Low-altitude flight scene edge computing node multi-constraint deployment method and system

Abstract

The application discloses a low-altitude flight scene edge computing node multi-constraint deployment method and system, and belongs to the technical field of edge computing infrastructure deployment. The method comprises the following steps: acquiring regional basic data of a target low-altitude flight area; dividing the target area into four types of low-altitude service units, and differentially configuring the number of coverage nodes and the main-backup redundancy requirement for different types of units; determining candidate deployment carriers and calculating feasibility identifiers, and eliminating unfeasible carriers; calculating a comprehensive coverage evaluation value based on three dimensions, generating a coverage relationship matrix; establishing a coverage overlap model to generate a main-backup overlap feasibility identifier; calculating a node deployment adaptation degree; performing a feasibility pre-check; establishing and solving a 0-1 site selection optimization model to determine the deployment position of the edge computing node and the main-backup relationship; and outputting a deployment scheme. The method can improve the coverage effectiveness, reliability and scheme implementability of the edge computing node deployment in the low-altitude flight scene.

Description

Technical Field

[0001] This invention relates to the fields of edge computing infrastructure deployment, low-altitude flight infrastructure planning, low-altitude operation support network planning, and multi-constraint location optimization technology, specifically to a multi-constraint deployment method and system for edge computing nodes in low-altitude flight scenarios. Background Technology

[0002] With the rapid development of scenarios such as drone logistics, low-altitude inspection, low-altitude security, and low-altitude traffic management, low-altitude flight routes, take-off and landing points, flight hotspots, and risk areas have created an urgent need for the proximity and computing capabilities of edge computing nodes. Against this backdrop, how to scientifically deploy edge computing nodes to effectively cover various low-altitude service areas while simultaneously considering multiple practical constraints such as power supply, network backhaul, installation space, structural load-bearing capacity, operational accessibility, and construction costs has become a critical issue that urgently needs to be addressed in the construction of low-altitude flight infrastructure.

[0003] Currently, edge computing node deployment mainly relies on three methods: First, deployment is based on communication base station or data center locations, primarily considering ground network conditions and data center resources, without fully taking into account the spatial orientation of low-altitude flight routes, the distribution of take-off and landing points, and the actual needs of flight hotspots, resulting in insufficient spatial matching between deployment locations and low-altitude flight demands; second, drones are used as mobile edge servers to provide computing services, but drones are limited by onboard computing power, flight time, and communication bandwidth, making it difficult to undertake large-scale, continuous edge computing services, and focusing on task offloading optimization rather than fixed node site selection planning; third, general facility site selection optimization methods such as integer programming and set coverage models are used, but these are directly applied to low-altitude areas. In flight scenarios, differentiated modeling and configuration were not performed for different types of areas such as routes, takeoffs and landings, hotspots, and risks. 3D coverage constraints such as horizontal distance, vertical height deviation, and building / obstacle obstruction were not fully considered. Before site selection, a unified feasibility pre-screening was not conducted on the installation space, structural load-bearing capacity, power supply capacity, network backhaul capacity, security protection, maintenance accessibility, and restrictions on installation of candidate carriers. There was a lack of primary and backup redundancy guarantee mechanisms for key areas such as takeoffs and landings, hotspots, and risks. Furthermore, multiple constraints such as distance, altitude, obstruction, infrastructure, maintenance, cost, number of nodes, and primary / backup redundancy were not solved collaboratively in the same optimization model, making it difficult to obtain a globally optimal and practically feasible deployment plan.

[0004] To address the aforementioned issues, existing patent literature covers aspects such as edge computing for low-altitude mapping, edge management of flight data, safety monitoring and scheduling, and resource allocation for task offloading. However, none of these patents address the deployment and site selection methods for edge computing nodes on fixed ground platforms, particularly regarding technical solutions for service unit differentiation, feasibility pre-screening of candidate platforms, primary / backup overlapping coverage modeling, and multi-constraint collaborative optimization. Therefore, it is necessary to propose an edge computing node deployment method and system for low-altitude flight scenarios. By incorporating coverage, distance, altitude, occlusion, infrastructure, operation and maintenance, cost, node quantity, and primary / backup redundancy into a unified site selection optimization framework, the coverage effectiveness, reliability, and feasibility of the node deployment scheme can be improved. Summary of the Invention

[0005] The purpose of this invention is to address the aforementioned problems by providing a multi-constraint deployment method and system for edge computing nodes in low-altitude flight scenarios. This involves acquiring basic data of the low-altitude flight area, dividing it into low-altitude service units, identifying candidate deployment carriers, establishing a coverage relationship matrix and primary / backup overlap relationships, and determining the deployment location, node type, coverage area, and primary / backup relationships of the edge computing nodes based on a multi-constraint location optimization model that can be implemented by a computer program.

[0006] The technical solution of the present invention is as follows: A multi-constraint deployment method for edge computing nodes in low-altitude flight scenarios includes the following steps: Acquire basic regional data for the target low-altitude flight area. The basic regional data should include at least the low-altitude flight path, take-off and landing points, flight altitude range, building distribution, obstacle distribution, power supply conditions, network backhaul conditions, construction costs, and prohibited and restricted areas. Based on the spatial distribution of low-altitude flight routes and take-off and landing points, the target low-altitude flight area is divided into multiple low-altitude service units. Each low-altitude service unit includes an en-route service unit and a take-off and landing service unit, and a minimum number of coverage nodes is configured for each low-altitude service unit. and maximum allowed service distance The minimum number of coverage nodes for take-off and landing service units. Greater than the minimum number of coverage nodes for an airway service unit; Candidate deployment carriers are identified within or near the low-altitude service unit, and the feasibility index of each candidate deployment carrier is calculated based on installation space, structural load-bearing capacity, power supply capacity, network backhaul capacity, and prohibition sign. Candidate deployment carriers with a feasibility index of 0 are eliminated. For each low-altitude service unit and each candidate deployment carrier, based on distance adaptation score High matching score And occlusion correction score Calculate the comprehensive coverage evaluation value ,in And based on this, a covering relation matrix is ​​generated; Establish a 0-1 integer programming model with candidate deployment carriers as decision variables. The 0-1 integer programming model should include at least coverage constraints. Feasibility constraints and cost constraints ,in Indicates an overlay relationship. Indicates feasibility. Indicates deployment cost; Solve the 0-1 integer programming model to determine the selected candidate deployment carriers as the deployment locations of edge computing nodes, and output the deployment scheme.

[0007] By comprehensively determining the coverage relationship through three dimensions—distance, height, and occlusion—and constructing a 0-1 integer programming model with coverage constraints, feasibility constraints, and cost constraints, the deployment scheme can simultaneously meet the three-dimensional spatial coverage requirements, the actual deployable conditions of the carrier, and the construction cost limitations. This effectively solves the problems of large deviations in two-dimensional coverage analysis, the inability to implement the scheme due to the lack of screening of candidate carriers, and the inability to coordinate and optimize multiple constraints in existing technologies.

[0008] Furthermore, the low-altitude service unit also includes a hotspot service unit and a risk service unit; the route service unit generates a spatial buffer zone based on the low-altitude route centerline and divides it according to segment length; the take-off and landing service unit generates a spatial buffer zone centered on the location of the take-off and landing point or apron; the hotspot service unit aggregates and divides continuous grids with density exceeding a threshold according to the flight density grid; the risk service unit is defined according to the density of obstacles or airspace management sensitive areas; the minimum number of coverage nodes for the hotspot service unit and the risk service unit are specified. All of these are greater than the minimum number of coverage nodes for an airway service unit.

[0009] By employing different partitioning algorithms (spatial buffer, grid density aggregation, and sensitive area delineation) for four types of service units—routes, takeoffs and landings, hotspots, and risks—and configuring differentiated coverage node numbers, the redundancy support level of key areas is higher than that of ordinary route areas, achieving tiered support with limited resources.

[0010] Furthermore, the distance adaptation score according to Confirmed, the high matching score according to Confirmed, the occlusion correction score according to Confirmed, among which For the amplitude limiting function, The distance between the candidate deployment vehicle and the low-altitude service unit. The altitude deviation between the installation height and the flight altitude range, For the maximum permissible height deviation, To mask the impact value.

[0011] By using a bounding normalization method, distance matching degree, height matching degree, and occlusion correction degree are unified to a comparable scale, enabling the three-dimensional spatial coverage quality to be quantitatively evaluated and weighted and fused, thus solving the problem that two-dimensional planar analysis cannot truly reflect the low-altitude coverage effect.

[0012] Furthermore, the occlusion impact value The target low-altitude flight area is determined as follows: a three-dimensional geometric scene of the target area is generated based on the distribution of buildings and obstacles; a straight line segment is launched from the spatial position of the candidate deployment vehicle to the geometric center of the low-altitude service unit; if the straight line segment intersects with any building or obstacle, then... If they do not intersect, then Comprehensive coverage evaluation value The score also includes infrastructure points. , ,in , , ; in, Indicates the available installation space for the candidate deployment carrier. This represents the maximum available installation space across the current candidate carrier set. The structural bearing capacity of the candidate deployment carrier, This represents the maximum load-bearing capacity of all structures in the current candidate carrier set. The available power supply capacity for candidate deployment carriers, This represents the maximum available power supply capacity among all candidate carriers in the current set. For the available network backhaul capabilities of candidate deployment carriers, This represents the maximum value of the available network backhaul capacity in the current candidate carrier set.

[0013] By detecting line-of-sight occlusion through the intersection of rays with the 3D scene, the impact value of occlusion has a clear spatial geometric basis. At the same time, the six engineering conditions of the carrier, namely installation space, load-bearing capacity, power supply, return route, safety protection and operation and maintenance accessibility, are quantified into infrastructure scores, which solves the problem that the engineering conditions of candidate carriers cannot be uniformly evaluated.

[0014] Furthermore, the feasibility identifier according to:

[0015] Confirmed, among which For indicator functions, Available installation space, For structural load-bearing capacity, For available power supply capacity, For available network backhaul capabilities, For safety protection rating, To score operational accessibility, This is a "No Installation" sign; , , , These are the corresponding feasibility thresholds.

[0016] The feasibility indicator is the product of seven engineering conditions (installation space, load-bearing capacity, power supply capacity, return range capacity, safety protection, operation and maintenance accessibility, and prohibition signage). If any condition is not met, the indicator is zero, which realizes a hard threshold screening with "one vote veto" to ensure that all candidate carriers participating in subsequent optimization have the actual deployment conditions.

[0017] Furthermore, it also includes calculating the node deployment adaptability of candidate deployment carriers. :

[0018] in to These are the weighting coefficients, and their sum is 1; To contribute points to coverage, Low-altitude service unit Service weight, Indicates an overlay relationship; The overall score is based on distance. For a highly comprehensive score; The overall score is based on the degree of occlusion. Score for cost. To score for operations and maintenance, For deployment costs, Due to the difficulty of operation and maintenance, To cap deployment costs, This represents the upper limit of operational and maintenance difficulty. This is a limiting function; For redundant collaboration scoring, Primary and backup overlap feasibility identifier. As a low-altitude service unit, As a key low-altitude service unit cluster, This represents the number of candidate carriers other than the candidate deployment carrier itself.

[0019] The eight dimensions of coverage contribution, distance matching, height matching, occlusion correction, infrastructure conditions, cost, operation and maintenance difficulty, and redundancy coordination capability are uniformly quantified into node deployment adaptability, so that the comprehensive deployment value of each candidate carrier can be fully evaluated and compared, providing a basis for decision-making in optimizing the model.

[0020] Furthermore, the 0-1 integer programming model is solved with the following optimization objective:

[0021] in to These are weighting coefficients, corresponding to the priority weights of deployment adaptability, primary / standby redundancy capability, deployment cost penalty, operation and maintenance difficulty penalty, and number of nodes, respectively. as candidate deployment carriers Node deployment adaptability, as candidate deployment carriers The choice of variables, To select a deployment carrier With candidate deployment carriers low-altitude service units The coverage overlap ratio, Primary and backup paired variables, To optimize the objective function value, the 0-1 integer programming model also includes a node number constraint. .

[0022] The five objectives of deployment adaptability, primary / standby redundancy capability, cost, operation and maintenance difficulty, and number of nodes are incorporated into the same optimization function. The priority of each item is flexibly configured through weight coefficients, so that the coupling relationship between multiple constraints is handled in a unified manner, avoiding local optima and constraint conflicts caused by separate optimization.

[0023] Furthermore, it also includes establishing a primary / backup overlapping coverage relationship: Low-altitude service unit Discretize into multiple sampling points and calculate candidate deployment carriers. and Coverage overlap ratio ,in, Indicates candidate deployment carrier Does it cover low-altitude service units? The One sampling point, as candidate deployment carriers Does it cover low-altitude service units? The One sampling point, If and only if To the The distance between sampling points does not exceed Height deviation not exceeding And the line of sight must be unobstructed; otherwise, the value is 0. This represents the total number of sampling points; when , and At that time, enable the primary and backup overlap to be feasible. ,in Minimum primary / backup coverage overlap ratio; 0-1 integer programming models also include primary and backup paired variables. In addition, there are primary / backup overlap constraints for takeoff and landing service units, hotspot service units, and risky service units: ,and , , , , as candidate deployment carriers The choice of variables.

[0024] By calculating the coverage overlap ratio of sampling points, a feasible identifier for primary and backup overlap is established. The primary and backup redundancy relationship is encoded into a hard constraint of the optimization model through primary and backup paired variables and their linearization constraints. This ensures that key service units are simultaneously covered by at least one pair of nodes that meet the coverage overlap requirements, thus solving the problem of lack of quantitative redundancy guarantee in key areas.

[0025] Furthermore, determining the primary service node and the backup service node includes: For route service units, within the selected candidate deployment carrier set Select to satisfy and The largest edge computing node serves as the primary service node; For takeoff and landing service units, hotspot service units, and high-risk service units, within the selected candidate deployment carrier set... Enumerate all that satisfy Candidate deployment carriers Calculate pairwise scores Select the candidate deployment pair with the highest pair score as the primary / backup candidate pair. One node from Sun Yat-sen University was designated as the primary service node, and the other node was designated as the backup service node; among them... as candidate deployment carriers Node deployment adaptability, as candidate deployment carriers low-altitude service units The comprehensive coverage evaluation value.

[0026] Different allocation strategies are adopted for route units and key units—route units can be covered by a single node, while key units must be allocated in pairs—and coverage quality and node compatibility are considered when selecting pairs to ensure that the primary and backup allocation has both coverage effectiveness and engineering feasibility in actual deployment.

[0027] This application also includes a multi-constraint deployment system for edge computing nodes in low-altitude flight scenarios, and executes a multi-constraint deployment method for edge computing nodes in low-altitude flight scenarios, including: The data acquisition module is used to acquire basic regional data of the target's low-altitude flight area; The service unit partitioning module is used to divide the target low-altitude flight area into multiple low-altitude service units based on the spatial distribution of the low-altitude routes and take-off and landing points. Each low-altitude service unit includes at least a route service unit and a take-off and landing service unit, and a minimum number of coverage nodes is configured for each low-altitude service unit. and maximum allowed service distance Among them, the take-off and landing service unit Larger than the route service unit ; The candidate carrier screening module is used to determine candidate deployment carriers and calculate feasibility indicators based on installation space, structural load-bearing capacity, power supply capacity, network backhaul capacity and prohibition indicators, and eliminate infeasible carriers; Coverage relationship building module, used for distance-based adaptation score High matching score And occlusion correction score Calculate the comprehensive coverage evaluation value And generate a coverage matrix; The site selection optimization module is used to establish and solve a 0-1 integer programming model with candidate deployment carriers as decision variables. The model includes at least coverage constraints. Feasibility constraints and cost constraints ; The solution output module is used to output the installation location of the selected edge computing nodes and the low-altitude service units they cover.

[0028] Compared with existing technologies, the advantages of this invention are: 1. To address the deficiency in existing methods in the background art that do not perform differentiated modeling for different types of areas, this application divides the low-altitude flight area into four service units based on low-altitude routes, take-off and landing points, flight hotspots, and risk areas. Furthermore, it differentiates the minimum number of coverage nodes for each type of service unit. Minimum primary / backup coverage overlap ratio This results in higher requirements for coverage redundancy and backup support for key service units such as takeoffs and landings, hotspots, and high-risk areas compared to ordinary airway service units. Consequently, a tiered support strategy of "prioritizing key areas and providing basic coverage for ordinary areas" can be implemented with limited deployment resources, avoiding insufficient support in key areas or wasted resources in ordinary areas due to a "one-size-fits-all" deployment approach.

[0029] 2. Addressing the shortcomings of existing methods in the background art, which are primarily based on two-dimensional planar analysis and do not adequately consider height deviation and line-of-sight obstruction, this application improves upon these shortcomings by enhancing the comprehensive coverage evaluation value. Distance adaptation score is also introduced in the middle. High matching score And occlusion correction score Furthermore, the coverage determination at the sampling point level requires that three conditions—distance, height deviation, and line-of-sight obstruction—must be met simultaneously for coverage to be recognized. This ensures that the deployment plan accurately reflects the actual coverage of low-altitude aircraft in three-dimensional space, avoiding the significant discrepancies between theoretical and actual coverage caused by neglecting height deviation and building obstruction in traditional two-dimensional planning.

[0030] 3. To address the deficiency in existing methods in the background art that do not conduct a unified feasibility screening of candidate carriers before site selection, this application uses feasibility identification... Candidate deployment carriers are pre-screened, eliminating those that do not meet any of the following conditions: installation space, structural load-bearing capacity, power supply capacity, network backhaul capability, security protection, maintenance accessibility, and prohibition of installation. Only those carriers that do not meet these conditions are considered. Feasible carriers are then incorporated into the subsequent optimization model. This ensures that all selected candidate carriers have the practical deployment conditions, preventing the optimization model from selecting deployment nodes on infeasible carriers and thus avoiding outputting solutions that cannot be implemented.

[0031] 4. To address the deficiency in existing methods in the background art of lacking a primary / backup redundancy guarantee mechanism for key areas, this application discretizes low-altitude service units into sampling points and calculates the coverage overlap ratio between candidate deployment carrier pairs. The degree of physical coverage overlap is quantified into a calculable mathematical index, and a feasible identifier for primary and backup overlap is established based on this index. This is incorporated as a hard constraint into the 0-1 location optimization model, and a primary / backup redundancy capability term is set in the optimization objective. This ensures that key service units such as takeoff and landing, hotspots, and risks have both a primary service node and a backup service node that meets the coverage overlap requirements. When the primary node is unavailable due to failure or maintenance, the backup node can quickly take over the service, avoiding service interruption in critical areas due to single point of failure and improving the reliability of edge computing services in critical low-altitude flight areas.

[0032] 5. Addressing the shortcomings of existing methods in the background art, which separately handle multiple constraints such as distance, height, occlusion, infrastructure, operation and maintenance, cost, number of nodes, and primary / backup redundancy, or employ simplified weighting methods, making it difficult to obtain a globally optimal solution, this application incorporates all the above constraints into a single 0-1 site selection optimization model. Through the joint solution of coverage constraints, feasibility constraints, cost constraints, number of nodes constraints, and primary / backup overlap constraints, the optimization objective—comprehensively improving deployment adaptability and primary / backup redundancy capabilities in key areas, while reducing deployment costs, operation and maintenance difficulty, and number of nodes—is achieved optimally or near optimally, while satisfying all constraints. The coupling relationships between various constraints are uniformly handled by the optimization model, avoiding local optima and constraint conflicts caused by separate optimization.

[0033] 6. This application performs a feasibility pre-check before optimization and a final verification of coverage, primary / backup, cost, number of nodes, and feasibility after the solution is obtained. The pre-check identifies unsolvable conditions and outputs the reasons for infeasibility before solving, avoiding wasting computational resources on infeasible problems and preventing the output of incorrect solutions. The final verification verifies whether the solution meets all constraints after solving, preventing infeasibility due to solver numerical errors or model simplification. This dual verification mechanism ensures the reliability of the deployment solution throughout the entire process from model solving to actual output.

[0034] 7. The 0-1 location optimization model of this application is a linear 0-1 integer programming model, which can be solved by various methods such as branch and bound method, cutting plane method, mixed integer programming solver, genetic algorithm or other heuristic algorithm. It can select an appropriate solution strategy according to the actual deployment scale and has good engineering applicability and scalability. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the edge computing node deployment method for this application.

[0036] Figure 2 This is a schematic diagram illustrating the spatial relationship between the low-altitude service unit division and candidate deployment carriers in this application.

[0037] Figure 3 This is a schematic diagram of the coverage relationship matrix between candidate deployment carriers and low-altitude service units in this application.

[0038] Figure 4 This is a schematic diagram of the multi-constraint addressing optimization model structure of this application.

[0039] Figure 5 This is a schematic diagram illustrating the overlapping coverage relationship between the primary service node and the backup service node in this application.

[0040] Figure 6 A schematic diagram of the system architecture for deploying edge computing nodes in this application. Detailed Implementation

[0041] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0042] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0043] This application focuses on determining the deployment location of edge computing nodes, screening candidate deployment carriers, low-altitude service unit coverage, node type configuration, generation of primary and backup redundancy relationships, and multi-constraint site selection optimization. It does not involve video recognition, artificial intelligence recognition algorithms, video compression, specific communication protocols, task migration, or cloud scheduling.

[0044] Please see Figure 1-6 A multi-constraint deployment method for edge computing nodes in low-altitude flight scenarios, such as Figure 1 As shown, it includes the following steps: Acquire regional basic data of the target low-altitude flight area. Regional basic data includes low-altitude routes, take-off and landing points, and flight altitude range. In addition, it may also include one or more of the following: helipad, drone airport, flight mission area, flight density, flight hotspot area, risk area, building distribution, obstacle distribution, communication infrastructure distribution, power supply conditions, network backhaul conditions, operation and maintenance conditions, construction costs, management boundaries, and prohibited and restricted areas. Based on low-altitude flight paths, take-off and landing points, aprons, drone airports, flight hotspots, and risk areas, the target low-altitude flight area is divided into multiple low-altitude service units, forming a set of low-altitude service units:

[0045] Each low-altitude service unit At least include service unit type, spatial range, flight altitude range, and service weight. Minimum number of covered nodes Maximum allowed service distance Maximum permissible height deviation Maximum permissible shading impact Minimum coverage evaluation threshold Minimum primary / backup coverage overlap ratio , ; Within or near the low-altitude service unit, candidate deployment carriers are identified to form a set of candidate deployment carriers:

[0046] Each candidate deployment carrier At least include location Installation height Available installation space Structural bearing capacity Available power supply capacity Available network backhaul capability Safety protection score Operation and maintenance accessibility score No Installation Signs Deployment costs and maintenance difficulty , ; The adjacent area specifically refers to: within the low-altitude service unit, or within the spatial boundary of the low-altitude service unit not exceeding the maximum permissible service distance. Within the designated area, candidate deployment carriers are identified.

[0047] The feasibility indicators of candidate deployment carriers are calculated based on factors such as installation space, structural load-bearing capacity, power supply capacity, network backhaul capacity, security protection, maintenance accessibility, and prohibition signs. ,in:

[0048] in, This is an indicator function that takes the value 1 when the condition is true and 0 otherwise; when At that time, candidate deployment carriers Not involved in subsequent site selection; For each low-altitude service unit With each candidate deployment carrier Calculate the distance adaptation score High matching score Occlusion correction score and infrastructure score And calculate the comprehensive coverage evaluation value:

[0049] in, , , and These are the weight coefficients corresponding to the distance adaptation score, height matching score, occlusion correction score, and infrastructure score, respectively, and they satisfy the following: .

[0050] Generate a coverage relationship matrix based on the comprehensive coverage evaluation value:

[0051] Among them, when candidate deployment carrier Simultaneously satisfy:

[0052] season Otherwise ; Each low-altitude service unit The sample is discretized into multiple sampling points, and the candidate deployment carrier is calculated based on the coverage of the sampling points by the candidate deployment carrier. With candidate deployment carriers low-altitude service units Coverage overlap ratio:

[0053]

[0054] in, Indicates candidate deployment carrier Does it cover low-altitude service units? The One sampling point, Low-altitude service unit The total number of sampling points, as candidate deployment carriers To the The distance between each sampling point as candidate deployment carriers Installation height and the first The height deviation between each sampling point; Indicates candidate deployment carrier With the Whether the line of sight between sampling points is obstructed by buildings or obstacles, 1 indicates no obstruction and 0 indicates obstruction. The determination of line of sight obstruction is related to the coverage at the service unit level. A consistent line-of-sight model is used.

[0055] when:

[0056] At that time, candidate deployment carriers were determined. With candidate deployment carriers low-altitude service units It has a primary / backup overlapping coverage relationship and enables primary / backup overlapping to be feasible. Otherwise ; The node deployment suitability of candidate deployment carriers is calculated based on coverage contribution, distance adaptability, height matching, occlusion correction, infrastructure conditions, cost, operation and maintenance difficulty, and primary / backup overlapping coverage capability. ; A feasibility pre-check was conducted before establishing the site selection optimization model, including for each low-altitude service unit. Determine whether the following conditions are met:

[0057] For takeoff and landing service units, hotspot service units, or high-risk service units, further determine whether there is at least one set of candidate deployment carriers. , so that:

[0058] If any low-altitude service unit does not meet the corresponding feasibility conditions, output the reason for infeasibility; After the feasibility pre-check is passed, a 0-1 site selection optimization model is established; the 0-1 site selection optimization model includes candidate deployment carrier selection variables:

[0059] And primary and backup paired variables:

[0060]

[0061] The 0-1 site selection optimization model includes at least coverage constraints, feasibility constraints, cost constraints, node number constraints, and primary / backup overlap constraints. Solve the 0-1 location optimization model to obtain the set of candidate deployment carriers. ; In the selected set of candidate deployment carriers Within this framework, a primary service node is determined for the low-altitude service unit, and a backup service node is determined for the take-off and landing service unit, the hot spot service unit, or the risk service unit. The backup service node and the corresponding primary service node satisfy the primary and backup overlapping coverage relationship. The obtained deployment plan is verified for coverage, primary / backup, cost, number of nodes, and feasibility. If the verification passes, the edge computing node deployment plan is output; if the verification fails, the reason for infeasibility is output or the solution is recalculated.

[0062] Low-altitude service units include route service units, takeoff and landing service units, hotspot service units, and risk service units. Route service units are defined based on the low-altitude route centerline and route width, with the route width determined by project deployment. Takeoff and landing service units are defined based on the spatial location of takeoff and landing points, aprons, or UAV airports and their surrounding buffer zones. Hotspot service units are defined based on areas with high flight activity or route intersections. Risk service units are defined based on areas with high obstacle density, management-sensitive areas, or areas with high flight risk. Furthermore, a minimum number of coverage nodes is required for each takeoff and landing service unit, hotspot service unit, and risk service unit. It is greater than the minimum number of coverage nodes for a typical air route service unit.

[0063] Distance fit score High matching score Occlusion correction score and infrastructure score Determine as follows:

[0064]

[0065]

[0066] in, To limit the value to Amplitude limiting function for an interval; as candidate deployment carriers With low-altitude service units The distance between them; as candidate deployment carriers Installation height and low-altitude service unit Altitude deviation between the flight altitude range; To determine the impact value of occlusion; Infrastructure score:

[0067] in , , , , , , which is the maximum value of each infrastructure indicator in the current candidate carrier set, used to normalize each indicator to . interval; , , , , and These are the weighting coefficients corresponding to installation space, structural load-bearing capacity, power supply capacity, network backhaul capacity, security protection score, and operation and maintenance accessibility score, respectively, and they satisfy the following: .

[0068] Node deployment adaptability Determine as follows:

[0069] in, , , , , , , and These are the weighting coefficients corresponding to the scores for coverage contribution, distance integration, height integration, occlusion integration, infrastructure, cost, operation and maintenance, and redundancy collaboration, respectively, and they satisfy the following:

[0070]

[0071]

[0072]

[0073] in, To contribute points to coverage, The overall score is based on distance. For a highly comprehensive score, To determine the overall score for occlusion, To score for infrastructure, Score for cost. To score for operations and maintenance, Scoring for redundant collaboration; Redundancy Coordination Score Based on candidate deployment carriers The primary and backup overlap coverage relationship for the same key low-altitude service unit was determined with other candidate deployment carriers. The set of key low-altitude service units includes one or more of the following: takeoff and landing service units, hotspot service units, and risky service units. To cap deployment costs, This represents the upper limit of operational and maintenance difficulty. and Determined by project deployment requirements, constraints imposed by the construction party, or constraints imposed by the operator; For the set of candidate deployment carriers The total number of elements.

[0074] The 0-1 location optimization model includes the following constraints: Coverage constraints:

[0075] Feasibility constraints:

[0076] Cost constraints:

[0077] Node count constraint:

[0078] In addition, there are primary / backup overlap constraints for takeoff and landing service units, hotspot service units, or high-risk service units:

[0079] and:

[0080]

[0081] Among them, the primary and backup overlap constraint is used to ensure that both the primary service node and the backup service node are selected, and that they meet the coverage overlap requirements for the corresponding low-altitude service units. and These represent the lower and upper limits for the number of nodes, respectively. As the upper limit of cost constraints, , and It is determined by the project deployment requirements, the constraints of the construction party, or the constraints of the operator.

[0082] The optimization objective of the 0-1 location optimization model is:

[0083] in, To optimize the objective function value; The set of key service units includes one or more of the following: takeoff and landing service units, hotspot service units, and risky service units. , , , and These are the weighting coefficients corresponding to deployment adaptability, primary / backup redundancy capability, deployment cost penalty, operation and maintenance difficulty penalty, and node quantity penalty, respectively. The first item represents the deployment adaptability of the selected node, the second item represents the primary / backup redundancy capability of the key service unit, the third item represents the deployment cost penalty, the fourth item represents the operation and maintenance difficulty penalty, and the fifth item represents the node quantity penalty.

[0084] The primary and backup service nodes are determined as follows: For route service units, within the selected candidate deployment carrier set Select to satisfy and The largest edge computing node serves as the primary service node; For takeoff and landing service units, hotspot service units, or high-risk service units, within the selected candidate deployment carrier set Enumerate all that satisfy Candidate deployment carriers And calculate the pairwise scores:

[0085] Select the candidate deployment pair with the highest pair score as the primary and backup candidate pairs; among the primary and backup candidate pairs, The node with the higher elevation is designated as the primary service node, and the other node is designated as the backup service node.

[0086] The output edge computing node deployment scheme includes at least the node installation location, node type, coverage area, corresponding low-altitude service unit, and primary / backup relationship, and may further include one or more of the following: installation height, installation carrier, power supply conditions, network access conditions, operation and maintenance conditions, deployment priority, and deployment cost estimate; wherein, the node type is determined based on the coverage contribution of the selected edge computing node to the route service unit, take-off and landing service unit, hotspot service unit, and risk service unit; if a node is assigned as a backup service node for any low-altitude service unit, then the node is also marked as a backup service node.

[0087] This application also includes a multi-constraint deployment system for edge computing nodes in low-altitude flight scenarios, comprising: The regional basic data acquisition module is used to acquire regional basic data of the target low-altitude flight area. The regional basic data includes at least low-altitude routes, take-off and landing points, and flight altitude range, and may further include one or more of the following: helipad, UAV airport, flight mission area, flight density, flight hotspot area, risk area, building distribution, obstacle distribution, communication infrastructure distribution, power supply conditions, network backhaul conditions, operation and maintenance conditions, construction costs, management boundaries, and prohibited and restricted areas. The low-altitude service unit division module is used to divide low-altitude service units according to low-altitude routes, take-off and landing points, aprons, flight hotspots and risk areas, and to configure service weight, minimum number of coverage nodes, maximum allowable service distance, maximum allowable altitude deviation, maximum allowable obstruction impact and minimum coverage evaluation threshold for low-altitude service units, and to further configure the minimum primary and backup coverage overlap ratio for key low-altitude service units. The candidate deployment carrier determination module is used to determine candidate deployment carriers and calculate the feasibility index of candidate deployment carriers based on installation space, structural load-bearing capacity, power supply capacity, network backhaul capacity, security protection, operation and maintenance accessibility and prohibition sign. The coverage relationship construction module is used to calculate the distance adaptation score, height matching score, occlusion correction score, infrastructure score and comprehensive coverage evaluation value of the candidate deployment carrier to the low-altitude service unit, and generate a coverage relationship matrix based on the comprehensive coverage evaluation value; The primary and backup overlap calculation module is used to discretize the low-altitude service unit into sampling points, calculate the coverage overlap ratio between candidate deployment carriers for the same low-altitude service unit, and generate a primary and backup overlap feasibility identifier. The adaptability evaluation module is used to calculate the node deployment adaptability of candidate deployment carriers based on coverage contribution, distance adaptability, height matching, occlusion correction, infrastructure conditions, cost, operation and maintenance difficulty, and redundancy coordination capability. The feasibility pre-check module is used to check whether the low-altitude service units meet the minimum coverage requirements before optimization, and to check whether the take-off and landing service units, hot spot service units or risk service units can form a primary and backup overlapping coverage relationship. The multi-constraint location optimization module is used to establish and solve a 0-1 location optimization model that includes candidate deployment carrier selection variables and primary / backup pair variables to determine the deployment location of edge computing nodes. The primary / backup relationship generation module is used to determine the primary and backup service nodes among the selected edge computing nodes and to verify the coverage overlap relationship between the primary and backup service nodes. The deployment scheme output module is used to output the edge computing node deployment scheme after the coverage, primary / backup, cost, number of nodes and feasibility verification are passed.

[0088] The following is a specific implementation example illustrating the process of the method described in this application: Regional basic data acquisition: Acquire regional basic data within the target low-altitude flight area, including low-altitude flight paths, take-off and landing points, aprons, UAV airports, flight mission areas, flight altitude ranges, flight density, building distribution, obstacle distribution, communication infrastructure distribution, power supply conditions, network backhaul conditions, operation and maintenance conditions, construction costs, management boundaries, and one or more of the prohibited and restricted areas.

[0089] In one implementation, regional basic data can be acquired and integrated from one or more of the following data sources: (1) low-altitude flight route planning data, flight mission database, or flight plan data; (2) historical flight trajectory data or simulated flight data; (3) geographic information data, including administrative boundaries, terrain, and elevation data; (4) building and obstacle distribution data; (5) communication infrastructure ledger, including communication base stations, building computer rooms, park computer rooms, and pole information; (6) power supply resource ledger and network backhaul resource ledger; (7) operation and maintenance management data and inspection station ledger; (8) airspace control boundaries, prohibited and restricted areas, and mission management boundaries issued by airspace management departments. Before use, regional basic data can be registered, cleaned, and rasterized according to a unified spatial reference system to facilitate subsequent low-altitude service unit division and feasibility screening of candidate deployment carriers.

[0090] Low-altitude service unit (LAU) segmentation: Based on low-altitude routes, takeoff and landing points, flight hotspots, and risk areas, the target low-altitude flight area is divided into multiple LAU service units. LAU service units include route service units, takeoff and landing service units, hotspot service units, and risk service units. The minimum number of coverage nodes for a route service unit can be set to 1; the minimum number of coverage nodes for takeoff and landing service units, hotspot service units, and risk service units can be set to 2 or higher.

[0091] In one implementation, the low-altitude service unit allocation includes the following processing flow: (1) Division of route service units: A spatial buffer zone is generated based on the centerline of the low-altitude route, and the spatial buffer zone is segmented according to the preset segment length or key nodes on the route to obtain route service units; wherein, the route width and segment length of the spatial buffer zone can be determined according to airspace management regulations, mission safety interval requirements or project deployment configuration.

[0092] (2) Division of take-off and landing service units: Taking the spatial location of the take-off and landing point, apron or UAV airport as the center, a circular or polygonal spatial buffer zone is generated as the take-off and landing service unit; wherein, the buffer radius or buffer shape of the spatial buffer zone can be determined according to the type of take-off and landing facility, take-off and landing track envelope or safe operation requirements.

[0093] (3) Division of hotspot service units: Map flight mission data, flight plan data or historical flight trajectory data to spatial grids, count the flight density within a unit grid, and aggregate continuous grids with flight density greater than the preset hotspot density threshold into a hotspot service unit; wherein, the scale of the spatial grid and the hotspot density threshold can be determined according to the flight traffic distribution characteristics or project deployment requirements.

[0094] (4) Division of risk service units: risk service units are delineated based on the distribution of obstacles, the density of tall buildings, sensitive areas of airspace management, or flight risk assessment results; among them, the risk judgment threshold or risk assessment method can be determined according to airspace management regulations or project deployment requirements.

[0095] Service weights can be configured separately for each type of low-altitude service unit. Minimum number of covered nodes Maximum allowed service distance Maximum permissible height deviation Maximum permissible shading impact and minimum coverage evaluation threshold For key low-altitude service units such as takeoff and landing service units, hotspot service units, and high-risk service units, further configure the minimum primary and backup coverage overlap ratio. The above configuration can be determined based on the flight mission importance, operational safety level, and redundancy requirements of the corresponding low-altitude service units, ensuring that the minimum number of coverage nodes and the minimum primary / backup coverage overlap ratio for takeoff and landing service units, hotspot service units, and high-risk service units are not lower than the corresponding values ​​for en-route service units. The configuration can generate project-level or industry-level low-altitude service unit parameter configuration tables, which can be adjusted according to project deployment requirements and actual operational data.

[0096] In one embodiment, the low-altitude service unit partitioning process and parameters can adopt the following exemplary values: (i) The width of the airway service unit is 50m on each side of the centerline of the low-altitude route (applicable to general low-altitude logistics routes, and can be adjusted between 30m and 100m in controlled airspace or special mission scenarios), and the segment length is 500m (which can be appropriately shortened when the curvature of the airway is large or there are key nodes on the airway). (ii) The buffer radius of the take-off and landing service unit is 200m (applicable to aprons and small UAV airports; for large take-off and landing facilities, it can be 300m to 500m). (iii) The spatial grid of the hotspot service unit is 50m×50m, and the hotspot density threshold is twice the average flight density of the target low-altitude flight area. Continuous grids with a density greater than the threshold are aggregated into a hotspot service unit. (iv) Risk service units are defined based on obstacle density, airspace management sensitivity level, or flight risk assessment results. For example, areas where the proportion of obstacles in a unit grid exceeds a preset obstacle density threshold, or continuous areas belonging to no-fly / restricted-fly management boundaries, are designated as risk service units. The obstacle density threshold or risk assessment method can be determined according to airspace management regulations or project deployment requirements.

[0097] For the aforementioned low-altitude service units, parameters , , , , , , The exemplary values ​​shown in Table 1 can be referred to respectively: Table 1. Example Value Table

[0098] The above parameter values ​​are only one embodiment and do not constitute a limitation on the scope of protection of this application. They can be adjusted according to the flight scenario, flight mission type, operational safety level and project deployment requirements.

[0099] Candidate deployment carriers are determined within or near the low-altitude service unit. Candidate deployment carriers include one or more of the following: buildings, communication base stations, helipads, drone airports, poles, campus data centers, road facilities, inspection stations, and other fixed facilities with installation, power supply, and network access capabilities.

[0100] Feasibility indicator calculation: for each candidate deployment carrier Feasibility indicators are calculated based on installation space, load-bearing capacity, power supply capacity, network backhaul capacity, security protection, maintenance accessibility, and prohibition signs. .when At that time, the candidate deployment carrier will not participate in the subsequent site selection.

[0101] The minimum installation space threshold used in the above feasibility assessment calculation Minimum structural bearing capacity threshold Minimum power supply capacity threshold Minimum network backhaul capacity threshold Minimum security protection score threshold and minimum operational accessibility score threshold The threshold can be determined according to one or more of the following methods: (1) Based on the equipment model requirements of the edge computing node to be deployed, such as the minimum installation space required for the equipment cabinet or rack, the minimum overall load requirement, the minimum power supply requirement, and the minimum network backhaul bandwidth requirement; (2) Based on the installation specifications, structural safety requirements, or relevant industry standards corresponding to the carrier type; (3) Based on the security protection level requirements and operation and maintenance management level requirements, including the minimum security protection score and the minimum operation and maintenance accessibility score; (4) Based on the deployment strategy document, construction constraints, or operation constraints of the specific project. In different application scenarios, each threshold can be configured with different values ​​and form a project-level or industry-level feasibility threshold configuration table, so that the feasibility identification calculation has repeatability and traceability.

[0102] In one embodiment, for a small, integrated edge computing node designed for low-altitude flight scenarios, the following exemplary values ​​can be used as a reference: Take 0.5m² as the installation space required for a small equipment chassis; Take 50kg as the minimum structural load-bearing capacity of the entire equipment and auxiliary accessories; Take 500W, which corresponds to the minimum power required for stable operation of edge computing devices; Take 100Mbps as the minimum effective bandwidth required for a single node to transmit low-altitude flight service data back to the backbone network; A score of 0.6 (out of 0 to 1) corresponds to the lowest safety protection level of the carrier. A score of 0.6 (out of 0 to 1) corresponds to the lowest operational accessibility level for the carrier. The above thresholds can be adjusted based on equipment model, carrier type, and project deployment requirements, and do not constitute a limitation on the scope of protection of this invention.

[0103] Coverage relationship matrix generation: Calculate the distance adaptation score, height matching score, occlusion correction score, and infrastructure score of candidate deployment carriers to low-altitude service units, and calculate the comprehensive coverage evaluation value based on these scores. When candidate deployment carriers meet the requirements for distance, altitude, obstruction, infrastructure, and minimum coverage evaluation thresholds, the coverage relationship is determined. Otherwise .

[0104] The above comprehensive coverage evaluation value The weighting coefficients involved , , and The determination method includes one or more of the following methods: (1) equal weight initialization method, that is, let And adjust according to the effect data after trial operation; (2) Preset the weight distribution method according to the type of low-altitude service unit, for example, increase the weight corresponding to the distance adaptation score for the route service unit, increase the weight corresponding to the occlusion correction score for the risk service unit, and increase the weight corresponding to the infrastructure score for the scenario that is strongly constrained by ground return and power supply resources; (3) The method of scoring and normalizing by domain experts according to the low-altitude operation requirements; (4) The method of solving the normalized weight after constructing the judgment matrix by using the analytic hierarchy process or similar methods; (5) The method of calibrating the weight according to historical operation data or simulation data. Regardless of the method used, the weight coefficient should meet the requirements. The weighting coefficients and their determination methods can be used as part of the project deployment parameters and configured separately for different low-altitude flight scenarios, and do not constitute a limitation on the scope of protection of this invention.

[0105] In one embodiment, the weighting coefficient , , and The following are some examples of possible values: Basic configuration values , , , This makes the distance adaptation score slightly higher than the other three items; for the route service unit, it can... Adjusted upwards to 0.35-0.40. and Adjustments can be made appropriately to highlight distance matching capabilities along the flight route; for takeoff and landing service units, adjustments can be made... The value was adjusted upwards to 0.30-0.35 to highlight its ability to match takeoff and landing altitudes; for high-traffic service units, the basic configuration can be maintained or adjusted. Adjustments can be made to accommodate intensive flight operations; for high-risk service units, adjustments can be made... The value was adjusted upwards to 0.30–0.35 to highlight its ability to correct for obstacle occlusion. These values ​​maintain... Under the premise of [previous conditions], adjustments can be made according to the deployment needs of the project.

[0106] Similarly, infrastructure score The weighting coefficients used in , , , , and Can be followed with The coefficients can be determined using one or more of the following methods: equal-weight initialization, pre-defined weight distribution based on the deployment scenario, expert score normalization, analytic hierarchy process (AHP), or historical / simulation data calibration. In one embodiment, the following can be used: As a basic configuration for equal-weight initialization; for carrier types with larger equipment scale or higher load-bearing capacity requirements, the weighting can be appropriately increased. , This refers to the weights corresponding to installation space and structural load-bearing capacity; for deployment scenarios that rely on external power supply and network backhaul, the weights can be appropriately increased. , The weights corresponding to power supply capacity and network backhaul capacity; for scenarios with limited operation and maintenance and security resources, the weights can be appropriately increased. , The corresponding weights. All weight coefficients should satisfy... The method of determining the weighting coefficients does not constitute a limitation on the scope of protection of this invention.

[0107] Primary / backup coverage overlap calculation: The low-altitude service unit is discretized into several sampling points, and the coverage overlap ratio is calculated based on the coverage of the sampling points by the candidate deployment carriers. If both candidate deployment carriers can cover the same low-altitude service unit, and the coverage overlap ratio meets the minimum primary / backup coverage overlap ratio requirement, then it is determined that the two have a primary / backup overlapping coverage relationship.

[0108] In one implementation, the sampling point coverage determination function Simultaneously considering three conditions: distance, height deviation, and line-of-sight obstruction: if and only if the candidate deployment carrier To the Distance of each sampling point Not exceeding the maximum allowed service distance The installation height of the candidate deployment carrier and the first Height deviation between sampling points Not exceeding the maximum permissible height deviation And candidate deployment carriers to the first If the line of sight to a sampling point is unobstructed, then the candidate deployment carrier is determined to cover that sampling point. ;otherwise .

[0109] Line-of-sight occlusion determination is based on the coverage relationship at the service unit level. A consistent line-of-sight model. Specifically, line-of-sight occlusion determination may include the following processing: generating a 3D geometric scene of the target low-altitude flight area based on building distribution, obstacle distribution, and terrain elevation data; and using candidate deployment carriers... Spatial position towards the first Line segments are emitted from the spatial locations of each sampling point, representing the line propagation path. It is determined whether each line segment intersects with any building, obstacle, or terrain in the 3D geometric scene. If no intersection occurs, the line is considered unobstructed. The line propagation function is then defined. If there is an intersection, then the line of sight is considered obstructed. The sampling point discretization method may include sampling at equal intervals along the centerline of the route service unit, or uniform sampling of takeoff and landing service units and hotspot service units within a buffer or grid. The total number of sampling points for each low-altitude service unit is [not specified]. It can be set according to the service unit scale and coverage determination accuracy requirements.

[0110] Node deployment adaptability evaluation: The node deployment adaptability of candidate deployment carriers is calculated based on coverage contribution, distance adaptability, height matching, occlusion correction, infrastructure conditions, cost, operation and maintenance difficulty, and redundancy coordination capability. .

[0111] Node deployment adaptability The weighting coefficients used in , , , , , , and It can be determined according to one or more of the following methods: (1) Equal weight initialization method, that is, let each All And adjust according to the effect data after trial operation; (2) Preset weight distribution method according to deployment scenario, for example, the coverage priority scenario can be appropriately increased. , This refers to the weighting of coverage contribution and infrastructure score; in scenarios prioritizing primary and backup redundancy, this weighting can be appropriately increased. The weight corresponding to the redundancy collaboration score can be appropriately increased in cost and operation and maintenance priority scenarios. , That is, the weights corresponding to the cost score and the operation and maintenance score; (3) the domain experts score and normalize the scores according to the requirements of low-altitude operation; (4) the normalized weights are solved after constructing the judgment matrix using the analytic hierarchy process or a similar method; (5) the weights are calibrated according to historical operation data or simulation data. In one embodiment, the following exemplary values ​​can be used as reference: , , , , , , , This allows the weights of coverage contribution, infrastructure, and redundancy coordination to be slightly higher than the other items. Regardless of the method used, the weighting coefficients should meet the following requirements. The method of determining the weighting coefficients does not constitute a limitation on the scope of protection of this invention.

[0112] Node deployment adaptability Cost ceiling involved in the calculation and the upper limit of operation and maintenance difficulty It can be determined according to one or more of the following methods: (1) Take the candidate deployment carrier set Deployment costs of various candidate deployment carriers in China The maximum value is used as The operational and maintenance difficulty of each candidate deployment platform The maximum value is used as Thus and Normalization to (2) The overall cost ceiling and the operation and maintenance difficulty ceiling shall be determined by the project deployment requirements, the constraints of the construction party or the constraints of the operator, and determined accordingly. and The method of determination does not constitute a limitation on the scope of protection of this invention.

[0113] Feasibility Pre-Check: Before establishing the site selection optimization model, a coverage feasibility check is performed on all low-altitude service units; a primary / backup coverage overlap feasibility check is performed on takeoff and landing service units, hotspot service units, and high-risk service units. If the conditions are not met, the reasons for infeasibility are output, and erroneous deployment plans are not generated.

[0114] 0-1 Site Selection Optimization Model: After the feasibility pre-check is passed, a system is established that includes variables for selecting candidate deployment carriers. Paired variables with primary and backup A 0-1 site selection optimization model is proposed. This model includes at least coverage constraints, feasibility constraints, cost constraints, node quantity constraints, and primary / backup overlap constraints. The optimization objective of this model is to comprehensively improve node deployment adaptability and primary / backup redundancy capabilities in key areas, while reducing deployment costs, operational complexity, and the number of nodes.

[0115] Weight coefficients in the optimization objective of the 0-1 location optimization model , , , and The configuration principles can be determined according to the deployment strategy, including: (1) normalizing the configuration proportionally based on the priority relationship between deployment adaptability, primary and backup redundancy capabilities, deployment costs, operation and maintenance difficulty, and the number of nodes; (2) when the deployment scenario is mainly for the protection of key low-altitude service units, the configuration can be appropriately increased. That is, the weight corresponding to the primary and backup redundancy capabilities; (3) When the deployment scenario is sensitive to construction investment, the weight can be appropriately increased. and That is, the weights corresponding to deployment cost penalties and node quantity penalties; (4) When the deployment scenario is sensitive to operation and maintenance resources, the weights can be appropriately increased. This refers to the weight corresponding to the penalty for operational difficulty. In one embodiment, the penalties can be applied proportionally to each... Initialization is performed, followed by iterative adjustments based on trial calculations or project deployment strategies. The method for determining the weighting coefficients does not constitute a limitation on the scope of protection of this application.

[0116] In one embodiment, the weighting coefficient , , , and The following are some examples of possible values: Basic configuration values , , , , This prioritizes deployment adaptability and primary / backup redundancy over cost, operational complexity, and node quantity penalties; in scenarios where priority is given to ensuring key low-altitude service units, it can... Adjusted upwards to 0.30-0.35; in investment-sensitive scenarios, it can be... , The sum is adjusted to around 0.40; in scenarios where operational resources are strained, it can be... Adjusted upwards to 0.20-0.25.

[0117] The solution process of the 0-1 location optimization model includes: (1) using feasibility indicators Coverage relation matrix Primary and backup overlap feasibility identifier Cost of candidate deployment carriers Operation and maintenance difficulty Upper and lower limits of the number of nodes and Cost constraint upper limit and key low-altitude service unit collection As input, where , and (1) The deployment requirements of the project, the constraints of the construction party or the constraints of the operator can be determined; (2) Construct an objective function that includes deployment adaptability, primary and backup redundancy, deployment cost penalty, operation and maintenance difficulty penalty and node number penalty; (3) Introduce coverage constraints, feasibility constraints, cost constraints, node number constraints and primary and backup overlap constraints; (4) Solve the 0-1 location optimization model through an integer programming solver, a mixed integer programming solver or a heuristic solving algorithm to obtain the candidate deployment carrier selection variables. Paired variables with primary and backup (5) Output the set of selected candidate deployment carriers. This serves as the input for subsequent primary-secondary relationship generation and final verification. Optionally, the solution algorithm may employ one or more of the following: branch and bound algorithm, cutting plane algorithm, mixed integer programming solver, greedy repair algorithm, genetic algorithm, or other heuristic search algorithm. This application does not limit the specific solution algorithm or specific solution tool.

[0118] Primary / backup relationship generation: Solving for the set of selected candidate deployment carriers Afterwards, The system generates primary and backup service nodes for low-altitude service units. For key service units, the backup service node must have a primary-backup overlapping coverage relationship with the corresponding primary service node.

[0119] Final verification and deployment plan output: The deployment plan is verified for coverage, primary / backup, cost, number of nodes, and feasibility. If the verification passes, the edge computing node deployment plan is output, including node installation location, node type, coverage area, corresponding low-altitude service unit, primary / backup relationship, installation height, installation carrier, power supply conditions, network access conditions, operation and maintenance conditions, deployment priority, and deployment cost estimate.

[0120] In one implementation, the node type may include one or more of route nodes, take-off and landing nodes, hotspot nodes, and risk nodes, determined by the type of low-altitude service unit that the node primarily serves; if the same edge computing node provides coverage for multiple types of low-altitude service units simultaneously, its node type is determined according to the type that contributes the most to the coverage of each type of low-altitude service unit; if the edge computing node is simultaneously assigned as a backup service node for any key low-altitude service unit, in addition to the node type, the edge computing node is also marked as a backup service node.

[0121] In urban low-altitude logistics scenarios, the target area includes multiple low-altitude air routes, multiple take-off and landing points, several helipads, drone airports, logistics distribution points, and flight hotspots. First, the following data is obtained: low-altitude air routes, take-off and landing points, flight altitude range, flight density, building distribution, communication infrastructure distribution, power supply conditions, network backhaul conditions, operation and maintenance conditions, construction costs, and restricted / prohibited areas. Then, low-altitude air routes are divided into airway service units, the areas surrounding helipads and drone airports are divided into take-off and landing service units, logistics distribution points and areas where air routes intersect are divided into hotspot service units, and areas with dense high-rise buildings or management-sensitive areas are divided into risk service units. Subsequently, candidate deployment carriers are extracted, their feasibility is screened, a coverage relationship matrix and primary / backup overlap relationships are established, a 0-1 site selection optimization model is solved, and a node deployment scheme is output.

[0122] In low-altitude air patrol scenarios within industrial parks, the target area includes patrol routes, patrol stations, helipads, key facility areas, and risk areas. Service units are divided based on patrol routes, takeoff and landing service units based on helipads and patrol stations, hotspot service units based on key facility areas, and risk service units based on risk facility areas. Candidate deployment carriers include park data centers, power poles, buildings, road infrastructure, patrol stations, and helipad ancillary facilities. This invention's method can meet the primary / backup redundancy coverage requirements of key areas while reducing deployment costs and operational complexity.

[0123] Simulation Implementation Examples and Expected Results: To further illustrate the expected technical effects of this application, taking an urban low-altitude logistics scenario as an example, an implementation method that can be verified on a simulation platform is given. In this simulation implementation method, the target low-altitude flight area is approximately 5km × 5km, including 12 low-altitude routes, 5 UAV airports, 20 take-off and landing points, and approximately 150 candidate deployment carriers; the target low-altitude flight area is divided into 30 route service units, 5 take-off and landing service units, 4 hotspot service units, and 6 risk service units. The simulation implementation method sets up two sets of comparative schemes: Scheme A serves as a reference scheme, representing an edge node deployment method that does not distinguish between low-altitude service unit types, does not perform feasibility pre-screening of candidate deployment carriers, and does not establish primary and backup overlapping coverage modeling, such as directly selecting deployment locations based on the location of ground communication infrastructure; Scheme B follows the method of this invention to perform feasibility screening of candidate deployment carriers, generation of coverage relationship matrix, calculation of primary and backup overlapping coverage, evaluation of node deployment adaptability, feasibility pre-check, solution of 0-1 site selection optimization model, and final verification.

[0124] The evaluation indicators used for comparison include: low-altitude service unit coverage, primary / backup redundancy fulfillment rate of key low-altitude service units (takeoff and landing, hotspots, and risks), number of nodes, deployment cost estimation, and operation and maintenance difficulty score. Based on the technical principles of the method of this invention, under the aforementioned simulation implementation method, compared with scheme A, scheme B is expected to achieve the following directional technical effects: (1) As Scheme B differentiates the modeling of the four types of low-altitude service units and explicitly introduces coverage constraints in the 0-1 location optimization model, the coverage of all low-altitude service units is significantly improved, and the coverage of key low-altitude service units is improved even more. (2) Since Scheme B establishes primary and backup overlapping coverage modeling and explicitly introduces primary and backup overlapping constraints, the primary and backup redundancy satisfaction rate of key low-altitude service units is significantly improved compared with Scheme A; the reference method has low redundancy guarantee capability for key low-altitude service units such as take-off and landing, hot spots and risks because it does not establish primary and backup overlapping coverage modeling. (3) Under the condition that the coverage and redundancy satisfaction rate are both improved, the number of edge computing nodes required is comparable to or slightly increased with that of Scheme A, and the deployment cost estimate and operation and maintenance difficulty score can be kept within an acceptable range. (4) Before solving Scheme B, the conditions for no solution can be identified through feasibility pre-check. After solving, the deployment schemes that do not meet the constraints can be identified through final verification, thus avoiding the output of unfeasible schemes.

[0125] It should be understood that the simulation scale, comparison scheme, evaluation indicators, and directional effects described above are only used to exemplify the beneficial effects of the method of the present invention compared to existing deployment methods. Specific effect values ​​should be based on the actual simulation platform calculation results or actual deployment case data. The simulation implementation methods and directional effects do not constitute a limitation on the scope of protection of the present invention. Under different low-altitude flight scenarios, different sets of candidate deployment carriers, and different low-altitude service unit configurations, the specific effect values ​​of the method of the present invention may vary.

[0126] The embodiments described above merely illustrate specific implementation methods of this application, and while the descriptions are detailed and specific, they should not be construed as limiting the scope of protection of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the technical solution of this application, and these modifications and improvements all fall within the scope of protection of this application.

Claims

1. A multi-constraint deployment method for edge computing nodes in a low-altitude flight scenario, characterized in that, Includes the following steps: Acquire basic regional data for the target low-altitude flight area. The basic regional data should include at least the low-altitude flight path, take-off and landing points, flight altitude range, building distribution, obstacle distribution, power supply conditions, network backhaul conditions, construction costs, and prohibited and restricted areas. Based on the spatial distribution of low-altitude flight routes and take-off and landing points, the target low-altitude flight area is divided into multiple low-altitude service units. Each low-altitude service unit includes an en-route service unit and a take-off and landing service unit, and a minimum number of coverage nodes is configured for each low-altitude service unit. and maximum allowed service distance The minimum number of coverage nodes for take-off and landing service units. Greater than the minimum number of coverage nodes for an airway service unit; Candidate deployment carriers are identified within or near the low-altitude service unit, and the feasibility index of each candidate deployment carrier is calculated based on installation space, structural load-bearing capacity, power supply capacity, network backhaul capacity, and prohibition sign. Candidate deployment carriers with a feasibility index of 0 are eliminated. For each low-altitude service unit and each candidate deployment carrier, based on distance adaptation score High matching score And occlusion correction score Calculate the comprehensive coverage evaluation value ,in And based on this, a covering relation matrix is ​​generated; Establish a 0-1 integer programming model with candidate deployment carriers as decision variables. The 0-1 integer programming model should include at least coverage constraints. Feasibility constraints and cost constraints ,in Indicates an overlay relationship. Indicates feasibility. Indicates deployment cost; Solve the 0-1 integer programming model to determine the selected candidate deployment carriers as the deployment locations of edge computing nodes, and output the deployment scheme.

2. The multi-constraint deployment method for edge computing nodes in a low-altitude flight scenario according to claim 1, characterized in that, The low-altitude service unit also includes a hotspot service unit and a risk service unit; the route service unit generates a spatial buffer zone based on the centerline of the low-altitude route and divides it according to segment length; the take-off and landing service unit generates a spatial buffer zone centered on the location of the take-off and landing point or apron; the hotspot service unit aggregates and divides continuous grids with density exceeding a threshold according to the flight density grid; the risk service unit is defined according to the density of obstacles or airspace management sensitive areas; the minimum number of coverage nodes for the hotspot service unit and the risk service unit are specified. All of these are greater than the minimum number of coverage nodes for an airway service unit.

3. The multi-constraint deployment method for edge computing nodes in a low-altitude flight scenario according to claim 1, characterized in that, Distance adaptation score according to Confirmed, high match score according to Confirmed, occlusion correction score. according to Confirmed, among which For the amplitude limiting function, The distance between the candidate deployment vehicle and the low-altitude service unit. The altitude deviation between the installation height and the flight altitude range, For the maximum permissible height deviation, To mask the impact value.

4. The multi-constraint deployment method for edge computing nodes in a low-altitude flight scenario according to claim 3, characterized in that, The occlusion impact value The target low-altitude flight area is determined as follows: a three-dimensional geometric scene of the target area is generated based on the distribution of buildings and obstacles; a straight line segment is launched from the spatial position of the candidate deployment vehicle to the geometric center of the low-altitude service unit; if the straight line segment intersects with any building or obstacle, then... If they do not intersect, then Comprehensive coverage evaluation value The score also includes infrastructure points. , ,in , , ; in, Indicates the available installation space for the candidate deployment carrier. This represents the maximum available installation space across the current candidate carrier set. The structural bearing capacity of the candidate deployment carrier, This represents the maximum load-bearing capacity of all structures in the current candidate carrier set. The available power supply capacity for candidate deployment carriers, This represents the maximum available power supply capacity among all candidate carriers in the current set. For the available network backhaul capabilities of candidate deployment carriers, This represents the maximum value of the available network backhaul capacity in the current candidate carrier set.

5. The multi-constraint deployment method for edge computing nodes in a low-altitude flight scenario according to claim 1, characterized in that, The feasibility indicator according to: Confirmed, among which For indicator functions, Available installation space, For structural load-bearing capacity, For available power supply capacity, For available network backhaul capabilities, For safety protection rating, To score operational accessibility, This is a "No Installation" sign; , , , These are the corresponding feasibility thresholds.

6. The multi-constraint deployment method for edge computing nodes in a low-altitude flight scenario according to claim 1, characterized in that, It also includes calculating the node deployment adaptability of candidate deployment carriers. : in to These are the weighting coefficients, and their sum is 1; To contribute points to coverage, Low-altitude service unit Service weight, Indicates an overlay relationship; The overall score is based on distance. For a highly comprehensive score; The overall score is based on the degree of occlusion. Score for cost. To score for operations and maintenance, For deployment costs, Due to the difficulty of operation and maintenance, To cap deployment costs, This represents the upper limit of operational and maintenance difficulty. This is a limiting function; For redundant collaboration scoring, Primary and backup overlap feasibility identifier. As a low-altitude service unit, As a key low-altitude service unit cluster, This represents the number of candidate carriers other than the candidate deployment carrier itself.

7. The multi-constraint deployment method for edge computing nodes in a low-altitude flight scenario according to claim 6, characterized in that, The 0-1 integer programming model is solved with the following optimization objective: in to These are weighting coefficients, corresponding to the priority weights of deployment adaptability, primary / standby redundancy capability, deployment cost penalty, operation and maintenance difficulty penalty, and number of nodes, respectively. as candidate deployment carriers Node deployment adaptability, as candidate deployment carriers The choice of variables, To select a deployment carrier With candidate deployment carriers low-altitude service units The coverage overlap ratio, Primary and backup paired variables, To optimize the objective function value, the 0-1 integer programming model also includes a node number constraint. .

8. The multi-constraint deployment method for edge computing nodes in a low-altitude flight scenario according to claim 1, characterized in that, This also includes establishing a primary / backup overlapping coverage relationship: Low-altitude service unit Discretize into multiple sampling points and calculate candidate deployment carriers. and Coverage overlap ratio ,in, Indicates candidate deployment carrier Does it cover low-altitude service units? The One sampling point, as candidate deployment carriers Does it cover low-altitude service units? The One sampling point, If and only if To the The distance between sampling points does not exceed Height deviation not exceeding And the line of sight must be unobstructed; otherwise, the value is 0. This represents the total number of sampling points; when , and At that time, enable the primary and backup overlap to be feasible. ,in Minimum primary / backup coverage overlap ratio; 0-1 integer programming models also include primary and backup paired variables. In addition, there are primary / backup overlap constraints for takeoff and landing service units, hotspot service units, and risky service units: ,and , , , , as candidate deployment carriers The choice of variables.

9. A multi-constraint deployment method for edge computing nodes in a low-altitude flight scenario according to claim 8, characterized in that, Determining the primary and backup service nodes includes: For route service units, within the selected candidate deployment carrier set Select to satisfy and The largest edge computing node serves as the primary service node; For takeoff and landing service units, hotspot service units, and high-risk service units, within the selected candidate deployment carrier set... Enumerate all that satisfy Candidate deployment carriers Calculate pairwise scores Select the candidate deployment pair with the highest pair score as the primary / backup candidate pair. One node from Sun Yat-sen University was designated as the primary service node, and the other node was designated as the backup service node; among them... as candidate deployment carriers Node deployment adaptability, as candidate deployment carriers low-altitude service units The comprehensive coverage evaluation value.

10. A multi-constraint deployment system for edge computing nodes in a low-altitude flight scenario, characterized in that, The method for multi-constraint deployment of edge computing nodes in a low-altitude flight scenario as described in any one of claims 1-9 includes: The data acquisition module is used to acquire basic regional data of the target's low-altitude flight area; The service unit partitioning module is used to divide the target low-altitude flight area into multiple low-altitude service units based on the spatial distribution of low-altitude routes and take-off and landing points. Each low-altitude service unit includes at least an en-route service unit and a take-off and landing service unit, and a minimum number of coverage nodes is configured for each low-altitude service unit. and maximum allowed service distance Among them, the take-off and landing service unit Larger than the route service unit ; The candidate carrier screening module is used to determine candidate deployment carriers and calculate feasibility indicators based on installation space, structural load-bearing capacity, power supply capacity, network backhaul capacity and prohibition indicators, and eliminate infeasible carriers; Coverage relationship building module, used for distance-based adaptation score High matching score And occlusion correction score Calculate the comprehensive coverage evaluation value And generate a coverage matrix; The site selection optimization module is used to establish and solve a 0-1 integer programming model with candidate deployment carriers as decision variables. The model must include at least coverage constraints. Feasibility constraints and cost constraints ; The solution output module is used to output the installation location of the selected edge computing nodes and the low-altitude service units they cover.

Images