Recycling control method of underwater cluster equipment, equipment and medium

Abstract

The invention discloses a recovery control method and device for underwater cluster equipment and a medium, and relates to the technical field of unmanned equipment recovery, and the method comprises the steps: distributing virtual berths for all underwater equipment under the condition of determining that all underwater equipment receives a recovery instruction and travels to a preset recovery range, each underwater device is controlled to run to the corresponding virtual berth; based on the equipment poses of the underwater equipment and the docking poses of the docking platform, determining a target docking path and approaching guide time; and based on the genetic algorithm and the approach guide time, determining the recovery docking time under the condition that the underwater devices in the target docking path meet the shortest total consumed time, and generating a recovery scheduling timetable based on the recovery time. The method is used for solving the problems of long time consumption, low efficiency and poor recovery effect when the underwater cluster equipment is recovered in the prior art, and the recovery of the underwater cluster equipment is rapidly and efficiently completed.

Description

Technical Field

[0001] This application relates to the field of unmanned equipment recovery technology, and in particular to a recovery control method, equipment and medium for underwater cluster equipment. Background Technology

[0002] In response to the increasing complexity of marine operating environments, the diversification of mission scenarios, and the continuous expansion of operating radii, marine unmanned systems are accelerating their development towards clustering and distributed collaboration in order to achieve the strategic goal of fully autonomous large-scale unmanned swarm equipment.

[0003] Currently, the recovery of underwater unmanned equipment mainly relies on submarines or surface ships as mother ships for docking and capture. A common operational procedure is as follows: after receiving a recovery command, the target equipment maneuvers towards the mother ship using its navigation and positioning system. Upon reaching the designated location, it completes a physical connection via a mechanical hook or docking device, thus achieving capture and recovery. In this process, the recovery command is only used to instruct single-unit operations. When applied to unmanned swarms, operators still need to rely on experience to repeatedly determine the recovery order and task time for the smallest unit to be recovered within the swarm until the entire swarm is recovered.

[0004] Current underwater unmanned swarm recovery methods remain in a single-machine command mode. During swarm recovery, operators must manually determine the recovery priority and timing of individual equipment based on experience, which is difficult to meet the needs of complex scenarios such as multi-machine collaborative path planning and dynamic obstacle avoidance. Furthermore, the lack of scheduling algorithm support means that only one equipment recovery task can be processed per round, requiring repeated iterations until the entire swarm is recovered, which seriously affects operational efficiency and is difficult to adapt to the heterogeneous and concurrent characteristics of swarm recovery. Summary of the Invention

[0005] In response to the aforementioned problems and technical requirements, the applicant has proposed a method, equipment, and medium for the recovery and control of underwater cluster equipment. This method aims to solve the problems of long recovery time, low efficiency, and poor recovery effect in the existing technology for underwater cluster equipment recovery, and to achieve rapid and efficient recovery of underwater cluster equipment.

[0006] This application provides a method for controlling the recovery of underwater cluster equipment, the method comprising:

[0007] Once it is determined that each underwater device has received a recovery command and has traveled to the preset recovery range, a virtual berth is assigned to each underwater device based on the Hungarian algorithm, and each underwater device is controlled to travel to the corresponding virtual berth.

[0008] Based on the equipment pose of each underwater device at the virtual berth and the docking pose of the docking platform, the target docking path and approach guidance time are determined. The target docking path is the shortest path for the underwater device to reach the docking platform from the virtual berth.

[0009] Based on genetic algorithms and approach guidance time, the recovery docking time of each underwater device under the target docking path is determined to minimize the total time. A recovery scheduling timetable is generated based on the recovery time, and the underwater cluster equipment is recovered based on the recovery scheduling timetable.

[0010] The underwater cluster equipment recovery control method provided in this application, based on a genetic algorithm and approach guidance time, determines the recovery docking time for each underwater device under the target docking path while minimizing the total time, including:

[0011] Based on the approach guidance time, determine the earliest and latest docking times corresponding to the target docking path;

[0012] Based on the earliest docking time and the latest docking time, the recovery docking time of each underwater device is determined, resulting in a set of docking times;

[0013] The docking time set is optimized based on a genetic algorithm and a preset objective function to minimize the total time for each underwater device, thus obtaining the final docking time set.

[0014] According to the underwater cluster equipment recovery control method provided in the embodiments of this application, the objective function includes:

[0015] min f=∑ i∈EPLAN (Aim i -plan i )×PE i +∑ j∈LPLAN (plan j -Aim j )×PL j ;

[0016] Where f represents the target value, EPLAN represents the set of early-arriving AUVs, LPLAN represents the set of late-arriving AUVs, and Aim... i This represents the target docking time for the i-th AUV, obtained based on the earliest docking time. i Let represent the docking time of the i-th AUV, and plan represent the docking time of the i-th AUV. j Let Aim represent the docking time of the j-th AUV. j PE represents the target docking time of the j-th AUV. i PL represents the penalty coefficient for underwater equipment arriving early. j This represents the penalty coefficient for underwater equipment arriving late.

[0017] According to the underwater cluster equipment recovery control method provided in the embodiments of this application, the docking time set is optimized based on a genetic algorithm and a preset objective function to minimize the total time consumed by each underwater device, resulting in the final docking time set, including:

[0018] A random set of docking times is initialized, and docking times with a fitness greater than a preset fitness are selected from the set of docking times based on the roulette wheel selection method.

[0019] The iterative execution involves cross-operations between the target docking time and the selected docking time to be processed to obtain the mutated docking time to be processed, and determining the total time consumption corresponding to the new set of docking times, until the set of docking times satisfies the objective function.

[0020] The underwater cluster equipment recovery control method provided in the embodiments of this application determines the target docking path based on the equipment pose of each underwater device at the virtual berth and the docking pose of the docking platform, including:

[0021] Input the equipment pose and docking pose into multiple preset path calculation formulas to obtain the path length output by each path formula;

[0022] The path corresponding to the smallest path length extracted from multiple path lengths is taken as the target docking path;

[0023] Among them, the multiple path calculation formulas include: the first calculation formula, the second calculation formula, the third calculation formula, and the fourth calculation formula. Different path calculation formulas correspond to different docking methods.

[0024] The first calculation formula includes:

[0025] L LSL =t L +p S +q L =-α+β+p S ;

[0026] Among them, L LSL t represents the first path length of the first docking method. L p represents the length of the first curve of the first segment in the first docking method. S q represents the length of the first straight line in the middle segment of the first type of docking. L The length of the second curve in the second segment of the first docking method is indicated by α, which represents the angle between the initial heading of the underwater equipment and the direction of the target connection line. The initial heading is the direction of travel at the virtual berth, and the direction of the target connection line is the direction from the virtual berth to the docking platform. β represents the angle between the docking heading of the underwater equipment and the direction of the target connection line. The docking heading is the direction of the docking platform corresponding to the underwater equipment.

[0027] The second calculation formula includes:

[0028] L LSR =t L +p S +q R =α-β+2t L +p S ;

[0029] Among them, L LSR q represents the second path length for the second docking method. R This indicates the length of the second curve in the second segment of the second docking method;

[0030] The third calculation formula includes:

[0031] L RSL =t R +p S +q L = -α + β + 2t R +p S ;

[0032] Among them, L RSL t represents the length of the third path in the third docking method. R This indicates the length of the first curve in the first segment of the third docking method;

[0033] The fourth calculation formula includes:

[0034] L RSR =t R +p S +q R =α-β+p S ;

[0035] Among them, L RSR This indicates the fourth path length for the fourth docking method.

[0036] According to the underwater cluster equipment recovery control method provided in the embodiments of this application, the approach guidance time is determined based on the equipment pose of each underwater device at the virtual berth and the docking pose of the docking platform, including:

[0037] Determine the average velocity of the underwater equipment under the target docking path, where the target docking path is obtained based on the equipment pose and docking pose;

[0038] Input the path length of the target docking path and the average speed into the preset approach guidance time calculation formula to obtain the approach guidance time output by the approach guidance time calculation formula;

[0039] The approach guidance time calculation formula includes:

[0040] t i =L total-i / V i ;

[0041] Among them, t i L represents the approach guidance time of the i-th underwater device. total-i V represents the path length of the target docking path for the i-th underwater equipment. i Let represent the average speed of the i-th underwater equipment.

[0042] The underwater cluster equipment recovery control method provided in the embodiments of this application allocates virtual berths to each underwater device based on the Hungarian algorithm, including:

[0043] Obtain the equipment location of each underwater device and the berth location of each virtual berth;

[0044] A cost matrix is ​​constructed based on the equipment location and berth location, and the target matrix corresponding to the cost matrix is ​​solved based on the Hungarian algorithm, wherein the target matrix includes multiple zero elements;

[0045] Based on the position of the zero element in the target matrix, the correspondence between underwater equipment and berth positions is determined.

[0046] The underwater cluster equipment recovery control method provided in this application embodiment solves for the target matrix corresponding to the cost matrix based on the Hungarian algorithm, including:

[0047] Subtract the first minimum value from each row of the target matrix and subtract the first minimum value from each column of the target matrix to obtain the matrix to be processed, wherein the first minimum value is obtained from the target matrix;

[0048] Cover all zero elements in the matrix to be processed using the smallest horizontal and vertical lines;

[0049] If the first number of horizontal and vertical lines to be covered is less than the second number of underwater equipment, the second minimum value is obtained from the uncovered areas. The second minimum value is then subtracted from each row of the uncovered areas and added to each column of the covered areas until the first number equals the second number, thus obtaining the target matrix.

[0050] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the underwater cluster device recovery control method as described above.

[0051] This application also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the underwater cluster device recovery control method as described above.

[0052] The underwater swarm equipment recovery control method, equipment, and medium provided in this application, after determining that each underwater device has received a recovery command and traveled to a preset recovery range, allocates virtual berths to each underwater device based on the Hungarian algorithm and controls each underwater device to travel to the corresponding virtual berth, thus making preliminary preparations for the coordinated recovery planning and swarm scheduling of the underwater swarm equipment. Based on the equipment pose of each underwater device at the virtual berth and the docking pose of the docking platform, the target docking path and approach guidance time are determined, wherein the target docking path is the shortest path for the underwater device to reach the docking platform from the virtual berth. Based on the genetic algorithm and approach guidance time, the recovery docking time of each underwater device under the target docking path is determined to meet the condition of the shortest total time, and a recovery scheduling timetable is generated based on the recovery time, so as to complete the recovery of the underwater swarm equipment based on the recovery scheduling timetable. It can be seen that this application aims to minimize the total time and total distance of all underwater devices in the swarm to complete the recovery, ensuring the coordinated control of the swarm, improving the recovery efficiency, and realizing the rapid and efficient recovery of underwater swarm equipment. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 This is one of the flowcharts illustrating the recovery control method for underwater cluster equipment provided in this application embodiment;

[0055] Figure 2 This is a schematic diagram illustrating the relationship between underwater equipment, virtual berth, and docking platform provided in the embodiments of this application;

[0056] Figure 3 This is a second schematic flowchart of the underwater cluster equipment recovery control method provided in the embodiments of this application;

[0057] Figure 4 This is a schematic diagram of the AUV depth-fixed operation space provided in the embodiments of this application;

[0058] Figure 5 This is a schematic diagram of the optimal allocation provided in the embodiments of this application;

[0059] Figure 6 This is a Gantt chart of unmanned cluster recovery provided in the embodiments of this application;

[0060] Figure 7This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0062] This application provides a method for controlling the recovery of underwater cluster equipment. This method can be applied to smart terminals, servers, and recovery systems (which include docking platforms, controllers, underwater equipment, communication modules, etc.). This application uses the application of this method in a server as an example for illustration, and some other descriptions in the embodiments are for illustrative purposes only and are not intended to limit the scope of protection of this application, and will not be described in detail thereafter. The specific implementation of the method is as follows... Figure 1 As shown:

[0063] Step 101: After determining that each underwater equipment has received a recovery command and traveled to the preset recovery range, a virtual berth is assigned to each underwater equipment based on the Hungarian algorithm, and each underwater equipment is controlled to travel to the corresponding virtual berth.

[0064] Step 102: Based on the equipment pose of each underwater device at the virtual berth and the docking pose of the docking platform, determine the target docking path and the approach guidance time.

[0065] The target docking path is the shortest path for underwater equipment to reach the docking platform from the virtual berth.

[0066] Step 103: Based on the genetic algorithm and approach guidance time, determine the recovery docking time of each underwater equipment under the target docking path that meets the condition of the shortest total time, and generate a recovery scheduling timetable based on the recovery time, so as to complete the recovery of the underwater cluster equipment based on the recovery scheduling timetable.

[0067] Specifically, through Figure 2 The relationship between underwater equipment, virtual berths, and docking platforms is illustrated. Figure 2 The diagram uses six AUVs as an example. The landing area represents the docking platform, and the star shape in the diagram represents a virtual berth.

[0068] The underwater swarm equipment recovery control method provided in this application, after determining that each underwater device has received a recovery command and traveled to a preset recovery range, allocates virtual berths to each underwater device based on the Hungarian algorithm, and controls each underwater device to travel to the corresponding virtual berth, thus making preliminary preparations for the coordinated recovery planning and swarm scheduling of the underwater swarm equipment. Based on the equipment pose of each underwater device at the virtual berth and the docking pose of the docking platform, the target docking path and approach guidance time are determined, wherein the target docking path is the shortest path for the underwater device to reach the docking platform from the virtual berth. Based on the genetic algorithm and approach guidance time, the recovery docking time of each underwater device under the target docking path is determined to meet the condition of the shortest total time, and a recovery scheduling timetable is generated based on the recovery time, so as to complete the recovery of the underwater swarm equipment based on the recovery scheduling timetable. It can be seen that this application aims to minimize the total time and total distance of all underwater devices in the swarm to complete the recovery, ensuring the coordinated control of the swarm, improving the recovery efficiency, and realizing the rapid and efficient recovery of underwater swarm equipment.

[0069] Specifically, the recovery control process of this application is divided into four stages: return to base, centralized takeover, approach and docking, and capture and landing.

[0070] Returning to base. After receiving the recovery command, the unmanned swarm equipment exits the work area and returns to the recovery platform. Using a submarine or surface ship as the recovery platform, targets within a certain range can be identified, and AUVs entering the recovery identification zone can be monitored in real time and taken over.

[0071] Centralized takeover. After successful takeover by the recovery system, the AUV's control mode changes from autonomous return to takeover waiting berth, with multiple AUVs forming an unmanned cluster awaiting recovery. Due to limited platform recovery equipment resources, multiple AUVs cannot be recovered simultaneously. Therefore, at a certain distance from the platform, virtual berths are allocated to the unmanned cluster and planned and scheduled, with each AUV awaiting individual approach instructions.

[0072] Approach and docking. Guided by the recovery system, the unmanned swarm sequentially performs the recovery. During the approach, the AUV adjusts its attitude and gradually approaches the target location, completes docking, and is captured by the equipment on the recovery platform.

[0073] Capture and landing. After capturing the AUV, the platform immediately transfers the AUV and resets the recovery device to release its occupancy and prepare for the next AUV docking and capture.

[0074] In one specific embodiment, the specific implementation of allocating virtual berths to various underwater equipment based on the Hungarian algorithm includes:

[0075] Obtain the equipment location of each underwater device and the berth location of each virtual berth; construct a cost matrix based on the equipment location and berth location, and solve the target matrix corresponding to the cost matrix using the Hungarian algorithm; determine the correspondence between underwater devices and berth locations based on the position of the zero element in the target matrix.

[0076] The target matrix includes multiple zero elements.

[0077] Specifically, assuming there are n AUVs forming an unmanned swarm, the equipment position of the i-th AUV is determined as P. i The position of the j-th berth is g. j Let X represent the solution matrix for target allocation in an unmanned swarm, where each element C ij This indicates a correspondence. Specifically, each AUV has one and only one corresponding berth location. If X... ij If P equals 1, then P i The corresponding AUV is assigned to g j If X ij An equal value of 0 indicates that P i The corresponding AUV was never assigned to g. j Place.

[0078] For details, please refer to formulas (1) and (2):

[0079]

[0080]

[0081] Use C ij Let $\frac{i}{j}$ represent the distance cost for the i-th AUV to reach the j-th berth position, as shown in formula (3):

[0082] C ij =||P i -g j || 2 ……………………………(3)

[0083] Based on the above information, a mathematical model for target allocation can be obtained, as shown in formula (4):

[0084]

[0085] Where s represents the numerical value output by the mathematical model for target allocation.

[0086] Among them, by minimizing s, the correspondence between underwater equipment and berth locations is obtained.

[0087] This application transforms the cluster target allocation problem into a linear allocation problem and introduces an optimal allocation algorithm to solve for matrix X. ijThis minimizes the sum of the squares of the straight-line distances from each AUV to its corresponding berth.

[0088] In one specific embodiment, the specific implementation of solving the target matrix corresponding to the cost matrix based on the Hungarian algorithm includes:

[0089] Subtract the first minimum value from each row of the target matrix and subtract the first minimum value from each column of the target matrix to obtain the matrix to be processed; cover all zero elements in the matrix to be processed with the minimum horizontal and vertical lines; if the first number of covered horizontal and vertical lines is less than the second number of underwater equipment, obtain the second minimum value from the uncovered areas, and subtract the second minimum value from each row of the uncovered areas and add the second minimum value to each column of the covered areas until the first number equals the second number, thus obtaining the target matrix.

[0090] The first minimum value is obtained from the target matrix.

[0091] Specifically, the Hungarian algorithm is an efficient combinatorial optimization algorithm for solving multi-task assignment problems. It has been widely proven to effectively solve single-objective assignment problems where the objective function is a linear function and the assignment strategy is one-to-one. Its core idea is that the maximum number of independent zero elements in the cost matrix equals the minimum number of straight lines that can cover all zero elements. The following... Figure 3 Here are the steps of the Hungarian algorithm in detail:

[0092] Step 301: Solve for the values ​​of each element in the cost matrix.

[0093] Specifically, the cost is assessed based on the distance between the initial position of the unmanned cluster and the virtual target berth at the takeover moment. An n×n cost matrix C is constructed based on the mathematical model of target allocation, namely:

[0094]

[0095] Step 302: Perform a linear transformation on the cost matrix: subtract the minimum value from each row and each column to ensure that there are zero elements in each row and column.

[0096] Step 303: Cover all zero elements in the matrix with the fewest possible horizontal or vertical lines.

[0097] Step 304: If the total number of covered horizontal and vertical lines is less than n, find the smallest uncovered element a, and subtract a from each uncovered row and add a to each covered column.

[0098] Step 305: Repeat steps 203 and 204 until the total number of covered lines equals n.

[0099] Step 306: After the transformations in steps 202 to 205, the optimal solution for the matrix is ​​obtained.

[0100] The position of the zero element determines the optimal task allocation.

[0101] Compared to other allocation algorithms, this algorithm, based on mathematical principles, can quickly generate multi-task allocation schemes in a shorter time, meeting the real-time and consistency requirements of cluster target allocation. Furthermore, in situations such as changes in cluster tasks or incomplete data, the Hungarian algorithm can also find suboptimal task allocation schemes, ensuring the normal operation of the system.

[0102] In one specific embodiment, the specific implementation of determining the target docking path based on the equipment poses of each underwater device at the virtual berth and the docking pose of the docking platform includes:

[0103] The equipment pose and docking pose are input into multiple preset path calculation formulas to obtain the path length output by each path formula; the path corresponding to the smallest path length extracted from multiple path lengths is taken as the target docking path.

[0104] The multiple path calculation formulas include: the first calculation formula, the second calculation formula, the third calculation formula, and the fourth calculation formula. Different path calculation formulas correspond to different docking methods.

[0105] Specifically, the Dubins curve is introduced to perform approach planning for AUV recovery.

[0106] like Figure 4 As shown, the AUV's fixed-depth operating space is simplified to a two-dimensional plane. Let the AUV's pose in the OE-XEYE coordinate system be η = [xyψ]. T The docking pose is η G =[x g y g ψ g ] T Establish an Ob-XbYb coordinate system with the AUV origin, where the Xb axis points towards the target direction. Rotate the Xb axis counterclockwise by 90° to obtain the Yb axis. The pose of the AUV in this coordinate system can then be expressed as η. b =[00α] T The docking posture is θ is the difference in heading angle between the AUV and the target.

[0107] Where x represents the X-coordinate of the AUV, y represents the Y-coordinate of the AUV, and ψ represents the orientation of the AUV. g This represents the X-axis coordinate value of the docking platform, y g ψ represents the Y-axis coordinate value of the docking platform. gThe orientation of the docking platform is indicated by α, the angle between the initial heading of the underwater equipment and the direction of the target connection line is indicated by β, the angle between the docking heading of the underwater equipment and the direction of the target connection line is indicated by β, the angle between the docking heading of the underwater equipment and the direction of the target connection line is indicated by β, and the distance between the AUV and the docking position is indicated by D.

[0108] The formulas for calculating the motion azimuth angles of the AUV and docking pose after coordinate transformation are shown in formulas (5), (6), and (7):

[0109]

[0110] α=mod(ψ-θ,2π)………………(6)

[0111] β = mod(ψ) g -θ,2π)………………(7)

[0112] The formula for the distance between the AUV and the docking position is given in formula (8):

[0113]

[0114] To further simplify the calculation, the distance is normalized, as shown in formula (9):

[0115]

[0116] Where r is the minimum turning radius, which makes it easy to calculate the arc length using angles.

[0117] Let the length of the first curve segment of the Dubins path be t, the length of the middle straight line segment be p, and the length of the second curve segment be q. Then, the formula for calculating the total length of the planned Dubins path is shown in formula (10):

[0118] L=t+p+q……………………(10)

[0119] The formula for the total length of an LSL type Dubins path (first calculation formula) is shown in formula (11):

[0120]

[0121] Among them, L LSL t represents the first path length of the first docking method. L p represents the length of the first curve of the first segment in the first docking method. S q represents the length of the first straight line in the middle segment of the first type of docking. L This indicates the length of the second curve in the second segment of the first docking method.

[0122] The formula for the total length of the LSR type Dubins path (second calculation formula) is shown in formula (12):

[0123]

[0124] Among them, L LSR q represents the second path length for the second docking method. R This indicates the length of the second curve in the second segment of the second docking method.

[0125] The formula for the total length of an RSL type Dubins path (the third calculation formula) is shown in formula (13):

[0126]

[0127] Among them, L RSL t represents the length of the third path in the third docking method. R This indicates the length of the first curve of the first segment in the third docking method.

[0128] The formula for the total length of the RSR type Dubins path (fourth calculation formula) is shown in formula (14):

[0129]

[0130] Among them, L RSR This indicates the fourth path length for the fourth docking method.

[0131] In one specific embodiment, the specific implementation of determining the approach guidance time based on the equipment pose of each underwater device at the virtual berth and the docking pose of the docking platform includes:

[0132] Determine the average speed of the underwater equipment along the target docking path; input the path length and average speed of the target docking path into the preset approach guidance time calculation formula to obtain the approach guidance time output by the approach guidance time calculation formula.

[0133] The target docking path is obtained based on the equipment pose and the docking pose.

[0134] The approach guidance time calculation formula is shown in formula (15):

[0135] t i =L total-i / V i ………………………………(15)

[0136] Among them, t i L represents the approach guidance time of the i-th underwater device. total-i V represents the path length of the target docking path for the i-th underwater equipment.i Let represent the average speed of the i-th underwater equipment.

[0137] Specifically, given the fixed poses of the AUV and the docking platform, the optimal Dubins path for the recovery approach depends on their relative poses. During recovery, assuming the recovery platform is stationary and each AUV travels at a constant speed, the shortest path for the i-th AUV from its berth position to its assigned docking pose is calculated. Then, by controlling the average speed V of the i-th AUV traversing this path... i The time consumed by approach guidance is determined. Based on the AUV's recovery and docking time and approach guidance time, the berthing scheduling time for each AUV is determined while minimizing the total recovery time.

[0138] In one specific embodiment, the specific implementation of determining the recovery and docking time of each underwater device under the target docking path while minimizing the total time, based on genetic algorithms and approach guidance time, includes:

[0139] Based on the approach guidance time, the earliest and latest docking times corresponding to the docking path with the target are determined; based on the earliest and latest docking times, the recovery docking times of each underwater equipment are determined, resulting in a set of docking times; based on a genetic algorithm and a preset objective function, the set of docking times is optimized to minimize the total time for each underwater equipment, resulting in the final set of docking times.

[0140] Specifically, each underwater device includes multiple time parameters, such as landing occupancy time, approach window time, and docking target time. The following explanation uses the i-th underwater device as an example:

[0141] Landing occupancy time: Considering the time required for the recovery platform to capture and transfer the AUV, as well as the time consumed by the buffer docking and recovery mechanism, a time interval (buf) is required between consecutive AUV docking and recovery operations. i Since this application studies the cluster recovery problem of multiple isomorphic AUVs, the landing occupancy time of the AUVs (time interval buf) can be used as a reference. i ) are considered consistent and are constants.

[0142] Approach Window Time: Since the positions and attitudes of each AUV relative to the recovery platform are not identical, each AUV has its own approach window time. Based on the approach planning described above, the Dubins curve is introduced to calculate the shortest path from the berth position to the assigned docking position for each AUV in the cluster. During the assembly and waiting process, it is assumed that the platform remains relatively stationary with respect to the cluster, and that each AUV approaches and docks at a constant speed along its Dubins shortest path. The approach speed V for each AUV is then determined. iThis allows us to obtain the time consumed for approach guidance of each AUV, the docking time that the AUV can dock with, and the Earliest docking time for that AUV. i .

[0143] The total time for an unmanned swarm recovery mission refers to the time from the issuance of the recovery command to the completion of recovery of the last AUV in the swarm. Therefore, the landing time of each AUV and the swarm size will affect the swarm recovery time. For example, when recovering an unmanned swarm consisting of n AUVs, the latest docking time of the i-th AUV is... i It can be defined as the sum of its earliest docking time and the total landing time of the cluster. See formula (16):

[0144] Latest i =Earliest i +n×buf i ………………(16)

[0145] Latest i Earliest represents the latest docking time of the i-th AUV. i This represents the earliest docking time of the i-th AUV, where n represents the number of AUVs, and buf i This represents the landing time of the i-th AUV, which is a constant.

[0146] Docking Target Time: To optimize the retrieval and scheduling of unmanned aerial vehicles (AUVs), each AUV sets a docking target time (Aim) within its approach time window. i We obtain the result using formula (17):

[0147] Aim i =Earliest i +buf i ……………………(17)

[0148] Among them, Aim i This represents the docking target time corresponding to the i-th AUV.

[0149] Based on this time, arriving early or late will incur additional scheduling "penalty" costs. Each AUV is defined with a penalty coefficient of PE per second for arriving early and late. i and PL j .

[0150] Therefore, the unmanned aerial vehicle (AUV) cluster recycling and scheduling problem can be described as follows: A cluster consisting of n AUVs needs to be recycled, and the recycling scheduling scheme variable is defined as PLAN = {plan1, plan2, plan3, ... plan...}. n plan iThis represents the recovery and docking time of the i-th AUV, and satisfies the earliest and latest approach time constraints.

[0151] The actual docking time interval between any two AUVs must be greater than the shortest interval, satisfying formula (18):

[0152] |plan i -plan j |≥buf ij …………………(18)

[0153] Among them, plan i Indicates the recovery and docking time of the i-th AUV, plan j buf represents the docking time of the j-th AUV. ij This represents the shortest interval predetermined based on the recovery docking time of each AUV, and is a constant.

[0154] Define the set of early-arriving AUVs (EPLAN) and the set of late-arriving AUVs (LPLAN) in the scheduling scheme, as shown in formulas (19) and (20):

[0155] EPLAN = {i∈n|plan i ≤Aim i}………(19)

[0156] LPLAN = {j∈n|plan j ≥Aim j}………(20)

[0157] Among them, Aim j This represents the docking target time corresponding to the j-th AUV.

[0158] Based on the additional scheduling "penalty" cost caused by early and late arrival of AUVs, an objective function for recycling scheduling is designed.

[0159] In one specific embodiment, the objective function is shown in formula (21):

[0160] minf = ∑ i∈EPLAN (Aim i -plan i )×PE i +∑ j∈LPLAN (plan j -Aim j )×PL j ……………………………………………(twenty one)

[0161] Where f represents the target value, used to characterize the total time taken by underwater equipment; EPLAN represents the set of early-arriving AUVs; LPLAN represents the set of late-arriving AUVs; and Aim... i This represents the target docking time for the i-th AUV, obtained based on the earliest docking time. i Let represent the docking time of the i-th AUV, and plan represent the docking time of the i-th AUV. j Let Aim represent the docking time of the j-th AUV. j PE represents the target docking time of the j-th AUV. i PL represents the penalty coefficient for underwater equipment arriving early. j This represents the penalty coefficient for underwater equipment arriving late.

[0162] Wherein, the smaller f is, the shorter the total time of each underwater device in the docking time set is. When f reaches its minimum, it means that the total time of each underwater device in the docking time set is the shortest.

[0163] In one specific embodiment, the optimization of the docking time set based on a genetic algorithm and a preset objective function to minimize the total time for each underwater device, resulting in the final docking time set, includes the following specific implementations:

[0164] The docking time set is randomly initialized, and docking times with fitness greater than the preset fitness are selected from the docking time set based on the roulette wheel selection method. The steps of cross-operation based on the docking target time and the selected docking times to be processed are iteratively executed to obtain the mutated docking times to be processed, and to determine the total time corresponding to the new docking time set are determined, until the docking time set satisfies the objective function.

[0165] The fitness is dynamically changed based on the crossover operation.

[0166] Specifically, first, the population is randomly initialized, and then a roulette wheel selection method is used to select individuals with high fitness (pending mating time) from the population for use in generating the next generation. To make the fitness value higher as the objective function value is smaller, a scaling transformation is performed on the fitness, as shown in formula (22):

[0167]

[0168] Where F(fit) i Let be the fitness of the i-th individual after transformation, and fit iLet $\frac{i}{i}$ be the current fitness of the i-th individual, $\maxfit$ be the maximum fitness of the population, and $\minfit$ be the minimum fitness of the population. When using the roulette wheel selection algorithm, the individual selection strategy needs to be adjusted. The fitness of individuals not in the window time is set to $\maxfit$. During the algorithm selection process, the probability of these individuals being selected is zero, i.e., they are selected and eliminated, thus avoiding the phenomenon of fitness mismatch.

[0169] Crossover is the core step of genetic algorithms, generating new individuals to introduce new genetic information. Traditional single-point crossover involves randomly selecting two individuals from the population for crossover. This application, however, knows the target docking time for each AUV, thus incorporating this prior knowledge into the individual's chromosome. During iteration, individuals from the population are directly crossovered with the chromosomes formed by the target docking times.

[0170] By performing mutation operations on selected individuals, population diversity is introduced to prevent getting trapped in local optima. A uniform mutation operator is designed, using a uniformly distributed random number within an interval to replace the original gene value with a certain probability. Here, this interval is defined as the interval between a certain dimension value of an individual and the optimal target value. The mutated gene value is shown in formula (23):

[0171] plan'(i)=plan(i)+rand×(Aim(i)-plan(i))......(23)

[0172] Where plan′(i) is the value of the i-th gene on the chromosome after mutation (the retrieval docking time after mutation, or the docking time to be processed after mutation), plan(i) is the gene value before mutation (the retrieval docking time before mutation, or the docking time to be processed before mutation), and rand is a random number. It can be seen that during the mutation process, based on the given prior knowledge, the gene will continuously approach the unconstrained docking target time.

[0173] Where plan(i) is equivalent to plan i .

[0174] Furthermore, to prevent the loss of superior individuals during the evolution process, the proposed algorithm adopts a strategy of retaining elites from past iterations. This means that the best individuals in the population are retained, do not participate in the selection, and determine whether the best individuals need to be replaced after each crossover and mutation. This further improves the performance of the algorithm and maintains the superior individuals in the population.

[0175] In one specific embodiment, a recycling schedule is generated based on the recycling scheme described in the above embodiments to control the unmanned swarm to complete the recycling task.

[0176] For example, the simulation experiment randomly generates the takeover starting point of the unmanned swarm within the recovery and identification area, and marks the coordinates of six target berths. Regarding target allocation in the unmanned swarm task planning method, the Hungarian algorithm is introduced to solve for the optimal allocation, such as... Figure 5 As shown.

[0177] To verify the algorithm's versatility, the relevant times in the simulation examples were converted to integer form, i.e., the hour-minute-second format was not used, thus meeting the basic requirement of numerically incrementing to represent time increases. The approach window time, initial docking target time, landing occupancy time, and penalty coefficient for each AUV were set, as shown in Table 1. The relevant parameters for the genetic algorithm were set, as shown in Table 2. The simulation results were mapped to a timetable, with the cluster's approach scheduling starting at 0:00:00. The simulated recovery scheduling timetable for each AUV is shown in Table 3.

[0178] / AUV1 AUV2 AUV3 AUV4 AUV5 AUV6 Earliest 25 46 52 78 81 88 Aim 55 76 82 108 111 118 Latest 185 196 202 228 231 238 buf 30 30 30 30 30 30 PE 10 10 10 10 10 10 PL 10 10 10 10 10 10

[0179] Table 1 Basic parameters of the example

[0180] Table 1 uses six AUVs as an example for illustration only and is not intended to limit the scope of protection.

[0181] parameter numerical values Population size 100 Crossover probability 0.9 Mutation probability 0.1 Number of iterations 100

[0182] The values ​​of the genetic algorithm parameters in Table 2 are for illustrative purposes only and are not intended to limit the scope of protection.

[0183]

[0184] Table 3 Recycling Scheduling Schedule

[0185] The values ​​in Table 3 are for illustrative purposes only and are not intended to limit the scope of protection.

[0186] The recovery of each AUV in the unmanned swarm involves four processes: waiting for berth, approach, docking, and landing. Analysis of the simulation example with AUV2 shows that from 0:00:00 to 0:00:09, AUV2 assembles and waits for berth at its corresponding virtual berth node; at 0:00:09, it receives an approach command, adjusts its attitude, and gradually approaches the recovery platform; at 0:00:55, it receives the final docking command and is captured by the recovery platform; and at 0:01:25, it lands on the ship, meaning AUV2 completes its recovery. Figure 6 The Gantt chart shown here illustrates the scheduling scheme generated by the planning algorithm for unmanned cluster recycling.

[0187] The cluster recovery scheduling in this application determines the order and timing of each AUV's approach and docking tasks, solving the problem that multiple AUV clusters cannot be directly recovered, improving the recovery efficiency of unmanned clusters, and avoiding mutual interference between individual AUVs during recovery.

[0188] This application, based on minimizing recovery time and maximizing resource utilization, comprehensively considers multiple constraints such as AUV cluster, target location, and recovery resources, and establishes a suitable cluster recovery process model. It decouples the cluster task into three parts: target allocation, path planning, and task scheduling. Optimization is performed with the goal of minimizing the total recovery time and total flight distance for all individuals in the cluster. Appropriate flight targets are assigned to each AUV, and a scientific recovery plan is formulated, ensuring cluster order, avoiding mutual interference between AUVs during recovery, and improving the robustness and flexibility of unmanned cluster recovery tasks.

[0189] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 7 As shown, the electronic device may include a processor 701, a communications interface 702, a memory 703, and a communication bus 704. The processor 701, communications interface 702, and memory 703 communicate with each other via the communication bus 704. The processor 701 can call logical instructions stored in the memory 703 to execute the underwater cluster device recovery control method.

[0190] Furthermore, the logical instructions in the aforementioned memory 703 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0191] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, and when the program instructions are executed by a computer, the computer is able to execute the underwater cluster equipment recovery control method provided by the above methods.

[0192] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the underwater cluster device recovery control method provided in the above embodiments.

[0193] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0194] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0195] Finally, it should be noted that the above descriptions are merely preferred embodiments of this application, and this application is not limited to the above embodiments. It is understood that other improvements and variations directly derived or conceived by those skilled in the art without departing from the spirit and concept of this application should be considered to be included within the protection scope of this application.

Claims

1. A method for recovering and controlling underwater cluster equipment, characterized in that, The method comprises: Once it is determined that each underwater equipment has received a recovery command and has traveled to the preset recovery range, a virtual berth is assigned to each underwater equipment based on the Hungarian algorithm, and each underwater equipment is controlled to travel to the corresponding virtual berth. Based on the equipment pose of each underwater device at the virtual berth and the docking pose of the docking platform, the target docking path and approach guidance time are determined. The target docking path is the shortest path for the underwater device to reach the docking platform from the virtual berth. Based on genetic algorithms and approach guidance time, the recovery docking time of each underwater device under the target docking path is determined to minimize the total time. A recovery scheduling timetable is generated based on the recovery time to complete the recovery of the underwater cluster equipment.

2. The recovery control method for underwater cluster equipment according to claim 1, characterized in that, Based on genetic algorithms and approach guidance time, the recovery and docking time for each underwater device under the target docking path is determined to minimize the total time, including: Based on the approach guidance time, determine the earliest and latest docking times corresponding to the target docking path; Based on the earliest docking time and the latest docking time, the recovery docking time of each underwater device is determined, resulting in a set of docking times; The docking time set is optimized based on a genetic algorithm and a preset objective function to minimize the total time for each underwater device, thus obtaining the final docking time set.

3. The recovery control method for underwater cluster equipment according to claim 2, characterized in that, The objective function includes: min f=∑ i∈EPLAN (Aim i -plan i )×PE i +∑ j∈LPLAN (plan j -Aim j )×PL j ; Where f represents the target value, EPLAN represents the set of early-arriving AUVs, LPLAN represents the set of late-arriving AUVs, and Aim... i This represents the target docking time for the i-th AUV, obtained based on the earliest docking time. i Let represent the docking time of the i-th AUV, and plan represent the docking time of the i-th AUV. j Let Aim represent the docking time of the j-th AUV. j Pe represents the target docking time of the j-th AUV. i PL represents the penalty coefficient for underwater equipment arriving early. j This represents the penalty coefficient for underwater equipment arriving late.

4. The recovery control method for underwater cluster equipment according to claim 2, characterized in that, The docking time set is optimized based on a genetic algorithm and a preset objective function to minimize the total time for each underwater device, resulting in the final docking time set, which includes: A random set of docking times is initialized, and docking times with a fitness greater than a preset fitness are selected from the set of docking times based on the roulette wheel selection method. The iterative execution involves cross-operations between the target docking time and the selected docking time to be processed to obtain the mutated docking time to be processed, and determining the total time consumption corresponding to the new set of docking times, until the set of docking times satisfies the objective function.

5. The method for recovering and controlling underwater cluster equipment according to any one of claims 1-4, characterized in that, Based on the equipment poses of each underwater device at the virtual berth and the docking pose of the docking platform, the target docking path is determined, including: Input the equipment pose and docking pose into multiple preset path calculation formulas to obtain the path length output by each path formula; The path corresponding to the smallest path length extracted from multiple path lengths is taken as the target docking path; Among them, the multiple path calculation formulas include: the first calculation formula, the second calculation formula, the third calculation formula, and the fourth calculation formula. Different path calculation formulas correspond to different docking methods. The first calculation formula includes: 50 LSL =t L +p S +q L =-α+β+p S ; Among them, L LSL t represents the first path length of the first docking method. L p represents the length of the first curve of the first segment in the first docking method. S q represents the length of the first straight line in the middle segment of the first type of docking. L The length of the second curve in the second segment of the first docking method is indicated by α, which represents the angle between the initial heading of the underwater equipment and the direction of the target connection line. The initial heading is the direction of travel at the virtual berth, and the direction of the target connection line is the direction from the virtual berth to the docking platform. β represents the angle between the docking heading of the underwater equipment and the direction of the target connection line. The docking heading is the direction of the docking platform corresponding to the underwater equipment. The second calculation formula includes: L LSR =t L +p S +q R =α-β+2t L +p S ; Among them, L LSR q represents the second path length for the second docking method. R This indicates the length of the second curve in the second segment of the second docking method; The third calculation formula includes: L RSL =t R +p S +q L =-α+β+2t R +p S ; Among them, L RSL t represents the length of the third path in the third docking method. R This indicates the length of the first curve in the first segment of the third docking method; The fourth calculation formula includes: 50 RSR =t R +p S +q R =α-β+p S ; Among them, L RSR This indicates the fourth path length for the fourth docking method.

6. The method for recovering and controlling underwater cluster equipment according to any one of claims 1-4, characterized in that, Based on the equipment poses of each underwater device at the virtual berth and the docking pose of the docking platform, the approach guidance time is determined, including: Determine the average velocity of the underwater equipment under the target docking path, where the target docking path is obtained based on the equipment pose and docking pose; Input the path length of the target docking path and the average speed into the preset approach guidance time calculation formula to obtain the approach guidance time output by the approach guidance time calculation formula; The approach guidance time calculation formula includes: t i =L total-i / V i ; Among them, t i L represents the approach guidance time of the i-th underwater device. total-i V represents the path length of the target docking path for the i-th underwater equipment. i Let represent the average speed of the i-th underwater equipment.

7. The method for recovering and controlling underwater cluster equipment according to any one of claims 1-4, characterized in that, Virtual berths are allocated to various underwater equipment based on the Hungarian algorithm, including: Obtain the equipment location of each underwater device and the berth location of each virtual berth; A cost matrix is ​​constructed based on the equipment location and berth location, and the target matrix corresponding to the cost matrix is ​​solved based on the Hungarian algorithm, wherein the target matrix includes multiple zero elements; Based on the position of the zero element in the target matrix, the correspondence between underwater equipment and berth positions is determined.

8. The recovery control method for underwater cluster equipment according to claim 7, characterized in that, Solving for the target matrix corresponding to the cost matrix using the Hungarian algorithm includes: Subtract the first minimum value from each row of the target matrix and subtract the first minimum value from each column of the target matrix to obtain the matrix to be processed, wherein the first minimum value is obtained from the target matrix; Cover all zero elements in the matrix to be processed using the smallest horizontal and vertical lines; If the first number of horizontal and vertical lines to be covered is less than the second number of underwater equipment, the second minimum value is obtained from the uncovered areas. The second minimum value is then subtracted from each row of the uncovered areas and added to each column of the covered areas until the first number equals the second number, thus obtaining the target matrix.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the recovery control method for underwater cluster equipment as described in any one of claims 1 to 8.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the recovery control method for underwater cluster equipment as described in any one of claims 1 to 8.

Images