Unmanned aerial vehicle maritime search and rescue path planning method based on tabu bee colony algorithm

Abstract

The application discloses a kind of unmanned aerial vehicle offshore search and rescue path planning method based on tabu bee colony algorithm, comprising: initialization task information, from rescue demand and environmental characteristics, construct search and rescue target model and objective function;Initialization constraint condition information, determine the resource constraint and environmental constraint in path planning process, construct constraint description model;The various types of information involved in the process of unmanned aerial vehicle search and rescue are analyzed, and the corresponding capability description model is constructed;According to task information and constraint condition information, the preliminary planning of unmanned aerial vehicle search and rescue path is carried out, and the preliminary search and rescue plan is generated;The objective function is used, and the preliminary search and rescue plan is iteratively updated by combining tabu search and bee colony algorithm until the optimized planning scheme is output.The method proposed in the application improves the exploration ability and convergence speed of the algorithm based on tabu bee colony algorithm, and realizes efficient solution to large-scale and complex offshore search and rescue path planning problem.

Description

Technical Field

[0001] This invention belongs to the field of path planning technology, specifically relating to a method for UAV maritime search and rescue based on the taboo bee colony algorithm, which can be used in maritime search and rescue and path planning scenarios. Background Technology

[0002] Maritime safety is a crucial component of national security emergency response systems in all countries. When incidents such as people falling overboard or vessels being damaged occur, requiring rapid searches of maritime accident sites to ensure personnel safety, relying solely on manual search and rescue operations is often inefficient across vast sea areas. Therefore, the application of drones in maritime rescue is gaining increasing importance. Consequently, developing a rational and efficient search and rescue plan to locate those in the water is paramount for maritime search and rescue. Drones, due to their flexibility, efficiency, and low cost, have already been widely adopted in many scenarios. The problem of drone-based maritime search and rescue path planning is the rational allocation of drone resources across a series of search and rescue missions, ensuring that drones complete tasks within their capabilities and maximize their effectiveness. When the number of search and rescue missions is small, all missions can be accomplished through simple planning.

[0003] Traditional UAV path planning and control models are based on centralized control by ground dispatch and flight control centers. The entire UAV search and rescue process is a closed loop from receiving emergency response requests to proposing reasonable flight path solutions.

[0004] Please see Figure 1 , Figure 1 This is a flowchart illustrating the implementation of a traditional maritime search and rescue route planning method. From... Figure 1 As can be seen, the maritime search and rescue process includes the following steps: ① The maritime search and rescue department first receives the rescue request; ② The maritime search and rescue department preprocesses the search and rescue mission based on meteorological information, converting it into a standard mission information format for path planning; ③ The maritime search and rescue department then combines information from drones and other maritime rescue equipment to perform mission planning, generating the final drone dispatch plan; ④ The maritime search and rescue department sends drone instructions and multi-equipment coordination tasks according to the mission dispatch plan to the command department; ⑤ The command department dispatches drones to carry out search and rescue according to the established plan; ⑥ The drones begin search and rescue; ⑦ The information acquired by the drones is transmitted to the maritime processing platform or command center via a transmission link; ⑧ The command center feeds back the processed rescue information to the user. This completes the closed-loop process of the entire drone search and rescue. Drone path planning mainly refers to steps ②, ③, and ④, which are the responsibility of the maritime search and rescue department.

[0005] from Figure 1It can be seen that this search and rescue process involves multiple units, including users, search and rescue departments, command centers, and data processing centers. Each unit has inconsistent operations, leading to inconvenient interaction. The meteorological and search and rescue resource information used for path planning is not real-time, resulting in a certain lag. Furthermore, my country has a long coastline, and current communication transmission links are not yet perfect. Limited link resources play a crucial role in the entire process, affecting the drone's ability to receive commands and transmit on-site information. Overall, while the traditional drone search and rescue process can acquire some environmental information, it is no longer adequate for new demands and challenges. The drawbacks of the traditional drone search and rescue process can be summarized as follows: the entire planning process is overly complex and cumbersome; the cycle from receiving emergency rescue information to the drone's feedback is often long, failing to meet users' timeliness requirements; drone planning is generally based on offline operations, and the planning scheme cannot adapt to real-time changes in the environment; it relies too heavily on communication links, preventing real-time communication between the drone and the command center; and it lacks auxiliary search and rescue methods and automated information processing processes, often requiring manual correction of information. The entire search process is complex, involving a large amount of human-machine interaction, making it prone to errors.

[0006] It can be seen that throughout the process, the drone only executed instructions such as path adjustment, environmental imaging, and data transmission according to the scheduling command. It could not respond in real time according to changes in the maritime search and rescue environment, which led to the failure of the search and rescue mission. This also reflects that the efficiency of the search and rescue was lower than expected.

[0007] For maritime rescue problems, excessively long solution times are unacceptable in practical search and rescue operations. Currently, there is considerable research on using bee colony algorithms to solve path planning problems or other algorithms to solve maritime rescue problems. For the path planning problem in maritime rescue, researchers have proposed a new form of bee colony algorithm to solve the path planning problem for indoor robots. This algorithm obtains an initial population of feasible paths in a static environment and then improves its superiority by avoiding collisions with dynamic obstacles through local search. Other researchers have used a two-stage bee colony algorithm to generate dynamic paths for mobile robots. First, simple rules are used to generate initial collision-free paths, and then the bee colony algorithm is used to optimize and obtain the robot's real-time path, effectively improving the algorithm's optimization speed and convergence performance. Furthermore, to address the problem that a single UAV cannot quickly cover the search area, a random search heuristic algorithm based on grid-based region decomposition is used to quickly solve the path planning problem for large-scale UAV area coverage. These technological inventions have significant research value and increasingly widespread application value in the fields of maritime rescue and path planning.

[0008] For the maritime search and rescue path planning problem, it is necessary to consider both the timeliness of the UAV search and rescue process and the global optimality of the algorithm during the solution process. Currently, there are many studies on solving the maritime search and rescue path planning problem. In these studies, evolutionary algorithms are commonly used, and these proposed algorithms are efficient in solving path planning for small-scale UAVs in static environments, but they lack a comprehensive understanding of the overall UAV search and rescue process objective. Their limitations mainly lie in: inadequacies in solving large-scale planning tasks and under complex environmental constraints; premature entrapment into local optima; limited search and rescue data; and insufficient consideration of constraints such as power constraints, motion performance constraints, and area coverage search for UAVs in the maritime flight environment. Because the objective function and constraints of the maritime search and rescue path planning problem are extremely complex, resulting in high computational complexity, current solution models and methods have slow convergence speeds, failing to achieve a balance between timeliness and global optimality, and require further improvement. Summary of the Invention

[0009] To address the aforementioned problems in existing technologies, this invention provides a method for unmanned aerial vehicle (UAV) maritime search and rescue path planning based on the taboo bee colony algorithm. Based on establishing a mathematical programming model for the maritime search and rescue path planning problem, the method leverages the taboo bee colony algorithm to enhance the algorithm's exploration capability and convergence speed, achieving efficient solutions to large-scale and complex maritime search and rescue path planning problems. The technical problem to be solved by this invention is achieved through the following technical solution:

[0010] This invention provides a method for unmanned aerial vehicle (UAV) maritime search and rescue path planning based on the taboo bee colony algorithm, including:

[0011] S1: Initialize mission information and construct search and rescue target model and objective function using rescue needs and environmental characteristics;

[0012] S2: Initialize constraint information, determine resource constraints and environmental constraints in the path planning process, and construct a constraint description model;

[0013] S3: Analyze various types of information involved in the drone search and rescue process and construct a drone search and rescue capability model;

[0014] S4: Based on mission information and constraint information, perform preliminary planning of the UAV search and rescue path and generate a preliminary search and rescue plan;

[0015] S5: Using the objective function, combined with tabu search and bee colony algorithm, the preliminary search and rescue plan is iteratively updated until the optimal planning scheme is output.

[0016] In one embodiment of the present invention, the objective function is:

[0017]

[0018] Among them, GT r Let represent the number of drones dispatched by the r-th search and rescue center, Q represent the number of search and rescue centers, SD represent the standard deviation of the number of drones dispatched by the search and rescue centers, α1 represent the influence factor of the number of drones in the objective function, α2 represent the influence factor of the standard deviation of the number of drones, and α1+α2=1.

[0019] In one embodiment of the present invention, S2 includes:

[0020] S2.1: Construct a drone search and rescue capability model by combining the drone working model:

[0021]

[0022] Among them, UAVCap i,j This represents the j-th capability of drone i. in, This represents a model describing the payload capacity of an unmanned aerial vehicle (UAV). Including search and rescue area and resolution, Model for describing the working mode of unmanned aerial vehicles Mode i,j Represents a vector consisting of a series of operations;

[0023] S2.2: Constructing the constraint description model:

[0024]

[0025] Among them, UAVCons i,j Let j represent the j-th constraint of UAV i, including the search and rescue area constraint C, the UAV battery power constraint E, the trajectory constraint L, and the range constraint D.

[0026] In one embodiment of the present invention, the expressions for the search and rescue area constraint, the UAV battery power constraint, the trajectory constraint, and the range constraint are respectively:

[0027] Search and rescue area constraints:

[0028]

[0029] Where Q represents the number of search and rescue centers, T represents the number of missions, and M represents the number of missions. r This indicates the number of drones included in the search and rescue center; This indicates that mission j will be carried out by drone u dispatched by search and rescue center r;

[0030] Drone power constraints:

[0031]

[0032] Where, d oj d represents the distance between task o and task j. ij e represents the distance between task i and task j. u This indicates the power consumption per unit distance for a single drone; Let be the decision variable, representing the outcome if the drone u dispatched by search and rescue center r first performs mission o and then mission j. otherwise Similarly Let be the decision variable, representing the outcome if the drone u dispatched by search and rescue center r performs task i first, and then task j. otherwise E u This indicates the maximum amount of electricity that the drone's battery can store. i(|T|+1) This represents the distance between task i and task |T|+1; Let be the decision variable, representing that if the drone u dispatched by search and rescue center r first executes task i, and then executes task |T|+1, then otherwise

[0033] Track constraints:

[0034]

[0035] Flight range constraints:

[0036]

[0037] In one embodiment of the present invention, the UAV search and rescue capability model includes UAV payload capability and executable mission capability, wherein the UAV payload capability includes airborne payload detection range and optical imaging resolution, and UAV dynamic description; the executable mission capability is used to describe search and rescue mission information, mission response capability, and search and rescue mission evaluation.

[0038] In one embodiment of the present invention, S4 includes:

[0039] S4.1: Perform preprocessing and adjustment on the set of tasks to be planned, and perform coarse planning on the set of tasks to be planned based on the results of the preprocessing and adjustment, and generate pre-planning results;

[0040] S4.2: The UAV search and rescue capability model and the constraint description model are used to plan the UAV flight path in a refined manner to generate a preliminary search and rescue plan.

[0041] In one embodiment of the present invention, S4.1 includes:

[0042] The set of planned tasks is preprocessed and adjusted, including checking the compliance of rescue task standards, merging identical rescue tasks, and splitting regional rescue tasks. Based on the preprocessing results, the set of planned tasks is coarsely planned, that is, according to the order of rescue requests submitted and the urgency of the rescue, without considering resource usage, energy consumption, or drone support, the requests to be arranged are sorted and a preliminary rescue plan is generated, and a preliminary search and rescue plan is produced.

[0043] In one embodiment of the present invention, S4.2 includes:

[0044] Based on the constraints in the constraint description model, and taking into account the planning of each search and rescue drone, the allocation of search and rescue tasks, the battery power and available range of each drone, the pre-planning results are adjusted until the generated scheme can meet the constraints, thereby generating a preliminary search and rescue plan.

[0045] In one embodiment of the present invention, S5 includes:

[0046] S5.1: Upon receiving the location information of the drone's pending mission, the system enters the peak hiring search phase, iterating through the selection algorithm to analyze and evaluate the current pending mission.

[0047] S5.2: The hiring bee search phase iterates repeatedly until the optimal search solution is output, at which point the planning ends.

[0048] In one embodiment of the present invention, S5.2 includes:

[0049] S5.2a: Load various initial information, including the search and rescue center set (Center), the number of tasks (Tasknum), the population size (Population), the population scale (LPopulation), the taboo list length (TBL), the swap segment length (GenL), and the control parameter (Thre).

[0050] S5.2b: Calculate population fitness based on the drone's flight path, flight time, and the arrangement of each drone.

[0051] S5.2c: Based on the population-based strategy, the fitness value of the current population satisfies the Local... i ≥Local i-1 Then, the current optimal function value replaces the original optimal function value, and the process proceeds to S5.2d;

[0052] S5.2d: Based on step S5.2c, obtain the current best individual to replace the original best individual Population. i ;

[0053] S5.2e: Adjust and optimize according to the tabu policy, if the tabu list length is TBL i If the length TBL is greater than the set maximum length, the taboo list is cleared, and the optimal sequence is maintained until the entire process is completed, generating the final path planning scheme.

[0054] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0055] 1. This invention takes into account the constraints of "timeliness" and regional coverage, as well as many practical factors such as the power constraints, motion performance constraints, and regional characteristics of UAVs in the maritime flight environment. It proposes an optimization function with the goal of minimizing the overall search and rescue execution time and balancing the task allocation among various search and rescue centers, taking into account various constraints such as UAV power and the comprehensive search of the mission area.

[0056] 2. The method of this invention comprehensively considers constraints such as UAV battery power and comprehensive search of the mission area, constructs a constrained description model and objective function for the maritime search and rescue path planning problem, studies the mathematical programming model of the maritime search and rescue path planning problem, and aims to maximize search and rescue efficiency under the premise of optimal planning scheme. Statistical analysis and evaluation are performed on key indicators, namely: mission response data, search and rescue response time, and resource utilization. The method proposed in this invention is based on the tabu-based bee colony algorithm. Through an improved algorithm combining tabu strategies and the bee colony algorithm, the algorithm's ability to solve complex combinatorial optimization problems is enhanced, avoiding the algorithm getting trapped in local optima, improving the algorithm's exploration ability and convergence speed, and achieving efficient solutions to large-scale and complex maritime search and rescue path planning problems.

[0057] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0058] Figure 1 This is a flowchart illustrating the implementation of a traditional maritime search and rescue route planning method.

[0059] Figure 2 This is a flowchart of a method for unmanned aerial vehicle (UAV) maritime search and rescue path planning based on the taboo bee colony algorithm provided in an embodiment of the present invention;

[0060] Figure 3 This is a schematic diagram illustrating the processing steps of a method for unmanned aerial vehicle (UAV) maritime search and rescue based on the taboo bee colony algorithm, provided in an embodiment of the present invention.

[0061] Figure 4 This is a schematic diagram of a search and rescue target model provided in an embodiment of the present invention;

[0062] Figure 5 This is a schematic diagram of a search and rescue capability model of a drone provided in an embodiment of the present invention;

[0063] Figure 6 This is a schematic diagram of constraint relationships in the path planning process of an unmanned aerial vehicle (UAV) provided by an embodiment of the present invention;

[0064] Figure 7 This is a task and UAV allocation constraint relationship diagram provided in an embodiment of the present invention;

[0065] Figure 8 This is a flowchart of an algorithm evaluation provided by an embodiment of the present invention;

[0066] Figure 9 This is a basic flowchart of task execution during a search and rescue process provided by an embodiment of the present invention;

[0067] Figure 10 This is a schematic diagram of UAV path planning before and after a task change, provided by an embodiment of the present invention;

[0068] Figure 11 This is a schematic diagram of a new task replanning provided in an embodiment of the present invention;

[0069] Figure 12 This is a flowchart of a bee colony algorithm based on tabu search provided in an embodiment of the present invention;

[0070] Figure 13 This is a comparison chart of the average values ​​of the results from various algorithms. Detailed Implementation

[0071] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following detailed description of the UAV maritime search and rescue path planning method based on the taboo bee colony algorithm proposed in accordance with the present invention is provided in conjunction with the accompanying drawings and specific embodiments.

[0072] The foregoing and other technical contents, features, and effects of the present invention will be clearly presented in the following detailed description of specific embodiments in conjunction with the accompanying drawings. Through the description of the specific embodiments, a more in-depth and concrete understanding can be gained of the technical means and effects adopted by the present invention to achieve its intended purpose. However, the accompanying drawings are for reference and illustration only and are not intended to limit the technical solutions of the present invention.

[0073] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations are intended to cover non-exclusive inclusion, such that an article or apparatus comprising a list of elements includes not only those elements but also other elements not expressly listed. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or apparatus that includes said element.

[0074] The complexity of the marine environment and the uncertainty of weather pose significant challenges to UAV-based maritime rescue. To formulate reasonable and efficient search and rescue plans, it is necessary to optimize the algorithm for solving the UAV rescue path planning problem to achieve high timeliness and stability. This invention, based on a mathematical programming model of the maritime search and rescue path planning problem, utilizes the taboo bee colony algorithm to enhance the algorithm's exploration capability and convergence speed, achieving efficient solutions to large-scale and complex maritime search and rescue path planning problems. Please see [link to related documentation]. Figure 2 and Figure 3 The UAV maritime search and rescue path planning method based on the taboo bee colony algorithm in this embodiment specifically includes the following steps:

[0075] S1: Initialize mission information and construct search and rescue target model and objective function using rescue needs and environmental characteristics.

[0076] In this step, it is first necessary to clarify the basic needs of search and rescue, taking the rescue needs as the guide and the rescue mission as the center, to assist in completing rapid, accurate, scientific, and efficient situation analysis and prediction. Furthermore, due to the complex environmental applications during the movement of UAVs at sea, the needs analysis phase also needs to consider various influencing factors in the UAV search and rescue process, such as the characteristics of the marine environment and the UAV search and rescue mission. Therefore, this embodiment of the invention starts from the characteristics of UAV planning in the needs analysis phase and defines the influencing parameters in the path planning process, including: target area coordinates (lon). i ,lat i ), drone flight speed (v u ), Power consumption per unit distance of a single drone (e u ), the maximum amount of electricity that a battery can store (E) u ), the distance between different search and rescue missions (D) ij )wait.

[0077] In this embodiment, the UAV path planning problem is described as having a given number of search and rescue centers of |Q|, each search and rescue center including |M rWhen a drone is deployed to perform a search and rescue mission, it needs to select a search and rescue center q from the pool of search and rescue centers. r Its coordinates are (Lon r ,Lat r ), and from the search and rescue center q r Select drone m u According to the velocity v u Fly to the coordinates (lon) in the task set T i ,lat i ) task t i The Euclidean distance between task i and task j in the given region is D. ij During this process, each drone can perform multiple tasks only if its battery power is sufficient to return to the search and rescue center. Therefore, if a drone is unable to return to the search and rescue center due to performing the next task, it must abandon the task and return directly to the search and rescue center.

[0078] Furthermore, due to the complexity of the maritime environment and the dynamic response of the mission, the search and rescue path planning process needs to consider the following preconditions: the entire search and rescue area is divided into several search and rescue task areas, and the search and rescue area and the tasks in these areas will not change throughout the entire search and rescue process; the UAV, from its departure from the search and rescue center to the task area to complete the search and rescue mission and return to the search and rescue center, will not be affected by external factors such as sea conditions or weather, nor will it experience unexpected situations such as malfunction or loss of control due to its own reasons; a search and rescue center can dispatch multiple UAVs simultaneously; a UAV can only conduct search and rescue operations in one task area at a time. These preconditions can reduce the possibility of ineffective algorithm evaluation due to uncertainties.

[0079] The initial mission information phase requires starting with rescue needs and environmental characteristics to construct a search and rescue target model and objective function. The objectives of drone rescue operations can be described from two perspectives: one is from the perspective of "timeliness," focusing on how to optimize the allocation of existing resources for a specific task to achieve rapid rescue operations; the other is from the perspective of "search area," studying how to allocate available resources to cover the area where the accident occurred to meet as many rescue needs as possible, such as... Figure 4 As shown.

[0080] This invention divides UAV flight into three stages: departure from the search center to the mission area, flight between mission areas, and return from the mission area to the search and rescue center. The search and rescue target model can be used to solve the path planning problem using a graph model. A directed graph (N,E) represents a network graph where a UAV departs from a search and rescue center, arrives at a mission area to perform one or more tasks, and returns. Here, node N = Q∪T represents the search and rescue center and the tasks (Q represents the number of search and rescue centers, and T represents the number of tasks), and E represents the edges of the directed graph, indicating the task nodes that need to be considered in the search and rescue path.

[0081] The objective of this invention is to find the optimal path planning scheme to minimize the overall search and rescue execution time, while maintaining a roughly balanced task allocation across search and rescue centers. The workload is measured by the number of drones dispatched by each search and rescue center, and the balance of task allocation is verified by calculating the standard deviation of the number of drones dispatched by each center. Furthermore, the final execution time is determined by the maximum flight time of the drones dispatched by each search and rescue center. The objective function in this embodiment is composed of two optimization dimensions, weighted by a sum of coefficients:

[0082]

[0083] Among them, GT r Let represent the number of drones dispatched by the r-th search and rescue center, SD represent the standard deviation of the number of drones dispatched by the search and rescue center, α1 represent the influence factor of the number of drones in the objective function, α2 represent the influence factor of the standard deviation of the number of drones, and α1+α2=1.

[0084] It should be noted that before step S1, it is necessary to first clarify the parameters describing the search and rescue mission, including drone information, search and rescue center information, and environmental information, considering the rescue request and the execution of the rescue mission. Furthermore, due to the complexity of the environment during the movement of drones at sea, various influencing factors during the drone search and rescue process, such as the marine environment and the characteristics of the drone search and rescue mission, must also be considered during the requirements analysis phase.

[0085] S2: Initialize constraint information, determine resource constraints and environmental constraints in the path planning process, and construct a constraint description model.

[0086] Step S2 in this embodiment specifically includes:

[0087] S2.1: Construct a drone search and rescue capability model by combining the drone working model:

[0088]

[0089] Among them, UAVCap i,j UAVCap represents the j-th capability of drone i.i,j The ability can be derived from Composed of binary pairs, where This represents a model describing the payload capacity of an unmanned aerial vehicle (UAV). Including search and rescue range and resolution, different configurations of drones have different payload capabilities. Model for describing the working mode of unmanned aerial vehicles Mode i,j This represents a vector consisting of a series of operations.

[0090] S2.2: Constructing the constraint description model:

[0091]

[0092] Among them, UAVCons i,j Let UAV_i represent the j-th constraint condition. For different UAVs, UAV_i = UAV_j. i,j The constraints can be represented by the quadruple {C,E,L,D}, where C represents the search and rescue area constraint, E represents the UAV battery power constraint, L represents the trajectory constraint, and D represents the range constraint. The expressions for the search and rescue area constraint, UAV battery power constraint, trajectory constraint, and range constraint are as follows:

[0093] Search and rescue area constraints:

[0094]

[0095] Where Q represents the number of search and rescue centers, T represents the number of missions, and M represents the number of missions. r This indicates the number of drones included in the search and rescue center; This indicates that mission j is carried out by drone u dispatched by search and rescue center r.

[0096] Drone power constraints:

[0097]

[0098] Where, d oj d represents the distance between task o and task j. ij e represents the distance between task i and task j. u This indicates the power consumption per unit distance for a single drone; Let be the decision variable, representing the outcome if the drone u dispatched by search and rescue center r first performs mission o and then mission j. otherwise Similarly Let be the decision variable, representing the outcome if the drone u dispatched by search and rescue center r performs task i first, and then task j. otherwise E uThis indicates the maximum amount of electricity that the drone's battery can store. i(|T|+1) This represents the distance between task i and task |T|+1; Let be the decision variable, representing that if the drone u dispatched by search and rescue center r first executes task i, and then executes task |T|+1, then otherwise

[0099] Track constraints:

[0100] Flight range constraints:

[0101] S3: Analyze various types of information involved in the drone search and rescue process, and construct a drone search and rescue capability model.

[0102] In this step, based on the basic requirements, objectives, and models of the path planning problem, we conduct in-depth analysis and research on various types of information involved in the UAV search and rescue process, and construct a corresponding search and rescue capability description model. This model is represented by characteristics of both the UAV's payload capacity and mission capabilities. Specifically, this includes:

[0103] S3.1: Construct a UAV search and rescue capability model by combining UAV payload and mission mode. The UAV search and rescue capability model describes the UAV's capabilities using UAV payload capacity and mission capabilities, as detailed below:

[0104] Please see Figure 5 , Figure 5 This is a schematic diagram of a search and rescue capability model for a drone provided in an embodiment of the present invention. The drone search and rescue capability model is defined in this embodiment as describing the drone's payload capacity and its ability to perform tasks. The drone's payload capacity includes the onboard payload detection range and optical imaging resolution. Taking a fixed-wing search and rescue drone as an example, the payload capacity model also includes a description of the drone's dynamics, described using information such as the speed of each axis and yaw angle. The drone's task capability mode in this embodiment is a model used to describe search and rescue task information (including target location, target characteristics, etc.), task response capability (shortest search and rescue time), and search and rescue task evaluation (combined search and rescue time and overall search and rescue task completion rate). Through the establishment of these two models (the drone search and rescue capability model and the constraint description model), this invention achieves refined management of search and rescue activities and improves search and rescue efficiency.

[0105] S3.2: Define the constraints that need to be considered during the UAV path planning process.

[0106] This invention defines the constraints that need to be considered in the UAV path planning problem, including search and rescue area constraints, UAV battery power constraints, trajectory constraints, and range constraints. Please refer to [link / reference]. Figure 6, Figure 6 This is a schematic diagram illustrating the constraint relationships in the UAV path planning process provided by an embodiment of the present invention. In this embodiment, the requirement for each UAV to depart from the search and rescue center, complete its mission, and return to the search and rescue center is represented as a trajectory constraint. Furthermore, considering the urgency of the situation, the search and rescue area restrictions for each search and rescue mission and the requirement that all search and rescue areas must be covered represent the search and rescue area constraint. Throughout this entire search and rescue process, battery power constraints and the UAV's maximum range constraints are applied. Specifically, the expressions for the search and rescue area constraint, UAV battery power constraint, trajectory constraint, and range constraint are as described above.

[0107] S4: Based on mission information and constraint information, perform preliminary planning of the UAV search and rescue path and generate a preliminary search and rescue plan.

[0108] Having clarified the basic objectives and constraints of the rescue problem through the above steps, we can now proceed to the task planning phase using a planning algorithm. This step specifically includes:

[0109] S4.1: Perform preprocessing and adjustment on the set of tasks to be planned, and perform coarse planning on the set of tasks to be planned based on the preprocessing results, and generate pre-planning results.

[0110] Specifically, the set of planned tasks undergoes preprocessing and adjustment, including checking the compliance of rescue mission standards (drone requirements, coordinates, rescue targets, rescue time, urgency, etc.), merging identical rescue missions, and splitting regional rescue missions. Based on the preprocessing results, the set of planned tasks is coarsely planned. That is, according to the order of submission of rescue requests and the urgency of the rescue, without considering resource usage, energy consumption, or drone support, the requests to be arranged are sorted and a preliminary rescue plan is generated, producing a pre-planning result.

[0111] S4.2: The drone's flight path is planned in detail using algorithms (drone search and rescue capability model and constraint description model) to generate a preliminary search and rescue plan.

[0112] Although the search and rescue task set was coarsely divided, the planning results still did not clearly reflect the overall effectiveness of the search and rescue plan. Therefore, an algorithm was used to fine-tune the drone flight paths. Several constraints were defined earlier in the path planning process. Clearly, under these constraints, the resource selection and planning results differ. Based on these constraints, and considering the planning of each search and rescue drone, the allocation of search and rescue tasks, the battery power of each drone, and its available range, the pre-planning results were adjusted until a usable plan satisfied the constraints was generated. Please refer to [link to relevant documentation]. Figure 7 , Figure 7 This is a task and UAV allocation constraint relationship diagram provided in an embodiment of the present invention. Figure 7 Starting from the allocation of UAV resources, and based on the basic characteristics of the mission, the constraints for generating UAV flight paths are further described. Path planning requires consideration of UAV payload and capability constraints. UAVs with different payloads perform different tasks; for example, the order in which a UAV carrying an imaging payload and a UAV carrying rescue equipment arrive at the target for rescue needs to be considered during path planning. Furthermore, different rescue needs place different capability requirements on the UAVs. Moreover, during the execution of mission j, UAV i needs to consider not only the power-on / off actions of its payload and the timing of the deployment of rescue supplies, but also the UAV's attitude maneuvers based on the actual situation. Therefore, the planning scheme also needs to combine the actual UAV capabilities and the environment to reasonably generate the action sequence. Generally, given the UAV resource allocation, the planned path satisfying the constraints is not unique. After preliminary planning by the algorithm, a preliminary search and rescue plan is generated.

[0113] The above steps complete the task preprocessing and algorithm invocation planning processes. The next step is the optimization process; please refer to [link to relevant documentation]. Figure 8 and Figure 9 , Figure 8 This is a flowchart of an algorithm evaluation provided by an embodiment of the present invention. Figure 9 This is a basic flowchart of mission execution during a search and rescue process provided by an embodiment of the present invention. In this flowchart, the dashed line A represents the initialization sequence, and the solid line B represents the flight path sequence that is replanned after evaluation. This process uses constraints to repeatedly calculate and iterate the candidate action set for the UAV to perform the mission, thereby achieving the goal of globally optimizing the UAV path.

[0114] After initial planning, the next step is to execute response actions according to the scheduling sequence. However, environmental changes, unforeseen circumstances, and economic considerations can cause the pre-set scheduling scheme to fail to meet actual needs, resulting in reduced search and rescue efficiency. To avoid this, the planning system needs to dynamically adjust actions based on current environmental characteristics to achieve optimal search and rescue benefits.

[0115] It should be noted that during the path planning process, unforeseen circumstances may necessitate adjustments to the original plan. Therefore, this process needs to exclude certain special cases to ensure the system is unaffected by unforeseen factors. Generally speaking, the constraints to be considered during the planning process remain the same. First, the entire search and rescue area is divided into several task areas to be searched, and these areas and tasks will not change throughout the search and rescue process. Second, the drone's journey from the search and rescue center to the task area, and back to the search and rescue center, will not be affected by external factors such as sea conditions or weather, nor will it experience malfunctions or loss of control due to its own inherent limitations. Third, a search and rescue center can dispatch multiple drones simultaneously. Finally, a single drone can only search one task area at a time.

[0116] When determining a flight path, an initial flight path input is required first. This input serves as the trigger for UAV path optimization, and the triggering conditions need to be set in advance before the UAV scheduling center begins planning and scheduling. During the UAV path optimization and scheduling process, once the replanning conditions are met, the scheduling center will complete the UAV flight path rescheduling response according to the system settings. Generally, the situations requiring replanning during UAV replanning can be divided into the following two categories:

[0117] 1) Task information changes. After the dispatch center corrects the previous task information based on real-time on-site dynamics, if the task information changes, replanning is required. For the task to be executed, due to limitations in the communication link between the UAV and the command center, the coordinate position of task T1 at a certain point in time may be incorrect. In this case, the UAV path needs to be replanned based on the actual location of the task to be rescued. During this process, due to the positional deviation, the original UAV performing the task may be replaced by another UAV. Therefore, the rescue paths of the UAVs involved in the task need to be replanned, and a UAV rescue path that meets the conditions needs to be generated. Please refer to [link to relevant documentation]. Figure 10 , Figure 10 This is a schematic diagram of UAV path planning before and after a mission change, provided by an embodiment of the present invention. Figure 10 As shown, the location information of Mission 5 has changed. According to the planning results, UAV 2, which was to carry out the search and rescue mission, can no longer continue to perform the mission, and UAV 1 will take over the mission.

[0118] 2) New Tasks. During search and rescue operations, unforeseen circumstances may arise, necessitating changes to the drone's flight path. In such cases, the path needs to be replanned to maximize effectiveness. Please refer to [link / reference]. Figure 11 , Figure 11 This is a schematic diagram of a new task replanning provided by an embodiment of the present invention, such as... Figure 11 As shown, based on the information feedback from the maritime rescue command center, a new rescue target is discovered. At this point, it is necessary to consider the execution of the new rescue target. Furthermore, it is necessary to consider whether there are any drones that can perform the mission under the premise of meeting the constraints. Therefore, it is necessary to replan the drone path, reasonably adjust the current path of the drone, and generate a drone rescue path that meets the conditions.

[0119] S5: Using the taboo bee colony algorithm and the objective function, iteratively update the preliminary search and rescue plan until the optimal planning scheme is output.

[0120] After clarifying the factors to be considered in task planning and the conditions for replanning, the specific planning method is described next. In this embodiment of the invention, tabu search and bee colony algorithm are combined as the core planning module to solve the path planning decision in the UAV path planning problem.

[0121] Tabu Search, proposed by Professor Fred Glover of the University of Colorado, is a solution method that uses continuous algorithmic changes to escape local optima. Bee Colony Algorithm (BSA) is an optimization algorithm that mimics the behavior of bees. Current algorithms for solving maritime path planning problems, particularly those dealing with large-scale planning tasks and complex environmental constraints, are insufficient, often getting trapped in local optima prematurely. This BSA-based BSA avoids this pitfall when solving complex combinatorial optimization problems. Furthermore, unlike traditional algorithms, this algorithm addresses the shortcomings of traditional algorithms for large-scale planning tasks and complex environmental constraints, making it applicable to maritime rescue planning scenarios and ultimately yielding the optimal planning solution.

[0122] Both bee colony optimization (BCO) and tabu search are classic path planning algorithms. BCO, a typical swarm intelligence algorithm, is built by simulating the collective behavior of bees foraging for nectar. Tabu search is a metaheuristic algorithm for solving optimization problems. Tabu is its most unique feature; during the optimization process, it discards previously encountered population features based on tabu strategies, effectively avoiding getting trapped in local optima. By using a tabu list to record previously found optimal solutions, and avoiding searching for optimal conditions in the tabu list in subsequent searches, it achieves the effect of avoiding getting trapped in local optima.

[0123] The advantage of the tabu search-based bee colony algorithm (TBBA, or tabu ant colony algorithm for short) is that by using the tabu search algorithm, the bee colony algorithm is guided to avoid the trap of local optima. This makes up for the shortcomings of the bee colony algorithm, such as poor local search ability, low convergence accuracy, and easy precession convergence. It can improve the convergence speed of the algorithm and realize rapid planning of search and rescue paths.

[0124] Please see Figure 12 , Figure 12 This is a flowchart of a bee colony algorithm based on tabu search provided in an embodiment of the present invention. The solution process of the bee colony algorithm based on tabu search includes the following steps:

[0125] 1) Initialize parameters and generate initial solutions. Load various initial information for maritime search and rescue path planning, including the search and rescue center set, number of tasks, population size, taboo list length, interchange segment length, and control parameters.

[0126] 2) Reconnaissance Peak Search Phase. During this phase, the reconnaissance peak search is used to select which drone to perform which rescue mission, and the search results are obtained.

[0127] 3) Receive the results of the reconnaissance peak search, mark some objects corresponding to the searched local optima, and try to avoid roundabout searches. Furthermore, a tabu-based strategy is considered to determine whether the current search process is within the tabu list; if so, the search is repeated; otherwise, the tabu strategy table information is updated to clarify the UAV execution schedule.

[0128] 4) Based on the search results from the previous step, update the taboo table and store the location information of the drone mission to be performed.

[0129] 5) Hired Swarm Search Phase. Based on the location information provided by the reconnaissance swarms, further exploration of the local space aims to improve algorithm performance. This process iterates through all eligible drones, calculates the objective function value obtained by different drones performing the task, and selects the final drone flight path based on the calculations.

[0130] 6) Determine if all tasks have been assigned. If all tasks have been assigned, end the process; otherwise, return to step 2.

[0131] The most important idea behind tabu search is to mark objects corresponding to locally optimal solutions already searched, thus avoiding roundabout searches as much as possible. In the scout bee search phase, this embodiment of the invention uses heuristic rules to generate an initial population, sorting the mission area from nearest to farthest according to its Euclidean distance from the search center, and generating chromosome sequences based on this sorting result. Before the scout bee search, the initial population is replicated, resulting in two populations with identical chromosome sequences. Then, a search is performed on each individual in the replicated population, and the length of the gene exchange segments is set.

[0132] Regarding the selection of the starting point for a set of interchange segments for each individual in the population, this patent considers and avoids two possible scenarios: whether the gene length of the remaining chromosome, including the starting point, meets the interchange length requirement, and whether a set of interchange segments overlaps. After each individual in the replicated population performs a search, the two interchange segment sequences of each individual are placed in a taboo list for comparison with interchange segments generated by subsequent new individuals, thus avoiding duplicate sequences through taboo search. If a new individual generates the same interchange segment as in the taboo list, a new set of interchange segments is generated for this individual and compared again with all existing interchange segments in the taboo list until no duplicates are found.

[0133] Next, the two interchange segments of each individual in both the original and replicated populations are recorded in the tabu list. Simultaneously, the path planning solution for each individual in the original population is recorded, and the fitness function corresponding to that path plan is calculated. The minimum fitness function of all individuals in both populations is recorded as the optimal value for the current generation, and the individual corresponding to this fitness function value is found. The population and reward are concatenated, and the fitness functions are sorted from smallest to largest. A new population is formed by randomly selecting the same number of individuals as the original population size after this selection process.

[0134] In the algorithm's hired bee phase, hired bees decide whether to follow scout bees to conduct a local search around the nectar source based on the quality of the nectar brought back by the scout bees. After evaluating the nectar in the flower bed, the hired bees roulette to choose the scout bee's nectar source and then search for new nectar sources around the chosen one. If the quality of the new nectar source found by the hired bee is better than that of the scout bee's nectar source, the hired bee becomes a scout bee, replacing the original best individual in the population. At the same time, new hired bees can be recruited to start the next round of searching.

[0135] Preferably, step S5 in this embodiment specifically includes:

[0136] S5.1: Upon receiving the location information of the drone's pending mission, the system enters the peak hiring search phase, iterates the selection algorithm, and analyzes and evaluates the current pending mission.

[0137] S5.1a) Based on the number of tasks currently completed by UAVComp i Based on the search and rescue request (RescueDemand) received by the search and rescue center i Search and rescue requests arranged by the command center. i By combining these three factors, the response requirements of the search and rescue center and the actual response requirements of the drones during their operation can be determined.

[0138] S5.1b) Based on the time it takes for UAV i to complete the search and rescue mission (UAVTaskTime) i Based on the search and rescue request (RescueDemand) received by the search and rescue centeri Combining the two factors, we analyze the timeliness of drone path planning from the perspective of drone search and rescue timeliness;

[0139] S5.1c) After analyzing the timeliness of drone search and rescue, this process will analyze the satisfaction of drone rescue missions and output the number of rescue missions that each drone can complete.

[0140] S5.2: The search phase involving hired bees is iterated repeatedly until the optimal planning solution is output, at which point the planning process ends;

[0141] S5.2a: Load various initial information, including the search and rescue center set (Center), the number of tasks (Tasknum), the population size (Population), the population scale (LPopulation), the taboo list length (TBL), the swap segment length (GenL), and the control parameter (Thre).

[0142] S5.2b: Calculate population fitness based on the drone's flight path, flight time, and the arrangement of each drone.

[0143] S5.2c: Based on the population-based strategy, the fitness value of the current population satisfies the Local... i ≥Local i-1 Then, the current optimal function value replaces the original optimal function value, and the process proceeds to S5.2d;

[0144] S5.2d: Based on step S5.2c, obtain the current optimal individual to replace the original optimal individual Population under the current circumstances. i ;

[0145] S5.2e: Adjust and optimize according to the tabu policy, if the tabu list length is TBL i If the length TBL is greater than the set maximum length, the taboo list is cleared, and the optimal sequence is maintained until the entire process ends, generating the final action sequence.

[0146] The effectiveness of the UAV maritime search and rescue path planning method based on the taboo bee colony algorithm of this invention can be further illustrated by the following comparative simulation experiments:

[0147] (I) Simulation Conditions

[0148] All simulation experiments conducted in this invention were performed on a computer with a Core™ i5-8265U CPU 1.60GHz, Windows 10 operating system, and Matlab 2020a coding environment.

[0149] (II) Simulation Content

[0150] This invention employs randomly generated scenarios for experimentation. The dataset allows for comprehensive evaluation of the algorithm's performance; therefore, this embodiment sets the task size to 100, 200, 300, and 400, with 3 search and rescue centers. To concisely and clearly represent the test set, we use the "T\C" format to reflect the task size and the number of search and rescue centers, where T represents the number of tasks and C represents the number of search and rescue centers. A larger ratio of the number of tasks to the number of search and rescue centers indicates a larger problem scale, and correspondingly, a greater difficulty in path planning.

[0151] For the comparison algorithm, evolutionary algorithm and search algorithm were used as comparison algorithms to verify the problem-solving performance of the TBBA algorithm in the embodiment of the present invention. Genetic algorithm (GA) constructs solutions based on spatial search, while neighborhood search algorithm (NS) finds better search solutions through continuous improvement of neighborhood structure. We set each test set to run 10 times, and recorded the minimum value (denoted as Min), average value (denoted as Ave), and standard deviation (denoted as Std) of each algorithm in the simulation experiment results table as indicators to evaluate the search performance of each algorithm.

[0152] Table 1. Experimental results for different task sizes

[0153]

[0154] The data comparison in the simulation experiment table reveals that, for different task scales, the TBBA algorithm proposed in this embodiment outperforms the comparative algorithms in all three aspects of the planning results. It can be seen that the algorithm in this embodiment can effectively solve the large-scale problem of increasing search and rescue missions, obtain better search and rescue scheduling and planning schemes, and provide auxiliary decision support for actual rescue operations.

[0155] For each scale of task quantity, a comparison chart of the average values ​​of the results from various algorithms is provided, such as... Figure 13 As shown, in all the test sets, the TBBA algorithm proposed in this embodiment can find the task scheduling and planning scheme more quickly, which is significantly better than the two comparison algorithms. This also reflects the superiority of the algorithm proposed in this embodiment, which can effectively handle regional searches including large-scale tasks.

[0156] Experimental results show that the UAV maritime search and rescue path planning method based on the tabu bee colony algorithm proposed in this invention can effectively solve the maritime rescue path planning problem, and its algorithmic benefits are significantly better than those of the comparison algorithm. During the scheduling process, the application of the tabu algorithm improves scheduling efficiency and reduces response time during search and rescue operations.

[0157] In summary, this invention addresses both timeliness and regional coverage, considering numerous practical factors such as the power constraints, motion performance constraints, and regional characteristics of UAVs in the maritime flight environment. It proposes an optimization function aimed at minimizing the overall search and rescue execution time and achieving balanced task allocation across search and rescue centers, taking into account various constraints such as UAV power consumption and the comprehensive search of the mission area. The method of this invention comprehensively considers constraints such as UAV power consumption and comprehensive mission area search, constructing a constraint description model and objective function for the maritime search and rescue path planning problem. It studies the mathematical programming model of the maritime search and rescue path planning problem, aiming to maximize search and rescue efficiency under the premise of optimal planning. Statistical analysis and evaluation are performed on key indicators, namely: mission response data, search and rescue response time, and resource utilization. The proposed method is based on the tabu-based bee colony algorithm. Through an improved algorithm combining tabu strategies and the bee colony algorithm, it enhances the algorithm's ability to solve complex combinatorial optimization problems, avoids getting trapped in local optima, improves the algorithm's exploration ability and convergence speed, and achieves efficient solutions to large-scale and complex maritime search and rescue path planning problems.

[0158] Another embodiment of the present invention provides a storage medium storing a computer program for executing the steps of the UAV maritime search and rescue path planning method based on the tabu bee colony algorithm described in the above embodiments. A further aspect of the present invention provides an electronic device including a memory and a processor. The memory stores a computer program, and the processor, when calling the computer program in the memory, implements the steps of the UAV maritime search and rescue path planning method based on the tabu bee colony algorithm described in the above embodiments. Specifically, the integrated modules implemented as software functional modules can be stored in a computer-readable storage medium. The software functional modules stored in a storage medium include several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) or processor to execute some 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 a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0159] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A method for unmanned aerial vehicle (UAV) maritime search and rescue path planning based on the taboo bee colony algorithm, characterized in that, include: S1: Initialize mission information and construct search and rescue target model and objective function using rescue needs and environmental characteristics; S2: Initialize constraint information, determine resource constraints and environmental constraints in the path planning process, and construct a constraint description model; S3: Analyze various types of information involved in the drone search and rescue process and construct a drone search and rescue capability model; S4: Based on mission information and constraint information, perform preliminary planning of the UAV search and rescue path and generate a preliminary search and rescue plan; S5: Using the objective function, combined with tabu search and bee colony algorithm, the preliminary search and rescue plan is iteratively updated until the optimal planning scheme is output; The objective function is: ， in, Indicates the first r The number of drones dispatched by each search and rescue center Q Indicates the number of search and rescue centers. This represents the standard deviation of the number of drones dispatched by the search and rescue center. This represents the influence factor of the number of drones in the objective function. The influencing factor representing the standard deviation of the number of drones. ; The UAV search and rescue capability model is based on the description of UAV payload capacity and executable mission capability. The UAV payload capacity includes the onboard payload detection range and optical imaging resolution, and the UAV dynamics description. The executable mission capability is used to describe search and rescue mission information, mission response capability, and search and rescue mission evaluation. S4 includes: S4.1: Perform preprocessing and adjustment on the set of tasks to be planned, and perform coarse planning on the set of tasks to be planned based on the results of the preprocessing and adjustment, and generate pre-planning results; S4.2: The UAV search and rescue capability model and the constraint description model are used to perform refined planning of the UAV flight path and generate a preliminary search and rescue plan; S4.1 includes: The set of planned tasks is preprocessed and adjusted, including checking the compliance of rescue task standards, merging identical rescue tasks, and splitting regional rescue tasks. Based on the preprocessing results, the set of planned tasks is coarsely planned, that is, according to the order of rescue requests and the urgency of the rescue, without considering resource usage, energy consumption, or drone support, the requests to be arranged are sorted and a preliminary rescue plan is generated, and a preliminary search and rescue plan is produced. S4.2 includes: Based on the constraints in the constraint description model, and taking into account the planning of each search and rescue drone, the allocation of search and rescue tasks, the battery power and available range of each drone, the pre-planning results are adjusted until the generated scheme can meet the constraints, thereby generating a preliminary search and rescue plan.

2. The method for UAV maritime search and rescue path planning based on taboo bee colony algorithm according to claim 1, characterized in that, S2 includes: S2.1: Constructing a UAV capability model by combining the UAV working model: ， in, Indicates drone i The j One ability, = ,in, This represents a model describing the payload capacity of a drone. Including search and rescue area and resolution, Model for describing the working mode of unmanned aerial vehicles , Represents a vector consisting of a series of operations; S2.2: Constructing the constraint description model: ， in, Indicates drone i The j Several constraints, including search and rescue area constraints. C Drone power constraints E Track constraints L and range constraints D .

3. The method for UAV maritime search and rescue path planning based on taboo bee colony algorithm according to claim 2, characterized in that, The expressions for the search and rescue area constraint, the UAV battery power constraint, the flight path constraint, and the flight range constraint are as follows: Search and rescue area constraints: ， in, Indicates the number of search and rescue centers. Indicates the number of tasks. This indicates the number of drones included in the search and rescue center; Indicates task j By Search and Rescue Center r dispatched drones u implement; Drone power constraints: in, Indicates task With the task The distance between them Indicates task With the task The distance between them This indicates the power consumption per unit distance for a single drone; Let be the decision variable, representing the decision if made by the search and rescue center. dispatched drones Execute the task first Then execute the task. ,but ,otherwise Similarly Let be the decision variable, representing the decision if made by the search and rescue center. dispatched drones Execute the task first Then execute the task. ,but ,otherwise ; This indicates the maximum amount of electricity that the drone's battery can store; Indicates task With the task The distance between them; Let be the decision variable, representing the decision if made by the search and rescue center. dispatched drones Execute the task first Then execute the task. ,but ,otherwise ; Track constraints: ； Flight range constraints: 。 4. The method for UAV maritime search and rescue path planning based on taboo bee colony algorithm according to claim 3, characterized in that, S5 includes: S5.1: Upon receiving the location information of the drone's pending mission, the system enters the peak hiring search phase, iterating through the selection algorithm to analyze and evaluate the current pending mission. S5.2: The hiring bee search phase iterates repeatedly until the optimal search solution is output, at which point the planning ends.

5. The method for UAV maritime search and rescue path planning based on taboo bee colony algorithm according to claim 4, characterized in that, S5.2 includes: S5.2a: Load various initial information and input it into the search and rescue center set. Number of tasks Population size Population size Length of the taboo table Swap segment lengths , control parameter Thre; S5.2b: Calculate population fitness based on the drone's flight path, flight time, and the arrangement of each drone. ; S5.2c: Based on the strategy implemented for the population, the fitness value of the current population satisfies... Then, the current optimal function value replaces the original optimal function value, and the process proceeds to S5.2d; S5.2d: Based on step S5.2c, obtain the current best individual to replace the original best individual. ; S5.2e: Adjustments and optimizations are made based on the tabu policy, if the tabu list length is... Greater than the set maximum length If the taboo list is cleared, the optimal sequence is maintained until the entire process ends, generating the final path planning scheme.

Images