CN-122009420-A - Spherical unmanned ship anti-rolling method and spherical unmanned ship

CN122009420ACN 122009420 ACN122009420 ACN 122009420ACN-122009420-A

Abstract

The application discloses a spherical unmanned ship stabilizing method and a spherical unmanned ship, which relate to the technical field of ships, the application adopts the cooperative control of double-shaft Active Mass Damping (AMD) and a vector propeller, the method is used for carrying out distribution treatment on low-frequency and high-frequency wave disturbance, and the anti-rolling effect of the spherical Unmanned Ship (USV) under complex sea conditions is obviously improved. Through a clear frequency separation mechanism, the application can efficiently treat low-frequency large-amplitude wave disturbance by the active mass damper, and high-frequency instantaneous disturbance by the propeller, thereby realizing effective complementary control of wave disturbance and remarkably improving the anti-rolling efficiency. The dual-mode cooperative control method not only overcomes the limitation of the traditional single control mechanism, but also remarkably improves the overall stability and the anti-rolling effect of the system.

Inventors

JI MINGYAO
JI JUNWEI
LI GUOQUAN
LIU YE
YIN ZHI
HE HONGKUN
HUANG XINYI
Sun Pinyuan
Guo Guyuhang

Assignees

江苏海洋大学
连云港沐航科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260403

Claims (10)

1. A method of stabilizing a spherical unmanned boat, comprising: acquiring the attitude and environmental information of the spherical unmanned ship by utilizing various sensors; Predicting future wave disturbance by utilizing a long-short-period memory network according to the gesture and the environmental information of the spherical unmanned ship to form a wave prediction disturbance moment; decomposing the wave prediction disturbance moment into a low-frequency disturbance component and a high-frequency disturbance component; And controlling the low-frequency disturbance component to be processed by active mass damping, and controlling the high-frequency disturbance component to be processed by a propeller.
2. The method of claim 1, wherein the acquiring pose and environmental information of the unmanned spherical vessel using a plurality of sensors comprises: An inertial measurement unit is adopted as a main attitude sensor, and the triaxial angular velocity and the triaxial acceleration of the gyroscope in the spherical unmanned ship are collected; And estimating the posture of the spherical unmanned ship from a measurement sequence containing noise by an extended Kalman filtering mode according to the triaxial angular speed and the triaxial acceleration.
3. The method of claim 2, wherein estimating the pose of the unmanned spherical vessel from the measurement sequence containing noise by means of extended kalman filtering based on the three-axis angular velocity and the three-axis acceleration comprises: Setting the state vector of the spherical unmanned ship as ,C(q)=[q 0 ,q 1 ,q 2 ,q 3 ], Wherein C (q) is a gesture rotation matrix of the spherical unmanned aerial vehicle, q 0 、q 1 、q 2 、q 3 represents a rotation gesture vector of the spherical unmanned aerial vehicle in a space coordinate system in four directions of a plane of the spherical unmanned aerial vehicle, b g represents zero bias of a gyroscope in the spherical unmanned aerial vehicle, the zero bias of the gyroscope is used for correcting measurement errors of the gyroscope, Zero offset of the gyroscope, which is positioned in the directions of an x axis, a y axis and a z axis in a space coordinate system, of the gyroscope in the spherical unmanned ship is represented, and T is a transpose; the method comprises the following steps of: Wherein q takes the value of q 0 、q 1 、q 2 or q 3 , As a time derivative of q, Omega m is the angular velocity measurement value of the gyroscope, omega m is the triaxial angular velocity omega x 、ω y or omega z ,n g and is the gyroscope noise; the three-axis acceleration is utilized to restrict the gravity direction, and an observation residual error is constructed The method is used for estimating the attitude error, wherein a m is an acceleration measurement value of a gyroscope, a m is triaxial acceleration a x 、a y or a z , and g is a gravity vector; Estimating the attitude of the spherical unmanned ship according to the time-dependent change model of the attitude of the spherical unmanned ship and the observation residual error.
4. The method for stabilizing the unmanned spherical vessel according to claim 1, wherein the predicting future wave disturbances to form a predicted disturbance moment by using a long-short-term memory network according to the attitude and the environmental information of the unmanned spherical vessel comprises: constructing a wave prediction model based on a long-term and short-term memory network; taking the attitude deviation, the angular velocity and the wave moment estimated value at the historical moment as historical actual sea state data, and performing supervised learning on the wave prediction model by utilizing the historical actual sea state data and high sea state wave data generated by computational fluid dynamics simulation to obtain a trained wave prediction model; And predicting future wave disturbance according to the trained wave prediction model and the attitude and environmental information of the spherical unmanned ship to form a wave prediction disturbance moment.
5. The method of claim 1, wherein said decomposing the wave predicted disturbance moment into a low frequency disturbance component and a high frequency disturbance component comprises: Setting a low-pass path and a high-pass path by adopting a second-order Butterworth filter, wherein transfer functions corresponding to the low-pass path and the high-pass path are respectively as follows: Wherein H LP (s) is the transfer function of the low-pass filter, H HP (s) is the transfer function of the high-pass filter, omega c is the cut-off frequency, omega c represents the demarcation point of the low-pass filter and the high-pass filter, and s is the complex variable of the Laplace transform; And carrying out frequency domain decomposition on the real-time attitude deviation and the wave prediction disturbance moment according to the low-pass path and the high-pass path to obtain a low-frequency disturbance component and a high-frequency disturbance component.
6. The method of claim 5, wherein said controlling the low frequency disturbance component to be handled by active mass damping and the high frequency disturbance component to be handled by a propeller comprises: Determining the attitude error after low-pass filtering and the attitude error after high-pass filtering based on the attitude error according to the low-pass filtering operation and the high-pass filtering operation respectively; acquiring an attitude error after low-pass filtering Acquiring an attitude error after high-pass filtering Wherein θ error (t) is an attitude error representing a difference between an actual attitude and a desired attitude of the spherical unmanned boat, [ ] Representing the operation of the low-pass filtering, [ ] Representing a high pass filtering operation; Taking the gesture error after low-pass filtering and the gesture error after high-pass filtering as input references of an active mass damper and a propeller, and optimizing control inputs of the active mass damper and the propeller for cooperatively controlling the spherical unmanned ship in a preset time period in the future; And dynamically adjusting the control weights of the active mass damping and the propeller according to the energy distribution of the low-frequency disturbance component and the high-frequency disturbance component, and cooperatively controlling the spherical unmanned ship according to the active mass damping, the control input of the propeller and the control weights.
7. The method of claim 6, wherein optimizing the control inputs for the active mass damping and the propeller to cooperatively control the unmanned spherical vessel for a predetermined time period comprises: dynamically constructing a prediction model based on rigid body dynamics and an actuating mechanism of the spherical unmanned ship in a model prediction control mode, wherein the prediction model is Wherein, eta= [ phi, theta, phi ] T is Euler angle vector and represents the roll angle phi, pitch angle theta and bow angle phi of the spherical unmanned ship; the first derivative of Euler angle is used for representing the angular speed; The second derivative of Euler angle is used for representing angular acceleration, M (eta) is an inertial matrix used for describing inertial characteristics of the spherical unmanned ship, C (eta, ) Is a coriolis force matrix for describing coriolis force generated by the movement of the spherical unmanned ship, D (eta, ) The system comprises a damping matrix, a control moment, a thrust force and a wave prediction disturbance moment, wherein the damping matrix is used for describing the damping force suffered by the movement of the spherical unmanned ship, g (eta) is a recovery moment and used for describing the recovery moment generated after the spherical unmanned ship is disturbed, tau AMD is a control moment generated by active mass damping, tau thr is the thrust force generated by a propeller, and tau wave is the wave prediction disturbance moment; obtaining mechanism constraint conditions of the spherical unmanned ship, wherein the mechanism constraint conditions comprise a mass block stroke of active mass damping, a propeller thrust, a thrust change rate and a state constraint, wherein the mass block stroke of the active mass damping is x AMD ∈[−x max , x max ], y AMD ∈[−y max , y max , the propeller thrust is F i ∈[0,F max , the thrust change rate is |delta F i ∣≤ΔF max , the state constraint is eta E H, and H is a safety gesture range; Solving an optimization problem in each control period according to the mechanism constraint conditions Wherein u t:t+N-1 is a control input sequence comprising an active mass damping position and a propeller thrust of the future N steps, eta t+k is a predicted Euler angle of the future moment k, eta ref is a desired Euler angle, Q is a weight matrix of an attitude error for adjusting the accuracy of attitude control, R is a weight matrix of the control input for limiting the energy of the control input, S is a weight matrix of the change rate of the control input for limiting the smoothness of the control input; and determining the control input of the driving mass damping and the propeller for cooperatively controlling the spherical unmanned ship in a preset time period in the future according to the output result of solving the optimization problem.
8. The method of claim 6, wherein dynamically adjusting the control weights of the active mass damping and the propeller according to the energy distribution of the low frequency disturbance component and the high frequency disturbance component comprises: acquiring short-time energy E LP of the low-frequency disturbance component and short-time energy E HP of the high-frequency disturbance component according to the energy distribution of the low-frequency disturbance component and the high-frequency disturbance component; By passing through And adjusting the control weight lambda AMD of the active mass damping and the control weight lambda thr of the propeller.
9. The spherical unmanned ship is characterized by comprising a base and an aircraft anti-roll device, wherein the aircraft anti-roll device comprises: A central fixed platform (1) fixed to the base; The propeller (2) comprises a Y-direction propeller (21) and an X-direction propeller (22), and the propeller (2) is used for driving the spherical unmanned ship to move towards the Y direction and the X direction; an active mass damper (3) mounted on the lower surface of the central fixed platform (1) for damping vibrations; wherein the active mass damping (3) comprises: an X-direction movement mechanism (31) mounted on the lower surface of the central fixed platform (1); the Y-direction movement mechanism (33) is slidably arranged on the lower surface of the X-direction movement mechanism (31), the Y-direction movement mechanism (33) is arranged in parallel with the central fixed platform (1), and the Y-direction movement mechanism (33) is perpendicular to the extending direction of the X-direction movement mechanism (31); an X-direction motor (32) mounted on the central fixed platform (1) and connected to the X-direction movement mechanism (31); a Y-direction motor (34) mounted on the central fixed platform (1) and connected to the Y-direction movement mechanism (33); And a mass ball (35) slidably mounted on the Y-direction movement mechanism (33).
10. The unmanned spherical vessel of claim 9, further comprising a controller electrically connected to the propeller and the active mass damped X-direction motor (22) and Y-direction motor (24), the controller comprising: A memory for storing a computer program; A processor for implementing the steps of the spherical unmanned aerial vehicle roll-reduction method according to any one of claims 1 to 8 when executing the computer program.

Description

Spherical unmanned ship anti-rolling method and spherical unmanned ship Technical Field The application relates to the technical field of ships, in particular to a spherical unmanned ship stabilizing method and a spherical unmanned ship. Background Existing Unmanned Surface Vehicles (USV) often face severe challenges in high-sea environments when performing high-precision tasks such as ocean monitoring, exploration or emergency rescue. Under the sea condition of 4-5 levels, the wave height is obvious, wave energy is concentrated, the traditional single or double-hull ship is easy to generate large-amplitude rolling, pitching and heaving motions due to insufficient wave resistance, the gesture stability is obviously reduced, and the precision requirements of sea monitoring, exploration or rescue tasks are difficult to meet. The spherical USV has a unique full-symmetrical appearance structure and omni-directional maneuvering capability, and has great potential for coping with complex wave environments. The appearance has no fixed part of the bow and the stern, can reduce the directional impact of waves, has better wave compliance and recovery characteristics theoretically, and is regarded as an important development direction of a high sea condition adaptive platform. However, the key technology of the current spherical USV is still immature, and the practical application of the current spherical USV still has obvious limitations, and the prior art has the following limitations: The existing spherical USV is characterized by being controlled by a single propeller moment, being good at responding to high-frequency disturbance components rapidly, and being stable in posture through moment adjustment, but being very poor in propeller moment when facing to continuous and large-offset disturbance generated by low-frequency large waves (long period and large energy), and being difficult to provide enough and efficient compensation moment, the existing spherical USV is caused to generate accumulated posture deviation and even instability. Few studies have attempted to introduce uniaxial active mass dampers (ACTIVE MASS DAMPER abbreviated AMD) that create anti-tipping moments by movement of an internal mass. But is limited by the internal space, the stroke of the mass block is limited, the generated correction moment is smaller, only partial disturbance under lower sea conditions can be dealt with, instantaneous disturbance cannot be considered, and the control efficiency is limited as a whole. The prior art lacks clear wave frequency separation mechanism, lacks effective identification and separation of low-frequency and high-frequency components with obvious difference in wave frequency spectrum, mixes low-frequency and high-frequency disturbance component treatment, has low control efficiency and insufficient cooperativity, and severely limits the potential exertion of an actuating mechanism and the overall stability of a system. The prediction and control limitation is that the existing control is dependent on a simple feedback algorithm (such as PID), the wave disturbance which is nonlinear, time-varying and has strong correlation lacks effective advanced prediction and learning ability, the control lag is obvious, the high sea condition performance is limited, and the precise and feedforward type gesture stability is difficult to realize under the high sea condition. Disclosure of Invention The application provides a spherical unmanned ship stabilizing method and a spherical unmanned ship, which are used for solving the problems of insufficient stability and poor stabilizing effect of the existing spherical unmanned ship under the condition of complex wave sea, and improving the attitude stability of the spherical unmanned ship under the condition of high sea, so that the spherical unmanned ship can reliably execute high-precision tasks such as sea monitoring, exploration or rescue. The application provides a spherical unmanned ship stabilizing method, which comprises the following steps: acquiring the attitude and environmental information of the spherical unmanned ship by utilizing various sensors; Predicting future wave disturbance by utilizing a long-short-period memory network according to the gesture and the environmental information of the spherical unmanned ship to form a wave prediction disturbance moment; decomposing the wave prediction disturbance moment into a low-frequency disturbance component and a high-frequency disturbance component; controlling the low frequency disturbance component to be processed by Active Mass Damping (AMD) and controlling the high frequency disturbance component to be processed by a propeller. The application also provides a spherical unmanned ship, which comprises a base and an aircraft anti-roll device, wherein the aircraft anti-roll device comprises: The central fixed platform is fixed on the base; The propeller comprises a Y-direction propeller and an X-direction prop