CN-121995959-A - Navigation channel ice condition detection-oriented ship unmanned aerial vehicle game guidance and control method

CN121995959ACN 121995959 ACN121995959 ACN 121995959ACN-121995959-A

Abstract

The invention discloses a ship unmanned aerial vehicle game guidance and control method for channel ice detection, which relates to the technical field of ship/unmanned aerial vehicle motion control research and comprises the steps of establishing a ship three-degree-of-freedom and unmanned aerial vehicle six-degree-of-freedom nonlinear model, designing a ship expected task point according to an actual channel point, generating an unmanned aerial vehicle task point according to an unmanned aerial vehicle detection radius and a water area width, constructing a local optimal motion strategy between a ship/unmanned aerial vehicle and a virtual body based on a Stark-berg game by taking the task point as expected coordinates, solving a global optimal motion strategy according to the local optimal motion strategy, designing a position controller, fusing the ship local optimal strategy with a real-time position of the unmanned aerial vehicle, and enabling the two-side gestures to converge by a total controller formed by a gesture virtual controller and the gesture controller, and driving the ship and the unmanned aerial vehicle to execute an optimal strategy by the synergistic effect of the position and the gesture controller to realize cooperative control.

Inventors

ZHANG GUOQING
YANG XUE
YIN SHILIN
Xing Yingshuo
ZHU MINGXU
XIAO HEYE
LI JIQIANG
ZHANG XIANKU
ZHANG WENJUN
WANG TINGYUE

Assignees

大连海事大学
西北工业大学

Dates

Publication Date: 20260508
Application Date: 20260410

Claims (8)

1. A ship unmanned aerial vehicle game guidance and control method for channel ice detection is characterized by comprising the following steps: S1, constructing a ship three-degree-of-freedom model and an unmanned aerial vehicle six-degree-of-freedom nonlinear mathematical model by adopting an Euler-Lagrange equation, and taking the model as a subsequent control object; S2, designing an ideal ship task point based on an actual ship route point, and generating an unmanned aerial vehicle task point according to the relative relation between the unmanned aerial vehicle detection radius and the real-time water area width; S3, respectively constructing local optimal motion strategies of the ship/unmanned aerial vehicle and the virtual ship/unmanned aerial vehicle based on Stark game by taking the ship task point and the unmanned aerial vehicle task point as expected coordinates, and acquiring the optimal motion strategies of the ship/unmanned aerial vehicle according to the local optimal motion strategies of the ship/unmanned aerial vehicle and the virtual ship/unmanned aerial vehicle so as to enable the ship/unmanned aerial vehicle to reach the expected coordinates; S4, constructing a ship/unmanned aerial vehicle position controller according to an optimal motion strategy, so that the actual positions of the ship and the unmanned aerial vehicle are converged to corresponding task point coordinates; And S5, constructing a ship/unmanned aerial vehicle attitude master controller based on a local optimal motion strategy of the ship and an actual position of the unmanned aerial vehicle so as to enable the ship/unmanned aerial vehicle attitude to converge to the requirement of the optimal motion strategy, wherein the ship/unmanned aerial vehicle attitude master controller comprises an attitude virtual controller and a ship/unmanned aerial vehicle attitude controller, and controlling the ship and the unmanned aerial vehicle to execute the optimal motion strategy according to the ship/unmanned aerial vehicle position controller and the ship/unmanned aerial vehicle attitude master controller so as to realize cooperative control of the ship/unmanned aerial vehicle.
2. The ship unmanned aerial vehicle game guidance and control method for channel ice detection according to claim 1, wherein the ship three-degree-of-freedom model and the unmanned aerial vehicle six-degree-of-freedom nonlinear mathematical model are constructed by using an Euler-Lagrange equation, and the method comprises the following steps: The ship three-degree-of-freedom model and the unmanned aerial vehicle six-degree-of-freedom nonlinear mathematical model are as follows: Wherein, the And Respectively, ship and the first The state matrix of each unmanned plane, wherein matrix elements respectively represent the ship in a fixed coordinate system A shaft(s), On-axis coordinates and unmanned aerial vehicle on A shaft(s), Shaft and method for producing the same Coordinates on the axis; the attitude matrix is used for respectively representing a ship heading angle, a transverse inclination angle, a pitch angle and a heading angle of the unmanned aerial vehicle; Is a linear velocity matrix, and respectively represents the advancing speed of the ship and the unmanned plane edge A shaft(s), Shaft and method for producing the same The speed of the shaft; representing the transverse floating speed of the ship; the ship bow speed and the roll, pitch and yaw speeds of the unmanned aerial vehicle are respectively represented by an angular speed matrix; is a quality matrix, wherein, For the mass of the vessel in the forward direction, Is the quality of the unmanned aerial vehicle; the mass of the ship in the bow-swing direction and the unmanned plane edge are respectively represented as a rotational inertia matrix A shaft(s), A shaft(s), The moment of inertia of the shaft; the mass of the ship in the horizontal drifting direction; for the position control input matrix, respectively representing the ship advancing control input and the unmanned plane A shaft(s), A shaft(s), A control input on the shaft; the system is a gesture control input matrix, and respectively represents ship bow control input and unmanned aerial vehicle roll, pitch and bow control input; Is a gravity acceleration matrix; for the position interference matrix, respectively representing the interference suffered by the ship in the advancing direction and the unmanned aerial vehicle in the advancing direction A shaft(s), A shaft(s), Interference on the shaft; The attitude interference matrix is used for respectively representing the interference of the ship in the yaw direction and the interference of the unmanned aerial vehicle in the roll, pitch and yaw directions; the nonlinear matrix is a nonlinear matrix of positions, and the nonlinear parameter of the ship in the advancing direction and the unmanned plane in the advancing direction are respectively represented A shaft(s), A shaft(s), System non-linear parameters on the shaft; the system nonlinear parameters of the ship in the bow and yaw directions and the system nonlinear parameters of the unmanned aerial vehicle in the roll, pitch and yaw directions are respectively represented as the attitude nonlinear parameters; The nonlinear parameter of the ship in the transverse drift direction is the system nonlinear parameter of the ship; Wherein, the Respectively show that unmanned plane is in A shaft(s), A shaft(s), Nonlinear parameters in axial, heel, toe, and yaw directions; the rotational inertia and the rotational speed of the unmanned aerial vehicle rotor are respectively; The parameters are respectively the first order, the second order and the third order of the nonlinear system of the ship advancing direction; The parameters are respectively the first order, the second order and the third order of the nonlinear system of the ship transverse drift direction; The parameters are respectively the first order, the second order and the third order of the ship bow-swing direction nonlinear system.
3. The method for game guidance and control of a ship unmanned aerial vehicle for channel ice detection according to claim 2, wherein designing ideal ship mission points based on actual ship channel points comprises: Definition of the first embodiment The navigation points and the coordinates thereof are The previous and next waypoints and coordinates adjacent thereto are expressed as And , ; Defining a first straight navigation section, a second straight navigation section and a steering navigation section among three continuous navigation points, and defining task points respectively positioned on the two straight navigation sections and two task points positioned on the steering navigation section; If the current navigation section is a straight navigation section, acquiring ship task points of the first straight navigation section and the second straight navigation section based on the navigation points, wherein the expression of the ship task points is as follows: Wherein, the Respectively as the task points of the ship A shaft(s), Defining the axial coordinate, heading angle, advancing speed and heading speed Is a state matrix of ship task points, and the initial state of the ship task points is set as follows ; Defining when the ship mission point is located in the first straight navigation section And The method comprises the following steps: defining when the ship mission point is located in the second straight navigation section And The method comprises the following steps: ; Wherein, the And The heading of the first straight navigation section and the heading of the second straight navigation section of the straight navigation path are respectively; Case2, if the current leg is in the steering leg, defining The method comprises the following steps: Wherein, the For the heading difference between the second straight navigation section and the first straight navigation section, ; For turning radius, the expression is as follows: Wherein, the And Respectively a maximum steering radius and a minimum steering radius; based on the steering radius, the ship task point and the way point of the straight navigation section, the expression for constructing the steering navigation section task point is as follows: Wherein, the Is that And (3) with The coordinates of the task points of the inter-turn leg, Is that And (3) with And (3) the task point coordinates of the steering leg.
4. The ship unmanned aerial vehicle game guidance and control method for channel ice detection according to claim 3, wherein generating unmanned aerial vehicle task points according to the relative relation between the unmanned aerial vehicle detection radius and the real-time water area width comprises: step one, judging the width of the navigable area Is of the size of (2): When unmanned aerial vehicles at two sides of the ship detect non-navigable areas, namely floating ice areas, the width of the navigable areas The values of (2) have the following: case 1 when In the time-course of which the first and second contact surfaces, the distribution formula of the unmanned aerial vehicle task points is designed as follows: Wherein, the Form reorganization parameters for a formation; case 2 when In the time-course of which the first and second contact surfaces, the distribution formula of the unmanned aerial vehicle task points is designed as follows: Wherein, the In order to assist in the calculation of the parameters, ; Case 3 when In the time-course of which the first and second contact surfaces, the distribution formula of the unmanned aerial vehicle task points is designed as follows: Wherein, the ; Case 4 when In the time-course of which the first and second contact surfaces, the distribution formula of the unmanned aerial vehicle task points is designed as follows: Wherein, the ; Case 5 when In the time-course of which the first and second contact surfaces, the distribution formula of the unmanned aerial vehicle task points is designed as follows: When unmanned aerial vehicles on two sides of a ship can not detect an area which can not be navigated, judging that the numerical value of the unmanned aerial vehicle formation detection range has the following conditions: Case 1 when the unmanned aerial vehicle formation detection range is When it will In the process, the unmanned aerial vehicle task points are distributed in a formula Replaced by ; Case 2 when the unmanned aerial vehicle formation detection range is When it will In the process, the unmanned aerial vehicle task points are distributed in a formula Replaced by , Replaced by ; Case 3 when the unmanned aerial vehicle formation detection range is When it will In the process, the unmanned aerial vehicle task points are distributed in a formula Replaced by ; Case 4 when the unmanned aerial vehicle formation detection range is When it will In the process, the unmanned aerial vehicle task points are distributed in a formula Replaced by ; When only one side of the ship is provided with the unmanned aerial vehicle to detect the non-navigable area, namely, the unmanned aerial vehicle cannot acquire the width of the navigable area, the unmanned aerial vehicle formation keeps the current formation until the unmanned aerial vehicles at the two sides of the ship detect the non-navigable area or cannot detect the non-navigable area, and the distribution calculation of the unmanned aerial vehicle task points is carried out; step two, defining a mathematical expression of the unmanned aerial vehicle task points based on the distribution of the unmanned aerial vehicle task points, wherein the mathematical expression is as follows: Wherein, the For unmanned aerial vehicle task point state matrix, respectively representing unmanned aerial vehicle task points A shaft(s), A shaft(s), An axis coordinate; representing a rotation matrix; And representing the relative position of the unmanned plane task point from the ship task point, and forming a structural matrix.
5. The method for guidance and control of a ship unmanned aerial vehicle game for channel ice detection according to claim 4, wherein the ship task point and the unmanned aerial vehicle task point are used as desired coordinates, and local optimal motion strategies of the ship/unmanned aerial vehicle and the virtual ship/unmanned aerial vehicle are respectively constructed based on stark-bird game, and the method comprises the following steps: s31, constructing a local optimal motion strategy of the ship/unmanned aerial vehicle based on the Stark game: the following layer performance indexes of the ship/unmanned plane and the virtual ship/machine stark game are constructed as follows: Wherein, the Representing the performance index of the following layer, The time is represented by the time period of the time, As an error between the ship/drone and the virtual ship/drone, For a ship/drone state matrix, In the form of a virtual state matrix, Respectively adopting motion strategies of a ship/unmanned aerial vehicle and a virtual ship/machine; are positive symmetric matrixes respectively representing the ship/unmanned aerial vehicle and the virtual ship/machine The weight occupied in (b); based on differential game principle, build the game The minimum ship/unmanned aerial vehicle local optimal motion strategy is: Wherein, the Is that Is set to be a minimum value of (c), Is that Is a function of the estimated value of (2); is a locally optimal motion strategy for a ship/unmanned aerial vehicle, wherein, The local optimal motion strategies of the ship and the unmanned aerial vehicle are respectively; Is that Wherein, Respectively is Is a function of the estimated value of (2); the weight and the activation function of the radial basis function neural network are respectively; Respectively is Critic and actor estimates; Controlling a parameter matrix for kinematic control of the ship/unmanned aerial vehicle; s32, constructing a local optimal motion strategy of the virtual ship/machine based on the Stark game: the performance indexes of the pilot layer for designing the ship/unmanned plane and the virtual ship/machine stark game are as follows: Wherein, the Representing the performance index of the leading layer; Errors from the virtual ship/machine to the service point; are positive symmetric matrixes respectively representing the virtual ship/machine and the ship/unmanned aerial vehicle The weight occupied in (b); representing a desired formation queue, i.e. a desired coordinate of the ship/unmanned aerial vehicle, for the task matrix; based on differential game principle, build the game The minimum virtual ship/machine local optimum motion strategy is: Wherein, the Is that Is set to be a minimum value of (c), Then is Is a function of the estimated value of (2); For a locally optimal motion strategy of the virtual ship/machine, An estimated value thereof; the weight and the activation function of the radial basis function neural network are respectively; Respectively is Critic and actor estimates; and controlling the parameter matrix for the kinematic control of the virtual ship/machine.
6. The method for game guidance and control of a ship unmanned aerial vehicle for channel ice detection according to claim 5, wherein the obtaining the optimal motion strategy of the ship/unmanned aerial vehicle according to the local optimal motion strategies of the ship/unmanned aerial vehicle and the virtual ship/unmanned aerial vehicle comprises: combining the local optimal motion strategies of the ship/unmanned aerial vehicle and the virtual ship/unmanned aerial vehicle to obtain an estimated value of the local optimal motion strategy of the ship/unmanned aerial vehicle, namely, the optimal motion strategy of the ship/unmanned aerial vehicle is as follows: Constructing an update law of the neural network weight estimation value so that Respectively converge to ; Respectively converge to Make the following Respectively converge to The update law is as follows: Wherein, the Is a matrix of units which is a matrix of units, Are all positive constants, respectively represent Is updated at a higher rate than the update rate of (a); Is a positive constant and represents a minimum update law.
7. The method for guidance and control of a ship unmanned aerial vehicle game for channel ice detection of claim 6, wherein constructing a ship/unmanned aerial vehicle position controller according to an optimal motion strategy comprises: s41, defining an expression of position virtual control through a first-order filter as follows: Wherein, the Is that By the time constant of the first order low pass filter, ; Is that A filtered signal; S42, defining an expression of a position dynamic error and a derivative thereof as follows: s43, constructing a ship/unmanned aerial vehicle position controller according to the position dynamic error and the derivative thereof, wherein the ship/unmanned aerial vehicle position controller comprises the following components: Wherein, the For the matrix of the position control parameters, Weights and activation functions, respectively, of a location radial basis neural network for approximating a location nonlinear matrix ; Is that Is a function of the estimated value of (2); For position interference matrix Is shown below: Wherein, the Observing parameters for position interference; is a position interference auxiliary parameter; The update law of (2) is: in the formula, Diagonal matrices, all positive, respectively representing And its minimum update rate.
8. The method for game guidance and control of a ship unmanned aerial vehicle for channel ice detection according to claim 7, wherein constructing a ship/unmanned aerial vehicle attitude master controller based on a local optimal motion strategy of the ship and an actual position of the unmanned aerial vehicle comprises: S51, local optimal motion strategy based on ship And drone position control input Calculating the reference attitude of the ship/unmanned aerial vehicle as follows: Wherein, the The reference attitude matrix of the ship/unmanned aerial vehicle respectively represents a reference heading angle of the ship and a reference roll, pitch and heading angle of the unmanned aerial vehicle, Is a preset value; Respectively represent a matrix A first row element and a second row element of (a); Representing the lift of the unmanned aerial vehicle; s52, defining a ship/unmanned aerial vehicle attitude kinematic error as follows: The construction of the attitude virtual controller based on the attitude kinematic error of the ship/unmanned aerial vehicle is as follows: Wherein, the Is a gesture virtual controller; Is a positive symmetric matrix and represents a gesture virtual control parameter matrix; S53, filtering the gesture virtual controller through a first-order low-pass filter to obtain ; In the formula, Is that By the time constant of the first order low pass filter, Is that A filtered signal; Defining the attitude kinetic error as: The construction of the ship/unmanned aerial vehicle attitude controller based on the attitude dynamic error is as follows: in the formula, Is a gesture control parameter matrix; weights and activation functions of the pose radial basis function neural network, respectively, for approximating a positional nonlinear matrix ; Is that Is a function of the estimated value of (2); for an attitude disturbance matrix As shown below, In the formula, Observing parameters for attitude disturbance; The attitude disturbance auxiliary parameter is; In addition, in the case of the optical fiber, The update law of (c) can be expressed as In the formula, Diagonal matrices, all positive, respectively representing And its minimum update rate.

Description

Navigation channel ice condition detection-oriented ship unmanned aerial vehicle game guidance and control method Technical Field The invention relates to the technical field of ship/unmanned aerial vehicle motion control research, in particular to a ship unmanned aerial vehicle game guidance and control method for channel ice detection. Background Traditional ice detection and navigability analysis rely on polar orbit satellites, and the obtained ice information is generally not time-efficient. The ship/unmanned plane cooperative system fully plays the air advantage, and the unmanned plane can continuously monitor the ice conditions of the ship in the forward direction and the surrounding water area in real time. How to reorganize the formation of unmanned aerial vehicle realizes the high-efficient coverage to the waters around the boats and ships, avoids the detection area overlapping and to the redundant monitoring of the area of can not sailing ice, is the key problem that promotes unmanned aerial vehicle ice condition detection efficiency. In the field of multi-agent formation system control, the core for realizing formation type recombination and change is to change formation structures, common design ideas are distributed control based on affine transformation ideas, artificial potential field method based on task allocation, leading-following control based on a preset formation library, and the like. Although the existing method can realize the recombination and the change of the formation of the cooperative body, the problems of dependence on manual adjustment, preset time, preset formation, unstable movement of the cooperative body in the recombination process and the like exist. So that the existing formation reorganization method does not meet the actual engineering requirements of complexity and variability. In the control engineering, the controller of the vessel/drone is usually directly designed from the error of the vessel/drone to the reference path. However, when the initial point of the ship/unmanned aerial vehicle deviates far from the reference path, problems of slow error convergence speed and saturated overload of the actuator occur. Although the conventional method commonly uses an artificial potential field method for planning a smooth track of a reference path for a ship/unmanned aerial vehicle in real time, the artificial potential field method does not consider factors such as control input, operability and inertia of the ship/unmanned aerial vehicle, and the generated smooth track is not beneficial to the ship/unmanned aerial vehicle to realize high-precision tracking, so that the artificial potential field method is difficult to apply to engineering practice. Based on the analysis, the existing ship/unmanned aerial vehicle formation recombination control algorithm mainly has the following 2 point defects: 1. The existing formation reorganization scheme is excessively dependent on manual and preset adjustment of a formation structure, and lacks autonomous and adaptive adjustment of a navigation environment, so that an unmanned aerial vehicle cannot reorganize formations dynamically according to an ice environment, and more efficient detection coverage is realized. And because the ship/unmanned aerial vehicle is required to change the sailing speed in the formation reorganization process, instability of the formation system can be caused. 2. Existing ship/drone control methods typically directly utilize the error of the ship/drone to the reference path to design the controller. When the deviation of the initial position of the ship/unmanned aerial vehicle from the reference path is large, the problems of slow convergence speed, overshoot, actuator saturation and the like are easily caused. Disclosure of Invention The invention provides a ship unmanned aerial vehicle game guidance and control method for channel ice detection, which aims to overcome the technical problems. In order to achieve the above object, the technical scheme of the present invention is as follows: A ship unmanned aerial vehicle game guidance and control method for channel ice detection comprises the following steps: S1, constructing a ship three-degree-of-freedom model and an unmanned aerial vehicle six-degree-of-freedom nonlinear mathematical model by adopting an Euler-Lagrange equation, and taking the model as a subsequent control object; S2, designing an ideal ship task point based on an actual ship route point, and generating an unmanned aerial vehicle task point according to the relative relation between the unmanned aerial vehicle detection radius and the real-time water area width; S3, respectively constructing local optimal motion strategies of the ship/unmanned aerial vehicle and the virtual ship/unmanned aerial vehicle based on Stark game by taking the ship task point and the unmanned aerial vehicle task point as expected coordinates, and acquiring the optimal motion strategies of the ship/unmanned