Flat type full-drive AUV anti-delay energy-saving motion control method and system

Abstract

The application discloses a kind of flat full-drive AUV anti-latency energy-saving motion control method and system.The application includes: constructing the six-degree-of-freedom hydrodynamics model of AUV based on the theory of submarine maneuverability;Through multi-condition test identification, the constant value of the pure lag time of propeller and the speed acceleration characteristic are identified, and the nonlinear propeller dynamics model including the pure lag link and the nonlinear speed dynamic response model is constructed;A composite control architecture integrating double-loop PID, Smith predictor and model reference adaptive controller (MRAC) is constructed;Smith predictor is used to compensate for the explicit time delay of the vertical plane motion subsystem, and MRAC is used to update the adaptive gain online to compensate for the model mismatch.The application can effectively suppress the high-frequency jitter of pitch angle, reduce the invalid commutation action and energy consumption of the propeller, and significantly improve the high-speed operation stability and endurance of the flat AUV.

Description

Technical Field

[0001] This invention belongs to the field of underwater robot motion control technology, and relates to a time-delay-resistant and energy-saving motion control method and system for a flat-type all-drive AUV. Background Technology

[0002] Flat-type fully driven autonomous underwater vehicles (AUVs) are more prone to continuous pitch oscillations during high-speed, constant-depth cruising due to the combined effects of thruster response delay and strong hydrodynamic coupling in the vertical plane, compared to conventional rotating AUVs. This high-frequency jitter not only severely affects the detection accuracy of onboard sensors but also causes the thrusters to perform frequent, ineffective acceleration, deceleration, and reversing maneuvers, leading to increased system energy consumption and shortened operational endurance.

[0003] Existing AUV control technologies have significant limitations: traditional PID control requires reduced gain when dealing with large time delays and is difficult to eliminate high-speed jitter; while existing Smith predictor and model reference adaptive control (MRAC) composite schemes (such as the invention patent application number CN120517395A) are mostly designed for single-degree-of-freedom (SISO) or general industrial processes, ignoring the strong coupling characteristics of vertical thrust and pitch attitude in the six-degree-of-freedom dynamics of all-drive AUVs, and lacking energy consumption optimization mechanisms for high-frequency thruster movements.

[0004] Therefore, considering the physical characteristics of all-wheel-drive flat-type AUVs, developing a composite motion control method that can explicitly compensate for multivariable time delays, correct nominal model mismatch online, and effectively reduce energy consumption has significant application value for improving the high-speed operation performance and energy efficiency of AUVs. Summary of the Invention

[0005] The purpose of this invention is to overcome the problems of severe pitch angle jitter, low control accuracy, and high energy consumption caused by thruster delay during high-speed, constant-depth motion in existing flat-type all-wheel-drive AUVs, and to provide a delay-resistant, energy-saving motion control method and system for flat-type all-wheel-drive AUVs.

[0006] This invention constructs a six-degree-of-freedom dynamic model that includes time delay parameters, uses a six-degree-of-freedom Smith predictor to explicitly compensate for thruster time delay, and combines this with Model Reference Adaptive Control (MRAC) to correct nominal model mismatch online, thereby effectively reducing the ineffective energy consumption of the propulsion system while suppressing pitch jitter.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] In a first aspect, the present invention provides a time-delay-resistant and energy-saving motion control method for a flat-type all-wheel-drive AUV, comprising the following steps:

[0009] Step 1: Construct a six-degree-of-freedom spatial motion equation based on the Newton-Euler rigid body dynamics principle and a six-degree-of-freedom hydrodynamic model of an AUV based on submarine maneuverability theory, as a reference for the subsequent Smith predictor's prediction model with time delay and the nominal model without time delay;

[0010] Step 2: Construct various static mapping models for the thruster; determine the setpoint pure time delay of the thruster control system; construct a piecewise function mapping model of the rotational acceleration with respect to the change in rotational speed and the current rotational direction; and establish a thruster thrust prediction model composed of a pure time delay element and a nonlinear rotational speed dynamic response model connected in series.

[0011] Step 3: Construct the Smith-MRAC composite time-delay controller architecture, which is a composite control architecture of a six-degree-of-freedom dual-loop PID controller, a six-degree-of-freedom Smith predictor, and a model reference adaptive controller;

[0012] Step 4: Simultaneously input the nominal control force / torque output by the six-degree-of-freedom dual-loop PID controller to the actual AUV propulsion system and the six-degree-of-freedom Smith predictor. Calculate the ideal current state using the time-delay-free nominal model and calculate the predicted lag state using the time-delay-inducible prediction model. Subtract the predicted lag state from the actual AUV measurement state and add the ideal current state to obtain the predicted compensation signal, which is then sent back to the six-degree-of-freedom dual-loop PID controller as a feedback value.

[0013] Step 5: Linearize the uncertainty terms of the AUV six-degree-of-freedom nonlinear dynamic equation model, design an adaptive parameter update law, update the parameter vector estimate in real time and generate an adaptive compensation force / torque, and superimpose it with the nominal control force / torque output by the six-degree-of-freedom dual-loop PID controller to form the final control quantity.

[0014] Step Six: Execute control and energy consumption optimization. Distribute the final control quantity to each thruster for execution. By suppressing high-frequency pitch jitter and reducing the thrusters' ineffective frequent commutation and acceleration / deceleration actions, energy-saving control of the AUV can be achieved.

[0015] Secondly, the present invention provides a flat-type all-drive AUV anti-time-delay energy-saving motion control system for implementing the above method, comprising:

[0016] The sensing module is used to collect AUV depth, pitch angle and motion speed information in real time through IMU, depth gauge and Doppler velocimeter;

[0017] The control calculation module has a built-in Smith-MRAC composite anti-delay controller, which is used to calculate the thrust control commands for each thruster;

[0018] The drive execution module includes six thrusters and their drivers arranged on a flat AUV fuselage for responding to the thrust control commands.

[0019] Compared with the prior art, the present invention has the following significant advantages:

[0020] (1) Unlike the traditional single-degree-of-freedom time delay control method, this invention designs a multivariable, multi-degree-of-freedom six-degree-of-freedom Smith predictor for the six-degree-of-freedom coupled model of flat-type all-wheel drive AUV. In particular, it solves the problem of strong coupling oscillation between depth and pitch caused by the time delay of vertical thruster under high-speed motion, fundamentally and effectively suppressing pitch angle oscillation and significantly improving the stability of high-speed constant-depth cruise.

[0021] (2) This invention takes energy consumption optimization as one of the control objectives. By eliminating unnecessary high-frequency jitter caused by time delay, it significantly reduces the frequent acceleration, deceleration, and forward / reverse switching of the thrusters (especially the vertical thrusters responsible for attitude adjustment), thereby reducing the average power of the propulsion system. Experimental data show that, under the same operating conditions, it effectively reduces the standard deviation of the vertical thruster duty cycle and the total energy consumption, and improves the endurance operation capability of the AUV.

[0022] (3) The present invention combines MRAC adaptive control, which effectively makes up for the defect of Smith predictor being sensitive to model parameters. Even when the nominal model is mismatched due to changes in hydrodynamic coefficients caused by changes in speed or external ocean currents, it can still maintain excellent control accuracy and depth holding capability, and reduce steady-state error.

[0023] (4) The method of the present invention is based on the mature ROS framework and has been verified by actual sea trials, proving its effectiveness in actual complex marine environments. It is easy to deploy and promote on existing flat AUV platforms. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This invention provides an overall flowchart of a time-delay-resistant and energy-saving motion control method for a flat-type all-drive AUV.

[0026] Figure 2 : Schematic diagram of the shape and coordinate system definition of the flat all-wheel drive AUV involved in the embodiments of the present invention.

[0027] Figure 3The principle block diagram of the Smith-MRAC composite anti-delay control architecture provided in this embodiment of the invention.

[0028] Figure 4 The following is a comparison curve of the pitch response of the AUV under high-speed constant-depth cruise conditions provided in the embodiments of the present invention.

[0029] Figure 5 : A schematic diagram comparing the duty cycles of vertical thrusters provided in this embodiment of the invention.

[0030] Figure 6 : A schematic diagram comparing the energy consumption of vertical thrusters provided in this embodiment of the invention. Detailed Implementation

[0031] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention.

[0032] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0033] Reference Figure 1 This embodiment proposes a time-delay-resistant and energy-saving motion control method for a flat-type, all-wheel-drive, six-DOF AUV. The AUV platform on which this method is based adopts a flat, streamlined fuselage design and is equipped with six thrusters. The specific layout includes: three vertical thrusters distributed in a "two in front, one in rear" configuration, with their drive shafts perpendicular to the horizontal plane. Two of the vertical thrusters are symmetrically arranged about the central axis at the front of the fuselage (referred to as "left front vertical thruster" and "right front vertical thruster"), and the other vertical thruster is located at the central axis at the rear of the fuselage (referred to as "rear vertical thruster"), which are used together for heave and pitch control; two horizontal main thrusters are symmetrically arranged at the rear of the fuselage (referred to as "left main thruster" and "right main thruster"), with their drive shafts parallel to the horizontal plane, which are used for forward, backward, and nose-turn control; and a lateral thruster is arranged above the geometric center of the fuselage (referred to as "lateral thruster"), which is used for lateral movement control.

[0034] The control method in this embodiment mainly addresses the pitch angle jitter problem caused by thruster response delay and hydrodynamic coupling during high-speed constant-depth cruise of an AUV at 1.5 m / s. The specific steps are as follows:

[0035] Step 1: Construct a six-degree-of-freedom equation of motion based on the principles of Newton-Euler rigid body dynamics, and build a six-degree-of-freedom hydrodynamic model of the AUV using Taylor series expansion based on submarine maneuverability theory. This model will serve as a reference for the subsequent design of the Smith predictor's prediction model with and without time delay, as well as the nominal model. This step aims to establish a mathematical model that accurately describes the dynamic characteristics of a flat-type AUV under high-speed unsteady motion, providing a high-fidelity "nominal model" for the subsequent six-degree-of-freedom Smith predictor.

[0036] In some embodiments, this step specifically includes:

[0037] First, define the inertial coordinate system and the carrier coordinate system to establish the dynamic model of the AUV.

[0038] Specifically, refer to Figure 2 Define the inertial coordinate system and the carrier coordinate system In an inertial coordinate system, Represents the origin. The axis points vertically towards the Earth's center; in the carrier coordinate system, It is the AUV's buoyancy center. The axis points towards the bow. The axis points to the starboard side. The axis is vertically downward.

[0039] The six degrees of freedom motion of an AUV are surge, sway, heave, roll, pitch, and yaw, corresponding to six degrees of freedom position / attitude in an inertial coordinate system. Velocity in a six-degree-of-freedom carrier coordinate system External forces / torques acting on six degrees of freedom The symbols for each parameter are defined as shown in Table 1.

[0040] Table 1. Definitions of Standard Symbols for Motion Parameters

[0041]

[0042] The transformation formula between the rate of change of state of the AUV in the inertial coordinate system and the carrier coordinate system, i.e., the kinematic model of the AUV, is as follows:

[0043]

[0044] Second, establish the six-degree-of-freedom motion equations for the AUV.

[0045] Specifically, establish the motion of the AUV and the external forces / torques it experiences. The relationship, that is, the general form of the equations of motion of an AUV in six degrees of freedom in space, is as follows:

[0046]

[0047] in, Indicates along Directional velocity Size changes Directional acceleration, Indicates along Directional velocity Size changes Directional acceleration, Indicates along Directional velocity The magnitude change produces along Directional acceleration, Indicates circling angular velocity in direction The size change produces the winding Angular acceleration, Indicates circling Directional angular velocity The size change produces the winding Angular acceleration, Indicates circling angular velocity The size change produces the winding Angular acceleration; This indicates the coordinates of the AUV's center of gravity in the carrier coordinate system; , Indicates AUV quality Moment of inertia of the shaft.

[0048] Third, based on submarine maneuverability theory, the six-degree-of-freedom dynamic equations of the AUV are constructed using the Taylor series expansion method.

[0049] Specifically, the flat-type all-wheel-drive AUV involved in this embodiment belongs to the all-drive model, and the forces acting on the AUV do not take into account the rudder angle and propeller. The formulas for the external forces / torques acting on the AUV are as follows:

[0050]

[0051] in The external forces acting on the AUV The static forces and torques acting on the AUV The hydrodynamic forces acting on the AUV The force and torque vectors generated by the AUV propulsion are given by the following formulas:

[0052]

[0053] in For the thrust vector of the AUV propulsion system, This is a thruster configuration matrix determined based on the thruster's installation position (lever arm) and direction.

[0054] Static forces acting on an AUV This includes gravity, buoyancy, and their moments. The static force in the carrier coordinate system is expressed as follows:

[0055]

[0056] Referring to the analysis method of hydrodynamics of unsteady motion of submersible bodies in Shi Shengda's book "Submarine Maneuverability", an overall model is adopted. It is assumed that the maneuvering motion of the AUV is slow. Taking the set speed as the equilibrium point, the hydrodynamics is expanded into a second-order polynomial of motion parameters using the Taylor series expansion method.

[0057] Furthermore, when an AUV undergoes unsteady motion (acceleration or deceleration) in a fluid, it causes the surrounding fluid to move, generating a fluid inertial force proportional to the acceleration, i.e., the added mass effect. According to potential flow theory, for flat AUVs, the smaller coupling terms are neglected, and the added mass coefficient on the diagonal is mainly considered. To facilitate the construction of the dynamic equations, these acceleration-related fluid inertial forces are usually moved to the left-hand side of the equations and combined with the rigid body mass matrix to form the overall inertial matrix of the AUV. Therefore, the subsequent focus is on constructing expressions describing velocity-related viscous hydrodynamic forces (i.e., damping forces).

[0058] Furthermore, considering that flat-fuselage AUVs have high lateral stability, the hydrostatic restoring moment dominates the roll dynamics; and that the flat fuselage has large roll damping, resulting in minimal changes in roll velocity and angle during flight. Therefore, the coupled hydrodynamic terms in the roll direction are ignored in the hydrodynamic modeling, and only the hydrostatic restoring moment in the roll direction is retained. and linear damping term The focus is on the motion coupling between the horizontal and vertical planes.

[0059] Furthermore, considering the asymmetric characteristics of flat AUVs, the six degrees of freedom motion is decomposed into horizontal and vertical planes and constructed separately:

[0060] For vertical plane motion (involving) , , Considering the asymmetry of the flat fuselage both vertically and horizontally, zero lift is retained. With zero lift torque The expression for the viscous hydrodynamics and torque of the term is constructed as follows:

[0061]

[0062] For horizontal plane motion (involving) , , The resistance of a submarine when traveling in a straight line Considering the left-right symmetry of the fuselage, its viscous hydrodynamic and torque expressions are constructed as follows:

[0063]

[0064] In the formula, The hydrodynamic coefficient of an AUV under a specific motion state; , , , This represents the linear damping coefficient that is proportional to the first power of the velocity. , , , , This represents the nonlinear damping coefficient that is proportional to the square of the velocity. , , , , , , , , , , , , , , , , , The coupling hydrodynamic coefficients between motions of different degrees of freedom, for example Describe vertical velocity and pitch angular velocity Coupling of longitudinal force The impact.

[0065] Substituting the aforementioned components, we obtain the force and torque vector generated by the thruster. Expressed as force / torque component in six degrees of freedom The specific six-degree-of-freedom dynamic equations are as follows (the left side represents the inertial term, and the right side represents the resultant external force):

[0066] Surge (longitudinal movement):

[0067]

[0068] Lateral movement (Sway):

[0069]

[0070] Vertical motion (Heave):

[0071]

[0072] Roll motion:

[0073]

[0074] Pitch motion:

[0075]

[0076] Yaw motion:

[0077]

[0078] Fourth, key dynamic parameters are identified through CFD simulation.

[0079] Specifically, the parameters to be identified mainly include: the hydrodynamic coefficients of the AUV under specific motion conditions. ; represents the linear damping coefficient that is proportional to the first power of the velocity. , , , , ; represents the nonlinear damping coefficient that is proportional to the square of the velocity. , , , ; represents the coupling hydrodynamic coefficient between motions of different degrees of freedom. , , , , , , , , , , , , , , , , , Additional quality coefficient .

[0080] Furthermore, considering the large projected area of ​​a flat body moving in a vertical plane, potential flow theory combined with CFD simulation is used to simulate the stable flow field at different angles of attack through CFD angle-of-attack oblique simulation. This allows for the calculation of the linear damping coefficient proportional to the linear velocity and the nonlinear damping coefficient proportional to the square of the velocity in the vertical plane damping matrix (e.g., ...). By using CFD straight-line towing simulation, longitudinal drag data at different speeds were obtained, and the nonlinear damping coefficient of longitudinal motion was calibrated. By using CFD forced oscillation experiments, coefficients related to acceleration, such as longitudinal and vertical added mass, in the inertial matrix were accurately identified.

[0081] Step 2: Construct various static mapping models and thrust prediction models for the thruster.

[0082] The various static mapping models for constructing the thruster include:

[0083] Under constant voltage power supply conditions, the steady-state mapping relationship between the control commands received by the thruster, actual speed, thrust, duty cycle and current is tested and calibrated. Based on this, a static mapping model of thruster energy consumption, a static mapping model of thruster speed-thrust, and a static mapping model of thruster duty cycle-speed are constructed.

[0084] The thruster thrust prediction model includes:

[0085] First, the step response data of the thruster under different speed references were tested under the conditions of same-direction speed change, deceleration and stop, and forward and reverse switching.

[0086] Specifically, a thrust test platform was built to test the thruster's response characteristics under different initial speed states and different amplitude step speed commands for the same-direction speed change, braking and stopping, and reverse switching conditions. The data of the thruster's speed change over time during the test was recorded.

[0087] Furthermore, the same-direction speed change condition is when the current speed and the target speed have the same sign; the braking and stopping condition is when the current speed command jumps to zero speed; and the reverse switching condition is when the current speed and the target speed have opposite signs.

[0088] Second, by using step response data, the setpoint pure time delay, the time difference of the thrust response stage, and the corresponding speed change of the thruster control system are obtained, and a piecewise function mapping model of speed acceleration with respect to speed change and current speed direction is constructed.

[0089] Specifically, it records the time from when the instruction is issued. The moment when the rotational speed begins to change And the moment when the rotational speed reaches the target value and the steady-state region is established. ; Calculate the setpoint pure time delay of the thruster control system This parameter characterizes the sum of the lag time caused by command transmission and motor dead zone, etc.; utilizing the time difference during the thrust response phase and the corresponding change in rotational speed Back-calculate the rotational speed acceleration under various operating conditions And construct rotational acceleration Piecewise function mapping model for the change in rotational speed and the direction of the current rotational speed. It is used to characterize the controlled acceleration characteristics under different operating conditions.

[0090] Third, based on the setpoint pure time delay of the thruster control system, the piecewise function mapping relationship of the speed acceleration with respect to the speed change and the current speed direction, and the thruster speed-thrust static mapping model, a thruster thrust prediction model is established, which consists of a pure time delay element and a nonlinear speed dynamic response model connected in series.

[0091] The thrust prediction model operates as follows: The input target rotational speed command... First, the time elapsed is... The pure time delay results in a delay instruction. Based on the current instantaneous rotational speed of the thruster With delay instructions The difference and sign are used to determine the current operating condition, and the corresponding rotational speed and acceleration are obtained in real time from the piecewise function mapping model. ;by As the slope, calculate the thruster's rotational speed from the current instantaneous speed. Predict the rotational speed at the next moment before the target rotational speed changes; when a new rotational speed command is received, the thruster does not wait for the process of changing from the current instantaneous rotational speed to the next target rotational speed to end, but directly takes the current instantaneous rotational speed as the starting point and re-plans the acceleration according to the new command to achieve an accurate description of the dynamic transition process.

[0092] Step 3: As Figure 3 As shown, a Smith-MRAC composite delay-resistant control architecture is constructed.

[0093] The Smith-MRAC composite anti-delay controller architecture is a composite control architecture consisting of a six-degree-of-freedom dual-loop PID controller, a six-degree-of-freedom Smith predictor, and a model reference adaptive controller (MRAC), used to suppress the impact of time delay on AUV motion control.

[0094] Specifically, a six-degree-of-freedom dual-loop PID controller serves as the main feedback controller, and a six-degree-of-freedom Smith predictor is embedded in parallel within the feedback channel of the six-degree-of-freedom dual-loop PID controller. The MRAC is based on the deviation between the speed in the actual measured state of the AUV and the state of the reference model. An adaptive compensation force / torque is generated in real time. This adaptive compensation force / torque is superimposed on the output of the six-degree-of-freedom dual-loop PID controller to form the final control quantity.

[0095] Furthermore, the six-degree-of-freedom dual-loop PID controller includes an outer-loop six-degree-of-freedom position PID controller and an inner-loop six-degree-of-freedom velocity PID controller, used to generate nominal control torque.

[0096] Specifically, the outer-loop six-degree-of-freedom position PID controller calculates the desired velocity target value in the inertial coordinate system based on the error between the target position / attitude value and the position / attitude feedback value, and outputs it. The six-degree-of-freedom Smith predictor provides it with a time-delay compensated position / attitude feedback signal to eliminate the phase lag caused by the integral element. The velocity target value in the inertial coordinate system is converted into a velocity target value in the carrier coordinate system, which serves as the target value input for the inner-loop six-degree-of-freedom velocity PID controller. The inner-loop six-degree-of-freedom velocity PID controller calculates the nominal control force / torque based on the velocity target value and velocity feedback value in the carrier coordinate system, and outputs it as the six-degree-of-freedom dual-loop PID controller. The six-degree-of-freedom Smith predictor provides it with a time-delay compensated velocity feedback value.

[0097] Furthermore, the six-degree-of-freedom Smith predictor is embedded in the feedback channel of the six-degree-of-freedom dual-loop PID controller in parallel, using the nominal control force / torque output by the six-degree-of-freedom dual-loop PID controller as the input of the six-degree-of-freedom Smith predictor, and providing the six-degree-of-freedom dual-loop PID controller with time-delay compensated position / attitude and velocity feedback values.

[0098] Step 4: Perform explicit time delay compensation based on the multi-degree-of-freedom Smith predictor with six degrees of freedom.

[0099] The six-DOF Smith predictor contains two parallel models: a prediction model with time delay and a prediction model with time delay. No-delay nominal model This includes a time delay prediction model. Using the six-DOF hydrodynamic model and thruster prediction model of the AUV constructed in steps one and two, in the time-delay nominal model upper series pure time delay element Right now No-delay nominal model Combined with the fixed pure time delay obtained in step two In prediction models with time delay Remove pure time delays ,Right now .

[0100] Furthermore, the explicit delay compensation specifically includes:

[0101] The nominal control force / torque output from the six-degree-of-freedom dual-loop PID controller is used as the input to the six-degree-of-freedom Smith predictor; the prediction model with time delay is used. The estimated hysteresis state output after inputting the nominal control force / torque. (Including estimated lag position / attitude and estimated lag velocity), respectively subtract the time-delay nominal model. The ideal current state output after inputting the nominal control force / torque. (Including ideal current position / attitude and ideal current velocity) is the output of the six-degree-of-freedom Smith predictor (including position / attitude prediction error under time delay difference and velocity prediction error under time delay difference).

[0102] Then measure the actual state of the AUV. The output of the six-DOF Smith predictor (including position, attitude, and velocity) is added to the corresponding values ​​(including position / attitude prediction error under time delay difference and velocity prediction error under time delay difference). The final result is the prediction compensation signal. The position / attitude feedback value and the velocity feedback value are fed back to the outer loop six-degree-of-freedom position PID controller and the inner loop six-degree-of-freedom velocity PID controller, respectively.

[0103] The formula for predicting the compensation signal is:

[0104]

[0105] From the perspective of the transfer function, this step will reduce the pure time delay term. By removing the denominator from the characteristic equation of the closed-loop system, the six-degree-of-freedom Smith predictor constructs a "delay-free" virtual feedback channel within the control loop, thereby achieving explicit delay compensation.

[0106] Step 5: Perform parameter adaptation and nominal model mismatch compensation based on MRAC.

[0107] To address the issues of inaccurate hydrodynamic modeling of flat-type AUVs and nominal model mismatch caused by changes in hydrodynamic parameters and environmental disturbances during high-speed motion, this paper linearizes the uncertainty terms of the six-degree-of-freedom nonlinear dynamic equations of the AUV and designs an adaptive parameter update law based on Lyapunov stability theory. The estimated parameter vector is then updated online in real-time using this adaptive update law, generating an adaptive compensation force / torque. Finally, the adaptive compensation force / torque is added to the nominal control force / torque output by the six-degree-of-freedom dual-loop PID controller as the output of the MRAC (Magnetic Reduction Actuator).

[0108] Specifically, because the six-degree-of-freedom Smith predictor is highly dependent on model accuracy, there are deviations between the actual hydrodynamic parameters of the AUV and the nominal model. Therefore, this embodiment linearizes the uncertainty terms of the six-degree-of-freedom nonlinear dynamic equations of the AUV, transforming them into the form of "regression matrix (known information matrix) × unknown parameter vector":

[0109]

[0110] The regression matrix Includes all known combinations of linear and nonlinear state variables (such as...) etc.), parameter vector Includes coefficients that require adaptive adjustment (such as hydrodynamic damping coefficients). wait).

[0111] The predicted hysteresis velocity output by the time-delay prediction model in the six-DOF Smith predictor after inputting nominal control force / torque is selected as the reference model state. The deviation between the velocity in the actual measured state of the AUV and the reference model state is then used. As an adaptive driving signal.

[0112] Assume the nominal control force / torque output by the six-degree-of-freedom dual-loop PID controller is The designed adaptive compensation force / torque is Ignoring higher-order disturbances, the dynamic change of the error can usually be expressed as:

[0113]

[0114] in, The matrix represents the stability of the reference model, which is an ideal closed-loop system consisting of a time-delay prediction model and a six-degree-of-freedom dual-loop PID controller. The control input matrix is ​​the inverse of the AUV inertia matrix. ; The mismatch value of the parameter vector, i.e., the true value of the parameter vector. The nominal model parameter vector identified offline in step one True value of error ; This is the disturbance term caused by inaccurate nominal model parameters. Since it is impossible to know... Therefore, the adaptive compensation force / torque is designed based on the error estimate. To perform calculations, that is, to let Define the parameter estimation error. For parameter vector estimates With error true value Difference:

[0115]

[0116] Will Formula for dynamic change of input error We can obtain:

[0117]

[0118] The scalar function V, which includes the state error energy and the parameter error energy, is constructed as follows:

[0119]

[0120] in It is a positive definite symmetric matrix and satisfies the Lyapunov equation. , It is a positive definite symmetric adaptive gain matrix.

[0121] Taking the time derivative of V, we get as follows:

[0122]

[0123] Error dynamics equation and Lyapunov equations Substituting and rearranging, we get:

[0124]

[0125] To ensure system stability, that is , need to The adaptive update law for parameters can then be solved:

[0126]

[0127] During the control process, MRAC constructs a regression matrix based on the speed in the actual measured state of the AUV, and then performs real-time integration and online updating of the estimated value according to the parameter adaptive update law. Then generate adaptive compensation force / torque. Finally, the adaptive compensation force / torque is added to the nominal control force / torque output of the six-degree-of-freedom dual-loop PID controller as the output of the MRAC. This process is performed online in real time, automatically suppressing the reduction in control performance caused by nominal model mismatch and preventing system instability.

[0128] Step Six: Execution Control and Energy Optimization.

[0129] The final six-degree-of-freedom adaptive control force / torque, after Smith prediction compensation and MRAC correction, is calculated into thrust control commands for the six thrusters through the thrust distribution matrix of the flat AUV, and is then sent out as the output of the control system. By suppressing high-frequency pitch jitter caused by time delay and reducing ineffective commutation of the vertical thrusters, energy-saving control of high-speed constant-depth motion of the AUV is achieved.

[0130] Furthermore, this application also provides a flat-type all-wheel drive AUV anti-delay energy-saving motion control system, including:

[0131] The sensing module is used to collect the depth, pitch angle and motion speed information of the AUV in real time through the onboard IMU, depth gauge and Doppler velocimeter (DVL).

[0132] The control calculation module has a built-in Smith-MRAC composite anti-delay controller, which is used to calculate the thrust control commands of each thruster based on the set target values ​​such as position / attitude and speed.

[0133] The drive execution module includes six thrusters and their drivers arranged on the flat AUV fuselage, for driving the AUV to move in response to thrust control commands from each thruster.

[0134] Reference Figure 3 This paper presents a verification process for a time-delay-resistant and energy-saving motion control method for all-wheel-drive flat-type AUVs based on Smith-MRAC, as one embodiment of the present invention. To verify the beneficial effects of the present invention, a physical experiment was conducted in open shallow sea areas with good sea conditions and relatively stable currents.

[0135] Step 1: Conduct offline parameter identification experiments for the thruster in a water tank environment to obtain dynamic model parameters under multiple operating conditions.

[0136] Specifically, it includes:

[0137] First, a water tank test platform was set up. The thruster under test was disassembled from the AUV body, rigidly connected to the two-dimensional force sensor, and fixed to the water tank test bracket to ensure that the thruster was completely submerged and that there was no obvious boundary interference from the surrounding water. The data acquisition system was connected to ensure that the thrust signal of the force sensor and the speed feedback signal of the thruster could be collected synchronously.

[0138] Secondly, a multi-condition step response test of the thruster was conducted. Under constant voltage power supply, with the thruster stable at different speeds, the host computer sent step speed commands of different amplitudes, focusing on the following three conditions: same-direction speed change condition (current speed and target speed have the same sign), braking and stopping condition (target speed is zero), and reverse switching condition (current speed and target speed have opposite signs).

[0139] In this example, the moment the command was issued was recorded by analyzing the thrust step response curve measured by the force sensor. The timing of thrust or rotational speed generation And the moment when the thrust or rotational speed reaches the target value and the steady-state region is established. Through calculation, the pure constant time delay parameter reflecting the dead zone characteristics was identified. ; Utilizing the time difference during the thrust response phase and the corresponding change in rotational speed Back-calculate the rotational speed acceleration under various operating conditions And construct rotational acceleration Piecewise function mapping model for the change in rotational speed and the direction of the current rotational speed. .

[0140] Calculate the time span of the dynamic response phase and the corresponding change in rotational speed Using formulas The average rotational acceleration under this operating condition was calculated in reverse; finally, the data from all operating conditions were summarized, and the rotational acceleration was obtained by fitting the data using the least squares method. Regarding the change in rotational speed and the piecewise function mapping model in the current rotational speed direction .

[0141] Based on the above identification, the fixed pure time delay is obtained. By combining the piecewise acceleration function model, a nonlinear dynamic model was constructed that can accurately describe the "pure lag-slope acceleration" characteristics of the thruster.

[0142] Step 2: Conduct high-speed, constant-depth straight-line navigation experiments to compare the effects of different control algorithms.

[0143] Specifically, it includes:

[0144] The target depth was set at 2 meters, the forward velocity at 1.5 m / s, and the travel distance at 200 meters. A synchronous acquisition and storage system was used to ensure alignment of motion status and energy consumption data timestamps, and a current recorder with a sampling rate of 10 Hz was used to accurately measure system power consumption. Tests were conducted under the following two control modes:

[0145] Original PID control mode (benchmark test): using only the original six-degree-of-freedom dual-closed-loop PID algorithm to complete the constant-depth straight-line cruise task;

[0146] Smith-MRAC composite control mode (method of this invention): A six-degree-of-freedom Smith predictor is added to the PID controller. The predictor model is initialized using the state-dependent time delay parameters and hydrodynamic parameters identified in step one. The MRAC adaptive law is activated, and the adaptive parameters are adjusted online to compensate for model mismatch.

[0147] Step 3: Experimental data analysis and effect verification.

[0148] The collected data were processed, and experimental data from the AUV's constant-depth straight-line navigation were selected to calculate the steady-state depth error, pitch standard deviation, vertical thruster duty cycle standard deviation, and energy consumption per unit distance (Wh / m).

[0149] In this example, the experimental results show that:

[0150] Vibration suppression effect: Under the original PID control, when the desired forward speed is 1.5 m / s, the AUV exhibits significant pitch angle jitter. (Refer to...) Figure 4 After adopting the Smith-MRAC composite control of this invention, the standard deviation of the pitch angle is reduced by 25% compared with the original PID, effectively suppressing the pitch angle oscillation during high-speed cruising.

[0151] Control accuracy: When the desired depth is 2m, the steady-state error of depth maintenance is reduced from -0.29m to -0.22m, a reduction of 24%, and the depth fluctuation range is controlled within ±12cm.

[0152] Energy saving effect: Refer to Figure 5 By suppressing high-frequency pitch oscillations, the ineffective movements of the vertical thrusters were significantly reduced, and the standard deviations of the duty cycles of the three vertical thrusters were reduced by 33%, 32%, and 27% respectively compared to the original PID control. (Refer to...) Figure 6 Energy consumption data for each thruster can be obtained through current recorder data. After adopting the method of the present invention, the energy consumption of the vertical thruster is significantly reduced. The energy consumption per unit distance (Wh / m) of the thruster using the method of the present invention is reduced by about 17% compared with the original PID control, which verifies the energy-saving effectiveness of the algorithm.

[0153] This invention also provides an embodiment that differs from the previous embodiments in that: if the function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0154] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects.

[0155] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A time-delay-resistant and energy-saving motion control method for a flat-type all-drive AUV, characterized in that, Includes the following steps: Step 1: Construct a six-degree-of-freedom spatial motion equation based on the Newton-Euler rigid body dynamics principle and a six-degree-of-freedom hydrodynamic model of an AUV based on submarine maneuverability theory, as a reference for the subsequent Smith predictor's prediction model with time delay and the nominal model without time delay; Step 2: Construct various static mapping models of the thruster, determine the setpoint pure time delay of the thruster control system; construct a piecewise function mapping model of the rotational acceleration with respect to the change in rotational speed and the current rotational direction, and establish a thruster thrust prediction model composed of a pure time delay element and a nonlinear rotational speed dynamic response model connected in series. Step 3: Construct the Smith-MRAC composite time-delay controller architecture, which is a composite control architecture of a six-degree-of-freedom dual-loop PID controller, a six-degree-of-freedom Smith predictor, and a model reference adaptive controller; Step 4: Simultaneously input the nominal control force / torque output by the six-degree-of-freedom dual-loop PID controller to the actual AUV propulsion system and the six-degree-of-freedom Smith predictor. Calculate the ideal current state using the time-delay-free nominal model and calculate the predicted lag state using the time-delay-inducible prediction model. Subtract the predicted lag state from the actual AUV measurement state and add the ideal current state to obtain the predicted compensation signal, which is then sent back to the six-degree-of-freedom dual-loop PID controller as a feedback value. Step 5: Linearize the uncertainty terms of the AUV six-degree-of-freedom nonlinear dynamic equation model, design an adaptive parameter update law, update the parameter vector estimate in real time and generate an adaptive compensation force / torque, and superimpose it with the nominal control force / torque output by the six-degree-of-freedom dual-loop PID controller to form the final control quantity. Step Six: Execute control and energy consumption optimization. Distribute the final control quantity to each thruster for execution. By suppressing high-frequency pitch jitter and reducing the thrusters' ineffective frequent commutation and acceleration / deceleration actions, energy-saving control of the AUV can be achieved.

2. The method according to claim 1, characterized in that, The construction of the six-degree-of-freedom hydrodynamic model of the AUV mentioned in step one includes: Define the inertial coordinate system and the carrier coordinate system, and establish the kinematic model of the AUV; Establish the six-degree-of-freedom motion equations for the AUV in space; Based on submarine maneuverability theory, the six-degree-of-freedom dynamic equations of an AUV are constructed using Taylor series expansion. Key dynamic parameters were identified through CFD simulation.

3. The method according to claim 1, characterized in that, Step two involves constructing various static mapping models for the thruster, including: Under constant voltage power supply conditions, the steady-state mapping relationship between the control commands received by the thruster and its actual speed, thrust, duty cycle and current is tested and calibrated, and a static mapping model of thruster energy consumption, thruster speed-thrust, and thruster duty cycle-speed is constructed.

4. The method according to claim 1 or 3, characterized in that, Step two, which involves determining the setpoint pure time delay of the thruster control system, includes: The step response data of the thruster under different speed references were tested under the conditions of same-direction speed change, deceleration and stop, and forward and reverse switching. Record the time from the moment the command is issued to the moment the speed begins to change, and calculate the pure time delay of the setpoint.

5. The method according to claim 1, characterized in that, In step three, the six-degree-of-freedom Smith predictor is embedded in the feedback channel of the six-degree-of-freedom dual-loop PID controller in parallel.

6. The method according to claim 5, characterized in that, The six-degree-of-freedom dual-loop PID controller mentioned in step three includes an outer-loop six-degree-of-freedom position PID controller and an inner-loop six-degree-of-freedom velocity PID controller; The outer ring six-degree-of-freedom position PID controller calculates the desired inertial coordinate system velocity target value based on the error between the position / attitude target value and the feedback value; The inner-loop six-degree-of-freedom velocity PID controller calculates the nominal control force / torque based on the error between the target velocity value and the feedback value in the carrier coordinate system.

7. The method according to claim 1, characterized in that, The formula for calculating the estimated compensation signal in step four is as follows: ； in This represents the actual measurement status of the AUV. This represents the estimated lag state output by the time-delay prediction model. This represents the ideal current state output by the no-delay nominal model.

8. The method according to claim 7, characterized in that, The linear parameterization form described in step five is the nonlinear dynamic term = ,in For the regression matrix, This is a vector of unknown parameters including the hydrodynamic damping coefficient; The adaptive update law for the parameters is: ； in It is a positive definite symmetric adaptive gain matrix. To control the input matrix, To satisfy the positive definite symmetric matrix of the Lyapunov equation. This represents the deviation between the actual measured speed and the state of the reference model.

9. The method according to claim 1, characterized in that, The allocation described in step six is ​​achieved through the thrust allocation matrix of the flat-type AUV; the flat-type AUV is equipped with six thrusters, including three vertical thrusters distributed in a front-two-rear-one configuration, two horizontal main thrusters symmetrically arranged at the tail, and one lateral thruster arranged above the geometric center.

10. A flat, all-wheel-drive AUV anti-time-delay energy-saving motion control system, characterized in that, To implement the method according to any one of claims 1-9, comprising: The sensing module is used to collect AUV's depth, pitch angle, and motion speed information in real time through IMU, depth gauge, and Doppler velocimeter; The control calculation module has a built-in Smith-MRAC composite anti-delay controller, which is used to calculate the thrust control commands for each thruster; The drive execution module includes six thrusters and their drivers arranged on a flat AUV fuselage for responding to the thrust control commands.

Images