A collaborative path planning method for multiple UAVs in complex environments

Abstract

This invention discloses a collaborative path planning method for multiple UAVs in complex environments, comprising the following steps: Step 1: Constructing a multi-priority task collaborative scheduling model and preparing UAV and task datasets for model adaptation; Step 2: Inputting the UAV and task parameters into the pre-constructed multi-priority task collaborative scheduling model, which then initially allocates tasks according to priority; Step 3: Real-time collection of the UAV's status parameters during task execution, including current position, remaining payload, remaining battery power, and distance to the target point, to determine the triggering conditions for material delivery, hovering, and return; Step 4: Completing material delivery, reconnaissance hovering, collaborative following, and battery replenishment through quantitative methods of payload update, hovering control, and charging scheduling; Step 5: Adjusting the subsequent task execution plan according to the task completion progress and UAV resource status based on collaborative scheduling principles, and outputting the final task completion result.

Description

Technical Field

[0001] This invention relates to a method for unmanned aerial vehicle (UAV) path planning, and more particularly to a method for collaborative path planning of multiple UAVs in complex environments. Background Technology

[0002] Multi-UAV collaborative path planning technology is increasingly being used in military reconnaissance, material delivery and other fields. Through multi-UAV collaborative operations, it can achieve large-scale target coverage and efficient mission execution, significantly improving operational efficiency in complex environments.

[0003] However, existing technologies still have many key problems in practical applications: In static multi-target coverage scenarios, it is difficult to optimize the total flight path while meeting the constraints of UAV communication distance and minimum safe distance, often leading to unbalanced task allocation or path redundancy; In dynamic environments, when facing sudden obstacles or new emergency targets, there is a lack of efficient real-time obstacle avoidance and task rescheduling mechanisms, which can easily prolong task completion time; In multi-priority task scenarios, due to insufficient coordination of UAV payload, endurance and charging scheduling, problems such as high-priority task delays, untimely UAV power warning responses or low resource utilization often occur.

[0004] Therefore, in response to the core technical bottlenecks in the aforementioned complex environments, developing a multi-UAV collaborative path planning system that takes into account path optimization, dynamic response, and resource coordination is of great significance for improving the stability and efficiency of multi-UAV swarm operations. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the present invention aims to provide a collaborative path planning method for multiple UAVs in complex environments. This method can achieve efficient coverage of static targets, real-time avoidance of dynamic obstacles, rapid response to emergency tasks, and collaborative resource scheduling for multiple priority tasks, while balancing task coverage, path efficiency, and communication stability in low-altitude complex environments.

[0006] To achieve the above objectives, this invention provides a method for cooperative path planning of multiple UAVs in complex environments, characterized by the following steps:

[0007] Step 1: Construct a multi-priority task collaborative scheduling model and prepare drone and task datasets for model adaptation;

[0008] Step 2: Input the UAV and mission parameters into a pre-built multi-priority mission collaborative scheduling model, and the model will initially allocate missions according to priority;

[0009] Step 3: Collect the status parameters of the UAV in real time when it is performing the mission, including the current position, remaining payload, remaining battery power and distance to the target point, and determine the trigger conditions for material delivery, hovering and return.

[0010] Step 4: Through quantitative methods of load updating, hovering control, and charging scheduling, complete material delivery, reconnaissance hovering, coordinated following, and power replenishment;

[0011] Step 5: Based on the task completion progress and the status of the UAV resources, adjust the subsequent task execution plan according to the principle of collaborative scheduling, and output the final task completion result.

[0012] Preferably, the multi-priority task collaborative scheduling model includes an optimized task allocation module, an optimized dynamic execution module, and an optimized resource coordination module.

[0013] The optimized task allocation module adopts a combination structure of bottleneck method, binary search method and Hungarian algorithm to perform three-level priority matching processing on multi-priority tasks. Priority tasks 2 and 3 will be transmitted once before matching. The task status information transmitted in the first transmission is named high-priority task status information, and the task status information transmitted in the second transmission is medium-priority task status information.

[0014] The optimized dynamic execution module performs a real-time status information transmission, named real-time execution status information; the optimized dynamic execution module includes 4 core functional sub-modules, 1 status monitoring module and constraint adaptation module.

[0015] Preferably, the task allocation initial stage of the multi-UAV collaborative path planning method based on complex environments introduces a progressive constraint adaptation method to optimize the collaboration of the initial path;

[0016] The optimized resource coordination module receives high-priority task status information, medium-priority task status information, and real-time execution status information. Simultaneously, it uses the resource scheduling sub-module in the optimized resource coordination module to fuse the high-priority task status information, medium-priority task status information, and real-time execution status information in the fusion layer. The fused scheduling scheme is then iteratively optimized to form a final path planning result that satisfies full task coverage, resource adaptation, and constraint compliance.

[0017] Preferably, the specific steps of the UAV collaborative scheduling method in step two include the following:

[0018] ① Determine the set of variables:

[0019] Maximum payload of a drone: C = {C1, C2...C} m};

[0020] At time t, the drone's current payload is: c = {c1, c2, ..., c} m};

[0021] Target point resource requirements: D = {d1, d2, ... d} n}, where d j Represents the j-th task point t j The demand for supplies;

[0022] At time t, the feasible set of drones is: {t 11 ,t 12 ...}{t 21 ,t 22 ...}...{t v1 ,t v2 ...};

[0023] Among them, {t v1 ,t v2 ...} indicates that the v-th drone satisfies c i ≦d j The target point is (1≦j≦n);

[0024] Unmanned aerial vehicles {u1,u2...u m The corresponding unique assignment is: {t} 1y1 ,t 2y2 ...t mym}, where t iyi Indicates drone u i The only selected target point, t iyi ∈ Feasible set {t i1 ,t i2 ...},t iyi It can be empty (i.e., a target cannot be assigned);

[0025] ② Optimization model:

[0026] Establish decision variables:

[0027] Based on the decision variables, the objective function is further obtained:

[0028] Constraints: ;

[0029] ③ Bottleneck Method Model Establishment:

[0030] Optimal allocation: {t 1x1 ,t 2x2 ...t mxm}

[0031] According to the optimal allocation binary search: Define the threshold θ, and find the minimum θ through binary search so that the following allocation scheme satisfies D ij ≤ 0, ∀ x ij = 1, where D ij is the geometric distance between the UAV u i and the target t j . The search and rescue process satisfies

[0032] The termination condition is: ;

[0033] ④. Hungarian method modeling:

[0034] For the binary search, we can obtain the maximum threshold θ, and solve the optimal allocation of UAVs according to the Hungarian algorithm: {t 1x1 , t 2x2 ... t mxm},

[0035] For the UAV u i at time t and the distance matrix with the feasible set {tu1, tu2... tu m} is:

[0036]

[0037] Among them, the targets that do not meet the load requirements are set to inf, and inf represents a very large number. D ij represents the geometric distance between the UAV u i and the target point t j .

[0038] Construct a bipartite graph to solve: G=(U∪T,E), and the edge set is E= {( i , j )| j ∈ F i ∧ D ij ≤ θ},

[0039] Find the maximum matching through the Hungarian algorithm M ⊆ E satisfies |M| = min(m,n),

[0040] where m represents the number of UAVs, n represents the number of target points. When m>n, n UAVs get the optimal unique allocation, and the remaining m - n UAVs are ∅; when m<n, all m UAVs get the optimal unique allocation;

[0041] ⑤. Rejection + retry strategy:

[0042] For subset {t j1 x j1 ,t j2 x j2 ,t j3 x j3 ...}∈{t 1x1 ,t 2x2 ...t mxm Given k drones, we use a "removal" strategy to make k drones {u j1 ,u j2 ,u j3 ...u jk Target point: { t j1 x j1 , t j2 x j2 ,..., t jk x jk}=∅,

[0043] Among them, t j1 x j1 Indicates drone u j1 target point

[0044] For {t 1x1 ,t 2x2 ...t mxm}\{t j1 x j1 ,t j2 x j2 ,t j3 x j3 ...}, we adopt a "retry" strategy, repeatedly solving the problem until a feasible allocation is found, such that |M| = min(mk) in the formula. min ,n),

[0045] Where, k min For the minimum number of k=0,1,...m that satisfy the condition,

[0046] The target allocation model is now fully established. Through this model, we can ultimately obtain the following allocation:

[0047] mk drones u={u1,u2...u m-k The unique optimal target point is: { t 1, t 2... t m-k},

[0048] k drones u={u m-k+1 ,u m-k+2 ,...,u m The empty set point of} is {t} m-k+1 ,...t m}=∅.

[0049] Preferably, the specific steps of the material delivery method in step four include the following:

[0050] ① Deployment Trigger: When the real-time position of the UAV d is close to the target point (p) k ,q k Satisfy distance constraints | p(t)-p |< E delivery ( E delivery When p = 20m, the deployment preparation process is triggered. d (t) represents the position of the drone d at time t, ε delivery This is a small tolerance value, indicating when the drone reaches the target point;

[0051] ② Payload Update: After the UAV d is deployed, the payload ld of UAV d is updated. t *)=max(ld( t ) - qk ,0), where l d (t) represents the current payload of UAV d. After the mission is completed, the target point is removed from the set T.

[0052] ③ Return-to-home trigger: If the current payload of all the aforementioned drones cannot meet the requirements of the next mission. (p k ,g k )∈ T d satisfy l d ( t ) < q k If this happens, all drones will enter return-to-home mode, and the target point will be set to b.

[0053] ④ Return-to-home motion: The speed of the UAV during the return-to-home motion is as follows: , where k attract The attraction coefficient, ε, is a minimum value to prevent division by zero;

[0054] ⑤ Resupply Completed: The conditions for completing drone resupply are | p d ( t )- b |< ε ( ε =20m), if resupply is complete, reset the load and exit the return-to-base state. l d ( t + ) = C d , T d ←Reassigned tasks.

[0055] Preferably, the specific steps of the reconnaissance hovering method in step four include the following:

[0056] ① Hovering determination: When the distance between the UAV and the target meets the following condition: | p d ( t ) - p target If | < 20 (hovering radius), activate the flyaround mode; otherwise, continue moving towards the target point.

[0057] ② Position update: After time Δt p d ( t +Δ t ) =p d ( t ) +v d ( t +Δ t ) · Δ t .

[0058] Preferably, the specific steps of the cooperative following method in step four include the following:

[0059] ① Selection: Select all empty target UAVs. i Select the nearest target drone. j , ensure min{|| u i - u j ||}, any u j ∈{u1,...,u m-k}, where u i ∈{u m-k+1 ,u m-k+2 ,...,u m}, u j ∈{u1,u2...u m-k};

[0060] ② Copy: Copy the most recent drone u j The target, the air target drone ui ,Right now t ixi = t jxj ;

[0061] ③ Clear: When target t jxj Complete the deployment of supplies and clear all drones in the entire region: { t 1, t 2... t m-k}{ t jxj}

[0062] Preferably, the specific steps of the power replenishment method in step four include the following:

[0063] ① Charging warning: When any drone meets the charging requirements... At that time, the drone will issue a charging warning; among them T safe For a safe buffer period (20-30 seconds), D i (t) represents the Euclidean distance of UAV i from the origin, T i remain (t) represents the remaining flight time of drone i at time t. When the above conditions are triggered, all drones will return to base.

[0064] ② Real-time monitoring and alarm: Each of the aforementioned drones periodically broadcasts its own status (x i ,y i ,T i remain Perform global most critical calculations. If T critical <T safe Execute the return-to-base instruction;

[0065] ③ Coordinated Return-to-Home Control: Return-to-Home Trigger: If the current payload of all the aforementioned UAVs cannot meet the requirements of the next mission... (p k ,g k )∈T d satisfy l d ( t ) < q k Then all drones enter return-to-home mode, with the target point set as b; Return-to-home motion: the speed of the drones during the return-to-home motion is as follows , where k attract The attraction coefficient, ε, is a minimum value to prevent division by zero;

[0066] ④ Latest Return Time Calculation: For each UAV i, the shortest return time is: t i return =D i / v max

[0067] To ensure safety, the following must be met: .

[0068] Preferably, the specific steps of the position update calculation method in step four are as follows:

[0069] ① Speed ​​of updates: v d ( t +Δ t ) = v d (t)+ γ ( t ),

[0070] The control function γ(t) is defined as follows:

[0071] ② Components of attraction:

[0072]

[0073] ③ Speed ​​constraint: .

[0074] Preferably, the drone resource status in step five includes payload and battery power.

[0075] The technical effects of this invention are as follows:

[0076] 1. This invention improves the multi-UAV collaborative path planning model, including an optimized task allocation module, dynamic execution module, and resource coordination module, which can adapt to more complex task scenarios, including multi-priority task scheduling, dynamic obstacle avoidance, and resource constraint adaptation. This enables the model to optimize path planning logic from multiple dimensions such as task priority, real-time status, and resource adaptation, thereby improving the efficiency and stability of collaborative path planning.

[0077] 2. This invention also provides a path planning strategy based on a hierarchical optimization framework, which, combined with a multi-UAV collaborative path planning model, can improve the adaptability to complex environments, enabling efficient coverage of static targets, real-time response to dynamic scenarios, and orderly execution of multi-priority tasks. This not only improves the task completion rate and resource utilization of multi-UAV collaborative operations, but also adapts to various complex low-altitude environments such as military reconnaissance and emergency rescue, enhancing the practical application value of multi-UAV swarm operations. Attached Figure Description

[0078] Figure 1 This is a map showing the generated drone and mission location.

[0079] Figure 2 The matching graph for the Hungarian algorithm;

[0080] Figure 3 This is a rendering of the return and resupply operation.

[0081] Figure 4 This is a hovering effect image;

[0082] Figure 5 This is an example of a device returning to base when the battery is low.

[0083] Figure 6 The image shows the effect of task recovery after charging.

[0084] Figure 7 Here is a flowchart of a multi-UAV collaborative path planning method for complex environments;

[0085] Figure 8 A diagram of the allocation model based on the bottleneck improvement method;

[0086] Figure 9 For the cooperative following model diagram;

[0087] Figure 10 Supplementing the model diagrams for deployment and return;

[0088] Figure 11 This is a hovering model diagram;

[0089] Figure 12 A model diagram for early warning of synchronized return-to-base charging of drone swarms. Detailed Implementation

[0090] To further illustrate the technical implementation and effects of this invention in multi-priority task and charging scheduling scenarios, the following detailed explanation of the specific implementation method, operation process and key parameters of the multi-UAV cooperative path planning strategy is provided in conjunction with simulation data, model steps and experimental results.

[0091] Reference Figure 1-12 As shown, the present invention provides a method for collaborative path planning of multiple UAVs in complex environments, comprising the following steps:

[0092] Step 1: Construct a multi-priority task collaborative scheduling model and prepare drone and task datasets for model adaptation;

[0093] The multi-priority task collaborative scheduling model includes an optimized task allocation module, an optimized dynamic execution module, and an optimized resource coordination module.

[0094] The optimized task allocation module adopts a combination structure of "bottleneck method + binary search method + Hungarian algorithm" to perform three-level priority matching processing on multi-priority tasks. Priority tasks 2 and 3 will be transmitted once before matching. The task status information transmitted in the first transmission is named high-priority task status information, and the task status information transmitted in the second transmission is medium-priority task status information.

[0095] The optimized dynamic execution module performs a real-time status information transmission, named real-time execution status information; the optimized dynamic execution module includes 4 core functional sub-modules, 1 status monitoring module and constraint adaptation module.

[0096] Step 2: Input the UAV and mission parameters into a pre-built multi-priority mission collaborative scheduling model, and the model will initially allocate missions according to priority;

[0097] The specific steps of the UAV collaborative scheduling method include the following:

[0098] ① Determine the set of variables:

[0099] Maximum payload of a drone: C = {C1, C2...C} m};

[0100] At time t, the drone's current payload is: c = {c1, c2, ..., c} m};

[0101] Target point resource requirements: D = {d1, d2, ... d} n}, where d j Represents the j-th task point t j The demand for supplies;

[0102] At time t, the feasible set of drones is: {t 11 ,t 12 ...}{t 21 ,t 22 ...}...{t v1 ,t v2 ...};

[0103] Among them, {t v1 ,t v2 ...} indicates that the v-th drone satisfies c i ≦d j The target point is (1≦j≦n);

[0104] Unmanned aerial vehicles {u1,u2...u m The corresponding unique assignment is: {t} 1y1 ,t 2y2 ...t mym}, where t iyi Indicates drone u i The only selected target point, t iyi ∈ Feasible set {t i1 ,t i2 ...},t iyi It can be empty (i.e., a target cannot be assigned);

[0105] ② Optimization model:

[0106] Establish decision variables:

[0107] Based on the decision variables, the objective function is further obtained:

[0108] Constraints: ;

[0109] ③ Bottleneck Method Model Establishment:

[0110] Optimal allocation: {t 1x1 ,t 2x2 ...t mxm}

[0111] Based on the optimal allocation binary search: Define a threshold θ, and use binary search to find the minimum θ such that the following allocation scheme is satisfied. D ij ≤ 0,∀ x ij =1, where D ij It is a drone. i With target t j The geometric distance satisfies the requirements of the search and rescue process.

[0112] The termination condition is: ;

[0113] ④ Hungarian modeling:

[0114] For the binary search, we can obtain the maximum threshold θ, and then use the Hungarian algorithm to solve for the optimal allocation of UAVs: {t 1x1 ,t 2x2 ...t mxm},

[0115] For the drone u at t seconds i With feasible set {tu1,tu2,...tu} m The distance matrix of} is:

[0116]

[0117] Among them, the target that does not meet the load requirement is set to inf, where inf represents a very large number, D ij represents the drone u i and the target point t j geometric distance

[0118] Construct a bipartite graph to solve: G=(U∪T,E), and the edge set is E= {( i , j )| j ∈ F i ∧ D ij ≤θ},

[0119] Find the maximum matching through the Hungarian algorithm M ⊆ E Satisfy |M| = min(m,n),

[0120] where m represents the number of drones, n represents the number of target points. When m>n, n drones get the optimal unique allocation, and the remaining m - n drones are ∅; when m<n, m drones all get the optimal unique allocation;

[0121] ⑤. Rejection + Retry Strategy:

[0122] For the subset {t j1 x j1 ,t j2 x j2 ,t j3 x j3 ...} ∈ {t 1x1 ,t 2x2 ...t mxm} of k drones, we use the "rejection" strategy to make the target points of the k drones {u j1 ,u j2 ,u j3 ...u jk}: { t j1 x j1 , t j2 x j2 ,..., t jk x jk} = ∅,

[0123] where t j1 x j1 represents the target point of the drone u j1 ,

[0124] For {t 1x1 ,t 2x2 ...t mxm}\{t j1 x j1 ,t j2 x j2 ,t j3 x j3 ...}, we adopt a "retry" strategy, repeatedly solving the problem until a feasible allocation is found, such that |M| = min(mk) in the formula. min ,n),

[0125] Where, k min For the minimum number of k=0,1,...m that satisfy the condition,

[0126] The target allocation model is now fully established. Through this model, we can ultimately obtain the following allocation:

[0127] mk drones u={u1,u2...u m-k The unique optimal target point is: { t 1, t 2... t m-k},

[0128] k drones u={u m-k+1 ,u m-k+2 ,...,u m The empty set point of} is {t} m-k+1 ,...t m}=∅.

[0129] Step 3: Collect the status parameters of the UAV in real time when it is performing the mission, including the current position, remaining payload, remaining battery power and distance to the target point, and determine the trigger conditions for material delivery, hovering and return.

[0130] Step 4: Through quantitative methods of load updating, hovering control, and charging scheduling, complete material delivery, reconnaissance hovering, coordinated following, and power replenishment;

[0131] The specific steps of the material distribution method in step four above include the following:

[0132] ① Deployment Trigger: When the real-time position of the UAV d is close to the target point (p) k ,q k Satisfy distance constraints | p(t)-p |< E delivery ( E delivery When p = 20m, the deployment preparation process is triggered. d (t) represents the position of the drone d at time t, εdelivery This is a small tolerance value, indicating when the drone reaches the target point;

[0133] ② Payload Update: After the UAV d is deployed, the payload ld of UAV d is updated. t *)=max(ld( t ) - qk ,0), where l d (t) represents the current payload of UAV d. After the mission is completed, the target point is removed from the set T.

[0134] ③ Return-to-home trigger: If the current payload of all the aforementioned drones cannot meet the requirements of the next mission. (p k ,g k )∈ T d satisfy l d ( t ) < q k If this happens, all drones will enter return-to-home mode, and the target point will be set to b.

[0135] ④ Return-to-home motion: The speed of the UAV during the return-to-home motion is as follows: , where k attract The attraction coefficient, ε, is a minimum value to prevent division by zero;

[0136] ⑤ Resupply Completed: The conditions for completing drone resupply are | p d ( t )- b |< ε ( ε =20m), if resupply is complete, reset the load and exit the return-to-base state. l d ( t + )= C d , T d ←Reassigned tasks.

[0137] The reconnaissance hovering method in step four above (refer to...) Figure 11 The specific steps include the following:

[0138] ① Hovering determination: When the distance between the UAV and the target meets the following condition: | p d ( t ) - p targetIf | < 20 (hovering radius), activate the flyaround mode; otherwise, continue moving towards the target point.

[0139] ② Position update: After time Δt p d ( t +Δ t ) =p d ( t ) +v d ( t +Δ t ) · Δ t .

[0140] The cooperative following method in step four above (refer to...) Figure 9 The specific steps include the following:

[0141] ① Selection: Select all empty target UAVs. i Select the nearest target drone. j , ensure min{|| u i - u j ||}, any u j ∈{u1,...,u m-k}, where u i ∈{u m-k+1 ,u m-k+2 ,...,u m}, u j ∈{u1,u2...u m-k};

[0142] ② Copy: Copy the most recent drone u j The target, the air target drone u i ,Right now t ixi = t jxj ;

[0143] ③ Clear: When target t jxj Complete the deployment of supplies and clear all drones in the entire region: { t 1, t 2... t m-k}{ t jxj}

[0144] The specific steps of the power replenishment method in step four include the following:

[0145] ① Charging warning: Refer to Figure 12When any drone satisfies At that time, the drone will issue a charging warning; among them T safe For a safe buffer period (20-30 seconds), D i (t) represents the Euclidean distance of UAV i from the origin, T i remain (t) represents the remaining flight time of drone i at time t. When the above conditions are triggered, all drones will return to base.

[0146] ② Real-time monitoring and alarm: Each of the aforementioned drones periodically broadcasts its own status (x i ,y i ,T i remain Perform global most critical calculations. If T critical <T safe Execute the return-to-base instruction;

[0147] ③ Coordinated Return-to-Home Control: Return-to-Home Trigger: If the current payload of all the aforementioned UAVs cannot meet the requirements of the next mission... (p k ,g k )∈T d satisfy l d ( t ) < q k Then all drones enter return-to-home mode, with the target point set as b; Return-to-home motion: the speed of the drones during the return-to-home motion is as follows , where k attract The attraction coefficient, ε, is a minimum value to prevent division by zero;

[0148] ④ Latest Return Time Calculation: For each UAV i, the shortest return time is: t i return =D i / v max

[0149] To ensure safety, the following must be met: .

[0150] The specific steps for calculating the position update in step four above are as follows:

[0151] ① Speed ​​of updates: v d ( t +Δ t ) = v d (t)+ γ (t ),

[0152] The control function γ(t) is defined as follows:

[0153]

[0154] ② Components of attraction:

[0155]

[0156] ③ Speed ​​constraint: .

[0157] Step 5: Based on the task completion progress and the UAV resource status (including payload and battery level), adjust the subsequent task execution plan according to the collaborative scheduling principle, and output the final task completion result.

[0158] The above-mentioned collaborative path planning method for multiple UAVs in complex environments introduces a "progressive constraint adaptation" method in the initial stage of task allocation to optimize the collaboration of the initial path;

[0159] The optimized resource coordination module receives high-priority task status information, medium-priority task status information, and real-time execution status information. Simultaneously, it uses the resource scheduling sub-module in the optimized resource coordination module to merge the high-priority task status information, medium-priority task status information, and real-time execution status information in the fusion layer. The merged scheduling scheme is then iteratively optimized to form a final path planning result that meets the requirements of "full task coverage, resource adaptation, and constraint compliance".

[0160] Reference Figure 1-12 As shown, the multi-UAV cooperative path planning system provided by this invention, based on a multi-UAV cooperative path planning method in complex environments, mainly includes the following components: an allocation model based on the bottleneck improvement method, a cooperative following model, a deployment and return replenishment model, a hovering model, and a UAV swarm synchronous return-to-base charging early warning model. The system implementation is divided into the following four components, and optimization proofs are provided for each of these four components. The following derivation is based on the core formulas in the specification combined with embodiments to reveal the inherent mathematical principles of the technical effect:

[0161] I. Proof of the optimization of the task allocation module

[0162] 1. Optimality of the Hungarian Algorithm (see performance reference) Figure 2 )

[0163] Original formula:

[0164] Derivation process:

[0165] Step 1: Construct a bipartite graph (G=( U ∪ T , E )), border rights ( w ij = - D ij (The minimization problem is transformed into maximizing negative weight matching).

[0166] Step 2: According to König's theorem, the maximum weight matching of a bipartite graph is equivalent to the sum of the weights of the minimum vertex cover;

[0167] Step 3: The augmenting path search of the Hungarian algorithm guarantees finding the globally optimal matching, with a time complexity of O( ). mn 2 )).

[0168] Technical effects:

[0169] The theoretical minimum of the total distance of task allocation is guaranteed by the duality of linear programming. In Example 1, the total flight distance of 4500m (A+B+C) has reached the lower bound.

[0170] 2. Convergence of the bisection threshold search

[0171] Original formula: Termination conditions θ low ≥ θ high ,

[0172] Derivation process:

[0173] Step 1: Proof of monotonicity. For any θ1 < θ2, if θ1 is feasible, then θ2 must be feasible (because...). D ij ≤ θ1⇒ D ij ≤θ2);

[0174] Step 2: Halve the search space in each iteration to converge the number of steps.

[0175] Step 3: Finally, θ satisfies ∀ x ij =1, D ij ≤ θ, and there is no smaller θ´ that satisfies the condition.

[0176] Technical effects:

[0177] In Example 1 θ =500 m The solution requires only 7 iterations, improving efficiency by 57%.

[0178] II. Stability Proof of the Dynamic Obstacle Avoidance Module

[0179] 1. Lyapunov function for virtual force model

[0180] Original formula:

[0181] Derivation process:

[0182] Step 1: Define Lyapunov functions .

[0183] Step 2: Take the derivative .

[0184] Step 3: Substitute into the velocity formula to prove... when P d ≠ b .

[0185] Technical effects:

[0186] The drone will inevitably converge to the target point. Figure 3 (Return trajectory verification), and steady-state error | p d - b |< ε =20m.

[0187] III. Robustness Proof of the Resource Collaboration Module

[0188] 1. Markov property of battery warning

[0189] Original formula:

[0190] Derivation process:

[0191] Step 1: Model as a continuous-time Markov chain, with state (s=( T remain , D i ));

[0192] Step 2: Define the set of hazardous states;

[0193] Step 3: Using the transfer rate matrix Q Calculate the absorption probability.

[0194] Technical effects:

[0195] when T safe =30s, in Example 4 the warning trigger probability is >99.9%, refer to Figure 5 The battery is depleted.

[0196] 2. The existence of Nash equilibrium in load balancing

[0197] Original formula: l d ( t * )=max( l d ( t ) - q k ,0)

[0198] Derivation process:

[0199] Step 1: Define the utility function ( U i = C i - l d ( t ));

[0200] Step 2: Prove that the game Γ = <{ u i},{ U i},{ C i}〉 is a potential game, with the potential function (Φ = ∑ U i );

[0201] Step 3: Potential games must have a pure strategy Nash equilibrium (Monderer & Shapley, 1996).

[0202] Technical effects:

[0203] In Example 2, the variance of load allocation after the return decision decreased from 12.3 to 2.1. Figure 3 ).

[0204] IV. Emergent Proof of System-Level Performance

[0205] 1. Upper bound of queuing theory for task completion time

[0206] Related formula: t i return ≤ T i remain - T safe

[0207] Derivation process:

[0208] Step 1: Model as an M / M / m queue, task arrival rate λ, service rate μ=1 / E[ D i / V max ];

[0209] Step 2: Apply Kingman's formula to calculate the average waiting time:

[0210] Step 3: Calculate the total time

[0211] Technical effects:

[0212] The total time of Examples 1-4, 656s, has an error of <5% compared to the theoretical value of 623s.

[0213] Reference Figures 1-12 As shown, this invention relates to a multi-UAV cooperative path planning system based on complex environments. A dataset is created to verify the accuracy of the UAV path planning. The dataset includes two main parts: UAV parameters and mission parameters. The verification process is as follows:

[0214] Example 1: Multi-priority task allocation and basic scheduling

[0215] Scenario setting: 3 drones need to perform 10 tasks (including 3 emergency delivery tasks of priority 1, 5 normal delivery tasks of priority 2, and 2 reconnaissance and hovering tasks of priority 3), which must be performed in priority order and meet the constraints of payload, communication distance (≤1000m) and power consumption.

[0216] Data preparation: Drone parameters and mission parameters were randomly generated using a computer seed. Drone parameters included the number of drones, maximum payload, maximum endurance, and reconnaissance hovering time. Mission parameters included mission priority, mission coordinates, material weight, or hovering requirements. Target points and drones were generated, such as... Figure 1 As shown.

[0217] Drone parameters:

[0218]

[0219] Task parameters:

[0220]

[0221] Implementation steps:

[0222] 1. Task Prioritization

[0223] Execute tasks in the order of priority 1→2→3, with priority 1 tasks processed first.

[0224] 2. Bottleneck Method Task Allocation (Refer to) Figure 8 )

[0225] Objective: To find the optimal allocation that minimizes the maximum mission distance while satisfying the drone's payload constraints, using a combination of bisection and Hungarian algorithm.

[0226] Priority 1 task assignment:

[0227] Feasible set selection: T1 (7kg) can be performed by A (8kg), B (14kg), and C (15kg); T2 (14kg) and T3 (14kg) can be performed by B (14kg) and C (15kg);

[0228] The binary search threshold θ is determined by using the distance matrix (Euclidean distance between the UAV and the mission) to ensure that all missions are available for assignment.

[0229] Hungarian algorithm matching: A→T1 (distance 450m), B→T3 (distance 600m), C→T2 (distance 550m);

[0230] 3. Cooperative following and communication constraint maintenance (refer to...) Figure 9 )

[0231] All drones start at the same position and maintain a distance (50m≤d≤1000m) during mission execution using a virtual force model:

[0232] Attraction: Directed towards the target point (e.g., A is attracted by T1, the direction is T1 - A's current position);

[0233] Repulsive force: Avoid drones with a distance of less than 50m (e.g., when the distance between A and B is 40m, a repulsive force will push them apart).

[0234] 4. Result Verification

[0235] Priority 1 task completion time: A completes T1 at step=86, B completes T3 at step=146, and C completes T2 at step=110; The effect is as follows. Figure 2 As shown.

[0236] Total flight distances: A=1200m, B=1800m, C=1500m, all of which meet the communication constraints (maximum spacing 900m < 1000m).

[0237] Example 2: Material Deployment and Return Supply Scheduling (Based on the deployment and return replenishment model, referring to...) Figure 10 )

[0238] Scenario setting: After the priority 1 task is completed, the drone's remaining payload is insufficient to execute the priority 2 task (e.g., A has 1kg remaining and cannot execute the 8kg requirement of T6), so a return to home for resupply is required.

[0239] Data preparation:

[0240] Drone remaining status: A (payload 1kg), B (payload 0kg), C (payload 1kg).

[0241] Priority 2 task requirements: T4 (14kg), T5 (5kg), T6 (8kg), T7 (9kg), T8 (12kg).

[0242] Implementation steps:

[0243] 1. Determining if deployment is complete

[0244] Deployment is triggered when the distance between the drone and the target point is less than 20m, and the payload is updated.

[0245] After A is deployed to T1, the load capacity decreases from 8kg to 1kg.

[0246] After B was deployed to T3, the load capacity decreased from 14kg to 0kg.

[0247] After T2 is deployed, the load capacity decreases from 15kg to 1kg.

[0248] 2. Return-to-home trigger condition detection

[0249] Check if the remaining load capacity meets the requirements of the next task:

[0250] A (1kg) < T6 (8kg), B (0kg) < T4 (14kg), C (1kg) < T7 (9kg), triggering a return to base for all aircraft.

[0251] 3. Return and Resupply Execution

[0252] Return target: Origin (0,0), guided by the attraction model (speed is proportional to the distance from the origin);

[0253] Resupply completion conditions: The drone reaches the origin point and the distance is <20m, then the load is reset to the maximum (A=8kg, B=14kg, C=15kg).

[0254] Time taken: A reaches the origin at step=212, B reaches it at step=248, and C reaches it at step=219. The result is as follows: Figure 3 .

[0255] 4. Secondary task allocation and execution

[0256] Priority 2 task allocation: A→T5 (5kg), B→T7 (9kg), C→T8 (12kg);

[0257] Completion time: A completes T5 at step=295, B completes T7 at step=305, and C completes T8 at step=305.

[0258] Result verification:

[0259] Total return time: 38 steps (from completion of mission 1 to the second departure);

[0260] The total flight distances for the two deployments are: A=900m, B=1100m, and C=1300m, all of which meet the communication constraints.

[0261] Example 3: Hovering Reconnaissance Mission and Cooperative Scheduling (Based on Hovering Model, Refer to...) Figure 4 , Figure 11 )

[0262] Scenario setting: After the priority 2 task is completed, the priority 3 reconnaissance task is executed (T9 needs to hover for 40s, T10 needs to hover for 25s), and the hovering time constraint and communication distance must be met.

[0263] Data preparation:

[0264] Drone status: A (payload 3kg, remaining flight time 500s), B (payload 5kg, remaining flight time 450s), C (payload 3kg, remaining flight time 400s).

[0265] Reconnaissance missions: T9 (hovering for 40 seconds), T10 (hovering for 25 seconds).

[0266] Implementation steps:

[0267] 1. Hovering task assignment

[0268] Based on the maximum hovering time: A (60s) can execute T9 (40s), B (30s) can execute T10 (25s), and C is a cooperative follower drone;

[0269] Path planning: A goes from the current position (701.72, 83.55) to T9, and B goes from (280.72, 744.17) to T10.

[0270] 2. Hover mode activation and execution

[0271] Activation conditions: The drone reaches the vicinity of the target point (distance < 20m) and enters a hovering flight phase;

[0272] Attraction: Pointed towards the center of the target point (maintaining the orbital radius);

[0273] Speed ​​constraint: Hovering speed ≤ 10m / s to avoid deviation;

[0274] Time counting: A hovers from step=613 for 40 steps; B hovers from step=631 for 25 steps.

[0275] 3. Collaborative follow-up maintenance

[0276] C follows the nearest B and maintains a distance (800m < 1000m) through virtual force, without interfering with the main task.

[0277] 4. Result Verification

[0278] Hover completion time: A completes T9 at step=653, B completes T10 at step=656;

[0279] Communication stability during hovering: The maximum distance between A and B is 950m, which meets the constraints, and the overall effect is as follows. Figure 4 .

[0280] Example 4: Charging Early Warning and Synchronous Return-to-Home Scheduling (Based on the Synchronous Return-to-Home Charging Early Warning Model, refer to...) Figure 12 )

[0281] Scenario setting: During mission execution, the drone's battery power is consumed over time. It is necessary to monitor the battery power in real time and return to the drone to recharge when an early warning is triggered.

[0282] Data preparation:

[0283] Battery parameters: A: 200s remaining, B: 180s remaining, C: 150s remaining;

[0284] Safety parameters: Safety buffer time Tsafe = 30s, average distance of the drone's current position from the origin is 800m, and maximum speed is 50m / s.

[0285] Implementation steps:

[0286] 1. Real-time power monitoring

[0287] Collect drone status every step (1 second): remaining battery life, current position distance from origin Di.

[0288] 2. Determining Early Warning Trigger Conditions

[0289] Based on the formula: Remaining battery life

[0290] C of D i =800m, return time is 800 / 50=16s, Tremain=150s ≤ 16+30=46s? This does not satisfy the condition;

[0291] When step=375, C's remaining battery life = 45s, 45 ≤ 16+30=46 → trigger warning.

[0292] 3. Synchronous return flight execution

[0293] All drones abandoned their current mission and returned to their original position; the effect was as follows. Figure 5

[0294] Charging time: 60 steps (from step=409 to 469), battery level reset to maximum (A=600s, B=500s, C=450s).

[0295] 4. Task resumes after charging.

[0296] Remaining tasks: T4 (14kg), T6 (8kg);

[0297] Allocation: B→T4, A→T6, Completion Time: B at step=550, A at step=594, Result as follows Figure 6 .

[0298] Result verification:

[0299] Warning trigger timeliness: All drones safely returned to base before their battery ran out (C remaining 45s > 16s);

[0300] Task completion time after charging: Total time 656 steps, which meets the overall scheduling requirements.

[0301] Summary of Implementation Results

[0302] 1. Constraints satisfied: In all embodiments, the UAV satisfies constraints such as communication distance, payload, and battery power;

[0303] 2. Efficiency metrics: The total completion time for multi-priority tasks was 656 seconds, and the path planning and scheduling strategies effectively balanced task priority and resource utilization.

[0304] 3. Robustness: Through mechanisms such as return-to-base resupply and charging early warning, dynamic issues such as insufficient payload and insufficient power are addressed to ensure mission continuity.

[0305] The above embodiments fully cover the entire process of multi-priority task allocation, material deployment, hovering reconnaissance and charging scheduling, verifying the feasibility and effectiveness of the model of the present invention in complex environments.

[0306] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for cooperative path planning of multiple UAVs in complex environments, characterized by: Includes the following steps: Step 1: Construct a multi-priority task collaborative scheduling model and prepare drone and task datasets for model adaptation; Step 2: Input the UAV and mission parameters into a pre-built multi-priority mission collaborative scheduling model, and the model will initially allocate missions according to priority; Step 3: Collect the status parameters of the UAV in real time when it is performing the mission, including the current position, remaining payload, remaining battery power and distance to the target point, and determine the trigger conditions for material delivery, hovering and return. Step 4: Through quantitative methods of load updating, hovering control, and charging scheduling, complete material deployment, reconnaissance hovering, coordinated following, and power replenishment; Step 5: Based on the task completion progress and the status of the UAV resources, adjust the subsequent task execution plan according to the principle of collaborative scheduling, and output the final task completion result; The specific steps of the UAV collaborative scheduling method in step two include the following: ① Determine the set of variables: Maximum payload of a drone: C = {C1, C2...C} m }; At time t, the drone's current payload is: c = {c1, c2, ..., c} m }; Target point resource requirements: D = {d1, d2, ... d} n }, where d j Represents the j-th task point t j The demand for supplies; At time t, the feasible set of drones is: {t 11 ,t 12 ...}{t 21 ,t 22 ...}...{t v1 ,t v2 ...}; Among them, {t v1 ,t v2 ...} indicates that the v-th drone satisfies c i ≦d j The target points are 1≦i≦m and 1≦j≦n. Unmanned aerial vehicles {u1,u2...u m The corresponding unique assignment is: {t} 1y1 ,t 2y2 ...t mym }, where t iyi Indicates drone u i The only selected target point, t iyi ∈ Feasible set {t i1 ,t i2 ...},t iyi Empty, meaning no target can be assigned; ② Optimization model: Establish decision variables: Based on the decision variables, the objective function is further obtained: Constraints: ; ③ Bottleneck method model establishment: Optimal allocation: Based on the optimal allocation binary search: Define a threshold θ, and use binary search to find the minimum θ such that the following allocation scheme is satisfied. D ij It is a drone. i With target t j The geometric distance satisfies the requirements of the search and rescue process. The termination condition is: ; ④ Hungarian modeling: For the binary search, the maximum threshold θ is obtained, and the optimal allocation of drones is solved using the Hungarian algorithm: , For the drone u at time t seconds i With feasible set {tu1,tu2,...tu} m The distance matrix of} is: Among them, the target that does not meet the load-bearing requirements is set to inf, where inf represents a very large number, D. ij Indicates drone u i With target point t j Geometric distance Construct a bipartite graph to solve: G = (U∪T, E), with the edge set as follows: , Finding the maximum matching using the Hungarian algorithm Satisfying |M| = min(m,n), Among them, m represents the number of UAVs, n represents the number of target points. When m > n, n UAVs get the optimal and unique allocation, and the remaining m - n UAVs are ; when m < n, all m UAVs get the optimal and unique allocation; ⑤ Elimination + Retry Strategy: For subset {t j1 x j1 ,t j2 x j2 ,t j3 x j3 ...}∈{t 1x1 ,t 2x2 ...t mxm Given k drones, we use a "removal" strategy to make k drones {u j1 ,u j2 ,u j3 ...u jk Target point: , Among them, t j1 x j1 Indicates drone u j1 target point For {t 1x1 ,t 2x2 ...t mxm }\{t j1 x j1 ,t j2 x j2 ,t j3 x j3 ...}, we adopt a "retry" strategy, repeatedly solving the problem until a feasible allocation is found, such that |M| = min(mk) in the formula. min ,n), Where, k min For the minimum number of k=0,1,...m that satisfy the condition, The target allocation model is now fully established. Based on this model, the following allocation is ultimately obtained: mk drones u={u1,u2...u m-k The unique optimal objective point is: , k drones u={u m-k+1 ,u m-k+2 ,...,u m The empty set point of} is .

2. The method for cooperative path planning of multiple UAVs in complex environments according to claim 1, characterized in that: The multi-priority task collaborative scheduling model includes an optimized task allocation module, an optimized dynamic execution module, and an optimized resource coordination module. The optimized task allocation module adopts a combination structure of bottleneck method, binary search method and Hungarian algorithm to perform three-level priority matching processing on multi-priority tasks. Priority tasks 2 and 3 will be transmitted once before matching. The task status information transmitted in the first transmission is named high-priority task status information, and the task status information transmitted in the second transmission is medium-priority task status information. The optimized dynamic execution module performs a real-time status information transmission, named real-time execution status information; the optimized dynamic execution module includes 4 core functional sub-modules, 1 status monitoring module and constraint adaptation module.

3. The method for cooperative path planning of multiple UAVs in complex environments according to claim 2, characterized in that: The proposed method for collaborative path planning of multiple UAVs in complex environments introduces a progressive constraint adaptation method in the initial stage of task allocation to optimize the collaboration of the initial path. The optimized resource coordination module receives high-priority task status information, medium-priority task status information, and real-time execution status information. Simultaneously, it uses the resource scheduling sub-module in the optimized resource coordination module to fuse the high-priority task status information, medium-priority task status information, and real-time execution status information in the fusion layer. The fused scheduling scheme is then iteratively optimized to form a final path planning result that satisfies full task coverage, resource adaptation, and constraint compliance.

4. The method for cooperative path planning of multiple UAVs in complex environments according to claim 1, 2, or 3, characterized in that: The specific steps of the material delivery method in step four include the following: ① Deployment Trigger: When the real-time position of the UAV d is close to the target point (p) k ,q k Satisfy distance constraints At that time, the deployment preparation process is triggered. The preset value is 20m, where p d (t) represents the position of the drone d at time t, ε delivery This is a small tolerance value, indicating when the drone reaches the target point; ② Payload Update: After UAV d is deployed, the payload of UAV d is updated. , where l d (t) represents the current payload of UAV d. After the mission is completed, the target point is removed from the set T. ③ Return-to-home trigger: If the current payload of all the aforementioned drones cannot meet the requirements of the next mission. satisfy If this happens, all drones will enter return-to-home mode, and the target point will be set to b. ④ Return-to-home motion: The speed of the UAV during the return-to-home motion is as follows: , where k attract The attraction coefficient, ε, is a minimum value to prevent division by zero; ⑤ Resupply Completed: The conditions for completing drone resupply are as follows: Where ε can be preset to 20m, and b is the target point position. If resupply is completed, the load is reset and the return-to-base state is exited. .

5. The method for cooperative path planning of multiple UAVs in complex environments according to claim 1, 2, or 3, characterized in that: The specific steps of the reconnaissance hovering method in step four include the following: ① Hovering determination: When the distance between the UAV and the target meets the following conditions: If p is positive, the orbital flight mode is activated; otherwise, the movement continues toward the target point. d (t) represents the position of the drone d at time t. The position of the hovering target point; ② Position update: After time Δt .

6. The method for cooperative path planning of multiple UAVs in complex environments according to claim 1, 2, or 3, characterized in that: The specific steps of the cooperative following method in step four include the following: ① Selection: Select all empty target UAVs. i Select the nearest target drone. j ,ensure any , where u i ∈{u m-k+1 ,u m-k+2 ,...,u m }, u j ∈{u1,u2...u m-k }; ② Copy: Copy the most recent drone u j The target, the air target drone u i ,Right now ; ③ Clear: When target t jxj The objective is to complete the deployment of supplies and clear out all drones in the region. .

7. The method for cooperative path planning of multiple UAVs in complex environments according to claim 4, characterized in that: The specific steps of the power replenishment method in step four include the following: ① Charging warning: When any drone meets the charging requirements... At that time, the drone will issue a charging warning; among them T safe For a safe buffer time, set it between 20-30 seconds, D i (t) represents the Euclidean distance of UAV i from the origin, T i remain (t) represents the remaining flight time of drone i at time t. When the above conditions are triggered, all drones will return to base. ② Real-time monitoring and alarm: Each of the aforementioned drones periodically broadcasts its own status (x i ,y i ,T i remain Perform global most critical calculations. If T critical <T safe Execute the return-to-base instruction; ③ Coordinated return control: Refer to steps ③ and ④ in step four, which are exactly the same as the coordinated return for material delivery and replenishment; ④ Latest Return Time Calculation: For each UAV i, the shortest return time is: To ensure safety, the following must be met: .

8. The method for cooperative path planning of multiple UAVs in complex environments according to claim 1, 2, or 3, characterized in that: The specific steps of the position update calculation method in step four are as follows: ① Speed ​​of updates: , The control function γ(t) is defined as follows: in Let d be the position of the drone at time t. The position of the hovering target point; ② Components of attraction: ③ Speed ​​constraint: .

9. The method for cooperative path planning of multiple UAVs in complex environments according to claim 1, 2, or 3, characterized in that: The drone resource status in step five includes payload and battery power.

Images