Energy consumption optimization path smoothing generation method and system for underwater glider robot with asymmetric turning radius

Abstract

The application discloses a kind of underwater glider robot energy consumption optimization path smooth generation method and system of asymmetric turning radius, first, establish considering oil bag excursion roll model, by analyzing oil bag and movable mass momentum influence, construct system kinematics equation and dynamics equation containing oil bag excursion;Based on the model, the forgetting factor recursive least squares algorithm is used to identify the oil bag gravity center excursion amount on line;Real-time left and right turning radius of robot is calculated using the identification result;Finally, based on asymmetric turning radius, generate suboptimal Dubins-Helix path, the path is composed of pitch angle smooth section and Dubins-Helix trajectory section, by optimizing pitch angle and helix section parameters, while meeting the motion constraint, realize energy consumption optimization.Modeling and parameter identification accurately, solve the model mismatch problem, the path generated is more in line with the actual maneuverability of robot, while ensuring the feasibility of the path, significantly reduce the system energy consumption.

Description

Technical Field

[0001] This invention belongs to the field of underwater robot path planning technology, and in particular relates to a method and system for energy consumption optimization and smooth path generation for underwater gliding robots with asymmetric turning radius. Background Technology

[0002] Path planning is the process of providing feasible and optimal (or suboptimal) motion paths for autonomous systems. It is an effective way to achieve requirements such as obstacle avoidance, coverage, multi-target tracking, and energy saving. For vehicles with full degrees of freedom (such as drones, remotely operated underwater vehicles, autonomous underwater vehicles, and Mecanum wheel robots), path planning only needs to consider paths that meet the requirements. However, for Dubins vehicles (such as fixed-wing aircraft, unmanned surface vessels, and underwater gliders), path planning not only needs to find feasible paths connecting the starting point and the ending point, but also needs to consider path smoothness to meet motion constraints.

[0003] In 1957, Lebensborn mathematically proved that the two-dimensional shortest path with initial pose, final pose, and minimum radius of curvature constraints is a Dubins curve. Since then, Dubins curves have been widely used in two-dimensional path planning. In recent years, related research has continued to deepen, such as discussing the reachability of the required length of an infinite number of Dubins curves between any two points, proposing a multi-objective coverage path planning reorganization framework, and using the Dubins algorithm to smooth trajectories.

[0004] The 3D path smoothing problem requires consideration of more constraints. Existing research has proposed various methods, such as parameterizing paths using k-order non-uniform rational B-spline curves, designing cooperative trajectory planning methods for biomimetic underwater vehicles, and generating 3D trajectories for multi-rotor UAVs using Bézier curves. However, these methods are computationally expensive, making them difficult to meet the needs of embedded real-time applications, and may get trapped in local minima. The Dubins-Helix method is a real-time computational 3D path smoothing method, with the RDDH method widely used in optimization and real-time dynamic planning tasks. However, for vehicles with asymmetric left / right turning radii, the RDDH method may lose its advantage. Furthermore, most energy-efficient path planning methods do not consider the impact of specific motion processes on energy consumption, while optimization targeting specific motion processes is crucial for the application of these technologies. Summary of the Invention

[0005] To address the above technical problems, this invention provides a method for optimizing the energy consumption of an underwater glider with asymmetrical left / right turning radii due to oil bladder offset, resulting in a smooth 3D path generation method. This method solves the problem that the existing RDDH method is ineffective when applied to underwater gliders with asymmetrical left / right turning radii.

[0006] The technical solution adopted by this invention to solve its technical problem is:

[0007] A method for energy-optimized path smoothing generation of an underwater glider with asymmetric turning radius, the method comprising the following steps:

[0008] S100: Establish a side tilt model of an underwater glider considering oil bladder offset. By analyzing the position vectors of the oil bladder mass and movable mass in the body coordinate system and their influence on the system momentum, construct the kinematic equations of the system, and derive the dynamic equations including the influence of oil bladder offset as well as the steady-state dynamic equations of the horizontal and vertical planes.

[0009] S200: Based on the roll model, a discrete state equation for roll angle acceleration is constructed. The forgetting factor recursive least squares algorithm is adopted, and the oil bladder center of gravity offset is identified online through multi-step differentiation and iterative update.

[0010] S300: Calculates the left and right turning radii by using the identified center of gravity offset and combining it with the real-time motion parameters of the underwater glider.

[0011] S400: Based on the calculated left and right turning radii, an algorithm is designed to generate a suboptimal Dubins-Helix path. The path consists of a pitch angle smoothing segment and a Dubins-Helix trajectory segment. The projection on the horizontal plane satisfies the radius of curvature constraint, and the pitch angle constraint is satisfied in three-dimensional space. By optimizing the pitch angle and helix segment parameters, an energy-optimized path is generated with path length, time consumption, and energy consumption as adversarial indicators. The energy consumption optimization considers the energy consumption components of the buoyancy drive system, attitude adjustment system, communication control and measurement system, and heading control system.

[0012] Preferably, in S100, the position vectors of the oil bladder mass and the movable mass in the body coordinate system and their motion are analyzed to influence the system momentum, and the dynamic equations are derived, including:

[0013] Define the mass momentum of the oil bladder and movable mass momentum Specifically:

[0014] ;

[0015] The forces generated by the mass of the oil bladder and its movable mass and They are represented as follows:

[0016] ;

[0017] ;

[0018] The total internal mass forces and moments are expressed as follows:

[0019] ;

[0020] in, For oil bladder quality, For movable mass; The linear velocity in the body coordinate system; Angular velocity in body coordinate system; , These are the position vectors of the oil bladder mass and the movable mass in the volume coordinate system, respectively. , These are the relative velocities of the oil bladder mass and the movable mass in the volume coordinate system, respectively. These are the linear velocity components along the x, y, and z axes in the body coordinate system, respectively. These are the angular velocity components about the x, y, and z axes in the body coordinate system, respectively. These are the coordinates of the oil bladder mass on the x, y, and z axes of the volume coordinate system; These are the coordinates of the movable mass on the x, y, and z axes of the volume coordinate system; The coordinates of the center of gravity on the z-axis of the volume coordinate system; Let be the velocity of the oil bladder mass moving along the y-axis of the body coordinate system; Let be the velocity of the movable mass along the x-axis of the body coordinate system. Let x be the linear acceleration along the x-axis of the body coordinate system; Angular acceleration about the z-axis of the body coordinate system; and The y-axis and z-axis components of the mass momentum of the oil bladder are respectively defined; The time derivative of the y-axis component of the oil bladder's mass momentum; The force vector generated by the mass of the oil bladder; The force vector generated by the movable mass; For the mass momentum of the oil bladder The time derivative; Momentum of movable mass The time derivative; For the total internal mass force; The torque generated by the mass of the oil bladder; The torque generated by the movable mass; This represents the total internal mass torque; It is the acceleration due to gravity; It is the direction of gravity relative to the body coordinate system. It is the unit vector pointing in the direction of gravity in the Earth coordinate system;

[0021] The total internal mass forces and torques are used as inputs to the rigid body dynamics equations of the underwater glider, in order to establish a complete dynamic model that includes the effects of the oil bladder and movable mass.

[0022] Preferably, the kinematic equations of the underwater glider tilt model considering oil bladder offset in S100 are used to describe the transformation relationship between the body coordinate system and the ground coordinate system, specifically including:

[0023] The kinematic relationship between roll angle and yaw angle is as follows:

[0024] ;

[0025] ;

[0026] in, The angular velocity is the roll angle. Let be the angular velocity of rotation about the x-axis of the body coordinate system. Let be the angular velocity of rotation about the z-axis of the body coordinate system. Angular velocity is the yaw angle. This refers to the roll angle;

[0027] Rotation matrix from volume coordinate system to ground coordinate system Defined as:

[0028] ;

[0029] in, , , , , These are roll angle, yaw angle, and pitch angle, respectively.

[0030] The kinematic model also includes differential equations for the rotation matrix and position, which are used to describe the evolution of the robot's posture and calculate the robot's position change in the ground coordinate system, respectively.

[0031] ;

[0032] ;

[0033] in, The time derivative of the rotation matrix. Let be the angular velocity about the x-axis of the body coordinate system. Let be the angular velocity about the y-axis of the body coordinate system. Let be the angular velocity about the z-axis of the body coordinate system. Let be the antisymmetric matrix of angular velocity. This is the time derivative of the position vector in the Earth coordinate system. Let be the linear velocity vector represented in the volume coordinate system.

[0034] Preferably, the steady-state dynamic equations in S100 include horizontal plane equations and vertical plane equations, with the horizontal plane equations being:

[0035] ;

[0036] The equation of the vertical plane is:

[0037] ;

[0038] in, This refers to the total mass of the underwater gliding robot. For drainage quality; Add mass to the water body For oil bladder quality, For movable mass; , , Let x represent the coordinates of the water body, oil sac, and movable mass on the x-axis of the volume coordinate system. , , , , For additional quality coefficients; , Let x be the moment of inertia about the x-axis and z-axis; It is net buoyancy; It is the resultant velocity on the vertical plane of the underwater gliding robot in a body coordinate system; It is excluding movable mass The sum of all masses except those mentioned above; and They represent Coordinates along the z-axis and x-axis; and They represent Coordinates along the z-axis and x-axis; X(), Y(), K(), N() are hydrodynamic terms; and It is the hydrodynamic coefficient.

[0039] Preferably, S200 includes:

[0040] S210: Construct system equations for parameter identification, expressing the relationship between the system's input, output, and identified parameters as follows:

[0041] ;

[0042] in , , and These represent the time steps. The system output, input, parameters to be identified, and deviation at that time. It is typically considered to be white noise with a mean of 0;

[0043] S220: Based on the roll model, a discrete state equation for roll angle acceleration is established, specifically:

[0044] ;

[0045] in It is the offset of the oil bladder's center of gravity to be identified. and For time step linear velocity, and For time steps angular velocity, to These are the model coefficients. For time steps rudder angle;

[0046] By matching the discrete state equations with the system equations of FFRLS, the following is determined:

[0047] ;

[0048] ;

[0049] ;

[0050] Roll acceleration is calculated using a multi-step differential method:

[0051] ;

[0052] in It is the sampling time. It is the time step interval oriented towards differentiation;

[0053] S230: In each sampling period Within, perform the following calculations in sequence:

[0054] Measured or calculated and ;

[0055] Calculate the deviation Specifically:

[0056] ;

[0057] Calculate the correction gain as follows:

[0058] ;

[0059] in It is a time step Algorithm gain at that point It is a time step The noise covariance matrix at that location, It is a forgetting factor;

[0060] Update the parameter estimates at the end of each period. Specifically:

[0061] ;

[0062] Update the noise covariance matrix Specifically:

[0063] .

[0064] Preferably, the formulas for calculating the left-turn radius and right-turn radius in S300 are as follows:

[0065] ;

[0066] And satisfy the following constraints:

[0067] ;

[0068] ;

[0069] ;

[0070] in, For the three-dimensional radius of curvature, Turn left; Turn right. and These are the minimum and maximum roll angles allowed for underwater gliders, respectively. and These are the minimum and maximum allowable deflection angles of the rudder, respectively.

[0071] Preferably, the suboptimal Dubins-Helix path in S400 consists of five sub-path segments, including a pitch angle smoothing segment and a Dubins-Helix trajectory segment; wherein, the generation process of the three-dimensional spiral curve in the Dubins-Helix trajectory segment includes:

[0072] Assuming the helical curve segment maintains a constant pitch angle during flight. The center of the arc projected onto the horizontal plane by the spiral curve is Then any point on the spiral curve The coordinates are represented as:

[0073] ;

[0074] in, The starting point of this spiral curve segment The height coordinates, yes The normalized unit vector, It is the starting point on the horizontal projection plane. arrive The rotation angle, when the spiral is left-handed. When the spiral is right-handed, ; The radius of curvature of the horizontal projection of the helical curve;

[0075] The value of the turning radius of the robot's three-dimensional motion and pitch angle The following relationship must be satisfied:

[0076] ;

[0077] in, It means turn left. This means turn right;

[0078] Horizontal projected radius of curvature The minimum turning radius constraint based on the robot's maneuverability must be satisfied:

[0079] (1) Curvature radius constraint of the horizontal projection of the left-handed helical curve: ;

[0080] (2) Curvature radius constraint of the horizontal projection of the right-hand helical curve: ;

[0081] in, , These are the minimum radii of curvature for left and right turns, respectively.

[0082] Preferably, in S400, a suboptimal Dubins-Helix path is generated, first based on the starting point and the ending point. With the two-dimensional Dubins path length and maximum pitch angle The relationship is determined, including the two-dimensional Dubins path length. , and These are the rotation angles corresponding to the arcs. It is the length of a straight line; relationship and path generation fall into the following two categories:

[0083] Case 1: When the conditions are met At this time, the path must contain a spiral segment, and its parameters must satisfy the basic height relationship: Its optimization is divided into two modes: no attitude switching and attitude switching.

[0084] (1) For the shortest path without attitude switching, its total length The pitch angles for each segment are determined by the following formula:

[0085] ;

[0086] ;

[0087] in and The optimal value is determined by the following optimization problem:

[0088] ;

[0089] The total length of the path can also be expressed as: ;

[0090] (2) For the shortest path with attitude switching, its total length The pitch angles and attitude switching point heights for each segment are determined by the following formula:

[0091] ;

[0092] ;

[0093] ;

[0094] in and The optimal value is determined by the following optimization problem:

[0095] ;

[0096] The total length of the path can also be expressed as: ;

[0097] in, This is the length of the attitude-free switching path. The length of the attitude switching path; , The projections of the left and right turning radii onto the two-dimensional plane; This represents the length of the Dubins path. As the starting height, The final height; , , For each segment of pitch angle; This is the absolute value of the maximum pitch angle; The height difference between the attitude switching points; , , , It is a natural number;

[0098] Case 2: When the conditions are met At this time, the path tilts the two-dimensional Dubins path by a single pitch angle to adapt to the height difference, forming the shortest path without spiral segments, and its total length is... With uniform pitch angle Determined by the following formula:

[0099] ;

[0100] .

[0101] Preferably, energy consumption optimization in S500 is achieved by calculating the total energy consumption E3D, specifically as follows:

[0102] ;

[0103] ;

[0104] ;

[0105] ;

[0106] ;

[0107] in, For fluid density; The power of the hydraulic pump with respect to depth; The flow rate of the hydraulic pump with respect to depth; The power of the solenoid valve; The flow rate of the solenoid valve; For movable mass stroke; The moving speed of movable mass; Input power to the motor; Power for communication, control, and measurement systems; The energy consumption of the servo motor when turning with the minimum turning radius.

[0108] An energy-optimized path smoothing generation system for underwater gliders with asymmetric turning radius, comprising:

[0109] The tilt model building module is used to build a tilt model of an underwater glider considering the oil bladder offset. By analyzing the position vectors of the oil bladder mass and movable mass in the body coordinate system and their motion on the system momentum, the kinematic equations of the system are constructed, and the dynamic equations including the influence of oil bladder offset, as well as the steady-state dynamic equations in the horizontal and vertical planes, are derived.

[0110] The oil bladder center of gravity offset identification module is used to construct the discrete state equation of roll angle acceleration based on the roll model. It adopts the forgetting factor recursive least squares algorithm and identifies the oil bladder center of gravity offset online through multi-step differentiation and iterative update.

[0111] Left and right turn radius determination module: Calculates the left and right turn radii by using the identified center of gravity offset and the real-time motion parameters of the underwater glider.

[0112] The path generation module is used to design an algorithm to generate a suboptimal Dubins-Helix path based on the calculated left and right turning radii. The path consists of a pitch angle smoothing segment and a Dubins-Helix trajectory segment. The projection on the horizontal plane satisfies the radius of curvature constraint, and the pitch angle constraint is satisfied in three-dimensional space. By optimizing the pitch angle and helix segment parameters, an energy-optimized path is generated with path length, time consumption, and energy consumption as adversarial indicators. The energy consumption optimization takes into account the energy consumption components of the buoyancy drive system, attitude adjustment system, communication control and measurement system, and heading control system.

[0113] The aforementioned method and system for smoothing energy-efficient paths for underwater gliders with asymmetric turning radii significantly improves the accuracy and practicality of path planning by establishing a dynamic model that accurately considers fuel bladder offset. Traditional methods often rely on the assumption of symmetrical turning radii, while this invention derives a novel dynamic equation incorporating internal mass forces by analyzing the momentum effects of the fuel bladder and movable mass, enabling the model to realistically reflect the robot's roll-yaw coupling characteristics. Furthermore, an innovative forgetting factor recursive least squares algorithm is employed to identify the center of gravity offset in real time, resolving model mismatch issues caused by fuel bladder installation errors and buoyancy variations, thus enhancing the system's adaptability under different operating conditions. Based on the identification results, this invention achieves accurate calculation of asymmetric left / right turning radii, overcoming the idealized limitation of symmetrical turning radii in traditional path planning. This improvement makes the generated path more consistent with the robot's actual maneuverability, significantly improving path feasibility and tracking accuracy. Finally, by integrating the accurate dynamic model and asymmetric maneuverability into 3D path generation, a suboptimal Dubins-Helix path with optimized energy consumption is designed. This path not only satisfies motion constraints but also comprehensively considers the energy consumption characteristics of systems such as buoyancy drive and attitude adjustment, achieving synergistic optimization of path length and energy consumption. This method, while ensuring path smoothness, reduces overall energy consumption and effectively extends the operational endurance of underwater gliders, providing reliable technical support for their long-term ocean observation missions. Attached Figure Description

[0114] Figure 1 This is a flowchart of a method for energy consumption optimization and path smoothing generation for an underwater gliding robot with asymmetric turning radius, according to an embodiment of the present invention.

[0115] Figure 2 This is a diagram illustrating a blended wing-body underwater glider and its coordinate system definition in one embodiment of the present invention.

[0116] Figure 3 This is a graph showing the changes in the values ​​of different terms during the turning process of an underwater gliding robot in one embodiment of the present invention;

[0117] Figure 4 This is an ATRDH path description diagram in one embodiment of the present invention;

[0118] Figure 5 This is an asymmetric left-handed / right-handed Dubins path diagram in one embodiment of the present invention; wherein, (a) is an external tangential path, (b) is an internal tangential path, (c) is an LRL path, and (d) is an RLR path, where "L" represents left-handed and "R" represents right-handed;

[0119] Figure 6 This is an unfolded ATRDH path plan diagram according to one embodiment of the present invention;

[0120] Figure 7 This is a schematic diagram of hardware-in-the-loop simulation results in one embodiment of the present invention; wherein, (a) is the center of gravity offset identification result, (b) is the visualization result of the ATRDH path, RDDH path and optimal path generated based on the first pair of poses, (c) is the projection diagram of the ATRDH path and RDDH path on the xy plane, and (d) is the relationship diagram of the heading angle and path length of the two paths. Detailed Implementation

[0121] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0122] This embodiment takes a certain type of blended wing-body underwater glider as the research object and adopts the three-dimensional path smoothing generation optimization method of the present invention, which considers fuel bladder offset and energy consumption and has asymmetric left / right turning radius.

[0123] In one embodiment, such as Figure 1 As shown, a method for energy consumption optimization and path smoothing generation of an underwater gliding robot with asymmetric turning radius is disclosed. The method includes the following steps:

[0124] S100: Establish a side tilt model of an underwater glider considering oil bladder offset. By analyzing the position vectors of the oil bladder mass and movable mass in the body coordinate system and their influence on the system momentum, construct the kinematic equations of the system, and derive the dynamic equations including the influence of oil bladder offset as well as the steady-state dynamic equations of the horizontal and vertical planes.

[0125] S200: Based on the roll model, a discrete state equation for roll angle acceleration is constructed. The forgetting factor recursive least squares algorithm is adopted, and the oil bladder center of gravity offset is identified online through multi-step differentiation and iterative update.

[0126] S300: Calculates the left and right turning radii by using the identified center of gravity offset and combining it with the real-time motion parameters of the underwater glider.

[0127] S400: Based on the calculated left and right turning radii, an algorithm is designed to generate a suboptimal Dubins-Helix path. The path consists of a pitch angle smoothing segment and a Dubins-Helix trajectory segment. The projection on the horizontal plane satisfies the radius of curvature constraint, and the pitch angle constraint is satisfied in three-dimensional space. By optimizing the pitch angle and helix segment parameters, an energy-optimized path is generated with path length, time consumption, and energy consumption as adversarial indicators. The energy consumption optimization considers the energy consumption components of the buoyancy drive system, attitude adjustment system, communication control and measurement system, and heading control system.

[0128] Specifically, this invention significantly improves the path planning performance of underwater gliding robots by establishing an accurate oil bladder offset dynamic model. An adaptive identification method is used to obtain the center of gravity offset in real time, effectively solving the model mismatch problem and enhancing system robustness. For the first time, accurate calculation of asymmetric left and right turning radii is achieved, making the path more consistent with actual maneuverability. Based on this, an energy-optimized 3D path generation method is developed, comprehensively considering the energy consumption characteristics of the system, such as buoyancy drive and attitude adjustment, effectively reducing overall energy consumption while ensuring path feasibility and extending the robot's underwater operation time.

[0129] Furthermore, the research object of this invention is a blended wing-body underwater glider that uses differential rudder control for roll-yaw coupling, such as... Figure 2 As shown. Figure 2 The definitions of the body coordinate system and the ground coordinate system are shown. The origin of the body coordinate system is located at the center of buoyancy of the underwater glider. The linear velocity and angular velocity in the body coordinate system are expressed as follows: and The position and Euler angles in the Earth coordinate system are expressed as follows: and .

[0130] In one embodiment, the kinematic equations of the underwater glider tilt model considering oil bladder offset in S100 are used to describe the transformation relationship between the body coordinate system and the ground coordinate system, specifically including:

[0131] The kinematic relationship between roll angle and yaw angle is as follows:

[0132] ;

[0133] ;

[0134] in, The angular velocity is the roll angle. Let be the angular velocity of rotation about the x-axis of the body coordinate system. Let be the angular velocity of rotation about the z-axis of the body coordinate system. Angular velocity is the yaw angle. This refers to the roll angle;

[0135] Rotation matrix from volume coordinate system to ground coordinate system Defined as:

[0136] ;

[0137] in, , , , , These are roll angle, yaw angle, and pitch angle, respectively.

[0138] The kinematic model also includes differential equations for the rotation matrix and position, which are used to describe the evolution of the robot's posture and calculate the robot's position change in the ground coordinate system, respectively.

[0139] ;

[0140] ;

[0141] in, The time derivative of the rotation matrix. Let be the angular velocity about the x-axis of the body coordinate system. Let be the angular velocity about the y-axis of the body coordinate system. Let be the angular velocity about the z-axis of the body coordinate system. Let be the antisymmetric matrix of angular velocity. This is the time derivative of the position vector in the Earth coordinate system. Let be the linear velocity vector represented in the volume coordinate system.

[0142] Unlike AUV (Autonomous Underwater Vehicle) models, the movable mass and fuel bladder mass of underwater gliders account for a large proportion of their total mass. Therefore, their linear momentum has a significant impact on the dynamics of underwater gliders.

[0143] In one embodiment, S100 analyzes the position vectors of the oil bladder mass and the movable mass in the body coordinate system and the influence of their motion on the system momentum, deriving the dynamic equations, including:

[0144] Considering the motion on the horizontal plane and the very slow speed of the oil bladder mass and movable mass, i.e., assuming... , , , , and Define the mass momentum of the oil bladder. and movable mass momentum Specifically:

[0145] ;

[0146] The forces generated by the mass of the oil bladder and its movable mass and They are represented as follows:

[0147] ;

[0148] ;

[0149] The total internal mass forces and moments are expressed as follows:

[0150] ;

[0151] in, For oil bladder quality, For movable mass; The linear velocity in the body coordinate system; Angular velocity in body coordinate system; , These are the position vectors of the oil bladder mass and the movable mass in the volume coordinate system, respectively. , These are the relative velocities of the oil bladder mass and the movable mass in the volume coordinate system, respectively. These are the linear velocity components along the x, y, and z axes in the body coordinate system, respectively. These are the angular velocity components about the x, y, and z axes in the body coordinate system, respectively. These are the coordinates of the oil bladder mass on the x, y, and z axes of the volume coordinate system; These are the coordinates of the movable mass on the x, y, and z axes of the volume coordinate system; The coordinates of the center of gravity on the z-axis of the volume coordinate system; Let be the velocity of the oil bladder mass moving along the y-axis of the body coordinate system; Let be the velocity of the movable mass along the x-axis of the body coordinate system. Let x be the linear acceleration along the x-axis of the body coordinate system; Angular acceleration about the z-axis of the body coordinate system; and The y-axis and z-axis components of the mass momentum of the oil bladder are respectively defined; The time derivative of the y-axis component of the oil bladder's mass momentum; The force vector generated by the mass of the oil bladder; The force vector generated by the movable mass; For the mass momentum of the oil bladder The time derivative; Momentum of movable mass The time derivative; For the total internal mass force; The torque generated by the mass of the oil bladder; The torque generated by the movable mass; This represents the total internal mass torque; It is the acceleration due to gravity; It is the direction of gravity relative to the body coordinate system. It is a unit vector pointing in the direction of gravity in the Earth coordinate system; the total internal mass force and torque are used as input terms and introduced into the rigid body dynamics equations of the underwater glider to establish a complete dynamic model that includes the influence of the oil bladder and movable mass.

[0152] Specifically, in order to quantitatively assess The contributions of each item in the simulation, and the values ​​of each item collected from the simulation of attitude transition and steering operations of the underwater glider, are as follows: Figure 3 As shown in Table 1, their absolute proportions in the simulation are illustrated. Therefore, , and It can be ignored. It can also be handled in the same way.

[0153] Table 1:

[0154]

[0155] In one embodiment, the steady-state dynamic equations in S100 include horizontal plane equations and vertical plane equations. The horizontal plane equations are as follows:

[0156] ;

[0157] The equation of the vertical plane is:

[0158] ;

[0159] in, This refers to the total mass of the underwater gliding robot. For drainage quality; Add mass to the water body For oil bladder quality, For movable mass; , , Let x represent the coordinates of the water body, oil sac, and movable mass on the x-axis of the volume coordinate system. , , , , For additional quality coefficients; , Let x be the moment of inertia about the x-axis and z-axis; It is net buoyancy; It is the resultant velocity on the vertical plane of the underwater gliding robot in a body coordinate system; It is excluding movable mass The sum of all masses except those mentioned above; and They represent Coordinates along the z-axis and x-axis; and They represent Coordinates along the z-axis and x-axis; X(), Y(), K(), N() are hydrodynamic terms; and It is the hydrodynamic coefficient.

[0160] Furthermore, before deployment, the underwater glider undergoes ballast calibration in a shore-based pool, adjusting the counterweight mass to achieve static equilibrium while maintaining zero buoyancy. However, errors in bladder installation may lead to uneven mass distribution when the underwater glider is under positive / negative buoyancy, affecting its center of gravity position. Changes in the center of gravity position cause changes in the rotational inertia matrix, thus affecting steering efficiency. Therefore, the horizontal position of the center of gravity is crucial for roll-yaw coupled dynamics. The Forgetting Factor Recursive Least Squares (FFLS) algorithm is widely used in many scenarios requiring rapid and accurate estimation of parameters or states due to its efficiency, real-time performance, robustness, and strong adaptability. Therefore, in this invention, FFRLS is selected as an embedded algorithm to identify center of gravity shifts.

[0161] In one embodiment, S200 includes:

[0162] S210: Construct system equations for parameter identification, expressing the relationship between the system's input, output, and identified parameters as follows:

[0163] ;

[0164] in , , and These represent the time steps. The system output, input, parameters to be identified, and deviation at that time. It is typically considered to be white noise with a mean of 0;

[0165] S220: Based on the roll model, a discrete state equation for roll angle acceleration is established, specifically:

[0166] ;

[0167] in It is the offset of the oil bladder's center of gravity to be identified. and For time step linear velocity, and For time step angular velocity, to These are the model coefficients. For time step rudder angle;

[0168] By matching the discrete state equations with the system equations of FFRLS, the following is determined:

[0169] ;

[0170] ;

[0171] ;

[0172] Most of the state variables in the matching formula can be obtained from the IMU and the observer. However, by... Numerical differentiation yields This is inaccurate because this method introduces greater numerical noise and discrete errors, hence the introduction of the multi-step differentiation method;

[0173] Roll angle acceleration Calculated using the multi-step differential method:

[0174] ;

[0175] in It is the sampling time. It is the time step interval for differentiation; L cannot be set too large, otherwise it may cause delay and reduce the accuracy of parameter identification.

[0176] S230: In each sampling period Within, perform the following calculations in sequence:

[0177] Measured or calculated and ;

[0178] Calculate the deviation Specifically:

[0179] ;

[0180] Calculate the correction gain as follows:

[0181] ;

[0182] in It is a time step Algorithm gain at the location, It is a time step The noise covariance matrix at that location, It is a forgetting factor;

[0183] Update the parameter estimates at the end of each period. Specifically:

[0184] ;

[0185] Update the noise covariance matrix Specifically:

[0186] .

[0187] In one embodiment, based on the center of gravity offset identification results, the tilt model of the underwater glider is complete and solvable. The formulas for calculating the left turning radius and right turning radius in S300 are as follows:

[0188] ;

[0189] And satisfy the following constraints:

[0190] ;

[0191] ;

[0192] ;

[0193] in, For the three-dimensional radius of curvature, Turn left; Turn right. and These are the minimum and maximum roll angles allowed for underwater gliders, respectively. and These are the minimum and maximum allowable deflection angles of the rudder, respectively.

[0194] In one embodiment, the suboptimal Dubins-Helix path in S400 consists of five sub-path segments, including a pitch angle smoothing segment and a Dubins-Helix trajectory segment; such as Figure 4 As shown, where and It is a smooth segment of the pitch angle. , and This is the Dubins-Helix trajectory segment. Its projection on the horizontal plane and the z-plot of its three-dimensional trajectory are Dubins curves that satisfy the radius of curvature constraint. and It is a spiral curve that satisfies the pitch angle constraint and the radius of curvature constraint. , , , , and These are the initial pitch smoothing point and the starting point. arrive The point of tangency of the segment arrive The tangent point, endpoint, and final pitch smoothing point of the segment.

[0195] The generation process of the three-dimensional spiral curve in the Dubins-Helix trajectory segment includes:

[0196] Assuming the helical curve segment maintains a constant pitch angle during flight. The center of the arc projected onto the horizontal plane by the spiral curve is Then any point on the spiral curve The coordinates are represented as:

[0197] ;

[0198] in, The starting point of this spiral curve segment The height coordinates, yes The normalized unit vector, It is the starting point on the horizontal projection plane. arrive The rotation angle, when the spiral is left-handed. When the spiral is right-handed, ; The radius of curvature of the horizontal projection of the helical curve;

[0199] The value of the turning radius of the robot's three-dimensional motion and pitch angle The following relationship must be satisfied:

[0200] ;

[0201] in, This means turn left. It means turn right;

[0202] Horizontal projected radius of curvature The minimum turning radius constraint based on the robot's maneuverability must be satisfied:

[0203] (1) Curvature radius constraint of the horizontal projection of the left-handed helical curve: ;

[0204] (2) Curvature radius constraint of the horizontal projection of the right-hand helical curve: ;

[0205] in, , These are the minimum radii of curvature for left and right turns, respectively.

[0206] Specifically, the parameterized description of ATRDH is as follows: the Durbins curve connecting the start and end points has 4 CSC paths (e.g., Figure 5 (as shown in (a) and (b)) and 2 CCC paths (as shown in (a) and (b)). Figure 5 As shown in (c) and (d), where "C" represents a curve and "S" represents a straight line. In this invention, a path planning problem with the following two conditions is discussed: 1. 2. and The projection of the distance between them onto the horizontal plane is greater than It is worth noting that for the upward path or projection distance less than... The planning problem can also be solved by the proposed method.

[0207] There are two cases regarding the parameters and the length of the ATRDH path: 1. 2. The difference between the two cases is that the former requires a spiral path, while the latter does not.

[0208] In one embodiment, in S400, a suboptimal Dubins-Helix path is generated, first based on the start and end points. With the two-dimensional Dubins path length and maximum pitch angle The relationship is determined, including the two-dimensional Dubins path length. , and These are the rotation angles corresponding to the arcs. It is the length of a straight line; relationship and path generation fall into the following two categories:

[0209] Case 1: When the conditions are met At this time, the path must contain a spiral segment, and its parameters must satisfy the basic height relationship: Its optimization is divided into two modes: no attitude switching and attitude switching.

[0210] (1) For the shortest path without attitude switching, its total length The pitch angles for each segment are determined by the following formula:

[0211] ;

[0212] ;

[0213] in and The optimal value is determined by the following optimization problem:

[0214] ;

[0215] The total length of the path can also be expressed as: ;

[0216] (2) For the shortest path with attitude switching, its total length The pitch angles and attitude switching point heights for each segment are determined by the following formula:

[0217] ;

[0218] ;

[0219] ;

[0220] in and The optimal value is determined by the following optimization problem:

[0221] ;

[0222] The total length of the path can also be expressed as: ;

[0223] in, This is the length of the attitude-free switching path. The length of the attitude switching path; , The projections of the left and right turning radii onto the two-dimensional plane; This represents the length of the Dubins path. As the starting height, The final height; , , For each segment of pitch angle; This is the absolute value of the maximum pitch angle; The height difference between the attitude switching points; , , , For natural numbers; furthermore, the generation of the shortest ATRDH path (AS-ATRDH) with attitude switching is shown in Figure 6 The red dotted line in the middle;

[0224] Case 2: When the conditions are met At this time, the path tilts the two-dimensional Dubins path by a single pitch angle to adapt to the height difference, forming the shortest path without spiral segments, and its total length is... With uniform pitch angle Determined by the following formula:

[0225] ;

[0226] .

[0227] In one embodiment, energy consumption optimization in S500 is achieved by calculating the total energy consumption E3D, specifically as follows:

[0228] ;

[0229] ;

[0230] ;

[0231] ;

[0232] ;

[0233] in, Energy consumption for the buoyancy-driven system; Energy consumption for the attitude adjustment system; Energy consumption for communication, control and measurement systems; Energy consumption for the heading control system; For fluid density; The power of the hydraulic pump with respect to depth; The flow rate of the hydraulic pump with respect to depth; The power of the solenoid valve; The flow rate of the solenoid valve; For movable mass stroke; The moving speed of movable mass; Input power to the motor; Power for communication, control, and measurement systems; This represents the energy consumption of the servo motor when turning with the minimum turning radius. A pair of poses can generate a maximum of 8 ATRDH paths, and the path with the lowest energy consumption is selected as the energy-optimal path.

[0234] Next, verification was performed. The proposed algorithm was developed based on an STM32F429 microcontroller with an ARM Cortex-M4 32-bit RISC core, and simulations were run on a hardware-in-the-loop (HIL) simulation platform for underwater gliders. The proposed algorithm can generate trajectories in real time, with an average execution time of less than 0.1 ms. The proposed algorithm is of great significance for real-time path optimization. In this invention, , Set. and The method used in this study is the same as the Dubins curve design for smoothing pitch angles. Figure 7 (a) shows the error in identifying the center of gravity shift.

[0235] To evaluate the effectiveness and superiority of the ATRDH method, this invention provides a comparison of the metrics of three methods for path generation: the ATRDH method, the RDDH method, and the numerical method. The smooth generation results of eight ATRDH paths with initial pose (0,0,0,180°) and final pose (100,0,-276,180°) are shown in Table 2. Among them, the shortest and least time-consuming path is the AS-ATRDH (LSL) path. The path length mainly depends on the initial depth. final depth Dubins path length Left and right helical radius and The time consumed depends not only on the parameters mentioned above, but also on the net buoyancy. The least energy-consuming path is the NAS-ATRDH (LSR or RSL) path. The AS-ATRDH path, derived from the same Dubins path type, will always consume additional energy from the BD and AR systems compared to the NAS-ATRDH path, but may consume less energy from the CMC system.

[0236] Table 2:

[0237]

[0238] Table 2 shows the 3D trajectory smoothing generation results of the three methods under four pairs of pose requirements. The visualization results based on the first pair of poses are shown in Table 2. Figure 7 (b)-(d), the black segments at both ends of the path in the diagram represent the pitch angle smoothing segments. It is worth noting that the path with the lowest energy consumption is considered the optimal ATRDH path, and the baseline of the difference is based on a numerical method. When The optimal ATRDH path is the same as the RDDH path (as shown in Table 2, based on the results generated from the second pair of poses). Generally, the generated NAS-ATRDH path set includes RDDH paths (the NAS-ATRDH (RSR) type paths in Table 2 are RDDH paths). It can be seen that the optimal ATRDH path is closer to the optimal path generated by the numerical method. That is, the AS-ATRDH type path achieves the maximum descent velocity at the cost of a longer vertical displacement, while the NAS-ATRDH type path ensures no increase in vertical displacement at the cost of a reduced descent velocity.

[0239] The method for smoothing three-dimensional paths for underwater gliding robots provided by this invention forms a complete technical solution by establishing an accurate dynamic model, designing an adaptive parameter identification mechanism, and developing an energy-optimized path planning algorithm, and has the following significant advantages:

[0240] First, this invention establishes an accurate roll dynamics model that incorporates the influence of the fuel bladder and the momentum of the movable mass. Compared with traditional AUV models, this model fully considers the significant impact of large-scale internal mass movement on system dynamics. By deriving the steady-state dynamic equations that include fuel bladder displacement, it accurately describes the roll-yaw coupling characteristics of the underwater glider, providing a reliable physical basis for subsequent path planning.

[0241] Secondly, to address the model mismatch problem caused by oil bladder installation errors, an innovative forgetting factor recursive least squares algorithm is employed to achieve online real-time identification of the center of gravity offset. This method constructs a discrete state equation for roll acceleration and utilizes multi-step differentiation techniques to improve computational accuracy, enabling rapid tracking of system parameter changes and significantly enhancing the model's adaptability and robustness under different buoyancy states.

[0242] Third, the robot's asymmetric left / right turning radius is accurately calculated based on the identification results, breaking through the idealized assumption of symmetric maneuverability in traditional path planning. This makes the generated path more consistent with the robot's actual maneuverability, effectively solving the problem of path infeasibility caused by the mismatch between theoretical models and actual capabilities.

[0243] Finally, by integrating precise dynamic models and asymmetric maneuverability into 3D path generation, a suboptimal Dubins-Helix path planning method with optimized energy consumption was developed. This method not only satisfies the radius of curvature and pitch angle constraints but also comprehensively considers key energy-consuming aspects such as the buoyancy drive system and attitude adjustment system, achieving synergistic optimization of path length and energy consumption. Practical applications show that this method effectively reduces overall energy consumption while ensuring path smoothness and accessibility, significantly extending the operational endurance of underwater gliders and providing reliable technical support for long-term ocean observation missions.

[0244] In one embodiment, an energy-optimized path smoothing generation system for underwater gliding robots with asymmetric turning radius is also provided, comprising:

[0245] The tilt model building module is used to build a tilt model of an underwater glider considering the oil bladder offset. By analyzing the position vectors of the oil bladder mass and movable mass in the body coordinate system and their motion on the system momentum, the kinematic equations of the system are constructed, and the dynamic equations including the influence of oil bladder offset, as well as the steady-state dynamic equations in the horizontal and vertical planes, are derived.

[0246] The oil bladder center of gravity offset identification module is used to construct the discrete state equation of roll angle acceleration based on the roll model. It adopts the forgetting factor recursive least squares algorithm and identifies the oil bladder center of gravity offset online through multi-step differentiation and iterative update.

[0247] Left and right turn radius determination module: Calculates the left and right turn radii by using the identified center of gravity offset and the real-time motion parameters of the underwater glider.

[0248] The path generation module is used to design an algorithm to generate a suboptimal Dubins-Helix path based on the calculated left and right turning radii. The path consists of a pitch angle smoothing segment and a Dubins-Helix trajectory segment. The projection on the horizontal plane satisfies the radius of curvature constraint, and the pitch angle constraint is satisfied in three-dimensional space. By optimizing the pitch angle and helix segment parameters, an energy-optimized path is generated with path length, time consumption, and energy consumption as adversarial indicators. The energy consumption optimization takes into account the energy consumption components of the buoyancy drive system, attitude adjustment system, communication control and measurement system, and heading control system.

[0249] Specific limitations regarding the energy consumption optimization path smoothing generation system for underwater gliders with asymmetric turning radii can be found in the limitations of the energy consumption optimization path smoothing generation method for underwater gliders with asymmetric turning radii described above, and will not be repeated here. Each module in the aforementioned energy consumption optimization path smoothing generation system for underwater gliders with asymmetric turning radii can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0250] In one embodiment, a computer device is also provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of a method for generating a smooth path for energy consumption optimization of an underwater gliding robot with asymmetric turning radius.

[0251] A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of a method for generating a smooth path for energy consumption optimization of an underwater gliding robot with asymmetric turning radius are disclosed.

[0252] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0253] The above provides a detailed description of the energy consumption optimization path smoothing generation method and system for underwater gliding robots with asymmetric turning radius provided by this invention. Specific examples have been used to illustrate the principles and implementation methods of this invention, and the descriptions of the embodiments are merely for the purpose of helping to understand the core ideas of this invention. It should be noted that those skilled in the art can make various improvements and modifications to this invention without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this invention.

Claims

1. A method for energy consumption optimization and path smoothing generation of an underwater gliding robot with asymmetric turning radius, characterized in that, The method includes the following steps: S100: Establish a side tilt model of an underwater glider considering oil bladder offset. By analyzing the position vectors of the oil bladder mass and movable mass in the body coordinate system and their influence on the system momentum, construct the kinematic equations of the system, and derive the dynamic equations including the influence of oil bladder offset as well as the steady-state dynamic equations of the horizontal and vertical planes. S200: Based on the roll model, a discrete state equation for roll angle acceleration is constructed. The forgetting factor recursive least squares algorithm is adopted, and the oil bladder center of gravity offset is identified online through multi-step differentiation and iterative update. S300: Calculates the left and right turning radii by using the identified center of gravity offset and combining it with the real-time motion parameters of the underwater glider. S400: Based on the calculated left and right turning radii, an algorithm is designed to generate a suboptimal Dubins-Helix path. The path consists of a pitch angle smoothing segment and a Dubins-Helix trajectory segment. Its projection on the horizontal plane satisfies curvature radius constraints, and its pitch angle constraints in three-dimensional space. By optimizing the pitch angle and helix segment parameters, an energy-optimized path is generated, with path length, time, and energy consumption as adversarial indicators. Energy consumption optimization considers the energy consumption components of the buoyancy propulsion system, attitude adjustment system, communication control and measurement system, and heading control system. In S400, the suboptimal Dubins-Helix path is generated first based on the starting and ending points... With the two-dimensional Dubins path length and maximum pitch angle The relationship is determined, including the two-dimensional Dubins path length. , and These are the rotation angles corresponding to the arcs. It is the length of a straight line; relationship and path generation fall into the following two categories: Case 1: When the conditions are met At this time, the path must contain a spiral segment, and its parameters must satisfy the basic height relationship: Its optimization is divided into two modes: no attitude switching and attitude switching. (1) For the shortest path without attitude switching, its total length The pitch angles for each segment are determined by the following formula: ； ； in and The optimal value is determined by the following optimization problem: ； The total length of the path can also be expressed as: ; (2) For the shortest path with attitude switching, its total length The pitch angles and attitude switching point heights for each segment are determined by the following formula: ； ； ； in and The optimal value is determined by the following optimization problem: ； The total length of the path can also be expressed as: ; in, This is the length of the attitude-free switching path. The length of the attitude switching path; , The projections of the left and right turning radii onto the two-dimensional plane; This represents the length of the Dubins path. As the starting height, The final height; , , For each segment of pitch angle; This is the absolute value of the maximum pitch angle; The height difference between the attitude switching points; , , , It is a natural number; Case 2: When the conditions are met At this time, the path tilts the two-dimensional Dubins path by a single pitch angle to adapt to the height difference, forming the shortest path without spiral segments, and its total length is... With uniform pitch angle Determined by the following formula: ； ； Energy consumption optimization in S500 is achieved by calculating the total energy consumption E3D, specifically as follows: ； ； ； ； ； in, For fluid density; The power of the hydraulic pump with respect to depth; The flow rate of the hydraulic pump with respect to depth; The power of the solenoid valve; The flow rate of the solenoid valve; For movable mass stroke; The moving speed of movable mass; Input power to the motor; Power for communication, control, and measurement systems; The energy consumption of the servo motor when turning with the minimum turning radius.

2. The method according to claim 1, characterized in that, In S100, the position vectors of the oil bladder mass and movable mass in the body coordinate system and their motion are analyzed to influence the system momentum, and the dynamic equations are derived, including: Define the mass momentum of the oil bladder and movable mass momentum Specifically: ； The forces generated by the mass of the oil bladder and its movable mass and They are represented as follows: ； ； The total internal mass forces and moments are expressed as follows: ； in, For oil bladder quality, For movable mass; The linear velocity in the body coordinate system; Angular velocity in body coordinate system; , These are the position vectors of the oil bladder mass and the movable mass in the volume coordinate system, respectively. , These are the relative velocities of the oil bladder mass and the movable mass in the volume coordinate system, respectively. These are the linear velocity components along the x, y, and z axes in the body coordinate system, respectively. These are the angular velocity components about the x, y, and z axes in the body coordinate system, respectively. These are the coordinates of the oil bladder mass on the x, y, and z axes of the volume coordinate system; These are the coordinates of the movable mass on the x, y, and z axes of the volume coordinate system; The coordinates of the center of gravity on the z-axis of the volume coordinate system; Let be the velocity of the oil bladder mass moving along the y-axis of the body coordinate system; Let be the velocity of the movable mass along the x-axis of the body coordinate system. Let x be the linear acceleration along the x-axis of the body coordinate system; Angular acceleration about the z-axis of the body coordinate system; and The y-axis and z-axis components of the mass momentum of the oil bladder are respectively defined; The time derivative of the y-axis component of the oil bladder's mass momentum; The force vector generated by the mass of the oil bladder; The force vector generated by the movable mass; For the mass momentum of the oil bladder The time derivative; Momentum of movable mass The time derivative; For the total internal mass force; The torque generated by the mass of the oil bladder; The torque generated by the movable mass; This represents the total internal mass torque; It is the acceleration due to gravity; It is the direction of gravity relative to the body coordinate system. It is the unit vector pointing in the direction of gravity in the Earth coordinate system; The total internal mass forces and torques are used as inputs to the rigid body dynamics equations of the underwater glider, in order to establish a complete dynamic model that includes the effects of the oil bladder and movable mass.

3. The method according to claim 2, characterized in that, The kinematic equations of the tilt model of the underwater glider considering oil bladder offset in S100 are used to describe the transformation relationship between the body coordinate system and the ground coordinate system, specifically including: The kinematic relationship between roll angle and yaw angle is as follows: ； ； in, The angular velocity is the roll angle. Let be the angular velocity of rotation about the x-axis of the body coordinate system. Let be the angular velocity of rotation about the z-axis of the body coordinate system. Angular velocity is the yaw angle. This refers to the roll angle; Rotation matrix from volume coordinate system to ground coordinate system Defined as: ； in, , , , , These are roll angle, yaw angle, and pitch angle, respectively. The kinematic model also includes differential equations for the rotation matrix and position, which are used to describe the evolution of the robot's posture and calculate the robot's position change in the ground coordinate system, respectively. ； ； in, The time derivative of the rotation matrix. Let be the angular velocity about the x-axis of the body coordinate system. Let be the angular velocity about the y-axis of the body coordinate system. Let be the angular velocity about the z-axis of the body coordinate system. Let be the antisymmetric matrix of angular velocity. This is the time derivative of the position vector in the Earth coordinate system. Let be the linear velocity vector represented in volume coordinates.

4. The method according to claim 3, characterized in that, The steady-state dynamic equations in S100 include horizontal plane equations and vertical plane equations. The horizontal plane equations are as follows: ； The equation of the vertical plane is: ； in, This refers to the total mass of the underwater gliding robot. For drainage quality; Add mass to the water body For oil bladder quality, For movable mass; , , Let x represent the coordinates of the water body, oil sac, and movable mass on the x-axis of the volume coordinate system. , , , , For additional quality coefficients; , Let x be the moment of inertia about the x-axis and z-axis; It is net buoyancy; It is the resultant velocity on the vertical plane of the underwater gliding robot in a body coordinate system; It is excluding movable mass The sum of all masses except those mentioned above; and They represent Coordinates along the z-axis and x-axis; and They represent Coordinates along the z-axis and x-axis; X(), Y(), K(), N() are hydrodynamic terms; and It is the hydrodynamic coefficient.

5. The method according to claim 4, characterized in that, S200 includes: S210: Construct system equations for parameter identification, expressing the relationship between the system's input, output, and identified parameters as follows: ； in , , and These represent the time steps. The system output, input, parameters to be identified, and deviation at that time. It is considered to be white noise with a mean of 0; S220: Based on the roll model, a discrete state equation for roll angle acceleration is established, specifically: ； in It is the offset of the oil bladder's center of gravity to be identified. and For time steps linear velocity, and For time steps angular velocity, to These are the model coefficients. For time steps rudder angle; By matching the discrete state equations with the system equations of FFRLS, the following is determined: ； ； ； Roll acceleration is calculated using a multi-step differential method: ； in It is the sampling time. It is the time step interval oriented towards differentiation; S230: In each sampling period Within, perform the following calculations in sequence: Measured or calculated and ; Calculate the deviation Specifically: ； Calculate the correction gain as follows: ； in It is a time step Algorithm gain at that point It is a time step The noise covariance matrix at that location, It is a forgetting factor; Update the parameter estimates at the end of each period. Specifically: ； Update the noise covariance matrix Specifically: 。 6. The method according to claim 5, characterized in that, The formulas for calculating the left-turn radius and right-turn radius in S300 are as follows: ； And satisfy the following constraints: ； ； ； in, For the three-dimensional radius of curvature, Turn left; Turn right. and These are the minimum and maximum roll angles allowed for underwater gliders, respectively. and These are the minimum and maximum allowable deflection angles of the rudder, respectively.

7. The method according to claim 6, characterized in that, The suboptimal Dubins-Helix path in S400 consists of five sub-path segments, including a pitch smoothing segment and a Dubins-Helix trajectory segment; the generation process of the three-dimensional spiral curve in the Dubins-Helix trajectory segment includes: Assuming the helical curve segment maintains a constant pitch angle during flight. The center of the arc projected onto the horizontal plane by the spiral curve is Then any point on the spiral curve The coordinates are represented as: ； in, The starting point of this spiral curve segment The height coordinates, yes The normalized unit vector, It is the starting point on the horizontal projection plane. arrive The rotation angle, when the spiral is left-handed. When the spiral is right-handed, ; The radius of curvature of the horizontal projection of the helical curve; The value of the turning radius of the robot's three-dimensional motion and pitch angle The following relationship must be satisfied: ； in, It means turn left. This means turn right; Horizontal projected radius of curvature The minimum turning radius constraint based on the robot's maneuverability must be satisfied: (1) Curvature radius constraint of the horizontal projection of the left-handed helical curve: ; (2) Curvature radius constraint of the horizontal projection of the right-hand helical curve: ; in, , These are the minimum radii of curvature for left and right turns, respectively.

8. An energy-optimized path smoothing generation system for an underwater glider with asymmetric turning radius, used to perform the method as described in any one of claims 1 to 7, characterized in that, include: The tilt model building module is used to build a tilt model of an underwater glider considering the oil bladder offset. By analyzing the position vectors of the oil bladder mass and movable mass in the body coordinate system and their motion on the system momentum, the kinematic equations of the system are constructed, and the dynamic equations including the influence of oil bladder offset, as well as the steady-state dynamic equations in the horizontal and vertical planes, are derived. The oil bladder center of gravity offset identification module is used to construct the discrete state equation of roll angle acceleration based on the roll model. It adopts the forgetting factor recursive least squares algorithm and identifies the oil bladder center of gravity offset online through multi-step differentiation and iterative update. The left and right turn radius determination module uses the identified center of gravity offset and the real-time motion parameters of the underwater glider to calculate the left and right turn radii. The path generation module is used to design an algorithm to generate a suboptimal Dubins-Helix path based on the calculated left and right turning radii. The path consists of a pitch angle smoothing segment and a Dubins-Helix trajectory segment. The projection on the horizontal plane satisfies the radius of curvature constraint, and the pitch angle constraint is satisfied in three-dimensional space. By optimizing the pitch angle and helix segment parameters, an energy-optimized path is generated with path length, time consumption, and energy consumption as adversarial indicators. The energy consumption optimization takes into account the energy consumption components of the buoyancy drive system, attitude adjustment system, communication control and measurement system, and heading control system.

Images