A trajectory tracking control method for autonomous underwater robot system

Abstract

The application relates to the technical field of underwater robot control, in particular to an autonomous underwater robot system trajectory tracking control method, which can effectively solve the problems of waste of communication resources of the autonomous underwater robot, influence of model uncertainty and external environment, and ensure accurate position and speed tracking of the autonomous underwater robot.

Description

Technical Field

[0001] This invention relates to the field of underwater robot control technology, specifically to a trajectory tracking control method for an autonomous underwater robot system. Background Technology

[0002] Autonomous underwater robots (AUVs) are robotic systems capable of autonomous movement, perception, and task execution underwater. In recent years, AUVs have received increasing attention, with a wide range of applications including marine exploration, deep-sea scientific research, seabed resource development, and marine environmental monitoring.

[0003] However, due to the environment in which autonomous underwater vehicles (AUVs) operate and uncertainties in their system models, they often fail to issue correct commands. On one hand, the underwater environment is complex and ever-changing, requiring AUVs to sense, analyze, and react. However, influenced by factors such as water flow, temperature, and marine life, their actuators are constrained to varying degrees, making it difficult to accurately perceive the surrounding environment, resulting in suboptimal tracking and control performance. On the other hand, underwater communication is challenging, easily subject to interference and signal loss, leading to low communication efficiency and poor transmission reliability. Limited by the underwater environment, data transmission speeds are slow, and transmission capacity is limited. Furthermore, the continuous communication between the actuators and control devices in an AUV system means that tracking and control require real-time calculation and operation, placing high demands on the computer's computing power and storage capacity. This also causes wear and tear on the system structure and wastes communication resources, resulting in poor tracking performance and inaccurate tracking data. Summary of the Invention

[0004] The purpose of this invention is to provide a trajectory tracking and control method for an autonomous underwater robot system.

[0005] The technical solution of this invention is as follows:

[0006] This invention provides a trajectory tracking control method for an autonomous underwater robot system, comprising the following steps:

[0007] S1: Based on the kinematic and dynamic models of the autonomous underwater robot, construct an autonomous underwater robot control system containing uncertainties and external disturbances to obtain the reference trajectory of the tracked target;

[0008] S2: The autonomous underwater robot control system combines the error transformation function to obtain the error system model; based on the error system model, the reference trajectory is combined with the actual output trajectory of the autonomous underwater robot control system to obtain the design variables related to the position tracking error; the velocity vector of the autonomous underwater robot is combined with the control input signal of the kinematic model to obtain the design variables related to the velocity tracking error.

[0009] S3: Based on the design variables related to the position tracking error, construct the first obstacle Lyapunov function, combine the control input signal of the kinematic model and the position adaptive law to obtain the derivative of the first obstacle Lyapunov function, control the range of the derivative of the first obstacle Lyapunov function to not exceed the first threshold, and realize the trajectory position tracking control of the autonomous underwater robot system.

[0010] S4: Using the Lyapunov function based on the first obstacle and the design variables related to the speed tracking error, a second obstacle Lyapunov function is constructed. Combining the continuous control input signal and the speed adaptive law, the derivative of the second obstacle Lyapunov function is obtained. The range of the derivative of the second obstacle Lyapunov function is controlled to not exceed the second threshold, thereby realizing the trajectory speed tracking control of the autonomous underwater robot system.

[0011] The trajectory tracking control method for an autonomous underwater robot system as described above is characterized in that the operation of S2 specifically includes:

[0012] The position tracking error vector is obtained by subtracting the reference trajectory from the actual output trajectory of the kinematic model in the autonomous underwater robot control system. The position tracking error vector is then multiplied by the error transformation function to obtain the design variables related to the position tracking error.

[0013] The velocity tracking error vector is obtained by subtracting the control input signal of the kinematic model from the velocity vector of the dynamic model in the autonomous underwater robot control system. The velocity tracking error vector is then multiplied by the error transformation function to obtain the design variables related to the velocity tracking error.

[0014] As described above, in the trajectory tracking control method for an autonomous underwater robot system, the reference trajectory of the autonomous underwater robot control system is:

[0015] ,

[0016] For reference trajectory, For time.

[0017] The error transformation function is:

[0018]

[0019] For error transformation function, Given a positive constant, For time, The stable time set by the user.

[0020] In the trajectory tracking control method for the autonomous underwater robot system described above, the operation of step S3 is specifically as follows:

[0021] Based on the design variables related to the position tracking error, a barrier Lyapunov function containing upper and lower error constraint boundary functions is constructed. Combined with the control gain matrix of the position adaptive law and the neural network weight error of the dynamic model, a first barrier Lyapunov function is constructed.

[0022] By combining the first barrier Lyapunov function with Young's inequality and taking the derivative, we obtain the derivative of the initial first barrier Lyapunov function.

[0023] After defining the control input signal and position adaptive law of the kinematic model, they are input into the derivative of the initial first obstacle Lyapunov function to obtain the derivative of the first obstacle Lyapunov function. The range of the derivative of the first obstacle Lyapunov function is controlled to not exceed a first threshold to realize the trajectory and position tracking control of the autonomous underwater robot system.

[0024] The control input signal in the kinematic model is:

[0025] For the control input signal of the kinematic model, For adjustable normals of the kinematic model, The control gain matrix of the kinematic model. For position adaptive law The control gain matrix, The control gain vector of the kinematic model. Let Lyapunov function vectors contain boundary function barriers with upper and lower error constraints. For neural network weight estimation of kinematic models, For the first calculation function, Error transformation function The derivative, The proportional function vector of the kinematic model. Design variables related to position tracking error;

[0026] The position adaptive law for:

[0027]

[0028] For position adaptive law, for The control gain vector.

[0029] In the trajectory tracking control method for the autonomous underwater robot system described above, the operation of step S4 is specifically as follows:

[0030] The first obstacle Lyapunov function and the design variables related to velocity tracking error are combined with the positive design parameters and the neural network weight estimation of the dynamic model to construct the second obstacle Lyapunov function.

[0031] By combining the second barrier Lyapunov function with Young's inequality and taking the derivative, we obtain the derivative of the initial second barrier Lyapunov function.

[0032] After defining the continuous control input signal and the speed adaptive law, they are input into the derivative of the initial second obstacle Lyapunov function to obtain the derivative of the second obstacle Lyapunov function. The range of the derivative of the second obstacle Lyapunov function is controlled to not exceed the second threshold, thereby realizing the trajectory and speed tracking control of the autonomous underwater robot system.

[0033] The continuous control input signal is:

[0034]

[0035] For continuous control input signals, This is the event trigger gain constant. For the adjustable constants of the dynamic model, The control gain matrix of the dynamic model. For the speed tracking error vector, The control gain matrix of the speed adaptive law. Error transformation function The derivative, For neural network weight estimation of the dynamic model, The second known function, Design variables related to tracking error, The direction of motion control in the dynamic model is a known constant;

[0036] Wherein, the continuous control input signal satisfies , for Continuous control input signal at any time for , , The event triggering mechanism function, for Continuous control input at any time and Actual control input at any time Measurement error, , for Current control input signal , For time, Indicates that the controller has occurred. The moment of the next trigger For the controller to occur the first The moment of the next trigger For the controller to occur the first The second event trigger and occurrence The time interval for each event to be triggered. The event triggering mechanism function. , This is the event trigger gain constant. This is the event trigger threshold constant.

[0037] The adaptive velocity law is as follows:

[0038]

[0039] For the velocity adaptive law, for The control gain vector.

[0040] The beneficial effects of this invention are as follows:

[0041] This invention provides a trajectory tracking control method for an autonomous underwater vehicle (AUV) system. Based on kinematic and dynamic models, it models the AUV with model uncertainties and external disturbances to obtain an AUV control system that reduces model uncertainties and the influence of the external environment. Based on an error transformation function, it constructs a first barrier Lyapunov function and a second barrier Lyapunov function containing radial basis function neural network parameters. This overcomes the influence of different initial values ​​on the AUV system, ensuring that the system does not violate asymmetric time-varying constraint boundaries after reaching a stable time. Through continuous control input signals related to the design event triggering mechanism, the AUV system can reduce data transmission frequency and save communication resources, solving the problem of continuous communication between the actuator and control device in the AUV system. Furthermore, by controlling the derivative ranges of the first and second barrier Lyapunov functions to not exceed a first threshold and a second threshold, respectively, the system can operate stably and achieve accurate tracking of the system's position and velocity.

[0042] This invention provides a trajectory tracking control method for an autonomous underwater robot system. Compared with existing time-triggered control and periodic-triggered control, the event-triggered mechanism designed in this invention saves computing resources, avoids unnecessary calculations, and improves computing efficiency; the system response speed is faster, thus making the system more stable and reliable; it can achieve high-performance tracking while avoiding frequent updates of control inputs, reducing the accumulation of tracking errors, and improving the tracking accuracy of the system.

[0043] This invention provides a trajectory tracking control method for an autonomous underwater vehicle (AUV) system. By introducing an error transformation function and a radial basis function (RBF) neural network function, it overcomes the influence of different bounded initial values ​​on the AUV system. On the one hand, compared with the traditional constraint problem handling, it further relaxes the constraints on the initial error of the system, ensuring that the system does not violate the asymmetric time-varying constraint boundary after reaching a steady time. On the other hand, the introduction of the RBF neural network function realizes the approximation of the model uncertainty and external disturbances of the system, solves the complex relationships of various functions in the system, and enhances the robustness of the system. Attached Figure Description

[0044] The solutions and advantages of this application will become clear to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of the invention.

[0045] In the attached diagram:

[0046] Figure 1 This is a flowchart of the control method in the embodiment;

[0047] Figure 2 shows scenarios 1 and 2 in the embodiments. The position and attitude curve changes over time. Figure 2a This is a graph showing the change in longitudinal motion position and attitude over time. Figure 2b This is a graph showing the change in lateral movement position and attitude over time. Figure 2c This is a graph showing the change in the bow roll angle position and attitude over time.

[0048] Figure 3 shows scenarios 1 and 2 in the embodiments. The tracking error curve changes over time. Figure 3a This is a graph showing the change in longitudinal motion position and attitude tracking error over time. Figure 3b This is a graph showing the change in lateral motion position and attitude tracking error over time. Figure 3c The graph shows the change in bow roll angle position attitude tracking error over time.

[0049] Figure 4 shows scenarios 1 and 2 in the embodiments. A schematic diagram of the velocity curve changing over time. Figure 4a This is a graph showing the change in longitudinal motion velocity and attitude over time. Figure 4b This is a graph showing the change in lateral motion velocity and attitude over time. Figure 4c The diagram shows the change in bow roll angular velocity attitude over time.

[0050] Figure 5 As shown in Case 1 of the embodiment A schematic diagram of the control input curves based on event-triggered control and time-triggered control;

[0051] Figure 6 Case 2 in the example A schematic diagram of the control input curves based on event-triggered control and time-triggered control.

[0052] Figure 7 As shown in Case 1 of the embodiment A schematic diagram of the event-triggered control of the trigger time and event interval.

[0053] Figure 8 Case 2 in the example A schematic diagram of the event-triggered control of the trigger time and event interval.

[0054] Figure 9 For cases 1 and 2 in the examples The parameter adaptive law curve.

[0055] Figure 10 This is a schematic diagram of the control device in the embodiment. Detailed Implementation

[0056] This embodiment provides a trajectory tracking control method for an autonomous underwater robot system. (See also...) Figure 1 This includes the following steps:

[0057] S1: Based on the kinematic and dynamic models of the autonomous underwater robot, construct an autonomous underwater robot control system containing uncertainties and external disturbances to obtain the reference trajectory of the tracked target;

[0058] S2: The autonomous underwater robot control system combines the error transformation function to obtain the error system model; based on the error system model, the reference trajectory is combined with the actual output trajectory of the autonomous underwater robot control system to obtain the design variables related to the position tracking error; the velocity vector of the autonomous underwater robot is combined with the control input signal of the kinematic model to obtain the design variables related to the velocity tracking error.

[0059] S3: Based on the design variables related to the position tracking error, construct the first obstacle Lyapunov function, combine the control input signal of the kinematic model and the position adaptive law to obtain the derivative of the first obstacle Lyapunov function, control the range of the derivative of the first obstacle Lyapunov function to not exceed the first threshold, and realize the trajectory position tracking control of the autonomous underwater robot system.

[0060] S4: Based on the first obstacle Lyapunov function and the design variables related to the velocity tracking error, a second obstacle Lyapunov function is constructed. Combining the continuous control input signal and the velocity adaptive law, the derivative of the second obstacle Lyapunov function is obtained. The range of the derivative of the second obstacle Lyapunov function is controlled to not exceed the second threshold, thereby realizing the trajectory velocity tracking control of the autonomous underwater robot system.

[0061] S1 Based on the kinematic and dynamic models of the autonomous underwater robot, an autonomous underwater robot control system containing uncertainties and external disturbances is constructed to obtain the reference trajectory of the tracked target.

[0062] Kinematic model. The kinematic model is obtained by multiplying the position and attitude vectors of the autonomous underwater vehicle by its velocity vectors. The kinematic model can be obtained by the following formula:

[0063]

[0064] This represents the position and attitude derivative vector of an autonomous underwater robot in an inertial coordinate system. For rotation matrix, ; This represents the position and attitude vector of the autonomous underwater robot in the inertial coordinate system. The longitudinal position in the inertial coordinate system. The lateral position in the inertial coordinate system. The position of the bow roll angle in the inertial coordinate system; Let V be the velocity vector of the autonomous underwater robot in rigid body coordinates. , The longitudinal velocity in the rigid body coordinate system. Let be the lateral velocity in the rigid body coordinate system. Let be the angular velocity of the bow roll motion in the rigid body coordinate system.

[0065] Dynamics Model. Based on the inertial matrix, velocity vector derivative, Coriolis and centripetal force matrices, hydrodynamic damping matrix, gravity recovery vector of position and attitude vector, input torque, external disturbances, and uncertainties of the autonomous underwater vehicle, a dynamics model is obtained. The dynamics model can be obtained from the following calculation formulas:

[0066] ,

[0067] The inertia matrix, This represents the derivative of the velocity vector of an autonomous underwater vehicle in a rigid body coordinate system. For the Coriolis and centripetal force matrices.

[0068]

[0069] satisfy , , , For the quality of autonomous underwater robots, The coefficient of the longitudinal velocity derivative in fluid dynamics; The additional mass during longitudinal motion, The coefficient of the derivative of the lateral velocity in fluid dynamics; This refers to the additional mass during lateral motion. It represents the derivative coefficient of the angular velocity of the bow roll motion in hydrodynamics.

[0070] The fluid dynamics damping matrix can be obtained from the following formula:

[0071] ,

[0072] The coefficient of friction is the linear friction coefficient for longitudinal motion velocity. The linear friction coefficient is the coefficient of friction for lateral motion velocity. The linear friction coefficient is the angular velocity of the bow roll motion. The longitudinal velocity is the hydrodynamic coefficient. The lateral velocity is the hydrodynamic coefficient. The angular velocity of the bow roll is the hydrodynamic coefficient. This is the gravity recovery vector. .

[0073] This represents the model uncertainty term in the control system of an autonomous underwater robot. . This represents the input torque of the autonomous underwater robot's control system.

[0074] Due to external interference,

[0075] .

[0076] By introducing model uncertainties and external disturbances into the control system of autonomous underwater robots, the impact of model uncertainties and the external environment on the autonomous underwater robot system can be reduced.

[0077] make Set the system initial values ​​respectively , . The above parameters are taken from empirical data, are widely used, and have high modeling accuracy.

[0078] To further simplify the control system of autonomous underwater vehicles (AUVs), the physical model of the AUV control system is transformed into a mathematical model. Specifically, let... but

[0079]

[0080] in, Let be the position and attitude derivative vector of the autonomous underwater robot in the inertial coordinate system. Let be the velocity derivative vector of the autonomous underwater robot in a rigid body coordinate system.

[0081] The autonomous underwater robot control system is as follows:

[0082]

[0083] in, Inertia matrix An invertible matrix, Let be a nonlinear function containing the Coriolis and centripetal force matrices, the hydrodynamic damping matrix, the gravity restoration vector, external disturbances, and uncertainty terms. Let be a vector containing the position, attitude, and velocity of the autonomous underwater robot. This is the actual output trajectory of the system.

[0084] Reference trajectory of autonomous underwater robot control system It can be obtained through the following calculation formula:

[0085]

[0086] For time, reference trajectory The curve showing the change over time is shown in Figure 2.

[0087] The S2 autonomous underwater robot control system combines the error transformation function to obtain the error system model. Based on the error system model, the reference trajectory is combined with the actual output trajectory of the autonomous underwater robot control system to obtain the design variables related to the position tracking error. The velocity vector of the autonomous underwater robot and the control input signal of the kinematic model are combined to obtain the design variables related to the velocity tracking error.

[0088] Traditional autonomous underwater vehicle (AUV) tracking control systems require that the tracking error of the target trajectory be limited to within the constraint boundary. However, due to the complexity of the marine environment, the tracking error of the target trajectory often cannot be guaranteed to remain within the constraint range, thus causing traditional control algorithms to fail.

[0089] To address this technical problem, an error transformation function is introduced into the autonomous underwater robot control system to obtain an error system model. Based on the error system model, the actual output trajectory of the kinematic model is subtracted from the reference trajectory to obtain the position tracking error vector. The position tracking error vector is multiplied by the error transformation function to obtain the design variables related to the position tracking error. The velocity vector of the dynamic model is subtracted from the control input signal of the kinematic model to obtain the velocity tracking error vector. The velocity tracking error vector is multiplied by the error transformation function to obtain the design variables related to the velocity tracking error.

[0090] Specifically, define the tracking error and coordinate transformation:

[0091]

[0092] This represents the position tracking error vector of the autonomous underwater robot. This represents the velocity tracking error vector of the autonomous underwater robot. Control input signals for the kinematic model Design variables related to tracking error, Design variables related to position tracking error, Design variables related to speed tracking error.

[0093] Among them, the error transformation function For time The formula for calculating a continuous piecewise function is as follows:

[0094]

[0095] in, Given positive constants, satisfying . For time, The stable time set for the user To track error in The initial value at time, Let the upper and lower constraint boundary functions be represented respectively. The initial value at time.

[0096] Based on the mathematical model of the autonomous underwater robot control system containing uncertainties and external disturbances, combined with the reference trajectory Actual output of the autonomous underwater robot control system Sum and error transformation function Define the tracking error and coordinate transformation of the autonomous underwater robot system, and use the error transformation function. In time Internally, the tracking error of the autonomous underwater robot system exhibits a relatively fast convergence rate, and it achieves convergence when reaching... Subsequently, the tracking error of the autonomous underwater robot system remained within the constraint boundary function.

[0097] S3 constructs a first obstacle Lyapunov function based on design variables related to position tracking error. Combining the control input signal of the kinematic model and the position adaptive law, the derivative of the first obstacle Lyapunov function is obtained. The range of the derivative of the first obstacle Lyapunov function is controlled to not exceed a first threshold, thereby realizing trajectory and position tracking control of the autonomous underwater robot system.

[0098] S31 Based on the design variables related to the position tracking error, a barrier Lyapunov function containing upper and lower error constraint boundary functions is constructed. This function is then combined with the control gain matrix of the position adaptive law and the neural network weight error of the dynamic model to construct the first barrier Lyapunov function.

[0099] In practical applications, autonomous underwater vehicle (AUV) systems may be affected by various nonlinear constraints. For example, complex hydrodynamic effects such as water flow and pressure in the underwater environment can make the motion and control of AUV systems difficult; the attitude stability of AUV systems requires consideration of both the complexity and nonlinear characteristics of the underwater environment; and the motion of AUV systems is affected by factors such as mechanical structure, underwater friction, and inertia. Therefore, AUV systems need to consider multiple factors including physical structure, motion constraints, and sensor errors.

[0100] To address the nonlinearity problem in autonomous underwater robot systems, a barrier Lyapunov function containing upper and lower error constraint boundary functions is constructed based on design variables related to position tracking error. Specifically, a barrier Lyapunov function is designed to address position tracking error.

[0101] ,

[0102] in, Design variables related to position tracking error, The design variable components representing the longitudinal motion tracking error, The design variable components representing the lateral motion tracking error, The design variable components represent the tracking error of the bow roll motion. For Lyapunov function vectors that are boundary functions constrained by upper and lower errors for positional errors, Let Lyapunov functions represent the obstacle functions at the longitudinal, lateral, and yaw motion positions, respectively.

[0103] and Let be the upper and lower error constraint boundary functions of the Lyapunov function as a barrier to position error, respectively, and let the initial values ​​of the constraint boundary functions satisfy . Constraint boundary functions , The expressions are respectively . To achieve system stability within a certain timeframe. Afterwards, the system state does not exceed the state constraint boundary. The error constraint boundary function should satisfy , ,in, These are the upper and lower constraint boundary functions, respectively. These represent the maximum and minimum values ​​of the reference signal, respectively.

[0104] Differentiating the Lyapunov function, which is a barrier to position tracking error, yields:

[0105] ,

[0106]

[0107] in, This represents the control gain vector of the kinematic model. This represents the longitudinal motion control gain of the kinematic model. This represents the lateral motion control gain of the kinematic model. This represents the control gain for the direction of the yaw motion in the kinematic model. The vector representing the proportional function of the kinematic model. The longitudinal motion proportional function representing the kinematic model. The lateral motion proportional function of the kinematic model. Let the yaw motion proportional function of the kinematic model represent the boundary function of the Lyapunov function with upper and lower error constraints for position tracking error be rewritten as:

[0108]

[0109] . , .

[0110] Constructing a barrier Lyapunov function containing upper and lower error constraints for position tracking error can effectively address the adverse effects of different initial values ​​on autonomous underwater robot systems during task execution.

[0111] Based on the derivative of the Lyapunov function for the obstacle of position tracking error, the autonomous underwater robot control system can be written in the following form:

[0112]

[0113] Construct the first obstacle Lyapunov function. The obstacle Lyapunov function vector for position tracking error is then used. F 1 By combining the control gain matrix of the position adaptive law and the neural network weight error of the dynamic model, a first barrier Lyapunov function based on the position tracking error and the error transformation function is constructed. The calculation formula is:

[0114]

[0115] in, for transpose, For the neural network weight error of the kinematic model, The optimal values ​​for the neural network weights of the kinematic model. This is for estimating the neural network weights of a kinematic model.

[0116] The autonomous underwater vehicle (AUV) system incorporates radial basis function (RBF) neural networks. RBF networks are used to extract, classify, and approximate model uncertainties and the effects of external disturbances (wind speed, ocean currents) within the dynamic system, mapping the dynamic system model data to achieve simplified data processing. An adaptive neural network tracking controller is used within the RBF network to facilitate rapid computation and processing within the AUV system.

[0117] By combining the first obstacle Lyapunov function with Young's inequality, the derivative of the initial first obstacle Lyapunov function can be obtained.

[0118] Young's inequality is:

[0119] ,

[0120] in All represent positive real numbers and satisfy the following conditions: It is used to handle strong coupling terms, bounded disturbances, neural network weights and other nonlinear terms in the control system of autonomous underwater robots, which simplifies the design of the control mechanism of the autonomous underwater robot control system and obtains a more accurate error tracking range.

[0121] Therefore, combining Young's inequality, we can apply this to the first barrier Lyapunov function. Differentiation yields the derivative of the Lyapunov function with respect to the initial first barrier of position tracking error. :

[0122]

[0123] in, Let be a first computable function whose expression satisfies , It is the first Gaussian function. For the weight vector, The derivative of the reference trajectory is expressed as follows: . For the error vector, The derivative of the error transformation function is expressed as follows:

[0124]

[0125] S33 defines the control input signal and position adaptive law of the kinematic model, and inputs them into the derivative of the initial first obstacle Lyapunov function to obtain the derivative of the first obstacle Lyapunov function. The range of the derivative of the first obstacle Lyapunov function is controlled to not exceed the first threshold, so as to realize the trajectory and position tracking control of the autonomous underwater robot system.

[0126] The control input signal in the kinematic model can be defined as:

[0127] For the control input signal of the kinematic model, For adjustable normals of the kinematic model, The control gain matrix of the kinematic model. For position adaptive law The control gain matrix, The control gain vector of the kinematic model. Let Lyapunov function vectors contain boundary function barriers with upper and lower error constraints. For neural network weight estimation of kinematic models, For the first calculation function, Error transformation function The derivative of The proportional function vector of the kinematic model. These are design variables related to position tracking error.

[0128] Position Adaptive Law Defined as:

[0129]

[0130] For position adaptive law estimation, for The control gain vector.

[0131] in, .

[0132] Control input signal of kinematic model And adaptive law The derivative of the Lyapunov function, the first obstacle, is input into the control system of the autonomous underwater robot. In this process, the derivative of the first obstacle Lyapunov function is obtained. for:

[0133]

[0134] in, Setting the first threshold to 0 means that the derivative of the Lyapunov function controlling the first obstacle is never greater than 0. This implies that the autonomous underwater robot control system will gradually approach the target trajectory position while continuously consuming energy, thus achieving precise tracking of the autonomous underwater robot system's trajectory position.

[0135] In a real marine environment, the hardware structure of an autonomous underwater vehicle (AUV) control system includes an inertial measurement unit (IMU), temperature sensors, humidity sensors, and water quality sensors to acquire information such as the system's position, attitude, angle, and water density. The control mechanism uses the information acquired by the sensors to perform calculations and inputs control signals through a kinematic model. And adaptive law The autonomous underwater robot's control system adjusts its motion posture, position, and angle to achieve precise tracking of the target trajectory.

[0136] S4 Based on the first obstacle Lyapunov function and the design variables related to the velocity tracking error, construct the second obstacle Lyapunov function. Combine the continuous control input signal and the velocity adaptive law to obtain the derivative of the second obstacle Lyapunov function. Control the range of the derivative of the second obstacle Lyapunov function to not exceed the second threshold, thereby realizing the trajectory velocity tracking control of the autonomous underwater robot system.

[0137] S41 combines the first obstacle Lyapunov function and the design variables related to velocity tracking error with the positive design parameters and the neural network weight estimation of the dynamic model to construct the second obstacle Lyapunov function. for:

[0138]

[0139] in, For positive design parameters, The neural network weight error of the dynamic model, This is for estimating the neural network weights of the dynamic model.

[0140] By combining the second barrier Lyapunov function with Young's inequality, the derivative of the initial second barrier Lyapunov function can be obtained.

[0141] when At that time, combining the second obstacle Lyapunov function with Young's inequality, after differentiation, the derivative of the initial second obstacle Lyapunov function is obtained as follows:

[0142]

[0143] S43 defines the continuous control input signal and the speed adaptive law, and inputs them into the derivative of the initial second obstacle Lyapunov function to obtain the derivative of the second obstacle Lyapunov function. The range of the derivative of the second obstacle Lyapunov function is controlled to not exceed the second threshold, thereby realizing the trajectory and speed tracking control of the autonomous underwater robot system.

[0144] definition Continuous control input signal at any time ,

[0145]

[0146] For continuous control input signals, This is the event trigger gain constant. For the adjustable constants of the dynamic model, The control gain matrix of the dynamic model. For the speed tracking error vector, The control gain matrix of the speed adaptive law. Error transformation function The derivative, For neural network weight estimation of the dynamic model, The second known function, , It is the second Gaussian function. Design variables related to tracking error, The direction of motion control in the dynamic model is a known constant.

[0147] Current autonomous underwater vehicle (AUV) systems employ time-triggered communication methods, typically updating the controller state and calculating control algorithms at fixed time intervals. However, as control systems become increasingly complex, time-triggered control methods are becoming less practical and may lack sufficient precision in terms of control effectiveness.

[0148] To solve this technical problem, a continuous control input signal was designed. satisfy , for Continuous control input signal at any time for , , The event triggering mechanism function, for Continuous control input at any time and Actual control input at any time Measurement error, , for Current control input signal , For time. Among them, Indicates that the controller has occurred. The moment of the next trigger For the controller to occur the first The moment of the next trigger For the controller to occur the first The second event trigger and occurrence The time interval between events.

[0149] Event triggering mechanism function , This is the event trigger gain constant. This is the event trigger threshold constant.

[0150] By introducing event-triggered functions into the autonomous underwater vehicle (AUV) control system, an event-triggered mechanism is designed. The continuous control input signals tracking the AUV system's speed satisfy the conditions of this mechanism. This mechanism effectively reduces the number of control state updates and calculations, conserving system resources and improving the reliability and stability of the AUV control system. It also more accurately reflects the system's dynamic behavior, only updating control when the tracking error exceeds a certain threshold during task execution. This avoids overly frequent controller responses and reduces the risk of system instability.

[0151] For autonomous underwater vehicle (AUV) systems with nonlinear and strongly coupled characteristics, traditional time-triggered control methods may not be suitable. However, AUV systems with event-triggered mechanisms can better control nonlinear systems, such as constraint control. Event-triggered control methods can provide more accurate and effective control schemes.

[0152] Define the adaptive velocity law for:

[0153]

[0154] for The control gain vector.

[0155] Will Continuous control signal at any time Adaptive velocity law Input Second Barrier: Lyapunov Function Derivative In this process, the derivative of the second obstacle Lyapunov function is obtained. The range is:

[0156]

[0157] in, Setting the second threshold to 0 means that the derivative of the Lyapunov function controlling the second obstacle is always no greater than 0. This implies that the autonomous underwater vehicle system can achieve precise tracking of the system's trajectory and velocity while continuously consuming energy.

[0158] In a real marine environment, the hardware structure of an autonomous underwater vehicle (AUV) control system includes an inertial measurement unit (IMU), humidity sensor, and water quality sensor to acquire information such as the system's longitudinal motion, lateral velocity, and yaw angular velocity. The control mechanism uses the information acquired by the sensors to perform calculations and input control signals as designed. Continuous control signal at any time Adaptive velocity law By adjusting the changes in the longitudinal motion speed, lateral motion speed, and yaw motion angular velocity of the system, precise tracking of the target trajectory velocity can be achieved.

[0159] To verify the effectiveness of the trajectory tracking control method for an autonomous underwater robot system proposed in this application, a closed-loop system stability analysis was conducted. Specifically:

[0160] Stability analysis of closed-loop system based on the design in this application From Young's inequality, we get Further scaling yields:

[0161]

[0162] in, , further

[0163] .

[0164] From the above equation and the definition of the barrier Lyapunov function, we can obtain the second barrier Lyapunov function. and It is bounded. According to the definition of coordinate transformation, we can obtain... The boundedness of is similar to the analysis above, and we eventually obtain . Since the signal is bounded, all signals in the closed-loop system are bounded; on the other hand, from the above equation, we get...

[0165] .

[0166] Furthermore, we can obtain from Young's inequality Furthermore, it is possible to obtain .when The convergence range of the tracking error can be derived as follows: .

[0167] On the other hand, definition . express Actual control input signal at time t and Control input at any time If the error is within the specified range and the event triggering condition is met, the control input will switch to... The control input at a given time is used to update the control input signal. Its derivative is...

[0168]

[0169] From the above formula, we get

[0170]

[0171] Further

[0172]

[0173] in, For positive integers, for Time and The time difference ensures that the system operates within the event interval. The number of triggers is limited. Therefore, the autonomous underwater vehicle avoids Zeno's behavior during operation.

[0174] To verify the effectiveness of the trajectory tracking control method for an autonomous underwater robot system proposed in this application, the system was considered in various aspects. The initial value of the autonomous underwater robot system state is within the constraints (Case 1) and There are two scenarios: the initial value of the autonomous underwater robot system is not within the constraints (Case 2).

[0175] Case 1 and Case 2 The changes in position and attitude, tracking error vector, and velocity vector over time are shown in Figures 2, 3, and 4, respectively. Cases 1 and 2 are also described. The control input curves for event-triggered control (the method proposed in this application) and time-triggered control (prior art) are respectively shown in [reference 1]. Figure 5 and Figure 6 .

[0176] Figure 1 A flowchart of the designed control algorithm is presented: the design process of the designed control algorithm is briefly explained by constructing kinematic and dynamic models, error transformation functions, obstacle Lyapunov functions, and other theories.

[0177] Figure 2a , Figure 2b , Figure 2c Case 1 and Case 2 are respectively This diagram illustrates the changes in longitudinal motion position and attitude, lateral motion position and attitude, and yaw angle position and attitude over time. Solid lines represent the curves when the system's initial values ​​are within the constraint boundary range (Case 1), dashed lines represent the curves when the system's initial values ​​are not within the constraint boundary range (Case 2), dotted lines represent the reference trajectory signal (reference signal) curve, and thick dashed and thick dotted lines represent the upper and lower constraint boundary functions, respectively. The curves are shown in the figure. As can be seen from the figure, due to the different initial values, the actuator cannot track the reference signal in time, resulting in significant position and attitude changes. When the system reaches stability, that is, when the system's position and attitude exhibit periodic changes, the curves of cases 1 and 2 coincide. When the system is in cases 1 and 2, the designed control scheme can completely avoid the constraints of the initial conditions and achieve accurate position trajectory tracking.

[0178] Figure 3a , Figure 3b , Figure 3c Case 1 and Case 2 are respectively The diagram illustrates the longitudinal motion position and attitude tracking errors, lateral motion position and attitude tracking errors, and yaw motion angle position and attitude tracking errors. The solid line represents the curve when the system's initial value is within the constraint boundary range (Case 1), and the dashed line represents the curve when the system's initial value is not within the constraint boundary range (Case 2). The thick dashed line and thick dotted line represent the upper and lower error constraint boundary functions, respectively. The curves illustrate this. As can be seen from the graph, at the initial moment of the system, due to different initial values, the actuator cannot track the reference signal in time, resulting in a large position tracking error. When the system reaches stability, the position tracking error exhibits periodic changes, thus achieving the coincidence of the curves in cases 1 and 2. When the system is in cases 1 and 2, the designed control scheme can completely avoid the constraints of the initial conditions, ensuring that the position tracking error remains within the constraint boundaries.

[0179] Figure 4a , Figure 4b , Figure 4c Case 1 and Case 2 are respectively The diagram illustrates the longitudinal velocity and attitude, lateral velocity and attitude, and yaw angular velocity and attitude within the system. The solid line represents the curve when the system's initial values ​​are within the constraint boundary range (Case 1), and the dashed line represents the curve when the system's initial values ​​are not within the constraint boundary range (Case 2). As can be seen from the diagram, at the initial moment of the system, due to the different initial values, the actuator cannot track the reference signal in time, resulting in significant velocity and attitude changes. When the system reaches stability, i.e., the system's velocity and attitude exhibit periodic changes, thus achieving the coincidence of the curves in Case 1 and Case 2. When the system is in Case 1 or Case 2, the designed control scheme ensures the boundedness of the velocity vector, achieving accurate velocity trajectory tracking.

[0180] Figure 5 For case 1 The diagram illustrates the control inputs for event-triggered control and time-triggered control. The solid line represents the control input curve under event-triggered control, and the dashed line represents the control input curve under time-triggered control. As can be seen from the diagram, during system startup, the control input generates a large input quantity. When the system achieves precise tracking, i.e., the control input reaches a stable range without significant fluctuations, the curves of the time-triggered control scheme and the time-triggered control scheme coincide. When the system is in state 1, the designed control scheme not only achieves the same control input effect and obtains good system tracking performance, but also further saves communication resources and reduces costs.

[0181] Figure 6 For case 2 The diagram illustrates the control inputs for event-triggered control and time-triggered control. The solid line represents the control input curve under event-triggered control, and the dashed line represents the control input curve under time-triggered control. As can be seen from the diagram, the control input generates a large amount of data during system startup. When the system achieves precise tracking, i.e., the control input reaches a stable range without significant fluctuations, the curves of the time-triggered control scheme and the time-triggered control scheme coincide. When the system is in state 2, the designed control scheme not only achieves the same control input effect and obtains good system tracking performance, but also further saves resources and reduces costs, effectively verifying the effectiveness of the proposed event-triggered control scheme.

[0182] Figure 7 For case 1 The diagram illustrates the triggering time and event interval of event-triggered control. As can be seen from the diagram, when the initial value of the system is within the constraint boundary range, compared with the time-triggered control strategy, the event-triggered control strategy designed in this invention has 593, 574, and 364 trigger times, which greatly reduces the number of triggers and saves communication resources.

[0183] Figure 8 For case 2 The diagram illustrates the triggering time and event interval of event-triggered control. It shows that when the system's initial value is outside the constraint boundary range, the event-triggered control strategy designed in this invention triggers 613, 587, and 394 times respectively, reducing the number of triggers by the control and execution mechanisms and saving communication resources.

[0184] Figure 9 For cases 1 and 2 The diagram illustrates the adaptive law curves for position and velocity within the system. The solid line represents the longitudinal position in the inertial coordinate system, and the longitudinal velocity adaptive law curve in the rigid body coordinate system represents the longitudinal position in the inertial coordinate system. The dashed line represents the lateral position in the inertial coordinate system, and the lateral velocity adaptive law curve in the rigid body coordinate system represents the yaw angle position in the inertial coordinate system, and the yaw angular velocity adaptive law curve in the rigid body coordinate system represents the yaw angle position in the inertial coordinate system. As can be seen from the diagram, when the system is in states 1 and 2, the designed adaptive law remains within a bounded range and exhibits a regular periodic change, achieving automatic adjustment of the system's position and velocity.

[0185] This embodiment also provides a trajectory tracking and control device for an autonomous underwater robot system, see [link / reference] Figure 10 ,include:

[0186] An autonomous underwater robot control system generation device is used to generate kinematic and dynamic models of autonomous underwater robots, construct an autonomous underwater robot control system containing uncertainties and external disturbances, and obtain the reference trajectory of the tracked target. ;

[0187] The device for generating design variables related to position tracking error and design variables related to velocity tracking error is used to realize the autonomous underwater robot control system combined with the error transformation function to obtain the error system model; based on the error system model, the reference trajectory is combined with the actual output trajectory of the autonomous underwater robot control system to obtain the design variables related to position tracking error, and the velocity vector of the autonomous underwater robot is combined with the control input signal of the kinematic model to obtain the design variables related to velocity tracking error.

[0188] The device for generating the range of the derivative of the first obstacle Lyapunov function is used to construct the first obstacle Lyapunov function based on the design variables related to the position tracking error, and obtain the derivative of the first obstacle Lyapunov function by combining the control input signal of the kinematic model and the position adaptive law. The range of the derivative of the first obstacle Lyapunov function is controlled to not exceed a first threshold, thereby realizing the trajectory and position tracking control of the autonomous underwater robot system.

[0189] The device for generating the derivative range of the second obstacle Lyapunov function is used to construct the second obstacle Lyapunov function based on the first obstacle Lyapunov function and design variables related to the velocity tracking error. By combining the continuous control input signal and the velocity adaptive law, the derivative of the second obstacle Lyapunov function is obtained. The range of the derivative of the second obstacle Lyapunov function is controlled to not exceed a second threshold, thereby realizing the trajectory and velocity tracking control of the autonomous underwater robot system.

[0190] This embodiment provides a trajectory tracking and control device for an autonomous underwater robot system, including a processor and a memory. When the processor executes the computer program stored in the memory, it implements the aforementioned trajectory tracking and control method for an autonomous underwater robot system.

[0191] This embodiment also provides a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the above-described trajectory tracking control method for an autonomous underwater robot system.

[0192] To address the constraints faced by autonomous underwater vehicle (AUV) systems, this invention employs adaptive control theory, a time-triggered control scheme, and radial basis function (RBF) neural network (RBN) technology to effectively solve existing problems. The proposed control scheme offers the following advantages: Based on adaptive theory, it enables online automatic adjustment of the system's controller parameters to adapt to dynamic changes and uncertainties. It can adjust the system's motion and control strategies in real time based on feedback signals to better adapt to the environment and improve control accuracy. The time-triggered control scheme effectively reduces the system's computational load and energy consumption. The event-triggered control scheme sends control signals to the control mechanism, thereby reducing power consumption and control cycle, further lowering the system's operating and maintenance costs, making it more energy-efficient and environmentally friendly. RBN technology can learn the system's dynamic behavior and response, improving its intelligence level. Simultaneously, RBN technology can adaptively adjust the system's parameters and model to adapt to different environments and applications, ensuring stable and efficient system operation, resulting in good tracking performance and high tracking data accuracy.

Claims

1. A trajectory tracking control method for an autonomous underwater robot system, characterized in that, Includes the following steps: S1: Based on the kinematic and dynamic models of the autonomous underwater robot, construct an autonomous underwater robot control system containing uncertainties and external disturbances to obtain the reference trajectory of the tracked target; S2: The autonomous underwater robot control system combines the error transformation function to obtain the error system model; based on the error system model, the reference trajectory is combined with the actual output trajectory of the autonomous underwater robot control system to obtain the design variables related to the position tracking error; the velocity vector of the autonomous underwater robot is combined with the control input signal of the kinematic model to obtain the design variables related to the velocity tracking error. S3: Based on the design variables related to the position tracking error, construct the first obstacle Lyapunov function, combine the control input signal of the kinematic model and the position adaptive law to obtain the derivative of the first obstacle Lyapunov function, control the range of the derivative of the first obstacle Lyapunov function to not exceed the first threshold, and realize the trajectory position tracking control of the autonomous underwater robot system. S4: Based on the first obstacle Lyapunov function and the design variables related to the velocity tracking error, a second obstacle Lyapunov function is constructed. Combining the continuous control input signal and the velocity adaptive law, the derivative of the second obstacle Lyapunov function is obtained. The range of the derivative of the second obstacle Lyapunov function is controlled to not exceed the second threshold, thereby realizing the trajectory velocity tracking control of the autonomous underwater robot system.

2. The trajectory tracking control method for an autonomous underwater robot system according to claim 1, characterized in that, The S2 operation specifically refers to: The position tracking error vector is obtained by subtracting the reference trajectory from the actual output trajectory of the kinematic model in the autonomous underwater robot control system. The position tracking error vector is then multiplied by the error transformation function to obtain the design variables related to the position tracking error. The velocity tracking error vector is obtained by subtracting the control input signal of the kinematic model from the velocity vector of the dynamic model in the autonomous underwater robot control system. The velocity tracking error vector is then multiplied by the error transformation function to obtain the design variables related to the velocity tracking error.

3. The trajectory tracking control method for an autonomous underwater robot system according to claim 1, characterized in that, The reference trajectory is: For reference trajectory, For time.

4. The trajectory tracking control method for an autonomous underwater robot system according to claim 1, characterized in that, The error transformation function is: The error transformation function is... Given positive constants, For time, The stable time set by the user.

5. The trajectory tracking control method for an autonomous underwater robot system according to claim 1, characterized in that, The specific operation of S3 is as follows: Based on the design variables related to the position tracking error, a barrier Lyapunov function containing upper and lower error constraint boundary functions is constructed. Combined with the control gain matrix of the position adaptive law and the neural network weight error of the dynamic model, a first barrier Lyapunov function is constructed. By combining the first barrier Lyapunov function with Young's inequality and taking the derivative, we obtain the derivative of the initial first barrier Lyapunov function. After defining the control input signal and position adaptive law of the kinematic model, they are input into the derivative of the initial first obstacle Lyapunov function to obtain the derivative of the first obstacle Lyapunov function. The range of the derivative of the first obstacle Lyapunov function is controlled to not exceed a first threshold to realize the trajectory and position tracking control of the autonomous underwater robot system.

6. The trajectory tracking control method for an autonomous underwater robot system according to claim 5, characterized in that, The control input signal in the kinematic model is: For the control input signal of the kinematic model, For adjustable normals of the kinematic model, The control gain matrix of the kinematic model. For position adaptive law The control gain matrix, The control gain vector of the kinematic model. Let Lyapunov function vectors contain boundary function barriers with upper and lower error constraints. For neural network weight estimation of kinematic models, For the first calculation function, Error transformation function The derivative, The proportional function vector of the kinematic model. Design variables related to position tracking error; The position adaptive law for: For position adaptive law, for The control gain vector.

7. The trajectory tracking control method for an autonomous underwater robot system according to claim 1, characterized in that, The specific operation of S4 is as follows: The first obstacle Lyapunov function and the design variables related to velocity tracking error are combined with the positive design parameters and the neural network weight estimation of the dynamic model to construct the second obstacle Lyapunov function. By combining the second barrier Lyapunov function with Young's inequality and taking the derivative, we obtain the derivative of the initial second barrier Lyapunov function. After defining the continuous control input signal and the speed adaptive law, they are input into the derivative of the initial second obstacle Lyapunov function to obtain the derivative of the second obstacle Lyapunov function. The range of the derivative of the second obstacle Lyapunov function is controlled to not exceed the second threshold, thereby realizing the trajectory and speed tracking control of the autonomous underwater robot system.

8. The trajectory tracking control method for an autonomous underwater robot system according to claim 7, characterized in that, The continuous control input signal is: For continuous control input signals, This is the event trigger gain constant. For the adjustable constants of the dynamic model, The control gain matrix of the dynamic model. For the speed tracking error vector, The control gain matrix of the speed adaptive law. Error transformation function The derivative, For neural network weight estimation of the dynamic model, The second known function, Design variables related to tracking error, The direction of motion control in the dynamic model is a known constant; The adaptive velocity law is as follows: For the velocity adaptive law, for The control gain vector.

9. The trajectory tracking control method for an autonomous underwater robot system according to claim 1 or 8, characterized in that, The continuous control input signal satisfies , for Continuous control input signal at any time for , , The event triggering mechanism function, for Continuous control input at any time and Actual control input at any time Measurement error, , for Current control input signal , For time, Indicates that the controller has occurred. The moment of the next trigger For the controller to occur the first The moment of the next trigger For the controller to occur the first The second event trigger and occurrence The time interval between events.

10. The trajectory tracking control method for an autonomous underwater robot system according to claim 9, characterized in that, The event triggering mechanism function , This is the event trigger gain constant. This is the event trigger threshold constant.

Images