. The spatial orientation of the path node sequence is consistent with the orientation of the main runoff line of the slope. The slope edge bypass path has more directed edges marked by transit and more relay nodes. After the double reduction coefficient is superimposed, the normalized priority ranks significantly lower in the transmission priority set. The priority difference between the two types of paths describes the substantial difference in failure transmission efficiency between the main runoff channel and the edge bypass channel. The dividing line usually corresponds to the spatial boundary between the main runoff line and the edge transit path.

[0040] A dependency matrix is ​​constructed by weighting the failure probability of the transmission priority set to determine its dependency strength. The failure probability weighting first determines the baseline failure probability for each level of zone, then multiplies it by the path priority. The baseline failure probability is determined based on the vegetation loss rate statistics of each level of zone in historical monitoring of mine slope restoration projects. Low-intensity zone units have a higher baseline failure probability, and high-intensity zone units have a lower baseline failure probability. The vegetation loss rate statistics cover continuous monitoring data for at least three complete rainy seasons. When the data volume is less than three rainy seasons, the baseline failure probability of the corresponding level of zone is replaced by a conservative estimate, which is 1.1 times the higher baseline failure probability among adjacent level zones. The priority value of each path in the transmission priority set is multiplied by the baseline failure probability of the level to which each colonized unit on the path belongs, yielding the failure probability contribution of each colonized unit on that path to its downstream units. When the same colonized unit appears on multiple paths, the maximum contribution of each path is taken instead of being summed, to avoid the dependency strength of the same unit being artificially inflated due to multiple path coverage. After weighting the failure probability, the dependency strength of each colonization unit pair is filled into the corresponding position M(i,j) of the dependency relationship matrix. The M(i,j) corresponding to the sub-regions in the low-confidence flow direction annotation set of the runoff directed graph is added as a whole with low-precision annotation after calibration, for subsequent failure cascade path tracking and weight reduction processing. The asymmetry of the dependency relationship matrix is ​​naturally determined by the directionality of the succession relationship graph. The transmission path priority from high-intensity colonization units to low-intensity colonization units at the slope foot is usually higher in the middle and lower slope sections with larger slopes, and the failure baseline probability of low-intensity units is higher. Therefore, the corresponding M(i,j) after weighting the failure probability is higher in the dependency relationship matrix.

[0041] Failure transmission path set is obtained by tracing the failure cascade path of the dependency matrix. A depth-first search algorithm starts from each colonized unit and advances step-by-step along the direction where M(i,j) is greater than zero. The path terminates when it reaches a leaf node unit with zero out-degree in the dependency matrix or when the path length exceeds the maximum transmission level. The maximum transmission level is the square root of the total number of colonized units on the entire slope, rounded off. Paths exceeding the maximum transmission level are truncated and marked as long-distance transmission. The failure transmission path set consists of a list of all traced paths. Each path records the starting unit identifier, the path node sequence, and the minimum value of M(i,j) on the path. The minimum value marks the weakest transmission link on the path. The upstream and downstream colonized units of the directed edge containing the weak link are marked as transmission bottleneck pairs in the failure transmission path set for manual verification. Slope areas with high density of transmission bottleneck pairs indicate the existence of a systematic transmission resistance concentration zone. The spatial location of the resistance concentration zone is usually related to the orientation of the level boundary of the colonization level zoning set. The difference in runoff erosion protection capacity caused by the difference in vegetation cover at the level switching boundary is the main cause of transmission resistance concentration. When high-intensity slope colonization units are used as starting units, the corresponding path node sequences in the failure propagation path set typically span multiple colonization level partitions from high to low intensity levels. The terminal leaf node units of these paths are usually located at the confluence endpoint of low-intensity slope partitions. These cross-level propagation paths are labeled in descending order of propagation level in the failure propagation path set. The more levels a path spans and the lower the level of its terminal leaf node, the higher its propagation importance weight in subsequent critical cut point unit identification. Depth-first search path segments corresponding to sub-matrices carrying low-precision regional labels in the dependency matrix are simultaneously labeled as low-confidence path segments in the failure propagation path set. These low-confidence path segments participate in the total number of propagation paths in a conservative manner during the critical cut point unit simulation removal phase to avoid overestimating the number of paths in the low-precision sub-matrix region, which could lead to deviations in the critical cut point unit identification results.

[0042] Critical cut point units are extracted from the set of failure propagation paths to determine the failure blocking succession table. A colony unit is identified as a critical cut point unit when the total number of propagation paths decreases by more than 20% after removal. The effect of cut point unit removal is quantified by re-counting the change in the number of paths in the failure propagation path set after unit-by-unit simulation. Colony units that do not exceed the threshold are not included in the failure blocking succession table. The 20% threshold is calibrated based on the elasticity of the impact of reducing the total number of propagation paths on the overall colonization stability in slope restoration projects. Historical monitoring shows that when the total number of propagation paths decreases by more than 20%, the slope failure propagation rate begins to accelerate significantly; below this threshold, the slope's self-healing ability can still restrain the spread of failure. Therefore, 20% is used as a practical criterion for identifying critical cut point units. After the critical cut point unit is identified, the blocking priority value is determined by the product of the occurrence frequency in the failure propagation path set and the path descent contribution rate. The failure blocking succession table also assigns a successor unit to each critical cut point unit. The successor unit is selected from the colonization units adjacent to the cut point unit in the colonization level partition set that are not on the same failure propagation path. Priority is given to neighboring units with high colonization level and high potential value in the colonization potential map. The spatial distance between the successor unit and the main unit does not exceed the grid spacing of two colonization units. Successor candidate units that exceed the distance are downgraded because the spatial span is too large and the actual succession response speed after the failure of the main unit is too low. The critical cut point units in the gully development zone of the slope appeared in a large number of failed transmission paths. After simulation removal, the total number of transmission paths decreased significantly, and they obtained the highest blocking priority in the failure blocking succession table. The successor units were specified from the high-intensity level colonization zone on the gully flank. The critical cut point units near the slope watershed ridge line appeared relatively less frequently and had a smaller contribution rate to the path decline. Their blocking priority values ​​were in the middle range. The difference in blocking priority between the two types of areas is presented as a numerical gradient in the failure blocking succession table. The slope spatial distribution gradient is consistent with the distribution gradient direction of the slope water catchment pattern.

[0043] The implementation priority is sorted according to the failure blocking succession table to generate an implementation sequence. The arrangement of the implementation sequence imposes a spacing constraint on the positional relationship between the main unit and the successor unit in the failure blocking succession table. The spacing between the two in the implementation sequence shall not exceed three positions to prevent the excessive insertion of other colonization units with high blocking priority, which would excessively widen the implementation sequence between the main unit and the successor unit. If the sequence spacing is too large, the successor unit may not have been completed after the main unit fails, resulting in a time gap of no blocking coverage in the slope failure transmission chain in that section. The blocking priority value of each colonization unit in the failure blocking succession table is quantified by the product of the transmission path coverage frequency and the path descent contribution rate. The unit with the larger product is ranked higher in the implementation sequence. When the blocking priority values ​​are the same, the order is determined by the level of the colonization level partition set. Key cut-point units in the catchment area at the bottom of the slope gully simultaneously appear in numerous failed conduction paths. These conduction paths have a high coverage frequency across the entire slope, resulting in a correspondingly high blocking priority value, and are ranked higher in the implementation sequence. Near the slope's watershed ridge, runoff convergence is low, leading to lower coverage frequency and path descent contribution rates for the corresponding key cut-point units, and they are ranked lower in the implementation sequence. The difference in blocking priority between these two types of areas directly reflects the spatial constraint of slope hydraulic path distribution on the implementation timeline. After prioritizing the arrangement of key cut-point units and successor units, non-cut-point units are added to the end of the implementation sequence in descending order of their colonization level. The implementation sequence is then arranged unit by unit according to the above rules to form an ordered list. The position of each colonization unit corresponds to the succession relationship in the failed blocking succession table. The length of the implementation sequence list is equal to the total number of colonization units on the slope, and the coverage area is consistent with the S150 root layer effectiveness attenuation monitoring target.

[0044] The construction execution configuration is formed based on the analysis of key lattice node parameters in the construction sequence. For each key lattice node in the construction sequence, the three parameters of the lattice beam cross-section, anchorage depth, and microbial agent carrier filling thickness are analyzed sequentially according to their position. The upper limit of the parameter values ​​corresponds to the lattice node with the highest blocking priority and the lowest cohesion after coupling feature set correction in the failure blocking succession table. These three parameters jointly cover the dual design requirements of mechanical support and ecological restoration. The cross-section size is determined by the blocking priority value and the corrected cohesion. The upper limit of the cross-section size is taken for lattice nodes with high blocking priority and low corrected cohesion. The cross-section size of lattice nodes in slope sections with significant mechanical deterioration can differ by one to two size grades compared to slope sections with normal cohesion. The anchorage depth is calculated using the limit equilibrium method based on the corrected internal friction angle φ and corrected cohesion c at the lattice node. The calculation results are taken as the upper envelope value within the fluctuation range of soil parameters in wet and dry seasons. The soil in the waterlogged area at the toe of the slope is in a near-saturated state for a long time, and the internal friction angle decreases significantly during the wet-dry cycle. The upper envelope value of the anchorage depth of the lattice node in this area is usually significantly greater than that in the dry area at the top of the slope. The difference in anchorage depth between the two types of areas is presented as a gradient of longitudinal parameters on the slope in the construction execution configuration. The filling thickness of the inoculant carrier corresponds to the level zone of the lattice node in the colonization level zoning set. The upper limit value is taken for high-strength level zones, and the lower limit value is taken for low-strength level zones. The difference in filling thickness determines the water-holding capacity of the inoculant and the initial diffusion matrix conditions of the mycelium at the lattice nodes in different level zones. After the three parameters of each lattice node are analyzed, they are arranged and summarized according to the construction sequence to form the construction execution configuration. The slope spatial gradient of cross-sectional dimensions and anchorage depth reflects the deterioration distribution of the corrected cohesion, and the spatial distribution of the inoculant carrier filling thickness is consistent with the orientation of the level boundary of the colonization level zoning set.

[0045] Step S140: Analyze the root layer water potential fluctuations of the construction execution configuration to generate a water potential characteristic map. Based on the water potential characteristic map, implement inoculation gradient planning to obtain a tiered mix ratio sequence. Based on the tiered mix ratio sequence, identify the enhanced section and the conventional section to construct a set of control schemes.

[0046] In some embodiments, the step of performing root layer water potential fluctuation analysis on the construction execution configuration to generate a water potential feature map includes: parsing the soil texture parameters of the lattice nodes based on the construction execution configuration to obtain a texture parameter set; calculating the root layer water holding capacity based on the texture parameter set to generate a water holding capacity distribution map; performing rainfall-evaporation disturbance simulation on the water holding capacity distribution map to obtain a water potential response curve set; and calculating the residence time of suitable water potential intervals based on the water potential response curve set to form a water potential feature map.

[0047] Based on the analysis of soil texture parameters at the lattice nodes in the construction execution configuration, a texture parameter set was obtained. The clay, silt, and sand contents were determined using laser particle size analysis of the undisturbed soil columns at the anchorage locations of each lattice node. The soil column sampling depth covered two layers: 0-20 cm and 20-40 cm. The active root layer was concentrated in the 0-20 cm layer, where the texture parameters had a higher weighting on the root zone's water-holding capacity than the layers below. The texture parameter set recorded the measured values ​​for both layers and added layer labels. The anchorage depth of each lattice node in the construction execution configuration determined the degree of construction disturbance to the surrounding soil at the sampling location. Grid nodes with larger anchorage depths showed significant changes in the pore structure of the soil after construction. The texture parameters of the corresponding nodes were corrected for construction disturbance after on-site sampling. The correction coefficient was quantified by the change in soil density within a range of one borehole diameter around the anchorage borehole. Increased density indicated higher compaction, and after compaction correction, the proportion of clay content effectively participating in water holding decreased accordingly. Organic matter content at each lattice node is preferentially recorded using the corresponding coupling feature set. For nodes with missing organic matter content, on-site measurements are conducted and the organic matter content is then incorporated into the texture parameter set. Organic matter content affects soil aggregate stability and consequently, the actual water-holding capacity of the root zone. Lattice nodes with low organic matter content typically correspond to areas with low slope vegetation cover or where heavy metals promote organic matter decomposition. These nodes are marked as low organic matter in the texture parameter set. For nodes with low organic matter content, a more conservative lower limit for water-holding capacity estimation is used during the water-holding capacity calculation phase to prevent the actual water-holding capacity from systematically falling below the estimated value due to soil aggregate instability under low organic matter conditions. When adjacent lattice nodes are close together during construction, the texture parameter set values ​​of the two nodes are cross-checked using Kriging interpolation. If the interpolation deviation exceeds 15% of the measured value, the corresponding node is resampled to ensure the spatial consistency and measurement accuracy of the texture parameter set across the entire slope.

[0048] A water-holding capacity distribution map is generated by calculating root zone water-holding capacity based on a set of texture parameters. Field water holding capacity θ_fc is estimated using the Saxton-Rawls equation based on the clay and organic matter content in the texture parameter set. The equation is θ_fc = a0 + a1 × Clay + a2 × OM, where Clay is the clay mass fraction, OM is the organic matter mass fraction, and a0, a1, and a2 are empirical coefficients calibrated according to the soil water-holding characteristics of the slope's climate zone. The wilting coefficient θ_wp is determined jointly by the sand and clay content in the texture parameter set. Lattice nodes with high clay content have larger wilting coefficients. Effective root zone water holding capacity is determined by the difference between field water holding capacity and the wilting coefficient. A larger effective water holding capacity means that the root zone at that node maintains available water for a longer period during water consumption. The effective water holding capacity of lattice nodes with low organic matter content in the texture parameter set is calculated using a reduction factor of 0.8. Because soil aggregate stability is poor under low organic matter conditions, the actual water holding capacity is lower than the estimated value of nodes with normal organic matter content in the same texture. The reduction factor further tightens as the degree of low organic matter increases, reaching 0.7 when the organic matter content is below 1%. The effective water holding capacity of each lattice node at the root layer is plotted sequentially using planar coordinates to form a complete water holding capacity distribution map. The overall spatial pattern is positively correlated with the clay content distribution in the texture parameter set. Slope areas with higher clay content have higher water holding capacity, while areas with higher sand content have lower water holding capacity. At the slope toe, due to long-term gravity water replenishment and high clay deposition, a high water holding capacity accumulation zone typically appears on the water holding capacity distribution map at the slope toe, forming a significant longitudinal gradient with the low water holding capacity of the sandy area at the slope top. In the water-holding capacity distribution map, lattice nodes with water holding capacity more than 30% lower than the average of the entire slope are designated as low water-holding areas. In the water potential response curve simulation stage, lattice nodes in low water-holding areas need to use a shorter soil moisture consumption time step to maintain the simulation accuracy of the rapid decline of water potential in low water-holding areas.

[0049] Rainfall-evaporation disturbance simulations were performed on the water-holding capacity distribution map to obtain a set of water potential response curves. The simulations were performed on a daily timescale. Rainfall input consisted of a ten-year daily rainfall sequence from the meteorological zone where the slope was located. Evaporation was calculated daily using the Penman formula based on temperature, wind speed, and radiation data. The net rainfall-evaporation water volume and the effective root water holding capacity of each lattice node in the water-holding capacity distribution map jointly determined the daily dynamic change of soil volumetric water content at that node. The process of converting soil volumetric water content into water potential Ψ was performed using the Brooks-Corey equation. Where θ is the volumetric water content, Ψ_e is the air intake value, θ_s is the saturated water content, and λ is the pore size distribution index. The latter three parameters are determined by referring to the parameter table based on the clay content and sand content in the texture parameter set. In the water-holding capacity distribution map, the simulation step size of the grid nodes in the low water-holding area is shortened to a half-day step size during the rapid soil moisture consumption phase after rainfall to avoid insufficient accuracy when the water potential rapidly declines and crosses the boundary of the suitable water potential interval. When there are differences in slope aspect, the evaporation calculation of grid nodes on sunny and shady slopes uses the radiation correction coefficient corresponding to the slope aspect. Sunny slopes have higher radiation intensity and thus higher evaporation, while shady slopes have lower evaporation. At the same elevation, the water potential decline rate of grid nodes on sunny slopes is significantly faster than that on shady slopes. The difference in curve shape between the two types of nodes is distinguished by the slope aspect label field, which is used as the basis for classification statistics during the water potential characteristic map statistical residence time stage. The daily water potential sequence of each lattice node is summarized according to the node identifier to form a set of water potential response curves. In the water holding capacity distribution map, the lattice nodes with high water holding capacity correspond to curves with small fluctuation amplitude and high trough water potential values ​​in the set of water potential response curves. The lattice nodes with low water holding capacity correspond to curves with violent fluctuation and low trough values. The duration of the flat section of the water potential curve of the node with high water holding capacity at the slope toe is longer than that of the node with low water holding capacity at the slope top.

[0050] A water potential characteristic map was generated based on the residence time within the suitable water potential range, calculated from the water potential response curve set. The suitable water potential range is defined as -0.01 to -0.3 MPa, representing the range of water potential ranges where arbuscular mycorrhizal fungi colonize actively. Above -0.01 MPa, the soil is close to saturation, and the oxygen supply to the root cortex is insufficient for arbuscular mycorrhizal fungi infection. Below -0.3 MPa, soil moisture stress is too strong, and arbuscular mycorrhizal fungi spore germination is inhibited. The thresholds at both ends were determined according to the infection rate versus water potential curve in a standardized pot inoculation experiment. The cumulative number of days falling within the suitable water potential range for each lattice node in the water potential response curve set was calculated. The cumulative number of days divided by the total number of simulated days yielded the proportion of residence time within the suitable water potential range for that node. A higher proportion of residence time indicates a larger percentage of the year that is suitable for arbuscular mycorrhizal fungi colonization at that node. The water potential response curves show interannual fluctuations in the residence time ratio of the same lattice node due to differences in rainfall across different years. lattice nodes with interannual fluctuations exceeding 30% of the mean are marked with high fluctuations. The minimum value of the ten-year series is taken as a conservative estimate of the residence time ratio of that node, thus ensuring that the inoculation gradient planning meets the basic water potential conditions for arbuscular mycorrhizal fungi colonization even in dry years. The residence time ratio of lattice nodes on sunny and shady slopes is affected by the aspect radiation correction coefficient. Sunny slope nodes experience strong evaporation and rapid water potential traverses the suitable range towards the stress zone below -0.3 MPa, resulting in a residence time ratio typically lower than that of shady slope nodes at the same elevation. The water potential characteristic map shows a systematic distribution of low residence time ratios in sunny slope areas, forming a stark contrast with the aspect differentiation pattern in shady slope areas. The proportion of dwell time of each lattice node is plotted on coordinates to form a water potential characteristic map. The proportion of dwell time of lattice nodes with high water holding capacity is higher in the water potential characteristic map, while that of nodes with low water holding capacity is lower. The spatial deviation of the two reflects the different effects of rainfall evaporation sequence on the water potential fluctuation pattern of nodes with different water holding capacities.

[0051] In some embodiments, the step of implementing inoculation gradient planning based on the water potential feature map to obtain a tiered ratio sequence includes: matching the water potential feature map to suitable water potential ranges for fungal agents to generate a suitable inoculation zoning map; analyzing the ratio gradient of Bacillus and arbuscular mycorrhizal fungi based on the suitable inoculation zoning map to obtain an initial ratio scheme; performing zoning gradient correction on the initial ratio scheme to form a corrected ratio scheme; and arranging the corrected ratio scheme in a time sequence to generate a tiered ratio sequence.

[0052] A suitable inoculation zoning map is generated by matching the suitable water potential range of the water potential feature map with the inoculant. Nodes in the water potential feature map with a residence time ratio exceeding 35% are classified as highly suitable for inoculation, those between 25% and 35% as moderately suitable, and those below 25% as lowly suitable. The grading thresholds for these three types of nodes are determined based on the field infection rate of arbuscular mycorrhizal fungi under different residence time ratios. The lower limit of the residence time ratio corresponding to an infection rate exceeding 40% is approximately 35%, and the upper limit corresponding to an infection rate below 15% is approximately 25%. Nodes in the water potential feature map marked with high fluctuations are downgraded one level in the suitable inoculation zoning map. Although the annual average residence time ratio of high-fluctuation nodes may be higher, the actual suitable inoculation period is significantly shortened under dry years. Downgrading them by one level makes the inoculation gradient planning more reliable. The zoning levels of each lattice node are plotted according to coordinates to form a suitable inoculation zoning map. High suitable inoculation nodes are usually concentrated near the slope catchment line and in areas with high water holding capacity at the slope toe, while low suitable inoculation nodes are concentrated near the slope water divide ridge and in the slope top section with low water holding capacity. Due to strong evaporation, the proportion of residence time in sunny slope areas is systematically low, and the proportion of low suitable inoculation nodes in sunny slope areas is significantly higher than that in shady slope areas in the suitable inoculation zoning map. When the zoning levels of adjacent lattice nodes in the suitable inoculation zoning map span two or more levels, the intermediate level buffer node is determined by spatial interpolation. The residence time ratio of the buffer node is taken as the linear interpolation of the values ​​of adjacent nodes. During the gradient correction stage, the buffer node enjoys a lower correction weight of 0.5 times to prevent the interpolated node from raising the zoning gradient of adjacent real measured nodes. The buffer node is recorded as an interpolation supplement node in the suitable inoculation zoning map.

[0053] The initial formulation scheme was obtained by analyzing the ratio gradient of Bacillus and arbuscular mycorrhizal fungi based on the suitable inoculation zoning map. The synergistic relationship between Bacillus and arbuscular mycorrhizal fungi was strongest when the soil water potential was -0.05 to -0.15 MPa. The diffusion efficiency of phosphorus-soluble organic acids secreted by Bacillus was the highest within this water potential range, and the mycelial extension rate of arbuscular mycorrhizal fungi also reached its maximum in this range. The optimal water potential range for the synergistic effect of the two coincides with the main water potential retention range corresponding to the high suitable inoculation node in the suitable inoculation zoning map. In the suitable inoculation zoning map, the residence time of highly suitable inoculation nodes is sufficient, allowing both types of inoculants enough time to exert a synergistic effect. The initial formulation at highly suitable inoculation nodes is set at a mass ratio of Bacillus to arbuscular mycorrhizal fungi (AMM), with AMM as the dominant fungi. At moderately suitable inoculation nodes, the residence time is shorter, and Bacillus exhibits stronger resistance to water stress than AMM; therefore, the ratio is set at 1:1. At low-suitable inoculation nodes, which are under prolonged water stress and have limited AMM colonization, the proportion of Bacillus is increased to maintain the basic soil improvement effect; therefore, the ratio is set at 2:1. After downgrading the zoning level of nodes marked with high fluctuations in the suitable inoculation zoning map, their formulations shift towards a Bacillus-heavy ratio, showing a higher proportion of Bacillus compared to nodes with stable water potential. This is because highly fluctuating nodes have a higher probability of entering a state of severe water stress during dry years, and the drought resistance and conservation function of Bacillus is more critical at these nodes. The initial proportioning scheme takes the zoning level of each grid node in the suitable vaccination zoning map as input, and assigns values ​​to each node according to the above three proportion rules. After all grid nodes are assigned values, an ordered list is formed, and the proportioning value of each node maintains a strict one-to-one correspondence with the zoning level of the suitable vaccination zoning map.

[0054] The initial inoculation scheme was subjected to zonal gradient correction based on the suitable inoculation zoning map to form a corrected inoculation scheme. In the initial scheme, the correction weight for interpolated nodes was 0.5, while the correction weight for actual measurement nodes was 1.0. This weight difference resulted in a smoother gradient in the corrected scheme in areas with dense interpolated nodes than in areas with dense actual measurement nodes. Consequently, the corrected inoculation values ​​of the interpolated nodes were closer to the average level of adjacent actual measurement nodes. The correction operation proceeded from high to low in the zoning gradient of the suitable inoculation map. The correction of each lattice node was recalculated based on the corrected inoculation values ​​of its four surrounding lattice nodes. The direction of advancement was the same as the decreasing direction of the residence time ratio in the water potential characteristic map, ensuring that the correction results in highly suitable inoculation areas were not underestimated due to reverse infiltration of inoculation values ​​from adjacent less suitable inoculation nodes. At the boundary between the sunny and shady slopes, there is a significant difference in zoning levels on both sides. The grid nodes at this boundary exhibit a marked abrupt change in zoning ratio in the initial scheme. Gradient correction at this location performs a harmonic mean transition, reducing the difference in zoning ratios on both sides of the boundary. After the transition, the zoning ratios on the sunny slope side are biased towards Bacillus, while those on the shady slope side are biased towards arbuscular mycorrhizal fungi. The width of the transition zone is determined by the difference in the proportion of residence time on both sides; a larger difference requires a larger number of grid cells. The corrected zoning scheme is formed by summing the zoning values ​​of each grid node after correction. Grid nodes with large correction amplitudes are usually located in the transition zone between the two zoning levels on the suitable inoculation zoning map. The absolute value of the difference between the corrected zoning scheme value and the initial zoning scheme value for such nodes is recorded separately in the difference field for manual verification. A larger difference field indicates that the precise application zoning ratio of that grid node is highly sensitive to changes in the zoning gradient.

[0055] A tiered mix design sequence was generated based on the corrected mix design scheme. The sequence was determined by the order in which each grid node in the suitable inoculation zone map entered its suitable inoculation period. Entering the suitable inoculation period refers to the date on which the water potential value of the node in the water potential response curve drops below -0.01 MPa. This date, during the soil moisture consumption process after rainfall, first appears at the low suitable inoculation nodes with low water holding capacity at the slope top and last appears at the high suitable inoculation nodes with high water holding capacity at the slope toe. The tiered mix design sequence was gradually incorporated into the grid nodes of the corrected mix design scheme from morning to night, following the daily order of entering the suitable inoculation period. When the difference in the date of entering the suitable inoculation period for grid nodes in different areas of the slope is within 1 to 5 days, they are classified as the same application batch; when the difference exceeds 5 days, they are classified as independent application batches. This batch division ensures sufficient overlap of the inoculation time windows for grid nodes within the same batch, preventing some nodes from leaving the suitable water potential range while others have not yet been inoculated. When there are differences in the corrected mix proportions of grid nodes within the same application batch, the mix proportions within the batch are weighted and averaged according to the residence time of each node. The mix proportions of grid nodes with higher weights contribute more to the calculation of the batch average, and the batch average serves as the unified application mix for that batch. When the mix proportion of individual grid nodes within a batch deviates from the batch average by more than 20%, that node is listed separately in the tiered mix proportion sequence, so that grid nodes with larger mix proportion deviations obtain independent and accurate application mixes. Due to strong evaporation, grid nodes on sunny slopes enter the suitable inoculation period earlier than nodes on shady slopes at the same elevation. In the tiered mix proportion sequence, sunny slope application batches are arranged earlier and shady slope application batches are arranged later. The time difference reflects the direct impact of slope aspect radiation differences on the soil water potential consumption rate. The unified mix proportions of each application batch and the accurate mix proportions of individually listed nodes are summarized in batch order to form the tiered mix proportion sequence.

[0056] Based on the tiered mix proportion sequence, enhanced and conventional sections were identified to construct a set of control schemes. The identification threshold for enhanced sections was set as the total mix proportion of each grid node in the tiered mix proportion sequence exceeding one standard deviation of the average value for the entire slope. Adjacent grid nodes consecutively exceeding the threshold were merged into one enhanced section. When the interval exceeded one grid grid, each enhanced section formed independently. This merging rule avoided spatial fragmentation of enhanced sections, preventing continuous application of enhancement measures. Grid nodes with a total mix proportion below the average were identified as conventional sections. Transitional markings were added to grid nodes at the boundary between the two types of sections, and the enhanced section scheme was prioritized during implementation. Enhanced sections, based on conventional sections, incorporated three additional enhancement measures: increasing the inoculant carrier filling thickness to 1.2 to 1.5 times, extending the inoculation time window to 1.5 times, and establishing an inoculant penetration buffer zone two grid grid widths at the boundary. The total mix proportion within the buffer zone linearly transitioned from the enhanced section value to the conventional section value to prevent abrupt changes in inoculant density at the boundary from affecting the continuity of colonization and diffusion. In the middle and lower slopes with steeper gradients, the suitable water potential residence time is relatively short. The total proportion of the tiered ratio sequence of the grid nodes in these sections is relatively high. These sections are classified as reinforced sections in the control scheme set, forming a clear longitudinal zoning pattern with the grid nodes at the top of the slope, which have ample residence time and are classified as regular sections. The spatial boundaries of these two types of sections are usually parallel to the contour lines of the slope, reflecting the contour line distribution characteristics of water potential residence time on the slope. The control scheme set uses two types of templates as its core: reinforced section schemes and regular section schemes. Each grid node calls the corresponding template according to its section affiliation. The three control parameters in the template—total inoculant ratio, filling thickness, and inoculation window—are arranged chronologically according to the tiered ratio sequence and expanded for each grid node. The expanded control parameters of all grid nodes are then summarized to form the control scheme set.

[0057] Step S150: Perform coverage-infection rate effectiveness decay trend analysis on each colonization unit corresponding to the application sequence to generate a decay trend map. Based on the decay trend map, identify the characteristics of the decay phase transition precursor and determine the critical value of tiered reseeding and inoculum replenishment. Based on the control scheme set and the critical value of tiered reseeding and inoculum replenishment, perform multi-threshold hierarchical decision-making for reseeding and inoculum replenishment and output slope collaborative restoration control instructions.

[0058] Specifically, a decay trend analysis of the coverage-infection rate effectiveness was performed on each colonization unit corresponding to the application sequence to generate a decay trend map. The vegetation coverage and arbuscular mycorrhizal fungal infection rate of each colonization unit during the operation and maintenance period were obtained through regular monitoring and archived according to the application sequence position. The coverage decay rate was quantified by the time-series linear regression slope of the coverage values ​​in each colonization unit in previous vegetation surveys. The slope with the largest negative absolute value indicates that the coverage of the colonization unit decreases rapidly over time. The infection rate decay rate was simultaneously quantified by the time-series linear regression slope of the arbuscular mycorrhizal fungal infection rate values ​​in the root slice detection of each colonization unit. The two types of slopes were calculated separately on the same time-series coordinate axis, with the time starting point being the time of data collection for the first survey after the colonization unit was completed. After pairing the coverage decay rate and infection rate decay rate for each colonization unit, the product of the absolute values ​​of their slopes serves as the overall efficacy decay intensity for that colonization unit. A larger product indicates a faster rate of simultaneous decline in both efficacy indicators. Colonization units placed earlier in the implementation sequence typically have higher overall efficacy decay intensities due to greater failure-blocking pressure and stronger environmental stress, forming a significant decay intensity gradient with the later-placed supplementary colonization units. When the intervals between surveys of colonization units are uneven, the time-series linear regression uses the weighted least squares method, with weights determined by the reciprocal of the time intervals between adjacent surveys. Survey data with shorter time intervals have higher weights, making the decay trend diagram more sensitive to recent monitoring data. The overall effectiveness attenuation intensity of each colonization unit, along with its coverage and infection rate from previous surveys, was integrated into an attenuation trend map using spatial coordinates. In the map, nodes with high attenuation intensity were concentrated in slope sections with high heavy metal concentrations and relatively low cohesion after correction, which had a high degree of overlap with the spatial distribution of low intensity level zones in the colonization level zoning set. Due to the continuous disturbance of the root matrix by runoff in the gully development area of ​​the slope, high attenuation intensity clusters usually appeared in the attenuation trend map in this area, which became the key area for subsequent attenuation precursor identification.

[0059] In some embodiments, the step of identifying precursor features of decay phase transition based on the decay trend map and determining the critical value for tiered reseeding and inoculum replenishment includes: performing a time-series analysis of the synchronicity between coverage and infection rate on the decay trend map to generate a synchronicity sequence; analyzing the increasing divergence trend of coverage-infection rate based on the synchronicity sequence to obtain a divergence trend parameter; jointly identifying precursors of underground pre-decay based on the synchronicity sequence and the divergence trend parameter to form a precursor feature set; and determining the critical value for tiered reseeding and inoculum replenishment according to the severity of decay based on the precursor feature set.

[0060] A synchronicity sequence is generated by performing a time-series analysis of the synchronicity between coverage and infection rate on the decay trend chart. The absolute value of the normalized difference between coverage and infection rate at each survey time for each colonization unit is taken as the econometric basis. Normalization is performed by using the maximum time-series value of each colonization unit as the divisor, ensuring that coverage and infection rate are comparable within the range of 0 to 1. A smaller absolute value of the difference indicates that the two performance indicators are currently at similar relative levels and have high synchronicity. The normalized difference is calculated by pairing the coverage and infection rate data of each colonization unit at each time point in the decay trend chart. The differences are arranged according to the survey time sequence to form the synchronicity time sequence for that colonization unit. Periods with persistently low differences correspond to a stage where coverage and infection rate decay together, and their relative levels decline synchronously. Periods with persistently high differences correspond to a stage where their relative levels diverge and synchronicity is impaired. At colonization units with high heavy metal concentrations, arbuscular mycorrhizal fungi generally exhibit weaker resistance to heavy metal toxicity than grass species. Their infection rate shows a unilateral decline even before the overall cover decreases significantly during the same period. The synchronicity time series shows a rapid increase in difference during this period, contrasting with the stable period of low difference in colonization units with lower heavy metal concentrations. The synchronicity time series of each colonization unit are sequentially arranged according to the survey time and then concatenated to form a synchronicity sequence. The spatial dimension corresponds to each colonization unit in the application sequence, and the temporal dimension corresponds to each survey time. The synchronicity sequence also inherits the comprehensive effectiveness decay intensity value of each colonization unit from the decay trend diagram, retaining it as a spatial identifier of the early decay degree of each colonization unit.

[0061] The divergence trend parameter is obtained by analyzing the increasing divergence trend between coverage and infection rate based on synchronous sequence analysis. The increasing divergence trend is confirmed by the sign of the difference between the absolute values ​​of the differences between adjacent survey times in the synchronous sequence. The existence of a divergence trend is confirmed when the absolute value of the difference increases for three consecutive survey times. A single survey time with a reversed difference does not interrupt the confirmed divergence trend, thus accommodating occasional fluctuations in the monitoring data. After the divergence trend is confirmed, the divergence trend parameter is quantified by the linear regression slope of the absolute values ​​of the differences within the continuous divergence period of the colonization unit. A larger slope indicates a faster rate of separation between the relative levels of coverage and infection rate. The divergence trend parameter synchronously records the starting survey time and the current duration of the divergence trend. Longer durations indicate a higher degree of stability in the divergence trend of the colonization unit. Short-term and continuous divergences are distinguished by the duration field for manual verification. Colonization units with an absolute difference value consistently below 0.05 in the synchronicity sequence are classified as being in a highly synchronic stable state. In this state, divergence trend parameters are not calculated, and are recorded as zero, clearly distinguishing them from colonization units that have triggered divergence trends in the parameter list. For colonization units on mid-slopes affected by intermittent runoff, the absolute difference value in their synchronicity sequence briefly increases after each heavy rainfall event, then decreases. The divergence trend difference sign shows a single positive abrupt change near the time of heavy rainfall, but subsequent consecutive differences remain negative, failing to meet the confirmation condition of increasing for three consecutive moments. The divergence trend parameter is not erroneously triggered, and the continuous requirement of the confirmation condition effectively filters out occasional separation phenomena caused by rainfall disturbances.

[0062] A precursor feature set is formed by jointly identifying underground pre-decay precursors using synchronicity sequences and divergence trend parameters. An underground pre-decay precursor is identified when the infection rate decline time series precedes the coverage decline time series. The lead time is determined by the difference between the survey time when the normalized value of the infection rate time series first shows a monotonically decreasing trend and the time when the normalized value of the coverage time series first shows a monotonically decreasing trend. A positive difference indicates that infection precedes coverage decline, while a negative difference indicates that coverage precedes infection rate decline. Underground pre-decay precursors are only valid when the difference is positive. Joint identification requires both the synchronicity sequence and the divergence trend parameter to meet the identification conditions simultaneously. Colony units with a confirmed upward trend in the absolute value of the difference in the synchronicity sequence and a divergence trend parameter slope exceeding one standard deviation of the global mean are included in the underground pre-decay precursor identification. Colony units that do not meet the divergence trend parameter threshold are not included in the lead time calculation to avoid misjudgment due to occasional synchronicity fluctuations with low divergence intensity. The product of the lead time of underground pre-decay and the slope of the divergence trend parameter forms the precursor intensity index for each colonization unit. A larger precursor intensity index indicates a stronger precursor signal of declining underground remediation efficiency in that colonization unit. Colonization units with long lead times and large divergence slopes signify that the root mycorrhizal symbiotic structure has undergone systematic degradation, the aboveground cover is still in the inertial maintenance stage, but underground support has begun to be lost. The precursor intensity index, lead time, slope of the divergence trend parameter, and initial trigger time of each colonization unit are compiled into a precursor feature set by unit, organized with the colonization unit identifier as the row and each precursor index as the column. The distribution pattern of the precursor intensity index in all colonization units characterizes the spatial concentration of the decline in slope underground remediation efficiency. The clustering trend of high precursor intensity areas on the slope is consistent with the high attenuation intensity area in the attenuation trend map.

[0063] The critical values ​​for tiered reseeding and inoculation are determined based on the severity of attenuation according to the precursor feature set. The severity of attenuation is quantified by the precursor intensity index (the product of the slope of the divergence trend parameter and the leading time of underground pre-attenuation) in the precursor feature set. The precursor intensity index is arranged from high to low in all colonization units in the precursor feature set and divided into three levels of high severity, medium severity, and low severity by the tertiles. The tertile boundaries are dynamically determined according to the actual distribution of the precursor feature set. When the distribution is skewed, the tertile thresholds are shifted accordingly to reflect the concentration of attenuation on the slope. For colony units at high severity levels, the critical value for tiered reseeding and bacterial supplementation is set at the third percentile of the high severity level, triggering an immediate reseeding and bacterial supplementation response. This response includes doubling the reseeding density and increasing the bacterial supplementation amount to the upper limit of the enhanced section in the control scheme set. For colony units at medium severity levels, the critical value is set at the time the precursor intensity index enters the medium severity range, triggering an early warning reseeding. The reseeding density is 1.3 times the parameter of the normal section in the control scheme set. For colony units at low severity levels, the critical value is set at the time the slope of the deviation trend parameter exceeds the mean, but the precursor intensity index has not yet entered the medium severity range, triggering intensified monitoring without triggering substantive reseeding and bacterial supplementation. Intensified monitoring shortens the next survey interval to half the normal interval. Colony units with a precursor intensity index of zero in the precursor feature set are considered to be in a precursor-free state, and the corresponding tiered reseeding and bacterial supplementation critical value is set to execute according to the normal monitoring rhythm, without triggering any level of reseeding and bacterial supplementation response. The response level threshold of each colonization unit and the corresponding response content of each level are combined to form a systematic tiered reseeding and re-inoculation threshold. The spatial distribution of high-severity colonization units is consistent with the high-intensity area of ​​the precursor feature set. The spatial consistency between the two confirms the complete inheritance of the precursor feature set information by the threshold grading results.

[0064] Based on the control scheme set and the critical values ​​for tiered reseeding and inoculum replenishment, a multi-threshold hierarchical decision-making process for reseeding and inoculum replenishment is implemented to output slope collaborative restoration control instructions. The multi-threshold hierarchical decision-making uses the hierarchical critical threshold of each colonization unit in the tiered reseeding and inoculum replenishment critical values ​​as the judgment benchmark. The current precursor intensity index of each colonization unit is compared step-by-step with the corresponding hierarchical critical threshold in the tiered reseeding and inoculum replenishment critical values, proceeding from the highest severity critical value to the lowest severity critical value. The hierarchical scheme corresponding to the first critical threshold exceeded is the reseeding and inoculum replenishment response level for that colonization unit. The segment assignment of each colonization unit in the control scheme set is used as the basis for adjusting control parameters in the multi-threshold hierarchical decision-making. When colonization units in a reinforced segment trigger the same response level, the reseeding and inoculum replenishment amounts are increased by 20% based on the corresponding template parameters in the control scheme set; for colonization units in a regular segment, the template parameters of the control scheme set are directly adopted without additional adjustment. The colonization units near the slope catchment line are subject to continuous runoff scouring, resulting in a higher overall effectiveness attenuation intensity in the attenuation trend chart and consequently higher precursor intensity indices. This leads to a higher frequency of triggering high-severity response levels compared to other areas of the slope. The multi-threshold grading decision-making process increases the adjustment of its control scheme template parameters by 30% after more than three cumulative triggers, indicating that the restoration needs of the high-frequency triggering area exceed the conventional coverage of the template parameters. The output of the multi-threshold grading decision-making process includes the current response level, the reseeding and inoculum application amounts in the corresponding grading scheme, and the inoculation time window. The inoculation time window is read from the tiered ratio sequence to determine the appropriate inoculation period for the application batch to which the colonization unit belongs, ensuring that reseeding and inoculum application are performed within the appropriate water potential range. The three decision outputs for each colonization unit are arranged sequentially according to the prescribed order in the application sequence to form a complete version of the slope collaborative restoration control instruction. This control instruction drives the execution of on-site grid construction and reseeding / inoculum application operations. The colony units that did not trigger any critical thresholds in the slope collaborative restoration and control instructions will execute the routine maintenance parameters specified in the control scheme set and will only re-enter the multi-threshold comparison process after the next monitoring survey.

[0065] To implement the above-described method embodiments, a method for the synergistic restoration of mine slopes using turf and microorganisms is proposed to achieve the corresponding functional and technical effects. See also... Figure 2 , Figure 2 This diagram illustrates a structural block diagram of a mine slope turf grid and microbial synergistic restoration system 200 provided in an embodiment of this application. For ease of explanation, only the parts relevant to this embodiment are shown. The mine slope turf grid and microbial synergistic restoration system 200 provided in this embodiment includes: Data parsing module 201 is used to collect heavy metal pollution detection data and slope mechanical parameters, and to perform pollution-slope stability coupling analysis on the heavy metal pollution detection data and the slope mechanical parameters to construct a coupling feature set; The partitioning construction module 202 is used to generate a colonization potential map by performing colonization self-enhancing analysis based on the coupled feature set, identify colonization core nodes in the colonization potential map to generate a colonization node set, and construct a colonization level partitioning set by calibrating the colonization diffusion range based on the colonization node set. The sequence planning module 203 is used to perform colonization unit failure propagation analysis on the colonization level partition set to determine the failure blocking succession table, sort the implementation priority according to the failure blocking succession table to generate an implementation sequence, and parse the key lattice node parameters based on the implementation sequence to form a construction execution configuration. The proportioning control module 204 is used to perform root layer water potential fluctuation analysis on the construction execution configuration to generate a water potential feature map, implement inoculation gradient planning based on the water potential feature map to obtain a tiered proportioning sequence, and identify the enhanced section and the conventional section according to the tiered proportioning sequence to construct a set of control schemes; The instruction output module 205 is used to perform coverage-infection rate effectiveness decay trend analysis on each colonization unit corresponding to the application sequence to generate a decay trend map, identify the decay phase transition precursor characteristics based on the decay trend map to determine the critical value of tiered reseeding and inoculation, and perform multi-threshold hierarchical decision-making for reseeding and inoculation based on the control scheme set and the critical value of tiered reseeding and inoculation to output slope collaborative restoration control instructions.

[0066] The aforementioned mine slope turf grid and microbial synergistic restoration system 200 can implement the mine slope turf grid and microbial synergistic restoration method described in the above-described method embodiments. The options in the above method embodiments are also applicable to this embodiment and will not be detailed here. The remaining contents of this application embodiment can be referred to the contents of the above method embodiments, and will not be repeated in this embodiment.

[0067] The purpose of the above embodiments is to reproduce and derive the technical solution of the present invention by way of example, and to fully describe the technical solution, purpose and effect of the present invention. The purpose is to enable the public to have a more thorough and comprehensive understanding of the disclosure of the present invention, and not to limit the scope of protection of the present invention.

Claims

1. A method for the synergistic restoration of mine slopes using turf lattice and microorganisms, characterized in that, include: Collect heavy metal pollution detection data and slope mechanical parameters, and construct a coupled feature set by performing pollution-slope stability coupling analysis on the heavy metal pollution detection data and the slope mechanical parameters. Based on the coupled feature set, a colonization self-enhancing analysis is performed to generate a colonization potential map. The colonization core node is identified in the colonization potential map to generate a colonization node set. Based on the colonization node set, the colonization diffusion range is calibrated to construct a colonization level partition set. A failure propagation analysis of the colonization unit is performed on the colonization level partition set to determine the failure blocking succession table. The construction priority is sorted according to the failure blocking succession table to generate a construction sequence. The key lattice node parameters are parsed based on the construction sequence to form a construction execution configuration. A root layer water potential fluctuation analysis is performed on the construction execution configuration to generate a water potential feature map. Based on the water potential feature map, inoculation gradient planning is implemented to obtain a tiered mix ratio sequence. Based on the tiered mix ratio sequence, an enhanced section and a conventional section are identified to construct a set of control schemes. A decay trend map is generated by analyzing the coverage-infection rate effectiveness decay trend of each colonization unit corresponding to the application sequence. Based on the decay trend map, the characteristics of the decay phase transition precursor are identified to determine the critical value of tiered reseeding and inoculum replenishment. Based on the control scheme set and the critical value of tiered reseeding and inoculum replenishment, a multi-threshold hierarchical decision-making process for reseeding and inoculum replenishment is performed to output the slope collaborative restoration control command.

2. The method according to claim 1, characterized in that, The step of generating a colonization potential map based on the coupled feature set through colonization self-enhancement analysis includes: The inhibition release threshold of the coupled feature set was determined by arbuscular mycorrhizal fungi, and an inhibition release threshold map was generated. Based on the inhibition release threshold map, the grass seed tillering diffusion radius is calculated to obtain the grass seed diffusion radius map; The inhibition relief threshold map and the grass seed diffusion radius map are overlaid to form a colonization potential map.

3. The method according to claim 1, characterized in that, The step of identifying core colonization nodes and generating a colonization node set from the colonization potential map includes: The colonization potential map is subjected to potential throat point detection to generate a potential throat point set; Based on the potential throat point set, perform lattice node position matching to obtain a candidate node set; The candidate node set is subjected to diffusion blank area connectivity gain evaluation to generate connectivity gain parameters; The optimal connectivity throat node is selected based on the connectivity gain parameter to determine the colony node set.

4. The method according to claim 1, characterized in that, The step of performing colonization unit failure propagation analysis on the colonization level partition set to determine the failure blocking succession table includes: Slope runoff paths are calculated for the colonization level zoning set to generate a runoff directed graph; Based on the runoff directed graph, the dependency relationship between upstream and downstream colonization units is identified and a dependency relationship matrix is ​​constructed; Failure cascade path tracing is performed on the dependency matrix to obtain a set of failure propagation paths; Based on the set of failure propagation paths, key cut point units are extracted to determine the failure blocking succession table.

5. The method according to claim 1, characterized in that, The step of performing root layer water potential fluctuation analysis and generating a water potential characteristic map on the construction execution configuration includes: Based on the construction execution configuration, the soil texture parameters of the lattice nodes are analyzed to obtain a texture parameter set; Based on the set of texture parameters, the water-holding capacity of the root layer is calculated to generate a water-holding capacity distribution map. A set of water potential response curves was obtained by performing a rainfall-evaporation disturbance simulation on the water-holding capacity distribution map; Based on the set of water potential response curves, a water potential characteristic map is formed by statistically analyzing the residence time within the appropriate water potential range.

6. The method according to claim 1, characterized in that, The step of identifying the precursor features of the decay phase transition based on the decay trend diagram and determining the critical value for tiered reseeding and inoculation includes: A time-series analysis of the synchronicity between coverage and infection rate is performed on the decay trend graph to generate a synchronization sequence; Based on the analysis of the synchronous sequence, the divergence trend of the coverage-infection rate divergence is increased, and the divergence trend parameter is obtained. A precursor feature set is formed by jointly identifying underground pre-decay precursors from the synchronicity sequence and the divergence trend parameters. The critical values ​​for tiered reseeding and bacterial replenishment are determined based on the severity of attenuation according to the aforementioned precursor feature set.

7. The method according to claim 1, characterized in that, The step of implementing inoculation gradient planning based on the water potential feature map to obtain the tiered matching sequence includes: The water potential feature map is used to match the suitable water potential range for the inoculant to generate a suitable inoculation zone map; Based on the analysis of the suitable inoculation zone map, the ratio gradient of Bacillus and Arbuscular Mycorrhizal Fungi was obtained to obtain an initial ratio scheme; The initial formulation scheme is subjected to appropriate inoculation zoning gradient correction to form a corrected formulation scheme; The corrected proportioning scheme is used to generate a tiered proportioning sequence by arranging the proportions in a time sequence.

8. The method according to claim 3, characterized in that, The step of evaluating the connectivity gain of the candidate node set by expanding the blank area to generate connectivity gain parameters includes: A penetration reachability map is generated by extrapolating the bidirectional penetration path of grass and bacteria in the surrounding blank areas of the candidate node set. The non-redundant coverage increment set is obtained by removing multi-node overlapping redundant segments from the penetration reachability map; The non-redundant coverage increment set is corrected for reachability domain boundary distance to obtain the correction gain set; The connectivity gain parameters are determined by selecting the node with the optimal connectivity contribution based on the corrected gain set.

9. The method according to claim 4, characterized in that, The process of identifying the dependency relationships between upstream and downstream colonization units and constructing a dependency matrix based on the runoff directed graph includes: Extract the runoff intensity classification characteristics of each colonization unit in the runoff directed graph to generate a unit runoff net flux distribution map; Based on the above unit runoff net flux distribution map, the runoff connection relationship between upstream and downstream units is marked to obtain the connection relationship diagram; Based on the aforementioned connection diagram, a set of transmission priority is obtained by prioritizing the cross-unit failure propagation paths. A dependency matrix is ​​constructed by weighting the failure probability of the transmission priority set and calibrating the dependency strength.

10. A synergistic restoration system for turf and microorganisms on mine slopes, characterized in that, include: The data parsing module is used to collect heavy metal pollution detection data and slope mechanical parameters, and to perform pollution-slope stability coupling analysis on the heavy metal pollution detection data and the slope mechanical parameters to construct a coupling feature set; The partitioning construction module is used to generate a colonization potential map by performing colonization self-enhancing analysis based on the coupled feature set, identify colonization core nodes in the colonization potential map to generate a colonization node set, and construct a colonization level partitioning set by calibrating the colonization diffusion range based on the colonization node set. The sequence planning module is used to perform colonization unit failure propagation analysis on the colonization level partition set to determine the failure blocking succession table, sort the implementation priority according to the failure blocking succession table to generate an implementation sequence, and parse the key lattice node parameters based on the implementation sequence to form a construction execution configuration. The mix proportion control module is used to perform root layer water potential fluctuation analysis on the construction execution configuration to generate a water potential feature map, implement inoculation gradient planning based on the water potential feature map to obtain a tiered mix proportion sequence, and identify the enhanced section and the conventional section according to the tiered mix proportion sequence to construct a set of control schemes; The instruction output module is used to perform coverage-infection rate effectiveness decay trend analysis on each colonization unit corresponding to the application sequence to generate a decay trend map, identify the characteristics of the decay phase transition precursor based on the decay trend map to determine the critical value of tiered reseeding and inoculation, and perform multi-threshold hierarchical decision-making for reseeding and inoculation based on the control scheme set and the critical value of tiered reseeding and inoculation to output slope collaborative restoration control instructions.

Images