Emergency treatment method for aviation oil pipeline leakage accident in mountainous area

Abstract

The invention discloses a mountainous area aviation oil pipeline leakage accident emergency processing method, which comprises the following steps: establishing a unified multi-attribute two-dimensional network model and a mountainous area aviation oil leakage finite element model FEM, and carrying out risk matching and model updating based on data of the mountainous area aviation oil leakage finite element model FEM and a support vector machine SVM; planning an evacuation path based on a multi-attribute two-dimensional network model and an improved ant colony algorithm; based on the multi-attribute two-dimensional network model, a support vector machine (SVM) and an A-Star algorithm, planning an approach repair route; and finally, integrating and outputting the evacuation path and the first-aid repair path. According to the scheme, on the basis of image recognition and fusion path planning, two path planning algorithm processes aiming at different targets are provided by constructing the two-dimensional network model, tight combination and collaborative planning of an evacuation path and a repair path are achieved, and the accuracy of emergency processing of the aviation oil pipeline leakage accident in the mountainous area is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of emergency response and geographic information processing, and particularly relates to a mountainous aviation oil pipeline leakage accident emergency processing method. BACKGROUND

[0002] The aviation oil pipeline (aviation kerosene pipeline) is an important energy supply line, which often needs to pass through the mountainous area with large terrain undulation, complex geological conditions and densely distributed environmental sensitive points (such as water sources and population gathering areas). Once pipeline leakage occurs in such an area, the highly volatile and flammable aviation oil will spread rapidly, posing a serious threat to the safety of downstream personnel and the ecological environment. How to guide the threatened personnel downstream and around the leakage point to quickly and safely evacuate to the designated safe area, and how to guide the professional repair team to quickly and efficiently arrive at the leakage source from the safe gathering point are two core problems that must be solved in emergency response.

[0003] Most of the prior art uses a single path planning algorithm to solve the emergency path problem. These technologies are difficult to quickly predict the spread of the leakage; secondly, they usually use a unified cost function (such as "shortest distance"), which cannot meet the needs of the two completely different and sometimes contradictory goals of "evacuation" and "repair"; finally, the existing solutions often regard evacuation and repair as two independent processes, lack a unified data model, and result in a lack of overall and coherence in the emergency plan. In particular, how to combine the predicted leakage model with the path planning algorithm in depth and automation, and based on this unified data model, derive the optimal evacuation and repair routes, is a technical problem that needs to be solved in the emergency processing of mountainous aviation oil pipeline leakage accidents. SUMMARY

[0004] In view of the above technical problems, the present application provides a mountainous aviation oil pipeline leakage accident emergency processing method based on image recognition and fusion path planning.

[0005] The present application is implemented by using the following technical solutions: The mountainous aviation oil pipeline leakage accident emergency processing method comprises the following steps: Step S1: constructing a unified multi-attribute two-dimensional network model and a mountainous aviation oil leakage finite element model FEM; Step S2: risk matching and model updating based on the mountainous aviation oil leakage finite element model FEM data and a support vector machine SVM; Step S3: evacuation path planning based on the multi-attribute two-dimensional network model and an improved ant colony algorithm; Step S4: approach repair route planning based on the multi-attribute two-dimensional network model, a support vector machine SVM and an A-Star algorithm ​​Step S5: Determine the evacuation routes Repair route Perform integrated output.

[0006] Specifically, the construction of the multi-attribute two-dimensional network model in step S1 includes: Step S11: Obtain detailed geographic information system data within the emergency response area and combine it with data from the high-precision digital elevation model (DEM); Step S12: Divide the entire emergency response area into A grid cell array, where the center of each grid cell is abstracted as a node. There are edges between adjacent nodes. Together, they form a two-dimensional network model diagram. : ; in, For nodes gather, For the edge gather; Step S13: Draw the two-dimensional network model diagram Each node in is Assign a multidimensional attribute vector, where For row index, For column indexing; the multidimensional attribute vector includes: nodes obtained from high-precision digital elevation model (DEM) data. elevation of The nodes are calculated using the elevation difference between adjacent nodes. slope at the location Nodes with values ​​{0,1} Static obstacle coefficient at the location Where 1 indicates that the node is a permanent, impassable obstacle, and 0 indicates that it is passable; nodes are assigned values ​​based on GIS data. Environmental sensitivity index The value ranges from [0, 100], with higher values ​​indicating higher environmental sensitivity; the nodes have discrete integer values. Highway grade ;node The leakage pollution index at the location .

[0007] Specifically, the construction of the finite element model (FEM) of the aviation fuel leak in the mountainous area in step S1 includes: For all potential leak points along the pipeline, finite element simulation is carried out using the geographic information system data, high-precision digital elevation model (DEM), and multiple preset leak scenarios to obtain a distribution map of leaked oil. The pre-simulation results are stored in a database that can be queried quickly, and each result is associated with the corresponding leakage parameter.

[0008] Specifically, step S2 includes: Step S21: Obtain the actual location of the leak point Based on the actual leakage type, query the FEM simulation results of the best matching mountain aviation fuel leakage finite element model in the FEM data of the mountain aviation fuel leakage finite element model; Step S22: Query the leakage pollution index As features, they are input into a pre-trained SVM classifier; Step S23: The SVM classifier classifies the FEM values ​​of the mountain aviation fuel leak finite element model based on the path planning risk level, and outputs the node... Risk classification, updated to the two-dimensional network model diagram The attributes include real-time updates of predicted leak and contamination data. .

[0009] Specifically, the SVM classifier and pre-training include: Leakage pollution index using extensive finite element model (FEM) simulation results. Using these as input features and labeling the degree of hazard, an SVM classifier is trained to convert continuous leakage and pollution indices into discrete risk levels. The SVM classifier outputs an integer value in the range [0, 100], with larger values ​​representing higher risk levels; and its output is assigned to the two-dimensional network model graph. Corresponding node .

[0010] Specifically, the improved ant colony algorithm in step S3 includes: Step S31: From node Move to adjacent node The evacuation cost is defined as , represented as: ; in, The Euclidean distance between the two nodes represents the path length. Adjacent nodes The slope; Adjacent nodes The static obstacle coefficient, if it is 1, means the cost is infinitely large, indicating that they cannot travel together; Adjacent nodes The risk index of leakage and pollution; Adjacent nodes Environmental sensitivity index; , , , , These are the weighting coefficients for each item; Step S32: When the ants Located at node When calculating the selection of the next node. probability for: ; In the formula, allowed k For ants From node The set of reachable and unvisited neighboring nodes; This is a pheromone importance factor, controlling the degree of influence of pheromones on path selection; The importance factor of the heuristic function controls the degree to which heuristic information influences path selection; pheromones In order to be in Time, from the node To the node The amount of pheromone accumulated along the path; heuristic function Guide the ants' local search, and ; Step S33: Simulate the evaporation and enhancement of pheromones in each iteration. Pheromones evaporation is represented as: ; Pheromones are enhanced as follows: ; in, The pheromone evaporation coefficient has a value of (0, 1). The number of ants; It is an ant In the path The amount of pheromones released and ; For pheromone constants, For ants k The total cost of the path found.

[0011] Specifically, the evacuation route The plan specifically includes: Set pheromones for all paths Initial value Set algorithm parameters and number of iterations ; Each ant starts from the evacuation starting point, selects the next node according to the state transition probability, until it reaches the evacuation endpoint or gets stuck in a dead end, and records the path and total cost of each ant. The pheromones on all paths are volatilized and enhanced, and the path with the lowest total cost is selected as the current optimal evacuation route. .

[0012] Specifically, step S4 involves the on-site emergency repair route. The plan specifically includes: Step S41: Based on the two-dimensional network model diagram and the predicted leakage contamination data updated in real time in step S23. A second SVM classifier is introduced to determine the passability of repair vehicles and equipment; Step S42: For each node in the network graph Extract a feature vector containing multi-dimensional information. And integrate predicted leakage pollution data eigenvectors Represented as: ; in, For slope, Road classification It is a static obstacle; Step S43: Process the feature vector using an SVM classifier Perform classification and output a binary result. And make a judgment, when When the value is 0, it indicates that the path is not feasible. When =1, it will start from node To the node The emergency repair cost is defined as: ; in, and These are the weights for distance and slope, respectively. Step S44: Apply the A-Star algorithm to the two-dimensional network model graph It automatically avoids all grids deemed impassable by the SVM classifier, thus obtaining the optimal access and repair path. .

[0013] Specifically, the access and emergency repair route The planning process also includes an evaluation function that guides the search, which is expressed as: ; in, The total estimate represents the total cost from the starting node through the nodes. Total estimated cost to the target node; The actual cost represents the cost from the starting node to the current node. The actual cost of emergency repairs; For heuristic valuation, it means starting from the current node. Estimated repair costs to the target node.

[0014] The beneficial effects of this invention are as follows: This invention provides an emergency response method for aviation fuel pipeline leak accidents in mountainous areas based on image recognition and fusion path planning. It constructs a unified, multi-attribute two-dimensional network model as the data foundation for the entire emergency response plan, integrating static geographic data and dynamic finite element simulation prediction data. Based on this, two path planning algorithms targeting different objectives are proposed, both sharing the same data model. This achieves a close integration and collaborative planning of evacuation and repair routes, ensuring the accuracy of emergency response to aviation fuel pipeline leak accidents in mountainous areas. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0016] Figure 1 This is a flowchart of an emergency response method for a leak in a mountainous aviation fuel pipeline, as described in this invention. Figure 2 This is a schematic diagram of the evacuation path based on the multi-attribute two-dimensional network model and the improved ant colony algorithm in this embodiment. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0018] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0019] The following is in conjunction with the appendix Figures 1-2The following describes some embodiments of the present invention in detail. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0020] This invention proposes an emergency response method for aviation fuel pipeline leaks in mountainous areas. In a preferred embodiment, the process is as follows: Figure 1 As shown, it includes: Step S1: Construct a unified multi-attribute two-dimensional network model and a finite element model (FEM) of aviation fuel leakage in mountainous areas; Step S2: Risk matching and model updating based on finite element model (FEM) data of aviation fuel leaks in mountainous areas and support vector machine (SVM); Step S3: Determine the evacuation path based on the multi-attribute two-dimensional network model and the improved ant colony algorithm. The plan; Step S4: Based on the multi-attribute two-dimensional network model, support vector machine (SVM), and A-Star algorithm, plan the emergency repair route. ; Step S5: Determine the evacuation routes Repair route Perform integrated output.

[0021] The specific steps are described in detail below with reference to specific embodiments: In this embodiment, constructing a unified multi-attribute two-dimensional network model includes the following steps: Detailed geographic information system (GIS) data of the emergency response area is acquired, covering road networks, river distribution, settlement locations, and various environmentally sensitive areas, and combined with high-precision digital elevation model (DEM) data. Next, the entire area is divided into a fine-grained... A grid cell array, where the center of each grid cell is abstracted as a node. There are edges between adjacent nodes. Together, they form a two-dimensional network diagram: ; In the formula, For nodes gather, For the edge gather.

[0022] For the image Each node in ,in For row index, For column indexes, assign a multidimensional attribute vector, including static attributes: :node The elevation at that location was obtained from DEM data; :node The slope at a given location is calculated using the elevation difference between that location and its adjacent nodes. :node The static obstacle coefficient at the node takes a value of {0,1}, where 1 indicates that the node is a permanent obstacle that is impassable, and 0 indicates that it is passable. :node The environmental sensitivity index is assigned a value based on the distribution of rivers, water sources, residential areas, nature reserves, etc. in GIS data. The range is usually [0,100], and the higher the value, the higher the environmental sensitivity. :node The road grade is represented by discrete integers: 0 indicates no road, 1 indicates a rural path (for walking only), 2 indicates a road that can be used by light vehicles (such as a rural road), and 3 indicates a road that can be used by heavy vehicles (such as a national highway or main road). :node The leakage pollution index at a location indicates the degree to which that node is affected by the leaked oil.

[0023] In this embodiment, the two-model network model provides a clear data-driven and objective visualization of the entire accident response process and communication coordination, helping the team make more accurate and rapid emergency response decisions and improving the ability and efficiency of responding to emergencies. The establishment of the two-model network model visually displays the entire emergency response process after an accident occurs, allowing for a clear and intuitive view of the actual division of labor in the actual emergency response coordination process. This facilitates more effective communication and collaboration among team members. Through simulation results, team members can gain a clearer understanding of the situation, improving the efficiency of communication and coordination. Furthermore, the two-model network model allows for the simulation of various scenarios and solutions on a computer, enabling the emergency response team to promptly identify potential problems in the response process. This, combined with daily monitoring and inspection, allows for timely updates and optimization of contingency plans, improving the ability to respond to different situations and ultimately helping to determine the optimal response process.

[0024] In this embodiment, the technical solution for constructing a finite element model of aviation fuel leakage in mountainous areas includes: For all potential leak points along the pipeline, finite element simulations were conducted using geographic information system data, digital elevation model data, and various preset leak scenarios (combinations of different leak volumes, leak durations, and downstream water flow velocities) to obtain a distribution map of the leaked oil, i.e., for each node. The predicted leakage pollution index is denoted as These pre-simulation results are stored in a database that can be queried quickly, with each result associated with parameters such as "leak point", "leak type", and "time".

[0025] In this embodiment, risk matching and model updating based on FEM data and SVM specifically include: Obtain the actual location of the leak. Based on the actual leakage type, query the FEM simulation result that best matches the data constructed in step one. Then, retrieve the found... As features, these are input into a pre-trained SVM classifier. The SVM is responsible for intelligently classifying complex FEM numerical values ​​into discrete "risk levels" usable for path planning, and outputting node-specific risk levels. The risk classification is updated in the two-dimensional network model. G Prediction is made based on the attributes.

[0026] The pre-trained SVM classifier is specifically as follows: Before an emergency occurs, a large number of FEM simulation results are used. The data is used as input features and labeled (e.g., concentration values ​​of 0-0.1 are "safe", 0.1-0.5 are "medium risk", and 0.5-1.0 are "high risk") to train a classification model that can convert continuous leakage pollution indices into discrete risk levels. Its output is an integer value in the range [0, 100], with larger values ​​representing higher risk levels. This output is then fed into a two-dimensional network model. G Corresponding node .

[0027] In this embodiment, the evacuation path planning based on a two-dimensional network model and an improved ant colony algorithm specifically includes: From node Move to adjacent node v j The evacuation cost is defined as This cost, taking into account multiple risk factors, is expressed as: ; In the formula, The Euclidean distance between the two nodes represents the path length. For nodes The slope; For nodes v j The static obstacle coefficient, if it is 1, means the cost is infinitely large, indicating that they cannot travel together; For nodes The risk index of leakage and pollution; For nodes Environmental sensitivity index; , , , , These are the weighting coefficients for each item. In evacuation route planning, leakage risk ( ) and environmentally sensitive areas, The weights of these areas are set relatively high to ensure that the path strongly avoids these dangerous areas.

[0028] In this embodiment, pheromones Indicates in t Time, from the node To the node v j The amount of pheromone accumulated along the path is typically initialized to a small constant. Heuristic function. The local search used to guide ants is usually defined as the reciprocal of the evacuation cost, representing the "visibility" or "attractiveness" of the path, and is expressed as: ; In the formula, It is a very small positive number, used to avoid division by zero.

[0029] When ants Located at node At that time, it selects the next section. probability The calculation is as follows: ; In the formula, For ants From node The set of reachable and unvisited neighboring nodes; This is a pheromone importance factor, controlling the degree of influence of pheromones on path selection; This is the importance factor of the heuristic function, which controls the degree of influence of heuristic information on path selection.

[0030] After each iteration, pheromones undergo two processes: evaporation and enhancement. Pheromones along all paths evaporate in a certain proportion, simulating pheromone evaporation, represented as: ; In the formula, The pheromone evaporation coefficient has a value of (0,1).

[0031] Ants that have completed their pathfinding will leave pheromones along their path to enhance their attractiveness, represented as: ; In the formula, The number of ants; It is an ant In the path The amount of pheromone released is commonly calculated as follows: ; In the formula, It is a pheromone constant; For ants The total cost of the path found.

[0032] When calculating the optimal evacuation route, firstly, set the pheromones for all routes. Initial value Set algorithm parameters α , β , p , Q , N and number of iterations n Subsequently, each ant starts from the evacuation starting point and selects the next node based on the state transition probability, until it reaches the evacuation endpoint or gets stuck in a dead end. The path and total cost of each ant are recorded. According to the above rules, pheromones are volatilized and enhanced on all paths. In all iterations, the path with the minimum total cost is selected as the current optimal evacuation path. ,like Figure 2 As shown.

[0033] In this embodiment, the on-site emergency repair route planning based on a two-dimensional network model and SVM and A-Star algorithms specifically includes: To ensure timely delivery while also protecting the safety of repair personnel and equipment, this step utilizes the two-dimensional network model constructed in step one. G And utilize the real-time updated predicted leakage pollution data from step two. A second SVM classifier is introduced to determine the "accessibility" of repair vehicles and equipment.

[0034] For each node in the network graph Extract a feature vector containing multi-dimensional information. Vectors encompass static geographic features, such as slope. Road grade Static obstacles It also incorporates the predicted leakage pollution index of the node obtained in step two. This is represented as: .

[0035] By drawing on a large amount of historical data, experience, and experimental results, a model is trained that can extract feature vectors. The SVM model is classified as either feasible or infeasible. This classifier classifies SVM models by... Perform classification and output a binary result. The symbol 0 represents impassable and 1 represents passable. Therefore, even if a highway has a high road classification... However, if the finite element method predicts that it will be covered by a high concentration of oil at the current moment, that is... Even with a high value, the classifier will still classify it as... =0, thus effectively avoiding the risks of explosion, poisoning or equipment damage faced by the emergency repair team.

[0036] Beneficiency assessment based on SVM classifier ,when When =0, it means the path is not feasible, and When =1, it will start from node To the node The emergency repair cost is defined as: ; In the formula, and These are the weights for distance and slope, respectively, and are usually set to higher values ​​to pursue speed.

[0037] The A-Star algorithm was then applied to the network graph, leveraging its efficient heuristic search capabilities to automatically avoid all grids deemed impassable by the SVM (including terrain obstacles, low-grade roads, and areas predicted as high-risk leaks), ultimately planning an optimal repair path. .

[0038] Throughout the process, the evaluation function guiding the search is: ; In the formula, For the total estimate, starting from the starting node and passing through the nodes... Total estimated cost to the target node; For actual costs, from the starting node to the current node The actual cost of emergency repairs; For heuristic valuation, starting from the current node The estimated repair cost to the target node is usually calculated using Euclidean distance as a heuristic function.

[0039] Finally, in this embodiment, the evacuation route is... Repair route The integrated output includes: Finally, the evacuation routes planned in step three will be implemented. and the emergency repair route planned in step four. The plans are integrated and displayed graphically. These planned routes can be visually presented on a map and quickly distributed to emergency response teams and threatened personnel on site, providing timely and accurate guidance for emergency rescue operations.

[0040] For the foregoing embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to this application.

[0041] The above embodiments describe the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Modifications and variations made by those skilled in the art without departing from the spirit and scope of the invention should be within the protection scope of the appended claims.

Claims

1. Emergency handling method for aviation fuel pipeline leaks in mountainous areas, characterized in that, Includes the following steps: Step S1: Construct a unified multi-attribute two-dimensional network model and a finite element model (FEM) of aviation fuel leakage in mountainous areas; Step S2: Risk matching and model updating based on finite element model (FEM) data of aviation fuel leaks in mountainous areas and support vector machine (SVM); Step S3: Determine the evacuation path based on the multi-attribute two-dimensional network model and the improved ant colony algorithm. The plan; Step S4: Based on the multi-attribute two-dimensional network model, support vector machine (SVM), and A-Star algorithm, plan the emergency repair route. ; Step S5: Determine the evacuation routes Repair route Perform integrated output.

2. The emergency response method for aviation fuel pipeline leaks in mountainous areas as described in claim 1, characterized in that, The construction of the multi-attribute two-dimensional network model in step S1 specifically includes: Step S11: Obtain detailed geographic information system data within the emergency response area and combine it with data from the high-precision digital elevation model (DEM); Step S12: Divide the entire emergency response area into A grid cell array, where the center of each grid cell is abstracted as a node. There are edges between adjacent nodes. Together, they form a two-dimensional network model diagram. : ； in, For nodes gather, For the edge gather; Step S13: Draw the two-dimensional network model diagram Each node in is Assign a multidimensional attribute vector, where For row index, For column indexing; the multidimensional attribute vector includes: nodes obtained from high-precision digital elevation model (DEM) data. elevation of The nodes are calculated using the elevation difference between adjacent nodes. slope at the location Nodes with values ​​{0,1} Static obstacle coefficient at the location Where 1 indicates that the node is a permanent, impassable obstacle, and 0 indicates that it is passable; nodes are assigned values ​​based on GIS data. Environmental sensitivity index The value ranges from [0, 100], with higher values ​​indicating higher environmental sensitivity; the nodes have discrete integer values. Highway grade ;node The leakage pollution index at the location .

3. The emergency response method for aviation fuel pipeline leaks in mountainous areas as described in claim 2, characterized in that, The construction of the finite element model (FEM) of the aviation fuel leak in the mountainous area in step S1 specifically includes: For all potential leak points along the pipeline, finite element simulation is carried out using the geographic information system data, high-precision digital elevation model (DEM), and multiple preset leak scenarios to obtain a distribution map of leaked oil. The pre-simulation results are stored in a database that can be queried quickly, and each result is associated with the corresponding leakage parameter.

4. The emergency response method for aviation fuel pipeline leaks in mountainous areas as described in claim 3, characterized in that, Step S2 specifically includes: Step S21: Obtain the actual location of the leak point Based on the actual leakage type, query the FEM simulation results of the best matching mountain aviation fuel leakage finite element model in the FEM data of the mountain aviation fuel leakage finite element model; Step S22: Query the leakage pollution index As features, they are input into a pre-trained SVM classifier; Step S23: The SVM classifier classifies the FEM values ​​of the mountain aviation fuel leak finite element model based on the path planning risk level, and outputs the node... Risk classification, updated to the two-dimensional network model diagram The attributes include real-time updates of predicted leak and contamination data. .

5. The emergency response method for aviation fuel pipeline leaks in mountainous areas as described in claim 4, characterized in that, The SVM classifier and pre-training include: Leakage pollution index using extensive finite element model (FEM) simulation results. Using these as input features and labeling the degree of hazard, an SVM classifier is trained to convert continuous leakage and pollution indices into discrete risk levels. The SVM classifier outputs an integer value in the range [0, 100], with larger values ​​representing higher risk levels; and its output is assigned to the two-dimensional network model graph. Corresponding node .

6. The emergency response method for aviation fuel pipeline leaks in mountainous areas as described in claim 5, characterized in that, The improved ant colony algorithm in step S3 specifically includes: Step S31: From node Move to adjacent node The evacuation cost is defined as , is represented as: ； in, The Euclidean distance between the two nodes represents the path length. Adjacent nodes The slope; Adjacent nodes The static obstacle coefficient, if it is 1, means the cost is infinitely large, indicating that they cannot travel together; Adjacent nodes The risk index of leakage and pollution; Adjacent nodes Environmental sensitivity index; , , , , These are the weighting coefficients for each item; Step S32: When the ants Located at node When calculating the selection of the next node. probability for: ； In the formula, allowed k For ants From node The set of reachable and unvisited neighboring nodes; This is a pheromone importance factor, controlling the degree of influence of pheromones on path selection; The importance factor of the heuristic function controls the degree to which heuristic information influences path selection; pheromones In order to be in Time, from the node To the node The amount of pheromone accumulated along the path; heuristic function Guide the ants' local search, and ; Step S33: Simulate the evaporation and enhancement of pheromones in each iteration. Pheromones evaporation is represented as: ； Pheromones are enhanced as follows: ； in, The pheromone evaporation coefficient has a value of (0, 1). The number of ants; It is an ant In the path The amount of pheromones released and ; For pheromone constants, For ants k The total cost of the path found.

7. The emergency response method for aviation fuel pipeline leaks in mountainous areas as described in claim 6, characterized in that, The evacuation route The plan specifically includes: Set pheromones for all paths Initial value Set algorithm parameters and number of iterations ; Each ant starts from the evacuation starting point, selects the next node according to the state transition probability, until it reaches the evacuation endpoint or gets stuck in a dead end, and records the path and total cost of each ant. The pheromones on all paths are volatilized and enhanced, and the path with the lowest total cost is selected as the current optimal evacuation route. .

8. The emergency response method for aviation fuel pipeline leaks in mountainous areas as described in claim 7, characterized in that, The emergency repair route in step S4 The plan specifically includes: Step S41: Based on the two-dimensional network model diagram and the predicted leakage contamination data updated in real time in step S23. A second SVM classifier is introduced to determine the passability of repair vehicles and equipment; Step S42: For each node in the network graph Extract a feature vector containing multi-dimensional information. And integrate predicted leakage pollution data eigenvectors Represented as: ； in, For slope, Road classification It is a static obstacle; Step S43: Process the feature vector using an SVM classifier Perform classification and output a binary result. And make a judgment, when When the value is 0, it indicates that the path is not feasible. When =1, it will start from node To the node The emergency repair cost is defined as: ； in, and These are the weights for distance and slope, respectively. Step S44: Apply the A-Star algorithm to the two-dimensional network model graph It automatically avoids all grids deemed impassable by the SVM classifier, thus obtaining the optimal access and repair path. .

9. The emergency response method for aviation fuel pipeline leaks in mountainous areas as described in claim 8, characterized in that, The emergency repair route The planning process also includes an evaluation function that guides the search, which is expressed as: ； in, The total estimate represents the total cost from the starting node through the nodes. Total estimated cost to the target node; The actual cost represents the cost from the starting node to the current node. The actual cost of emergency repairs; For heuristic valuation, it means starting from the current node. Estimated repair costs to the target node.

Images