A multi-unmanned vehicle cooperative trajectory planning method for deep sea resource exploration

Abstract

The present application relates to the technical field of cooperative trajectory planning in the application scene of deep-sea resource space-time distribution detection, in particular to a multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource detection, comprising the following steps: an underwater carrier carries an AUV to keep constant depth and uniform straight sailing at a fixed bow angle, and reaches a predetermined position; an unmanned aerial vehicle maintains uniform flight at a fixed height and reaches the predetermined position at the same time as the underwater carrier, the unmanned aerial vehicle keeps a fixed heading before flying to the predetermined position, and performs circular flight around the underwater carrier after reaching the predetermined position; after the underwater carrier reaches the predetermined position, the AUV is released, and the underwater carrier sends a trajectory node of global trajectory planning to the AUV; the AUV performs local trajectory planning based on collision avoidance and deep-sea resource detection. The problem of real-time instruction transmission and state interaction between the test command station staff and the AUV during the deep-sea operation task is solved, which is beneficial to the intelligent implementation of the deep-sea operation task.

Description

Technical Field

[0001] This invention relates to the field of collaborative trajectory planning technology in application scenarios such as deep-sea resource spatiotemporal distribution detection, specifically a collaborative trajectory planning method for multiple unmanned vehicles for deep-sea resource detection. Background Technology

[0002] Currently, in the field of marine resource application and development, tasks such as dam inspection and resource exploration still rely heavily on manual intervention, and the overall level of intelligence needs to be improved. This reduces operational efficiency and increases operational costs to some extent. In complex operating environments where divers and other underwater tools cannot reach, intelligent underwater robots (AUVs) play an irreplaceable role. However, for offshore and deep-sea operations, due to the long distance and limited underwater signal transmission conditions, the problem of real-time information interaction between AUVs and test command station personnel urgently needs to be solved.

[0003] To address the communication difficulties between AUVs and test command stations caused by long communication distances in offshore missions, researchers proposed a method of collaborative trajectory planning between AUVs and unmanned aerial vehicles (UAVs). Utilizing UAVs as communication relays, this method effectively solved the problem of real-time status interaction and command transmission between AUVs and test command stations when operating at shallow depths. However, in deep-sea operations, due to the limited underwater communication distance between UAVs and AUVs, relying solely on UAVs as the communication medium is still insufficient to effectively solve the problem of command transmission and status interaction between AUVs and test command stations during deep-sea operations.

[0004] Therefore, this invention aims to propose a collaborative trajectory planning method for multiple unmanned vehicles (UUVs) for deep-sea resource exploration, realizing collaborative trajectory planning of AUVs, underwater vehicles (UCs) and unmanned aerial vehicles, in order to meet the needs of deep-sea engineering applications, thereby ensuring real-time command transmission and status interaction between AUVs and test command station personnel in operations such as deep-sea resource spatiotemporal distribution exploration. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a multi-unmanned vehicle (UUV) collaborative trajectory planning method for deep-sea resource exploration. By employing collaborative work between UUVs and underwater vehicles, and between underwater vehicles and AUVs, deep-sea operational communication is achieved. By using UUVs and underwater vehicles as communication relays, the method indirectly enables real-time command transmission and status interaction between test command station personnel and AUVs.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a collaborative trajectory planning method for multiple unmanned aerial vehicles (UAVs) for deep-sea resource exploration, comprising the following steps:

[0007] S1. The underwater vehicle carrying the AUV maintains a constant depth and travels straight at a constant speed with a fixed heading angle to reach the predetermined position;

[0008] S2. The unmanned aerial vehicle maintains a fixed altitude and flies at a constant speed, arriving at the predetermined position simultaneously with the underwater vehicle. The unmanned aerial vehicle maintains a fixed heading before arriving at the predetermined position, and after arriving at the predetermined position, it performs circular flight with the underwater vehicle as the center.

[0009] S3. After the underwater vehicle arrives at the predetermined position, it performs global trajectory planning and generates trajectory nodes, releases the AUV, and sends the trajectory nodes of the global trajectory planning to the AUV.

[0010] After receiving the trajectory nodes from the global trajectory planning, the S4 AUV performs collision avoidance-based local trajectory planning to explore deep-sea resources.

[0011] Furthermore, the collaborative trajectory planning process for unmanned aerial vehicles and underwater vehicles includes:

[0012] S1. Determine the trajectory start point and predetermined location information;

[0013] S2. The underwater vehicle departs from the test command station first, maintaining a constant depth, orientation, and speed while traveling in a straight line.

[0014] S3. The unmanned aerial vehicle then departs from the test command station and maintains a constant altitude, orientation, and speed of flight.

[0015] S4. A novel whale algorithm is used to optimize the parameters of the unmanned aerial vehicle;

[0016] S5. The underwater vehicle and the unmanned aerial vehicle arrive at the predetermined position simultaneously.

[0017] S6. After reaching the predetermined position, the underwater vehicle will begin a hovering and depth-fixed motion.

[0018] S7. At the same time, the unmanned aerial vehicle hovers in a circle with the underwater vehicle as the center.

[0019] S8. The unmanned aerial vehicle (UAV) and the underwater vehicle communicate in two directions. The UAV receives the instruction transmission information from the test command station, forwards it to the underwater vehicle, and sends it to the AUV through the underwater vehicle. The AUV transmits the status information to the underwater vehicle, forwards it to the UAV through the underwater vehicle, and then the UAV reports it to the test command station.

[0020] S9. Determine if the mission is over. If the mission is not over, continue the underwater vehicle's levitation and depth-fixed motion; otherwise, end the mission.

[0021] Furthermore, a novel whale algorithm is used to optimize the parameters of the unmanned aerial vehicle, including:

[0022] S1. Set the parameters for the whale algorithm;

[0023] S2. Calculate the fitness value of an individual whale;

[0024] S3. Update the adaptive inertia weights to obtain the current global optimal solution;

[0025] S4. Perform a Levy flight search on the current global optimal solution;

[0026] S5. Determine if the maximum number of iterations has been reached. If the maximum number of iterations has not been reached, recalculate the fitness value of the individual whale. If the maximum number of iterations has been reached, output the global optimal solution and end the process.

[0027] Furthermore, the parameters to be optimized include: the time difference between the departure of the underwater vehicle and the unmanned aerial vehicle. Flight altitude of unmanned aerial vehicles The radius of the circular motion around the underwater vehicle and the initial heading angle deviation between unmanned aerial vehicles and underwater vehicles .

[0028] Furthermore, the function expression of the whale algorithm is as follows:

[0029] (18)

[0030] in, This represents the number of iterations of the whale algorithm. Representing the The position of the optimal individual in the next iteration. Representing the The current position of the individual in the next iteration. and Indicates the adjustable coefficient. Indicates the distance between the humpback whale and its prey. Represents the natural constant. Represents a constant. This represents a random parameter, and it is between [-1, 1]. Represents the cosine function. This represents a random parameter, and it is between [0, 1].

[0031] Introducing the adaptive inertia weight factor into the whale algorithm yields the following function expression:

[0032] (twenty one)

[0033] in, This represents the adaptive inertia weighting factor;

[0034] The positional change based on Lévy's flight strategy is expressed as follows:

[0035] (twenty three)

[0036] in, Represents the step size transformation parameter. Represents point-to-point operations. This represents Levi's flight path.

[0037] Furthermore, the collaborative trajectory planning process between the underwater vehicle and the AUV includes:

[0038] S1. Construct an underwater surrounding environment model;

[0039] S2. Set the information for the starting point and target point of the trajectory;

[0040] S3. Select preliminary nodes for trajectory planning based on experience, and avoid identified obstacles;

[0041] S4. An improved artificial rabbit algorithm based on chaotic mapping initialization population, nonlinear inertial weight parameters and Gaussian walk strategy is used to optimize the global trajectory planning model and obtain global trajectory planning nodes.

[0042] S5. The underwater vehicle sends the trajectory nodes of the global trajectory planning to the AUV via underwater acoustic communication.

[0043] S6. The AUV uses detection equipment to detect obstructions in real time. If no obstruction is found, it switches to S8; otherwise, it switches to S7.

[0044] S7. Local trajectory planning is carried out using a novel artificial potential field method based on an improved repulsion function and a multi-objective weighted reward function method to avoid obstacles.

[0045] S8. Using the current heading angle and position provided by the navigation system, calculate the heading angle and depth of the target at the next trajectory node;

[0046] S9. The heading angle and depth information of the next trajectory node target are sent to the motion control system for control calculation and motion implementation;

[0047] The sensor data of S10 and AUV are updated;

[0048] S11, the AUV communicates bidirectionally with the underwater vehicle, indirectly realizing command transmission and status interaction between the test command station and the AUV;

[0049] S12. Determine whether the target point has been reached. If the target point has not been reached, proceed to S6; otherwise, end the process.

[0050] Furthermore, the global trajectory planning process for underwater vehicles includes the following steps:

[0051] S1. Initialize the initial node information for trajectory planning;

[0052] S2. Initialize the parameters of the new artificial rabbit algorithm;

[0053] S3, Calculate the energy factor;

[0054] S4. Introduce chaotic mapping to initialize the population;

[0055] S5. Calculate the nonlinear inertia weight parameters;

[0056] S6. Update the location of the artificial rabbit individual;

[0057] S7. Apply Gaussian random walk strategy;

[0058] S8. Determine if the maximum number of iterations has been reached. If the maximum number of iterations has not been reached, go to S2; otherwise, go to S9.

[0059] S9. Output the global trajectory planning nodes and end.

[0060] Furthermore, the functional expression of the artificial rabbit optimization algorithm is as follows:

[0061] (twenty four)

[0062] in, It is the first During the nth iteration, the 1st The alternative locations for a rabbit, and They represent the first During the nth iteration, the 1st Only rabbits and the first The rabbit's current location. Represents an exponential function. The maximum number of iterations, Represents the sine function. For random numbers that follow a standard normal distribution, Represents the floor function. and Represents a random number between [0, 1];

[0063] The function expression for initializing the population using chaotic mapping is as follows:

[0064] (28)

[0065] in, For the corresponding dimension, This is the serial number of the individual rabbit. Here are the chaos coefficients. A random number between (0, 1) This refers to the population size of artificially bred rabbits.

[0066] The functional expression for the nonlinear inertia weight parameter is as follows;

[0067] (29)

[0068] in, For nonlinear inertia weighting parameters, This represents the adjustment parameter that controls the upper and lower boundaries of the nonlinear inertia weight. Represents the tangent function. Used to control the smoothness of nonlinear inertia weight parameters;

[0069] By introducing the nonlinear inertia weight parameter into the artificial rabbit optimization algorithm, we can obtain equation (30).

[0070] (30)

[0071] The position change based on the Gaussian random walk strategy is shown in equation (31).

[0072] (31)

[0073] in, Represents the Gaussian factor. Indicates the first The optimal individual in the next iteration. For mean and variance, Indicates the first The th iteration in the Individuals, and This represents a random number that follows a uniform distribution in the range [0, 1].

[0074] Furthermore, the AUV local trajectory planning process includes:

[0075] S1, AUV uses detection equipment to detect obstructions in real time, and has detected obstructions;

[0076] S2, Enter local trajectory planning;

[0077] S3. The resultant repulsive force is calculated using a novel artificial potential field method with an improved repulsive function.

[0078] S4. Determine if the system has entered a local extremum. If it has not entered a local extremum, proceed to S6; otherwise, proceed to S5.

[0079] S5. Employ a multi-objective weighted reward function method to guide the escape from local extremum traps;

[0080] S6. Calculate the new target heading angle and depth based on the current actual heading angle and position information provided by the navigation system;

[0081] S7, End.

[0082] Furthermore, the functional expression for the target heading angle is as follows:

[0083] (33)

[0084] (34)

[0085] in, For the target heading angle, The current actual heading angle, ( , ) represents the coordinates of the first trajectory node of the AUV. , ( ) represents the current actual position coordinates of the AUV. This is the straight-line distance between the current actual position of the AUV and the lines connecting the previous trajectory node and the next target trajectory node. The angle between the line connecting the previous trajectory node and the next target trajectory node and the horizontal coordinate axis of the geodetic coordinate system;

[0086] The functional expression of the resultant repulsive force is as follows:

[0087] (35)

[0088] in, The resultant force representing the repulsive force, Indicates the number of detection devices. Indicates the first The repulsive gain coefficient of each detection device For the first The distance between the detection device and the obstacle. To maintain an absolutely safe distance, For a safe distance, It is the tangent function. It is the cotangent function. For the first The angle between the detection device and the geodetic coordinate system;

[0089] The expression for the multi-objective weighted reward function is as follows:

[0090] (36)

[0091] in, The sum of weighted rewards, Indicates goal-oriented reward. This indicates a collision avoidance safety reward. This indicates a very small local escape reward. This represents the reward control coefficient related to goal guidance; it is a positive coefficient. This indicates the distance between the previous position and the target trajectory node. This indicates the distance between the current position and the target trajectory node. This represents the reward control coefficient related to collision avoidance safety; it is a positive coefficient. Indicates a safe distance. For the first The distance between the detection device and the obstacle. This represents the reward control coefficient related to local minimum escape, and is a positive coefficient. The resultant force representing the repulsive force, This represents the threshold value for the minimum resultant repulsive force. It is a small positive value.

[0092] Compared with the prior art, the technical solution of this application has the following beneficial effects:

[0093] 1. This multi-unmanned vehicle collaborative trajectory planning method for deep-sea resource exploration uses AUVs, underwater vehicles, and unmanned aerial vehicles (UAVs) as communication mediums to solve the problem of real-time command transmission and status interaction between the test command station staff and AUVs during deep-sea operations. This method is conducive to the intelligent realization of deep-sea operations, reduces labor costs, and improves operational safety.

[0094] 2. This collaborative trajectory planning method for multiple unmanned aerial vehicles (UAVs) for deep-sea resource exploration introduces an artificial rabbit optimization algorithm, uses chaotic mapping to initialize the population and nonlinear inertial weight parameters, and replaces traditional random behavior with a Gaussian walk strategy to achieve fast global trajectory planning. By adopting a novel artificial potential field method with an improved repulsion function and introducing a multi-objective weighted reward function, it considers multiple factors such as target trajectory points, navigational obstacles, and local extreme value traps to achieve local trajectory planning for AUVs under precise collision avoidance conditions, thus solving the problem of collaborative trajectory planning between AUVs and underwater vehicles. The collaborative trajectory planning between underwater vehicles and unmanned aerial vehicles is achieved by introducing a novel whale optimization algorithm.

[0095] 3. This multi-unmanned vehicle (UUV) collaborative trajectory planning method for deep-sea resource exploration employs an underwater vehicle to send global trajectory planning node information commands to the AUV, eliminating the need for additional surface vessels and saving mission implementation costs. This invention is suitable for deep-sea engineering applications and can be applied to trajectory planning and control for tasks such as detecting the spatiotemporal distribution of resources in complex deep-sea environments. Attached Figure Description

[0096] Figure 1 This is a schematic diagram of a multi-unmanned vehicle collaborative trajectory planning task scenario for deep-sea resource exploration according to the present invention;

[0097] Figure 2 This is a flowchart of the collaborative trajectory planning process for the unmanned aerial vehicle and underwater vehicle according to the present invention.

[0098] Figure 3 This is a flowchart illustrating the parameter optimization process for the unmanned aerial vehicle of the present invention.

[0099] Figure 4 This is a flowchart of the collaborative trajectory planning process between the underwater vehicle and the AUV of the present invention.

[0100] Figure 5 This is a flowchart of the global trajectory planning for the underwater vehicle of the present invention;

[0101] Figure 6 This is a flowchart of the AUV local trajectory planning process of the present invention;

[0102] Figure 7 The figure shows the simulation results of the AUV trajectory tracking based on underwater vehicle commands according to the present invention. Detailed Implementation

[0103] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0104] Please see Figure 1-7 This embodiment presents a collaborative trajectory planning method for multiple unmanned vehicles (UAVs) for deep-sea resource exploration, comprising the following steps:

[0105] S1. The underwater vehicle carrying the AUV maintains a constant depth and travels straight at a constant speed with a fixed heading angle to reach the predetermined position;

[0106] S2. The unmanned aerial vehicle maintains a fixed altitude and constant speed and arrives at the predetermined position at the same time as the underwater vehicle. Before arriving at the predetermined position, the unmanned aerial vehicle maintains a fixed heading. After arriving at the predetermined position, it performs circular flight with the underwater vehicle as the center.

[0107] S3. After the underwater vehicle carrying the AUV arrives at the predetermined location, it performs global trajectory planning and generates trajectory nodes, releases the AUV, and sends the trajectory nodes of the global trajectory planning to the AUV.

[0108] After receiving the trajectory nodes from the global trajectory planning, the S4 AUV performs collision avoidance-based local trajectory planning to explore deep-sea resources.

[0109] like Figure 1 As shown, the underwater vehicle departs first from the test command station, followed by the unmanned aerial vehicle (UAV). B represents the departure phase. The underwater vehicle, carrying the AUV, travels straight at a fixed heading angle at a certain depth with a constant speed, while the UAV maintains a fixed altitude, fixed heading, and constant speed. The UAV and the underwater vehicle arrive at the predetermined position simultaneously. M represents the simultaneous arrival phase. Subsequently, the UAV hovers in a circle around the underwater vehicle as the center.

[0110] After the underwater vehicle carrying the AUV arrives at the predetermined location, it remains suspended at a fixed depth, releases the AUV, and then performs global trajectory planning, sending the trajectory nodes of the global trajectory planning to the AUV; the AUV performs collision avoidance-based local trajectory planning and carries out deep-sea resource exploration.

[0111] The letter "E" represents the phase where the AUV (Unmanned Aerial Vehicle) hovers in a circle around the underwater vehicle. It also represents the phase where the AUV receives the global trajectory planning node, executes collision-avoidance-based local trajectory planning, and conducts deep-sea resource exploration. To indirectly facilitate command transmission and status interaction between the test command station personnel and the AUV, the command station personnel transmit commands to the AUV, which then forwards them to the underwater vehicle, and finally to the AUV, thus issuing commands. Simultaneously, the AUV transmits status information to the underwater vehicle, which feeds back to the AUV, and finally, the AUV transmits the information to the test command station personnel. This allows for online operation optimization and facilitates intelligent operation.

[0112] In this embodiment, for a scenario where an underwater vehicle and an unmanned aerial vehicle (UAV) work collaboratively, a novel whale algorithm is introduced to complete collaborative trajectory planning. This enables the underwater vehicle to hover at a predetermined position and communicate with the UAV, and the UAV to communicate with personnel at the test command station. Because the collaborative work between the underwater vehicle and the AUV enables deep-sea communication, it indirectly facilitates real-time status interaction and command transmission between the test command station personnel and the AUV, allowing for online optimization of predetermined tasks and promoting the intelligent implementation of deep-sea operations.

[0113] like Figure 2 As shown, the collaborative trajectory planning process for unmanned aerial vehicles and underwater vehicles includes the following steps:

[0114] S1. Determine the trajectory start point and predetermined location information;

[0115] S2. The underwater vehicle departs from the test command station first, maintaining a constant depth, orientation, and speed while traveling in a straight line.

[0116] S3. The unmanned aerial vehicle then departs from the test command station and maintains a constant altitude, orientation, and speed of flight.

[0117] S4. A novel whale algorithm is used to optimize the parameters of the unmanned aerial vehicle;

[0118] S5. The underwater vehicle and the unmanned aerial vehicle arrive at the predetermined position simultaneously.

[0119] S6. After reaching the predetermined position, the underwater vehicle will begin a hovering and depth-fixed motion.

[0120] S7. At the same time, the unmanned aerial vehicle hovers in a circle with the underwater vehicle as the center.

[0121] S8. The unmanned aerial vehicle (UAV) and the underwater vehicle communicate bidirectionally. The UAV receives the command transmission information from the test command station, forwards it to the underwater vehicle, and sends it to the AUV through the underwater vehicle. The AUV transmits the status information to the underwater vehicle, which forwards it to the UAV. The UAV then reports the information to the test command station, thereby indirectly realizing the command transmission and status interaction between the test command station and the AUV.

[0122] S9. Determine if the mission is over. If the mission is not over, continue the underwater vehicle's levitation and depth-fixed motion; otherwise, end the mission.

[0123] Among them, such as Figure 3 As shown, a novel whale algorithm is used to optimize the parameters of the unmanned aerial vehicle, including the following steps:

[0124] S1. Set the parameters for the new whale algorithm;

[0125] S2. Calculate the fitness value of an individual whale;

[0126] S3. Update the adaptive inertia weights to obtain the current global optimal solution;

[0127] S4. Perform a Levy flight search on the current global optimal solution;

[0128] S5. Determine if the maximum number of iterations has been reached. If the maximum number of iterations has not been reached, recalculate the fitness value of the individual whale. If the maximum number of iterations has been reached, output the global optimal solution and end the process.

[0129] In this embodiment, the unmanned aerial vehicle (UAV) maintains a fixed altitude and uniform speed, arriving at the predetermined position simultaneously with the underwater vehicle. Before reaching the predetermined position, the UAV maintains a fixed heading. After arriving at the predetermined position, it performs circular hovering flight with the underwater vehicle as the center, specifically including:

[0130] 1) Establish a kinematic model:

[0131] After the underwater vehicle reaches the predetermined position, it will perform a hovering and fixed-depth motion. The motion of the underwater vehicle at the predetermined position can be considered to remain basically unchanged. Therefore, the collaborative trajectory planning of the underwater vehicle and the unmanned aerial vehicle focuses on the realization of the unmanned aerial vehicle itself.

[0132] Based on the above description and general assumptions, the kinematic model of the unmanned aerial vehicle is constructed as shown in equation (1):

[0133] (1)

[0134] in, Indicates the flight speed of the unmanned aerial vehicle. Indicates normal control force. Represents the quality of unmanned aerial vehicles. Represents the sine function. Represents the cosine function. Represents the tangent function. , , The meaning is the same in the following text. Represents gravitational acceleration. This indicates the distance between the Earth's center and the unmanned aerial vehicle. Indicates the path angle of the unmanned aerial vehicle. Indicates the roll angle of an unmanned aerial vehicle. Represents the heading angle of the unmanned aerial vehicle. Represents the latitude of the unmanned aerial vehicle. This represents the longitude of the unmanned aerial vehicle.

[0135] Because before reaching the predetermined location, the underwater vehicle travels at a certain depth with a fixed heading angle and a constant speed, while the unmanned aerial vehicle maintains a fixed altitude and a constant speed, therefore we have equation (2):

[0136] (2)

[0137] Meanwhile, because the unmanned aerial vehicle maintains a fixed heading before reaching the predetermined location, we have equation (3):

[0138] (3)

[0139] Analysis of equation (1) shows that the normal control force and tilt angle Equation (2) has a decisive influence on the motion of the unmanned aerial vehicle (UAV), and Equation (4) can ensure that the UAV maintains a fixed altitude.

[0140] (4)

[0141] After the unmanned aerial vehicle reaches the predetermined location, it will conduct circular hovering flight with the underwater vehicle as the center. Let the radius of this hovering flight be denoted as... According to the laws of motion, we have equation (5):

[0142] (5)

[0143] Combining equations (4) and (5), taking into account the flight speed of the unmanned aerial vehicle... Not too large, relative to the distance between the Earth's center and the unmanned aerial vehicle. Since it is a small quantity, when the unmanned aerial vehicle hovers around the underwater vehicle, we can obtain equation (6):

[0144] (6)

[0145] 2) Constructing an optimization model:

[0146] Because unmanned aerial vehicles (UAVs) and underwater vehicles are needed as communication relays to indirectly enable real-time command transmission and status interaction between the test command station personnel and the AUV, the UAVs and underwater vehicles must meet some basic conditions:

[0147] Considering that when the unmanned aerial vehicle and the underwater vehicle arrive at the predetermined position at the same time, these two unmanned vehicles of different media should be able to communicate normally, that is, the maximum communication distance limit must be met, and equation (7) can be obtained.

[0148] (7)

[0149] in, This indicates the distance between the underwater vehicle and the unmanned aerial vehicle. This represents the maximum straight-line distance that allows for normal communication between an underwater vehicle and an unmanned aerial vehicle.

[0150] As mentioned above, when the unmanned aerial vehicle is flying in a circular motion around the underwater vehicle, the relative distance between the two can be expressed as Equation (8).

[0151] (8)

[0152] in, This represents the relative distance between the unmanned aerial vehicle and the underwater vehicle. Represents the flight altitude of the unmanned aerial vehicle. Represents the buoyancy depth of an underwater vehicle. This represents the flight radius of an unmanned aerial vehicle when it hovers around an underwater vehicle. Should be in The range of values.

[0153] To ensure that the UAV and the underwater vehicle arrive at their designated positions simultaneously, and that the UAV can subsequently perform circular maneuvers around the underwater vehicle, their initial headings should differ. Let the distance between the UAV's starting point at the test command station and its designated position be denoted as... The deviation of the initial heading angle between the unmanned aerial vehicle and the underwater vehicle is denoted as . Then we can obtain equation (9);

[0154] (9)

[0155] Based on the physical characteristics of the actuators actually carried by the unmanned aerial vehicle, its control capability cannot be infinite and should be within a certain range. Therefore, equation (10) can be obtained.

[0156] (10)

[0157] in, This indicates the maximum possible roll angle of the unmanned aerial vehicle. Indicates the controllability of an unmanned aerial vehicle. This indicates the maximum controllability that an unmanned aerial vehicle can achieve.

[0158] Considering that communication needs to be kept smooth between the test command station and the unmanned aerial vehicle, in order to avoid the interference of the curvature of the earth on communication, the flight altitude of the unmanned aerial vehicle should satisfy equation (11).

[0159] (11)

[0160] in, Represents the flight altitude of the unmanned aerial vehicle. Represents the average radius of the Earth. The geocentric angle between the test command station and the predetermined location is given by equation (12).

[0161] (12)

[0162] in, Indicates the latitude of the predetermined location. Indicates the longitude of the predetermined location. The latitude of the test command post. The longitude represents the test command station.

[0163] Since the underwater vehicle and the unmanned aerial vehicle (UAV) departed from the test command station separately, and the underwater vehicle's speed in water differs from the UAV's speed in the air, the parameters to be optimized should include the time difference between the underwater vehicle's and the UAV's departure. It also includes the flight altitude of unmanned aerial vehicles. and the radius of the circular motion around the underwater vehicle Simultaneously considering the initial heading angle deviation between the unmanned aerial vehicle and the underwater vehicle. .

[0164] 3) Cooperative trajectory planning for unmanned aerial vehicles and underwater vehicles:

[0165] The departure time difference between the unmanned aerial vehicle and the underwater vehicle is obtained by using a novel whale algorithm. Initial heading angle deviation Drone flight altitude and the radius of circular motion This enables collaborative trajectory planning for unmanned aerial vehicles and underwater vehicles.

[0166] The process of parameter solving using the whale algorithm is as follows. Assume the number of individual whales is... The algorithm includes three behaviors: surrounding prey, bubble net hunting, and random learning.

[0167] a) Encirclement of prey;

[0168] The prey-encircling behavior of the whale algorithm refers to imitating the humpback whale's discovery and encirclement of prey, as shown in equation (13).

[0169] (13)

[0170] in, This represents the number of iterations of the whale algorithm. Representing the The position of the optimal individual in the next iteration. Representing the The current position of the individual in the next iteration. and The adjustable coefficient is shown in equations (14) to (16).

[0171] (14)

[0172] (15)

[0173] (16)

[0174] in, This represents a coefficient that changes with the number of iterations. and This represents a random parameter, and it is between [0, 1]. This indicates the maximum number of iterations set.

[0175] b) Predation using bubble nets;

[0176] Observations revealed that humpback whales use bubble nets to surround their prey and use spiral motion to approach and catch them, as shown in equation (17).

[0177] (17)

[0178] in, Indicates the distance between the humpback whale and its prey. Represents the natural constant. Represents a constant. This represents a random parameter, and it is between [-1, 1]. This represents the cosine function.

[0179] Humpback whales randomly switch between two prey-attacking behaviors, as shown in equation (18).

[0180] (18)

[0181] in, This represents a random parameter, and it is between [0, 1].

[0182] c) Random learning;

[0183] Humpback whales engage in random learning when they are not hunting prey in a swarm or using bubble nets, as shown in equation (19).

[0184] (19)

[0185] in, Representing the In the next iteration, a position is randomly selected from the current population individuals.

[0186] Based on the whale algorithm, a new type of whale algorithm is formed by introducing an adaptive inertia weight factor and the Levy flight strategy to enhance the search range and achieve global optimization.

[0187] (a) Adaptive inertia weighting factor;

[0188] The inertia weight factor has a good effect on improving the convergence speed and search accuracy of the solution. A larger inertia weight factor has a stronger global search capability in the early stage, while a smaller inertia weight factor has a stronger local search capability in the later stage. By adopting an adaptive inertia weight factor, in the early stage of iteration, a larger inertia weight can achieve a certain convergence accuracy faster. As the inertia weight slowly decreases, in the later stage of iteration, a smaller inertia weight can enable the optimization to reach the optimal solution, as shown in equation (20).

[0189] (20)

[0190] in, This represents the adaptive inertia weighting factor. This indicates the minimum weight set. This indicates the maximum weight set.

[0191] By introducing the adaptive inertia weight factor into the whale algorithm, we can obtain equation (21).

[0192] (twenty one)

[0193] (b) Levy's flight strategy;

[0194] The positional change based on Levi's flight strategy can be expressed as equation (22);

[0195] (twenty two)

[0196] in, Represents the step size transformation parameter. Represents point-to-point operations. The flight path of Levi is represented by the following distribution:

[0197] (twenty three)

[0198] In this embodiment, considering the collaborative operation scenario of AUV and underwater vehicle, and given the limited research on trajectory planning in deep-sea hydrothermal vent areas, a global trajectory planning approach and a collision avoidance strategy-based local trajectory planning are proposed for deep-sea resource spatiotemporal distribution detection. The global trajectory planning incorporates an artificial rabbit optimization algorithm, employs a novel chaotic mapping initialization population and a nonlinear inertial weight adjustment strategy, and replaces traditional random behavior with a Gaussian random walk strategy to achieve fast global trajectory planning. Compared to previous AUV-based collision avoidance strategies, a modified artificial potential field method is used to achieve refined collision avoidance under specific local trajectory planning conditions.

[0199] like Figure 4 As shown, the collaborative trajectory planning process between the underwater vehicle and the AUV includes the following steps:

[0200] S1. Construct an underwater surrounding environment model;

[0201] S2. Set the information for the starting point and target point of the trajectory;

[0202] S3. Select preliminary nodes for trajectory planning based on experience, and avoid identified obstacles;

[0203] S4. An improved artificial rabbit algorithm based on chaotic mapping initialization population, nonlinear inertial weight parameters and Gaussian walk strategy is used to optimize the global trajectory planning model and obtain global trajectory planning nodes.

[0204] S5. The underwater vehicle sends the trajectory nodes of the global trajectory planning to the AUV via underwater acoustic communication.

[0205] S6. The AUV uses detection equipment to detect obstructions in real time. If no obstruction is found, it switches to S8; otherwise, it switches to S7.

[0206] S7. Local trajectory planning is carried out using a novel artificial potential field method based on an improved repulsion function and a multi-objective weighted reward function method to avoid obstacles.

[0207] S8. Using the current heading angle and position provided by the navigation system, calculate the heading angle and depth of the target at the next trajectory node;

[0208] S9. The heading angle and depth information of the next trajectory node target are sent to the motion control system for control calculation and motion implementation;

[0209] The sensor data of S10 and AUV are updated;

[0210] S11, the AUV communicates bidirectionally with the underwater vehicle, indirectly realizing command transmission and status interaction between the test command station and the AUV;

[0211] S12. Determine whether the target point has been reached. If the target point has not been reached, proceed to S6; otherwise, end the process.

[0212] In this embodiment, as Figure 5 As shown, the global trajectory planning process for an underwater vehicle includes the following steps:

[0213] S1. Initialize the initial node information for trajectory planning;

[0214] S2. Initialize the parameters of the new artificial rabbit algorithm;

[0215] S3, Calculate the energy factor;

[0216] S4. Introduce chaotic mapping to initialize the population;

[0217] S5. Calculate the nonlinear inertia weight parameters;

[0218] S6. Update the location of the artificial rabbit individual;

[0219] S7. Apply Gaussian random walk strategy;

[0220] S8. Determine if the maximum number of iterations has been reached. If the maximum number of iterations has not been reached, go to S2; otherwise, go to S9.

[0221] S9. Output the global trajectory planning nodes and end.

[0222] In this embodiment, as Figure 6As shown, the AUV local trajectory planning process specifically includes the following steps:

[0223] S1, AUV uses detection equipment to detect obstructions in real time, and has detected obstructions;

[0224] S2, Enter local trajectory planning;

[0225] S3. The resultant repulsive force is calculated using a novel artificial potential field method with an improved repulsive function.

[0226] S4. Determine if the system has entered a local extremum. If it has not entered a local extremum, proceed to S6; otherwise, proceed to S5.

[0227] S5. Employ a multi-objective weighted reward function method to guide the escape from local extremum traps;

[0228] S6. Calculate the new target heading angle and depth based on the current actual heading angle and position information provided by the navigation system;

[0229] S7, End.

[0230] As a preferred embodiment, based on pre-obtained information such as seabed topography, the AUV's navigation area is set in the underwater vehicle's software program. Seabed topography and navigational obstruction information are analyzed, and preliminary trajectory planning nodes are constructed based on experience. An improved artificial rabbit algorithm is then used to optimize the trajectory node information, achieving rapid trajectory generation. The specific implementation method of global trajectory planning is as follows:

[0231] An improved artificial rabbit algorithm was used to optimize the initial trajectory planning model and obtain global trajectory planning nodes. The whole process includes detour foraging, random hiding, and energy shrinkage.

[0232] a) Taking a detour to find food;

[0233] Rabbits are good at running into the territory of other individuals to find food, so each searching individual tends to randomly select another searching individual in the population to update its own position and increase the relevant perturbation. Its state at the next moment is as shown in equation (24).

[0234] (twenty four)

[0235] in, It is the first During the nth iteration, the 1st The alternative locations for a rabbit, and They represent the first During the nth iteration, the 1st Only rabbits and the first The rabbit's current location. Represents an exponential function. The maximum number of iterations, Represents the sine function. For random numbers that follow a standard normal distribution, Represents the floor function. and It represents a random number between [0, 1].

[0236] b) Randomly hide;

[0237] To avoid predators, rabbits often dig burrows near their dens for shelter. With each iteration, the rabbit generates burrows around itself along the dimensions of the search space. The first cave, to reduce the probability of being preyed upon, the second The rabbit's first The locations of the hiding caves are shown in equations (25) and (26);

[0238] (25)

[0239] (26)

[0240] Indicates from A randomly selected hiding place in one of the caves ; Represents a random number between [0, 1]; This refers to the population size of artificially bred rabbits.

[0241] c) Energy contraction;

[0242] Rabbits frequently forage by taking detours in the early stages of iteration, while they frequently hide randomly in the later stages. This is mainly determined by the rabbit's energy, which gradually decreases over time. An energy factor is used to simulate the transition from foraging by detours to hiding randomly. The energy factor is shown in equation (27).

[0243] (27)

[0244] Represents the natural logarithm function. Represents a random number between [0, 1], when At that time, the rabbit chose to take a detour to find food; when At that time, the rabbit will hide randomly.

[0245] Based on the artificial rabbit optimization algorithm, the search speed is improved and the search range is enhanced by introducing chaotic mapping to initialize the population, nonlinear inertial weight parameters and Gaussian walk strategy, so as to achieve global optimization.

[0246] (a) Chaotic mapping initializes the population;

[0247] (28)

[0248] in, For the corresponding dimension, This is the serial number of the individual rabbit. Here are the chaos coefficients. A random number between (0, 1) This refers to the population size of artificially bred rabbits.

[0249] (b) Nonlinear inertia weighting parameters;

[0250] Using nonlinear inertial weight parameters, in the early stage of iteration, a larger inertial weight can achieve a certain convergence accuracy faster. As the inertial weight gradually decreases, in the later stage of iteration, a smaller inertial weight can enable the optimization to reach the optimal solution, as shown in equation (29).

[0251] (29)

[0252] in, For nonlinear inertia weighting parameters, This represents the adjustment parameter that controls the upper and lower boundaries of the nonlinear inertia weight. Represents the tangent function. Used to control the smoothness of nonlinear inertial weight parameters.

[0253] By introducing the nonlinear inertia weight parameter into the artificial rabbit optimization algorithm, we can obtain equation (30).

[0254] (30)

[0255] (c) Gaussian walk strategy;

[0256] The position change based on the Gaussian random walk strategy is shown in equation (31).

[0257] (31)

[0258] in, Represents the Gaussian factor. Indicates the first The optimal individual in the next iteration. For mean and variance, Indicates the first The th iteration in the Individuals, and This represents a random number that follows a uniform distribution in the range [0, 1].

[0259] Mean and variance As shown in equation (32);

[0260] (32)

[0261] in, It is a logarithmic function.

[0262] An improved artificial rabbit algorithm is used to optimize the global trajectory planning model of the AUV, and the trajectory nodes of the global trajectory planning are obtained. The underwater vehicle sends the trajectory nodes of the global trajectory planning to the AUV through underwater acoustic communication, so that the AUV can use them to perform local trajectory planning and motion control calculation.

[0263] In this embodiment, after receiving the trajectory nodes from the global trajectory planning, the AUV performs collision avoidance-based local trajectory planning for deep-sea resource exploration, including:

[0264] After the underwater vehicle sends the trajectory nodes for global trajectory planning to the AUV, the AUV enters the local trajectory planning stage. Based on the trajectory node information, the target heading angle and target depth are calculated. The target heading angle is calculated as shown in equations (33) and (34).

[0265] (33)

[0266] (34)

[0267] In equation (33), For the target heading angle, The current actual heading angle, ( , ) represents the coordinates of the first trajectory node of the AUV. , ) represents the current actual position coordinates of the AUV. After the AUV passes the first trajectory node, the heading angle of the target will be calculated in a form similar to a straight line segment using equation (34), so that the AUV can navigate in a near-straight line, which helps to reduce power consumption.

[0268] In equation (34), This is the straight-line distance between the current actual position of the AUV and the lines connecting the previous trajectory node and the next target trajectory node. The target heading angle is the angle between the line connecting the previous trajectory node and the next target trajectory node and the horizontal coordinate axis of the geodetic coordinate system. The AUV sends the calculated target heading angle to the control system for calculation and controls the AUV to navigate according to the planned heading angle. At the same time, during navigation, the AUV uses its onboard detection equipment to detect nearby temporary obstacles in a timely manner. When the AUV encounters a temporary obstacle, it initiates collision avoidance-based local trajectory planning to calculate a new target heading angle and target depth, thereby gradually approaching the final trajectory node.

[0269] The AUV uses a novel artificial potential field method with an improved repulsion function to simulate the repulsion between the AUV and the obstruction. The resultant force of the repulsion can be expressed as equation (35).

[0270] (35)

[0271] in, The resultant force representing the repulsive force, Indicates the number of detection devices. Indicates the first The repulsive gain coefficient of each detection device For the first The distance between the detection device and the obstacle. The absolute safe distance (once the distance between the detection equipment and the obstruction approaches the absolute safe distance, the repulsive force approaches infinity). For a safe distance, It is the tangent function. It is the cotangent function. For the first The angle between the detection device and the geodetic coordinate system. This demonstrates the relative positional relationships between the AUV and all obstructions. A larger value indicates that the AUV is relatively close to the obstruction, while a smaller value indicates that the AUV is relatively far away from the obstruction.

[0272] If the AUV gets caught in a local extremum trap, a multi-objective weighted reward function is introduced to guide the AUV to escape the local extremum trap. The design of the multi-objective weighted reward function is as shown in equation (36).

[0273] (36)

[0274] in, The sum of weighted rewards, Indicates goal-oriented reward. This indicates a collision avoidance safety reward. This indicates a very small local escape reward. This represents the reward control coefficient related to goal guidance; it is a positive coefficient. This indicates the distance between the previous position and the target trajectory node. This indicates the distance between the current position and the target trajectory node. This represents the reward control coefficient related to collision avoidance safety; it is a positive coefficient. Indicates a safe distance. For the first The distance between the detection device and the obstacle. This represents the reward control coefficient related to local minimum escape, and is a positive coefficient. The resultant force representing the repulsive force, This represents the threshold value for the minimum resultant repulsive force. It is a small positive value.

[0275] By introducing a multi-objective weighted reward function, the reward situation of the AUV can be changed, allowing the AUV to escape the local extreme value trap. This allows for the calculation of new target heading angle and target depth, guiding the AUV to approach the target trajectory node until it reaches the final trajectory node.

[0276] In summary, this invention proposes a collaborative trajectory planning method for multiple unmanned vehicles (UUVs) for deep-sea resource exploration. By using AUVs, underwater vehicles, and UUVs as communication mediums for collaborative trajectory planning, and leveraging UUVs and underwater vehicles as communication mediums, this method solves the problem of real-time command transmission and status interaction between the test command station personnel and the AUVs during deep-sea operations. This facilitates the intelligent implementation of deep-sea operations, reduces labor costs, and improves operational safety.

[0277] By introducing an artificial rabbit optimization algorithm, employing chaotic mapping to initialize the population and nonlinear inertial weight parameters, and using a Gaussian walk strategy to replace traditional random behavior, fast global trajectory planning is achieved. Furthermore, by adopting a novel artificial potential field method with an improved repulsion function and introducing a multi-objective weighted reward function, considering multiple factors such as target trajectory points, obstacles, and local extremum traps, local trajectory planning for AUVs under precise collision avoidance conditions is achieved, solving the problem of collaborative trajectory planning between AUVs and underwater vehicles. Finally, a novel whale optimization algorithm is introduced to achieve collaborative trajectory planning between underwater vehicles and unmanned aerial vehicles.

[0278] By using an underwater vehicle to send global trajectory planning node information commands to the AUV, no additional surface vessels are required, saving mission implementation costs.

[0279] This invention is suitable for deep-sea engineering applications and can be applied to trajectory planning and control for tasks such as the detection of the spatiotemporal distribution of resources in complex deep-sea environments.

[0280] The following simulation experiments are conducted using an example of an unmanned aerial vehicle (UAV), an underwater vehicle, and an AUV as embodiments to verify the method. A cooperative trajectory planning method for multiple unmanned vehicles (UAVs) for deep-sea resource exploration is presented. The basic parameters of the cooperative trajectory planning method for the UAV and the underwater vehicle are as follows: the flight speed of the UAV... The speed is 50 m / s, and the mass is... For a weight of 100kg, the maximum achievable tilt angle is... The maximum controllable angle is 30°. The underwater vehicle's speed is 5 m / s, and its underwater depth is 50 m. The coordinates of the test command station are 41°N, 121°E, and the coordinates of the predetermined location are 41.22°N, 121.22°E. Based on the above scenario, what is the distance between the test command station and the predetermined location? Approximately 30,000 m. Assume the maximum communication distance between the unmanned aerial vehicle and the underwater vehicle. The number of individuals in the whale algorithm is 400m. The maximum number of iterations is 10. The minimum weight of the inertia weight factor is 15. The maximum weight is 0.45. The value is 0.85. The basic parameters of the cooperative trajectory planning method for underwater vehicles and AUVs are as follows: the number of individuals in the artificial rabbit algorithm. The maximum number of iterations is 12. The adjustment parameter for the nonlinear inertia weighting parameter is 30. It is 0.4. The value is 0.6. The number of detection devices used for local trajectory planning. The reward control coefficient is 4, which is related to goal guidance. The reward control coefficient is 0.4, which is related to collision avoidance safety. The reward control coefficient is 0.25, which is related to local minimum escape. The threshold value for the minimum value of the net repulsive force is 0.35. It is 0.01.

[0281] The parameter optimization results for the cooperative trajectory planning of the unmanned aerial vehicle and the underwater vehicle are as follows, including the departure time difference between the underwater vehicle and the unmanned aerial vehicle. The initial heading angle deviation between the unmanned aerial vehicle and the underwater vehicle is 5402.6s. The altitude of the unmanned aerial vehicle is 0.706°. Approximately 93.4m, the radius of the circular motion of the unmanned aerial vehicle around the underwater vehicle. It is 369.6m.

[0282] The simulation results of AUV trajectory tracking based on underwater vehicle commands are shown in the figure below. Figure 7 As shown, the results demonstrate that global trajectory planning was achieved, while temporary obstacles were avoided through local trajectory planning, resulting in a good trajectory planning effect.

[0283] It should be noted that, in this document, relational terms such as "first" and "second" are used only 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 thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0284] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A collaborative trajectory planning method for multiple unmanned aerial vehicles (UAVs) for deep-sea resource exploration, characterized in that, Includes the following steps: S1. The underwater vehicle carrying the AUV maintains a constant depth and travels straight at a constant speed with a fixed heading angle to reach the predetermined position; S2. The unmanned aerial vehicle maintains a fixed altitude and flies at a constant speed, arriving at the predetermined position simultaneously with the underwater vehicle. The unmanned aerial vehicle maintains a fixed heading before arriving at the predetermined position, and after arriving at the predetermined position, it performs circular flight with the underwater vehicle as the center. S3. After the underwater vehicle arrives at the predetermined position, it performs global trajectory planning and generates trajectory nodes, releases the AUV, and sends the trajectory nodes of the global trajectory planning to the AUV. After receiving the trajectory nodes from the global trajectory planning, the S4 AUV performs collision avoidance-based local trajectory planning to explore deep-sea resources.

2. The multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource exploration according to claim 1, characterized in that: The collaborative trajectory planning process for unmanned aerial vehicles and underwater vehicles includes: S1. Determine the trajectory start point and predetermined location information; S2. The underwater vehicle departs from the test command station first, maintaining a constant depth, orientation, and speed while traveling in a straight line. S3. The unmanned aerial vehicle then departs from the test command station and maintains a constant altitude, orientation, and speed of flight. S4. A novel whale algorithm is used to optimize the parameters of the unmanned aerial vehicle; S5. The underwater vehicle and the unmanned aerial vehicle arrive at the predetermined position simultaneously. S6. After reaching the predetermined position, the underwater vehicle will begin a hovering and depth-fixed motion. S7. At the same time, the unmanned aerial vehicle hovers in a circle with the underwater vehicle as the center. S8. The unmanned aerial vehicle (UAV) and the underwater vehicle communicate in two directions. The UAV receives the instruction transmission information from the test command station, forwards it to the underwater vehicle, and sends it to the AUV through the underwater vehicle. The AUV transmits the status information to the underwater vehicle, forwards it to the UAV through the underwater vehicle, and then the UAV reports it to the test command station. S9. Determine if the mission is over. If the mission is not over, continue the underwater vehicle's levitation and depth-fixed motion; otherwise, end the mission.

3. The multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource exploration according to claim 2, characterized in that: A novel whale algorithm is used to optimize the parameters of the unmanned aerial vehicle, including: S1. Set the parameters for the whale algorithm; S2. Calculate the fitness value of an individual whale; S3. Update the adaptive inertia weights to obtain the current global optimal solution; S4. Perform a Levy flight search on the current global optimal solution; S5. Determine if the maximum number of iterations has been reached. If the maximum number of iterations has not been reached, recalculate the fitness value of the individual whale. If the maximum number of iterations has been reached, output the global optimal solution and end the process.

4. The multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource exploration according to claim 3, characterized in that: The parameters to be optimized include: the time difference between the departure of the underwater vehicle and the unmanned aerial vehicle. Flight altitude of unmanned aerial vehicles The radius of the circular motion around the underwater vehicle and the initial heading angle deviation between unmanned aerial vehicles and underwater vehicles .

5. The multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource exploration according to claim 4, characterized in that: The function expression of the whale algorithm is as follows: （18） in, This represents the number of iterations of the whale algorithm. Representing the The position of the optimal individual in the next iteration. Representing the The current position of the individual in the next iteration. and Indicates the adjustable coefficient. Indicates the distance between the humpback whale and its prey. Represents the natural constant. Represents a constant. This represents a random parameter, and it is between [-1, 1]. Represents the cosine function. This represents a random parameter, and it is between [0, 1]. Introducing the adaptive inertia weight factor into the whale algorithm yields the following function expression: （21） in, This represents the adaptive inertia weighting factor; The positional change based on Lévy's flight strategy is expressed as follows: （23） in, Represents the step size transformation parameter. Represents point-to-point operations. This represents Levi's flight path.

6. The multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource exploration according to claim 1, characterized in that: The collaborative trajectory planning process between underwater vehicles and AUVs includes: S1. Construct an underwater surrounding environment model; S2. Set the information for the starting point and target point of the trajectory; S3. Select preliminary nodes for trajectory planning based on experience, and avoid identified obstacles; S4. An improved artificial rabbit algorithm based on chaotic mapping initialization population, nonlinear inertial weight parameters and Gaussian walk strategy is used to optimize the global trajectory planning model and obtain global trajectory planning nodes. S5. The underwater vehicle sends the trajectory nodes of the global trajectory planning to the AUV via underwater acoustic communication. S6. The AUV uses detection equipment to detect obstructions in real time. If no obstruction is found, it switches to S8; otherwise, it switches to S7. S7. Local trajectory planning is carried out using a novel artificial potential field method based on an improved repulsion function and a multi-objective weighted reward function method to avoid obstacles. S8. Using the current heading angle and position provided by the navigation system, calculate the heading angle and depth of the target at the next trajectory node; S9. The heading angle and depth information of the next trajectory node target are sent to the motion control system for control calculation and motion implementation; The sensor data of S10 and AUV are updated; S11, the AUV communicates bidirectionally with the underwater vehicle, indirectly realizing command transmission and status interaction between the test command station and the AUV; S12. Determine whether the target point has been reached. If the target point has not been reached, proceed to S6; otherwise, end the process.

7. A multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource exploration according to claim 6, characterized in that: The global trajectory planning process for underwater vehicles includes the following steps: S1. Initialize the initial node information for trajectory planning; S2. Initialize the parameters of the new artificial rabbit algorithm; S3, Calculate the energy factor; S4. Introduce chaotic mapping to initialize the population; S5. Calculate the nonlinear inertia weight parameters; S6. Update the location of the artificial rabbit individual; S7. Apply Gaussian random walk strategy; S8. Determine if the maximum number of iterations has been reached. If the maximum number of iterations has not been reached, go to S2; otherwise, go to S9. S9. Output the global trajectory planning nodes and end.

8. The multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource exploration according to claim 7, characterized in that: The function expression for the artificial rabbit optimization algorithm is as follows: （24） in, It is the first During the nth iteration, the 1st The alternative locations for a rabbit, and They represent the first During the nth iteration, the 1st Only rabbits and the first The rabbit's current location. Represents an exponential function. The maximum number of iterations, Represents the sine function. For random numbers that follow a standard normal distribution, Represents the floor function. and Represents a random number between [0, 1]; The function expression for initializing the population using chaotic mapping is as follows: （28） in, For the corresponding dimension, This is the serial number of the individual rabbit. Here are the chaos coefficients. A random number between (0, 1) This refers to the population size of artificially bred rabbits; The functional expression for the nonlinear inertia weight parameter is as follows; （29） in, For nonlinear inertia weighting parameters, This represents the adjustment parameter that controls the upper and lower boundaries of the nonlinear inertia weight. Represents the tangent function. Used to control the smoothness of nonlinear inertia weight parameters; By introducing the nonlinear inertia weight parameter into the artificial rabbit optimization algorithm, we can obtain equation (30). （30） The position change based on the Gaussian random walk strategy is shown in equation (31). （31） in, Represents the Gaussian factor. Indicates the first The optimal individual in the next iteration. For mean and variance, Indicates the first The th iteration in the Individual, and This represents a random number that follows a uniform distribution in the range [0, 1].

9. A multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource exploration according to claim 1, characterized in that: The AUV local trajectory planning process includes: S1, AUV uses detection equipment to detect obstructions in real time, and has detected obstructions; S2, Enter local trajectory planning; S3. The resultant repulsive force is calculated using a novel artificial potential field method with an improved repulsive function. S4. Determine if the system has entered a local extremum. If it has not entered a local extremum, proceed to S6; otherwise, proceed to S5. S5. Employ a multi-objective weighted reward function method to guide the escape from local extremum traps; S6. Calculate the new target heading angle and depth based on the current actual heading angle and position information provided by the navigation system; S7, End.

10. A multi-unmanned vehicle cooperative trajectory planning method for deep-sea resource exploration according to claim 9, characterized in that: The functional expression for the target heading angle is as follows: （33） （34） in, For the target heading angle, The current actual heading angle, ( , ) represents the coordinates of the first trajectory node of the AUV. , ( ) represents the current actual position coordinates of the AUV. This is the straight-line distance between the current actual position of the AUV and the lines connecting the previous trajectory node and the next target trajectory node. The angle between the line connecting the previous trajectory node and the next target trajectory node and the horizontal coordinate axis of the geodetic coordinate system; The functional expression of the resultant repulsive force is as follows: （35） in, The resultant force representing the repulsive force, Indicates the number of detection devices. Indicates the first The repulsive gain coefficient of each detection device For the first The distance between the detection device and the obstacle. To maintain an absolutely safe distance, For a safe distance, It is the tangent function. It is the cotangent function. For the first The angle between the detection device and the geodetic coordinate system; The expression for the multi-objective weighted reward function is as follows: （36） in, The sum of weighted rewards, Indicates goal-oriented reward. This indicates a collision avoidance safety reward. This indicates a very small local escape reward. This represents the reward control coefficient related to goal guidance; it is a positive coefficient. This indicates the distance between the previous position and the target trajectory node. This indicates the distance between the current position and the target trajectory node. This represents the reward control coefficient related to collision avoidance safety; it is a positive coefficient. Indicates a safe distance. For the first The distance between the detection device and the obstacle. This represents the reward control coefficient related to local minimum escape, and is a positive coefficient. The resultant force representing the repulsive force, This represents the threshold value for the minimum resultant repulsive force. It is a small positive value.

Images