Unmanned dragging ship path planning method based on improved probability road map method

Abstract

The invention discloses an unmanned dragged ship path planning method based on an improved probability road map method, and relates to the technical field of ship motion control. The method aims at solving the problem that in the marine environment, PRM is prone to being affected by static obstacles in the path generation process, and an effective path cannot be generated easily. The method comprises the following steps: constructing a safety domain of unmanned dragged ships and a target ship in a movement process, and selecting a target dragging point for each unmanned dragged ship based on the safety domain; constructing a planning route map by adopting an improved probability route map method; an A * algorithm is adopted in the planned route map to inquire an expected path from the starting point to the corresponding target dragging point of each unmanned dragging ship; and smoothing the expected path to obtain a final path of the unmanned dragging ship.

Description

Technical Field

[0001] The present invention belongs to the technical field of ship motion control. Background Art

[0002] With the continuous advancement of global marine resource development, offshore towing operations are becoming increasingly frequent. As a key marine engineering equipment, unmanned towing vessels have been widely used in deep-sea operations, channel dredging, and offshore platform construction due to their high efficiency, intelligence, and unmanned characteristics. The research and application of unmanned towing vessels play an important role in improving marine development efficiency and reducing operation risks, and are also an important part of the marine strategic layout. However, when performing towing operations in restricted marine environments, how to plan the path of unmanned towing vessels to achieve efficient, safe, and low-energy-consuming towing tasks remains an urgent problem to be solved.

[0003] To solve this problem, path planning technology has become one of the core research directions of unmanned towing vessel systems. Path planning not only needs to consider the high-efficiency requirements of towing tasks, but also needs to take into account optimizing path length and execution efficiency, as well as various uncertain factors in complex marine environments. For example, the distribution of static obstacles, the existence of dynamic interferences such as wind waves, tides, and ocean currents, and the operation constraints of the vessel itself such as the maximum turning radius, minimum safety distance, and motion speed range also pose strict technical requirements on the adaptability and accuracy of path planning algorithms. In addition, path planning must also ensure the feasibility and smoothness of the path to reduce the maneuvering difficulty and energy consumption of the vessel, and ensure the stability and safety of towing operations.

[0004] In the field of path planning, the traditional probabilistic roadmap method (PRM), as a classic algorithm, has been widely used in various environments in recent years due to its high trajectory planning accuracy. However, in the face of complex marine environments, PRM is prone to be affected by static obstacles during the path generation process, resulting in the generation of invalid paths, increasing the computational burden, and even possibly unable to generate valid paths eventually due to its structural reasons. Summary of the Invention

[0005] The present invention aims to solve the problem that in the marine environment, PRM is prone to be affected by static obstacles during the path generation process, resulting in the generation of invalid paths, increasing the computational burden, and even possibly unable to generate valid paths eventually due to its structural reasons. Now, a path planning method for unmanned towing vessels based on an improved probabilistic roadmap method is provided. This method significantly reduces the generation probability of invalid paths by dynamically adjusting the value range of the probability factor and combining the positions and shapes of static obstacles in the actual marine environment, thereby improving the accuracy and computational efficiency of path planning.

[0006] A path planning method for an unmanned towing vessel based on an improved probabilistic roadmap method, comprising:

[0007] Construct a safety domain for the unmanned towing vessel and the target vessel during movement, and select a target towing point for each unmanned towing vessel based on the safety domain;

[0008] Use the improved probabilistic roadmap method to construct a planning roadmap. In the improved probabilistic roadmap method, the sampling interval of the sampling points is an annular region composed of two circles with different radii centered on the obstacle center. The radii of the two circles are the sum of the radius of the obstacle and the length of one ship of the unmanned towing vessel, and the sum of the radius of the obstacle and the length of two ships of the unmanned towing vessel, respectively;

[0009] Use the A* algorithm in the planning roadmap to query the expected path between each unmanned towing vessel from the starting point to its corresponding target towing point.

[0010] Further, the above path planning method for an unmanned towing vessel based on the improved probabilistic roadmap method further comprises: smoothing the expected path to obtain the final path of the unmanned towing vessel.

[0011] Further, the above construction of the safety domain for the unmanned towing vessel and the target vessel during movement includes:

[0012] The safety domain D safe (E) of the target vessel has the following expression:

[0013]

[0014] where (x E , y E ) is the central coordinate of the target vessel, and (x DE , y DE ) is the coordinate point within the safety domain of the target vessel. L and B are the length and width of the target vessel, respectively;

[0015] The safety domain D safe (P1) of the unmanned towing vessel in front of the target vessel has the following expression:

[0016]

[0017] where (x P1 , y P1 ) is the central position coordinate of the unmanned towing vessel in front of the target vessel, and (x DP1 , y DP1 ) is the coordinate point within the safety domain of the unmanned towing vessel in front of the target vessel. f is the radius of the safety domain of the unmanned towing vessel in front of the target vessel;

[0018] The safety domain D safeThe expression of (P2) is:

[0019]

[0020] where (x P2 , y P2 ) is the central position of the unmanned towing vessel behind the target vessel, (x DP2 , y DP2 ) are the coordinate points within the safety area of the unmanned towing vessel behind the target vessel, and g is the radius of the safety area of the unmanned towing vessel behind the target vessel.

[0021] Furthermore, a target towing point is selected for each unmanned towing vessel based on the safety area, including:

[0022] Setting the selection rule for the target towing point according to the range of the safety area, and selecting the target towing point in the target area according to the selection rule, and using the Hungarian algorithm for target towing point allocation so that each unmanned towing vessel has a corresponding unique target towing point.

[0023] Furthermore, the selection rule for the above target towing point includes:

[0024] When the unmanned towing vessel reaches the corresponding target towing point, the following conditions can be satisfied:

[0025] (1) The number of unmanned towing vessels in front of the target vessel is greater than or equal to the number of unmanned towing vessels behind;

[0026] (2) All the unmanned towing vessels in front of or behind the target vessel are arranged in a row and are symmetrically distributed in an axisymmetric form with the central axis of the target vessel as the axis of symmetry;

[0027] (3) The safety area of each unmanned towing vessel is tangent to the boundary or the extension of the boundary of the safety area of the target vessel;

[0028] (4) The safety areas of adjacent unmanned towing vessels are tangent to each other.

[0029] Furthermore, along the Y E axis of the north-east local coordinate system, the target towing points in a row are sequentially numbered from positive to negative, and j is the number of the target towing point;

[0030] The coordinates of the target towing point G1 corresponding to the unmanned towing vessel P1 in front of the target vessel are expressed as:

[0031]

[0032] where (x E , y E) is the central coordinate of the target ship, L and l are the lengths of the target ship and the unmanned towing ship P1 respectively, R is the total number of target towing points, and q is the number of unmanned towing ships in front of the target ship.

[0033] The coordinates of the target towing point G2 corresponding to the unmanned towing ship P2 behind the target ship Are expressed as:

[0034]

[0035] where p is the number of unmanned towing ships behind the target ship.

[0036] Furthermore, the above-mentioned Hungarian algorithm is used for target towing point allocation, so that each unmanned towing ship has a corresponding unique target towing point, including:

[0037] The expression of the objective function d for allocation is:

[0038]

[0039] where z ij is a decision variable, used to represent whether the target towing point Gi is allocated to the unmanned towing ship Pj. If so, z ij = 1, otherwise z ij = 0, i = 1, 2, …, R, j = 1, 2, …, R, and R is the total number of target towing points;

[0040] d ij is the distance between the target towing point Gi and the center point of the unmanned towing ship Pj.

[0041] The allocation uses an improved probabilistic roadmap method to construct a planning roadmap, including:

[0042] Taking the centers of each obstacle between the target towing point and the unmanned towing ship as the centers of two circles with different radii to form an annular region, and randomly sampling a set of nodes within the annular region to represent the possible positions that the unmanned towing ship may pass through;

[0043] Based on the safety domain, collision detection is performed on each node, and then the nodes that coincide with the obstacles are screened out;

[0044] Connect the remaining nodes according to the preset connection criteria to form a planning roadmap. The connection criteria include: two nodes with a distance greater than the preset threshold are not connected, and two nodes whose connection passes through the obstacle are not connected.

[0045] The allocation uses the A* algorithm to guide the search direction by setting an evaluation function.

[0046] The evaluation function f(n) is:

[0047] f(n) = g(n) + h(n),

[0048] where g(n) is the actual cost from the starting point to node n, and h(n) is the estimated cost from node n to the target dragging point;

[0049] The Euclidean distance is used to measure the movement cost d between two nodes n :

[0050]

[0051] where: (x1, y1) and (x2, y2) represent the coordinates of node n1 and node n2 respectively.

[0052] Allocating and smoothing the desired path to obtain the final path of the unmanned tugboat, including:

[0053] Drawing circles with each inflection point on the desired path as the center, and the circle intersects with the two broken line segments connected to its center to obtain two intersection points. One intersection point is used as the starting point of the Bezier curve, and the other intersection point is used as the end point of the Bezier curve, constructing the Bezier curve to obtain the final path of the unmanned tugboat.

[0054] The advantages of the method for determining the optimal path of the system based on the improved probabilistic roadmap method proposed by the present invention for the path planning task during the movement of the unmanned tugboat system to the target dragging point in a restricted marine environment compared with the prior art:

[0055] (1) The present invention improves the traditional probabilistic roadmap method. By designing the probability factor interval, it reduces the invalid path nodes, thereby improving the path search efficiency, better achieving the purpose of path planning, ensuring the rationality and efficiency of the path planning strategy, and at the same time avoiding the disadvantage that the traditional probabilistic roadmap method may not be able to generate a planned path.

[0056] (2) For the process of the unmanned tugboat system controlling the unmanned tugboat to move to the specified target dragging point, combined with the Hungarian algorithm for point selection, using the path planning of the improved probabilistic graph method and the A* algorithm, and using the Bezier curve for smoothing the planned path, designing the corresponding safe path planning strategy, the obtained planning result not only reduces the total length of the system planned path (i.e., reduces energy consumption) but also improves the efficiency of path planning (i.e., reduces the planning time), so it improves the practicality of path planning and is more conducive to solving practical engineering problems.

[0057] In summary, compared with traditional methods, the present invention can more effectively avoid obstacles, reduce the path length and energy consumption. In addition, through the smoothing process of the Bezier curve, the path better meets the actual operation requirements and reduces the problem of dragging instability that may be caused by the uneven path. Through simulation verification, it can prove the effectiveness and feasibility of the path planning strategy proposed by the present invention in practical applications. The present invention is applicable to the path planning problem of an unmanned towing vessel system moving towards a target towing point in a restricted marine environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is the flowchart of the path planning of an unmanned towing vessel based on the improved probabilistic roadmap method;

[0059] Figure 2 is the schematic diagram of the towing mode model;

[0060] Figure 3 is the path generation diagram after constraining the sampling points;

[0061] Figure 4 is the schematic diagram after using the Bezier curve for smoothing processing. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0062] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention. It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0063] Refer to Figures 1 to 4 to specifically describe this embodiment. A method for path planning of an unmanned towing vessel based on the improved probabilistic roadmap method described in this embodiment:

[0064] First, set the safety domain during the movement of the unmanned towing vessel according to the characteristics of the towing model to ensure that there is enough safety distance between ships during the path planning process; then, select a reasonable position of the target towing point and use the Hungarian algorithm to ensure that each unmanned towing vessel matches a unique target towing point.

[0065] Secondly, in order to improve the path planning efficiency and the ability to generate the optimal path, this embodiment improves the traditional probabilistic roadmap method. The improved probabilistic roadmap method effectively avoids the generation of invalid paths through sampling settings. Subsequently, in combination with the improved probabilistic roadmap method, the A* algorithm is used to search for the path. When the OPEN (open list) is empty or the last node is searched, the search ends. By reasonably designing the heuristic function g(n) and the path cost function h(n), it is ensured that the A* algorithm can always advance along the optimal direction during the search process and generate the global optimal path.

[0066] Finally, in order to make the generated path more in line with the actual requirements, the Bezier curve is used to smooth the optimal path, thereby obtaining a smooth, feasible and efficient planned path. This method improves the path planning efficiency of the system while ensuring path optimization, meeting the actual application requirements in complex environments.

[0067] The specific implementation steps are as follows:

[0068] In the first step, a certain point on the earth's surface is used as the coordinate origin O E to establish a north-east-earth coordinate system X E Y E Z E ; with the center of gravity of the ship as the coordinate origin O B to establish a ship hull coordinate system X B Y B Z B . Based on the coordinate system, a ship kinematics and dynamics model is established:

[0069]

[0070] Among them: represents the actual position (x, y) and attitude angle ψ of the ship in the north-east-earth coordinate system; represents the linear velocity (u, v) and angular velocity r of the ship in the ship hull coordinate system; is the transformation matrix between the two coordinate systems, and satisfies R T (ψ) = R -1 (ψ).

[0071] Set the initial north-east positions of the three unmanned tugboats to be: η1 = [100 65 0] T , υ1 = [0 0 0] T , η2 = [90 75 0] T , υ2 = [0 0 0] T , η3 = [100 75 0] T , υ3 = [0 0 0] T .

[0072] The position information of the target ship being towed is: the central position (x E , y E ) = (300, 295).

[0073] The information of the obstacles are respectively: the center position of obstacle 1 c1 = (162, 172), radius r1 = 20; the center position of obstacle 2 c2 = (198, 231), radius r2 = 30; the center position of obstacle 3 c3 = (260, 188), radius r3 = 15; the center position of obstacle 4 c4 = (284, 232), radius r4 = 10.

[0074] In the second step, considering the safety and collision avoidance issues during the movement and towing process of the unmanned towing ship system, it is necessary to design a corresponding safety area for the ship.

[0075] Set the safety area according to the outer contour size of the target ship E. The safety area of the target ship E is a rectangular area. In this embodiment, assume the length of the target ship L = 60m and the width of the target ship B = 40m, then the safety area of the target ship is a 60×40 rectangular area.

[0076] The safety area D safe (E) is expressed as:

[0077]

[0078] Where: (x E , y E ) is the central coordinate of the target ship, and (x DE , y DE ) is the coordinate point within the safety area of the target ship.

[0079] According to Figure 2 as shown, assume the towing ships in front of and behind the target ship E are P1 and P2 respectively. Set the safety areas of P1 and P2 according to the size of the unmanned towing ship and considering the distance from the target ship. The safety areas of P1 and P2 are circular areas with the center of the ship as the center and 1.5l and 0.5l as the radii respectively. The models of the R towing ships are the same, so the safety areas of the R unmanned towing ships are the same. In this embodiment, the length of the unmanned towing ship l = 1.225m and the width of the unmanned towing ship b = 0.29m.

[0080] The safety areas D safe (P1) and D safe (P2) are respectively expressed as follows:

[0081]

[0082] Where: (x P1 , y P1) is the central position of the unmanned tugboat P1, (x P2 , y P2 ) is the central position of the unmanned tugboat P2. f = 1.5l and g = 0.5l are the safety zone radii of the unmanned tugboats P1 and P2 respectively.

[0083] Thirdly, after establishing the safety zone, select the target tug point position and allocate the target tug point position.

[0084] The target tug point is the final target position of the center of the tugboat. Therefore, the selection of the target tug point is related to the arrangement method of the tugboats. The rules for selecting the target tug point are as follows:

[0085] When the unmanned tugboat reaches the corresponding target tug point, the following conditions can be met:

[0086] (1) The number of unmanned tugboats in front of the target ship is greater than or equal to the number of unmanned tugboats behind;

[0087] (2) All the unmanned tugboats in front of or behind the target ship are arranged in a row and are symmetrically distributed in an axisymmetric form with the central axis of the target ship as the axis of symmetry;

[0088] (3) The safety zone of each unmanned tugboat is tangent to the boundary or the extension line of the boundary of the safety zone of the target ship;

[0089] (4) The safety zones of adjacent unmanned tugboats are tangent to each other.

[0090] According to the tugging method as Figure 2 shown, select R target tug points for tugging operations. The unmanned tugboat P1 is located in front of the target ship, and its corresponding target tug point is G1; the unmanned tugboat P2 is located behind the target ship, and its corresponding target tug point is G2.

[0091] The target tug points corresponding to the tugboats located in front of the target ship satisfy: draw circles with the target tug points as the centers and 1.5 times the ship length as the radii. The adjacent circles are externally tangent to each other and are tangent to the front and rear boundaries of the target safety zone or their extension lines. At the same time, these circles are arranged in an axisymmetric form with the central axis of the target ship as the axis of symmetry to ensure the balance and controllability of the tugging force on the target ship.

[0092] The target tug points corresponding to the tugboats located behind the target ship satisfy: draw circles with the target tug points as the centers and 0.5 times the ship length as the radii. The adjacent circles are externally tangent to each other and are tangent to the front and rear boundaries of the target safety zone or their extension lines. At the same time, these circles are arranged in an axisymmetric form with the central axis of the target ship as the axis of symmetry to ensure the balance and controllability of the tugging force on the target ship.

[0093] Then the position coordinates of the target tug points G1 and G2 can be expressed as:

[0094]

[0095]

[0096] Among them, q and p are the numbers of unmanned tugboats in front of and behind the target ship respectively; the target tug points are sequentially numbered from positive to negative along the Y-axis of the northeast local coordinate system, and j is the number of the target tug point. E After the target tug point is determined, the Hungarian algorithm is used for target tug point allocation to ensure that each unmanned tugboat has a unique corresponding target tug point. Specifically, in order to achieve a reasonable allocation of tug points and unmanned tugboats, an allocation method with the minimization of the total straight-line distance d as the goal can be adopted. This method is optimized based on the Hungarian algorithm, and the straight-line distance d between the center points of the target tug point Gi and the unmanned tugboat Pj

[0097] forms a coefficient matrix. Define the decision variable z ij to represent whether the target tug point Gi is allocated to the unmanned tugboat Pj: if so, z ij = 1, otherwise z ij = 0. ij

[0098] Thus, the following objective function is established:

[0099]

[0100] Fourthly, as one of the classic algorithms for path planning, the Probabilistic Roadmap Method (PRM), its core idea is to randomly sample nodes in the configuration space and establish edges through feasible connections, thereby constructing the motion path of the unmanned tugboat. This method has strong adaptability. The basic steps of the PRM algorithm are as follows:

[0101] Node sampling: Randomly sample a group of nodes in the configuration space, and these nodes represent the possible positions of the unmanned tugboat.

[0102] Collision detection: Based on the safety domain, perform collision detection on each sampled node, and screen out the nodes located inside the obstacle to ensure that the path will not pass through the obstacle.

[0103] Connecting nodes: Connect the remaining sampled points with line segments according to the preset connection criterion. Sampled points with a distance greater than the set threshold are not connected, and sampled points passing through the obstacle are not connected.

[0104] Constructing a graph: After connecting the nodes, a network graph is formed. The nodes in the network graph represent the possible positions of the unmanned tugboat, and the edges represent the feasible paths between two nodes.

[0105] Since the sampling points of the traditional PRM algorithm are randomly selected, the planned path may deviate significantly from the expected path, and there may even be no expected path. Therefore, a sampling interval K is designed to satisfy the following constraints:

[0106] K = {ε|λ 1n <ε<λ 2n}, n = 1, 2, 3,..., N,

[0107] where: ε is the sampling point position, and λ 1n represents the sum of the radius of the nth obstacle and the length of one ship's hull of the tugboat, and λ 2n represents the sum of the radius of the nth obstacle and the length of two ship's hulls of the tugboat, and N is the total number of obstacles.

[0108] The improved PRM algorithm limits the sampling position of the sampling points within an annular region composed of two circles with different radii centered on the obstacle center. Through this design, the generated path is more in line with the actual requirements, improving the rationality and accuracy of path planning. When conducting simulation verification, the sampling point parameter k = 10 (when the line connecting the starting point and the target tugging point passes through n obstacles, the number of sampling points selected is n×k). As Figure 3 shown in the schematic diagram of the effect after setting the probability sampling factor.

[0109] Connect the sampling points within the sampling interval K according to the preset connection criterion to construct a planned roadmap, and then use the A* algorithm to query the expected path from the starting point to the target point. In the constructed roadmap, the A* algorithm guides the search direction by setting an evaluation function, and its evaluation function is defined as:

[0110] f(n) = g(n) + h(n),

[0111] where f(n) represents the evaluation function from the starting point through any node n to the target point, g(n) represents the actual cost from the starting point to node n, and h(n) represents the estimated cost from node n to the target point.

[0112] The Euclidean distance is used to measure the movement cost d n between two nodes:

[0113]

[0114] where: (x1, y1) and (x2, y2) represent the coordinates of nodes n1 and n2 respectively.

[0115] The A* algorithm starts from the starting point to expand the surrounding nodes. By gradually expanding the nodes, it selects the node with the minimum total cost as the next expansion point according to the evaluation function f(n), and repeats the expansion process until the target point is selected, thereby finding the optimal path from the starting point to the target point.

[0116] In the fifth step, the path planned using the PRM algorithm consists of the connections between discrete points, and the path inflection points are presented in the form of broken lines. This path form is not suitable for the driving conditions of an unmanned towboat in practical engineering applications because an unmanned towboat usually sails at sea in a straight line or a curve. To meet the characteristics of an actual unmanned towboat sailing in a straight line or a curve at sea and to improve the applicability of the path, it is necessary to smooth the broken-line path generated by the traditional PRM algorithm.

[0117] For the expected path obtained by the improved probabilistic roadmap method mentioned above, draw circles with each sampling point selected in each PRM algorithm as the center, and the circle intersects the two broken-line segments connected to its center, forming two intersection points. Take the selected sampling point and these two intersection points as control points, one intersection point as the starting point of the Bézier curve, and the other intersection point as the end point of the Bézier curve to construct the Bézier curve, obtaining the smoothed expected path, as Figure 4 shown in the path planning schematic diagram after path smoothing.

[0118] The formula for the mth-order Bézier curve:

[0119]

[0120] where B(t) represents the trajectory of the Bézier curve; t ∈ [0, 1]; P r is the coordinate of the rth control point, r = 0, 1, 2,..., m;

[0121] is the binomial coefficient and has:

[0122] (1 - t) m-r and t r both represent the weight factors of the binomial of the Bézier curve.

[0123] Verified by Matlab simulation, the improved probabilistic roadmap method reduces redundant path nodes by optimizing the sampling interval, improving the feasibility and accuracy of path generation. The time taken to generate a path is significantly shorter compared to that of the traditional probabilistic roadmap method, shortening the running time of path planning and improving the algorithm running efficiency. Therefore, the practicality of path planning is improved, better meeting the needs of actual engineering. At the same time, the generated path is more compact and efficient, with better adaptability in complex environments. In addition, the simulation results further verify the effectiveness and accuracy of the algorithm in practical applications. It can not only ensure the generation and query of the optimal path, but also demonstrates superior performance in avoiding the generation of invalid nodes, reducing the computing cost, and improving the path quality. This research provides an efficient and robust solution for the path planning of unmanned towing vessels, with important engineering application value.

[0124] In summary, in this embodiment, the target towing points are reasonably selected according to the selected towing mode. The Hungarian algorithm is used to optimize the allocation of the target towing points to ensure that each unmanned towing vessel has a unique and optimal towing target. Next, on the basis of the traditional probabilistic roadmap method (PRM), an improved probabilistic factor interval design method is proposed. By dynamically adjusting the value range of the probabilistic factor and combining the distribution characteristics of static obstacles in the actual marine environment, the probability of generating invalid paths is significantly reduced, improving the overall efficiency and accuracy of path planning. On this basis, to further improve the usability and smoothness of the path, the design principle of Bessel curves is introduced to smooth the expected path queried by the A* algorithm. Through this processing, not only the tortuosity and complexity of the path are effectively reduced, but also the applicability and safety of the path in actual engineering are improved.

[0125] Although the present invention has been described in reference to specific embodiments herein, it should be understood that these embodiments are merely examples of the principles and applications of the present invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed, as long as they do not depart from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the different dependent claims and the features described herein can be combined in a manner different from that described in the original claims. It should also be understood that the features described in connection with a single embodiment can be used in other described embodiments.

Claims

1. A path planning method for an unmanned towing vessel based on an improved probabilistic roadmap method, characterized in that: include: Constructing a safety domain for the unmanned towing vessel and the target vessel during the movement, and selecting a target towing point for each unmanned towing vessel based on the safety domain; An improved probabilistic roadmap method is used to construct a planning roadmap. The sampling interval of the sampling points in the improved probabilistic roadmap method is: a circular area is formed by two circles with different radii, with the center of the obstacle as the center. The radii of the two circles are the sum of the radius of the obstacle and one length of the unmanned towing ship, and the sum of the radius of the obstacle and two lengths of the unmanned towing ship. In the planning route map, the A* algorithm is used to query the expected path of each unmanned towing vessel from the starting point to its corresponding target towing point.

2. The unmanned towing vessel path planning method based on the improved probabilistic roadmap method according to claim 1 is characterized in that: Also includes: The expected path is smoothed to obtain the final path of the unmanned tugboat.

3. The unmanned towing vessel path planning method based on the improved probabilistic roadmap method according to claim 1 or 2, characterized in that: The construction of the safety domain of the unmanned towing vessel and the target vessel during the movement process includes: The target ship's safety zone D safe The expression of (E) is: Among them, (x E ,y E ) is the center coordinate of the target ship, (x DE ,y DE ) is the coordinate point within the safety zone of the target ship, L and B are the length and width of the target ship respectively; The safety zone D of the unmanned towing ship located in front of the target ship safe The expression of (P1) is: Among them, (x P1 ,y P1 ) is the center coordinate of the unmanned towing ship in front of the target ship, (x DP1 ,y DP1 ) is the coordinate point in the safety zone of the unmanned towing ship in front of the target ship, and f is the safety zone radius of the unmanned towing ship in front of the target ship; The safety zone D of the unmanned towing ship behind the target ship safe The expression of (P2) is: Among them, (x P2 ,y P2 ) is the center position of the unmanned towing ship behind the target ship, (x DP2 ,y DP2 ) is the coordinate point within the safety zone of the unmanned towing ship behind the target ship, and g is the safety zone radius of the unmanned towing ship behind the target ship.

4. The unmanned towing vessel path planning method based on the improved probabilistic roadmap method according to claim 1 is characterized in that: Selecting a target towing point for each unmanned towing vessel based on the safety domain includes: The target towing point selection rule is set according to the range of the safety zone, and the target towing point is selected in the target area according to the selection rule. The target towing point is allocated using the Hungarian algorithm, so that each unmanned towing ship has a corresponding unique target towing point.

5. The unmanned towing vessel path planning method based on the improved probabilistic roadmap method according to claim 4 is characterized in that: The selection rules of the target drag point include: When the unmanned towing vessel reaches the corresponding target towing point, the following conditions can be met: (1) The number of unmanned towing vessels in front of the target ship is greater than or equal to the number of unmanned towing vessels behind the target ship; (2) All unmanned tugboats in front of or behind the target ship are arranged in a row and distributed in an axisymmetric manner with the central axis of the target ship as the axis of symmetry; (3) The safety zone of each unmanned towing vessel is tangent to the safety zone boundary or the extended line of the boundary of the target vessel; (4) The safety zones of two adjacent unmanned towing vessels are tangent.

6. The unmanned towing vessel path planning method based on improved probabilistic roadmap method according to claim 5 is characterized in that: Y along the north-east coordinate system E The target drag points in a row are numbered sequentially from positive to negative along the axis, and j is the number of the target drag point; The coordinates of the target towing point G1 corresponding to the unmanned towing ship P1 in front of the target ship It is expressed as: Among them, (x E ,y E ) is the center coordinate of the target ship, L and l are the lengths of the target ship and the unmanned towing ship P1 respectively, R is the total number of target towing points, and q is the number of unmanned towing ships located in front of the target ship; The coordinates of the target towing point G2 corresponding to the unmanned towing ship P2 behind the target ship It is expressed as: Where p is the number of unmanned towing vessels behind the target ship.

7. The unmanned towing vessel path planning method based on improved probabilistic roadmap method according to claim 5 is characterized in that: The Hungarian algorithm is used to allocate the target towing points, so that each unmanned towing vessel has a corresponding unique target towing point, including: The expression of the objective function d assigned is: Among them, z ij is a decision variable, which is used to indicate whether the target towing point Gi is assigned to the unmanned towing ship Pj. If yes, then z ij =1, otherwise z ij =0, i=1,2,…,R, j=1,2,…,R, R is the total number of target drag points; d ij It is the distance between the target towing point Gi and the center point of the unmanned towing vessel Pj.

8. The unmanned towing vessel path planning method based on improved probabilistic roadmap method according to claim 1 or 2, characterized in that: The improved probabilistic roadmap method is used to construct a planning roadmap, including: A ring area is formed by two circles with different radii and the center of each obstacle between the target towing point and the unmanned towing ship. A group of nodes is randomly sampled in the ring area to represent the position where the unmanned towing ship may pass. Performing collision detection on each node based on the safety domain, thereby filtering out nodes that overlap with obstacles; The remaining nodes are connected according to the preset connection criteria to form a planning roadmap, and the connection criteria include: Two nodes whose distance is greater than a preset threshold are not connected, and two nodes whose connection line passes through an obstacle are not connected.

9. The unmanned towing vessel path planning method based on improved probabilistic roadmap method according to claim 1 or 2, characterized in that: The A* algorithm guides the search direction by setting an evaluation function. The valuation function f(n) is: f(n)=g(n)+h(n), Among them, g(n) is the actual cost from the starting point to node n, and h(n) is the estimated cost from node n to the target drag point; The Euclidean distance is used to measure the movement cost d between two nodes. n : Among them, (x1, y1) and (x2, y2) represent the coordinates of node n1 and node n2 respectively.

10. The unmanned towing vessel path planning method based on improved probabilistic roadmap method according to claim 2, characterized in that: The step of smoothing the expected path to obtain a final path of the unmanned towing vessel includes: A circle is drawn with each inflection point on the desired path as the center, and the circle intersects with two broken line segments connected to the center of the circle to obtain two intersection points, one of which is used as the starting point of the Bezier curve, and the other intersection point is used as the end point of the Bezier curve to construct a Bezier curve and obtain the final path of the unmanned towing ship.

Images