Multi-target unmanned ship hunting method and system

Abstract

The invention relates to a multi-target unmanned ship hunting method and system. The hunting method comprises the following steps: selecting a proper hunting unmanned ship for a target unmanned ship as a bid invitation person according to information of the hunting unmanned ship and the target unmanned ship; the bid invitation person captures unmanned ships around to issue information for bid invitation; the bidder bids, and calculates the distance between the bidder and the target unmanned ship and the wrap angle of the target unmanned ship; the bid invitation selects an optimal bidder combination as a distribution result according to the distance and the wrap angle; after all the target unmanned ships are distributed, the adjacent target unmanned ships carry out local exchange on the surrounding unmanned ships; performing hunting, wherein the hunting is divided into cooperative approaching and cooperative hunting; when the difference value between the distance between the hunting unmanned ship and the target unmanned ship and the surrounding radius Rs is smaller than a given threshold value epsilon and the total surrounding angle is larger than or equal to 360 degrees, hunting succeeds; and if a target unmanned ship is caught, task redistribution is carried out, and if all the target unmanned ships are caught successfully, the caught task is ended. According to the multi-target unmanned ship hunting method provided by the invention, task allocation of multiple unmanned ships is carried out through an improved contract net algorithm, the speed is increased, and the result is better; and meanwhile, the improved particle swarm optimization is combined with a staged hunting method to realize hunting control of multiple unmanned ships, the hunting success rate is improved, and the method is stable and good in adaptability.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned boat cluster capture, and in particular to a multi-target unmanned boat capture method and system. Background Art

[0002] Unmanned swarm capture refers to the mutual coordination of unmanned group systems such as unmanned vehicles, unmanned boats, drones and robots to jointly complete the capture of single or multiple targets. Compared with a single system, the unmanned group system has stronger robustness, scalability and flexibility, and can complete large-scale complex tasks more quickly.

[0003] In practical applications, the group roundup problem often involves the coordinated roundup of multiple targets. At present, there are many related works such as group task allocation and cluster roundup. However, the existing algorithms have some obvious shortcomings, such as the task allocation is mainly distributed and there are disadvantages such as large communication volume and poor results; the success rate of cluster roundup for unmanned boats with high-speed moving targets is low, and the motion constraints are rarely considered and obstacles are simply avoided. Summary of the invention

[0004] In view of some or all of the problems in the prior art, the present invention provides a multi-target unmanned boat capture method, which comprises the following steps:

[0005] According to the information of the capture unmanned boat and the target unmanned boat, a suitable capture unmanned boat is selected as the bidder for the target unmanned boat;

[0006] The bidder releases information to surrounding unmanned capture boats for bidding, and the unmanned capture boats that receive the message are bidders;

[0007] The bidder makes a bid and calculates the distance between the bidder and the target unmanned boat and the encirclement angle of the target unmanned boat;

[0008] The tenderer selects the best bidder combination according to the distance and the encirclement angle as the allocation result, and notifies the corresponding bidders;

[0009] After all target unmanned boats are allocated, adjacent target unmanned boats partially exchange the encirclement unmanned boats to obtain the optimized allocation result;

[0010] Execute the roundup, which is divided into coordinated approach and coordinated roundup; when the distance between the roundup unmanned boat and the target unmanned boat is greater than a set threshold, the coordinated approach is executed; when the distance between the roundup unmanned boat and the target unmanned boat is less than the set threshold, the coordinated roundup is executed, and the particle swarm algorithm is used to perform speed solution;

[0011] When the difference between the distance between the encircling unmanned boat and the target unmanned boat and the encirclement radius Rs is less than the given threshold ε and the total encirclement angle is greater than or equal to 360°, the encirclement is successful;

[0012] If a target unmanned boat is surrounded, the task will be redistributed. If all target unmanned boats are successfully surrounded, the surrounding mission will be completed.

[0013] Further, according to the positions of the capture unmanned boat and the target unmanned boat, the selection of a suitable capture unmanned boat as a tenderer for the target unmanned boat includes:

[0014] All the captured unmanned boats are represented by the set U.

[0015] U={U1,U2,...,U i ,...U n}

[0016] in,

[0017] in, is the captured unmanned boat U at time t+1 i location, is the captured unmanned boat U at time t+1 i speed, The number of unmanned boats captured at time t is i location;

[0018] All target unmanned boats are represented by set T,

[0019] T={T1,T2,...,T j ,...T m}

[0020] in,

[0021] in, is the target unmanned boat T at time t+1 j location, is the target unmanned boat T at time t+1 j speed, The number of unmanned boats captured at time t is T j location;

[0022] Target unmanned boat T j Use set D to select the smallest distance unmanned boat as the bidder.

[0023] D = min{d 1j ,d 2j ,...,d nj}

[0024] Among them, d nj Target unmanned boat T j To capture the unmanned boat U n distance.

[0025] Furthermore, the bidder releases information to surrounding unmanned boats for bidding, and the bidder participates in the bidding including:

[0026] The bidder releases information to the surroundings, with the initial number of times being 0. Each unmanned capture boat that receives the information will increase the number by 1. When the communication number limit is reached, no more communication will be performed. The unmanned capture boats that receive the message are represented by a set M.

[0027] M={U1,U2,...,U z}

[0028] The number of communications is limited to 3 times.

[0029] Furthermore, the bidder makes a bid and calculates the distance between the bidder and the target unmanned boat and the encirclement angle of the target unmanned boat, including:

[0030] When there is no obstacle between the bidder and the target unmanned boat, the bidder calculates the distance d between the bidder and the target unmanned boat.

[0031]

[0032] Wherein, ΔX and ΔY are the horizontal coordinate difference and vertical coordinate difference between the bidder and the target unmanned boat respectively;

[0033] When there is an obstacle between the bidder and the target unmanned boat, the bidder calculates the distance d between the bidder and the target unmanned boat.

[0034]

[0035] Wherein, ΔX and ΔY are the horizontal coordinate difference and vertical coordinate difference between the bidder and the target unmanned boat, respectively, and L is the maximum size of the obstacle;

[0036] The bidder calculates the relative angle α between itself and the target unmanned boat,

[0037] α=tanh- 1 (ΔY / ΔX)

[0038] The bidder calculates the enveloping angle S between itself and the target unmanned boat,

[0039] S=sin- 1 (Vu / Vt)

[0040] Wherein, Vu is the speed of the bidder, and Vt is the speed of the target unmanned boat.

[0041] Furthermore, the tenderer selects the best bidder combination as the allocation result according to the distance and the encirclement angle, and notifies the corresponding bidders including:

[0042] The best bidder combination satisfies the following formula:

[0043] ∑S≥360°

[0044] Among them, ∑S is the sum of the encirclement angles S between the bidder and the target unmanned boat in the best bidder combination;

[0045] The best bidder combination satisfies the minimum target unmanned boat function, and the target unmanned boat function is,

[0046] f=min(w1D sum +w2(-max|Δα|))

[0047] Among them, min is the minimum value, D sum is the sum of the distances d between the bidders and the target unmanned boat in the best bidder combination, max|Δα| is the maximum value of the difference in angles between the bidders relative to the target unmanned boat, w1 and w2 are weights, and satisfy w1+w2=1.

[0048] Furthermore, after all target unmanned boats are allocated, adjacent target unmanned boats partially exchange the encirclement unmanned boats, and the optimized allocation results include:

[0049] The capture unmanned boats assigned to each target unmanned boat are exchanged with each other, and the change Δf of the total target unmanned boat function before and after the exchange is calculated ij ,

[0050]

[0051] Among them, f i 、f j To capture the unmanned boat U i , U j The target unmanned boat function value before the exchange, To capture the unmanned boat U i , U j The target unmanned boat function value after exchange, Δf ij Satisfy Δf ij >0;

[0052] The difference of the exchange is represented by the set E,

[0053] E={e1,e2,...e e}

[0054] Among them, e i ={Δf ij ,i,j}, select the largest difference max{e i} as the optimized allocation result.

[0055] Further, the encirclement is performed, and the encirclement is divided into coordinated approach and coordinated encirclement; when the distance between the encirclement unmanned boat and the target unmanned boat is greater than the set threshold, the coordinated approach is performed; when the distance between the encirclement unmanned boat and the target unmanned boat is less than the set threshold, the coordinated encirclement is performed, and the particle swarm algorithm is used to perform the speed solution, including:

[0056] The threshold value D th satisfy,

[0057] D detect >D th >D alert

[0058] Among them, D detect To capture the unmanned boat detection range, D alert It is the warning range of the target unmanned boat.

[0059] When the total encirclement angle G satisfies the following formula, the capture unmanned boats shrink in coordination:

[0060] G>TH

[0061] Among them, TH is the encirclement angle threshold, G = ∑S-Gr, ∑S ≥ 360°, Gr = ∑r, ∑S is the sum of the encirclement angles S of the capture unmanned boat and the target unmanned boat, Gr is the total overlap angle, and r is the overlap angle;

[0062] The overlap angle r is expressed as follows:

[0063]

[0064] Among them, α i To capture the unmanned boat U i The relative angle with the target unmanned boat, α j To capture the unmanned boat U j The relative angle with the target unmanned boat, S i To capture the unmanned boat U i The encirclement angle with the target unmanned boat, S j To capture the unmanned boat U j The encirclement angle with the target unmanned boat;

[0065] In the collaborative capture phase, when G>TH, the capture unmanned boats shrink collaboratively. During collaborative shrinkage, the particle swarm algorithm is used to perform speed calculation. The particle swarm algorithm is as follows:

[0066]

[0067]

[0068] in, is the unmanned boat U captured after k+1 iterations i speed, is the number of unmanned boats captured after k iterations i speed, ω represents the inertia weight, c1 and c2 represent the individual learning factor and the group learning factor, r1 and r2 are random numbers between 0 and 1, is the unmanned boat U captured after k+1 iterations i location, is the number of unmanned boats captured after k iterations i location, is the number of unmanned boats captured after k iterations i The historical optimal fitness value of The historical optimal fitness value of the group of unmanned boats for encirclement;

[0069] The fitness value p is calculated using the following formula:

[0070] p=a1(Gr-r min )+a2∑(D alert -d i ),G<TH,a1> a2

[0071] p=a1(Gr-r min )+a2∑(Rs-d i ),G≥TH,a1 <a2

[0072] Among them, r min =∑S-360°, a1 and a2 are weights, satisfying a1+a2=1, Rs is the enclosing radius; d i To capture the unmanned boat U i The distance to the target unmanned boat.

[0073] Furthermore, when the difference between the distance between the encircling unmanned boat and the target unmanned boat and the encirclement radius Rs is less than a given threshold ε and the total encirclement angle is greater than or equal to 360°, the encirclement is successful including:

[0074] The conditions for a successful roundup are:

[0075] ∑(d i -Rs)≤ε

[0076] G≥360°

[0077] Among them, d i To capture the unmanned boat U i The distance to the target unmanned boat, Rs is the encirclement radius, ε is the given threshold, and G is the total encirclement angle.

[0078] The present invention also provides a multi-target unmanned boat capture system for the multi-target unmanned boat capture method, the system comprising the following modules:

[0079] The tenderer selection module is configured to select a suitable capture unmanned boat as a tenderer for the target unmanned boat according to the information of the capture unmanned boat and the target unmanned boat;

[0080] The bidding module is configured so that the bidder releases information to surrounding unmanned capture boats for bidding, and the unmanned capture boats that receive the message are bidders;

[0081] A bidding module is configured to calculate the distance between the bidder and the target unmanned boat and the encirclement angle of the target unmanned boat for the bidder to bid;

[0082] A bidder combination determination module is configured to select the best bidder combination as the allocation result according to the distance and the encirclement angle by the tenderer, and notify the corresponding bidders;

[0083] The optimization allocation module is configured such that after all target unmanned boats are allocated, adjacent target unmanned boats partially exchange the encirclement unmanned boats to obtain an optimized allocation result;

[0084] The encirclement and capture module is configured to execute the encirclement and capture, and the encirclement and capture are divided into coordinated approach and coordinated encirclement and capture; when the distance between the encirclement and capture unmanned boat and the target unmanned boat is greater than a set threshold, the coordinated approach and capture are executed; when the distance between the encirclement and capture unmanned boat and the target unmanned boat is less than the set threshold, the coordinated encirclement and capture are executed, and the particle swarm algorithm is used to perform the speed solution;

[0085] The capture success judgment module is configured to capture successfully when the difference between the distance between the capture unmanned boat and the target unmanned boat and the encirclement radius Rs is less than a given threshold ε and the total encirclement angle is greater than or equal to 360°;

[0086] The capture mission completion judgment module is configured to reallocate tasks if a target unmanned boat is captured, and the capture mission is completed if all target unmanned boats are successfully captured.

[0087] The present invention also provides a computer system, comprising:

[0088] a processor configured to execute machine-readable instructions;

[0089] A graphics card having an artificial intelligence chip configured to train a multi-target unmanned boat capture method; and

[0090] The memory is configured to store machine-readable instructions, which, when executed by the processor and / or the graphics card, perform the steps of the multi-target unmanned boat capture method.

[0091] Compared with the prior art, the present invention has the following advantages:

[0092] 1. The multi-target unmanned boat encirclement method proposed in the present invention uses an improved contract network algorithm to perform multi-unmanned boat task allocation, reduces the communication volume by limiting the number of times the tenderer communicates with the surrounding areas, and optimizes the allocation results and improves the quality of the solution by exchanging encirclement unmanned boats between adjacent target unmanned boats. Compared with the traditional contract network algorithm, it speeds up the process and achieves better results.

[0093] 2. The multi-target unmanned boat capture method proposed in the present invention regards obstacles as static unmanned boats with a certain encirclement angle to the target unmanned boat. At the same time, the improved particle swarm optimization algorithm (PSO) is combined with the staged capture method to realize the capture control of multiple unmanned boats and consider motion constraints. By improving the population initialization method and adopting different fitness functions for different stages, the capture success rate is improved. The method is stable and has good adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] To further illustrate the above and other advantages and features of various embodiments of the present invention, a more specific description of various embodiments of the present invention will be presented with reference to the accompanying drawings. It will be understood that these drawings only depict typical embodiments of the present invention and are therefore not to be considered as limiting the scope thereof. In the accompanying drawings, for clarity, identical or corresponding parts will be represented by identical or similar reference numerals.

[0095] Figure 1 A schematic diagram showing a process flow of a multi-target unmanned boat capture method according to an embodiment of the present invention;

[0096] Figure 2 A schematic diagram of tenderer selection according to an embodiment of the present invention is shown;

[0097] Figure 3 A schematic diagram showing the communication times limitation according to an embodiment of the present invention is shown;

[0098] Figure 4 A schematic diagram of relative angles and enclosing angles according to an embodiment of the present invention is shown;

[0099] Figure 5 A schematic diagram of bidder combination selection according to an embodiment of the present invention is shown;

[0100] Figure 6 A schematic diagram of exchanging unmanned capture boats according to an embodiment of the present invention is shown;

[0101] Figure 7 A schematic diagram of the phased capture of an embodiment of the present invention is shown;

[0102] Figure 8 A schematic diagram of overlap angle according to an embodiment of the present invention is shown;

[0103] Fig. 9A schematic diagram showing the total enclosing angle overlap of an embodiment of the present invention is shown; and

[0104] Fig.10 A schematic diagram of a multi-target unmanned boat capture system according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0105] In the following description, the present invention is described with reference to various embodiments. However, those skilled in the art will recognize that various embodiments may be implemented without one or more specific details or with other replacement and / or additional methods or components. In other cases, well-known structures or operations are not shown or described in detail to avoid obscuring the inventive point of the present invention. Similarly, for the purpose of explanation, specific numbers and configurations are set forth to provide a comprehensive understanding of embodiments of the present invention. However, the present invention is not limited to these specific details.

[0106] In this specification, reference to "one embodiment" or "the embodiment" means that a particular feature, structure, or characteristic described in conjunction with the embodiment is included in at least one embodiment of the present invention. The phrase "in one embodiment" appearing in various places in this specification does not necessarily all refer to the same embodiment.

[0107] It should be noted that the embodiments of the present invention describe the method steps in a specific order, but this is only for the purpose of illustrating the specific embodiment, rather than limiting the order of the steps. On the contrary, in different embodiments of the present invention, the order of the steps can be adjusted according to actual needs.

[0108] In the present invention, each module of the system according to the present invention can be implemented using software, hardware, firmware or a combination thereof. When the module is implemented using software, the function of the module can be implemented by a computer program flow, for example, the module can be implemented by a code segment (such as a code segment of a language such as C, C++) stored in a storage device (such as a hard disk, a memory, etc.), wherein the corresponding function of the module can be implemented when the code segment is executed by a processor. When the module is implemented using hardware, the function of the module can be implemented by setting a corresponding hardware structure, for example, the function of the module can be implemented by hardware programming a programmable device such as a field programmable gate array (FPGA), or the function of the module can be implemented by designing an application-specific integrated circuit (ASIC) including electronic devices such as a plurality of transistors, resistors and capacitors. When the module is implemented using firmware, the function of the module can be written into a read-only memory such as an EPROM or EEPROM of the device in the form of a program code, and the corresponding function of the module can be implemented when the program code is executed by a processor. In addition, certain functions of the module may need to be implemented by separate hardware or by collaboration with the hardware, for example, the detection function is implemented by corresponding sensors (such as proximity sensors, acceleration sensors, gyroscopes, etc.), the signal transmission function is implemented by corresponding communication devices (such as Bluetooth devices, infrared communication devices, baseband communication devices, Wi-Fi communication devices, etc.), the output function is implemented by corresponding output devices (such as displays, speakers, etc.), and so on.

[0109] There have been many studies on the task allocation of multiple unmanned boats, and the task allocation methods are mainly divided into centralized task allocation and distributed task allocation. Centralized task allocation collects all information from a central unit, has a simple structure, good globality, and can obtain high-quality static allocation solutions, but has high communication requirements and is prone to single point failure problems. Distributed task allocation does not need to rely on a central unit, and the unmanned boats are independent of each other, so it is widely used in dynamic uncertainty environments. However, most of the existing studies are static task allocation, and there is less research on dynamic environments. Contract network, as a distributed task allocation method, has the advantages of robustness and flexibility, and the amount of computation does not increase with the increase of the problem scale. It is the most widely used method today. However, the large amount of communication is its main disadvantage, and a large amount of communication time may lead to untimely task allocation. The present invention proposes an improved contract network algorithm to perform multi-unmanned boat task allocation, which reduces the communication volume by limiting the number of times the tenderer communicates with the surrounding area, and optimizes the allocation results and improves the quality of the solution by exchanging unmanned boats between adjacent target unmanned boats. Compared with the traditional contract network algorithm, it speeds up and has better results.

[0110] Regarding the research on the capture of multiple unmanned boats, the previous research environment was relatively simple and the capture efficiency was reduced when facing fast target unmanned boats. There were fewer studies in complex environments, and the capture in an obstacle environment was usually just a simple obstacle avoidance, while obstacles were also obstacles to the target unmanned boats. The multi-target unmanned boat capture method proposed in this paper regards obstacles as static unmanned boats with a certain encirclement angle to the target unmanned boat. At the same time, the improved particle swarm algorithm combined with the staged capture method is used to realize the capture control of multiple unmanned boats and consider motion constraints. By improving the population initialization method, different fitness functions are adopted for different stages, which improves the capture success rate. The method is stable and has good adaptability.

[0111] The multi-target unmanned boat encirclement method proposed in this paper is divided into two parts: multi-target unmanned boat task allocation and single-target unmanned boat encirclement. First, a suitable unmanned boat is assigned to each target unmanned boat and then the assigned target unmanned boat is encircled.

[0112] Figure 1 The following is a schematic diagram of a multi-target unmanned boat capture method according to an embodiment of the present invention. Figure 1 , the multi-target unmanned boat capture method proposed by the present invention is described. In one embodiment of the present invention, the multi-target unmanned boat capture method can be executed by a computer. Figure 1 As shown, the multi-target unmanned boat capture method includes the following steps:

[0113] First, according to the information of the capture unmanned boat and the target unmanned boat, a suitable capture unmanned boat is selected as the bidder for the target unmanned boat. In the initial stage, each capture unmanned boat has no target unmanned boat assigned. In one embodiment of the present invention, the number of capture unmanned boats and the number of target unmanned boats meet a minimum ratio of 3:1.

[0114] All the captured unmanned boats are represented by the set U.

[0115] U={U1,U2,...,U i ,...U n}

[0116] in, 1,2,...,n are the numbers of the unmanned boats for capture.

[0117] in, is the captured unmanned boat U at time t+1 i location, is the captured unmanned boat U at time t+1 i speed, The number of unmanned boats captured at time t is i location.

[0118] All target unmanned boats are represented by set T,

[0119] T={T1,T2,...,T j ,...T m}

[0120] in, 1,2,...,m are the target unmanned boat numbers.

[0121] in, is the target unmanned boat T at time t+1 j location, is the target unmanned boat T at time t+1 j speed, The number of unmanned boats captured at time t is T j location.

[0122] Target unmanned boat T j Use set D to select the smallest distance unmanned boat as the bidder.

[0123] D = min{d 1j ,d 2j ,...,d nj}

[0124] Among them, d nj Target unmanned boat T j To capture the unmanned boat U n distance. Figure 2 FIG. 2 shows a schematic diagram of the selection of tenderers according to an embodiment of the present invention. Figure 2 As shown, the capture unmanned boat with the smallest distance to the target unmanned boat is selected as its bidder, and the selected unmanned boat cannot be selected again.

[0125] Next, the bidder releases information to the surrounding unmanned capture boats for bidding, and the unmanned capture boats that receive the message are bidders.

[0126] Figure 3 FIG. 2 shows a schematic diagram of limiting the number of communications according to an embodiment of the present invention. Figure 3 As shown, the bidder releases information to the surroundings, the initial number is 0, and each capture unmanned boat that receives the information will increase the number by 1. When the communication number limit is reached, no more communication will be performed. The capture unmanned boats that receive the message are represented by the set M.

[0127] M={U1,U2,...,U z}

[0128] In one embodiment of the present invention, the number of communications is limited to 3. z is the number of unmanned boat combinations. In one embodiment of the present invention, z is generally 2-4.

[0129] Next, the bidder makes a bid and calculates the distance between the bidder and the target unmanned boat and the surrounding angle of the target unmanned boat. First, the distance d and the relative angle α between the bidder and the target unmanned boat are calculated, and then the surrounding angle S of the bidder to the target unmanned boat is calculated. If there is an obstacle between the target unmanned boat and the bidder, the distance must be added to the maximum size of the obstacle L. The obstacle can be represented by a set O, O = {O1, O2, ..., O O},1,2,...,O is the target unmanned boat number.

[0130] When there is no obstacle between the bidder and the target unmanned boat, the bidder calculates the distance d between the bidder and the target unmanned boat.

[0131]

[0132] Wherein, ΔX and ΔY are the horizontal coordinate difference and vertical coordinate difference between the bidder and the target unmanned boat respectively;

[0133] When there is an obstacle between the bidder and the target unmanned boat, the bidder calculates the distance d between the bidder and the target unmanned boat.

[0134]

[0135] Wherein, ΔX and ΔY are the horizontal coordinate difference and vertical coordinate difference between the bidder and the target unmanned boat, respectively, and L is the maximum size of the obstacle.

[0136] The bidder calculates the relative angle α between itself and the target unmanned boat,

[0137] α=tanh- 1 (ΔY / ΔX)

[0138] The bidder calculates the enveloping angle S between itself and the target unmanned boat.

[0139] S=sin- 1 (Vu / Vt)

[0140] Among them, Vu is the speed of the bidder and Vt is the speed of the target unmanned boat.

[0141] Figure 4 FIG. 2 shows a schematic diagram of relative angle and enclosing angle of an embodiment of the present invention. Figure 4 As shown, assuming that the relative angle α is 60° and the encirclement angle S is 50°, the bidder's effective encirclement range of the target unmanned boat is 60°±50°, i.e. 10°-110°.

[0142] Next, the tenderer selects the best bidder combination according to the distance and the encirclement angle as the allocation result, and notifies the corresponding bidders.

[0143] Figure 5FIG. 2 shows a schematic diagram of a bidder combination selection according to an embodiment of the present invention. Figure 5 As shown, the tenderer selects the best bidder combination as the allocation result according to the bidders in the set M and the distance, relative angle and encirclement angle data calculated above. The best bidder combination satisfies the following formula:

[0144] ∑S≥360°

[0145] Among them, ∑S is the sum of the encirclement angles S of the bidders and the target unmanned boat in the best bidder combination.

[0146] The best bidder combination satisfies the minimum target unmanned boat function, and the target unmanned boat function is,

[0147] f=min(w1D sum +w2(-max|Δα|))

[0148] Among them, min is the minimum value, D sum is the sum of the distances d between the bidders in the best bidder combination and the target unmanned boat, max|Δα| is the maximum value of the difference in angles between the bidders relative to the target unmanned boat, w1 and w2 are weights, and w1+w2=1. In one embodiment of the present invention, w1 and w2 are 0.6 and 0.4 respectively.

[0149] Next, after all target unmanned boats are allocated, adjacent target unmanned boats partially exchange the encirclement unmanned boats to obtain the optimized allocation result.

[0150] Figure 6 FIG. 1 shows a schematic diagram of the exchange of unmanned capture boats according to an embodiment of the present invention. Figure 6 As shown, the unmanned boats are exchanged locally, and the capture unmanned boats of the adjacent target unmanned boats are exchanged, and the others are not exchanged. The capture unmanned boats assigned to each target unmanned boat are exchanged with each other, and the change Δf of the total target unmanned boat function before and after the exchange is calculated ij ,

[0151]

[0152] Among them, f i 、f j To capture the unmanned boat U i , U j The target unmanned boat function value before the exchange, To capture the unmanned boat U i , U j The target unmanned boat function value after the exchange. The local exchange needs to satisfy Δf ij Satisfy Δf ij>0, because the target unmanned boat function is essentially to find the minimum value. If the exchange is better than the original, the target unmanned boat function will be smaller. However, considering that there are more than one exchange times for the capture of unmanned boats, the set E is used to represent the difference in exchange.

[0153] E={e1,e2,...e e}

[0154] Among them, e i ={Δf ij ,i,j}, select the largest difference max{e i} as the optimized allocation result.

[0155] The above steps are to use the improved contract network algorithm to perform multi-unmanned boat task allocation. After all adjacent target unmanned boats are exchanged, the multi-unmanned boat task allocation step enters the encirclement step.

[0156] Next, execute the roundup. Figure 7 FIG. 2 shows a schematic diagram of the phased capture of an embodiment of the present invention. Figure 7 As shown in the figure, the roundup is divided into the coordinated approach stage and the coordinated roundup stage. When the distance between the roundup unmanned boat and the target unmanned boat is greater than the set threshold, the coordinated approach is performed. When the target unmanned boat is first discovered, the roundup unmanned boat may be far away from the target unmanned boat. At this time, the roundup unmanned boats collectively approach the target unmanned boat but do not enter the warning range of the target unmanned boat. When the distance between the roundup unmanned boat and the target unmanned boat is less than the set threshold, the coordinated roundup is performed.

[0157] The threshold value D th satisfy,

[0158] D detect >D th >D alert

[0159] Among them, D detect To capture the unmanned boat detection range, D alert It is the warning range of the target unmanned boat.

[0160] In the round-up phase, the particle swarm algorithm is combined with the Apollo circle to solve the velocity.

[0161] In the Apollo circle, each capture unmanned boat has an encirclement angle S to the target unmanned boat, and there may be overlaps between them, such as Figure 8 The total encirclement angle is the sum of the encirclement angles minus Gr. When the total encirclement angle G satisfies the following formula, the capture unmanned boats shrink in coordination:

[0162] G>TH

[0163] Wherein, TH is the surrounding angle threshold, G = ∑S - Gr, ∑S ≥ 360°, Gr = ∑r, ∑S is the sum of the surrounding angles S of the chasing unmanned boat and the target unmanned boat, Gr is the total overlapping angle, and r is the overlapping angle.

[0164] In an embodiment of the present invention, the surrounding angle threshold TH is 330°.

[0165] The overlapping angle r is expressed by the following formula

[0166]

[0167] Wherein, α i is the relative angle between the chasing unmanned boat U i and the target unmanned boat, α j is the relative angle between the chasing unmanned boat U j and the target unmanned boat, S i is the surrounding angle between the chasing unmanned boat U i and the target unmanned boat, S j is the surrounding angle between the chasing unmanned boat U j and the target unmanned boat.

[0168] Fig. 9 Shows the schematic diagram of the total surrounding angle overlap in an embodiment of the present invention. The smaller the total surrounding angle G is, the more overlap there is. When G is less than 360°, it means there is too much overlap, and the position of the chasing boat needs to be adjusted because the size of the overlapping part is related to the difference in the phase angles between the chasing boats. The smaller the difference, the more overlap. Fig. 9 The overlapping part on the left is smaller than that on the right, and the relative G is larger.

[0169] The total surrounding angle G < TH indicates that the overlapping angle is large, and the position of the chasing unmanned boat needs to be adjusted. The minimum overlapping angle r min = ∑S - 360°. If the chasing unmanned boats can form a surrounding circle, then r min ≥ 0. When cooperating to surround, ensure that Gr is not too large, that is, Gr - r min should be as small as possible. In addition, when cooperating to surround, it is also necessary to avoid entering the warning range of the target unmanned boat, and ensure that it is not discovered by the target unmanned boat before forming a surrounding trend, that is, d > D alert . The cooperative contraction process does not need to be considered.

[0170] During the cooperative chasing stage, when G > TH, the chasing unmanned boats cooperate to contract. When cooperating to contract, the particle swarm algorithm is used to solve the velocity. The particle swarm algorithm is as follows:

[0171]

[0172]

[0173] Wherein, is the unmanned boat U captured after k+1 iterations i speed, is the number of unmanned boats captured after k iterations i speed; ω represents the inertia weight, which balances the globality and speed of convergence. A larger ω is conducive to global search, jumping out of local extreme values ​​and not falling into local optimality, while a smaller ω is conducive to local search, allowing the algorithm to converge quickly to the optimal solution; k represents the number of iterations; c1 and c2 represent individual learning factors and group learning factors; r1 and r2 are random numbers between 0 and 1, which increase the randomness of the search; is the unmanned boat U captured after k+1 iterations i location, is the number of unmanned boats captured after k iterations i location, is the number of unmanned boats captured after k iterations i The historical optimal fitness value of It is the historical optimal fitness value of the unmanned boat group.

[0174] The fitness value p is calculated using the following formula:

[0175] p=a1(Gr-r min )+a2∑(D alert -d i ),G<TH,a1> a2

[0176] p=a1(Gr-r min )+a2∑(Rs-d i ),G≥TH,a1 <a2

[0177] Among them, r min =∑S-360°, a1 and a2 are weights, satisfying a1+a2=1, Rs is the enclosing radius, d i To capture the unmanned boat U i The distance to the target unmanned boat. For the same target unmanned boat, there can be multiple encirclement unmanned boats.

[0178] In one embodiment of the present invention, ω takes the value (0.4, 2), c1 and c2 take the values ​​(0, 4), a1 and a2 take the values ​​(0, 1), and the number of iterations k takes the value (50, 100).

[0179] Set a safe distance D between unmanned boats safe , if there is |d ij -D safe|≤∈, the capture unmanned boat is reinitialized to remove the unsatisfactory solutions and meet the collision avoidance requirement with the target unmanned boat. In one embodiment of the present invention, ∈ can be a very small value. Once the distance between the capture unmanned boat and the target unmanned boat is very close to the safety distance, especially when it is very small, it does not meet the collision avoidance requirement and needs to be reinitialized.

[0180] The specific process of particle swarm algorithm is as follows:

[0181] 1. Initialize the population, taking the current position of the unmanned boat as the benchmark and the maximum speed v of the unmanned boat as the benchmark. max is the range, θ takes the value (0°, 360°), and the position of each particle is calculated according to x1=X+vcosθ, x2=X+vsinθ. The population size is N, the particle dimension is Z*2, and there are Z unmanned boats. Each unmanned boat contains two coordinates. Consider the current moving direction of the unmanned boat θ c , divided into (θ c ,θ c +180°)、(θ c ,θ c -180°) to select the initial velocity direction of the individuals in the population to prevent the random population from concentrating on one side and limiting the range.

[0182] 2. Calculate the fitness value, record the global optimal individual of the first generation, and the historical optimal individual of each particle;

[0183] 3. The number of iterations k does not reach the maximum number k max , then start updating the speed and position according to the speed and position update formula.

[0184] 4. Recalculate the fitness value.

[0185] 5. Update individual optimality and global optimality.

[0186] 6. According to the formula ω=ω max -(ω max -ω min )k / k max Update the inertia weight ω, ω max ,ω min The upper and lower bounds of the weight value range.

[0187] 7. When the maximum number of iterations is reached, output the optimal solution v.

[0188] 8. Constrain the optimal solution and use the relative angle θ between the current position of the unmanned boat and the last position as the current speed v of the unmanned boat c Direction, the optimal solution v should satisfy:

[0189]

[0190] where θ m is the corner constraint.

[0191] Next, when the difference between the distance between the encircling unmanned boat and the target unmanned boat and the encirclement radius Rs is less than a given threshold ε and the total encirclement angle is greater than or equal to 360°, the encirclement is successful.

[0192] The conditions for a successful roundup are:

[0193] ∑(d i -Rs)≤ε

[0194] G≥360°

[0195] Among them, d i To capture the unmanned boat U i The distance to the target unmanned boat, Rs is the encirclement radius, ε is the given threshold, and G is the total encirclement angle.

[0196] Finally, if a target unmanned boat is captured, the task will be redistributed, and the capture mission will be completed if all target unmanned boats are successfully captured.

[0197] The particle swarm algorithm proposed in the present invention first randomly selects a series of positions, and calculates the total encirclement angle and the sum of the distances of the new positions, and then calculates through the above formula, and then iterates the particle swarm algorithm to find an optimal position as the optimal position for encircling the unmanned boat. If the optimal position meets the conditions, the encirclement process is terminated, otherwise it continues to be executed.

[0198] In one embodiment of the present invention, the present invention can use C++ or Python as the programming language, and implement specific decision-making functions through the contract network algorithm and the particle swarm algorithm. According to needs, some simulation platforms such as ROS, Unity-3D, etc. can also be used, and the communication method adopts UDP communication.

[0199] The multi-target unmanned boat capture method proposed in the present invention includes multi-target unmanned boat task allocation and single-target unmanned boat capture. The multi-target unmanned boat task allocation is performed through an improved contract network algorithm, and the communication volume is reduced by limiting the number of times the tenderer communicates with the surroundings. At the same time, the unmanned boat exchange between adjacent target unmanned boats is used to optimize the allocation results and improve the quality of the solution. Compared with the traditional contract network algorithm, the speed is accelerated and the results are better.

[0200] The obstacle is considered as a static unmanned boat with a certain encirclement angle to the target unmanned boat. At the same time, the improved particle swarm algorithm is combined with the staged encirclement method to realize the encirclement control of multiple unmanned boats and consider the motion constraints. By improving the population initialization method, the distance is used to judge whether to approach collaboratively or to encircle collaboratively. If the encirclement unmanned boat is far away from the target unmanned boat as a whole, it will approach collaboratively, otherwise it will be encircled collaboratively. According to the encirclement angle, the encirclement situation is calculated to use different fitness functions to calculate the best position. If the encirclement situation is good, the fitness function is mainly to reduce the distance to the target unmanned boat and to maintain the encirclement situation, otherwise it is the opposite. The encirclement success rate is improved, and the method is stable and adaptable.

[0201] In one embodiment of the present invention, the present invention also provides a multi-target unmanned boat capture system for use in the multi-target unmanned boat capture method as described above. Fig.10 FIG. 1 is a schematic diagram of a multi-target unmanned boat capture system according to an embodiment of the present invention. Fig.10 As shown, the system includes the following modules:

[0202] The tenderer selection module is configured to select a suitable capture unmanned boat as a tenderer for the target unmanned boat according to the information of the capture unmanned boat and the target unmanned boat;

[0203] The bidding module is configured so that the bidder releases information to surrounding unmanned capture boats for bidding, and the unmanned capture boats that receive the message are bidders;

[0204] A bidding module is configured to calculate the distance between the bidder and the target unmanned boat and the encirclement angle of the target unmanned boat for the bidder to bid;

[0205] A bidder combination determination module is configured to select the best bidder combination as the allocation result according to the distance and the encirclement angle by the tenderer, and notify the corresponding bidders;

[0206] The optimization allocation module is configured such that after all target unmanned boats are allocated, adjacent target unmanned boats partially exchange the encirclement unmanned boats to obtain an optimized allocation result;

[0207] The encirclement and capture module is configured to execute the encirclement and capture, and the encirclement and capture are divided into coordinated approach and coordinated encirclement and capture; when the distance between the encirclement and capture unmanned boat and the target unmanned boat is greater than a set threshold, the coordinated approach and capture are executed; when the distance between the encirclement and capture unmanned boat and the target unmanned boat is less than the set threshold, the coordinated encirclement and capture are executed, and the particle swarm algorithm is used to perform the speed solution;

[0208] The capture success judgment module is configured to capture successfully when the difference between the distance between the capture unmanned boat and the target unmanned boat and the encirclement radius Rs is less than a given threshold ε and the total encirclement angle is greater than or equal to 360°;

[0209] The capture mission completion judgment module is configured to reallocate tasks if a target unmanned boat is captured, and the capture mission is completed if all target unmanned boats are successfully captured.

[0210] In one embodiment of the present invention, the present invention also provides a computer system, which includes a processor, a graphics card with an artificial intelligence chip and a memory, the memory is configured to store machine-readable instructions, the graphics card is configured to train the multi-target unmanned boat encirclement method, and the processor is configured to execute machine-readable instructions. When the processor and / or the graphics card executes the machine-readable instructions, the following processing steps are implemented: according to the information of the encirclement unmanned boat and the target unmanned boat, a suitable encirclement unmanned boat is selected as the bidder for the target unmanned boat; the bidder publishes information to the surrounding encirclement unmanned boats for bidding, and the encirclement unmanned boats that receive the message are bidders; the bidders bid and calculate their distance from the target unmanned boat and the encirclement angle of the target unmanned boat; the bidder selects the best bidder combination as the allocation result according to the distance and encirclement angle, and notifies the corresponding bidders; after all target unmanned boats are allocated, the adjacent target unmanned boats locally exchange the encirclement unmanned boats to obtain the optimal The allocation result is optimized; the encirclement is executed, and the encirclement is divided into coordinated approach and coordinated encirclement; when the distance between the encirclement unmanned boat and the target unmanned boat is greater than the set threshold, the coordinated approach is executed; when the distance between the encirclement unmanned boat and the target unmanned boat is less than the set threshold, the coordinated encirclement is executed, and when the total encirclement angle is greater than the encirclement angle threshold, the encirclement unmanned boats are contracted in a coordinated manner, and the particle swarm algorithm is used for speed calculation; when the difference between the distance between the encirclement unmanned boat and the target unmanned boat and the encirclement radius Rs is less than a given threshold ε and the total encirclement angle is greater than or equal to 360°, the encirclement is successful; if a target unmanned boat is encircled, the task is redistributed, and the encirclement task ends when all target unmanned boats are successfully encircled.

[0211] The graphics card may preferably have a GPU computing power higher than 5.0. Since the amount of data to be trained is large, providing a graphics card configuration can significantly increase the training speed.

[0212] The memory includes: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store machine-readable instructions.

[0213] It can be understood that in addition to the memory and processor mentioned above, the above-mentioned computer system also includes other software and hardware components not listed in this specification. The specific components can be determined according to the model of specific data processing equipment in different application scenarios, and this specification will not list them one by one in detail.

[0214] Although various embodiments of the present invention are described above, it should be understood that they are presented as examples only and not as limitations. It is obvious to those skilled in the relevant art that various combinations, modifications and changes can be made thereto without departing from the spirit and scope of the present invention. Therefore, the breadth and scope of the present invention disclosed herein should not be limited by the exemplary embodiments disclosed above, but should be defined according to the technical solutions of the present invention and their equivalents.

Claims

1. A multi-target unmanned boat capture method, characterized in that: The steps include: According to the information of the capture unmanned boat and the target unmanned boat, a suitable capture unmanned boat is selected as the bidder for the target unmanned boat; The bidder releases information to surrounding unmanned capture boats for bidding, and the unmanned capture boats that receive the message are bidders; The bidder makes a bid and calculates the distance between the bidder and the target unmanned boat and the encirclement angle of the target unmanned boat; The tenderer selects the best bidder combination according to the distance and the encirclement angle as the allocation result, and notifies the corresponding bidders; After all target unmanned boats are allocated, adjacent target unmanned boats partially exchange the encirclement unmanned boats to obtain the optimized allocation result; Execute the roundup, which is divided into coordinated approach and coordinated roundup; when the distance between the roundup unmanned boat and the target unmanned boat is greater than a set threshold, the coordinated approach is executed; when the distance between the roundup unmanned boat and the target unmanned boat is less than the set threshold, the coordinated roundup is executed, and the particle swarm algorithm is used to perform speed solution; When the difference between the distance between the encircling unmanned boat and the target unmanned boat and the encirclement radius Rs is less than the given threshold ε and the total encirclement angle is greater than or equal to 360°, the encirclement is successful; If a target unmanned boat is surrounded, the task will be redistributed. If all target unmanned boats are successfully surrounded, the surrounding mission will be completed.

2. The multi-target unmanned boat capture method according to claim 1, characterized in that: According to the location of the capture unmanned boat and the target unmanned boat, the tenderer for selecting a suitable capture unmanned boat for the target unmanned boat includes: All the captured unmanned boats are represented by the set U. U={U1,U2,...,U i ,...U n } in, in, is the captured unmanned boat U at time t+1 i location, is the captured unmanned boat U at time t+1 i speed, The number of unmanned boats captured at time t is i location; All target unmanned boats are represented by set T, T={T1,T2,...T j ,...T m } in, in, is the target unmanned boat T at time t+1 j location, is the target unmanned boat T at time t+1 j speed, The number of unmanned boats captured at time t is T j location; Target unmanned boat T j Use the set D to select the smallest distance unmanned boat as the bidder, D = min{d 1j , d 2j , ..., d nj } Among them, d nj Target unmanned boat T j To capture the unmanned boat U n distance.

3. The multi-target unmanned boat capture method according to any one of claims 1-2, characterized in that: The bidder releases information to surrounding unmanned boats for bidding, and the bidder participates in the bidding including: The bidder releases information to the surroundings, with the initial number of times being 0. Each unmanned capture boat that receives the information will increase the number by 1. When the communication number limit is reached, no more communication will be performed. The unmanned capture boats that receive the message are represented by a set M. M={U1, U2, ..., U z } The number of communications is limited to 3 times.

4. The multi-target unmanned boat capture method according to claim 1, characterized in that: The bidder makes a bid and calculates the distance between the bidder and the target unmanned boat and the encirclement angle of the target unmanned boat, including: When there is no obstacle between the bidder and the target unmanned boat, the bidder calculates the distance d between the bidder and the target unmanned boat. Wherein, ΔX and ΔY are the horizontal coordinate difference and vertical coordinate difference between the bidder and the target unmanned boat respectively; When there is an obstacle between the bidder and the target unmanned boat, the bidder calculates the distance d between the bidder and the target unmanned boat. Wherein, ΔX and ΔY are the difference in horizontal coordinates and vertical coordinates between the bidder and the target unmanned boat, respectively, and L is the maximum size of the obstacle; The bidder calculates the relative angle α between itself and the target unmanned boat, α=tanh- 1 (ΔY / ΔX) The bidder calculates the enveloping angle S between itself and the target unmanned boat, S=sin -1 (Vu / Vt) Wherein, Vu is the speed of the bidder, and Vt is the speed of the target unmanned boat.

5. The multi-target unmanned boat capture method according to any one of claims 1 to 4, characterized in that: The tenderer selects the best bidder combination according to the distance and the encirclement angle as the allocation result, and notifies the corresponding bidders including: The best bidder combination satisfies the following formula: ∑S≥360° Among them, ∑S is the sum of the encirclement angles S between the bidder and the target unmanned boat in the best bidder combination; The best bidder combination satisfies the minimum target unmanned boat function, and the target unmanned boat function is, f=min(w1D sum +w2(-max|Dα|)) Among them, min is the minimum value, D sum is the sum of the distances d between the bidders and the target unmanned boat in the best bidder combination, max|Δα| is the maximum value of the difference in angles between the bidders relative to the target unmanned boat, w1 and w2 are weights, and satisfy w1+w2=1.

6. The multi-target unmanned boat capture method according to claim 1, characterized in that: After all target unmanned boats are allocated, adjacent target unmanned boats partially exchange the encirclement unmanned boats, and the optimized allocation results include: The capture unmanned boats assigned to each target unmanned boat are exchanged with each other, and the change Δf of the total target unmanned boat function before and after the exchange is calculated ij , Among them, f i 、f j To capture the unmanned boat U i , U j The target unmanned boat function value before the exchange, To capture the unmanned boat U i , U j The target unmanned boat function value after exchange, Δf ij Satisfy Δf ij >0; The difference of the exchange is represented by the set E, E={e1,e2,...e e } Among them, e i ={Δf ij , i, j}, select the largest difference max{e i } as the optimized allocation result.

7. The multi-target unmanned boat capture method according to claim 1, characterized in that: Execute a roundup, which is divided into a coordinated approach and a coordinated roundup; when the distance between the roundup unmanned boat and the target unmanned boat is greater than a set threshold, a coordinated approach is executed; When the distance between the capture unmanned boat and the target unmanned boat is less than the set threshold, the coordinated capture is performed, and the particle swarm algorithm is used to solve the speed, including: The set threshold Dt h satisfy, D detect >D th >D alert Among them, D detect To capture the unmanned boat detection range, D alert It is the warning range of the target unmanned boat. When the total encirclement angle G satisfies the following formula, the capture unmanned boats shrink in coordination: G>TH Among them, TH is the encirclement angle threshold, G = ∑S-Gr, ∑S ≥ 360°, Gr = ∑r, ∑S is the sum of the encirclement angles S of the capture unmanned boat and the target unmanned boat, Gr is the total overlap angle, and r is the overlap angle; The overlap angle r is expressed as follows: Among them, α i To capture the unmanned boat U i The relative angle with the target unmanned boat, α j To capture the unmanned boat U j The relative angle with the target unmanned boat, S i To capture the unmanned boat U i The encirclement angle with the target unmanned boat, S j To capture the unmanned boat U j The encirclement angle with the target unmanned boat; In the collaborative capture phase, when G>TH, the capture unmanned boats shrink collaboratively. During collaborative shrinkage, the particle swarm algorithm is used to perform speed calculation. The particle swarm algorithm is as follows: in, is the unmanned boat U captured after k+1 iterations i speed, is the number of unmanned boats captured after k iterations i speed, ω represents the inertia weight, c1 and c2 represent the individual learning factor and the group learning factor, r1 and r2 are random numbers between 0 and 1, is the unmanned boat U captured after k+1 iterations i location, is the number of unmanned boats captured after k iterations i location, is the number of unmanned boats captured after k iterations i The historical optimal fitness value of The historical optimal fitness value of the group of unmanned boats for encirclement; The fitness value p is calculated using the following formula: p=a1(Gr-r min )+a2∑(D alert -d i ),G<TH,a1>a2 p=a1(Gr-r min )+a2∑(Rs-d i ),G≥TH,a1<a2 Among them, r min =∑S-360°, a1 and a2 are weights, satisfying a1+a2=1, Rs is the enclosing radius; d i To capture the unmanned boat U i The distance to the target unmanned boat.

8. The multi-target unmanned boat capture method according to claim 1, characterized in that: When the difference between the distance between the capture unmanned boat and the target unmanned boat and the encirclement radius Rs is less than the given threshold ε and the total encirclement angle is greater than or equal to 360°, the capture is successful including: The conditions for a successful roundup are: ∑(d i -Rs)≤ε G≥360° Among them, d i To capture the unmanned boat U i The distance to the target unmanned boat, Rs is the encirclement radius, ε is the given threshold, and G is the total encirclement angle.

9. A multi-target unmanned boat capture system for use in the multi-target unmanned boat capture method as claimed in any one of claims 1 to 8, characterized in that: Includes the following modules: The tenderer selection module is configured to select a suitable capture unmanned boat as a tenderer for the target unmanned boat according to the information of the capture unmanned boat and the target unmanned boat; The bidding module is configured so that the bidder releases information to surrounding unmanned capture boats for bidding, and the unmanned capture boats that receive the message are bidders; A bidding module is configured to calculate the distance between the bidder and the target unmanned boat and the encirclement angle of the target unmanned boat for the bidder to bid; A bidder combination determination module is configured to select the best bidder combination as the allocation result according to the distance and the encirclement angle by the tenderer, and notify the corresponding bidders; The optimization allocation module is configured such that after all target unmanned boats are allocated, adjacent target unmanned boats partially exchange the encirclement unmanned boats to obtain an optimized allocation result; The encirclement and capture module is configured to execute the encirclement and capture, and the encirclement and capture are divided into coordinated approach and coordinated encirclement and capture; when the distance between the encirclement and capture unmanned boat and the target unmanned boat is greater than a set threshold, the coordinated approach and capture are executed; when the distance between the encirclement and capture unmanned boat and the target unmanned boat is less than the set threshold, the coordinated encirclement and capture are executed, and the particle swarm algorithm is used to perform the speed solution; The capture success judgment module is configured to capture successfully when the difference between the distance between the capture unmanned boat and the target unmanned boat and the encirclement radius Rs is less than a given threshold ε and the total encirclement angle is greater than or equal to 360°; The capture mission completion judgment module is configured to reallocate tasks if a target unmanned boat is captured, and the capture mission is completed if all target unmanned boats are successfully captured.

10. A computer system, characterized in that: include: a processor configured to execute machine-readable instructions; A graphics card with an artificial intelligence chip configured to train a multi-target unmanned boat capture method; as well as A memory configured to store machine-readable instructions, wherein the machine-readable instructions, when executed by a processor and / or a graphics card, perform the steps of the method according to any one of claims 1 to 8.

Images