Unmanned ship cluster landing target dynamic matching method and matching device

Abstract

The application discloses a kind of unmanned ship cluster landing target dynamic matching method and equipment, its matching method minimizes the model of unmanned aerial vehicle and unmanned ship distribution problem with the goal of establishing unmanned aerial vehicle cluster execution landing total energy consumption and using improved auction algorithm to solve model, auction algorithm sets return and energy consumption negative correlation, through multiple iterations and price update, and introduce variable near_count when each time updating price, in positive income, judge whether to satisfy |c i,q+1 -c i,q |<ε, if yes, then variable near_count is accumulated 1, and serial number q is accumulated 1, to determine the number of bids close to the highest yield, according to variable near_count Guiding selection of different pricing strategy, when near_count≤1, update the bid of unmanned ship J i,1 When near_count>1, update the bid of unmanned ship J i,1 Through the above method, each unmanned aerial vehicle can be quickly matched with an appropriate unmanned ship, and the energy consumption of the unmanned aerial vehicle cluster is minimized, improving the task execution efficiency.

Description

Technical Field

[0001] This invention belongs to the field of cross-domain heterogeneous unmanned equipment matching technology, and more specifically, relates to a dynamic matching method and matching device for unmanned aerial vehicle swarm landing targets. Background Technology

[0002] With the development of maritime exploration, increasingly stringent requirements are being placed on the operational capabilities of unmanned surface vessels (USVs). Single USV platforms are often limited by their own structure and the variability of the environment, making them inadequate for these tasks. However, by mounting freely moving unmanned aerial vehicles (UAVs) on USVs, the perception dimension of the USV can be expanded from two-dimensional to three-dimensional, giving it stereoscopic monitoring capabilities. Simultaneously, UAVs can leverage the USV's advantages of large payload capacity and long endurance for recovery and recharging, effectively extending its operational time. Therefore, organically combining UAVs and USVs into a cross-domain heterogeneous collaborative system can protect the vulnerabilities of both, achieving the goal of complementing each other's strengths and maximizing efficiency.

[0003] When a drone completes a phase of its mission or its remaining flight time falls below a certain threshold, it needs to be landed and recovered by an unmanned surface vessel (USV) for charging and resuming flight. Since drone and USV swarms typically cover a large area and are spaced far apart, how to allocate a suitable USV to each drone to reduce energy consumption and improve mission efficiency when multiple drones need to land simultaneously is a pressing issue that needs to be addressed. Summary of the Invention

[0004] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a dynamic target matching method and matching device for unmanned surface vessel (USV) swarm landing. The purpose is to assign a suitable USV to each USV during USV swarm landing to reduce the energy consumption of the USV swarm and improve mission execution efficiency.

[0005] To achieve the above objectives, according to one aspect of the present invention, a method for dynamic target matching of unmanned surface vessels (USVs) swarm landing is provided, comprising:

[0006] Step S1: Calculate the energy E consumed by each UAV i to reach each unmanned surface vessel j. ij ;

[0007] Step S2: Establish a mathematical model for the allocation problem of UAVs and unmanned surface vessels with the objective of minimizing the total energy consumption of UAV swarm landing;

[0008] Step S3: Solve the mathematical model using an auction algorithm to determine the allocation, including performing:

[0009] Step S31: Determine whether the selected drone i satisfies the following conditions. If yes, proceed to step S32; otherwise, proceed to step S35.ij As a decision variable, if unmanned surface vessel j is assigned to unmanned aerial vehicle i, then x ij =1, otherwise x ij =0, where N is the number of unmanned surface vessels;

[0010] Step S32: Calculate the revenue r of drone i bidding for each unmanned surface vessel j. ij -y ij (t), where r ij In return for the drone i landing on the unmanned surface vessel j, y ij (t) represents the bidding price of UAV i for UAV j in the t-th iteration, and the return r. ij With energy E ij Negative correlation;

[0011] Step S33: Among the positive returns, determine whether |c i,q+1 -c i,q | < ε, where ε is a set threshold value. If it is, the variable near_count is incremented by 1, and the index q is incremented by 1. Step S33 is repeated until |c i,q+1 -c i,q The comparison of elements in the range of |≥ε or positive returns has been completed, where c i,q This represents the q-th position of the positive revenue of drone i after sorting from high to low, with the initial value of q being 1;

[0012] Step S34: When near_count≤1, update the unmanned surface vessel J. i,1 The price is When near_count > 1, update the unmanned surface vessel J. i,1 The price is Among them, J i,1 The highest-profit unmanned surface vessel in the bidding for drones. For the unmanned surface vessel J i,1 The current price, σ is the set adjustable coefficient, and δ is the set minimum price increment;

[0013] Step S35: Obtain the latest quotes for each UAV, update the quote for UAV i for each UAV to the highest existing quote for the corresponding UAV, when UAV i quotes for UAV J i,1 When the current price is not the highest price, update Otherwise, maintain

[0014] Step S36: Update drone i and proceed to step S31 until all drones have been updated;

[0015] Step S37: Determine whether each UAV has been assigned to a different unmanned surface vessel. If yes, end the iteration; otherwise, proceed to step S31 to continue the iteration.

[0016] In one embodiment, after step S3, the method further includes:

[0017] Step S4: Control the drone to fly towards the matched unmanned surface vessel, and after the assignment is completed and the interval Δt is reached, determine whether there are any drones that have not yet successfully landed. If so, jump to step S1 to redistribute the tasks of the drones that have not yet successfully landed and the available unmanned surface vessels until all drones have successfully landed.

[0018] In one embodiment, the formula for calculating the interval time Δt is:

[0019]

[0020] Among them, L UAV D is the wheelbase of the drone. UAV The diameter of the drone's propeller blades, Let be the optimal constant speed of drone i obtained in the previous assignment.

[0021] In one embodiment, the return r ij The calculation formula is:

[0022]

[0023] Where ξ is a constant greater than zero.

[0024] In one embodiment, the minimum price increment δ = 1 / M, where M is the number of drones.

[0025] In one embodiment, in step S1, the optimal energy E ij The calculation formula is:

[0026]

[0027] Where, η tot For the overall efficiency of the drone's motor and propeller, E hov E consumes energy for drone hovering. acc E consumes energy due to changes in the kinetic energy of the drone. drag Energy is consumed to help drones overcome air resistance.

[0028] In one embodiment, the drone hovering consumes energy E. hov Energy consumption E due to changes in the kinetic energy of the drone acc And the energy consumed by drones to overcome air resistance E drag The calculation formulas are as follows:

[0029] When the drone i and the unmanned surface vessel j move in the same direction

[0030]

[0031]

[0032]

[0033]

[0034] When drone i and unmanned surface vessel j move toward each other...

[0035]

[0036]

[0037]

[0038]

[0039] Where m is the mass of the drone, g is the acceleration due to gravity, ρ is the air density, and C is the mass of the drone. D Let A be the drag coefficient, and A be the coverage area when the UAV propeller is rotating. eff Let d be the effective windward area of ​​the UAV, d be the initial distance between the UAV and the unmanned surface vessel, and v be the effective windward area of ​​the UAV. i v is the current flight speed of the drone. j v is the current speed of the unmanned surface vessel. m Let t be the uniform flight speed of the UAV during its uniform flight phase. tot Let be the time required for the drone to land on the unmanned surface vessel, and 'a' be the flight acceleration.

[0040] In one embodiment, the optimal uniform flight speed is solved. And order in,

[0041] When the drone and the unmanned surface vessel are moving in the same direction, the optimal uniform flight speed is... The solution formula is:

[0042]

[0043]

[0044] When the drone and the unmanned surface vessel are moving toward each other, the optimal uniform flight speed is... The solution formula is:

[0045]

[0046]

[0047] In one embodiment, in step S2, the mathematical model is expressed as follows:

[0048]

[0049]

[0050]

[0051]

[0052]

[0053] Where, x ij Let x be the decision variable. If drone i chooses unmanned surface vessel j to land, then x ij =1, otherwise x ij =0, M is the number of drones, and N is the number of unmanned surface vessels.

[0054] A dynamic target matching device for unmanned aerial vehicle (UAV) swarm landing includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the method described above.

[0055] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0056] This invention establishes a mathematical model for the allocation problem of UAVs and unmanned surface vessels with the total landing energy consumption as the objective, and solves the model using an improved auction algorithm. This auction algorithm uses x... ij x is the decision variable. ij This represents the matching status between drone i and unmanned surface vessel j. If unmanned surface vessel j is assigned to drone i, then x... ij =1, otherwise x ij =0, and with As the reward for UAV i landing on UAV j, through multiple iterative iterations and continuous updates of each UAV's bid for each UAV, the algorithm ultimately allocates an UAV to each UAV while maximizing the reward, thus minimizing total energy consumption. Furthermore, the improved auction algorithm introduces the variable `near_count` to guide the bid update strategy for each iteration. If `near_count` ≤ 1, where 0 indicates a large difference between the optimal and second-best rewards, and 1 indicates that only the second-best and optimal rewards are close; if `near_count` > 1, it means that the rewards obtained by UAVs when selecting some UAVs are very similar, indicating intense competition. In this case, a larger increase in the bid for the optimal UAV is needed to reduce the number of iterations. Therefore, different price update strategies are selected for different values ​​of `near_count`. In summary, this invention, by using an improved auction algorithm, can quickly allocate suitable UAVs to each UAV and minimize the energy consumption of the UAV swarm, thereby improving task execution efficiency. Attached Figure Description

[0057] Figure 1 This is a flowchart of the steps in a method for dynamically matching landing targets of an unmanned surface vessel swarm in one embodiment.

[0058] Figure 2 This is a schematic diagram illustrating the motion process of a drone flying towards an unmanned surface vessel, according to one embodiment.

[0059] Figure 3 A flowchart illustrating the steps of solving the mathematical model using an improved auction algorithm, as shown in one embodiment;

[0060] Figure 4 A flowchart of the steps for adding task reassignment to a method for dynamic target matching of unmanned aerial vehicle (UAV) swarm landing in one embodiment. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0062] like Figure 1 The diagram shows a flowchart of a method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing in one embodiment, which mainly includes the following steps:

[0063] Step S100: Calculate the energy E consumed by each UAV i to reach each unmanned surface vessel j. ij .

[0064] Eij This represents the energy consumed by drone i to reach unmanned surface vessel j.

[0065] Both drones and unmanned surface vessels (USVs) are equipped with GPS, inertial navigation, and wireless communication modules. A communication network is established between the USV and USV swarms based on the TCP / IP protocol and Socket communication. The drones can obtain the current position p of all USVs. j and speed v j .

[0066] To minimize energy consumption during the flight of the drone toward the unmanned surface vessel, the drone should move in a straight line and go through three stages: acceleration, constant speed, and deceleration. Assume that the magnitude of the drone's acceleration during the acceleration and deceleration stages is constant at a.

[0067] like Figure 2 As shown, the initial velocity of the drone is v. i It accelerates uniformly to v at time t1. m and with v m It moves at a constant speed until time t2, and then decelerates uniformly to v at time t3. j The distances traversed in the three stages are s1, l, and s2, respectively. The energy consumption mainly includes three parts: hovering energy consumption, kinetic energy change energy consumption, and energy consumption due to overcoming drag. Considering only multi-rotor UAVs, when the UAV maintains a constant speed, its fuselage tilts forward, weakening the vortex ring effect, resulting in energy consumption similar to that in hovering. Each UAV's energy consumption depends on its position p. i and speed v i Calculate the energy E consumed to reach the unmanned surface vessel j. ij for:

[0068]

[0069] Where, η tot For the overall efficiency of the drone's motor and propeller, E hov For energy consumption during hovering, E acc Energy is consumed for kinetic energy change, E drag To overcome resistance, energy is consumed.

[0070] Considering that there may be two situations between the matched drones and unmanned surface vessels, namely, moving in the same direction and moving in opposite directions, in one embodiment, different formulas are used to calculate energy consumption for different situations.

[0071] Detailed introduction is as follows:

[0072] By the laws of kinematics and the law of conservation of energy, let t tot =t3-t0, where v i v m v j Both refer to rate.

[0073] When the drone and the unmanned surface vessel move in the same direction, the following relationship exists:

[0074]

[0075]

[0076]

[0077]

[0078] s1+s2+l=d+v j (t3-t0)

[0079]

[0080] in, For hovering power, Given the air drag coefficient, we get:

[0081]

[0082]

[0083]

[0084]

[0085] When a drone and an unmanned surface vessel move toward each other, the following relationship exists:

[0086]

[0087]

[0088]

[0089]

[0090] s1+s2+l=dv j (t3-t0)

[0091]

[0092] The combined results are:

[0093]

[0094]

[0095]

[0096]

[0097] Where d=||p i -p j || represents the initial distance between the drone and the unmanned surface vessel, m is the mass of the drone, g is the acceleration due to gravity, ρ is the air density, and C is the mass of the unmanned surface vessel. D Let A be the drag coefficient, and A be the coverage area when the UAV propeller is rotating. eff This refers to the effective windward area of ​​the drone.

[0098] Among them, v i v j Given the quantities, it is easy to see that the drone's energy consumption E ij It's about speed v m The function.

[0099] In one embodiment, it can be achieved by adjusting E ij Taking the first derivative yields an equation for the uniform flight speed of the UAV. Solving this equation provides the optimal uniform flight speed.

[0100] Optimal constant speed of flight The calculation equation is as follows:

[0101] When the drone and the unmanned surface vessel are moving in the same direction:

[0102]

[0103]

[0104] When the drone and the unmanned surface vessel are moving toward each other:

[0105]

[0106]

[0107] The optimal constant speed of flight Substitute these values ​​into the energy consumption equation above to calculate the optimal energy consumption for each step.

[0108] Step S200: Establish a mathematical model for the allocation problem of UAVs and unmanned surface vessels with the goal of minimizing the total energy consumption of UAV swarm landing.

[0109] Assuming there are M drones and N unmanned surface vessels (USVs) performing landing missions, with N ≥ M, each drone needs to be assigned to one USV for landing, and each USV can carry at most one drone. The mathematical model for the dynamic matching problem of landing targets is as follows:

[0110]

[0111] Where:

[0112]

[0113] Where, x ij Let x be a 0-1 decision variable. If drone i chooses unmanned surface vessel j to land, then x ij =1, otherwise x ij =0.

[0114] Step S300: Solve the mathematical model using an auction algorithm to determine the allocation.

[0115] This invention is based on energy E ij Measure the reward r for drone i landing on unmanned surface vessel j ij , return r ij With energy E ij Negative correlation, required energy E ij The larger the value, the greater the return r. ij The smaller the value, the less energy E is required. ij The smaller the value, the higher the return r. ij The larger the value, the greater the total reward can be achieved by allocating unmanned surface vessels to each drone through an auction algorithm, thus minimizing total energy consumption.

[0116] Because the auction algorithm uses the reward r ij We can first calculate the return r for each drone bidding for each unmanned surface vessel. ij In one embodiment, the reward r can be... ij With energy E ij The relationship is set as follows:

[0117]

[0118] Where ξ is a constant greater than zero, and this constant needs to be determined according to E ij Adjusting the order of magnitude, when E ij When the unit is kJ, ξ can be taken as 10. 3 .

[0119] like Figure 3 As shown, the execution process of the auction algorithm specifically includes:

[0120] Step S310: Select drone i and determine whether the conditions are met. If yes, proceed to step S320; otherwise, proceed to step S350.

[0121] Understandably, when the algorithm starts running, it first performs an initialization action to initialize the drone's decision variables x. i (0), Auction price y i (0) is a zero vector of length N.

[0122] if This indicates that drone i has not yet won the bid for the unmanned surface vessel (USV). Therefore, it is necessary to increase drone i's bid for the USV with the highest potential return. Thus, the process proceeds to step S320 to perform a price adjustment. This indicates that drone i has already won the bid for the unmanned surface vessel, so there is no need to adjust the price further; therefore, the price adjustment action will be skipped.

[0123] Step S320: Calculate the revenue r of drone i bidding for each unmanned surface vessel j. ij -y ij (t).

[0124] Where, r ij In return for the drone i landing on the unmanned surface vessel j, y ij (t) represents the bidding price of UAV i for UAV j in the t-th iteration, and the return r. ij With energy E ij Negative correlation.

[0125] Step S330: Among the positive returns, determine whether |c i,q+1 -c i,q | < ε, where ε is a set threshold value. If it is, the variable near_count is incremented by 1, and the index q is incremented by 1. Step S330 is repeated until |c i,q+1 -c i,q The comparison of elements with ≥ε or positive returns has been completed.

[0126] Among them, c i,q This represents the q-th position of the positive revenue of drone i after sorting from high to low. The initial value of q is 1 for each price update, and the initial value of near_count is 0 for each price update.

[0127] The specific steps are as follows:

[0128] Step S331: Set the matching function:

[0129]

[0130] Here, Γ(·) is a discriminant function, taking the value 1 if the discriminant is true, and 0 otherwise. It can be seen that mat... ij =1 indicates that drone i has a positive return when bidding for unmanned surface vessel j, mat ij =0 indicates that drone i does not have a positive return when bidding for unmanned surface vessel j.

[0131] Step S332: Sort the positive rewards that each drone can obtain from largest to smallest, and update the decision state of drone i.

[0132] The optimal profit of drone i is c. i,1 The unmanned surface vessel corresponding to the optimal benefit is J.i,1 :

[0133]

[0134] in, The positive returns are ranked as k-th. i Positive returns for the position This concludes the calculation of positive returns from the best to the worst.

[0135] The unmanned surface vessel J with the highest positive returns i,1 Assign it to drone i, and update drone i's decision state as follows:

[0136]

[0137] Step S333: Initialize the integer variable near_count to zero, and update the variable near_count according to the proximity of positive drone revenues.

[0138] That is, in c i,1 Under the premise that >0, if:

[0139] abs(c i,q+1 -c i,q )<ε, for q=1,…,k i -1

[0140] The `near_count` function increments by one iteratively. If the above condition is not met, the loop ends, and the number of rewards closest to the optimal reward is `near_count`. Here, `abs(·)` performs the absolute value calculation, and `ε` is a constant that measures the proximity of the rewards. `ε` can be determined based on the reward `r`. ij The order of magnitude determines the range of values; specifically, the order of magnitude of the constant ε is greater than that of the return r. ij It is two orders of magnitude smaller.

[0141] Step S340: Adjust the value of UAV i to UAV J according to the value of variable near_count. i,1 The bidding.

[0142] If near_count ≤ 1, it means that the difference between the optimal and second-best returns is large, or that only the second-best and optimal returns are close. In this case, the auction price of the drone can be updated as follows:

[0143]

[0144] To prevent optimal return c i,1 and suboptimal return c i,2If the price difference is too small, the auction algorithm will get stuck. It is necessary to set the minimum price increment δ during the drone bidding process. In one embodiment, δ = 1 / M can be taken.

[0145] If near_count > 1, it means that the rewards obtained by the drone when selecting some unmanned surface vessels are very similar, and the competition is fierce. In this case, it is necessary to increase the bid for the best unmanned surface vessel by a larger margin to reduce the number of iterations.

[0146]

[0147] Wherein, 0 < σ < 1 is an adjustable coefficient, and in one embodiment, σ can be taken as 0.3.

[0148] Step S350: Obtain the latest quotes for each UAV, update the quote for UAV i for each UAV to the highest existing quote for the corresponding UAV, when UAV i quotes for UAV J i,1 When the current price is not the highest price, update Otherwise, maintain

[0149] The drone swarm communicates with each other, transmits bidding price lists, and obtains the current status of the drone swarm regarding the unmanned surface vessel mission point J. i,1 The highest bidder for the drone Since the drone will adjust its price for each unmanned surface vessel (USV) to the highest price among all USVs after communication, identical prices may occur. Therefore, the price will only be adjusted if another USV offers a higher price than drone i. Update to the UAV number corresponding to the higher bid. If all other UAV bids are less than or equal to the bid for UAV i, then... For i:

[0150]

[0151] In the preceding steps, if the selected drone i failed to win the bid for the unmanned surface vessel J with the optimal profit in the previous round of bidding due to its low bid, i,1 Then, by raising the unmanned surface vessel J through steps S320 to S340, i,1 The quoted price, however, after the current drone price increases, it may become the highest among all drones for unmanned surface vessels. i,1 The highest bidder successfully won the auction for the unmanned surface vessel J. i,1 ,at this time, maintain It's also possible that they didn't win the bid for the unmanned surface vessel J because it wasn't the highest bidder. i,1 ,at this time, renew

[0152] At the same time, after receiving the quotes from each drone for all unmanned surface vessels, the highest quote for each unmanned surface vessel can be determined, and the current quote from drone i for each unmanned surface vessel can be adjusted to the highest quote for each unmanned surface vessel.

[0153] Specifically, drone i will update the price quoted for each unmanned surface vessel to the highest price in the entire drone swarm:

[0154]

[0155] if This indicates that the drone i does not possess information about the unmanned surface vessel's mission point J. i,1 If the highest bid is received, then drone i will lose the mission, and the drone's decision status will be updated:

[0156]

[0157] Step S360: Update drone i and proceed to step S310 until all drones have been updated.

[0158] When the number of drones is M, steps S310 to S350 need to be repeated M times.

[0159] Step S370: Determine whether each drone has been assigned to a different unmanned surface vessel. If yes, end the iteration; otherwise, proceed to step S310 to continue the iteration.

[0160] if and This means that each drone has been assigned to a different unmanned surface vessel, and the assignment process ends. Otherwise, repeat steps S310 to S360 until the above conditions are met or the number of iterations reaches the upper limit.

[0161] In one embodiment, such as Figure 4 As shown, after step S300, the following steps are also included:

[0162] Step S400: Control the drone to fly towards the matched unmanned surface vessel, and after the assignment is completed and the interval Δt is reached, determine whether there are any drones that have not yet successfully landed. If so, proceed to step S100 to redistribute tasks to the drones that have not yet successfully landed and the available unmanned surface vessels, until all drones have successfully landed.

[0163] Based on the three-dimensional position error and velocity information between the UAV and the matched unmanned surface vessel (USV), a PID controller is designed to control the UAV to fly towards the USV and complete the landing. Since the positions and velocities of both the UAV and USV swarm are dynamically changing, and the UAVs have high flight speed and strong maneuverability, real-time task reallocation is essential. After one allocation is completed, the update time step for the task allocation is dynamically determined based on the UAVs' maneuverability.

[0164]

[0165] Among them, L UAV D is the wheelbase of the drone. UAV The diameter of the drone's propeller blades, Let be the optimal constant speed of drone i obtained in the previous assignment.

[0166] If the difference between the current time t_now and the last time the allocation was completed t_end is greater than Δt, it means that the task allocation state needs to be recalculated. In other words, the position and speed information of the UAV swarm at the current time should be obtained, and energy consumption calculation and allocation optimization should be performed again to obtain the optimal allocation that is suitable for the current state.

[0167] The time required for drones to perform landing missions varies, therefore, drones that have already landed need to be removed from the list of drones to be assigned. A drone is considered to have successfully landed when the distance between its center of mass and the center of the landing platform on its matching unmanned surface vessel is less than a sufficiently small constant ψ, and its information is removed from the list i = 1, ..., M. When the list of drones to be assigned is empty, it means that all drones have landed, and the overall mission ends.

[0168] Accordingly, the present invention also relates to a dynamic target matching device for unmanned aerial vehicle (UAV) swarm landing, characterized in that it includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method. This device can be a computing device such as an industrial control computer, desktop computer, laptop, handheld computer, or cloud server.

[0169] In summary, by employing an improved auction algorithm, iterating multiple times, and optimizing the price update strategy, this invention can quickly allocate suitable unmanned surface vessels to each drone and minimize the energy consumption of the drone swarm, thereby improving mission execution efficiency.

[0170] Those skilled in the art will readily understand that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing, characterized in that, include: Step S1: Calculate the energy E consumed by each UAV i to reach each unmanned surface vessel j. ij ; Step S2: Establish a mathematical model for the allocation problem of UAVs and unmanned surface vessels with the objective of minimizing the total energy consumption of UAV swarm landing; Step S3: Solve the mathematical model using an auction algorithm to determine the allocation, including performing: Step S31: Determine whether the selected drone i satisfies the following conditions. If yes, proceed to step S32; otherwise, proceed to step S35. ij As a decision variable, if unmanned surface vessel j is assigned to unmanned aerial vehicle i, then x ij =1, otherwise x ij =0, where N is the number of unmanned surface vessels; Step S32: Calculate the revenue r of drone i bidding for each unmanned surface vessel j. ij -y ij (t), where r ij In return for the drone i landing on the unmanned surface vessel j, y ij (t) represents the bidding price of UAV i for UAV j in the t-th iteration, and the return r. ij With energy E ij Negative correlation; Step S33: Among the positive returns, determine whether |c i,q+1 -c i,q If |<ε, where ε is a set threshold, then the variable near_count is incremented by 1, and the index q is incremented by 1. Step S33 is repeated until |c i,q+1 -c i,q The comparison of elements in the range of |≥ε or positive returns has been completed, where c i,q This represents the q-th position of the positive revenue of drone i after sorting from high to low, with the initial value of q being 1; Step S34: When near_count≤1, update the unmanned surface vessel J. i,1 The price is When near_count > 1, update the unmanned surface vessel J. i,1 The price is Among them, J i,1 The highest-profit unmanned surface vessel in the bidding for drones. For the unmanned surface vessel J i,1 The current price, σ is the set adjustable coefficient, and δ is the set minimum price increment; Step S35: Obtain the latest quotes for each UAV, update the quote for UAV i for each UAV to the highest existing quote for the corresponding UAV, when UAV i quotes for UAV J i,1 When the current price is not the highest price, update Otherwise, maintain Step S36: Update drone i and proceed to step S31 until all drones have been updated; Step S37: Determine whether each UAV has been assigned to a different unmanned surface vessel. If yes, end the iteration; otherwise, proceed to step S31 to continue the iteration.

2. The method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing as described in claim 1, characterized in that, Following step S3, the following is also included: Step S4: Control the drone to fly towards the matched unmanned surface vessel, and after the assignment is completed and the interval Δt is reached, determine whether there are any drones that have not yet successfully landed. If so, jump to step S1 to redistribute the tasks of the drones that have not yet successfully landed and the available unmanned surface vessels until all drones have successfully landed.

3. The method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing as described in claim 2, characterized in that, The formula for calculating the interval time Δt is: Among them, L UAV D is the wheelbase of the drone. UAV The diameter of the drone's propeller blades, Let be the optimal constant speed of drone i obtained in the previous assignment.

4. The method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing as described in claim 1, characterized in that, Return r ij The calculation formula is: Where ξ is a constant greater than zero.

5. The method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing as described in claim 1, characterized in that, The minimum price increment is δ = 1 / M, where M is the number of drones.

6. The method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing as described in claim 1, characterized in that, In step S1, the optimal energy E ij The calculation formula is: Where, η tot For the overall efficiency of the drone's motor and propeller, Eh ov E consumes energy for drone hovering. acc E consumes energy due to changes in the kinetic energy of the drone. drag Energy is consumed to help drones overcome air resistance.

7. The method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing as described in claim 6, characterized in that, Drone hovering energy consumption Eh ov Energy consumption E due to changes in the kinetic energy of the drone acc And the energy consumed by drones to overcome air resistance E drag The calculation formulas are as follows: When the drone i and the unmanned surface vessel j move in the same direction When drone i and unmanned surface vessel j move toward each other... Where m is the mass of the drone, g is the acceleration due to gravity, ρ is the air density, and C is the mass of the drone. D Let A be the drag coefficient, and A be the coverage area when the UAV propeller is rotating. eff Let d be the effective windward area of ​​the UAV, d be the initial distance between the UAV and the unmanned surface vessel, and v be the effective windward area of ​​the UAV. i v is the current flight speed of the drone. j v is the current speed of the unmanned surface vessel. m Let t be the uniform flight speed of the UAV during its uniform flight phase. tot Let be the time required for the drone to land on the unmanned surface vessel, and 'a' be the flight acceleration.

8. The method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing as described in claim 7, characterized in that, Solve for the optimal uniform flight speed And order in, When the drone and the unmanned surface vessel are moving in the same direction, the optimal uniform flight speed is... The solution formula is: When the drone and the unmanned surface vessel are moving toward each other, the optimal uniform flight speed is... The solution formula is:

9. The method for dynamic target matching for unmanned aerial vehicle (UAV) swarm landing as described in claim 1, characterized in that, In step S2, the mathematical model is expressed as follows: Where, x ij Let x be the decision variable. If drone i chooses unmanned surface vessel j to land, then x ij =1, otherwise x ij =0, M is the number of drones, and N is the number of unmanned surface vessels.

10. A dynamic target matching device for unmanned aerial vehicle (UAV) swarm landing, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the steps of the method as described in any one of claims 1 to 9.

Images