A method for predicting a dynamic target by a fleet of vessels

Abstract

The application discloses a kind of ship to dynamic target prediction method, comprising the following steps: establishing the trajectory model of target;Determine the ship P i Corresponding target point G i ;Determine the ship P i Corresponding prediction point Set each ship sailing direction.The prediction time of the present application when ship cluster carries out position prediction to dynamic target is not fixed value, but the prediction time of each time ship cluster about target is calculated by the present application, then the corresponding prediction point of each ship is obtained, so that each ship sails to the corresponding prediction point, when each ship reaches the prediction point, it also reaches the real target point, and the ship control is more continuous, and the ship trajectory is smoother.The present application can make the ship complete the task in a shorter time through the position prediction of dynamic target.

Description

Technical Field

[0001] This invention belongs to the field of autonomous control of surface vessels, and particularly relates to a method for predicting dynamic targets by a fleet of vessels. Background Technology

[0002] When ship swarms are performing tasks targeting dynamic targets, they often improve efficiency by predicting the targets to complete complex tasks. To predict dynamic targets, it is often necessary to first determine the prediction time, and the selection of the prediction time has always been a key factor affecting the accuracy of dynamic target prediction. Therefore, optimizing the selection of the dynamic target prediction time is of great significance for ship swarms performing tasks targeting dynamic targets.

[0003] Current ship swarm predictions of dynamic targets are based on fixed prediction times. This means the prediction time is pre-set to a fixed value, and the corresponding predicted target is then derived from that time. This disconnects the target from the predicted target, leading to ships failing to reach the predicted target location when they actually arrive. This results in suboptimal ship swarm trajectories and indicates room for improvement in swarm efficiency. Summary of the Invention

[0004] To address the problems existing in the prediction of dynamic targets by ship swarms in the prior art, this invention provides a method for predicting dynamic targets by ship swarms, which optimizes the prediction time so that the ship swarm arrives at the target location at the same time as the predicted target location.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows: A method for predicting dynamic targets by a ship, comprising the following steps:

[0006] Step A: Establish the trajectory model of the target.

[0007] In the geodetic coordinate system, t is collected k Ship P at time i Position p i (x i ,y i ) and sailing speed v i The position p of target G g (x g ,y g ) and sailing speed v g 0 <i≤n,i∈N + N + Represents a positive integer; the trajectory model of the target is established using the acquired target position and velocity, and the steps are as follows:

[0008] Step A1: Store the time when the target location is captured each time in the variable X, i.e.:

[0009] X = {t1, t2, ..., t} n} (1)

[0010] Step A2: Store the target location collected each time into variable Y, i.e.:

[0011]

[0012] Step A3: Process the variables X and Y using the least squares method, support vector regression, or multinomial fitting to obtain the target trajectory model as follows:

[0013] Y = F(X) (3)

[0014] Step B: Determine the vessel P i The corresponding target point G i

[0015] In t k At time t, for n ships and target G, the formation of the ship group is determined using the virtual structure method, where ship P i The desired position in the corresponding formation is called the target point G. i ,Right now:

[0016] G i =f(G) (4)

[0017] Step C: Determine the vessel P i Corresponding prediction point

[0018] In t k At time t, for n ships and target G, determine ship P. i The corresponding target point G i The position after the prediction time T, i.e., the prediction point The steps are as follows:

[0019] Step C1: Calculate t k Ship P at time i and the corresponding target point G i distance e i ,Right now:

[0020] e i =|P i G i | (5)

[0021] Step C2: Calculate t k Ship P at time i For target point G i Prediction time T i ,Right now:

[0022]

[0023] In the formula, K is the prediction coefficient, 0 <K<1,v g >0; when ship P i Reaching target point G i At the predicted time T i Predicted position points hour Where |P i G i |=e i , Therefore That is, ship P i Reaching target point G i At the predicted time T i Predicted position points The target point G was reached. i Location;

[0024] Step C3: If T i If the value is 0, then the ship P i It has reached its corresponding target point G. i The corresponding prediction point and target point G i Overlap, ship P i Then continue to follow the corresponding target point G i Movement; if T i If it is not equal to 0, then T i The predicted time T of the ship cluster with respect to target G is the minimum value in the predicted time set A of all ships that have not yet reached the corresponding target point, that is:

[0025] T = min{A} (7)

[0026] Step C4: Calculate t k The predicted target G corresponding to target G at time time T ,Right now:

[0027] G T =F(t) k +T) (8)

[0028] Step C5: Calculate t k Ship P that has not yet reached the corresponding target point i Corresponding prediction point Right now:

[0029]

[0030] Step D: Set the navigation direction for each ship

[0031] Based on the calculated predicted points corresponding to each ship, a navigation direction is set for each ship, and each ship travels in this direction until t. k+1 If any ships have not yet reached their corresponding predicted points, return to step A and calculate t according to steps AD. k+1 The system calculates the predicted points for each ship at each time point and provides the sailing direction for each ship until all ships reach their corresponding predicted points.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1) In this invention, the prediction time for the position prediction of a dynamic target by a ship cluster is not a fixed value. Instead, the prediction time of the ship cluster with respect to the target at each moment is calculated by this invention. Then, the prediction point corresponding to each ship is obtained, so that each ship sails to the corresponding prediction point. When each ship reaches the prediction point, it also reaches the actual target point. Ship control is more continuous and the ship trajectory is smoother.

[0034] 2) This invention reduces the time required for ships to complete their missions by predicting the position of dynamic targets. Attached Figure Description

[0035] Figure 1 This is a flowchart of the ship control process of the present invention.

[0036] Figure 2 This is a flowchart illustrating the ship control process according to an embodiment of the present invention. Detailed Implementation

[0037] To more clearly explain the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0038] The following will combine Figure 1-2 The invention will be described in detail with examples:

[0039] Step A: Establish the trajectory model of the target.

[0040] In the geodetic coordinate system, given t k Ship P at time i Position p i (x i ,y i ) and sailing speed v i The position p of target G g (x g ,y g ) and sailing speed v g0 <i≤n,i∈N + N + Represents a positive integer; the trajectory model of the target is established using the collected position and velocity information of the target, and the steps are as follows:

[0041] Step A1: Store the time when the target location is captured each time in the variable X, i.e.:

[0042] X = {t1, t2, ..., t} n} (10)

[0043] Step A2: Store the target location collected each time into variable Y, i.e.:

[0044]

[0045] In the formula

[0046] Step A3: Input the collected x and y values ​​from the target location into the variable Y. x and Y y In, that is

[0047]

[0048]

[0049] Step A4: Solve for the m-th degree polynomial Y x (t), such that at t i The value of is as close as possible to . Right now

[0050]

[0051] Where the coefficients a0, a1, ..., a m Obtained by the least squares method; at t k After a prediction time T, the target's x-coordinate is:

[0052] Step A5: Solve for the m-th degree polynomial Y y (t), such that at t i The value of should be as close as possible to . Right now

[0053]

[0054] Where the coefficients b0, b1, ..., b m Obtained by the least squares method; at t k After a prediction time T, the target's ordinate is:

[0055] Step B: Determine the vessel P i The corresponding target point G i

[0056] In t k At any given time, the ship P is obtained according to the virtual structure method. i (0 <i≤n,i∈N + The target point G corresponding to ) i ,Right now

[0057] G i =(x g +rcos(β i ),y g +rsin(β i (16)

[0058] In the formula (x g ,y g Let r be the position coordinates of the target G, where r (r∈R) + β is the radius of the circular formation of ships surrounding the target. i (β i ∈[0,360]) represents the positive x-axis direction in the geodetic coordinate system and G. i The connection between G and G i The angle of G in the counterclockwise direction;

[0059] Step C: Determine the vessel P i Corresponding prediction point

[0060] In t k At time t, for n ships and target G, determine ship P. i The corresponding target point G i The position after the prediction time T, i.e., the prediction point The steps are as follows:

[0061] Step C1: Calculate t k Ship P at time i and the corresponding target point G i distance e i ,Right now

[0062]

[0063] Step C2: Calculate t k Ship P at time i For target point G i Prediction time T i ,Right now

[0064]

[0065] In the formula, K is the prediction coefficient, 0 <K<1,vg >0; when ship P i Reaching target point G i At the predicted time T i Predicted position points hour Where |P i G i |=e i , Therefore That is, ship P i Reaching target point G i At the predicted time T i Predicted position points The target point G was reached. i Location;

[0066] Step C3: If T i If the value is 0, then the ship P i It has reached its corresponding target point G. i The corresponding prediction point and target point G i Overlap, ship P i Then continue to follow the corresponding target point G i Movement; if T i If it is not equal to 0, then T i The predicted time T of the ship cluster with respect to target G is the minimum value in the predicted time set A of all ships that have not yet reached the corresponding target point, i.e.

[0067] T = min{A} (19)

[0068] Step C4: Calculate t k The predicted target G corresponding to target G at time time T ,Right now

[0069]

[0070] Step C5: Calculate t k Ship P that has not yet reached the corresponding target point i Corresponding prediction point Right now

[0071]

[0072] Step D: Set the navigation direction for each ship

[0073] Based on the predicted points corresponding to each vessel, set the navigation direction for each vessel and sail in this direction until t k+1 time;

[0074] Step D1: Calculate ship P iCorresponding prediction point For P i Gravitational potential field of ship P i Corresponding prediction point The coordinates in the geodetic coordinate system are The gravitational potential field U generated by the predicted point on the ship att for:

[0075]

[0076] In the formula, k att This is the gravitational gain coefficient. For ship P i Location p i Reaching the corresponding prediction point The Euclidean distance between them, in the direction from position p i Pointing to position

[0077] Step D2: Calculate the impact of obstacles on ship P i Repulsive potential field: The coordinates of the obstacle in the geodetic coordinate system are p o =(x o ,y o If the obstacle exerts a repulsive potential field U on the ship, then... rep for:

[0078]

[0079] In the formula, k rep ρ is the repulsive gain coefficient. o A positive value represents the range of influence that the obstacle can have, ρ(p i ,p o ) for ship P i Location p i Reach the location p of the obstacle o The Euclidean distance between them, in the direction from position p o Pointing to position p i ;

[0080] Step D3: Calculate ship P i The overall potential field influenced by external factors: Ship P i Due to the combined effect of the gravitational potential field generated at the prediction point and the repulsive potential field generated by other ships, ship P i Affected by the overall external force field U all for:

[0081] U all =U att +U rep (twenty four)

[0082] Step D4: Calculate ship Pi Resultant force on the ship: P i The negative gradient of the total potential field is the ship's P. i The combined force received Right now:

[0083]

[0084] Step D5: Calculate ship P i heading angle Decompose the resultant force into components along the x-axis and y-axis. and and The ratio is the ship P. i The heading angle at the next moment Right now:

[0085]

[0086] Step D6: Control the ship to move towards the corresponding predicted point: The ship moves at a speed v i along the heading angle Proceed in the direction of t; k+1 If any ship has not yet reached its predicted point, return to step A and calculate t according to steps AD. k+1 The system calculates the predicted points for each ship at each time point and provides the sailing direction for each ship until all ships reach their corresponding predicted points.

[0087] The above description is merely an example of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for predicting dynamic targets by a ship, characterized in that: Includes the following steps: Step A: Establish the trajectory model of the target. In the geodetic coordinate system, t is collected k Ship P at time i Position p i (x i ,y i ) and sailing speed v i The position p of target G g (x g ,y g ) and sailing speed v g 0 <i≤n,i∈N + N + Represents a positive integer; the trajectory model of the target is established using the acquired target position and velocity, and the steps are as follows: Step A1: Store the time when the target location is captured each time in the variable X, i.e.: X={t1,t2,...,t n } (1) Step A2: Store the target location collected each time into variable Y, i.e.: Step A3: Process the variables X and Y using the least squares method, support vector regression, or multinomial fitting to obtain the target trajectory model as follows: Y = F(X) (3) Step B: Determine the vessel P i The corresponding target point G i In t k At time t, for n ships and target G, the formation of the ship group is determined using the virtual structure method, where ship P i The desired position in the corresponding formation is called the target point G. i ,Right now: G i =f(G) (4) Step C: Determine the vessel P i Corresponding prediction point In t k At time t, for n ships and target G, determine ship P. i The corresponding target point G i The position after the prediction time T, i.e., the prediction point The steps are as follows: Step C1: Calculate t k Ship P at time i and the corresponding target point G i distance e i ,Right now: e i =|P i G i | (5) Step C2: Calculate t k Ship P at time i For target point G i Prediction time T i ,Right now: In the formula, K is the prediction coefficient, 0 <K<1,v g >0; when ship P i Reaching target point G i At the predicted time T i Predicted location points hour Where |P i G i |=e i , Therefore That is, ship P i Reaching target point G i At the predicted time T i Predicted location points The target point G was reached. i Location; Step C3: If T i If the value is 0, then the ship P i It has reached its corresponding target point G. i The corresponding prediction point and target point G i Overlap, ship P i Then continue to follow the corresponding target point G i Movement; if T i If it is not equal to 0, then T i The predicted time T of the ship cluster with respect to target G is the minimum value in the predicted time set A of all ships that have not yet reached the corresponding target point, that is: T = min{A} (7) Step C4: Calculate t k The predicted target G corresponding to target G at time time T ,Right now: G T =F(t k +T) (8) Step C5: Calculate t k Ship P that has not yet reached the corresponding target point i Corresponding prediction point Right now: Step D: Set the navigation direction for each ship Based on the calculated predicted points corresponding to each ship, a navigation direction is set for each ship, and each ship travels in this direction until t. k+1 If any ships have not yet reached their corresponding predicted points, return to step A and calculate t according to steps AD. k+1 The system calculates the predicted points for each ship at each time point and provides the sailing direction for each ship until all ships reach their corresponding predicted points.

Images