Task allocation and route autonomous planning method for multi-unmanned aerial vehicle cooperative bird repelling

Abstract

The invention relates to a task allocation and route autonomous planning method for multi-unmanned aerial vehicle cooperative bird repelling. The method comprises the following steps: step 1, digital modeling; 2, collecting information; step 3, task list establishment: adding a bird flock needing to be driven into a task list according to a constraint condition; 4, calculating parameters; step 5, task allocation: using a genetic algorithm to call an objective function to allocate bird repelling tasks to each unmanned aerial vehicle platform, and formulating a bird repelling sequence; 6, flight path planning: unmanned aerial vehicle group flight path autonomous planning based on an artificial field potential method is adopted, an ambush method bird repelling strategy is called to estimate the expected position and speed of each unmanned aerial vehicle, and finally flight path planning is realized; 7, effect evaluation: returning to the step 2 to collect information, evaluating the bird repelling effect, judging whether to continue the step 3 or not, and re-planning the task; the method has the advantages of multi-UAV collaborative bird repelling, bird repelling task allocation optimization, autonomous route planning based on artificial field potential and introduction of path optimization.

Description

Technical Field

[0001] The present invention belongs to the technical field of bird repelling by unmanned aerial vehicles, and particularly relates to a method for task allocation and autonomous route planning for multi-unmanned aerial vehicle collaborative bird repelling. Background Technique

[0002] In recent years, bird-related faults in transmission lines, airports, substations, etc. have become the third largest transmission line fault problem after lightning strikes and external force damage. However, different from line fault problems such as lightning strikes and external force damage, bird-related faults often show a relationship with natural laws, bird living habits, and human production and life, and are also closely related to the insulation configuration, tower structure, size, etc. of substations and lines. Especially in airports, the harm caused by bird-related faults is even greater, and it is very likely to cause major disasters. Traditional bird repelling means mainly focus on ground equipment, including dummy people, colored wind wheels, high - pitched loudspeakers, bird repelling gas guns, remote - controlled firecrackers, bird repelling falcons, high - frequency flashlights, bird - blocking nets, manual bird hunting, bird repelling vehicles, and spraying repellent agents, etc. These means generally have defects such as high consumption of human and material resources, being restricted by time and airspace, environmental pollution, continuously increasing bird tolerance, and large fluctuations in bird repelling effects. For airport bird repelling, the effects are even more inconspicuous. In recent years, the rapid development of unmanned aerial vehicle technology has provided a new means for bird repelling. However, the current formation path planning schemes for bird - repelling unmanned aerial vehicle groups are mostly based on the "leader - follower" model or similar "centralized" control ideas. Such schemes can solve the coordination problem of unmanned aerial vehicle groups to a certain extent during flight tasks. However, since each unmanned aerial vehicle in the group needs to plan its path based on considering the positions of other individuals in the cluster and its own position in the formation, there are deficiencies in terms of autonomy, and the computational complexity will increase significantly, resulting in the inability of the unmanned aerial vehicle group to effectively complete the preset bird repelling tasks. Therefore, it is very necessary to provide a method for task allocation and autonomous route planning for multi - UAVs collaborative bird repelling, which optimizes bird repelling task allocation, autonomously plans routes based on artificial potential fields, and introduces path optimization. Summary of the Invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a method for task allocation and autonomous route planning for multi - UAVs collaborative bird repelling, which optimizes bird repelling task allocation, autonomously plans routes based on artificial potential fields, and introduces path optimization.

[0004] The purpose of the present invention is achieved as follows: A method for task allocation and autonomous route planning for multi - UAVs collaborative bird repelling, the method includes the following steps:

[0005] Step 1: Digital modeling: Taking the center of the airport runway as the origin, establishing a Cartesian coordinate system in three - dimensional space, and using the ambush - type bird repelling strategy to divide the task space into multiple sub - spaces such as early warning, driving, maneuvering, and standby.

[0006] Step 2: Information collection: After the airport bird detection radar detects a bird flock target, it transmits the bird situation information to the ground control station. The UAVs transmit their own position and speed information to the ground control station, and the air traffic control system provides the status of the aircraft at the airport.

[0007] Step 3: Task list establishment: Add the bird flocks to be driven away to the task list according to the constraints.

[0008] Step 4: Parameter calculation: For the bird flock targets in the task list, call the bird flock threat equation to evaluate their threat levels, and call the fuzzy calculator to estimate the efficiency of each UAV in driving away each target.

[0009] Step 5: Task allocation: Use the genetic algorithm to call the objective function to allocate the bird repelling tasks to each UAV platform and formulate a bird repelling sequence.

[0010] Step 6: Trajectory planning: Adopt the autonomous route planning of the UAV swarm based on the artificial potential field method, and call the bird repelling strategy of the ambush method to estimate the expected positions and speeds of each UAV, and finally realize the trajectory planning.

[0011] Step 7: Effect evaluation: Return to Step 2 to collect information, evaluate the bird repelling effect, judge whether to continue Step 3, and re-plan the task.

[0012] The specific task allocation mathematical model in Step 1 is as follows: Assume that there are m UAVs in the bird repelling UAV formation and n bird flock targets in the task list, and design the decision parameter V ij as: Allocate one UAV to execute each target in the task list in sequence.

[0013] The objective function of the task allocation model is: Take the maximum benefit of completing all tasks as the performance index. Among them, W j is the threat index of the jth batch of bird flock targets; P ij is the interception efficiency of the ith UAV in completing the task of driving away the jth bird flock; When a UAV is assigned multiple tasks, determine the execution order of the tasks according to the benefits of the UAV in executing these tasks, M ij =W j P ij , 2≤i≤m, 1≤j≤n (3).

[0014] The constraint conditions of the task allocation model include: working area constraint, working height constraint, time constraint, safety constraint, and threat constraint.

[0015] The bird flock threat equation in step 4 includes: bird strike risk parameters and the setting of bird strike risk parameters. Specifically, based on factors such as the position, speed, quantity, volume, flight altitude of the bird flock and the attitude of the aircraft at the current airport, an estimated equation for the collision threat of the i-th batch of bird flocks is established: W i = k1 / L i + k2 cosθ i + k3v i cosβ i + k4m i + k5s i + k6R(H i ) + k7z(20), where k1 is the distance threat parameter; k2 is the azimuth threat parameter; k3 is the speed threat parameter; k4 is the quantity threat parameter; k5 is the volume threat parameter; k6 is the altitude threat parameter; k7 is the aircraft state threat parameter.

[0016] The specific process of calling a fuzzy calculator to estimate the efficiency of each UAV in driving each target in step 4 is as follows:

[0017] Step 4.1: Description of the bird flock driving efficiency parameter: The interception efficiency of the i-th UAV to complete the task of driving the j-th bird flock: P ij = f(P z , P l )(24), where P z is the UAV turning efficiency factor; P l is the UAV path efficiency factor;

[0018] Step 4.2: Setting of the bird flock driving efficiency parameter: An intelligent algorithm introducing fuzzy logic is used to estimate the bird flock driving efficiency. By establishing fuzzy rules, it is ensured that when estimating the efficiency of UAV i in driving bird flock j, the UAV adopts the correct driving strategy.

[0019] The autonomous path planning of the UAV swarm based on the artificial potential field method in step 6 specifically includes the following steps:

[0020] Step 6.1: Improved single-UAV planning based on the artificial potential field method: For the individual UAV planning scheme, the resultant force field F n is obtained by orthogonally synthesizing and normalizing the gravitational field F ar of the target point on the UAV and the repulsive force field F re of the obstacle, and taking a fixed step length l, that is:

[0021] Step 6.2: UAV formation control:;

[0022] Step 6.3: Path optimization;

[0023] Step 6.4: Evaluation system: Use the evaluation function to score the path based on several parameters of the path.

[0024] The improved single - machine planning based on the artificial potential field method in Step 6.1 is specifically as follows:

[0025] Step 6.11: Target gravitational field: Define the gravitational field as follows: where, K a is a set coefficient value related to the environment and conditions; R is the distance between the UAV and the target point at this time; R r0 is the maximum influence range of the target point on the UAV; P a is the unit direction vector;

[0026] Step 6.12: Obstacle repulsive field: Define the repulsive field as follows: where, K r is a set coefficient value related to the environment and conditions; R is the distance between the UAV and the target point at this time; r is the distance between the UAV and the obstacle point at this time; R r0 is the maximum influence distance of the UAV's attraction; R a0 is the maximum influence distance of the obstacle; P r is the unit direction vector.

[0027] The path optimization in Step 6.3 uses a correction scheme to correct the path planned by the UAV, specifically as follows:

[0028] Step 6.31: Let the adjacent 3 points at the start of the path be A(X1, Y1, Z1), B(X2, Y2, Z2), C(X3, Y3, Z3), and O be the mid - point of A and C; if the angle ∠CBA is greater than a certain set value, then replace the coordinates of point B with those of point B', which is the mid - point of B and O; the set value can be adjusted according to the debugging situation, and if it is not greater than this set value, then these 3 points remain unchanged;

[0029] Step 6.32: Slide down one point, take the following 3 points B, C, D and repeat Step 6.31 for correction;

[0030] Step 6.33: Always slide down one point and repeat the above Steps 6.31 and 6.32 until all points on this path are corrected;

[0031] Step 6.34: Repeat Steps 6.31 - 6.33 again to perform another correction on this line until no point on this line needs to be corrected anymore, then complete all the correction work on this line.

[0032] A task allocation and autonomous route planning system for multi-UAV collaborative bird repelling, the system includes a hybrid formation composed of multiple fixed-wing, rotor / flapping-wing UAVs, an airport bird detection radar, and a UAV ground control station; the task allocation and autonomous route planning system for multi-UAV collaborative bird repelling is used to execute the task allocation and autonomous route planning method for multi-UAV collaborative bird repelling as described above; the UAVs in the system adopt a bionic eagle shape design and are equipped with ultrasonic wave, flash lamp, audio generator, pyrotechnic emitter, and other bird repelling modules.

[0033] The beneficial effects of the present invention: The present invention is a task allocation and autonomous route planning method for multi-UAV collaborative bird repelling. The present invention mainly aims at dealing with bird-related faults at airports. In actual work, the present invention can also be applied to dealing with bird-related faults in fields such as transmission lines, substations, and distribution substations. For example, a relevant model can be established with a certain point on a certain line or the midpoint of a transmission line tower as the origin for bird repelling task allocation, or a relevant model can be established with the midpoint of a substation or distribution substation as the origin for bird repelling task allocation and other bird repelling schemes, not limited to the embodiment scheme of establishing a model with the airport runway center as the origin provided by the present invention to achieve bird repelling; in use, first construct a bird repelling task model and adopt a bird repelling strategy with the "ambush method" as the core; then respectively use the inference tree analysis method and the fuzzy logic analysis method to optimize the parameters of two preconditions, namely bird strike threat and interception efficiency, use the proportional navigation method to estimate the shortest time for the UAV to reach the predicted attack point, and use the genetic algorithm to calculate the objective function of task allocation. The present invention can effectively complete the preset tasks of airport bird repelling work; in the process of multi-UAV collaborative bird repelling trajectory planning, in order to realize the autonomous path planning of the UAV formation from the starting point to the target point in the specified airspace, and the aircraft group has good coordination and high execution efficiency during this process, a method for autonomous route planning of UAV groups based on the artificial potential field method is proposed, and the advantages and disadvantages of the above paths are evaluated through the established mathematical evaluation system. The present invention can better correct the inherent defects of traditional planning schemes and is more conducive to the realization of UAV groups in reality; the present invention has the advantages of multi-UAVs collaborative bird repelling, optimized bird repelling task allocation, autonomous route planning based on the artificial potential field, and introduction of path optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic diagram of the multi-UAV collaborative bird repelling structure of the present invention.

[0035] Figure 2 It is a working flow chart of the present invention.

[0036] Figure 3 It is a schematic diagram of the UAV bird repelling strategy of the present invention.

[0037] Figure 4Schematic diagram for guiding the movement analysis of the UAV of the present invention.

[0038] Figure 5 Schematic diagram of the position, speed and threat of the bird flock of the present invention.

[0039] Figure 6 Schematic diagram of the inference tree for bird strike accidents of the present invention.

[0040] Figure 7 Schematic diagram of the principle of fuzzy calculation of the present invention.

[0041] Figure 8 Schematic diagram of the path correction scheme of the present invention.

[0042] Figure 9 Flow chart of the path planning of the present invention. Detailed implementation manners

[0043] The present invention will be further described below with reference to the accompanying drawings.

[0044] Embodiment 1

[0045] As Figures 1-9 shown, a method for task allocation and autonomous route planning of multi-UAV collaborative bird repelling, the method includes the following steps:

[0046] Step 1: Digital modeling: Taking the center of the airport runway as the origin, establishing a Cartesian coordinate system in three-dimensional space, and using the ambush method bird repelling strategy to divide the task space into multiple sub-spaces such as early warning, driving, maneuvering and standby;

[0047] Step 2: Information collection: After the airport bird detection radar detects the bird flock target, it transmits the bird situation information to the ground control station, the UAV transmits its own position and speed information to the ground control station, and the air traffic control system provides the status of the aircraft at this airport;

[0048] Step 3: Task list establishment: Adding the bird flocks that need to be repelled to the task list according to the constraint conditions;

[0049] Step 4: Parameter calculation: For the bird flock targets in the task list, calling the bird flock threat equation to evaluate their threat levels, and calling the fuzzy calculator to estimate the efficiency of each UAV in repelling each target;

[0050] Step 5: Task allocation: Using the genetic algorithm to call the objective function to allocate the bird repelling tasks to each UAV platform and formulate a bird repelling sequence;

[0051] Step 6: Trajectory planning: Adopting the autonomous route planning of the UAV group based on the artificial potential field method, and calling the ambush method bird repelling strategy to estimate the expected positions and speeds of each UAV, and finally realizing the trajectory planning;

[0052] Step 7: Effect evaluation: Return to Step 2 to collect information, evaluate the bird repelling effect, and determine whether to continue with Step 3 to re-plan the task.

[0053] In this embodiment, ① Task assignment model: Assume that there are m unmanned aerial vehicles (UAVs) in the UAV formation for bird repelling, and there are n bird flock targets in the task list. Design the decision parameter V ij as: Assign one UAV to execute each target in the task list in sequence.

[0054] 1) Bird repelling strategy: The ultimate goal of the multi-UAV collaborative bird repelling method of the present invention is to ensure the safety of the airspace above the runway, that is, neither the bird flock nor the UAVs are allowed to enter the specified no-fly airspace under any circumstances; when the method of the present invention is applied to the field of transmission lines, the purpose is to ensure the safety of the airspace above the transmission lines or transmission towers, that is, neither the bird flock nor the UAVs enter the area too close to the transmission lines or transmission towers (equivalent to setting a certain range of no-fly airspace), to avoid damage to the UAVs or damage to the transmission lines and transmission towers, etc.; similarly, when applied to the fields of substations and distribution stations, the purpose is to ensure the safety of the substations or distribution stations, that is, neither the bird flock nor the UAVs enter the area too close to the substations or distribution stations (equivalent to setting a certain range of no-fly airspace), to avoid interference with the UAV signals, etc. or damage to the substations and distribution stations; according to this strategy, it can be extended and applied to bird repelling work in other fields.

[0055] Therefore, the ambush method is adopted as the bird repelling strategy: 1. Plan the space above the airport, and set up a no-fly zone (neither UAVs nor bird flocks are allowed to enter), a UAV maneuvering area (bird flocks are not allowed to enter), a UAV working area (the UAVs complete bird repelling operations), and an outer working area (UAVs are not allowed to enter); 2) Predict the flight trend of the bird flock, determine the position and speed at which the bird flock enters the attackable area, and calculate the predetermined attack position and driving direction that the UAVs need to reach accordingly; 3) Before the bird flock reaches the warning position, command the UAVs to reach the attack position and drive the bird flock in the specified direction in a timely manner; 4) After completing the task, the UAVs update the task sequence and execute the next task or return to the standby position.

[0056] The UAV bird repelling strategy is as Figure 3 shown. Project the three-dimensional space onto the horizontal plane. The endpoints and midpoint of the airport runway are A, B, and 0 respectively; the area enclosed by the no-fly line is the no-fly zone; the area enclosed between the no-fly line C and the attack line D is defined as the UAV maneuvering area, and its size is determined by the maneuvering performance of the UAVs; the area enclosed between the attack line D and the warning line E is the UAV working area, and its size is determined by the effective range of the UAVs for bird repelling; the area outside E is the outer working area.

[0057] At the moment when the radar detects the target bird flock, the target bird flock is represented by T, its initial position is T0, and its initial speed is V T0; The unmanned aerial vehicle is represented by H, its initial position is H0H0, and its initial velocity is V H0 ; When predicting the moment when the unmanned aerial vehicle drives the bird flock away, the position of the bird flock T is T1, which is the intersection point with the warning line E; at this time, the position of the unmanned aerial vehicle H is H1, which is the intersection point of OT1 and the attack line D; the expected velocity of the unmanned aerial vehicle is and its direction is the same as OT1.

[0058] 2) Objective function: There are two commonly used objective functions for the collaborative multi-task allocation model: minimizing the total energy loss of the unmanned aerial vehicle team and minimizing the total time to complete all tasks; the objective of the present invention is to prevent the bird flock from breaking in, so the shortest time index for completing the task is not pursued, but the endurance performance of the unmanned aerial vehicle and the ability to return to the cruising position to perform the task again after completing the task are required to be relatively high. Therefore, taking the maximum benefit of completing all tasks as the performance index, 1≤i≤m,1≤j≤n(2) where, W j is the threat index of the target of the jth batch of bird flocks; P ij is the interception efficiency of the ith unmanned aerial vehicle to complete the task of driving away the jth bird flock; when a UAV (unmanned aerial vehicle) is assigned multiple tasks, the order of performing the tasks is determined according to the benefits of the UAV performing these tasks, M ij =W j P ij ,2≤i≤m,1≤j≤n(3).

[0059] 3) Constraint conditions: When the following constraint conditions are met, V ij is assigned a value of 1, otherwise it is assigned a value of 0.

[0060] 1. Working area constraint: The position of the bird flock j should be located within the working area of the unmanned aerial vehicle shown in Figure 3 , that is

[0061] 2. Working height constraint: The target task height assigned to the UAV must be within the working height of the unmanned aerial vehicle, that is H min ≤H1≤H max (5); Since the present invention consists of two types of fixed-wing unmanned aerial vehicles and rotary-wing unmanned aerial vehicles, considering the good maneuverability and low-altitude passing performance of the rotary-wing aircraft, the working height of the rotary-wing aircraft is assigned within the range of 1m - 20m; the fixed-wing aircraft has advantages in speed and endurance, and its working height is assigned within the range above 20m.

[0062] 3. Time constraint: Before the bird flock j arrives at the warning line E, the ith UAV can arrive at its predicted attack point, that is where, T ij is the shortest time for the ith UAV to drive away the jth bird flock and arrive at the predicted attack point; is Figure 3 the distance at which the j-th flock of birds in reaches the warning line E; is Figure 3 the current speed of the j-th flock of birds in; The present invention uses the proportional navigation method to estimate T ij as follows:

[0063] The motion analysis of the guided UAV is as shown in Figure 4 Assume that the UAV moves in a plane, its initial position is H0, the target position is H1, and the initial speed is γ represents the track angle, θ represents the line-of-sight angle, Φ represents the speed vector lead angle, R represents the straight-line flight distance, and a H represents the guidance command, (X H , Y H ) represents the position of the UAV. The magnitude of the UAV flight speed remains unchanged, and N is a proportionality constant. Then the flight motion relationship of the UAV guidance satisfies: Define the proportionality coefficient The initial conditions are Solving the above equations, we get: In the proportional navigation method, when the proportionality coefficient N takes the value of 3, the control energy is optimal; assuming that the lead angle Φ is a small angle, performing a second-order Taylor expansion on the above formula, we get the differential equation: In the proportional navigation method, the process of the UAV approaching the target is the process in which the initial lead angle gradually approaches 0 under the guidance command. Therefore, calculating the remaining flight time t ij can be converted to solving the time required for the lead angle to tend to 0; Integrating both sides of the above formula and substituting the termination condition t = t ij we get: Substitute into the solution formula (10), we get

[0064] 4) Safety constraint: Only one UAV is assigned to drive each target, which can save costs and improve efficiency. At the same time, the risk of UAV collision can be avoided, that is:

[0065] 5) Threat constraint: Due to different factors such as the position, speed, quantity, volume, flight altitude of each flock of bird targets and the attitude of current airport aircraft, their threats to aircraft near the airport are different; Set a bird strike risk threshold W f according to expert experience. When the threat level W i of the flock of bird targets does not exceed the threshold W f , it does not need to be included in the task list, but a continuous observation strategy is adopted, that is: W i ≥W f (13).

[0066] ②Parameter optimization and selection: 1) The schematic diagram of the positions and speeds of the bird flocks is as Figure 5 shown. The takeoff and landing lines of the aircraft are at both ends of the runway. The closer the current position of the bird flock is to the runway and the closer it is to both ends of the runway, the greater the threat; as Figure 5 can be seen, A, B, and O are the two ends and the midpoint of the runway. The area enclosed by A1 - A2 - C2 - B2 - B3 - C3 - A3 - A4 - C4 - B4 - B1 - C1 - A1 is the no - fly zone. T i is the i - th bird flock detected by the radar; L i is the minimum distance from T i to the no - fly zone, and θ i is the angle between OT i and the runway. Then the threat of the bird flock's position: W 1i = k1 / L i + k2 cosθ i (14); the greater the velocity component of the target bird flock pointing to the center of the runway, the greater the threat. As Figure 5 can be seen, v i is the flight velocity vector of T i , and β i is the angle between OT i and v i . Then the threat of the bird flock's velocity: W 2i = k3v i cosβ i (15); the more the number of flying birds in the target bird flock and the larger the volume of the space occupied by the bird flock, the greater the threat. Define m i as the number of individual flying birds in T i , and s i as the radar cross - sectional area of T i on the bird situation radar. Then the threat of the bird flock's number and volume: W 3i = k4m i + k5s i (16); the threat function of the bird flock's flight altitude: W 4i = k6R(H i )(17), where H i is the average flight altitude of the i - th bird flock from the ground; H d is the ground altitude, with a value of 0; H h is the altitude at which bird strikes occur most frequently; H t is the upper limit of the working altitude of the UAV; the threat of the current runway aircraft's attitude: W 5i = k7z(19). The specific assignment of the threat of the runway aircraft's attitude is: for aircraft takeoff and approach, z = 5; for aircraft climb, z = 4; for aircraft descent, z = 3; for aircraft landing, z = 2; for aircraft taxiing, z = 1.

[0067] Based on factors such as the position, velocity, quantity, volume, flight altitude of the bird flock, and the attitude of the aircraft at the current airport, establish the equation for estimating the collision threat of the \(i\)-th batch of bird flocks: \(W\) i = \(k_1 / L\) i + \(k_2\cos\theta\) i + \(k_3v\) i \(\cos\beta\) i + \(k_4m\) i + \(k_5s\) i + \(k_6R(H\) i ) + \(k_7z(20)\), where \(k_1\) is the distance threat parameter; \(k_2\) is the azimuth threat parameter; \(k_3\) is the velocity threat parameter; \(k_4\) is the quantity threat parameter; \(k_5\) is the volume threat parameter; \(k_6\) is the altitude threat parameter; \(k_7\) is the aircraft state threat parameter.

[0068] 2) Setting of bird strike risk parameters: Use the inference tree analysis method to set the 7 parameters of the bird strike threat estimation equation. The steps are as follows:

[0069] 1. Determine that "the bird flock breaks through the control line and collides with the aircraft, causing an accident" is the top event \(T\);

[0070] 2. Analyze the logical relationship between each bottom event, and establish an inference tree as Figure 6 shown, Figure 6 in which, "⊙" represents logical AND, "+" represents logical OR, \(\{x_1, x_2, x_3, x_4, x_5, x_6, x_7\}\) represents 7 bottom events, and these bottom events correspond one-to-one with the 7 bird strike threat parameters \(\{k_1, k_2, k_3, k_4, k_5, k_6, k_7\}\);

[0071] 3. Use the downward method to obtain the minimum cut sets of the bottom events that trigger the top event: \(K_1=\{x_1, x_4, x_6, x_7\}\), \(K_2=\{x_2, x_4, x_6, x_7\}\), \(K_3=\{x_3, x_4, x_6, x_7\}\), \(K_4=\{x_1, x_5, x_6, x_7\}\), \(K_5=\{x_2, x_5, x_6, x_7\}\), \(K_6=\{x_3, x_5, x_6, x_7\}\);

[0072] 4. Quantitatively calculate the importance of each bottom event: Use a questionnaire survey to obtain the occurrence probability \(P(x\) i ) of each bottom event, and then obtain the occurrence probability \(P(K\) i ) of each minimum cut set, and use equations (20) and (21) to calculate the occurrence probability \(P(T)\) of the top event, As can be seen from equation (20), \(P(T)\) is determined by the occurrence probability \(P(K\) i ) of each minimum cut set, and the occurrence probability \(P(K\) i ) of each minimum cut set is determined by the occurrence probability \(P(x\) i)Determined by the product. Therefore, the functional relationship between the independent variable P(x i ) and the dependent variable P(T) is defined as g, i.e.: Finally, the importance degree of each basic event can be calculated: I i = g(P(x1), P(x2),..., P(x7)) - g(P(x1), P(x2),..., P(x i-1 ), 0, P(x i+1 ),..., P(x7))(23), and then the required parameter values are obtained after normalization processing.

[0073] ③Description of the bird flock driving efficiency parameter: As defined in the objective function of Equation (2), the interception efficiency of the i-th UAV to complete the task of driving the j-th bird flock: P ij = f(P z , P l )(24), where P z is the UAV turning efficiency factor, determined by the relative velocity ΔV between the UAV and the bird flock target; P l is the UAV path efficiency factor, determined by the displacement L between the UAV and the bird flock target.

[0074] ④Setting of the bird flock driving efficiency parameter: According to the different relative positions and relative velocities of the runway, the UAV and the target bird flock, the relative postures between the UAV and the bird flock can be divided into: facing each other, back to back, moving in the same direction outwards and moving in the same direction inwards, etc. Therefore, the strategies that the UAV needs to adopt to drive the bird flock can also be divided into different strategies such as frontal attack, intercepting from the front, pursuit, and pursuing from the back; therefore, an intelligent algorithm introducing fuzzy logic is used to estimate the bird flock driving efficiency. By establishing fuzzy rules incorporating human experience judgment, it is ensured that when estimating the efficiency of the i-th UAV driving the j-th bird flock, the UAV adopts the correct driving strategy; it is stipulated that the direction from the center of the runway to the UAV is positive. The driving strategies of the UAV are defined under various relative postures, and the driving efficiency is judged according to human experience; the principle of the fuzzy calculator is as Figure 7 shown. The efficiency of the UAV driving the target bird flock is calculated using the Matlab fuzzy logic toolbox. By establishing triangular membership functions and "if-then" type fuzzy rules, and using the Mamdani fuzzy inference method, the accurate data measured by the bird situation radar is input, and according to the sorting of the task list, the interception efficiency P ij of the i-th UAV to complete the task of driving the j-th bird flock is calculated and output in sequence.

[0075] The present invention relates to a method for task allocation and autonomous route planning for multi-UAV collaborative bird repelling. In use, to solve the problem of low efficiency of traditional airport bird repelling, the present invention introduces multi-UAV collaborative technology, completes task planning modeling based on the bird repelling strategy of the "ambush method", and respectively introduces the inference tree analysis method and the fuzzy logic analysis method to optimize the selection of parameters, and evaluates two preconditions for task planning, namely bird strike threat and UAV interception efficiency; sets multiple constraint conditions to ensure the actual effect of task planning, and estimates the task execution time of the UAVs based on the proportional guidance system. The method of the present invention can effectively complete the task allocation work of multi-UAV collaborative bird repelling under multiple constraint conditions and when facing multiple bird targets at the same time; the present invention has the advantages of multi-UAVs collaborative bird repelling, optimized bird repelling task allocation, autonomous route planning based on artificial potential field, and introduction of path optimization.

[0076] Embodiment 2

[0077] As Figures 1-9 shown, a method for task allocation and autonomous route planning for multi-UAV collaborative bird repelling, the method comprising the following steps:

[0078] Step 1: Digital modeling: Taking the center of the airport runway as the origin, a Cartesian coordinate system is established in three-dimensional space, and the ambush method bird repelling strategy is used to divide the task space into multiple sub-spaces such as early warning, driving, maneuvering, and standby;

[0079] Step 2: Information collection: After the airport bird detection radar detects a bird flock target, it transmits the bird situation information to the ground control station, and the UAV transmits its own position and speed information to the ground control station, and the air traffic control system provides the status of the aircraft at the airport;

[0080] Step 3: Task list establishment: Add the bird flocks to be driven to the task list according to the constraint conditions;

[0081] Step 4: Parameter calculation: For the bird flock targets in the task list, call the bird flock threat equation to evaluate their threat levels, and call the fuzzy calculator to estimate the efficiency of each UAV in driving each target;

[0082] Step 5: Task allocation: Use the genetic algorithm to call the objective function to allocate the bird repelling tasks to each UAV platform and formulate a bird repelling sequence;

[0083] Step 6: Trajectory planning: Adopt the autonomous route planning of the UAV swarm based on the artificial potential field method, and call the ambush method bird repelling strategy to estimate the expected positions and speeds of each UAV, and finally realize the trajectory planning;

[0084] In this embodiment, ① Improved single-UAV planning based on the artificial potential field method: For the individual UAV planning scheme, the resultant force field F n is the gravitational field F of the target point on the UAV ar, the repulsive force field F of the obstacle on the UAV re This is obtained by orthogonally synthesizing and normalizing these two fields with a fixed step size, that is: Among them, l is the set fixed step size; it should be noted that because there are multiple obstacles, the repulsive force field is generated by the synthesis of all obstacles.

[0085] Considering that the UAV flies at full speed during the bird repelling process, the flight time and distance are linearly related. Since each point in the path planning is the position where the UAV should be in the next time frame, the step size can reflect the speed of the UAV at this time; therefore, after taking a fixed step size, it is ensured that there will be no speed oscillation and jump between frames during this section of flight of the UAV; the total field value of the UAV changes with the position of the UAV, and the UAV moves towards the target point under the action of the field.

[0086] 1) Target gravitational field: The target gravitational field attracts the UAV to move in the target direction, ensuring that the UAV can reach the target point smoothly. The gravitational field is defined as follows: Among them, K a is a set coefficient value related to the environment and conditions; R is the distance between the UAV and the target point at this time; to prevent the gravitational value from being too large and affecting the planning when the UAV is far from the target point, R r0 is the maximum influence range of the target point on the UAV. When the distance exceeds this distance, the gravitational value of the target point on the UAV remains the maximum and no longer increases; P a is the unit direction vector, and the direction points from the position of the UAV to the target point; it can be seen from the formula that the gravitational force is linearly related to the distance, and the farther the distance, the greater the gravitational force.

[0087] 2) Obstacle repulsive force field: The obstacle repulsive force field repels the UAV during the movement of the UAV towards the target point, so that the UAV can avoid obstacles to prevent collisions. The repulsive force field is defined as follows: Among them, K r is a set coefficient value related to the environment and conditions; R is the distance between the UAV and the target point at this time; r is the distance between the UAV and this obstacle point at this time; R r0 is the maximum influence distance of the UAV's attraction degree. When the UAV is far enough from a certain obstacle, it will no longer be affected by this obstacle, so R a0 is the maximum influence distance of the obstacle. When the distance between the UAV and the obstacle is greater than this value, the repulsive force is 0; P r is the unit direction vector, and the direction points from the obstacle to the position of the UAV; considering that there will be many obstacles in the flight environment, in the adjacent direction, only avoiding the nearest obstacle can achieve the purpose of obstacle avoidance, and there is no need to consider the distant obstacles. Therefore, it is stipulated that if there are several obstacles within a 15° sector centered on the position of the UAV, the nearest one to the UAV is taken to calculate the repulsion degree, and the others are no longer counted.

[0088] ② UAV formation control: Number all UAVs in the formation one by one. When starting the planning, start from the UAV numbered 1 and plan the positions that should be occupied in the first time frame one by one. After the UAV numbered at the end completes the planning of this time frame, restart the path planning for the next time frame (T = T + 1) of the UAV numbered 1, and so on until all UAVs reach the target position. If a UAV with a certain number has reached the target point, skip the planning of this UAV in this time frame.

[0089] To enable all UAVs in the formation to autonomously plan without colliding with other UAVs in the formation, the following method is adopted: Add a one-dimensional time parameter T when defining the obstacle coordinates, indicating that the obstacle exists in certain specific time frames. Note: Generally, obstacles detected by sensors, since the existence time cannot be determined, are all defined as T ∈ (0, ∞).

[0090] When the UAV numbered 1 completes the path planning for any time frame T x after that, add the coordinates of this point (with the time parameter T = T x ) to the obstacle information, that is, regard the position where this UAV is located at this moment as an obstacle that other UAVs in the formation should avoid during planning. All other UAVs in the formation should complete this step after completing the planning according to the new obstacle information. Note: Due to the time parameter, this obstacle only exists at this position at this moment.

[0091] ③ Path optimization: Adopt a correction scheme to correct the path planned by the UAV to avoid the defect of path point jitter caused by falling into the minimum value in some cases due to large turning angles. Specifically:

[0092] 1) Let the adjacent three points at the start of the path be A(X1, Y1, Z1), B(X2, Y2, Z2), C(X3, Y3, Z3), and O be the midpoint of A and C. If the angle ∠CBA is greater than a certain set value, then replace the coordinates of point B with those of point B', which is the midpoint of B and O. The correction schematic is as Figure 8 shown; the set value can be adjusted according to the debugging situation. If it is not greater than this set value, these three points remain unchanged.

[0093] 2) Slide down one point, and take the following three points B (or B')、C、D and repeat 1) for correction.

[0094] 3) Keep sliding down one point and repeat the above 1) and 2) until the correction of all points on this path is completed.

[0095] 4) Repeat 1)-3) again to perform the second correction of this line until no point on this line needs to be corrected anymore, then complete all the correction work for this line.

[0096] It should be noted that: This correction will cause the step length of the path to decrease at the correction point, that is, the flight speed will decrease, but the degree of decrease only depends on the angle of rotation of this section. The larger the angle of rotation between adjacent path points, the smaller the corrected value of the step length. Therefore, both the increase and decrease of the speed are smooth changes and will not cause sudden changes, that is, it will not affect the flight stability. At the same time, it can also avoid dangers and flight errors caused by excessive flight speed at places where the flight path turns too much.

[0097] ④ Evaluation system: Use an evaluation function to score the path through several parameters of the path; the rules are as follows:

[0098] 1) Obtain all discrete points of a path, calculate and quantify the score through a formula. The lower the score, the better the characteristics; its evaluation indicators include: "total length", "hazard degree", "smoothness";

[0099] Among them, total length: Connect adjacent two points and sum to get the total length Y1;

[0100] Hazard degree: Take the first point of the path and calculate its distances to all n static obstacles, denoted as d1 - d n ; Take the reciprocal and sum them, denoted as D1; Perform this calculation for all m path points to obtain D1 - D m And sum them to obtain the hazard degree value Y2;

[0101] Smoothness: Take the starting 3 points W1, W2, W3 of the path points, calculate the included angle between the two lines of these 3 points, denoted as φ, and calculate to get S1. Slide down one point and calculate S2 in the same way. Slide down sequentially until the calculation of all m points is completed to obtain S1 - S m-2 and sum them to obtain the smoothness value Y3;

[0102] 2) Normalize Y1, Y2, and Y3, and set coefficients K1, K2, and K3; then the goodness - badness score SC is SC = K1S1 + K2S2 + K3S3;

[0103] 3) Sum the path scores SC1 - SC u of all N Nu unmanned aerial vehicles in the fleet to obtain the goodness - badness score S a .

[0104] Step 7: Effect evaluation: Return to Step 2 to collect information, evaluate the bird - repelling effect, and judge whether to continue Step 3 to re - plan the task.

[0105] The present invention relates to a method for task allocation and autonomous path planning for multi-UAV collaborative bird repelling. In use, to solve the problem of low efficiency of traditional airport bird repelling, the present invention introduces multi-UAV collaborative technology, completes task planning modeling based on the bird repelling strategy of the "ambush method", respectively introduces the inference tree analysis method and the fuzzy logic analysis method to optimize the selection of parameters, and evaluates two preconditions for task planning, namely bird strike threat and UAV interception efficiency; sets multiple constraint conditions to ensure the actual effect of task planning, and estimates the UAV task execution time based on the proportional navigation system. The method of the present invention can effectively complete the multi-UAV collaborative bird repelling task allocation under multiple constraint conditions and when facing multiple bird targets; in addition, during the path planning process of multi-UAV collaborative bird repelling, the present invention adopts an autonomous path planning method for UAV formation based on the "artificial potential field method", introduces a time variable, treats the UAV as a dynamic obstacle for easy handling, and introduces a path optimization scheme to optimize large turning angles and oscillations in the path, which can ensure that the UAV formation smoothly avoids obstacles and reaches the target point; the present invention has the advantages of multi-UAVs collaborative bird repelling, optimized bird repelling task allocation, autonomous path planning based on artificial potential field, and introduction of path optimization.

[0106] Embodiment 3

[0107] As Figures 1-9 shown, a system for task allocation and autonomous path planning for multi-UAV collaborative bird repelling, as Figure 1 shown, the system includes a mixed formation composed of multiple fixed-wing, rotor / flapping-wing UAVs, an airport bird detection radar, and a UAV ground control station; the system for task allocation and autonomous path planning for multi-UAV collaborative bird repelling is used to execute the method for task allocation and autonomous path planning for multi-UAV collaborative bird repelling as described above; the UAVs in the system are designed with a bionic eagle shape and are equipped with ultrasonic wave, flash, audio generator, pyrotechnic emitter, and other bird repelling modules.

Claims

1. A task allocation and autonomous route planning method for multi-UAV collaborative bird repelling, characterized in that: The method includes the following steps: Step 1: Digital modeling: Taking the center of the airport runway as the origin, a Cartesian coordinate system is established in three-dimensional space, and the task space is divided into multiple sub-spaces such as early warning, driving, maneuvering, and standby by using the bird repelling strategy of the ambush method; Step 2: Information collection: After the airport bird detection radar detects a bird flock target, it transmits the bird situation information to the ground control station. The UAV transmits its own position and speed information to the ground control station, and the air traffic control system provides the status of the aircraft at the airport; Step 3: Task list establishment: Add the bird flocks to be driven to the task list according to the constraint conditions; Step 4: Parameter calculation: For the bird flock targets in the task list, call the bird flock threat equation to evaluate their threat levels, and call the fuzzy calculator to estimate the efficiency of each UAV in driving each target; Step 5: Task allocation: Use the genetic algorithm to call the objective function to allocate the bird repelling tasks to each UAV platform and formulate a bird repelling sequence; Step 6: Trajectory planning: Adopt the autonomous route planning of the UAV swarm based on the artificial potential field method, and call the bird repelling strategy of the ambush method to estimate the expected positions and speeds of each UAV, and finally realize the trajectory planning; Step 7: Effect evaluation: Return to step 2 to collect information, evaluate the bird repelling effect, judge whether to continue step 3, and re-plan the task.

2. The task allocation and autonomous route planning method for multi-UAV collaborative bird repelling according to claim 1, wherein: The task assignment mathematical model in the above step 1 is specifically as follows: Assume that there are m drones in the bird repelling drone formation and n bird flock targets in the task list, and design the decision parameter V ij as follows: Assign one drone to execute each target in the task list in sequence.

3. The task allocation and autonomous route planning method for multi-UAV collaborative bird repelling according to claim 2, wherein: The objective function of the task allocation model is: taking the maximum benefit of completing all tasks as the performance index, where W j is the threat index of the j-th flock target; P ij is the interception efficiency of the i-th UAV to complete the task of driving away the j-th flock of birds; when a UAV is assigned multiple tasks, the order of task execution is determined according to the benefits of the UAV executing these tasks, M ij =W j P ij , 2 ≤ i ≤ m, 1 ≤ j ≤ n (3).

4. The task assignment and autonomous route planning method for multi-UAV collaborative bird repelling according to claim 3, characterized in that: The constraint conditions of the task allocation model include: working area constraint, working height constraint, time constraint, safety constraint, and threat constraint.

5. The task allocation and autonomous route planning method for multi-UAV collaborative bird repelling according to claim 1, characterized in that: The bird flock threat equation in step 4 includes: bird strike risk parameters and the setting of bird strike risk parameters. Specifically, based on factors such as the position, speed, quantity, volume, flight altitude of the bird flock and the attitude of the current airport aircraft, an estimation equation for the collision threat of the i-th batch of bird flocks is established: W i = k1 / L i + k2cosθ i + k3v i cosβ i + k4m i + k5s i + k6R(H i ) + k7z(20), where k1 is the distance threat parameter; k2 is the azimuth threat parameter; k3 is the speed threat parameter; k4 is the quantity threat parameter; k5 is the volume threat parameter; k6 is the altitude threat parameter; k7 is the aircraft state threat parameter.

6. The task allocation and route autonomous planning method for multi-UAV collaborative bird repelling according to claim 5, characterized in that: The specific method of calling the fuzzy calculator in step 4 to estimate the efficiency of each UAV in driving each target is as follows: Step 4.1: Description of the bird flock driving efficiency parameter: The interception efficiency of the i-th UAV to complete the task of driving the j-th bird flock: P ij = f(P z , P l )(24), where P z is the UAV turning efficiency factor; P l is the UAV path efficiency factor; Step 4.2: Setting of bird flock driving efficiency parameters: Use an intelligent algorithm introducing fuzzy logic to estimate the bird flock driving efficiency. By establishing and adding fuzzy rules, ensure that when estimating the efficiency of UAV i in driving bird flock j, the UAV adopts the correct driving strategy.

7. A task assignment and route autonomous planning method for multi-UAV collaborative bird repelling according to claim 1, characterized in that: The autonomous route planning of the UAV swarm based on the artificial potential field method in step 6 specifically includes the following steps: Step 6.1: Improved single-vehicle planning based on the artificial potential field method: For the individual planning scheme of the UAV, the resultant force field F n is the gravitational field F of the target point on the UAV ar , and the repulsive force field F of the obstacle on the UAV re These two fields are orthogonally synthesized and then normalized, and obtained by taking a fixed step size l, that is: Step 6.2: UAV formation control:; Step 6.3: Path optimization; Step 6.4: Evaluation system: Use an evaluation function to score the path based on several parameters of the path.

8. The task allocation and autonomous route planning method for multi-UAV collaborative bird repelling according to claim 7, wherein: The improved single UAV planning based on the artificial potential field method in step 6.1 is specifically as follows: Step 6.11: Target gravitational field: Define the gravitational field as follows: where K a is a set coefficient value related to the environment and conditions; R is the distance of the UAV from the target point at this time; R r0 is the maximum influence range of the target point on the UAV; P a is the unit direction vector; Step 6.12: Obstacle repulsive force field: Define the repulsive force field as follows: Among them, K r is a setting coefficient value related to the environment and conditions; R is the distance of the UAV from the target point at this time; r is the distance of the UAV from this obstacle point at this time; R r0 is the maximum influence distance of the UAV attraction; R a0 is the maximum influence distance of the obstacle; P r is the unit direction vector.

9. The task assignment and autonomous route planning method for multi-UAV collaborative bird repelling according to claim 7, characterized in that: The path optimization in step 6.3 uses a correction scheme to correct the path planned by the UAV, specifically as follows: Step 6.31: Let the adjacent 3 points at the start of the path be A(X1, Y1, Z1), B(X2, Y2, Z2), C(X3, Y3, Z3), and O be the midpoint of A and C; if the angle ∠CBA is greater than a certain set value, then replace the coordinates of point B with those of point B′, that is, the midpoint of B and O; the set value can be adjusted according to the debugging situation. If it is not greater than this set value, then these 3 points remain unchanged; Step 6.32: Slide down one point, take the following 3 points B, C, D and repeat step 6.31 for correction; Step 6.33: Always slide down one point and repeat the above steps 6.31 and 6.32 until all points on this path are corrected; Step 6.34: Repeat steps 6.31 - 6.33 again to perform another correction of this line until no point on this line requires correction anymore, then the entire correction work of this line is completed.

10. A task allocation and route autonomous planning system for multi-UAV collaborative bird repelling, characterized in that: The system includes a mixed formation composed of multiple fixed-wing, rotor / flapping-wing UAVs, an airport bird detection radar, and a UAV ground control station; the multi-UAV collaborative bird repelling task allocation and route autonomous planning system is used to execute the multi-UAV collaborative bird repelling task allocation and route autonomous planning method according to any one of claims 1 - 9; the UAVs in the system adopt a bionic eagle shape design and are equipped with ultrasonic wave, flash lamp, audio generator, pyrotechnic emitter, and other bird repelling modules.

Images