Unmanned aerial vehicle cluster control design and simulation method and system based on consensus algorithm

Abstract

This application discloses a design and simulation method and system for UAV swarm control based on a consensus algorithm, relating to the field of UAV swarm control technology. The method includes: performing global path planning based on the start and end points of the UAV swarm in the current flight environment, with the goal of avoiding obstacle locations, to obtain several global path points for each UAV; based on the global path points and a preset communication topology of the UAV swarm, employing a consensus formation control algorithm to perform cooperative control of the UAV swarm, constructing a global controller for the UAV swarm; and constructing the potential function of the UAVs to perform local path planning and obstacle avoidance on the global controller, thereby achieving formation control of the UAV swarm. This application achieves efficient cooperative control of the UAV swarm through global path planning and consensus formation control, and effectively avoids obstacles during UAV movement by calculating the potential function, improving the control flexibility of the UAV formation.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) swarm control technology, specifically to a UAV swarm control design and simulation method and system based on a consensus algorithm. Background Technology

[0002] Unmanned aerial vehicle (UAV) swarm control technology refers to the use of advanced communication, navigation, and control algorithms to achieve collaborative, coordinated, and efficient mission execution among multiple UAVs. With the rapid development of UAV technology, swarm flight components have become a hot topic, especially in various industries such as military, agriculture, power line inspection, environmental monitoring, and logistics distribution. The application scenarios of UAV swarms are constantly expanding, demonstrating enormous potential and value.

[0003] Current drone swarm control technologies primarily rely on traditional path planning algorithms and basic formation control algorithms (such as centralized control, distributed control, and hybrid control). These technologies employ a single algorithm to handle global path planning or local obstacle avoidance, relying on communication between drones and sensor technology to perceive the surrounding environment and the status of other drones to prevent collisions. However, in complex environments with frequent sudden obstacles, existing technologies demonstrate poor collaborative optimization capabilities, making timely obstacle avoidance difficult. Furthermore, basic formation control algorithms face challenges in complex situations, including communication latency and data synchronization, hindering the accurate control of the entire drone swarm. Summary of the Invention

[0004] This application proposes a design and simulation method and system for UAV swarm control based on a consensus algorithm. It achieves efficient collaborative control of UAV swarms through global path planning and consensus formation control, and effectively avoids obstacles during UAV movement by calculating the potential function, thereby improving the control flexibility of UAV formation.

[0005] The first aspect of this application provides a method for the design and simulation of unmanned aerial vehicle (UAV) swarm control based on a consensus algorithm, the method comprising:

[0006] Based on the start and end points of the drone swarm in the current flight environment, global path planning is performed with the goal of avoiding obstacle locations, resulting in several global waypoints for each drone.

[0007] Based on the global path points and the preset communication topology of the UAV cluster, a consensus formation control algorithm is used to coordinate the control of the UAV cluster and construct a global controller for the UAV cluster.

[0008] By constructing the potential function of the UAVs to perform local path planning and obstacle avoidance for the global controller, formation control of the UAV swarm can be achieved.

[0009] The above scheme, based on the known current position and corresponding destination of each drone, performs global path planning for each drone, avoiding obstacle locations, to obtain the global path points traversed by the drone from the starting point to the destination. Then, a communication topology map of the drone swarm is constructed based on the communication relationships between drones. A global controller is built based on these communication relationships to coordinate the formation switching of the drone swarm, reducing communication latency and data asynchrony issues, improving the real-time performance of drone swarm control, and enhancing the flexibility of formation control. Finally, the potential functions between drones, the destination, and obstacles are calculated to measure the attractive and repulsive forces between these nodes, enabling the drone to effectively avoid obstacles in complex environments based on the planned global path.

[0010] In one possible implementation of the first aspect, global path planning is performed based on the start and end points of the UAV swarm in the current flight environment, with the goal of avoiding obstacle locations, to obtain several global waypoints for each UAV, specifically:

[0011] Starting from the current location of the drone, the global planning range is determined by the starting point and the ending point.

[0012] Based on the current location of the drone, a node search is performed within the global planning range with the goal of avoiding obstacle locations, to find the forced neighbors of each node within the global planning range; wherein, the forced neighbors are the nodes that force the algorithm to change its movement direction when performing global path planning;

[0013] Based on the forced neighbors, the drone's endpoint and origin, find the nodes belonging to the jump point within the global planning scope;

[0014] Based on the jump point and obstacle location, global path planning is performed starting from the current position of the drone to obtain several global path points for each drone.

[0015] The above scheme first plans the global planning range of the drone based on its starting point and ending point, providing data support for the drone's path planning. Then, it determines the drone's movement direction towards the ending point by finding the forced neighbors of each node within the global planning range. Finally, it performs global path planning using the jump points and obstacle locations to quickly find the global path from the drone to the ending point.

[0016] In one possible implementation of the first aspect, based on the current position of the UAV, a node search is performed within the global planning range with the objective of avoiding obstacle locations to find the forced neighbors of each node within the global planning range, specifically:

[0017] Starting from the current position of the drone, with the goal of reaching the destination, the search status of each node within the global planning range is determined by a preset node judgment rule; wherein, the search status includes unobstructed horizontal movement status, unobstructed diagonal movement status, obstructed horizontal movement status, and obstructed diagonal movement status.

[0018] Path selection is performed based on the search status and obstacle locations to determine the forced neighbors of each node.

[0019] In one possible implementation of the first aspect, path selection is performed based on the search state and obstacle locations to determine the forcing neighbors of each node, specifically as follows:

[0020] Based on the search state, and combined with the location of obstacles, a path is selected from the node to the corresponding adjacent node to determine the path trajectory between each node and its corresponding adjacent node.

[0021] Based on the search value of nodes in the path trajectory, the adjacent nodes are classified as natural neighbors or forced neighbors; wherein, the search value is used to measure whether a node is meaningful for search when making path selection.

[0022] The above scheme first determines the search status of each node, and then judges whether the neighboring nodes around the node have search value based on the search status. Neighboring nodes without search value will not be considered as nodes in path planning. Therefore, the retrieval efficiency can be improved by excluding nodes without search value.

[0023] In one possible implementation of the first aspect, based on the forced neighbors, the drone's endpoint, and the starting point, nodes belonging to the jump point are identified from the global planning scope, specifically as follows:

[0024] If a node within the global planning scope is the starting point or the ending point, then that node is determined as a jump point;

[0025] If a node within the global planning scope has the forced neighbor, then that node is determined to be a jump point;

[0026] If a node within the global planning scope is in an unobstructed diagonal movement state or an obstructed diagonal movement state, and the node can reach a certain jump point by moving horizontally or vertically, then the node is determined to be a jump point.

[0027] The above scheme determines which points in the path planning process require a change in direction by identifying jump points, thereby achieving accurate path planning.

[0028] In one possible implementation of the first aspect, based on the global pathpoints and a preset communication topology of the UAV swarm, a consensus formation control algorithm is used to coordinate the control of the UAV swarm, thereby constructing a global controller for the UAV swarm, specifically as follows:

[0029] Based on the global path points, a motion model for each drone is constructed using the drone's current position and the preset next drone position.

[0030] Based on the communication relationships between drones, construct the communication topology diagram of the drone cluster;

[0031] Based on the communication topology of the drone swarm and the motion model, and taking into account the formation switching of the drone swarm, a global controller for the drone swarm is constructed.

[0032] The above scheme takes into account that the behavior and interaction between drones affect the entire drone swarm. Therefore, a communication topology diagram of the drone swarm is constructed based on the communication relationship between drones to indicate whether there is connectivity between drones. Then, considering the formation switching of the drone swarm, a global controller for the drone swarm is constructed to record the status of each drone in the drone swarm. Through the global controller, the formation of drones can be flexibly switched, which greatly improves the flexibility of formation control.

[0033] In one possible implementation of the first aspect, the global controller is specifically:

[0034] ;

[0035] In the formula, x(t) is the current position of the UAV at time t. The flight speed of the controlled UAV is obtained by differentiating x(t), where D is the preset degree matrix, A is the adjacency matrix corresponding to the communication topology graph of the UAV cluster, and L is the Laplace matrix.

[0036] In one possible implementation of the first aspect, the global controller is used to perform local path planning and obstacle avoidance by constructing a potential function of the UAV, specifically:

[0037] Based on the starting and ending points of the drone swarm in the current flight environment, as well as the location of obstacles, the potential function of each drone is constructed using the artificial potential field method; wherein, the potential function includes the gravitational potential function and the repulsive potential function;

[0038] Based on the repulsive potential function between drones, the repulsive potential function between drones and obstacles, and the gravitational potential function between drones and their corresponding endpoints, the global controller performs local path planning and obstacle avoidance to obtain the drone cluster's direction of movement and speed at the next moment.

[0039] The formation control of the drone swarm is completed based on the direction and speed of movement of the drone swarm in the next moment.

[0040] The above scheme constructs potential functions between drones, obstacles, and the destination using an artificial potential field method to measure the repulsive and attractive forces between the drone and each node. Obstacle avoidance planning is performed as the drone moves along a global path using the combined effects of these repulsive and attractive forces, enabling it to effectively avoid obstacles in complex environments based on the planned global path.

[0041] In one possible implementation of the first aspect, the repulsive potential function and the gravitational potential function are specifically as follows:

[0042] The repulsive potential function between the drones is specifically expressed as follows:

[0043] ;

[0044] In the formula, U rep1 (p) is the repulsive potential function between the UAVs at node p, q i and q j K represents the positions of the i-th and j-th drones, respectively. r r is the repulsion coefficient. o This refers to the repulsive force range parameter;

[0045] The repulsive potential function between the drone and the obstacle is specifically expressed as follows:

[0046] ;

[0047] In the formula, U rep2 (p) represents the repulsive potential function between the UAV and the obstacle at node p, and q i For the location of the drone, q k Let K be the position of the k-th obstacle. ro r is the repulsion coefficient. o The repulsive force range parameter;

[0048] The gravitational potential function between the drone and the corresponding endpoint is specifically expressed as follows:

[0049] ;

[0050] In the formula, U att(q) is the gravitational potential function between the UAV at node q and the corresponding endpoint, where q i For the location of the drone, q target K represents the location of the endpoint. a This is the gravitational coefficient.

[0051] The second aspect of this application provides a design and simulation system for unmanned aerial vehicle (UAV) swarm control based on a consensus algorithm. The system includes: a global path planning module, a global controller generation module, and a local obstacle avoidance module.

[0052] The global path planning module is used to perform global path planning based on the start and end points of the UAV cluster in the current flight environment, with the goal of avoiding obstacle locations, and obtains several global path points for each UAV.

[0053] The global controller generation module is used to construct a global controller for the UAV cluster by using a consistent formation control algorithm to coordinate the control of the UAV cluster based on the global path points and the preset communication topology map of the UAV cluster.

[0054] The local obstacle avoidance module is used to construct the potential function of the UAV to perform local path planning and obstacle avoidance for the global controller, so as to realize the formation control of the UAV swarm. Attached Figure Description

[0055] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0056] Figure 1 This is a schematic flowchart illustrating a specific embodiment of a UAV swarm control design and simulation method based on a consensus algorithm provided in this application.

[0057] Figure 2 This is a search state diagram illustrating a design and simulation method for unmanned aerial vehicle (UAV) swarm control based on a consensus algorithm, provided in a certain embodiment of this application.

[0058] Figure 3 This is a communication topology diagram of a UAV swarm, provided in a certain embodiment of the UAV swarm control design and simulation method based on a consensus algorithm.

[0059] Figure 4 This is a drone swarm path diagram provided in a certain embodiment of the present application for a drone swarm control design and simulation method based on a consensus algorithm;

[0060] Figure 5 This is a structural diagram of a drone swarm control design and simulation system based on a consensus algorithm, provided in a certain embodiment of this application. Detailed Implementation

[0061] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0062] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0063] First Embodiment

[0064] Existing drone swarm control methods involve selecting a leader drone within the formation, with other drones following its trajectory at a certain relative distance. While this method is simple to implement and offers high formation accuracy, it is highly dependent on the leader, whose performance directly impacts the overall swarm performance. Therefore, how to effectively control the entire drone swarm, clearly define the state characteristics of each drone, and improve the overall control of the swarm is the main research direction of this application. Furthermore, the real-time performance and accuracy of path planning and obstacle avoidance in complex environments in existing drone swarm technologies still need improvement, and their ability to cope with sudden obstacles is insufficient.

[0065] like Figure 1 As shown, Figure 1 This application provides a flowchart illustrating a method for designing and simulating UAV swarm control based on a consensus algorithm, according to a specific embodiment. The method includes steps S1 to S3, detailed below:

[0066] Step S1: Based on the starting point and ending point of the drone cluster in the current flight environment, perform global path planning with the goal of avoiding obstacle locations, and obtain several global path points for each drone.

[0067] In this embodiment, the JPS jump point algorithm is used for global path planning of UAV swarm flight. First, the corresponding destination is divided into a 100m range from the current position of the UAV to determine the global planning range. Then, several global path points that can avoid obstacles and reach the destination are calculated by using the starting point, the destination, and the position of obstacles.

[0068] Among them, the JPS jump point algorithm is a graph-based search algorithm, which is... Improvements were made based on the algorithm, while retaining... The algorithm's framework, the cost function, can still be expressed as: But optimized The algorithm finds the successor node. Unlike... The JPS jump point algorithm, which traverses all reachable neighbors of the current node, pre-selects logically computationally valuable jump points as children of the current node through a jump point search strategy. This reduces the cost evaluation operations on intermediate nodes and further improves the algorithm's efficiency. The implementation of the jump point search strategy is mainly based on neighbor pruning rules and jump rules.

[0069] The neighbor pruning rule determines the natural and forced neighbors of a node based on the search status of nodes within the global planning scope, which can greatly improve path planning efficiency. First, starting from the drone's current position, with the destination as the goal, the search status of each node within the global planning scope is determined by a preset node judgment rule. This node judgment rule specifies four search states: unobstructed horizontal movement, unobstructed diagonal movement, obstructed horizontal movement, and obstructed diagonal movement. An obstruction is defined as the presence of an obstacle within the global planning scope. Specifically, as shown... Figure 2 As shown in Figure (a), point p(x) can reach point x without any obstacles by moving horizontally, thus showing the unobstructed horizontal movement state of point x; in Figure (b), point p(x) moves diagonally to point x, and there are no obstacles in the figure, thus showing the unobstructed diagonal movement state of point x; in Figure (c), point p(x) moves horizontally to point x, and there are black nodes indicating the location of obstacles in the figure, thus showing the obstructed horizontal movement state of point x; in Figure (d), point p(x) moves diagonally to point x, and there are black nodes indicating the location of obstacles around it, thus showing the obstructed diagonal movement state of point x.

[0070] Then, based on the search state and the location of obstacles, path selection is performed from the node to its corresponding neighboring node to determine the path trajectory between each node and its corresponding neighboring node. When determining the path trajectory, the search value of nodes within the global planning scope is also assessed, and the neighboring nodes of a node are classified as natural neighbors or forced neighbors based on the search value.

[0071] Among them, the natural neighbors are adjacent nodes that can be directly reached from the current node within the global planning scope, that is, nodes that are horizontally, vertically, and diagonally adjacent to the current node; the forced neighbors are nodes that force the movement direction to change during the path planning process; in addition, clipped neighbors with no search value will also be divided among the natural neighbors.

[0072] Furthermore, by distinguishing between natural neighbors and forced neighbors with search value, as well as pruned neighbors with no search value, the JPS jump point algorithm can avoid considering the pruned neighbors in global path planning, saving the computation of these worthless nodes and greatly improving path planning efficiency.

[0073] in addition, Figure 2 It also demonstrates how to select paths based on search status and obstacle locations, and the process of determining the forced neighbors of each node. Figure 2 Figures (a) and (b) show node x in an unobstructed horizontal movement state and an unobstructed diagonal movement state, respectively. Node x is the current node, node p(x) is the parent node of node x, and gray nodes are extended nodes. If the extended node can be reached through the parent node via a shorter or equal path without passing through the current node, then the extended node does not need to be reached through the current node. That is, the red path in Figures (a) and (b) that leads to the extended node via the current node x can be replaced by the blue path that does not pass through the current node x. In this case, the extended node can be determined as a pruned neighbor of the current node, indicating that the extended node has no search value.

[0074] Figure 2 In Figures (c) and (d), node x is shown in a horizontal movement state with obstacles and a diagonal movement state with obstacles, respectively. Black nodes represent nodes where obstacles exist, and blue nodes represent forced neighbors. If the shortest path from node p(x) to node m must pass through node x, then node m is called a forced neighbor of node x.

[0075] Based on the aforementioned neighbor pruning rules, forced neighbors and natural neighbors starting from the drone's current position within the global planning scope can be found. Then, according to the aforementioned jump rules, several jump points are found from the global planning scope to assist the JPS jump point algorithm in global path planning.

[0076] The jump rules define how to find jump points from all nodes within the global path planning. Specifically, for any node n, if node n is a location or destination, then node n is a jump point; if node n has a forced neighbor, then node n is a jump point; if the movement from node n's parent node p(n) to node n is a grid diagonal movement (i.e., node n's search state is an unobstructed diagonal movement state or an obstructed diagonal movement state), and node n can reach a certain jump point by moving horizontally or vertically, then node n is also a jump point.

[0077] After determining the jump point using the above method, the JPS jump point algorithm will search for the 8 directions of the starting point within the global planning scope. When a jump point or obstacle is found, the search in that direction will stop, and the found jump point will be added to the first list. After the current point search is completed, the jump point that can achieve the shortest path will be extracted from the first list, the jump point will be set as the current node for the next search, and the jump point will be added to the second list. This process will be repeated until the destination is added to the second list, thus completing the global path planning and finding several global path points for each drone.

[0078] Optionally, in this embodiment, the UAV flight control board incorporates the PX4 open-source flight software, responsible for the UAV's basic attitude and position control, possessing multiple flight modes and automation functions, and is customizable and expandable. The onboard computer is equipped with an Ubuntu operating system and deploys the ROS robot operating system framework. ROS manages the code of each functional module, improving code reusability. Its distributed processing framework loosely couples the modules, enabling data interaction between modules through various communication methods. The ground control console implements mission planning functions based on the open-source software QGroundControl, can run on various devices, and provides the UAV with complete flight control and parameter settings.

[0079] Step S2: Based on the global path points and the preset communication topology of the UAV cluster, a consensus formation control algorithm is used to coordinate the control of the UAV cluster and construct a global controller for the UAV cluster.

[0080] In this embodiment, during the collaborative formation of a drone swarm, the behavior and interactions between drones affect the entire formation. These drones are interconnected through a communication network; therefore, the communication relationships between these drones are represented as a graph structure to obtain the communication topology of the drone swarm. Each drone is modeled as a node in the graph, and the connections between nodes represent the information flow between drones, specifically as follows: Figure 3 As shown.

[0081] Then, based on the communication topology of the drone cluster, an adjacency matrix can be obtained. The elements in this adjacency matrix represent whether there is a communication link between drones, for example, a. ij It is the i-th and j-th elements of the adjacency matrix. A value of 1 indicates that there is communication between drone i and drone j, while a value of 0 indicates that there is no communication between drone i and drone j.

[0082] Then, based on the current position of drone i at time t and the position that drone i needs to reach in the next time step, a motion model for each drone is constructed, specifically expressed as follows:

[0083] ;

[0084] In the formula, x(t) is the current position of the UAV at time t. The flight speed of the control drone is obtained by differentiating x(t). t represents the control input for the drone at time t, and N represents the total number of drones in the swarm.

[0085] The flight speed of the corresponding drone can be adjusted through the control input. Therefore, a consistency controller for each drone can be constructed based on the above formula, specifically:

[0086] ;

[0087] In the formula, a ij It is the i-th and j-th terms of the adjacency matrix. Let be the current position of drone j at time t. Let be the current position of drone i at time t.

[0088] If the communication topology of the UAV swarm is undirected and connected, then according to the aforementioned consensus controller, the influence of the positions of UAVs on their own flight speed can be studied through the communication connections between UAVs provided by the adjacency matrix, thereby understanding the influence of their own positions and ultimately enabling the UAV formation to reach a consistent equilibrium. Based on the motion model, the communication topology of the UAV swarm, and the consensus controller, the global controller of the UAV swarm can be obtained, with the specific expression as follows:

[0089] ;

[0090] In the formula, x(t) is the current position of the UAV at time t. The flight speed of the controlled drone is obtained by differentiating x(t) with respect to t; D is a preset degree matrix, which is a diagonal matrix, where each diagonal element D ii Let A represent the degree of node i, which is the sum of the weights of all edges connected to node i; let A be the adjacency matrix corresponding to the communication topology graph of the UAV cluster; L is the Laplace matrix, which contains the connection information of the communication topology graph and is the core of the global controller. It can drive the system to reach consensus by aggregating the differences in neighbor states. When the communication topology graph is connected, the property of L ensures that the states of all UAVs eventually converge to the same value.

[0091] As can be seen from the global controller, the closed-loop dynamic characteristics of the UAV formation depend on the Laplace matrix of the formation. In discrete time, the method for updating the state parameters of each UAV in the formation is as follows:

[0092] ;

[0093] In the formula, dt represents the data sampling time interval. Based on the above theoretical foundation, the formation switching problem can be transformed into a problem of coordinating variable switching, that is, a set of coordinating variables corresponds to a formation, and by designing a certain coordinating variable switching mechanism, the formation switching process can be realized.

[0094] Step S3: By constructing the potential function of the UAV, local path planning and obstacle avoidance are performed on the global controller to achieve formation control of the UAV swarm.

[0095] In this embodiment of the application, collisions between drones and between drones and obstacles must also be considered during the formation of drone swarms. Therefore, the artificial potential field method is used for local path planning and obstacle avoidance.

[0096] The artificial potential field method designs the drone's motion as movement within a virtual force field. The target point / endpoint exerts an attractive force on the drone, drawing it towards it; obstacles and other drones exert a repulsive force, preventing collisions. The net force acting on the drone is the vector sum of the attractive and repulsive forces. The position is updated based on this net force, thus achieving path planning. This virtual force field, also called a potential field, comprises an attractive and a repulsive field. The attractive field is typically proportional to the distance between the drone and the target point; the greater the distance, the stronger the attraction. The repulsive field is typically inversely proportional to the distance between the drone and obstacles; the closer the distance, the stronger the repulsive force.

[0097] This application's embodiments construct a potential field by building potential functions, where the potential function is a differentiable function, and the gravitational and repulsive functions are the simplest potential functions. Given the starting and ending points of the drone swarm in the current flight environment, as well as the locations of obstacles, a potential function is constructed for each drone to enable it to move along a certain path and avoid obstacles under the combined influence of gravity and repulsion until it reaches the destination.

[0098] The potential function U(q) of a node q within the scope of global programming can be expressed as:

[0099] ;

[0100] In the formula, Let gravitational potential function be . Let be the repulsive potential function.

[0101] Then, based on the potential functions of the nodes, the repulsive potential function between UAVs, the repulsive potential function between UAVs and obstacles, and the gravitational potential function between UAVs and their corresponding endpoints are constructed.

[0102] The repulsive potential function between the drones is specifically expressed as follows:

[0103] ;

[0104] In the formula, U rep1 (p) is the repulsive potential function between the UAVs at node p, q i and q j K represents the positions of the i-th and j-th drones, respectively. r r is the repulsion coefficient. o This refers to the repulsive force range parameter;

[0105] The repulsive potential function between the drone and the obstacle is specifically expressed as follows:

[0106] ;

[0107] In the formula, U rep2 (p) represents the repulsive potential function between the UAV and the obstacle at node p, and q i For the location of the drone, q k Let K be the position of the k-th obstacle. ro r is the repulsion coefficient. o The repulsive force range parameter;

[0108] The gravitational potential function between the drone and the corresponding endpoint is specifically expressed as follows:

[0109] ;

[0110] In the formula, U att (q) is the gravitational potential function between the UAV at node q and the corresponding endpoint, where q i For the location of the drone, q target K represents the location of the endpoint. a This is the gravitational coefficient.

[0111] Based on the aforementioned repulsive potential function and gravitational potential function, the global controller performs local path planning and obstacle avoidance to obtain the movement direction and speed of the UAV swarm in the next moment. This allows the UAVs to move along a certain path and avoid obstacles under the combined action of gravity and repulsion until they reach the destination. Ultimately, this enables the UAV swarm to complete formation control and switching as a whole, improving the flexibility of formation control and enabling the UAVs to effectively avoid obstacles in complex environments.

[0112] Based on steps S1, S2, and S3 provided in the embodiments of this application, a simulation test of UAV swarm formation obstacle avoidance in a complex scenario was conducted in a simulation system. In the simulation system, a complex obstacle avoidance scenario consisting of 5 UAVs and 10 different obstacles was created. During the UAV swarm's journey to the destination, 10 different obstacles were placed. The UAV swarm maintained its formation while using an artificial potential field method to avoid obstacles, keeping a minimum distance of more than 1 meter from the obstacles. After successfully avoiding obstacles, it maintained its formation and reached the destination.

[0113] exist Figure 4 The image illustrates the path of a drone swarm to its destination in a simulation system. Red squares represent global path points planned by JPS (Journal of Path Planning), yellow squares represent the destination, blue balls represent the leader drone, green balls represent the followers, and black blocks represent obstacles of varying sizes. The process of the drone swarm reaching its destination is shown from the top left corner to the top right corner, then to the bottom left corner, and finally to the bottom right corner. During the flight, the global path points guide the drones to their destination. The bottom right corner shows the flight paths of the drone swarm, represented by five different colors, each color representing the flight path of one drone. It demonstrates how the five drones avoid obstacles throughout the flight, ultimately maintaining a cross formation to reach the destination after obstacle avoidance. The image clearly shows that the drone swarm effectively avoided obstacles while traveling in formation to the destination.

[0114] Implementing the embodiments of this application has the following beneficial effects:

[0115] This application embodiment performs global path planning for each drone based on its known current position and corresponding destination, prioritizing obstacle avoidance, thus obtaining the global path points traversed by the drone from the starting point to the destination. Then, a communication topology map of the drone swarm is constructed based on the communication relationships between drones. A global controller is built based on these communication relationships to coordinate the formation switching of the drone swarm, reducing communication latency and data asynchrony issues, improving the real-time performance of drone swarm control, and enhancing the flexibility of formation control. Finally, the potential functions between drones, the destination, and obstacles are calculated to measure the attractive and repulsive forces between these nodes, enabling the drone to effectively avoid obstacles in complex environments based on the planned global path.

[0116] Second Embodiment

[0117] Furthermore, in order to implement the UAV swarm control design and simulation system based on the consensus algorithm corresponding to the above method embodiments, and to achieve the corresponding functions and technical effects, Figure 5A structural diagram of a UAV swarm control design and simulation system based on a consensus algorithm is provided. For ease of explanation, only the parts relevant to this embodiment are shown. The UAV swarm control design and simulation system based on a consensus algorithm provided in this application embodiment includes:

[0118] The global path planning module 201 is used to perform global path planning based on the start and end points of the UAV cluster in the current flight environment, with the goal of avoiding obstacle locations, and obtain several global path points for each UAV.

[0119] In this embodiment of the application, the global planning range is determined by taking the current position of the UAV as the starting point and the ending point;

[0120] Based on the current location of the drone, a node search is performed within the global planning range with the goal of avoiding obstacle locations, to find the forced neighbors of each node within the global planning range; wherein, the forced neighbors are the nodes that force the algorithm to change its movement direction when performing global path planning;

[0121] Based on the forced neighbors, the drone's endpoint and origin, find the nodes belonging to the jump point within the global planning scope;

[0122] Based on the jump point and obstacle location, global path planning is performed starting from the current position of the drone to obtain several global path points for each drone.

[0123] The global controller generation module 202 is used to construct a global controller for the UAV cluster by using a consistent formation control algorithm to coordinate the control of the UAV cluster based on the global path points and the preset communication topology map of the UAV cluster.

[0124] In this embodiment of the application, a motion model of each drone is constructed based on the global path point, using the drone's current position and the preset next drone position.

[0125] Based on the communication relationships between drones, construct the communication topology diagram of the drone cluster;

[0126] Based on the communication topology of the drone swarm and the motion model, and taking into account the formation switching of the drone swarm, a global controller for the drone swarm is constructed.

[0127] The local obstacle avoidance module 203 is used to construct the potential function of the UAV to perform local path planning and obstacle avoidance for the global controller, so as to realize the formation control of the UAV swarm.

[0128] In this embodiment of the application, collisions between drones and between drones and obstacles must also be considered during the formation of drone swarms. Therefore, the artificial potential field method is used for local path planning and obstacle avoidance.

[0129] The artificial potential field method designs the drone's motion as movement within a virtual force field. The target point / endpoint exerts an attractive force on the drone, drawing it towards it; obstacles and other drones exert a repulsive force, preventing collisions. The net force acting on the drone is the vector sum of the attractive and repulsive forces. Position is updated based on this net force, thus enabling path planning. The gravitational field is typically proportional to the distance between the drone and the target point; the greater the distance, the stronger the attraction. The repulsive field is typically inversely proportional to the distance between the drone and obstacles; the closer the distance, the stronger the repulsive force.

[0130] This application's embodiments construct a potential field by building potential functions, where the potential function is a differentiable function, and the gravitational and repulsive functions are the simplest potential functions. Given the starting and ending points of the drone swarm in the current flight environment, as well as the locations of obstacles, a potential function is constructed for each drone to enable it to move along a certain path and avoid obstacles under the combined influence of gravity and repulsion until it reaches the destination.

[0131] The potential function U(q) of a node q within the scope of global programming can be expressed as:

[0132] ;

[0133] In the formula, Let gravitational potential function be . Let be the repulsive potential function.

[0134] Then, based on the potential functions of the nodes, the repulsive potential function between UAVs, the repulsive potential function between UAVs and obstacles, and the gravitational potential function between UAVs and their corresponding endpoints are constructed.

[0135] The repulsive potential function between the drones is specifically expressed as follows:

[0136] ;

[0137] In the formula, U rep1 (p) is the repulsive potential function between the UAVs at node p, q i and q j K represents the positions of the i-th and j-th drones, respectively. r r is the repulsion coefficient. o This refers to the repulsive force range parameter;

[0138] The repulsive potential function between the drone and the obstacle is specifically expressed as follows:

[0139] ;

[0140] In the formula, U rep2 (p) represents the repulsive potential function between the UAV and the obstacle at node p, and q i For the location of the drone, q k Let K be the position of the k-th obstacle. ro r is the repulsion coefficient. o This refers to the repulsive force range parameter.

[0141] The gravitational potential function between the drone and the corresponding endpoint is specifically expressed as follows:

[0142] ;

[0143] In the formula, U att (q) is the gravitational potential function between the UAV at node q and the corresponding endpoint, where q i For the location of the drone, q target K represents the location of the endpoint. a This is the gravitational coefficient.

[0144] Based on the aforementioned repulsive potential function and gravitational potential function, the global controller performs local path planning and obstacle avoidance to obtain the movement direction and speed of the UAV swarm in the next moment. This allows the UAVs to move along a certain path and avoid obstacles under the combined action of gravity and repulsion until they reach the destination. Ultimately, this enables the UAV swarm to complete formation control and switching as a whole, improving the flexibility of formation control and enabling the UAVs to effectively avoid obstacles in complex environments.

[0145] In some embodiments, the global path planning module 201 specifically comprises:

[0146] The JPS jump point algorithm was used for global path planning of UAV swarm flight. First, the corresponding destination was divided into a 100m range from the current position of the UAV to determine the global planning range. Then, several global path points that can avoid obstacles and reach the destination were calculated by using the start point, destination, and obstacle positions.

[0147] Among them, the JPS jump point algorithm is a graph-based search algorithm, which is... Improvements were made based on the algorithm, while retaining... The algorithm's framework, the cost function, can still be expressed as: But optimized The algorithm finds the successor node. Unlike... The JPS jump point algorithm, which traverses all reachable neighbors of the current node, pre-selects logically computationally valuable jump points as children of the current node through a jump point search strategy. This reduces the cost evaluation operations on intermediate nodes and further improves the algorithm's efficiency. The implementation of the jump point search strategy is mainly based on neighbor pruning rules and jump rules.

[0148] The neighbor pruning rule determines the natural and forced neighbors of a node based on the search status of nodes within the global planning scope, which can greatly improve path planning efficiency. First, starting from the drone's current position, with the destination as the goal, the search status of each node within the global planning scope is determined by a preset node judgment rule. This node judgment rule specifies four search states: unobstructed horizontal movement, unobstructed diagonal movement, obstructed horizontal movement, and obstructed diagonal movement. An obstructed state is defined as one where obstacles exist within the global planning scope; a horizontal movement state is defined as one where the previous movement was horizontal, and a diagonal movement state is defined as one where the previous movement was diagonal.

[0149] Then, based on the search state and the location of obstacles, path selection is performed from the node to its corresponding neighboring node to determine the path trajectory between each node and its corresponding neighboring node. When determining the path trajectory, the search value of nodes within the global planning scope is also assessed, and the neighboring nodes of a node are classified as natural neighbors or forced neighbors based on the search value.

[0150] Among them, the natural neighbors are adjacent nodes that can be directly reached from the current node within the global planning scope, that is, nodes that are horizontally, vertically, and diagonally adjacent to the current node; the forced neighbors are nodes that force the movement direction to change during the path planning process; in addition, clipped neighbors with no search value will also be divided among the natural neighbors.

[0151] Furthermore, by distinguishing between natural neighbors and forced neighbors with search value, as well as pruned neighbors with no search value, the JPS jump point algorithm can avoid considering the pruned neighbors in global path planning, saving the computation of these worthless nodes and greatly improving path planning efficiency.

[0152] Based on the aforementioned neighbor pruning rules, forced neighbors and natural neighbors starting from the drone's current position within the global planning scope can be found. Then, according to the aforementioned jump rules, several jump points are found from the global planning scope to assist the JPS jump point algorithm in global path planning.

[0153] The jump rules define how to find jump points from all nodes within the global path planning. Specifically, for any node n, if node n is a location or destination, then node n is a jump point; if node n has a forced neighbor, then node n is a jump point; if the movement from node n's parent node p(n) to node n is a grid diagonal movement (i.e., node n's search state is an unobstructed diagonal movement state or an obstructed diagonal movement state), and node n can reach a certain jump point by moving horizontally or vertically, then node n is also a jump point.

[0154] After determining the jump point using the above method, the JPS jump point algorithm will search for the 8 directions of the starting point within the global planning scope. When a jump point or obstacle is found, the search in that direction will stop, and the found jump point will be added to the first list. After the current point search is completed, the jump point that can achieve the shortest path will be extracted from the first list, the jump point will be set as the current node for the next search, and the jump point will be added to the second list. This process will be repeated until the destination is added to the second list, thus completing the global path planning and finding several global path points for each drone.

[0155] Optionally, in this embodiment, the UAV flight control board incorporates the PX4 open-source flight software, responsible for the UAV's basic attitude and position control, possessing multiple flight modes and automation functions, and is customizable and expandable. The onboard computer is equipped with an Ubuntu operating system and deploys the ROS robot operating system framework. ROS manages the code of each functional module, improving code reusability. Its distributed processing framework loosely couples the modules, enabling data interaction between modules through various communication methods. The ground control console implements mission planning functions based on the open-source software QGroundControl, can run on various devices, and provides the UAV with complete flight control and parameter settings.

[0156] In some embodiments, the global controller generation module 202 specifically comprises:

[0157] In the collaborative formation of a drone swarm, the behavior and interactions between drones affect the entire formation. These drones are interconnected through communication network links, so the communication relationships between these drones can be represented by a graph structure to obtain the communication topology of the drone swarm. Each drone is modeled as a node in the graph, and the lines between nodes represent the information flow between drones.

[0158] Then, based on the communication topology of the drone cluster, an adjacency matrix can be obtained. The elements in this adjacency matrix represent whether there is a communication link between drones, for example, a. ij It is the i-th and j-th elements of the adjacency matrix. A value of 1 indicates that there is communication between drone i and drone j, while a value of 0 indicates that there is no communication between drone i and drone j.

[0159] Then, based on the current position of drone i at time t and the position that drone i needs to reach in the next time step, a motion model for each drone is constructed, specifically expressed as follows:

[0160] ;

[0161] In the formula, x(t) is the current position of the UAV at time t. The flight speed of the control drone is obtained by differentiating x(t). t represents the control input for the drone at time t, and N represents the total number of drones in the swarm.

[0162] The flight speed of the corresponding drone can be adjusted through the control input. Therefore, a consistency controller for each drone can be constructed based on the above formula, specifically:

[0163] ;

[0164] In the formula, a ij It is the i-th and j-th terms of the adjacency matrix. Let be the current position of drone j at time t. Let be the current position of drone i at time t.

[0165] If the communication topology of the UAV swarm is undirected and connected, then according to the aforementioned consensus controller, the influence of the positions of UAVs on their own flight speed can be studied through the communication connections between UAVs provided by the adjacency matrix, thereby understanding the influence of their own positions and ultimately enabling the UAV formation to reach a consistent equilibrium. Based on the motion model, the communication topology of the UAV swarm, and the consensus controller, the global controller of the UAV swarm can be obtained, with the specific expression as follows:

[0166] ;

[0167] In the formula, x(t) is the current position of the UAV at time t. The flight speed of the controlled drone is obtained by differentiating x(t) with respect to t; D is a preset degree matrix, which is a diagonal matrix, where each diagonal element D ii Let A represent the degree of node i, which is the sum of the weights of all edges connected to node i; let A be the adjacency matrix corresponding to the communication topology graph of the UAV cluster; L is the Laplace matrix, which contains the connection information of the communication topology graph and is the core of the global controller. It can drive the system to reach consensus by aggregating the differences in neighbor states. When the communication topology graph is connected, the property of L ensures that the states of all UAVs eventually converge to the same value.

[0168] As can be seen from the global controller, the closed-loop dynamic characteristics of the UAV formation depend on the Laplace matrix of the formation. In discrete time, the method for updating the state parameters of each UAV in the formation is as follows:

[0169] ;

[0170] In the formula, dt represents the data sampling time interval. Based on the above theoretical foundation, the formation switching problem can be transformed into a problem of coordinating variable switching, that is, a set of coordinating variables corresponds to a formation, and by designing a certain coordinating variable switching mechanism, the formation switching process can be realized.

[0171] Implementing the embodiments of this application has the following beneficial effects:

[0172] This application embodiment performs global path planning for each drone based on its known current position and corresponding destination, prioritizing obstacle avoidance, thus obtaining the global path points traversed by the drone from the starting point to the destination. Then, a communication topology map of the drone swarm is constructed based on the communication relationships between drones. A global controller is built based on these communication relationships to coordinate the formation switching of the drone swarm, reducing communication latency and data asynchrony issues, improving the real-time performance of drone swarm control, and enhancing the flexibility of formation control. Finally, the potential functions between drones, the destination, and obstacles are calculated to measure the attractive and repulsive forces between these nodes, enabling the drone to effectively avoid obstacles in complex environments based on the planned global path.

[0173] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application. In particular, it should be noted that any modifications, equivalent substitutions, or improvements made by those skilled in the art within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for the design and simulation of unmanned aerial vehicle (UAV) swarm control based on a consensus algorithm, characterized in that, include: Based on the start and end points of the drone swarm in the current flight environment, global path planning is performed with the goal of avoiding obstacle locations, resulting in several global waypoints for each drone. Based on the global pathpoints and the preset communication topology of the UAV cluster, a consistent formation control algorithm is used to coordinate the control of the UAV cluster and construct a global controller for the UAV cluster. Specifically, based on the global pathpoints, a motion model of each UAV is constructed using the current position of the UAV and the preset next position of the UAV; based on the communication relationships between the UAVs, a communication topology of the UAV cluster is constructed; and based on the communication topology of the UAV cluster and the motion model, taking into account the formation switching of the UAV cluster, a global controller for the UAV cluster is constructed. Specifically, the global controller is: ； In the formula, x(t) is the current position of the UAV at time t. The flight speed of the controlled drone is obtained by differentiating x(t), where D is the preset degree matrix, A is the adjacency matrix corresponding to the communication topology graph of the drone cluster, and L is the Laplace matrix. By constructing potential functions for UAVs to perform local path planning and obstacle avoidance for the global controller, formation control of the UAV swarm is achieved. Specifically, based on the start and end points of the UAV swarm in the current flight environment and the positions of obstacles, a potential function for each UAV is constructed using an artificial potential field method. The potential function includes an gravitational potential function and a repulsive potential function. Based on the repulsive potential functions between UAVs, the repulsive potential functions between UAVs and obstacles, and the gravitational potential functions between UAVs and their corresponding endpoints, local path planning and obstacle avoidance are performed on the global controller to obtain the movement direction and speed of the UAV swarm at the next moment. Based on the movement direction and speed of the UAV swarm at the next moment, formation control of the UAV swarm is completed. Specifically, the repulsive potential function and the gravitational potential function are as follows: The repulsive potential function between the drones is specifically expressed as follows: ； In the formula, U rep1 (p) is the repulsive potential function between the UAVs at node p, q i and q j K represents the positions of the i-th and j-th drones, respectively. r r is the repulsion coefficient. o This refers to the repulsive force range parameter; The repulsive potential function between the drone and the obstacle is specifically expressed as follows: ； In the formula, U rep2 (p) represents the repulsive potential function between the UAV and the obstacle at node p, and q i For the location of the drone, q k Let K be the position of the k-th obstacle. ro r is the repulsion coefficient. o This refers to the repulsive force range parameter; The gravitational potential function between the drone and the corresponding endpoint is specifically expressed as follows: ； In the formula, U att (q) is the gravitational potential function between the UAV at node q and the corresponding endpoint, where q i For the location of the drone, q target K represents the location of the endpoint. a This is the gravitational coefficient.

2. The UAV swarm control design and simulation method based on consensus algorithm according to claim 1, characterized in that, The process involves global path planning based on the start and end points of the drone swarm in the current flight environment, with the goal of avoiding obstacle locations. This yields several global waypoints for each drone, specifically: Starting from the current location of the drone, the global planning range is determined by the starting point and the ending point. Based on the current location of the drone, a node search is performed within the global planning range with the goal of avoiding obstacle locations, to find the forced neighbors of each node within the global planning range; wherein, the forced neighbors are the nodes that force the algorithm to change its movement direction when performing global path planning; Based on the forced neighbors, the drone's endpoint and origin, find the nodes belonging to the jump point within the global planning scope; Based on the jump point and obstacle location, global path planning is performed starting from the current position of the drone to obtain several global path points for each drone.

3. The UAV swarm control design and simulation method based on consensus algorithm according to claim 2, characterized in that, Based on the drone's current location, a node search is performed within the global planning range with the goal of avoiding obstacle locations to find the forced neighbors of each node within the global planning range. Specifically: Starting from the current position of the drone, with the goal of reaching the destination, the search status of each node within the global planning range is determined by a preset node judgment rule; wherein, the search status includes unobstructed horizontal movement status, unobstructed diagonal movement status, obstructed horizontal movement status, and obstructed diagonal movement status. Path selection is performed based on the search status and obstacle locations to determine the forced neighbors of each node.

4. The UAV swarm control design and simulation method based on consensus algorithm according to claim 3, characterized in that, The process of selecting paths based on the search status and obstacle locations, and determining the forcing neighbors of each node, specifically involves: Based on the search state, and combined with the location of obstacles, a path is selected from the node to the corresponding adjacent node to determine the path trajectory between each node and its corresponding adjacent node. Based on the search value of nodes in the path trajectory, the adjacent nodes are classified as natural neighbors or forced neighbors; wherein, the search value is used to measure whether a node is meaningful for search when making path selection.

5. The UAV swarm control design and simulation method based on consensus algorithm according to claim 2, characterized in that, The step of finding nodes belonging to the jump point from the global planning scope based on the forced neighbors, the drone's endpoint, and the origin, specifically involves: If a node within the global planning scope is the starting point or the ending point, then that node is determined as a jump point; If a node within the global planning scope has the forced neighbor, then that node is determined to be a jump point; If a node within the global planning scope is in an unobstructed diagonal movement state or an obstructed diagonal movement state, and the node can reach a certain jump point by moving horizontally or vertically, then the node is determined to be a jump point.

6. A design and simulation system for unmanned aerial vehicle (UAV) swarm control based on a consensus algorithm, characterized in that, include: Global path planning module, global controller generation module, and local obstacle avoidance module; The global path planning module is used to perform global path planning based on the start and end points of the UAV cluster in the current flight environment, with the goal of avoiding obstacle locations, and obtains several global path points for each UAV. The global controller generation module is used to construct a global controller for the UAV cluster by using a consistent formation control algorithm to coordinate the control of the UAV cluster based on the global path points and a preset communication topology map of the UAV cluster. Specifically, it constructs a motion model for each UAV based on the global path points, the current position of the UAV, and the preset next position of the UAV; it constructs a communication topology map of the UAV cluster based on the communication relationships between the UAVs; and it constructs a global controller for the UAV cluster by considering the formation switching of the UAV cluster and the communication topology map of the UAV cluster. Specifically, the global controller is: ； In the formula, x(t) is the current position of the UAV at time t. The flight speed of the controlled drone is obtained by differentiating x(t), where D is the preset degree matrix, A is the adjacency matrix corresponding to the communication topology graph of the drone cluster, and L is the Laplace matrix. The local obstacle avoidance module is used to construct the potential functions of the UAVs to perform local path planning and obstacle avoidance for the global controller, thereby achieving formation control of the UAV swarm. Specifically, based on the start and end points of the UAV swarm in the current flight environment and the positions of obstacles, a potential function for each UAV is constructed using an artificial potential field method. The potential function includes an gravitational potential function and a repulsive potential function. Based on the repulsive potential functions between UAVs, the repulsive potential functions between UAVs and obstacles, and the gravitational potential functions between UAVs and their corresponding endpoints, local path planning and obstacle avoidance are performed on the global controller to obtain the movement direction and speed of the UAV swarm at the next moment. Based on the movement direction and speed of the UAV swarm at the next moment, formation control of the UAV swarm is completed. Specifically, the repulsive potential function and the gravitational potential function are as follows: The repulsive potential function between the drones is specifically expressed as follows: ； In the formula, U rep1 (p) is the repulsive potential function between the UAVs at node p, q i and q j K represents the positions of the i-th and j-th drones, respectively. r r is the repulsion coefficient. o This refers to the repulsive force range parameter; The repulsive potential function between the drone and the obstacle is specifically expressed as follows: ； In the formula, U rep2 (p) represents the repulsive potential function between the UAV and the obstacle at node p, and q i For the location of the drone, q k Let K be the position of the k-th obstacle. ro r is the repulsion coefficient. o This refers to the repulsive force range parameter; The gravitational potential function between the drone and the corresponding endpoint is specifically expressed as follows: ； In the formula, U att (q) is the gravitational potential function between the UAV at node q and the corresponding endpoint, where q i For the location of the drone, q target K represents the location of the endpoint. a This is the gravitational coefficient.

Images