A method, system, device and medium for fast deployment control of a floating platform in a random wind field

Abstract

A kind of airship platform in random wind field in the method, system, equipment and medium of fast deployment control of level flight;Method: based on Markov decision process, establish the mathematical model of airship platform under random wind field to the shortest time task of target point, combined with value iteration method, the optimal action sequence of airship platform is obtained, while establishing the horizontal plane dynamics model based on the six-degree-of-freedom dynamics model of airship platform, and join random disturbance to obtain the horizontal plane random dynamics model of airship platform, combined with the horizontal plane random dynamics model, based on the idea of backstepping method, design level flight random velocity control law and adaptive law, and combine optimal action sequence to control airship platform to level flight fast deployment;System, equipment and medium are used to realize a kind of airship platform in random wind field in the method of level flight fast deployment control;The airship platform deployed by the application reaches target area global time short, horizontal deployment precision is high, can effectively suppress the interference of random wind field to airship platform position deployment.

Description

Technical Field

[0001] This invention belongs to the field of mechanical control technology, specifically relating to a method, system, equipment, and medium for rapid deployment control of a floating platform in a random wind field, which can be used in the design of a floating platform control system. Background Technology

[0002] Aerials powered by helium, capable of sustained aerodynamic operation, possess advantages such as high payload capacity, long aerodynamic duration, wide coverage, and low energy consumption, gaining widespread attention and being applied to various high- and low-altitude missions. However, the air is inevitably affected by wind fields. Therefore, considering the stochastic characteristics of wind fields and addressing future practical application scenarios, achieving rapid deployment from the current starting point to the target area will be a key issue restricting the large-scale application of aerial platforms in the future.

[0003] Significant progress has been made both domestically and internationally in addressing this issue:

[0004] Zhang Heng et al. published a study on the control method of multi-vector thrust stratospheric airships (Zhang Heng, Yang Hui, Wang Yu. Research on control method of multi-vector thrust stratospheric airships based on backstepping [J]. Agricultural Equipment and Vehicle Engineering, 2023, 61(02): 117-121). Based on the backstepping theory, they designed an attitude controller for stratospheric airships with vector thrust, which showed significant performance in terms of dynamics, and was able to achieve good global stabilization of the system and progressive tracking of a given position and yaw angle. However, it did not consider the interference of stratospheric random wind fields on the motion process of the floating platform, resulting in low motion accuracy in actual trajectory planning.

[0005] Shen SP et al. published "Wind Speed ​​Estimation and Station-Keeping Control for Stratospheric Airships with Extended Kalman Filter" (Shen SP, Liu L, Huang BM, Lin XW, et al. Wind Speed ​​Estimation and Station-Keeping Control for Stratospheric Airships with Extended Kalman Filter. Proceedings of the 2015 Chinese Intelligent Automation Conference. 2015: 145-157). This paper describes a method using the airship's position information to design an extended Kalman filter to estimate the airship's speed and wind speed, reducing wind disturbances through speed compensation. However, based on existing sensor technology, this extended Kalman filter cannot accurately estimate time-varying wind disturbances, resulting in insufficient accuracy for trajectory planning of aerostated platforms.

[0006] Yang, XW, et al. published "Horizontal trajectory control of stratospheric airships in wind field using Q-learning algorithm" (Yang, XW, Yang, XX, Deng, XL, Horizontal trajectory control of stratospheric airships in wind field using Q-learning algorithm[J]. AEROSPACE SCIENCE AND TECHNOLOGY, 2020, 106: 67-82.), proposing an adaptive horizontal trajectory control method for stratospheric airships in uncertain wind fields based on the Q-learning algorithm. A Markov decision process (MDP) model of the airship is established, where the airship's motion strategy is determined by the wind direction, and a cerebellar model joint controller (CMAC) neural network is designed to optimize the motion strategy for each state. This control method exhibits good stability and intelligent decision-making ability in the horizontal trajectory control of stratospheric airships. However, it is not integrated with the dynamic model of the aerostat platform; trajectory planning for the aerostat platform is only performed in the kinematics, resulting in poor control performance in practice.

[0007] The aforementioned studies either failed to consider the interference of stratospheric random wind fields on the motion of the aerostat when designing controllers, or treated the stratospheric wind field as a constant without considering its time-varying random characteristics, or focused only on the kinematic level in trajectory planning without integrating it with the aerostat's dynamic model, resulting in poor actual control performance. In summary, there is currently no method that considers the impact of time-varying random wind fields on the aerostat to achieve effective, rapid, and interference-resistant deployment on the horizontal surface. Summary of the Invention

[0008] To overcome the shortcomings of the prior art, the present invention aims to provide a method, system, device, and medium for rapid deployment control of an airborne platform in random wind fields. Based on a Markov decision process, a mathematical model for rapid deployment of the airborne platform in random wind fields is established. Based on this model, the optimal action sequence of the airborne platform is obtained using a value iteration method. On the basis of the existing horizontal dynamics model of the airborne platform, a random disturbance term is added to obtain a horizontal random dynamics model of the airborne platform. Based on this model, a random speed control law for level flight is designed using the backstepping method. This control law drives the platform, enabling it to track different optimal action sequences and achieve rapid deployment control of the airborne platform in random wind fields. This improves the rapid deployment effect of the airborne platform in random wind fields and effectively suppresses the influence of random wind fields on the deployment accuracy of the airborne platform.

[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0010] A method for rapid deployment and control of a floating platform in a stochastic wind field during level flight includes the following steps:

[0011] Step 1: Based on the Markov decision process, establish a mathematical model for the task of the floating platform traveling to the target point in the shortest time within a specific area under a random wind field;

[0012] Step 2: Based on the mathematical model established in Step 1, the optimal action sequence of the floating platform is obtained by combining the value iteration solution method;

[0013] Step 3: Establish a six-degree-of-freedom dynamic model of the floating platform, and based on the six-degree-of-freedom dynamic model of the floating platform, establish a horizontal dynamic model of the floating platform;

[0014] Step 4: Based on the horizontal dynamic model of the floating platform established in Step 3, add a random perturbation term to the model to obtain the stochastic dynamic model of the horizontal surface of the floating platform;

[0015] Step 5: Combining the stochastic dynamics model of the airborne platform obtained in Step 4, design the level flight stochastic speed control law and adaptive law based on the backstepping method;

[0016] Step 6: Using the control law and adaptive law designed in Step 5, drive the aerobatic platform to navigate in the random wind field according to the action sequence solved in Step 2, so as to realize the rapid deployment control of the aerobatic platform in level flight.

[0017] The mathematical model described in step 1 is expressed as follows:

[0018] M = (S, A, P, R, γ),

[0019] Where S represents the set of states of the environment, and state s t ∈S is a state in the environmental state set at time t; A represents the action set of the floating platform, and action a t ∈A is an action in the action set at time t; P represents the set of state transition probabilities, and the state transition probability P(s) is... t+1 |s t a) indicates that the floating platform is in state s t The state s is reached by performing action a. t+1 The probability; R represents the instant reward received by the floating platform, R(s) t+1 |s t a) indicates that the floating platform is in state s t The next action a reaches state s. t+1 The immediate reward received; γ∈[0,1] represents the discount factor, used to calculate the cumulative reward of the floating platform, and its value indicates the value of the platform in the current state s. t With future states s t+1 The degree to which immediate rewards are considered; the larger γ is, the greater the impact of subsequent rewards on the current reward.

[0020] Step 2 specifically includes:

[0021] First, an environmental state set S is established using wind field data, followed by the introduction of a stochastic wind field model f. GS The system calculates the state transition probability set P by combining the input action set A and the target point position, and calculates the immediate reward R. Then, it introduces a value iteration algorithm, sets a discount factor γ and a value iteration convergence threshold ε, initializes the state value function in the value iteration algorithm, and enters a loop iteration. The difference between the two state value functions is then calculated. During the loop, if the difference of the state value function reaches the value iteration convergence threshold ε, the loop is exited; otherwise, the larger value of the state value function is taken and the iteration continues until the loop ends. The resulting state value function is the optimal state value function, and the optimal action sequence is extracted through the optimal state value function.

[0022] The six-degree-of-freedom dynamic model of the floating platform established in step 3 is represented as follows:

[0023]

[0024] Where, ξ=[uvwpqr] T Let u, v, and w represent the velocity vectors in the body coordinate system along the X, Y, and Z axes, respectively. Let p, q, and r represent the angular velocity vectors in the body coordinate system along the X, Y, and Z axes, respectively. Let U represent the control input of the six-degree-of-freedom dynamic model of the floating platform, and M represent the mass matrix of the floating platform, as shown below:

[0025]

[0026] Where m is the mass of the floating platform, m ij (i,j=1,2,...,6) represents the additional mass, I x ,I y ,I Z I represents the moments of inertia about the X, Y, and Z axes, respectively. xz Represents the product of inertia about the XOY plane, (x G ,y G ,z G () represents the coordinates of the center of gravity of the floating platform in the body coordinate system;

[0027] F f The dynamic force term is represented as follows:

[0028]

[0029] F A The aerodynamic term is represented as follows:

[0030]

[0031] in, This is the coordinate transformation matrix from the velocity coordinate system to the body coordinate system. For dynamic pressure, ρ 空气 air density, V represents the magnitude of the flight speed. a C represents the volume of the floating platform. X C is the drag coefficient. Y C is the lateral force coefficient. Z C is the lift coefficient. L C is the rolling moment coefficient. M C is the pitching moment coefficient. N These are the yaw moment coefficients, collectively referred to as aerodynamic coefficients;

[0032] F G The term representing its own gravity is as follows:

[0033]

[0034] Where G = mg is the magnitude of the gravity acting on the floating platform. Let r be the transformation matrix from the ground inertial coordinate system to the body coordinate system. G Let φ be the radius vector from the center of buoyancy to the center of gravity, and let φ, θ, and ψ be the roll angle, pitch angle, and yaw angle of the floating platform, respectively.

[0035] F B The buoyancy term is represented as follows:

[0036]

[0037] Where, B1=ρ 空气 gV a The magnitude of the buoyancy force on the floating platform is zero because the origin of the machine system's coordinate system coincides with the center of buoyancy.

[0038] The establishment of the horizontal dynamic model of the floating platform is achieved as follows:

[0039]

[0040] Where λ = [uvr] T u and v represent the components of the velocity vector on the X and Y axes in the coordinate system of the floating platform, respectively, and r represents the component of the angular velocity vector on the Z axis; M A =diag(m+m) 11 ,m+m 22 ,I Z +m 66 ), where diag(·) denotes a diagonal matrix; F a =[X a cosβ+Y a sinβ-X a sinβ+Y a cosβN a ] T The aerodynamic term representing the horizontal plane, X a =-(1 / 2)ρ h (u 2 +v 2 V a 2 / 3 C X , ρ h β represents the air density at height h, and β represents the sideslip angle of the floating platform. Both are aerodynamic derivatives, U λ This represents the control input to the dynamic model of the horizontal plane of the floating platform.

[0041] The stochastic dynamic model of the floating platform on the horizontal plane described in step 4 is expressed as follows:

[0042]

[0043] in, λ=[uvr] T u and v represent the components of the velocity vector on the X and Y axes in the coordinate system of the floating platform, respectively, and r represents the component of the angular velocity vector on the Z axis; M A =diag(m+m) 11 ,m+m 22 ,I Z +m 66 ), m 11 m 22 m 66 For added mass, I Z Let represent the moment of inertia about the Z-axis, respectively. F a =[X a cosβ+Y a sinβ-X a sinβ+Y a cosβN a ] T The aerodynamic term representing the horizontal plane, X a =-(1 / 2)ρ h (u 2 +v 2 V a 2 / 3 C X , ρ h β represents the air density at height h, and β represents the sideslip angle of the floating platform. All are aerodynamic derivatives. n λ ω represents the random perturbation intensity coefficient of the model. λ It is a 3-dimensional independent standard Wiener process vector.

[0044] The control law described in step 5 is expressed as follows:

[0045]

[0046] The adaptive law is expressed as:

[0047]

[0048] Among them, U λ To control the input, M A =diag(m+m) 11 ,m+m 22 ,I Z +m 66 ), where m is the mass of the floating platform, m11 m 22 m 66 For added mass, I Z Let represent the moment of inertia about the Z-axis, respectively. λ=[uvr] T u and v represent the components of the velocity vector on the X and Y axes in the coordinate system of the floating platform, respectively, and r represents the component of the angular velocity vector on the Z axis. F a =[X a cosβ+Y a sinβ-X a sinβ+Y a cosβN a ] T The aerodynamic term representing the horizontal plane, X a =-(1 / 2)ρ h (u 2 +v 2 V a 2 / 3 C X , ρ h β represents the air density at height h, and β represents the sideslip angle of the floating platform. All are aerodynamic derivatives. K λ k λ γ λ δ λ All are positive constants of the design. e represents the estimated value of the model's random disturbance intensity coefficient. λ λ represents the backstepping tracking error. d Indicates the desired speed.

[0049] Step 6 specifically includes:

[0050] The estimated values ​​of the random disturbance intensity coefficients of the model are obtained by solving the adaptive law. The velocity λ of the floating platform at different moments and the estimated value obtained from the solution are used. Substituting the values ​​into the control law, we obtain the control input U required by the aerodynamic system of the floating platform. λ Then, based on the different power system layouts, different horizontal thrust F is established. T The expression, and by U λ =F T The relationship is used to calculate the magnitude and direction of the horizontal thrust, so that the error between the actual speed and the desired speed approaches zero. By navigating in each wind field according to the optimal action sequence solved in step 2, the rapid deployment control of the airborne platform in level flight can be achieved.

[0051] A rapid deployment control system for a floating platform in a stochastic wind field includes:

[0052] The model building module, based on the Markov decision process, establishes a mathematical model of the shortest time task for an airborne platform to reach a target point in a specific region under a random wind field. It also uses the value iteration method to find the optimal action sequence of the airborne platform. At the same time, it establishes a six-degree-of-freedom dynamic model of the airborne platform, then establishes a horizontal dynamic model of the airborne platform, and finally adds a random perturbation term to the model to obtain a random dynamic model of the horizontal plane of the airborne platform.

[0053] The control law and adaptive law design module combines the stochastic dynamics model of the airborne platform on the horizontal plane and designs the level flight stochastic speed control law and adaptive law based on the backstepping method.

[0054] The deployment module utilizes control and adaptive laws to drive the aerobatic platform to navigate in a random wind field according to an action sequence, thereby achieving rapid deployment control of the aerobatic platform in level flight.

[0055] A rapid deployment control device for a floating platform in a random wind field during level flight includes:

[0056] Memory: Used to store the computer program that implements the method for rapid deployment control of a floating platform in a random wind field;

[0057] Processor: Used to implement the aforementioned method for rapid deployment control of a floating platform in a random wind field during the level flight when executing the computer program.

[0058] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of a method for rapid deployment control of a floating platform in a random wind field.

[0059] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0060] 1. In the horizontal deployment of the floating platform, this invention introduces a Markov decision process to consider the influence of random wind fields and uses a value iteration method to solve for the optimal action sequence, which shortens the global time to reach the target area and improves the speed of horizontal deployment of the floating platform.

[0061] 2. This invention utilizes a stochastic dynamic model of the horizontal plane of the floating platform and designs a control law based on the backstepping method to satisfy the stability of the system against random disturbances. The engine thrust obtained from the control law drives the floating platform, which can be rapidly deployed to the target location. This effectively suppresses the influence of random wind fields on the deployment of the floating platform and improves the horizontal deployment accuracy of the floating platform.

[0062] 3. This invention utilizes a stochastic dynamic model of the horizontal plane of the floating platform and designs an adaptive law based on the backstepping method to satisfy the stability of the system against random disturbances. This enables the controller to continuously adjust according to the adaptive law, better cope with changes in the dynamic characteristics of the system, and improve the robustness of the horizontal deployment scheme of the floating platform.

[0063] In summary, this invention improves the speed, accuracy, and robustness of horizontal deployment of floating platforms. Attached Figure Description

[0064] Figure 1 This is a flowchart of the method of the present invention.

[0065] Figure 2 This is a wind field distribution map of the study area of ​​this invention.

[0066] Figure 3 This is a diagram illustrating the wind direction angle division and direction definition of the present invention.

[0067] Figure 4 This is a basic diagram of the floating platform of the present invention.

[0068] Figure 5 This is a flowchart illustrating the value iteration algorithm of the present invention for solving the problem.

[0069] Figure 6 This is a diagram showing the optimal action sequence of the floating platform in each region when the driving speed is 10 m / s in a random wind field according to the present invention.

[0070] Figure 7 This is a wind field distribution map of the study area when the simulation time t = 45s is the present invention.

[0071] Figure 8 This is a diagram showing the optimal action sequence of the floating platform in each region when the simulation time t = 45s is the same as that of the present invention.

[0072] Figure 9 This is a graph showing the response curve of the velocity u in the X-axis direction during a simulation experiment of the horizontal deployment of a floating platform to resist random wind interference, according to the present invention.

[0073] Figure 10 This is a response curve of the Y-axis velocity v in a simulation experiment of the horizontal deployment of a floating platform to resist random wind interference, according to the present invention.

[0074] Figure 11 The diagram shows the response curve of the yaw rate r in the simulation experiment of the horizontal deployment of the floating platform to resist random wind interference according to the present invention. Detailed Implementation

[0075] This embodiment describes a method for controlling the resistance of aerostated platforms to random wind interference. The aerostated platform refers to an aircraft operating in near-space capable of performing specific tasks, including low-dynamic aircraft such as high-altitude balloons, stratospheric airships, and high-altitude long-endurance unmanned aerial vehicles (UAVs). Long endurance and regional loiter time are the most significant characteristics and advantages of stratospheric aerostats, and effective trajectory tracking control technology is one of the key technologies for leveraging the advantages of stratospheric airships. However, since the flight speed of the aerostated platform is on the same order of magnitude as the wind speed, wind interference has a significant impact during movement, greatly affecting the deployment accuracy of the aerostated platform. Against this background, this embodiment proposes a method for rapid deployment control of aerostated platforms in random wind fields during level flight, aiming to improve the rapid deployment effect of aerostated platforms in random wind fields and effectively suppress the impact of random wind fields on the deployment accuracy of aerostated platforms.

[0076] The present invention will be further described below with reference to the accompanying drawings. It should be noted that this embodiment is based on the present technical solution and provides detailed implementation methods and specific operation processes, but the protection scope of the present invention is not limited to this embodiment.

[0077] Reference Figure 1 The implementation steps for this example are as follows:

[0078] Step 1, the mathematical model M for the task of minimizing the travel time of an aerial platform from a specific area to a target point under random wind conditions, is implemented as follows:

[0079] Step 1.1: Establish a mathematical model for the environmental state set S:

[0080] like Figure 2 As shown, the research environment in this embodiment is a specific region within China where the airborne platform is performing a mission. The approximate latitude and longitude range is [102°, 113°] East longitude and [28°, 39°] North latitude. The research region is discretized into 12×12 grids. For a specific grid i, s i Indicates the current state, s j The state indicating moving to the next grid cell, i,j∈(0,1,2...12), is represented by r. i = (x, y) means that x is the longitude and y is the latitude.

[0081] Figure 2 The arrows in each grid region represent known wind field information for that region. The arrow direction represents the wind direction, the arrow size represents the wind magnitude, and the wind speed is represented by w(r). i )=[w x (r i ),w y (r i )] indicates that w x (r i) represents the velocity in the longitude direction, w y (r i The wind speed (r) represents the velocity in the latitudinal direction. As the floating platform moves from region i to region j, the wind speed remains constant at w(r). i (Unchanged, it also provides a certain driving speed V) i = [u,v], the velocity V of the floating platform i,a At this point, it equals the wind speed w(r) i ) and its own driving speed V i =The vector sum of [u,v], when it reaches the next region j, the wind field information changes to w(r) j The self-driving speed also needs to be changed according to the algorithm.

[0082] The environmental state S in the MDP is represented by the wind direction and speed in each grid cell. To ensure consistent angles during calculation, the wind direction angle is defined as [-180°, 180°]. Specific angle divisions and direction definitions are as follows: Figure 3 As shown, there are 8 directions: (1) North: [67.5°, 112.5°]; (2) East: [-22.5°, 22.5°]; (3) South: [-67.5°, -112.5°]; (4) West: [-180°, -157.5°]∪[157.5°, 180°]; (5) Northeast: [22.5°, 67.5°]; (6) Southeast: [-67.5°, -22.5°]; (7) Southwest: [-157.5°, -112.5°]; (8) Northwest: [112.5°, 157.5°].

[0083] Step 1.2: Establish the mathematical model of the floating platform motion set A:

[0084] By adjusting the multi-vector propulsion system through the controller, the aerostat can obtain eight driving velocity directions, namely: [0, v...]. m ], [u m [,0],[0,-v] m ],[-u m ,0],[u m ,v m ], [u m ,-v m ],[-u m ,-v m ],[-u m ,v m That is, 8 basic movements. Similarly, setting the positive x-axis direction as due east and the positive y-axis direction as due north, the basic movement directions of the floating platform are as follows: Figure 4 As shown, the angles of each movement direction are all... Figure 3 Consistent with the above.

[0085] Step 1.3: Establish the mathematical model of the state transition probability set P:

[0086] By unifying the random variations in the magnitude and direction of wind speed into the variation in the direction of the actual resultant velocity of the aerostat, the probability distribution of the resultant velocity direction can be considered to be a Gaussian distribution. A mathematical model of the variation in the resultant velocity direction yields the following probability density function:

[0087]

[0088] in, Indicates the direction of the current resultant velocity in the region. The entire directional distribution unfolds around it, σ i =k / u m The value is related to the maximum speed in a single direction that the floating platform can provide, and k is the intensity coefficient of the randomness of the wind field.

[0089] From the above analysis, it is easy to see that the state transition probability of the floating platform under random wind field can be obtained by integrating the above equation:

[0090]

[0091] Wherein, P(s) j |s i a) indicates that the floating platform is located in region s. i The probability of transitioning from action a to region sj, given by angle θ. ij for Figure 3 The boundary angles for dividing the direction can take values ​​of ±22.5°, ±67.5°, ±112.5°, and ±157.5°, for a total of 12×12×8×8 values. It should be noted that the following two cases need to be considered due to the angle division: when... When calculating θ ij When integrating θ = -157.5°, you must first convert θ to θ. ij Adding a 360° transformation to the [202.5°, 247.5°] range; when When calculating θ ij =112.5° and θ ij When integrating θ = 157.5°, you must first convert θ to 157.5°. ij Subtracting 360° converts the values ​​to the ranges [-247.5°, -202.5°] and [-202.5°, -180°], respectively.

[0092] Step 1.4: Establish the mathematical model of the reward function R:

[0093] The reward function needs to reflect the relationship between the state, action, and goal of the floating platform, so the floating platform in state s is used. iIf action a is selected, the state transitions to state s. j The time consumed is represented by a negative value, denoted as R(s). j |s i a) There are a total of 12×12×8 values. The floating platform starts from state s i Transfer to s j The time consumed in the process, Δt, can be expressed as:

[0094]

[0095] Where, d ij The distance between two grid cells, ||V i,a || represents the current actual speed of the floating platform, which is obtained by combining the current wind speed and the platform's own speed. Therefore, to minimize the global time required to reach the target area, the reward function is set as follows:

[0096]

[0097] Step 1.5: Based on the mathematical models of the environmental state set S, the floating platform action set A, the state transition probability set P, and the reward function R established in Steps 1.1 to 1.4, the mathematical model M of the task of reaching the target point in the shortest time from a specific area is expressed as:

[0098] M = (S, A, P, R, γ),

[0099] Where S represents the set of states of the environment, and state s t ∈S is a state in the environmental state set at time t; A represents the action set of the floating platform, and action a t ∈A is an action in the action set at time t; P represents the set of state transition probabilities, and the state transition probability P(s) is... t+1 |s t a) indicates that the floating platform reaches state s by performing action a in state st. t+1 The probability of the floating platform receiving an instant reward; R represents the instant reward R(s). t+1 |s t a) indicates that the floating platform is in state s t The next action a reaches state s. t+1 The immediate reward received; γ represents the discount factor, γ∈[0,1] is used to calculate the cumulative reward of the floating platform, and its value indicates the value of the platform in the current state s. t With future states s t+1 The degree to which immediate rewards are considered; the larger γ is, the greater the impact of subsequent rewards on the current reward.

[0100] Step 2: Based on the mathematical model established in Step 1, and using the value iteration method, the optimal action sequence of the floating platform is obtained, as follows:

[0101] The solution process is as follows: Figure 5 As shown, an environmental state set S is first established using wind field data, followed by the introduction of a stochastic wind field model f. GS The state transition probability P is calculated by combining the input action set A and the target point position, and the immediate reward R is calculated. Then, a value iteration algorithm is introduced, with a discount factor γ = 0.95 and a value iteration convergence threshold ε = 0.001. After initializing the value function in the value iteration algorithm, the loop iteration begins. During the loop, if the difference between the state value functions reaches the value iteration convergence threshold, the loop is exited; otherwise, the larger value of the state value function is taken and the iteration continues until the loop ends. The calculated value function is the optimal state value function. Finally, the optimal action sequence is extracted through the optimal state value function. Figure 7 This represents the optimal action sequence for the floating platform in each region under a random wind field with a driving speed of 10 m / s.

[0102] Step 3: Establish a six-degree-of-freedom dynamic model of the floating platform to facilitate the subsequent establishment of a horizontal dynamic model of the floating platform, as follows:

[0103] (1) By analyzing the forces acting on the floating platform during its motion and combining the structural parameters of the floating platform, the dynamic force term F is obtained. f Aerodynamic force F generated by relative motion in a fluid A Engine thrust F T Self-gravity F G And the buoyancy F of the displaced air B And define the control input U = F for the six-degree-of-freedom dynamic model of the floating platform. T ;

[0104] (2) Using the parameters obtained in (1), a six-degree-of-freedom dynamic model of the floating platform is established according to the Newton-Euler method:

[0105]

[0106] Where: ξ=[uvwpqr] T Let U represent the total velocity vector, where u, v, and w are the components of the velocity vector in the body coordinate system along the X, Y, and Z axes, respectively; p, q, and r are the components of the angular velocity vector in the body coordinate system along the X, Y, and Z axes, respectively; U represents the control input of the six-degree-of-freedom dynamic model of the floating platform; and M is the mass matrix of the floating platform, as shown below:

[0107]

[0108] Where m is the mass of the floating platform, m ij (i,j=1,2,...,6) represents the additional mass, I x ,I y ,I Z I represents the moments of inertia about the X, Y, and Z axes, respectively. xz Represents the product of inertia about the XOY plane, (x G ,y G ,z G () represents the coordinates of the center of gravity of the floating platform in the body coordinate system;

[0109] F f The dynamic force term is represented as follows:

[0110]

[0111] F A The aerodynamic term is represented as follows:

[0112]

[0113] in, This is the coordinate transformation matrix from the velocity coordinate system to the airship body coordinate system. For dynamic pressure, ρ 空气 air density, V represents the magnitude of the flight speed. a C represents the volume of the floating platform. X C is the drag coefficient. Y C is the lateral force coefficient. Z C is the lift coefficient. L C is the rolling moment coefficient. M C is the pitching moment coefficient. N These are the yaw moment coefficients, collectively referred to as aerodynamic coefficients;

[0114] F G The term representing its own gravity is as follows:

[0115]

[0116] Where G = mg is the magnitude of the gravity acting on the floating platform. Let r be the transformation matrix from the ground inertial coordinate system to the body coordinate system. G Let φ be the radius vector from the center of buoyancy to the center of gravity, and let φ, θ, and ψ be the roll angle, pitch angle, and yaw angle of the floating platform, respectively.

[0117] F B The buoyancy term is represented as follows:

[0118]

[0119] Where, B1=ρ 空气 gV a The magnitude of the buoyancy force on the floating platform is zero because the origin of the machine system's coordinate system coincides with the center of buoyancy.

[0120] Based on the six-degree-of-freedom dynamic model of the floating platform, a horizontal dynamic model of the floating platform is established as follows:

[0121] In the horizontal plane, we have w = p = q = 0 and z = θ = φ = 0. Substituting these into the six-degree-of-freedom dynamic model of the floating platform, we obtain the dynamic equations for the horizontal plane:

[0122]

[0123] Among them, X a =-(1 / 2)ρ h (u 2 +v 2 V a 2 / 3 C X , ρ h β represents the air density at height h, and β represents the sideslip angle of the floating platform. All are aerodynamic derivatives, which are written in matrix form for ease of description and derivation:

[0124]

[0125] Where λ = [uvr] T u and v represent the components of the velocity vector on the X and Y axes in the coordinate system of the floating platform, respectively, and r represents the component of the angular velocity vector on the Z axis; M A =diag(m+m) 11 ,m+m 22 ,I Z +m 66 ), diag(·) denotes a diagonal matrix, U λ =[U1 U2 U3] T This represents the control input to the horizontal plane dynamics model. F a =[X a cosβ+Y a sinβ-X a sinβ+Y a cosβN a ] T The aerodynamic term representing the horizontal plane, X a =-(1 / 2)ρ h (u 2 +v 2V a 2 / 3 C X , ρ h β represents the air density at height h, and β represents the sideslip angle of the floating platform. Both are aerodynamic derivatives, U λ This represents the control input to the dynamic model of the horizontal plane of the floating platform.

[0126] Step 4: Combining the horizontal dynamic model of the floating platform derived in Step 3, add a random perturbation term to obtain the stochastic dynamic model dλ of the horizontal surface of the floating platform. This facilitates the subsequent application of the optimal action sequence obtained in Step 2, as follows:

[0127] Step 4.1: Based on the dynamic model of the floating platform's horizontal plane established in Step 3, let... The dynamic model of the horizontal plane can then be simplified to:

[0128]

[0129] in,

[0130] Step 4.2: Introduce random perturbations into the dynamic model established in Step 4.1, and extend it to:

[0131]

[0132] Where, n λ The coefficients used to determine the intensity of random perturbations in the model;

[0133] Step 4.3: Set n λ If f(λ) is considered as a Gaussian random perturbation generated by white noise excitation, then the horizontal surface dynamic model in step 4.2 can be transformed into the following horizontal surface stochastic dynamic model:

[0134]

[0135] in, λ=[uvr] T u and v represent the components of the velocity vector on the X and Y axes in the coordinate system of the floating platform, respectively, and r represents the component of the angular velocity vector on the Z axis; M A =diag(m+m) 11 ,m+m 22 ,I Z +m 66 ), m 11 m 22 m 66 For added mass, I Z These represent the moments of inertia about the Z-axis. F a =[X a cosβ+Y a sinβ-X a sinβ+Y a cosβN a ] T The aerodynamic term representing the horizontal plane, X a =-(1 / 2)ρ h (u 2 +v 2 V a 2 / 3 C X , ρ h β represents the air density at height h, and β represents the sideslip angle of the floating platform. All are aerodynamic derivatives. n λ ω represents the random perturbation intensity coefficient of the model. λ It is a 3-dimensional independent standard Wiener process vector.

[0136] Step 5: Combining the horizontal stochastic dynamics model of the floating platform obtained in Step 4, design the level flight stochastic velocity control law and adaptive law based on the backstepping method, as follows:

[0137] Step 5.1: Based on the backstepping method, define the tracking error as:

[0138] e λ =λ-λ d ,

[0139] Where λ d The target speed value;

[0140] Step 5.2: Based on the stochastic dynamics model of the floating platform's horizontal plane, the tracking error e defined in Step 5.1 is... λ Differentiating, we get:

[0141]

[0142] Step 5.3: Based on the differential of the tracking error in Step 5.2, de λ Consider the following Lyapunov function in the form of:

[0143]

[0144] Where, γ λ For the design of positive constants, The estimation error of the random disturbance intensity coefficient of the model is . This is the error estimate;

[0145] Step 5.4: For the Lyapunov function V designed in Step 5.3... λ Find the generator of infinitesimals:

[0146]

[0147] in, For Lyapunov candidate functions V λ The infinitesimal generator, k λ It is a positive number;

[0148] Step 5.5: Apply the principle of control stability of stochastic systems to the infinitesimal generator obtained in Step 5.4. To meet the requirements, we analyze the expression on the right-hand side of the inequality sign for infinitesimal generators. It needs to satisfy the stochastic control stability condition. Therefore, the control law and adaptive law are designed as follows:

[0149]

[0150] Among them, U λ To control the input, M A =diag(m+m) 11 ,m+m 22 ,I Z +m 66 ), where m is the mass of the floating platform, m 11 m 22 m 66 For added mass, I Z Let f(λ) represent the moment of inertia about the Z-axis, and M represent the moment of inertia about the Z-axis, respectively. A -1 (G(λ)M A λ+F a ), λ=[uvr] T u and v represent the components of the velocity vector on the X and Y axes in the coordinate system of the floating platform, respectively, and r represents the component of the angular velocity vector on the Z axis. F a =[X a cosβ+Y a sinβ-X a sinβ+Y a cosβN a ] T The aerodynamic term representing the horizontal plane, X a =-(1 / 2)ρ h (u 2 +v 2 V a 2 / 3 C X , ρ hβ represents the air density at height h, and β represents the sideslip angle of the floating platform. All are aerodynamic derivatives. K λ k λ γ λ δ λ All are positive constants of the design. e represents the estimated value of the model's random disturbance intensity coefficient. λ λ represents the backstepping tracking error. d Indicates the desired speed.

[0151] Step 6: Using the control law and adaptive law designed in Step 5, drive the aerobatic platform to achieve rapid deployment in a random wind field during level flight.

[0152] The estimated values ​​of the random disturbance intensity coefficients of the model are obtained by solving the adaptive law. The velocity λ of the floating platform at different moments and the estimated value obtained from the solution are used. Substituting the values ​​into the control law, we obtain the control input U required by the aerodynamic system of the floating platform. λ Different thrust F are established based on the different power system layouts. T The expression, and by U λ =F T The relationship is solved to calculate the magnitude and direction of the thrust so that the error between the actual speed and the desired speed approaches zero. By navigating in each wind field according to the optimal action sequence solved in step 2, the floating platform can be rapidly deployed in a random wind field.

[0153] A rapid deployment control system for a floating platform in a stochastic wind field includes:

[0154] The model building module, based on the Markov decision process, establishes a mathematical model of the shortest time task for an airborne platform to reach a target point in a specific area under a random wind field. It also uses the value iteration method to find the optimal action sequence of the airborne platform. Simultaneously, it establishes a six-degree-of-freedom dynamic model of the airborne platform, then establishes a horizontal dynamic model of the airborne platform, and finally adds a random perturbation term to the model to obtain a horizontal random dynamic model of the airborne platform. This is used to implement steps 1 to 4 of a method for rapid deployment control of an airborne platform in a random wind field.

[0155] The control law and adaptive law design module, combined with the horizontal plane stochastic dynamics model of the aerostat platform, designs the level flight stochastic speed control law and adaptive law based on the backstepping method, which is used to realize a rapid deployment control method for level flight of an aerostat platform in a stochastic wind field; Step 5:

[0156] The deployment module, using control and adaptive laws, drives the aerobatic platform to navigate according to an action sequence in a random wind field, thereby achieving rapid deployment control of the aerobatic platform in level flight. This is step 6 of a method for rapid deployment control of an aerobatic platform in a random wind field.

[0157] A rapid deployment control device for a floating platform in a random wind field during level flight includes:

[0158] Memory: Used to store the computer program that implements the method for rapid deployment control of a floating platform in a random wind field;

[0159] Processor: Used to implement the aforementioned method for rapid deployment control of a floating platform in a random wind field during the level flight when executing the computer program.

[0160] This embodiment also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method for rapid deployment control of a floating platform in a random wind field.

[0161] The effects of this invention can be further illustrated by the following simulation experiments:

[0162] I. Simulation Parameters

[0163] 1. Assume the parameters of the floating platform are shown in Table 1, all of which are in the International System of Units (SI).

[0164] Table 1 Parameters of the Aerial Platform

[0165] parameter Value (unit) parameter Value (unit) m 1200 (kg) <![CDATA[z G ]]> 2(m) <![CDATA[k m1 ]]> 0.079 <![CDATA[I x ]]> <![CDATA[9024(kgm 2 )]]> <![CDATA[k m2 ]]> 0.869 <![CDATA[I y 、I z ]]> <![CDATA[16900(kgm 2 )]]> <![CDATA[k m3 ]]> 0.631 <![CDATA[I xz ]]> <![CDATA[2198(kgm 2 )]]> <![CDATA[x G ]]> 0(m) <![CDATA[V a ]]> <![CDATA[1800(m 3 )]]>

[0166] 2. Initial velocity values ​​are set to u0 = 2 m / s, v0 = 0 m / s, r = 0.1 rad / s, and altitude is set to 3 km. Controller parameters are set to: K λ =0.5I 3×3 k λ =0.01, the adaptive parameter is set to: γ λ =δ λ =0.001, adaptive initial value is 0, and the model random disturbance intensity coefficient is set to 10. -9 The simulation time is 90 seconds.

[0167] II. Simulation Content

[0168] To include scenarios where the wind field varies significantly over time, and to demonstrate the ability to control changes in airship speed, we assume the wind field changes once every 90 seconds, with a wind intensity coefficient of k = 1.5. The maximum propulsion speed the airship can achieve is then determined. m =v m=10 m / s. The environmental wind field state at t=0 s is as follows: Figure 2 As shown, the optimal action sequence obtained from the decision is as follows: Figure 6 As shown; the environmental wind field state at t=45s is as follows Figure 7 As shown, the optimal action sequence obtained using the decision algorithm described above is as follows: Figure 8 As shown.

[0169] In this simulation experiment, the control objective is to enable the airship to adjust its own speed according to the driving speed sequence obtained by solving the MDP in the above embodiment. The goal is simply to ensure that the airship reaches the required maximum speed within a short time. It should be noted that in the horizontal plane speed control, the airship's kinematic performance is not considered; the yaw angle ψ is assumed to be 0. Therefore, the velocity in the ground coordinate system is the same as the velocity in the airship's hull coordinate system, expressed as u. m =v m Taking 10m / s as an example, the expected speed value is set as follows:

[0170]

[0171] Since the yaw angle remains constant, the yaw rate r is also always set to 0. How the yaw angle changes in practice needs further analysis based on the specific circumstances. The response curves of the X-axis velocity u, Y-axis velocity v, and yaw rate r in the simulation experiment of horizontal deployment of an airborne platform against random wind interference are shown below. Figure 9 , 10 As shown in Figure 11, using the control law of this invention, the X-axis velocity u and Y-axis velocity v can both reach the desired velocity within 20 seconds, the yaw rate r reaches the desired angular velocity within 15 seconds and fluctuates up and down, and can completely track the desired angular velocity within 60 seconds. This control law is used for tracking... Figure 8 The optimal action sequence in each wind field region enables rapid deployment of the aerostat in horizontal flight. Compared with existing technologies, the innovation of this invention lies in: introducing a Markov decision process to consider the influence of random wind fields during the horizontal deployment of the aerostat, thus solving for the globally optimal action sequence; establishing a new stochastic dynamic model of the aerostat in the horizontal plane; and designing a control law that satisfies the stability of the system against random disturbances based on the backstepping method.

[0172] Based on the above simulation results, the method of the present invention can enable the airborne platform to effectively resist random wind interference during motion, quickly track the desired target, and reach the desired position in a short time, thereby improving the rapid deployment effect of the airborne platform in random wind fields, effectively suppressing the influence of random wind fields on the deployment accuracy of the airborne platform, improving the robustness of the level flight rapid deployment control of the airborne platform, and realizing the level flight rapid deployment control of the airborne platform.

[0173] The above description is merely a specific example of the present invention and does not constitute any limitation on the present invention. Obviously, those skilled in the art, after understanding the content and principles of the present invention, may make various modifications and changes in form and details without departing from the principles and structure of the present invention. However, these modifications and changes based on the ideas of the present invention are still within the scope of protection of the claims of the present invention.

[0174] It should be noted that the step numbers in the specification and claims of this invention are only for the purpose of clearly describing the embodiments of this invention and facilitating understanding, and their order is not limited.

Claims

1. A method for rapid deployment and control of a floating platform in a stochastic wind field, characterized in that, Includes the following steps: Step 1: Based on the Markov decision process, establish a mathematical model for the task of the floating platform traveling to the target point in the shortest time within a specific area under a random wind field; Step 2: Based on the mathematical model established in Step 1, the optimal action sequence of the floating platform is obtained by combining the value iteration solution method; Step 3: Establish a six-degree-of-freedom dynamic model of the floating platform, and based on the six-degree-of-freedom dynamic model of the floating platform, establish a horizontal dynamic model of the floating platform; Step 4: Based on the horizontal dynamic model of the floating platform established in Step 3, add a random perturbation term to obtain the stochastic dynamic model of the horizontal surface of the floating platform. The stochastic dynamic model of the horizontal surface of the floating platform is expressed as: in, λ=[uvr] T u and v represent the components of the velocity vector on the X and Y axes in the coordinate system of the floating platform, respectively, and r represents the component of the angular velocity vector on the Z axis; M A =diag(m+m) 11 ,m+m 22 ,I Z +m 66 ), m 11 m 22 m 66 For added mass, I Z These represent the moments of inertia about the Z-axis; F a =[X a cosβ+Y a sinβ-X a sinβ+Y a cosβN a ] T The aerodynamic term representing the horizontal plane, X a =-(1 / 2)ρ h (u 2 +v 2 V a 2 / 3 C X , ρ h β represents the air density at height h, and β represents the sideslip angle of the floating platform. All are aerodynamic derivatives; n λ ω represents the random perturbation intensity coefficient of the model. λ A 3-dimensional independent standard Wiener process vector; Step 5: Combining the stochastic dynamics model of the floating platform on the horizontal plane obtained in Step 4, design the level flight stochastic velocity control law and adaptive law based on the backstepping method. The control law is expressed as: The adaptive law is expressed as: Among them, U λ To control the input, M A =diag(m+m) 11 ,m+m 22 ,I Z +m 66 ), where m is the mass of the floating platform, m 11 m 22 m 66 For added mass, I Z Let represent the moment of inertia about the Z-axis, respectively. λ=[uvr] T u and v represent the components of the velocity vector on the X and Y axes in the coordinate system of the floating platform, respectively, and r represents the component of the angular velocity vector on the Z axis. F a =[X a cosβ+Y a sinβ-X a sinβ+Y a cosβN a ] T The aerodynamic term representing the horizontal plane, X a =-(1 / 2)ρ h (u 2 +v 2 V a 2 / 3 C X , ρ h β represents the air density at height h, and β represents the sideslip angle of the floating platform. All are aerodynamic derivatives; K λ k λ γ λ δ λ All are positive constants of the design. e represents the estimated value of the model's random disturbance intensity coefficient. λ λ represents the backstepping tracking error. d Indicates the desired speed; Step 6: Using the control law and adaptive law designed in Step 5, drive the aerobatic platform to navigate in the random wind field according to the action sequence solved in Step 2, so as to realize the rapid deployment control of the aerobatic platform in level flight.

2. The method for rapid deployment and control of a floating platform in a random wind field according to claim 1, characterized in that, The mathematical model described in step 1 is expressed as follows: M = (S, A, P, R, γ), Where S represents the set of states of the environment, and state s t ∈S is a state in the environmental state set at time t; A represents the action set of the floating platform, and action a t ∈A is an action in the action set at time t; P represents the set of state transition probabilities, and the state transition probability P(s) is... t+1 |s t a) indicates that the floating platform is in state s t The state s is reached by performing action a. t+1 The probability; R represents the instant reward received by the floating platform, R(s) t+1 |s t a) indicates that the floating platform is in state s t The next action a reaches state s. t+1 The immediate reward received; γ∈[0,1] represents the discount factor, used to calculate the cumulative reward of the floating platform, and its value indicates the value of the platform in the current state s. t With future states s t+1 The degree to which immediate rewards are considered; the larger γ is, the greater the impact of subsequent rewards on the current reward.

3. The method for rapid deployment and control of a floating platform in a random wind field during level flight according to claim 2, characterized in that, Step 2 specifically includes: First, an environmental state set S is established using wind field data, followed by the introduction of a stochastic wind field model f. GS The system calculates the state transition probability set P by combining the input action set A and the target point position, and calculates the immediate reward R. Then, it introduces a value iteration algorithm, sets a discount factor γ and a value iteration convergence threshold ε, initializes the state value function in the value iteration algorithm, and enters a loop iteration. The difference between the two state value functions is then calculated. During the loop, if the difference of the state value function reaches the value iteration convergence threshold ε, the loop is exited; otherwise, the larger value of the state value function is taken and the iteration continues until the loop ends. The resulting state value function is the optimal state value function, and the optimal action sequence is extracted through the optimal state value function.

4. The method for rapid deployment and control of a floating platform in a random wind field during level flight according to claim 1, characterized in that, The six-degree-of-freedom dynamic model of the floating platform established in step 3 is represented as follows: Where, ξ=[uvwpqr] T Let u, v, and w represent the velocity vectors in the body coordinate system along the X, Y, and Z axes, respectively. Let p, q, and r represent the angular velocity vectors in the body coordinate system along the X, Y, and Z axes, respectively. Let U represent the control input of the six-degree-of-freedom dynamic model of the floating platform, and M represent the mass matrix of the floating platform, as shown below: Where m is the mass of the floating platform, m ij (i,j=1,2,...,6) represents the additional mass, I x ,I y ,I Z I represents the moments of inertia about the X, Y, and Z axes, respectively. xz Represents the product of inertia about the XOY plane, (x G ,y G ,z G () represents the coordinates of the center of gravity of the floating platform in the body coordinate system; F f The dynamic force term is represented as follows: F A The aerodynamic term is represented as follows: in, This is the coordinate transformation matrix from the velocity coordinate system to the body coordinate system. For dynamic pressure, ρ 空气 air density, V represents the magnitude of the flight speed. a C represents the volume of the floating platform. X C is the drag coefficient. Y C is the lateral force coefficient. Z C is the lift coefficient. L C is the rolling moment coefficient. M C is the pitching moment coefficient. N These are the yaw moment coefficients, collectively referred to as aerodynamic coefficients; F G The term representing its own gravity is as follows: Where G = mg is the magnitude of the gravity acting on the floating platform. Let r be the transformation matrix from the ground inertial coordinate system to the body coordinate system. G Let φ be the radius vector from the center of buoyancy to the center of gravity, and let φ, θ, and ψ be the roll angle, pitch angle, and yaw angle of the floating platform, respectively. F B The buoyancy term is represented as follows: Where, B1=ρ 空气 gV a The magnitude of the buoyancy force on the floating platform is zero because the origin of the machine system's coordinate system coincides with the center of buoyancy. The establishment of the horizontal dynamic model of the floating platform is achieved as follows: Where λ = [uvr] T u and v represent the components of the velocity vector on the X and Y axes in the coordinate system of the floating platform, respectively, and r represents the component of the angular velocity vector on the Z axis; M A =diag(m+m) 11 ,m+m 22 ,I Z +m 66 ), where diag(·) denotes a diagonal matrix; F a =[X a cosβ+Y a sinβ-X a sinβ+Y a cosβN a ] T The aerodynamic term representing the horizontal plane, X a =-(1 / 2)ρ h (u 2 +v 2 V a 2 / 3 C X , ρ h β represents the air density at height h, and β represents the sideslip angle of the floating platform. Both are aerodynamic derivatives, U λ This represents the control input to the dynamic model of the horizontal plane of the floating platform.

5. The method for rapid deployment and control of a floating platform in a random wind field during level flight according to claim 1, characterized in that, Step 6 specifically includes: The estimated values ​​of the random disturbance intensity coefficients of the model are obtained by solving the adaptive law. The velocity λ of the floating platform at different moments and the estimated value obtained from the solution are used. Substituting the values ​​into the control law, we obtain the control input U required by the aerodynamic system of the floating platform. λ Then, based on the different power system layouts, different horizontal thrust F is established. T The expression, and by U λ =F T The relationship is used to calculate the magnitude and direction of the horizontal thrust, so that the error between the actual speed and the desired speed approaches zero. By navigating in each wind field according to the optimal action sequence solved in step 2, the rapid deployment control of the airborne platform in level flight can be achieved.

6. A system based on the rapid deployment control method for a floating platform in a random wind field during level flight according to any one of claims 1 to 5, characterized in that, include: The model building module, based on the Markov decision process, establishes a mathematical model of the shortest time task for an airborne platform to reach a target point in a specific region under a random wind field. It also uses the value iteration method to find the optimal action sequence of the airborne platform. At the same time, it establishes a six-degree-of-freedom dynamic model of the airborne platform, then establishes a horizontal dynamic model of the airborne platform, and finally adds a random perturbation term to the model to obtain a random dynamic model of the horizontal plane of the airborne platform. The control law and adaptive law design module combines the stochastic dynamics model of the airborne platform on the horizontal plane and designs the level flight stochastic speed control law and adaptive law based on the backstepping method. The deployment module utilizes control and adaptive laws to drive the aerobatic platform to navigate in a random wind field according to an action sequence, thereby achieving rapid deployment control of the aerobatic platform in level flight.

7. A rapid deployment control device for a floating platform in a random wind field, characterized in that, include: Memory: Used to store a computer program for implementing the rapid deployment control method for level flight of an airborne platform in a random wind field as described in any one of claims 1 to 5; Processor: Used to implement the rapid deployment control method for level flight of an airborne platform in a random wind field as described in any one of claims 1 to 5 when executing the computer program.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of a method for rapid deployment control of a floating platform in a random wind field as described in any one of claims 1 to 5.

Images