Cross-domain motion planning method and system for air-ground amphibious robot

Abstract

The invention discloses a cross-domain motion planning method and system for an air-ground amphibious robot, and relates to the technical field of robot control. The method comprises the following steps: acquiring operation constraints of the air-ground amphibious robot in a working space; performing path planning according to the operation constraint by using a particle swarm optimization algorithm; constructing a kinetic model, and constructing a wing folding angle smooth switching track by using a Gaussian pseudo-spectrum method and taking the minimum flight height decrease in the modal switching process as a target; and setting a nonlinear controller according to the kinetic model and a fuzzy algorithm, controlling the air-ground amphibious robot by using the nonlinear controller according to the optimal air-ground cross-domain path, and in addition, performing smooth switching in the flight process according to the wing folding angle smooth switching track. According to the method, an adaptive fuzzy position controller is designed to process nonlinear uncertainty in a dynamic model, and cross-domain motion planning is performed based on an energy optimal path and smooth mode switching control, so that the air-ground amphibious robot can realize stable motion and reliable operation in a cross-domain manner.

Description

Technical Field

[0001] The present invention relates to the technical field of robot control, and particularly to a cross-domain motion planning method and system for an amphibious land-air robot. Background Art

[0002] The statements in this section merely provide background technical information related to the present invention and do not necessarily constitute prior art.

[0003] Robots are widely used in industries such as power inspection, logistics transportation, and industrial production. However, traditional robots are limited to operating in a single environment and cannot meet the requirements of complex scenarios. To address these challenges, amphibious robots with excellent mobility, adaptability, and flexibility are used in applications such as disaster rescue, environmental monitoring, military operations, and agricultural and industrial fields. Among them, amphibious land-air robots have attracted much attention due to their wide application range, low energy consumption, and high operation efficiency.

[0004] Generally, an amphibious land-air robot is equipped with multiple rotors for flight in the air and multiple wheels for movement on the ground, so as to achieve cross-domain operation. However, due to the integration of the ground or air modality, the amphibious land-air robot faces various challenges, such as complex non-linear models, uncertain dynamic characteristics, system uncertainties, and instabilities caused by multi-modal operation.

[0005] The above problems make the motion planning and control of the amphibious land-air robot extremely difficult. In terms of the planning of the amphibious land-air robot, some researchers have proposed a hierarchical motion planner to generate a ground-air hybrid path and used the B-spline method to enhance safety, smoothness, and compliance with dynamic constraints. Some other researchers have formulated an optimal control problem and used the trapezoidal coordination method for numerical solution to obtain the shortest time trajectory under spatially varying constraints. Some have also proposed a non-linear model predictive control method considering hybrid dynamics complementary constraints and evaluated the uncertainty bounds along the nominal trajectory in a rolling horizon manner to ensure safety and robustness.

[0006] In addition, some progress has been made in the control of the amphibious land-air robot, and many efforts have achieved commendable results. In existing research, a non-linear controller with adaptive characteristics has been designed to offset the influence of non-linear uncertainties in the model and track air or ground trajectories. Some other research has developed a non-linear, multi-input, multi-output sliding mode controller based on the model and used the Lyapunov direct method to prove its robustness to bounded parameter uncertainties. Some research has also analyzed the coupled non-linear dynamics of the aircraft landing process, where the rotor is used as an active regulator to reduce the vibration of the tire suspension after landing. In addition, a model predictive control method with control allocation has been designed to improve the landing stability of the aircraft.

[0007] However, despite the above efforts, there are still the following problems with the amphibious land-air robot: 1) Due to the cross-domain movement characteristics, the motion planning of the amphibious land-air robot is a complex problem with constraints. 2) Due to the dynamic system conversion, the mode switching process during the cross-domain movement of the amphibious land-air robot is extremely unstable. However, few people have solved this problem. 3) The amphibious land-air robot exhibits highly nonlinear, strongly coupled, and uncertain dynamics.

[0008] Therefore, how to achieve precise and stable tracking control of the amphibious land-air robot for the complex cross-domain motion planning, dynamic mode switching, and nonlinear strong coupling dynamics during the motion process is a major challenge. Summary of the Invention

[0009] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide a cross-domain motion planning method and system for an amphibious land-air robot, design an adaptive fuzzy position controller to handle the nonlinear uncertainties in the dynamic model, and perform cross-domain motion planning based on the energy-optimal path and smooth mode switching control, so that the amphibious land-air robot can achieve stable motion and reliable operation across domains.

[0010] To achieve the above purpose, the present invention is implemented through the following technical solutions:

[0011] The first aspect of the present invention provides a cross-domain motion planning method for an amphibious land-air robot, including the following steps:

[0012] Determine the working space of the amphibious land-air robot according to the task requirements, and obtain the operation constraints of the amphibious land-air robot within the working space;

[0013] Use the particle swarm optimization algorithm to perform path planning according to the operation constraints to obtain the optimal land-air cross-domain path;

[0014] Construct a dynamic model according to the dynamic parameters of the amphibious land-air robot, and use the Gaussian pseudospectral method to construct a smooth switching trajectory of the wing folding angle with the goal of minimizing the flight altitude drop during the mode switching process;

[0015] Set a nonlinear controller according to the dynamic model and the fuzzy algorithm, use the nonlinear controller to control the amphibious land-air robot according to the optimal land-air cross-domain path, and perform smooth switching during the flight according to the smooth switching trajectory of the wing folding angle.

[0016] Furthermore, the operation constraints include energy consumption cost constraint, collision cost constraint, and distance cost constraint. Among them, the energy consumption cost constraint is set according to the selection and switching of motion modes, the collision cost constraint is set according to the positional relationship between the amphibious robot and the obstacles in the working space, and the distance cost constraint is determined according to the segmented path between the amphibious robot and the destination.

[0017] Furthermore, the specific steps for using the particle swarm optimization algorithm to perform path planning according to the operation constraints are as follows:

[0018] Define the optimization objective function for path planning;

[0019] Set the velocity and position update laws for each particle;

[0020] Design a new inertia weight considering the problem of being prone to falling into local optimum;

[0021] Based on the update law and the new inertia weight, perform iterative solution on the optimization objective function to obtain the optimal land-air cross-domain path.

[0022] Furthermore, the dynamic model is the dynamic model of the amphibious robot in the air flight mode when the wing folding angle is zero.

[0023] Furthermore, the specific steps for using the Gaussian pseudospectral method to construct a smooth switching trajectory of the wing folding angle with the goal of minimizing the flight altitude drop during the mode switching process are as follows:

[0024] Transform the dynamic model with the goal of minimizing the flight altitude drop during the mode switching process;

[0025] Set the dynamic constraints;

[0026] Based on the transformed dynamic model and dynamic constraints, construct a nonlinear programming function with algebraic constraints;

[0027] Solve the nonlinear programming function to obtain the optimal state sequence of the descent altitude.

[0028] Even further, the dynamic constraints include the folding angle constraint and the flight descent constraint of the amphibious robot.

[0029] Furthermore, the specific steps for setting the nonlinear controller according to the dynamic model and the fuzzy algorithm are as follows:

[0030] Construct the position control system function of the amphibious robot according to the dynamic model;

[0031] Introduce the fuzzy algorithm according to the nonlinear interference term in the position control system function to construct an adaptive fuzzy controller.

[0032] The second aspect of the present invention provides a cross-domain motion planning system for an amphibious robot, including:

[0033] A constraint determination module, configured to determine the working space of the amphibious aerial-ground robot according to the task requirements, and obtain the operation constraints of the amphibious aerial-ground robot within the working space;

[0034] A path optimization module, configured to use the particle swarm optimization algorithm to perform path planning according to the operation constraints to obtain an optimal aerial-ground cross-domain path;

[0035] A wing switching module, configured to construct a dynamic model according to the dynamic parameters of the amphibious aerial-ground robot, and use the Gaussian pseudospectral method to construct a smooth switching trajectory of the wing folding angle with the goal of minimizing the flight altitude drop during the mode switching process;

[0036] A motion control module, configured to set a non-linear controller according to the dynamic model and the fuzzy algorithm, and use the non-linear controller to control the amphibious aerial-ground robot according to the optimal aerial-ground cross-domain path. Additionally, perform smooth switching during the flight according to the smooth switching trajectory of the wing folding angle.

[0037] The third aspect of the present invention provides a medium, on which a program is stored, and when the program is executed by a processor, it implements the steps in the cross-domain motion planning method of the amphibious aerial-ground robot as described in the first aspect of the present invention.

[0038] The fourth aspect of the present invention provides a device, including a memory, a processor, and a program stored on the memory and executable on the processor. When the processor executes the program, it implements the steps in the cross-domain motion planning method of the amphibious aerial-ground robot as described in the first aspect of the present invention.

[0039] The above one or more technical solutions have the following beneficial effects:

[0040] The present invention discloses a cross-domain motion planning method and system for an amphibious aerial-ground robot, aiming to solve the problems of complex cross-domain motion planning, dynamic mode switching, and strong non-linear coupling dynamics during the motion process of the current amphibious aerial-ground robot. The present invention designs a new path planning method based on an improved particle swarm optimization algorithm to obtain an optimal aerial-ground cross-domain path with the lowest energy. Based on the Gaussian pseudospectral method and the dynamic model of the amphibious aerial-ground robot, with the goal of minimizing the flight altitude drop during the mode switching process, a smooth switching trajectory of the wing folding angle is obtained. Based on the dynamic model of the amphibious aerial-ground robot, a non-linear controller based on adaptive fuzzy for the system is obtained. The method of the present invention has a simple structure, easy-to-adjust parameters, and is easy to apply. It can realize the planning and control of the cross-domain motion of the amphibious aerial-ground robot, improve the working efficiency and operation stability of the amphibious aerial-ground robot, and is applicable to the motion planning and control of variable-structure amphibious aerial-ground robots, with the characteristics of convenient application and wide usage range.

[0041] Advantages of additional aspects of the present invention will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings forming a part of this specification are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not unduly limit the present invention.

[0043] Figure 1 is the working principle diagram of the cross-domain motion planning method for the land-air amphibious robot in the first embodiment of the present invention;

[0044] Figure 2 is the planning result diagram of the cross-domain path planning method in the first embodiment of the present invention;

[0045] Figure 3 is the cost convergence situation diagram of the cross-domain path planning method in the first embodiment of the present invention;

[0046] Figure 4 is the planning result diagram of the smooth switching trajectory planning method in the first embodiment of the present invention;

[0047] Figure 5 is the result diagram of the smooth switching trajectory method in the first embodiment of the present invention;

[0048] Figure 6 is the result diagram of the adaptive fuzzy control method in the first embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which the present invention belongs.

[0050] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof;

[0051] Embodiment 1:

[0052] The first embodiment of the present invention provides a cross-domain motion planning method for a land-air amphibious robot, as Figure 1 shown, including the following steps:

[0053] Step 1: Determine the working space of the amphibious robot according to the task requirements, and obtain the operation constraints of the amphibious robot within the working space.

[0054] Among them, the operation constraints include energy consumption cost constraint, collision cost constraint, and distance cost constraint. Among them, the energy consumption cost constraint is set according to the selection and switching of motion modes, the collision cost constraint is set according to the positional relationship between the amphibious robot and the obstacles in the working space, and the distance cost constraint is determined according to the segmented path between the amphibious robot and the destination.

[0055] Specifically,

[0056] Determine the working space of the amphibious robot according to the task requirements. The energy-optimal cross-domain path can be represented by a set of sequences p as follows:

[0057] p = {p s , p1, p2,... p k-2 , p k-1 , p g} (1).

[0058] Among them, p s is the starting point, p g is the target point, p1,..., p k-1 are the planned path nodes, and k is the number of planned path segments.

[0059] The objective of the present invention is to enable the amphibious robot to autonomously select the ground or air motion mode according to the environmental characteristics, thereby significantly improving the accessibility and efficiency. For this purpose, an energy consumption cost function is first proposed as follows:

[0060]

[0061] Among them, J energy is the total energy consumption cost of the path, E s is the energy consumption cost of the s-th path segment, 0 and 1 respectively correspond to the ground and air motion modes, E a , E t are respectively the air and ground motion energy consumption costs. Since the resistance to be overcome during air motion is much greater than that during ground motion, and the motor speed during flight is significantly higher than that during ground travel, therefore is the energy consumption coefficient.

[0062] Subsequently, in order to ensure the safety and stability of the amphibious robot during operation in a restricted working space and in the presence of complex multiple obstacles, a collision cost function is artificially set as follows:

[0063]

[0064] Among them, Jcollision is the total collision cost of the planned path, O s is the collision cost of the sth path, O t is the cost after a collision with an obstacle.

[0065] Finally, in order to ensure the shortest length of the planned path, the distance cost function is designed as follows:

[0066]

[0067] Among them, J distance is the total cost of the planned path distance, L s is the length of the sth path segment. During the planning process, collision avoidance inevitably incurs additional distance costs. However, in practice, safe robot operation is paramount. Therefore, when setting costs, the collision cost is artificially set to be much greater than the distance cost, making the additional distance cost negligible.

[0068] Step 2: Use the particle swarm optimization algorithm to plan the path according to the operation constraints and obtain the optimal land-air cross-domain path.

[0069] Step 2.1: Define the optimization objective function for path planning.

[0070] In a specific embodiment, in order to obtain the sequence p, the optimization goal is to minimize the total path cost J, which is defined as

[0071]

[0072] Among them, J c (p) represents the cost associated with each constraint, W c represents the weight coefficient of each constraint, NC is the number of constraints, and c is the constraint index. In this embodiment, NC is 3, and c = energy, collision, or distance.

[0073] Step 2.2: Set the speed and position update law for each particle.

[0074] In a specific embodiment, the energy-optimal cross-domain path is determined by cost functions (2)-(4). In order to solve the optimization problem (5), a cross-domain path planning method based on an improved particle swarm optimization algorithm is designed, in which the speed and position update law of each particle is designed as follows:

[0075]

[0076] Among them, V and X are the velocity and position of the particle, which are respectively determined by vi in the t+1th iteration. j and x ij Composition, P ib and Pgb They are the positions of the individual and global extrema respectively. ω is called the inertia weight and is a non - negative value. c1 and c2 are called learning factors and are non - negative constants. r1 and r2 are vectors composed of random values between 0 and 1.

[0077] Step 2.3: Design a new inertia weight considering the problem of being prone to falling into local optima.

[0078] The improvement of the present invention on the particle swarm optimization algorithm lies in that, considering the problem of being prone to falling into local optima, a more reasonable inertia weight is designed. Compared with the traditional inertia weight, in solving the problem of energy - optimal cross - domain path planning of the present invention, it can speed up the solving process and improve the solving accuracy. The new inertia weight is designed as follows:

[0079]

[0080] where ω M and ω m represent the upper and lower limits of the inertia weight respectively, T M is the maximum number of iterations, and λ is a set constant.

[0081] Step 2.4: Iteratively solve the optimization objective function based on the update law and the new inertia weight to obtain the optimal land - air cross - domain path.

[0082] Specifically, based on the update law (6) and the new inertia weight (7), the energy - optimal cross - domain path sequence p can be obtained by iteration.

[0083] Step 3: Construct a dynamic model according to the dynamic parameters of the land - air amphibious robot, and use the Gaussian pseudospectral method to construct a smooth switching trajectory of the wing folding angle with the goal of minimizing the flight altitude drop during the mode - switching process.

[0084] Step 3.1: Construct a dynamic model according to the dynamic parameters of the land - air amphibious robot.

[0085] The wing folding angle is the angle between the two wings of the land - air amphibious robot and the fuselage. In the normal mode, this folding angle is zero, that is, the wings and the fuselage are on the same horizontal line. During the air - flight process, when the wings fold downwards, it will inevitably cause a change in the lift force received by the robot, which will lead to an unstable running state or even crash of the robot. Therefore, this embodiment proposes a method for planning a smooth switching trajectory of the wing folding angle to achieve a smooth mode - switching of the land - air amphibious robot in the air and ensure its safe operation.

[0086] In the present invention, only the control method of the robot in the air mode is involved, so the mathematical model ignores the land mode. Among them, the dynamic model is the dynamic model of the land - air amphibious robot in the air - flight mode when the wing folding angle is zero, which is expressed as follows:

[0087]

[0088] Among them, {x n , y n , z n} is the position of the robot in the world coordinate system, and are the velocity and acceleration respectively, {φ, θ, ψ}, and are the attitude angles, angular velocity and angular acceleration of the robot, {I xx , I yy , I zz} is the moment of inertia in the world coordinate system, τ1 is the total thrust of 4 propellers, τ i , i = 2, 3, 4 are the torques in the body coordinate system, w i , i = 1, 2,..., 6 are the external uncertain disturbances, k fi , i = 1, 2, 3 are the wind resistance coefficients, m is the mass of the robot, g is the acceleration due to gravity, S Δ represents sinΔ, C Δ represents cosΔ. In the world coordinate system, the x-axis points east, the y-axis points north, and the z-axis points upward, with the starting position of the robot as the origin.

[0089] Subsequently, the z-axis dynamic model of the amphibious robot under the action of the PID controller can be expressed as follows:

[0090]

[0091] Among them, z e is the height tracking error, U0 = C ρ τ1 - mg, k p , k i , k d are the PID controller parameters, and ρ is the wing folding angle. In addition, at the initial moment, U0 = 0.

[0092] Step 3.2: Use the Gauss pseudospectral method to construct a smooth switching trajectory of the wing folding angle with the goal of minimizing the flight height drop during the mode switching process.

[0093] Step 3.2.1: Transform the dynamic model with the goal of minimizing the flight height drop during the mode switching process.

[0094] The Gauss pseudospectral method is a numerical method commonly used to solve the optimal control problems and trajectory optimization problems of dynamic systems. By analyzing the altitude dynamics model of the amphibious aerial-ground robot and considering a series of safety constraints, the present invention establishes an optimization problem. It is solved by using mathematical means such as Gaussian integration and Legendre polynomials to obtain the optimal variation trajectory of the wing folding angle, thereby reducing the impact of wing folding on the flight altitude and ensuring smooth switching.

[0095] In a specific embodiment, in order to obtain a smooth switching trajectory of the folding angle with the minimum descent height, state variables ξ and control input u are first defined to facilitate the transformation of the altitude dynamics model:

[0096]

[0097] where C ρ = cosρ, and are the first derivative and the second derivative of C ρ respectively, and is the second derivative of the altitude tracking error z e .

[0098] Subsequently, the altitude dynamics model (10) can be transformed as follows:

[0099]

[0100] where

[0101]

[0102] Step 3.2.2: Set the dynamic constraints. Among them, the dynamic constraints include the folding angle constraint and the flight descent constraint of the amphibious aerial-ground robot.

[0103] In a specific embodiment, next, in order to ensure the stability during the switching process, three groups of constraint conditions will be set.

[0104] First, set the initial values and termination values of the folding angle ρ, ξ, and u as follows:

[0105]

[0106] Then, set the folding angle change speed and acceleration constraints as follows:

[0107]

[0108] where and are the upper bounds of the folding angle change speed, the upper bound of the folding angle acceleration, the upper bound of the cosine value change speed of the folding angle, and the upper bound of the cosine value acceleration of the folding angle respectively.

[0109] In addition, the constraints on the descending height, descending speed, and descending acceleration of the robot are set as follows:

[0110]

[0111] where z emax , and are the upper bounds of the descending height, descending speed, and descending acceleration.

[0112] Step 3.2.3: Construct a non - linear programming function with algebraic constraints based on the transformed dynamic model and dynamic constraints.

[0113] In a specific implementation, select the Legendre - Gauss point sequence {t′0, t′1, t′2,..., t′ K}, and introduce the Lagrange interpolation polynomial:

[0114]

[0115] where t′ i , i = 0,..., K are the Legendre - Gauss points, represents the Lagrange interpolation polynomial function.

[0116] Then the planned trajectory can be approximately represented as

[0117]

[0118] where ξ(t′) and u(t′) represent the system state variables and inputs at t′, and the derivative of ξ(t′) with respect to time can be expressed as

[0119]

[0120] where is the derivative of the Lagrange interpolation polynomial function.

[0121] Based on the above, this optimization problem can be transformed into a non - linear programming problem with algebraic constraints, which is expressed as follows:

[0122]

[0123] where

[0124]

[0125] Step 3.2.4: Solve the non - linear programming function to obtain the optimal state sequence of the descending height.

[0126] In this embodiment, it can be solved by the quadratic programming method to obtain the optimal state sequence of the descending height:

[0127] {ξ a min (t′0), ξ a min (t′1),..., ξ a min (t′ K ), ξ a min (t′ K+1 )} (22).

[0128] Among them, ξ a min (t′ i ), i = 0,..., K + 1 are the optimal state sequence points obtained by planning.

[0129] Step 4: Set a non - linear controller according to the dynamic model and the fuzzy algorithm, and use the non - linear controller to control the amphibious robot along the optimal land - air cross - domain path. In addition, perform smooth switching during flight according to the smooth switching trajectory of the wing folding angle.

[0130] Step 4.1: Construct the position control system function of the amphibious robot according to the dynamic model.

[0131] In a specific implementation manner, according to the dynamic models (8)-(9), the position control system of the amphibious robot can be expressed as:

[0132]

[0133] Among them is the robot speed, is the robot acceleration, M is the robot mass matrix, U is the robot control input, is the non - linear interference term related to , and there is:

[0134]

[0135] The meanings of the variables have been explained in formulas (8) and (9).

[0136] Step 4.2: Introduce the fuzzy algorithm according to the non - linear interference term in the position control system function to construct an adaptive fuzzy controller.

[0137] In a specific implementation manner. To handle the non - linear interference term introduce a fuzzy structure:

[0138]

[0139] Among them, x is the fuzzy structure input, is the fuzzy structure output, is the fuzzy weight, is the fuzzy basis function, and

[0140]

[0141] Among them, and are respectively and B j member functions.

[0142] Next, set the controller control objectives as follows:

[0143]

[0144] And define the tracking error as

[0145]

[0146] Construct the auxiliary variable

[0147]

[0148] According to (25), there exists an optimal weight vector ω * satisfying:

[0149]

[0150] Among them, ω * is composed of ω ij and satisfies is the estimated value of ω.

[0151] Then, define the approximation error vector ε as follows:

[0152]

[0153] Among them, And the estimation error is expressed as

[0154]

[0155] Next, according to (30)-(32), we can get

[0156]

[0157] Rearrange (18) into the following form:

[0158]

[0159] Among them, the fuzzy basis function vector satisfies

[0160] To handle the approximation error in (31), a weight estimate value is designed with a time-varying gain γ ik (t) update law is as follows:

[0161]

[0162] where j = 1, 2,..., n, i = 1, 2,..., r, is the upper bound of the fuzzy basis function vector, k ωi > 0 is a set parameter, Ω ij > 0 forms the matrix Ω, and the time-varying gain γ ik (t) forms the matrix Γ(t), and its online update law is as follows:

[0163]

[0164] where γ0 > 0, Λ is composed of the positive parameter μ ik constitutes.

[0165] Then, according to the update law (35), the estimation error can be written as

[0166] Furthermore, the adaptive fuzzy controller can be designed in the following form:

[0167]

[0168] where each parameter satisfies the following set conditions:

[0169]

[0170] Based on the controller designed in (37) and the update law designed in (35), the system (34) is asymptotically stable, and there is:

[0171]

[0172] The present invention adopts a method for verifying the cross-domain movement of an amphibious land-air robot, which is applicable to the performance verification of the method proposed in the present invention. The specific technical solution is as follows:

[0173] Based on the MATLAB platform, a multi-obstacle movement scenario for the amphibious land-air robot is built, and the positions, sizes, quantities, etc. of the obstacles are set. According to equations (1)-(7), the cross-domain path planning algorithm code is written using MATLAB Function.

[0174] Based on the MATLAB platform, the wing folding angle smooth switching trajectory planning algorithm code is written based on (20), and constraint conditions such as the descent height and speed are set.

[0175] Based on the ROS environment and Python language, write the code for smooth switching of the wing folding angle, and verify the designed smooth switching trajectory algorithm on the amphibious robot platform.

[0176] Based on the ROS environment and Python language, write the code for the design of an adaptive fuzzy controller, track and control the designed cross-domain path, and verify it on the amphibious robot platform.

[0177] To verify the effectiveness of the method designed in the present invention, the verification can be carried out according to the above steps.

[0178] Verification 1: Cross-domain path planning: As shown in the appendix Figure 2 As shown, the method proposed in the present invention realizes the energy-optimal land-air cross-domain path planning. When encountering passable obstacles on the ground, ground movement is preferentially considered to reduce energy consumption; when the obstacles cannot be passed through, air flight movement is considered to improve the adaptability of the robot in different environments. At the same time, as shown in the appendix Figure 3 As shown, the proposed method realizes the rapid convergence of the cost function in the iterative solution process, avoiding the problem of falling into local optima.

[0179] Verification 2: Smooth switching trajectory planning: As shown in the appendix Figure 4 As shown, under the action of the planning method proposed in the present invention, a smooth switching trajectory of the wing folding angle is obtained. To verify the effectiveness of the planned trajectory, the upper bound of the descent height is set to 0.1 m, and the air mode switching verification is carried out on the amphibious robot platform. As shown in the appendix Figure 5 As shown, the robot executes mode switching at a position of 1 m height according to the planned wing folding angle trajectory, and the height of the robot always remains above 0.9 m during the switching process. This shows that the proposed smooth switching trajectory planning method of the wing folding angle has a good effect during the air mode switching process.

[0180] Verification 3: Adaptive fuzzy control tracking: To verify the effectiveness of the control method proposed in the present invention, set up the scenario according to the appendix Figure 1 and carry out path tracking verification on the amphibious robot platform. As shown in the appendix Figure 6 As shown, the robot tracks the planned path, and the tracking error converges to 0 within 15 s. This verifies the effectiveness of the control method proposed in the present invention.

[0181] In summary, the cross-domain motion planning method and control method of the amphibious robot proposed in the present invention have practical functions and good application effects.

[0182] Embodiment 2:

[0183] Embodiment 2 of the present invention provides a cross-domain motion planning system for an amphibious robot, including:

[0184] A constraint determination module, configured to determine the working space of the amphibious aerial-ground robot according to task requirements, and obtain the operation constraints of the amphibious aerial-ground robot within the working space;

[0185] A path optimization module, configured to use the particle swarm optimization algorithm to perform path planning according to the operation constraints to obtain an optimal aerial-ground cross-domain path;

[0186] A wing switching module, configured to construct a dynamic model according to the dynamic parameters of the amphibious aerial-ground robot, and use the Gaussian pseudospectral method to construct a smooth switching trajectory of the wing folding angle with the goal of minimizing the flight altitude drop during the mode switching process;

[0187] A motion control module, configured to set a non-linear controller according to the dynamic model and the fuzzy algorithm, and use the non-linear controller to control the amphibious aerial-ground robot according to the optimal aerial-ground cross-domain path. Additionally, perform smooth switching during the flight according to the smooth switching trajectory of the wing folding angle.

[0188] Embodiment III:

[0189] Embodiment III of the present invention provides a medium, on which a program is stored, and when the program is executed by a processor, it implements the steps in the cross-domain motion planning method of the amphibious aerial-ground robot as described in Embodiment I of the present invention.

[0190] Embodiment IV:

[0191] Embodiment IV of the present invention provides a device, including a memory, a processor, and a program stored on the memory and executable on the processor. When the processor executes the program, it implements the steps in the cross-domain motion planning method of the amphibious aerial-ground robot as described in Embodiment I of the present invention.

[0192] The steps involved in Embodiments II, III, and IV above correspond to those in Method Embodiment I, and the specific implementation manners can refer to the relevant description part of Embodiment I.

[0193] Those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computer device. Optionally, they can be implemented by program codes executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module to implement. The present invention is not limited to any specific combination of hardware and software.

[0194] Although the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, it is not a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications or deformations that can be made without creative efforts on the basis of the technical solution of the present invention are still within the protection scope of the present invention.

Claims

1. A cross-domain motion planning method for an amphibious land-air robot, characterized in that, It includes the following steps: Determine the working space of the amphibious robot according to the task requirements, and obtain the operation constraints of the amphibious robot in the working space; Use the particle swarm optimization algorithm to perform path planning according to the operation constraints to obtain the optimal land-air cross-domain path; Construct a dynamic model based on the dynamic parameters of the amphibious robot, and use the Gaussian pseudospectral method to construct a smooth switching trajectory of the wing folding angle with the goal of minimizing the flight altitude drop during the mode switching process; Set a nonlinear controller according to the dynamic model and the fuzzy algorithm, and use the nonlinear controller to control the amphibious robot according to the optimal land-air cross-domain path. In addition, perform smooth switching during the flight according to the smooth switching trajectory of the wing folding angle.

2. The cross-domain motion planning method for the land-air amphibious robot according to claim 1, wherein The operation constraints include energy consumption cost constraint, collision cost constraint, and distance cost constraint. Among them, the energy consumption cost constraint is set according to the selection and switching of the motion mode, the collision cost constraint is set according to the position relationship between the amphibious robot and the obstacles in the working space, and the distance cost constraint is determined according to the segmented path between the amphibious robot and the destination.

3. The cross-domain motion planning method of the land-air amphibious robot according to claim 1, characterized in that, The specific steps of using the particle swarm optimization algorithm to perform path planning according to the operation constraints are: Define the optimization objective function of path planning; Set the velocity and position update rules of each particle; Design a new inertia weight considering the problem of being prone to falling into local optimum; Iteratively solve the optimization objective function based on the update rule and the new inertia weight to obtain the optimal land-air cross-domain path.

4. The cross-domain motion planning method of the land-air amphibious robot according to claim 1, characterized in that, The dynamic model is the dynamic model of the amphibious robot in the air flight mode when the wing folding angle is zero.

5. The cross-domain motion planning method of the land-air amphibious robot according to claim 1, characterized in that, The specific steps of using the Gaussian pseudospectral method to construct a smooth switching trajectory of the wing folding angle with the goal of minimizing the flight altitude drop during the mode switching process are: Transform the dynamic model with the goal of minimizing the flight altitude drop during the mode switching process; Set the dynamic constraints; Construct a nonlinear programming function with algebraic constraints based on the transformed dynamic model and dynamic constraints; Solve the nonlinear programming function to obtain the optimal state sequence of the descent altitude.

6. The cross-domain motion planning method for the land-air amphibious robot according to claim 5, characterized in that, The dynamic constraints include the folding angle constraint and the flight descent constraint of the amphibious robot.

7. The cross-domain motion planning method of the land-air amphibious robot according to claim 1, characterized in that, The specific steps of setting a nonlinear controller according to the dynamic model and the fuzzy algorithm are: Construct the position control system function of the amphibious robot according to the dynamic model; Introduce the fuzzy algorithm according to the nonlinear interference term in the position control system function to construct an adaptive fuzzy controller.

8. A cross-domain motion planning system for an amphibious land-air robot, characterized in that, It includes: A constraint determination module configured to determine the working space of the amphibious robot according to the task requirements and obtain the operation constraints of the amphibious robot in the working space; A path optimization module configured to use the particle swarm optimization algorithm to perform path planning according to the operation constraints to obtain the optimal land-air cross-domain path; A wing switching module configured to construct a dynamic model based on the dynamic parameters of the amphibious robot and use the Gaussian pseudospectral method to construct a smooth switching trajectory of the wing folding angle with the goal of minimizing the flight altitude drop during the mode switching process; The motion control module is configured to set a non-linear controller according to a dynamics model and a fuzzy algorithm, and use the non-linear controller to control the amphibious aerial-ground robot along an optimal aerial-ground cross-domain path. Additionally, smooth switching is performed during flight according to the smooth switching trajectory of the wing folding angle.

9. A computer-readable storage medium, characterized in that, It stores multiple instructions, and the instructions are adapted to be loaded and executed by a processor of a terminal device for the cross-domain motion planning method of the amphibious aerial-ground robot according to any one of claims 1-7.

10. A terminal device, characterized in that, It includes a processor and a computer-readable storage medium. The processor is used to implement each instruction; the computer-readable storage medium is used to store multiple instructions, and the instructions are adapted to be loaded and executed by the processor for the cross-domain motion planning method of the amphibious aerial-ground robot according to any one of claims 1-7.

Images