A Leader-Based Collaborative Path Planning Method for Ground-Based Multi-Unmanned Systems

Abstract

This invention provides a collaborative path planning method for ground-based multi-unmanned systems in leader mode. It addresses two scenarios: turning and straight-line movement of the leader node. For turning, a consistency deviation index parameter is introduced to adjust the position of unmanned nodes, implementing a "shortcut" behavior to keep up with the leader node as much as possible while maintaining the basic formation. For straight-line movement, since the leader node and unmanned nodes have the same speed, the algorithm only fine-tunes the angle to reduce formation distance deviation. The autonomous path planning algorithm proposed in this invention, which allows unmanned nodes to follow manned nodes in weak network environments, can adaptively adjust the formation by changing its own direction of travel while maintaining the formation to the greatest extent possible. The planned path is smooth and can meet the special requirements of turning during unmanned system movement.

Description

Technical Field

[0001] This invention belongs to the field of unmanned system formation control technology, and particularly relates to a method for collaborative path planning of ground-based multi-unmanned system formations under leader mode. Background Technology

[0002] Compared to manned systems, unmanned systems possess advantages such as being unaffected by human casualties, high durability, and flexible operation. In recent years, with the rapid development of information and control technologies, unmanned systems have gained widespread attention worldwide and are playing an increasingly important role in fields such as military, transportation, remote sensing, and disaster prevention. Currently, human intervention is still required during the execution of tasks by unmanned systems, and for some complex tasks, operators need to intervene remotely via commands. At the same time, with the rapid development of artificial intelligence technology, the autonomous decision-making capabilities of unmanned systems have been greatly enhanced. Further reducing human involvement in unmanned system operations and enabling unmanned systems to collaborate with humans at a higher level of autonomy to complete pre-defined tasks has become a hot research topic in many fields.

[0003] There are two main scenarios for collaboration among unmanned systems: one is autonomous collaboration between unmanned systems, where unmanned devices achieve autonomous collaborative behavior according to pre-defined rules; the other is collaboration between unmanned systems and manned systems, with manned control equipment at the core, using its behavior as input data, and unmanned devices utilizing their limited intelligence to achieve collaboration around the manned equipment. Relatively speaking, manned / unmanned collaboration is currently more technically feasible. For manned / unmanned collaboration scenarios, a leader-following pattern can be introduced, where there is a leader node (manned), and followers (unmanned) respond accordingly using the leader's action data as input commands, thus completing collaboration between the two types of nodes.

[0004] Path planning for multiple unmanned systems is a key technical problem in the field of autonomous control of unmanned systems. It primarily addresses finding the optimal path from the starting point to the target point for each unmanned system within a specific mission context. Path design must simultaneously satisfy the physical constraints of the unmanned systems themselves and the external environment, avoiding obstacles or threats, and preventing collisions between different unmanned systems. Researching path planning under a leader mode is of significant importance for exploring the collaborative behavior of unmanned systems around manned systems.

[0005] However, compared to aerial unmanned systems, ground-based unmanned systems have a lower spatial dimension in path planning, which can be regarded as a path planning problem in a two-dimensional plane space, but they are more constrained by the environment and are more affected by environmental factors.

[0006] Many researchers have made valuable explorations into the path planning problem in the cooperative process of multi-unmanned systems under the Leader-Follower model. Kabalan A et al. studied the time-constrained control problem of multi-agent systems in directed communication graph networks using quantum communication under the Leader-Follower model, proposing a control scheme with predefined time and demonstrating its effectiveness in formation cooperative control through simulation. Liu Z et al. studied discrete-time formation control methods for linear multi-agent systems, assuming a directed spanning tree between nodes, and proposed a formation control protocol, based on which a multi-step control algorithm using the Lyapunov method and Riccati equations was adopted. Rehman FU et al. proposed a Leader-Follower model formation cooperative control strategy for the formation cooperative requirements of underwater unmanned vehicles and implemented it on an autonomous underwater robot. Cw A et al. proposed a Leader-Follower formation tracking control method for spacecraft formation cooperative requirements, achieving collision avoidance and connectivity maintenance through the development of a logarithmic function-based method. Their results mainly target unmanned systems in the aerospace field, providing an effective solution for the cooperative problem of set tracks. The above findings studied the collaborative problem of multi-agent systems under the Leader-Follower model and made adaptive designs for the characteristics of Agent objects in different domains. However, they lacked consideration for unstable terrestrial network environments and also lacked consideration for formation control capabilities. Summary of the Invention

[0007] To address the aforementioned issues, this invention provides a collaborative path planning method for ground-based multi-unmanned systems in leader mode. This method can adaptively adjust the formation by changing its own direction of travel while maintaining the formation to the greatest extent possible.

[0008] A collaborative path planning method for ground-based multi-unmanned system formations under a leader mode is proposed. The formation contains one manned leader node, with the rest being unmanned nodes. The path planning method for each unmanned node is as follows:

[0009] The leader node's direction of travel and position are captured at a fixed detection frequency, and its current heading angle α is determined. s With the heading angle α at the next moment s+1 If they are the same, and not the same, introduce the consistency deviation index corresponding to the current unmanned node at the current time t. Adjust the current unmanned node's direction of travel. If the directions are the same, adjust the current unmanned node's direction of travel based on its speed.

[0010] Furthermore, the consistency deviation index The calculation formula is as follows:

[0011]

[0012] Where i is the label of the currently unmanned node, (x o ,y o (x) represents the target position of the current unmanned node under ideal conditions at time t. s ,y s ) represents the actual position of the currently unmanned node at time t, D[(x o ,y o ),(x s ,y s )] i It means (x) o ,y o ) and (x s ,y s The distance between ), where α and μ are set weights, and β ij The actual position (x) of the current unmanned node i s ,y s The angle between the line connecting the actual position of any node j and the direction of travel of node j, Let x be the ideal position of the current unmanned node i. o ,y o The angle d between the line connecting the actual position of any node j and the direction of travel of node j. ij The actual position (x) of the current unmanned node i s ,y s The Euclidean distance between ) and any node j in the formation. Let x be the ideal position of the current unmanned node i. o ,y o ) is the Euclidean distance between the node and any node j in the formation, where m is the total number of nodes in the formation.

[0013] Furthermore, the consistency deviation index γ corresponding to the current unmanned node at the current time t is introduced. i The 't' parameter is used to adjust the current unmanned node's direction of travel as follows:

[0014] Calculate the consistency index con:

[0015]

[0016] Where Λ is a set constant, r o The current unmanned node moves along the Dubins path to the target position (x). o ,y o The radius of curvature at the end of the curve, α o+1Point the current unmanned node to the target location (x) o ,y o The direction of travel. This is the consistency deviation index for the currently unmanned node at the current time t.

[0017] Calculate the actual position (x′) of the current unmanned node at the next moment based on the consistency index con. s+1 ,y′ s+1 ):

[0018] x′ s+1 =x o+1 +con*sin(α o+1 )

[0019] y′ s+1 =y o+1 +con*cos(α o+1 )

[0020] Among them, (x o+1 ,y o+1 The ideal position of the current unmanned node at the next moment is calculated by introducing formation constraints by capturing the position of the leader node at the next moment.

[0021] Assuming the unmanned node is moving at a constant speed, calculate the distance from the current unmanned node's current actual position (x). s ,y s Move to the actual position (x′) s+1 ,y′ s+1 The reachable point (x) within the required time period. s+1 ,y s+1 );

[0022] Obtain the planned heading angle α′ of the current unmanned node at the next moment. s+1 :

[0023]

[0024] Where Δx is the current actual position (x s ,y s ) and reachable point (x) s+1 ,y s+1 The difference in the x-coordinates between the two points, where Δy is the current actual position (x, y). s ,y s ) and reachable point (x) s+1 ,y s+1 The difference in the ordinates between )

[0025] Based on the planned heading angle α′ s+1 and the leader node's current heading angle α sGet the current unmanned node from its current actual position (x) s ,y s ) to reachable point (x) s+1 ,y s+1 Dubins path between ) (s,s+1) :

[0026] path (s,s+1) =Dubins([x s ,y s ],[x s+1 ,y s+1 ]).

[0027] Furthermore, adjusting the current unmanned node's direction of travel based on its current speed specifically involves:

[0028] Calculate the actual position (x′) of the currently unmanned node in the next time step. s+1 ,y′ s+1 ):

[0029]

[0030]

[0031] Where v is the current velocity of the unmanned node; D is the current target position (x) of the unmanned node i. o ,y o ) and current actual position (x) s ,y s Distance between )

[0032] Assuming the unmanned node is moving at a constant speed, calculate the distance from the current unmanned node's current actual position (x). s ,y s ) Move to the actual position (x′) at the next moment s+1 ,y′ s+1 The reachable point (x) within the required time period. s+1 ,y s+1 );

[0033] Obtain the planned heading angle α′ of the current unmanned node at the next moment. s+1 :

[0034]

[0035] Where Δx is the actual position (x′) s+1 ,y′ s+1 ) and reachable point (x) s+1 ,y s+1 The difference in the x-coordinates between (x' and y'), where Δy is the actual position (x'). s+1 ,y′ s+1) and reachable point (x) s+1 ,y s+1 The difference in the ordinates between )

[0036] Based on the planned heading angle α′ s+1 And the leader node's heading angle α at the next moment s+1 Get the current unmanned node from its current actual position (x) s ,y s ) to reachable point (x) s+1 ,y s+1 Dubins path between ) (s,s+1) :

[0037] path (s,s+1) =Dubins([x s ,y s ],[x s+1 ,y s+1 ])

[0038] Repeat the above process, the unmanned node continuously detects the movement of the leader node from time t to time t+1, and obtains the changes in the leader node's position and direction of travel during this movement.

[0039] Furthermore, the Euclidean distance d ij The calculation method is as follows:

[0040]

[0041] Among them, (x i ,y i (x) represents the coordinates of the currently unmanned node i. j ,y j Let be the coordinates of any node j.

[0042] Beneficial effects:

[0043] 1. This invention provides a collaborative path planning method for ground-based multi-unmanned system formations under leader mode. It addresses two scenarios for the leader node: turning and straight-line movement. For the turning scenario, a consistency deviation index is introduced. The parameters are adjusted to position the unmanned nodes, implementing a "shortcut" behavior to keep up with the leader node as much as possible while maintaining the basic formation. For straight-line travel, since the leader node and unmanned nodes have the same speed, this invention only makes minor adjustments to the angle to reduce formation distance deviation. The autonomous path planning algorithm proposed in this invention for unmanned nodes following manned nodes in weak network environments can adaptively adjust the formation by changing its own direction of travel while maintaining the formation to the greatest extent. The planned path is smooth and can meet the special requirements of turning during unmanned system travel.

[0044] 2. This invention provides a collaborative path planning method for ground-based multi-unmanned system formations under a leader mode, by introducing a consistency deviation index. Adjusting the position of unmanned nodes by parameters enables fine-tuning of the travel direction angle after turning. This method helps to minimize the path length of unmanned nodes. At the same time, for multiple consecutive turns, this invention achieves the goal of maintaining formation as much as possible by shortening the distance with the Leader node. Therefore, it can effectively plan the adaptive path adjustment of unmanned nodes in the event of sudden turns by manned nodes, resulting in a smooth overall path and minimal formation deformation.

[0045] 3. This invention provides a method for collaborative path planning of ground multi-unmanned systems in leader mode. For straight-line travel, this invention only makes minor adjustments to the angle to reduce formation distance deviation and save computing resources. Attached Figure Description

[0046] Figure 1 A research framework for the path coordination problem in multi-unmanned systems;

[0047] Figure 2 This is a diagram illustrating the formation;

[0048] Figure 3 This is a schematic diagram of the formation's movement.

[0049] Figure 4 This is a schematic diagram of the formation obstacle clearance process;

[0050] Figure 5 This is a typical Dubins path;

[0051] Figure 6 Modeling is required for formation;

[0052] Figure 7 This is for simulating the experimental environment structure;

[0053] Figure 8 The simulation results are for 4 groups of Leader paths with a duration of 50 seconds each.

[0054] Figure 9 The result of the formation and collaborative path planning for path 1;

[0055] Figure 10 The result of the collaborative path planning for path 2;

[0056] Figure 11 The result of the formation and collaborative path planning for path 3;

[0057] Figure 12 The result of the collaborative path planning for path 4 is shown. Detailed Implementation

[0058] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0059] This invention takes the ground environment as the research background, regards the manned system as the leader and multiple unmanned systems as followers, studies the path planning problem when multiple unmanned systems on the ground cooperate in formation with the manned system as the core, and explores how unmanned systems can cooperate with manned systems in unstable network environments.

[0060] It should be noted that ground-based environments are far more complex than aerial environments, and network instability is a significant factor in the study of ground-based unmanned systems. In ground environments, especially during intense confrontations such as war, communication networks are difficult to guarantee effectively or are unstable; this type of network environment is commonly referred to as a weak network condition. When studying multi-unmanned system formation coordination under weak network conditions, this instability must be fully considered. Path coordination control methods used under strong network conditions cannot be simply applied; rather, the possibility of limited or even disconnected network bandwidth must be taken into account.

[0061] For weak network conditions, this invention designs a system architecture for the path coordination process of multi-unmanned systems, such as... Figure 1 As shown, in this system architecture, unmanned nodes are equipped with their own path planning algorithms and rules, while the leader node (manned equipment) sends control commands to the unmanned nodes. Considering the unstable network environment, the leader node implements intermittent control over the unmanned system, mainly by issuing commands to adjust the unmanned system's behavior. The unmanned system actively acquires environmental information, including the behavior information of manned nodes, using its own detection payloads. It then responds specifically to commands and plans its own route in real time. Its behavior should be proactive, rather than entirely passive and controlled; its coordination relies more on the autonomous action capabilities of the unmanned equipment. Therefore, the research problem can be viewed as a multi-agent coordination problem, with the behavioral algorithms primarily focused on path design for the agents.

[0062] In practical applications, multi-unmanned systems are typically designed in formation according to the requirements of the scenario to ensure their operational effectiveness. For example, five ground-based unmanned vehicles might be configured to move in formation to ensure the effective area of ​​environmental detection. Figure 2 The V-shape, straight line, and linear formations shown are used to group the equipment, specifying parameters such as the distance between them.

[0063] This presents new challenges to path coordination in unmanned systems. In leader mode, manned or remotely controlled equipment may perform complex behaviors, requiring unmanned equipment to respond accordingly based on different data. For path planning problems, formation is usually designed relative to linear motion, and other movements of manned equipment require unmanned equipment to maintain formation as much as possible and complete corresponding path changes.

[0064] like Figure 3 As shown, the formation mode corresponds to the consistency requirements of each node in the spatiotemporal coordinate system. Each node must meet the formation constraints under the spatiotemporal coordinates in order to form a formation. Therefore, the formation transformation of multiple unmanned nodes in the leader mode needs to comprehensively consider the action capabilities of the unmanned nodes and the consistency requirements of the formation, and design a method for planning the action paths of the unmanned nodes.

[0065] Planning considerations

[0066] It should be noted that path coordination mainly focuses on the location attributes of unmanned nodes. For node n i Its location information can be regarded as coordinates in two-dimensional space (x, y). i ,y i Meanwhile, unmanned nodes possess motion attributes such as direction of travel α and speed v. For manned nodes, the direction of travel α0 is an unknown variable, while the speed v0 is variable. Unmanned nodes move in a pre-defined formation centered on manned nodes, with each unmanned node's direction of travel being α. i The speed of travel is v i The formation coordination problem studied in this paper requires planning the action scheme of multiple unmanned nodes based on the action information of manned nodes and the position information of several unmanned nodes at a certain moment {(α i ,v i )|i∈(1,2…m)}.

[0067] Path planning needs to consider avoiding threats and obstacles in the ground environment. Impassable areas can be approximated by circular or rectangular obstruction regions. These obstruction regions are impassable regardless of whether there are people or unattended nodes. Figure 4 As shown.

[0068] The maneuvering speed of unmanned nodes varies within their performance range, and they can perform actions such as acceleration, deceleration, and stopping. Considering the current development of unmanned ground equipment, the maneuvering speed of unmanned ground nodes is usually lower than that of similar manned maneuvering equipment. Therefore, during cooperative operations, unmanned nodes typically travel at their maximum speed within a reliable range. Thus, this invention treats unmanned nodes as performing uniform-speed maneuvers, where this speed is their maximum reliable maneuvering speed. The path planning problem studied primarily explores path adjustment planning methods under uniform-speed maneuvers.

[0069] Path planning model

[0070] This invention views manned / unmanned collaboration as a leader mode, with manned nodes as the core of the unmanned system response. In this context, the path planning of the unmanned system should be real-time planning, and the movement trajectory of the unmanned nodes should meet certain constraints.

[0071] Specifically, this invention introduces the Dubins path to perform path planning for unmanned nodes. Existing research has proven that the Dubins path is the shortest path between two points. For example... Figure 5 As shown, the Dubins path consists of a straight line and two arcs. In practical applications, it is commonly used for path planning of aircraft. It is also applicable to path planning during the maneuvering of unmanned ground nodes, accurately reflecting the process of adjusting the unmanned node's direction of travel along the arcs while maintaining constant speed, and ensuring that the travel path is the shortest route. Therefore, when planning the path of an unmanned node based on the Dubins path, the known location (x...) is crucial. s ,y s ), direction of travel α s and target location (x) o ,y o ), direction of travel α o Factors that need to be planned include the arc radius r. s and r o And the direction of the two turning arcs.

[0072] Formation requires modeling

[0073] Furthermore, formation mode is an important implementation pattern of multi-unmanned systems cooperating with manned systems. Formation is equivalent to adding constraints to the cooperative behavior of multi-unmanned systems, with the goal of improving the stability of the effect of autonomous cooperative actions of unmanned systems.

[0074] The formation coordination in this invention is mainly reflected in the coordination of positions, that is, the node set {n} at time t. i The positional attributes of nodes ∈ N|i∈0,1,2…m} satisfy the formation constraints, where n0 represents a person node. For typical formation settings, the formation L can be described using a quadruple L={m,V,D,B}. Here, m is the number of nodes in the formation; V={v i |i=0,1,2…m} represents the maneuvering speed of each unmanned node. Unmanned equipment usually has variable speed capability, but the corresponding energy consumption load is relatively large. In this study, it is set as a constant. D={d ij |i,j∈0,,1,2…m} represents node n i With n j The distance between them, calculated using the Euclidean distance formula, is as follows:

[0075]

[0076] B is the angle between the line connecting the movement direction and position of the two nodes, represented by the set B = {β}. ij Let |i=1,2…m+1,j=1,2…m+1} represent this, where β ij Represents node n i Relative to node nj Maneuvering direction α i and The angle between the directions, such as Figure 6 As shown.

[0077] This invention studies the leader pattern, where a central node exists in the formation constraints, and this central node determines the formation constraints. Therefore, B in the 4-tuple can be converted into the maneuver direction deviation of each node relative to the central node, i.e., B = {β} ik |i=1,2…m,k=center}. Furthermore, the maneuvering directions of both unmanned formations and manned nodes are subject to certain constraints, which are described by a directional deviation angle, denoted as Δα={α ij |i=1,2…m,j=1,2…m},α ij This represents the angular deviation of the maneuvering direction between any two nodes in the formation. Therefore, this paper uses the quintuple L={m,V,D,B,Δα} to constrain the formation.

[0078] Target modeling

[0079] The core focus of this invention is the path planning problem when unmanned nodes maintain specific formation constraints in a mode where a manned node acts as the leader. Two objective elements can be identified: on the one hand, satisfying the path coordination requirements from the initial point to the target point, and on the other hand, satisfying the formation constraints.

[0080] In leader mode, the initial and target points of unmanned nodes are determined by manned nodes and change continuously as manned nodes move. Formation constraints, also centered on manned nodes, are similarly determined by them. Therefore, the path coordination requirement for unmanned nodes can be equated to the unmanned node reaching its corresponding position (x) centered on a manned node at any given time t. o ,y o In cases where the formation deviates from the set formation during coordination, the deviation (Var) can be measured.

[0081]

[0082] in, The formula for fitting the consistency deviation is given, where i represents the number of unmanned nodes in the formation and t represents any time.

[0083] The objective function of this invention is to find an effective means to minimize the consistency deviation of unmanned formations, that is, for Seeking a path planning method This causes consistency deviation Var t The expected value is the smallest.

[0084] Based on the above problem model, the path planning algorithm of this invention will be described in detail below.

[0085] The leader-mode multi-unmanned system formation coordination problem studied in this invention requires unmanned systems to respond in real time to the behavior of manned systems, thus belonging to a real-time path planning problem. Solving this problem under the system architecture proposed in this research background requires designing a real-time path planning algorithm for unmanned systems and a management node operation algorithm that can adapt to unstable communication.

[0086] To address this problem, this invention proposes a heuristic path planning algorithm guided by group consistency deviation. This algorithm can be used for unmanned system grouping action simulation and actual path planning under the problem context.

[0087] Algorithm Analysis

[0088] In the leader mode, the fundamental goal of unmanned equipment formation cooperative path planning is to respond to the behavior of manned nodes. The fundamental goal of cooperative behavior is to achieve coordinated cooperation with manned nodes. Therefore, it is necessary to analyze the formation changes caused by the behavior of manned nodes, mainly including the following situations:

[0089]

[0090] To maintain a set formation, unmanned formations need to adjust their planned paths. This can be achieved by predicting the target point's movement direction in advance and planning the shortest path to shorten the distance and quickly restore formation. However, this method is vulnerable if the predicted target point is too far from the current location, making it difficult to effectively respond to sudden maneuvers such as changes in direction by manned nodes. This presents a trade-off between finding the shortest path and the ability to adapt to changes.

[0091] In this invention, it is initially assumed that both the Leader and Follower (unmanned nodes) move at a constant speed without stopping. Under these conditions, the r path in the dubins path of the unmanned node... s and r o The minimum value is a definite value directly related to one's own control ability; the variable requiring decision-making is... Direction of travel at any time α t By reasonably controlling the direction, we can meet the formation requirements as much as possible.

[0092] To address this issue, this paper introduces a formation coordination consistency index to describe the maintenance of formation. The consistency index is used to guide the "shortcut" behavior of unmanned nodes. By controlling the formation, a balance is sought between short paths and the ability to respond to changes, thereby achieving autonomous coordination of multiple unmanned nodes in leader mode.

[0093] Algorithm Design

[0094] The consistency of autonomous collaborative behavior is described using the collaborative deviation index, defined as:

[0095]

[0096] The formula reflects the difference between the formation requirements centered on manned nodes and the current state of unmanned nodes. Furthermore, considering that formation is a collaborative behavior among multiple unmanned nodes, the formula is modified by incorporating differences within multiple unmanned nodes and introducing formation maintenance factors during path adjustment.

[0097]

[0098] For each node, this paper proposes a single-node inconsistency description parameter, expressed as:

[0099]

[0100] Will Introduced into path planning algorithms, this invention proposes a method... The guided autonomous path planning algorithm for a single unmanned node is as follows:

[0101] A collaborative path planning method for ground-based multi-unmanned system formations under a leader mode is proposed. The formation contains one manned leader node, with the rest being unmanned nodes. The path planning method for each unmanned node is as follows:

[0102] The leader node's direction of travel and position are captured at a fixed detection frequency, and its current heading angle α is determined. s With the heading angle α at the next moment s+1 If they are the same, and not the same, introduce the consistency deviation index corresponding to the current unmanned node at the current time t. Adjust the current unmanned node's direction of travel. If the directions are the same, adjust the current unmanned node's direction of travel based on its speed.

[0103] Furthermore, the consistency deviation index The calculation formula is as follows:

[0104]

[0105] Where i is the label of the currently unmanned node, (x o ,y o (x) represents the target position of the current unmanned node under ideal conditions at time t. s ,y s ) represents the actual position of the currently unmanned node at time t, D[(x o ,y o ),(x s,y s )] i It means (x) o ,y o ) and (x s ,y s The distance between ), where α and μ are set weights, and β ij The actual position (x) of the current unmanned node i s ,y s The angle between the line connecting the actual position of any node j and the direction of travel of node j, Let x be the ideal position of the current unmanned node i. o ,y o The angle d between the line connecting the actual position of any node j and the direction of travel of node j. ij The actual position (x) of the current unmanned node i s ,y s The Euclidean distance between ) and any node j in the formation. Let x be the ideal position of the current unmanned node i. o ,y o ) is the Euclidean distance between the node and any node j in the formation, where m is the total number of nodes in the formation.

[0106] The consistency deviation index corresponding to the current unmanned node at the current time t is introduced. The specific steps to adjust the current unmanned node's direction of travel are as follows:

[0107] Calculate the consistency index con:

[0108]

[0109] Where Λ is a set constant, r o The current unmanned node moves along the Dubins path to the target position (x). o ,y o The radius of curvature at the end of the curve, α o+1 Point the current unmanned node to the target location (x) o ,y o The direction of travel. This is the consistency deviation index for the currently unmanned node at the current time t.

[0110] Calculate the actual position (x′) of the current unmanned node at the next moment based on the consistency index con. s+1 ,y′ s+1 ):

[0111] x′ s+1 =x o+1 +con*sin(α o+1 )

[0112] y′ s+1 =y o+1 +con*cos(α o+1 )

[0113] Among them, (x o+1 ,y o+1 The ideal position of the current unmanned node at the next moment is calculated by introducing formation constraints by capturing the position of the leader node at the next moment.

[0114] Assuming the unmanned node is moving at a constant speed, calculate the distance from the current unmanned node's current actual position (x). s ,y s Move to the actual position (x′) s+1 ,y′ s+1 The reachable point (x) within the required time period. s+1 ,y s+1 );

[0115] Obtain the planned heading angle α′ of the current unmanned node at the next moment. s+1 :

[0116] α′ s+1 =arctan(Δx / Δy)

[0117] Where Δx is the current actual position (x s ,y s ) and reachable point (x) s+1 ,y s+1 The difference in the x-coordinates between the two points, where Δy is the current actual position (x, y). s ,y s ) and reachable point (x) s+1 ,y s+1 The difference in the ordinates between )

[0118] Based on the planned heading angle α′ s+1 and the leader node's current heading angle α s Get the current unmanned node from its current actual position (x) s ,y s ) to reachable point (x) s+1 ,y s+1 Dubins path between ) (s,s+1) :

[0119] path (s,s+1) =Dubins([x s ,y s ],[x s+1 ,y s+1 ])

[0120] The specific steps for adjusting the travel direction of the current unmanned node based on its current speed are as follows:

[0121] Calculate the actual position (x′) of the currently unmanned node in the next time step. s+1 ,y′ s+1 ):

[0122] x' s+1 =x o +v*(x o+1 -x s ) / D

[0123] y' s+1 =y o +v*(y o+1 -y s ) / D

[0124] Where v is the current velocity of the unmanned node; D is the current target position (x) of the unmanned node i. o ,y o ) and current actual position (x) s ,y s Distance between )

[0125] Assuming the unmanned node is moving at a constant speed, calculate the distance from the current unmanned node's current actual position (x). s ,y s ) Move to the actual position (x′) at the next moment s+1 ,y′ s+1 The reachable point (x) within the required time period. s+1 ,y s+1 );

[0126] Obtain the planned heading angle α′ of the current unmanned node at the next moment. s+1 :

[0127] α′ s+1 =arctan(Δx / Δy)

[0128] Where Δx is the actual position (x′) s+1 ,y′ s+1 ) and reachable point (x) s+1 ,y s+1 The difference in the x-coordinates between (x' and y'), where Δy is the actual position (x'). s+1 ,y′ s+1 ) and reachable point (x) s+1 ,y s+1 The difference in the ordinates between )

[0129] Based on the planned heading angle α′ s+1 And the leader node's heading angle α at the next moment s+1Get the current unmanned node from its current actual position (x) s ,y s ) to reachable point (x) s+1 ,y s+1 Dubins path between ) (s,s+1) :

[0130] path (s,s+1) =Dubins([x s ,y s ],[x s+1 ,y s+1 ])

[0131] Repeat the above process, the unmanned node continuously detects the movement of the leader node from time t to time t+1, and obtains the changes in the leader node's position and direction of travel during this movement.

[0132] Therefore, this invention addresses two scenarios: Leader turning and straight-line movement. In the turning scenario, the γ_i^t parameter is introduced to adjust the target point position, implementing a "shortcut" behavior to keep up with the Leader node as much as possible while maintaining the basic formation. In the straight-line scenario, since the Leader and Followers have the same speed, the algorithm only makes minor adjustments to the angle to reduce formation distance deviation.

[0133] Furthermore, to verify the effectiveness of the proposed method, we constructed a co-simulation experimental platform based on Matlab software. The platform is divided into the following sections: Figure 7 The diagram shows two modules: Leader behavior simulation and Follower behavior simulation.

[0134] The Leader behavior simulation uses a random variable generation method to generate the Leader's movement path. In real-world scenarios, the Leader node's movement path is not random; its direction is usually adjusted within a certain range. We assume its direction adjustment angle is a random number between [-60, +60], and its movement speed is constant. Given v(i) = 20, four Leader path simulation results with a duration of 50 were obtained through randomized simulation, as follows: Figure 8 As shown.

[0135] Using the simulated path described above, an experiment was conducted to construct a 3-node group, where node 1 is the Leader node, and nodes 2 and 3 are Follower nodes. The Leader and Follower nodes travel at the same speed, and the group formation is a rear triangle. The required group distance is described in matrix form according to the node order as follows:

[0136]

[0137] The azimuth angle requirements between formation nodes are described as follows:

[0138]

[0139] The numerical unit is angle, with positive values ​​representing clockwise angles and negative values ​​representing counterclockwise angles.

[0140] Since this invention is conducted in a weak network environment, the Leader node periodically synchronizes its position and direction of travel with the unmanned nodes. It is assumed that the unmanned nodes cannot anticipate any turning actions, but rather learn about them after they occur through their own detection capabilities. Based on the formation requirements, Algorithm 1 is used to plan the path for each unmanned node, resulting in cooperative path planning for four paths, as follows: Figures 9-12 As shown. From Figure 9 The results show that the algorithm can effectively plan adaptive path adjustments for unmanned nodes in response to sudden turns by manned nodes, resulting in a smooth overall path and minimal formation deformation. Figure 10 The results show that the algorithm can fine-tune the travel direction angle after turning, which is beneficial for minimizing the path length of unmanned nodes. Figure 11 and Figure 12 The results of paths 3 and 4 show that the algorithm maintains the formation as much as possible by shortening the distance from the Leader node when there are multiple consecutive turns.

[0141] The above experiments demonstrate that the autonomous path planning algorithm proposed in this invention, which allows unmanned nodes to follow manned nodes in weak network environments, can adaptively adjust its formation by changing its direction of travel while maintaining the formation to the greatest extent possible. The planned path is smooth and can meet the special requirements of turning during unmanned system travel.

[0142] In summary, this invention takes the ground environment as its research background and studies the collaboration problem between unmanned and manned nodes, transforming it into a Leader-Follower model. In this model, manned nodes act as Leaders, autonomously executing maneuvers, while unmanned nodes act as Followers, responding around the manned nodes. First, this paper analyzes and models the problem, systematically analyzing the collaboration process of unmanned systems on the ground and introducing the Dubins path as a coherent path planning model. Then, considering the unstable network connections in the ground environment, this paper proposes a ground-based multi-unmanned system autonomous collaboration algorithm incorporating a consistency index. Finally, simulation experiments demonstrate that the algorithm can autonomously make expected response actions based on the different behaviors of manned nodes, achieving path optimization as much as possible while maintaining the basic formation. Building on this research, future work will continue to explore the impact of unmanned equipment's variable speed behavior on its own performance and formation capabilities, and to propose a variable speed adjustment control method for unmanned nodes based on a formation deviation index to meet the requirements of multi-unmanned system formation collaboration.

[0143] Of course, the present invention may have other various embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art can make various corresponding changes and modifications according to the present invention, but these corresponding changes and modifications should all fall within the protection scope of the appended claims.

Claims

1. A method for collaborative path planning in a formation of multiple unmanned ground systems under a leader mode, characterized in that, The formation contains one leader node controlled by a human, and the rest are unmanned nodes. The path planning method for each unmanned node is as follows: The leader node's direction of travel and position are captured at a fixed detection frequency, and its current heading angle is determined. Heading angle at the next moment If they are the same, and if not, introduce the currently unattended node at the current time. t Corresponding consistency deviation index Adjust the current unmanned node's direction of travel. If the directions are the same, adjust the current unmanned node's direction of travel based on its speed. The introduction of currently unmanned nodes at the current time t Corresponding consistency deviation index The specific steps to adjust the current unmanned node's direction of travel are as follows: Calculate the consistency index : in, To set constants, The current unmanned node moves along the Dubins path to the target location. The radius of curvature at the end of the curve, Point the currently unmanned node to the target location. The direction of travel, For the currently unmanned node at the current moment t The corresponding consistency deviation index; According to the consistency index Calculate the actual position of the currently unmanned node in the next time step. : in, The ideal position of the current unmanned node at the next moment is calculated by introducing formation constraints to capture the position of the leader node at the next moment. Assuming the unmanned node is moving at a constant speed, calculate the distance from the current unmanned node's actual position. Move to actual position Reachable points within the required time period ; Obtain the planned heading angle of the current unmanned node at the next moment. : in, Current actual location Reachable points The difference in the x-coordinates between them Current actual location Reachable points The difference in the ordinates between them; Based on the planned heading angle and the leader node's current heading angle Get the current unmanned node from its current actual location To reach point Dubins path between : 。 2. The method for collaborative path planning of ground-based multi-unmanned systems in leader mode as described in claim 1, characterized in that, Consistency Deviation Index The calculation formula is as follows: in, i This is the label of the currently unmanned node. For the current unmanned node in the ideal situation at time... t The target location For the currently unmanned node at time t The actual location, express and The distance between them and To set weights, Currently unmanned node i actual location With any node j Actual location connection and node j The angle between the directions of travel Currently unmanned node i Ideal position With any node j Actual location connection and node j The angle between the directions of travel Currently unmanned node i actual location With any node in the formation j The Euclidean distance between them Currently unmanned node i Ideal position With any node in the formation j The Euclidean distance between them m This represents the total number of nodes in the formation.

3. A method for collaborative path planning of ground-based multi-unmanned systems in leader mode as described in claim 1 or 2, characterized in that, The specific steps for adjusting the travel direction of the current unmanned node based on its current speed are as follows: Calculate the actual position of the currently unmanned node in the next time step. : in, The speed of the currently unmanned node; D Currently unmanned node i Target location Current actual location Distance between; Assuming the unmanned node is moving at a constant speed, calculate the distance from the current unmanned node's actual position. Move to the actual position at the next moment Reachable points within the required time period ; Obtain the planned heading angle of the current unmanned node at the next moment. : in, For actual location Reachable points The difference in the x-coordinates between them For actual location Reachable points The difference in the ordinates between them; Based on the planned heading angle Next-moment heading angle of the leader node Get the current unmanned node from its current actual location To reach point Dubins path between : Repeating the above process, the unmanned node continuously probes the leader node from time [time]. t At the time t The movement process of +1 is used to obtain the position change and direction change of the leader node during the movement process.

4. The method for collaborative path planning of ground-based multi-unmanned systems in leader mode as described in claim 2, characterized in that, European distance The calculation method is as follows: in, Currently unmanned node i coordinates For any node j The coordinates.

Images